The market reform in Dutch health care: results, lessons and prospects

TH E M A R K ET R EFO R M IN D U TCH H EA LTH CA R E The market reform in Dutch health care Results, lessons and prospects 55 In 2006, the Netherlands embarked upon an ambitious reform of the Dutch health care system based upon the principles of regulated competition. Some 15 years later, it is an appropriate time to find out how this ‘market reform’ has worked out, and what the experience has been like for those involved in putting it into practice. The authors of this important new study review the reforms and their impact to date and ask whether the reforms merit being counted as a success. Did they alter the relationship between state, insurers, providers and patients? Has there been evidence of problems that market-based systems are often associated with, such as high administrative costs, restricted access to health care, rent-seeking, skimming and adverse selection? Whilst addressing these questions and suggesting possible answers, the authors also examine what can be learned from the Dutch experience with competition in health care and what changes might be expected in the near future in the Netherlands and more broadly, especially considering the context of the COVID-19 pandemic. The editors Patrick Jeurissen is professor of Sustainable Health Care at the Radboud University, Nijmegen Hans Maarse was professor of Health Policy Analysis at Maastricht University, Maastricht Health Policy Series Series No. 55 www.eurohealthobservatory.who.int Patrick Jeurissen Hans Maarse O bs er va to ry St ud ie s Se ri es 55 E dited by P atrick Jeurissen and H ans M aarse R E S U LTS , LE S S O N S A N D P R O S P E CTS OBS_55_A4_Dutch health care.qxp_WHO_A4 27/10/2021 12:55 Page 1 The market reform in Dutch health care The European Observatory on Health Systems and Policies supports and promotes evidence-based health policy-making through comprehensive and rigorous analysis of health systems in Europe. It brings together a wide range of policy-makers, academics and practitioners to analyse trends in health reform, drawing on experience from across Europe to illuminate policy issues. The Observatory is a partnership, hosted by WHO/Europe, which includes other international organizations (the European Commission); national and regional governments (Austria, Belgium, Finland, Ireland, Norway, Slovenia, Spain, Sweden, Switzerland, the United Kingdom and the Veneto Region of Italy with Agenas); other health system organizations (the French National Union of Health Insurance Funds (UNCAM), the Health Foundation); and academia (the London School of Economics and Political Science (LSE) and the London School of Hygiene & Tropical Medicine (LSHTM)). The Observatory has a secretariat in Brussels and it has hubs in London at LSE and LSHTM) and at the Berlin University of Technology. The market reform in Dutch health care Results, lessons and prospects Patrick Jeurissen Hans Maarse Keywords: FINANCING, HEALTH DELIVERY OF HEALTH CARE HEALTH CARE SYSTEMS HEALTH POLICY THE NETHERLANDS © World Health Organization 2021 (acting as the host organization for, and secretariat of, the European Observatory on Health Systems and Policies) All rights reserved. The European Observatory on Health Systems and Policies welcomes requests for permission to reproduce or translate its publications, in part or in full. The designations employed and the presentation of the material in this publication do not imply the expression of any opinion whatsoever on the part of the European Observatory on Health Systems and Policies concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. The mention of specific companies or of certain manufacturers’ products does not imply that they are endorsed or recommended by the European Observatory on Health Systems and Policies in preference to others of a similar nature that are not mentioned. Errors and omissions excepted, the names of proprietary products are distinguished by initial capital letters. All reasonable precautions have been taken by the European Observatory on Health Systems and Policies to verify the information contained in this publication. However, the published material is being distributed without warranty of any kind, either express or implied. The responsibility for the interpretation and use of the material lies with the reader. In no event shall the European Observatory on Health Systems and Policies be liable for damages arising from its use. The views expressed by authors, editors, or expert groups do not necessarily represent the decisions or the stated policy of the European Observatory on Health Systems and Policies or any of its partners. ISBN 978 92 890 5182 8 Printed in the United Kingdom. Address requests about publications to: Publications, WHO Regional Office for Europe, UN City, Marmorvej 51, DK-2100 Copenhagen Ø, Denmark Alternatively, complete an online request form for documentation, health information, or for permission to quote or translate, on the Regional Office web site (http://www.euro.who.int/ pubrequest). Table of contents Preface viii List of tables, figures and boxes x Chapter 1: Introduction to the reform of Dutch health care 1 1.1 Scope and purpose of the study 1 1.2 Is “market reform” a correct term? 3 1.3 Historical and political context 4 1.4 What is health care reform? 7 1.5 Structure of the study 10 Chapter 2. The market reform in Dutch health care: plan and strategy 13 2.1 Introduction 13 2.2 First round: a failed attempt to reform 14 2.3 Second round: the politics of the Health Insurance Act 20 2.4 A new policy paradigm: regulated competition 23 2.5 The shape of regulated competition in Dutch health care 25 2.6 Third round: the politics of market-making 31 2.7 New corporatism: collective framework agreements 37 2.8 Summary and discussion 38 Chapter 3. Health insurance reform in practice 41 3.1 Introduction 41 3.2 Insurers’ implementation strategy 42 3.3 Consumer response 51 3.4 Risk selection and risk solidarity 56 3.5 Income solidarity 59 3.6 Declining public support for solidarity? 60 3.7 Discussion and policy lessons 61 Chapter 4. Purchasing health care in practice 65 4.1 Introduction 65 4.2 Regulatory context 67 The market reform in Dutch health carevi 4.3 Models of purchasing 69 4.4 Evolution of purchasing 70 4.5 Contracts 72 4.6 Purchasing specialist medical care 73 4.7 Purchasing mental health care 78 4.8 Purchasing prescription medicines 80 4.9 Purchasing general practitioner care 82 4.10 Contracting other primary care providers 84 4.11 Purchasing and power 85 4.12 Discussion and policy lessons 86 Chapter 5. Health care provision in regulated competition 89 5.1 Introduction 89 5.2 Concentration and deconcentration 90 5.3 Hospital bankruptcies 95 5.4 The competition–price relationship 97 5.5 The competition–quality relationship 101 5.6 Access to medical care 104 5.7 Financial management and capital investments 107 5.8 Discussion and policy lessons 109 Chapter 6. Impact of the market reform on health care expenditures 113 6.1 Introduction 113 6.2 The international picture 115 6.3 Rise of total health care expenditure 116 6.4 Expenditure growth by sector 117 6.5 Expenditure on outpatient prescription medicines 121 6.6 Effectiveness of global spending constraints and increase of insurer risk 123 6.7 Production incentive? 125 6.8 Effect of co-payments on health care spending 127 6.9 Administrative costs 129 6.10 Discussion and policy lessons 00 Chapter 7. Critical assessment of the market reform and its future 133 7.1 Introduction 133 7.2 Overview of the market reform 133 7.3 The Dutch version of regulated competition 137 7.4 Market or quasi-market? 138 7.5 The critical role of insurers 139 Table of contents vii 7.6 Reasserting the steering role of the state? 144 7.7 The COVID-19 pandemic and regulated competition 146 7.8 Conclusions 150 References 155 Preface Good health is one of the most important preconditions for a good life. That is why countries spend a lot on health care and build complex institutions to create fair, effective and efficient ways to manage these resources. We call these health systems, but none of them are the same due to different historical pathways, different political preferences, different epidemiological challenges and many other differences. Nevertheless, we can broadly categorize health systems in two families: Bismarck’s social insurance and Beveridge’s tax-funded health systems. The idea that the private sector could have a major role in health care seemed highly questionable after the seminal analysis of Kenneth Arrow (1963) on information asymmetries and opportunistic rent-seeking strategies in health care. His analysis did not end discussions of alternatives, however, most notably in countries with a substantial private sector such as the United States of America. In the mid-1970s Alain Enthoven brought a proposal to the table that sought to create universal access and competition in the US private health care system. Called regulated or managed competition, at the time it was the most sophisticated proposal thus far for a “market” approach to health policy (Starr, 1982). The Netherlands seems to be the country that has gone farthest in implementing Enthoven’s model and his ideas on managed competition. The Netherlands’ system is respected among policy-makers. Routinely, it scores very well in the annual surveys of the US Commonwealth Fund think tank. It was very favourably reviewed by Emanuel (2020) in his recent scholarly search to find the best health care systems. The Dutch performance on universal access is much better than that of the United States, and it easily equals those of the better performing tax-funded European systems. In comparison to most other Organisation for Economic Co-operation and Development (OECD) countries, the growth of health care expenditures has slowed substantially in the last decade. And competition is more intense than in the high-priced Swiss health care system. Have the Dutch indeed found the Holy Grail of health systems governance? Or are there caveats or paradoxes to consider? Could one explain these questions for an international audience of policy-makers and researchers who seek to understand and learn from the Dutch? It was Josep Figueras, Director of the European Observatory on Health Systems and Policies, who first proposed this Preface ix book in mid-2018. We are grateful to Josep for asking and for his continuing support. Without him this book would not have been written. We would also like to thank the many reviewers who so generously took the time to read an earlier manuscript, offer many improvements and keep us from writing up too many details, which we were too blind to see: professors Martin McKee (London School of Hygiene and Tropical Medicine (LSHTM)), Kathy Swartz (Harvard School of Public Health), Tom Rice (University of California, Los Angeles (UCLA)), Scott Greer (University of Michigan), Reinhard Busse (Technische Universität Berlin), Elias Mossialos (London School of Economics and Political Science (LSE)), Peter Smith (Imperial College London), Luigi Siciliani (University of York) and Charles Normand (Trinity College Dublin). And thank you, team of the European Observatory: Ewout van Ginneken and Erin Webb. We also thank Gert- Anne van Pruissen (Dutch Ministry of Health) and Niek Stadhouders (National Institute for Public Health and the Environment (RIVM)/Radboudumc) for helping us to provide some of the tables. We were also lucky to be able to hold in-depth conversations on an earlier version of the book with some of the key players, and we are grateful to them: to Ab Klink, Edith Schippers and Martin van Rijn, who belong to different political parties but all were ministers of health and directly shaped many of the key institutions involved; to Ernst van Koesveld, Director-General of Long-Term Care at the Ministry of Health; to Bart Berden, CEO of St. Elisabeth Hospital in Tilburg, who has first-hand experience on what managed competition did to the hospital sector; and finally to Wynand van de Ven and Carolyn Tuohy, who were kind to give their comments on a few chapters. We take full responsibility for all remaining mistakes. Patrick Jeurissen & Hans Maarse List of figures, tables and boxes Figures Fig. 1.1 Scope and pace of policy change: four strategic types 9 Fig. 1.2 Analytical model for the investigation of health care reform 11 Fig. 2.1 Structure of regulated competition in Dutch health care 26 Fig. 2.2 Financial risk of health insurers, 2012–2018 35 Fig. 3.1 Decline in the number of health insurers, 1997–2019 43 Fig. 3.2 Average nominal premium and mandatory deductible, 2008–2019 (euros) 47 Fig. 3.3 Average annual financial result per insured (≥18 years) in basic health insurance, 2010–2018 (euros) 49 Fig. 3.4 Solvency rate of health insurers 50 Fig. 3.5 Percentage who switched insurance provider, 2008–2019 52 Fig. 3.6 Net result per insured for selected insurers in 2019 (euros) 58 Fig. 3.7 Number of recipients of a care allowance (in millions) 59 Fig. 3.8 Willingness to pay for the health care of other people and expected solidarity of other people (percentages) 61 Fig. 5.1 Number of general hospitals and university hospitals, 2000–2018 91 Fig. 5.2 Rise in the number of independent treatment centres, 2000–2016 93 Fig. 5.3 Relationship between the mandatory deductible and non-compliance with referrals 105 Fig. 5.4 Percentage of hospital departments above waiting time norms, 2010−2018 107 Fig. 5.5 Investments in major hospital construction works, 2010–2017 (billions of euros) 108 Fig. 5.6 Financial reserves of hospitals as a percentage of total turnover, 2007–2018 109 Fig. 6.1 Health care spending as a percentage of gross domestic product in selected countries, 2018 or nearest year 115 Fig. 6.2 Annual growth of per capita health care spending (real terms) in selected countries, 2008–2018 116 Fig. 6.3 Growth of total per capita health care expenditures, absolute (blue bars) and percentages (red line), 2000–2018 117 Fig. 6.4 Growth of health care expenditure and gross domestic product, 2009–2018 117 Fig. 6.5 Per capita expenditures on outpatient prescription medicines, absolute and percentage change, 2004–2017 122 Fig. 6.6 Price effects and volume effects of outpatient prescription medicines 122 List of figures, tables and boxes xi Fig. 6.7 Overspending and underspending as percentage of the global budget 123 Fig. 6.8 Overspending and underspending by type of health care 124 Fig. 6.9 Amount of mandatory deductible and percentage of health care financing 128 Fig. 6.10 Percentage of time spent by general practitioners on administrative issues/claiming payments 130 Fig. 7.1 Percentage of people who trusted in providers and insurers 143 Fig. 7.2 Number of weekly hospital admissions of COVID-19 patients 147 Tables Table 2.1 Policy goals and instruments of the market reform 27 Table 2.2 Size of A-segment and B-segment in hospital funding 32 Table 2.3 Overview of collective framework agreements on net expenditure growth (percentages) 38 Table 3.1 Consumer response to choice options 53 Table 3.2 Average under- and overcompensation per person in year t for selected groups based on information from year t–1, using the Dutch risk equalization formula of 2014 57 Table 5.1 Main categories of motives for mergers (multiple answers) 92 Table 6.1 Average nominal percentage growth of total health care spending and spending in three selected sectors 118 Boxes Box 2.1 Structure of health insurance before 2006 16 Box 3.1 Pre-authorization 46 Box 3.2 Comparison sites for health insurance 53 Box 4.1 Concise list of topics in purchasing documents 71 Box 4.2 Experiments with new purchasing models 76 Box 5.1 Bankruptcies of two hospitals in 2018 96 Box 5.2 Rise of quality management in Dutch health care 102 Box 7.1 The COVID-19 pandemic in the Netherlands 147

Chapter 1 Introduction to the reform of Dutch health care 1.1 Scope and purpose of the study In January 2006 the Dutch embarked upon a reform of their health care system based upon the principles of regulated competition (Enthoven, 1988). The flagship of the reform was the Health Insurance Act (Zorgverzekeringswet), which integrated statutory health insurance and all other (mainly private) health insurance schemes into a single mandated health insurance scheme with free consumer choice that covered the entire population. In the new system, consumers were given free choice of an insurer in order to trigger competition. Private insurers would act as prudent purchasers of health services on behalf of their clients, offering them an attractive health plan in terms of quality and costs. For their part, providers would compete for contracts with insurers. The main policy goals (in policy documents often referred to as public values) of this “market reform” were to achieve a health care system offering high-quality care to patients that would be accessible to every person (universal access), based upon solidarity and affordability (financial sustainability). Another goal of the reform was to enhance freedom of choice. The primary function of the state was to regulate health care and preserve the public values in health care. The introduction of the new health insurance legislation was the preliminary end-point of a political discussion that had started almost 20 years earlier with the publication of a report written by a group of independent experts at the request of the government. In its report “Willingness to change” (Bereidheid tot verandering), published in 1987, the Dekker Committee identified many structural problems in Dutch health care. Its main conclusion was that Dutch health care had evolved as an inefficient and inflexible system, lacking powerful incentives to replace expensive care with less expensive but equally effective care. To resolve these problems, the Committee recommended a market-based reform that would offer insurers and providers more room for entrepreneurship and consumers more freedom of choice. These recommendations sparked a The market reform in Dutch health care2 heated dispute on the merits of competition in health care, which is illustrated by the simple fact that it would take until 2006 before the Committee’s main recommendation of introducing a mandated basic health insurance scheme would come into effect. The political debate on competition in health care has never ended, and currently (2021) one can hear various voices calling for a reversal of the market reform and a reassertion of the role of the state in health care. The choice for market reform can be interpreted as a reaction to a period of ever-extending state intervention in health care. Growing concerns about escalating health care expenditures, particularly after the oil crises in the 1970s, had resulted in an avalanche of regulatory and budgetary instruments, including hospital planning, expenditure caps, price controls, user charges and various other policy instruments. After years of mostly disappointing experience with these instruments, the idea emerged that an alternative strategy was needed to make the direction of health care more effective and to establish a proper balance between equity and efficiency. Health care had to be transformed from a supply- driven system into a demand-led system, in which the state concentrated on the introduction of an effective regulatory system and an effective supervisory system. This concept of an alternative model for the organization of health care fit well with the ideas of the so-called New Public Management (Clarke & Newman, 1997; Pollitt, 1993), which had gained much popularity in the Netherlands. The advocates of this new wave in public policy-making postulated that the state had to transform itself from a “bureaucratic state” into a “managerial state”, in which it would carry system responsibility and delegate a great deal of its steering to regulatory agencies at arm’s length. Competition in health insurance and health care provision was depicted as a more effective instrument for achieving the state’s policy goals in health care than detailed bureaucratic intervention. Competition was certainly not presented as a goal in itself, but rather as an alternative institutional vehicle for achieving the state’s policy goals. In choosing a health care system based on regulated competition, the Netherlands changed its system in a more fundamental way than other western European countries. Belgium, Germany and Switzerland, each in their own way, restructured their health care systems by moving away from detailed hierarchical control towards systems with more freedom of choice and room for entrepreneurial behaviour for insurers and providers. This restructuring was done with the intention of improving performance in terms of quality of care, accessibility and financial sustainability (Thomson et al., 2013; van de Ven et al., 2013). The Netherlands now has 15 years of experience with regulated competition among private health insurers. This is enough time to find out how the system has worked in practice and what can be learned from the Dutch experience. In order to do so, we start, first, with an examination of the policy goals, policy Introduction to the reform of Dutch health care 3 instruments and the “assumptive world” (Vickers, 1965) of market reform. We will see that the path from abstract ideas to concrete changes was paved with many obstacles. The reform required the accommodation of a set of abstract ideas to the real world of diverging ideological views, conflicting interests, institutional constraints and complex power relations. Second, this study offers an analysis of the implementation of the market reform. How have insurers and providers put it into practice, and how did consumers respond to the reform? Our third purpose is to examine the results of the reforms. To what extent have the stated policy goals been achieved? Did the reforms establish new relationships among citizens, insurers and providers as envisaged in the government’s policy documents? What is the evidence for the problems that market-based systems have often been accused of, such as high administrative costs, restricted access to health care, rent-seeking, cream skimming and adverse selection? Fourth, we will briefly consider the future of the market reform. Is there any reason to expect a move away from regulated competition in the near future? What may be the impact of the coronavirus disease (COVID-19) pandemic in this respect? Finally, our purpose is to draw some policy lessons from the Dutch experience with health care reform that, we hope, will be valuable to health care policy- makers in other countries, in particular countries that are experimenting with regulated competition. Our analysis is based upon government documents, policy reports, research papers, monitors, market scans and articles published in peer-reviewed journals. Furthermore, we conducted semi-structured interviews about our findings and analyses with former ministers, policy-makers and administrators who are or have been closely involved in the reform of Dutch health care. Their comments helped us to get a better understanding of the essence of the reform. We also benefited from critical comments of foreign experts on an earlier draft of the study. The structure of this chapter is as follows: After a critical comment on the term “market reform” (section 1.2), we consider the historical and political context of the reform (section 1.3). Next follows a brief outline of an analytical model for the study of health care reform (section 1.4). The chapter ends with an overview of the structure of this book (section 1.5). 1.2 Is “market reform” a correct term? Throughout this study, we will use the terminology “market reform” and “competition” to describe the reform of Dutch health care. We use these terms for the practical reason that both terms are commonly used in the political debate and international literature. The terminology suggests that health care in the Netherlands has indeed adopted the characteristics of a market system. This study will lead us to conclude that this is only partially true. It is true for health The market reform in Dutch health care4 insurance, although we will see that this part of health care is heavily regulated to preserve the public values or interests of universal access, high quality and financial sustainability. It is a much more strictly regulated market than, say, the private market for car or civil liability insurance. By contrast, purchasing and provision take place in a system that, with some exceptions, can be depicted as a quasi-market. For instance, it will appear that the scope of price competition should not be overstated and that most providers of health care (in some sectors all providers) are contracted. In short, the term “market reform” suggests more competition than really exists in practice and is, for this reason, actually misleading. We hope to make clear in this study that the current Dutch health care system must be understood as very much a hybrid system and that practice differs in many respects from the policy paradigm of regulated competition. Nevertheless, the term “market reform” is still frequently used in the political debate on health care reform in the Netherlands. In the current political debate, competition is increasingly depicted as a source of persistent problems. Often- heard diagnoses are that competition conflicts with people’s badly needed cooperation and trust in health professionals and that there is a strong need for more state direction in health care. In politics competition appears to be an easy target for the blame game. Almost everything considered wrong is somehow linked to competition. Competition has become a container concept without any precise meaning. Obviously, this is not helpful. Therefore, it is necessary to investigate what is meant by competition in health care, how it has been shaped and how it has played out in terms of access, quality, efficiency costs and other issues. This is what this study aims to do. 1.3 Historical and political context Reforms can never be well understood without taking their historical and political context into account. At the beginning of the 19th century, state intervention in health care was largely non-existent. Since then, as in all countries on the European continent, the state gradually but deeply penetrated into Dutch health care. The state mainly focused its interventions on addressing public health problems, for instance by vaccinating children of poor families against smallpox, putting legal restrictions on child labour and improving public hygiene. Repeated outbreaks of cholera were another concern (de Swaan, 1988; Houwaart, 1991). The health policy agenda broadened in the 20th century, when the state became ever more involved in the regulation of health care. Policy attention increasingly shifted to the regulation of health care financing and provision of health services. Particularly after the Second World War (1939–1945), the organization of health care evolved as a critical element in the building of the welfare state. State responsibility for public health was explicitly formulated in Introduction to the reform of Dutch health care 5 Article 22 of the revised Constitution (1983), which states that the state shall take measures to promote public health. As an outcome of a long historical process, state responsibility for health care has become deeply institutionalized. The challenge of the reform of health care that we analyse in this book, thus, was not to question this responsibility, but rather to redefine and reorganize it in a changed health care landscape. The reform can be conceptualized as a complex mix of institutional change and continuity. A second contextual factor is associated with the specific structure of the public– private mix in Dutch health care. Simply stated, one may depict Dutch health care as a combination of public financing and private provision. While health care is largely financed by public resources (mainly social contributions), the provision of health services is almost completely left to private providers, most of which operate on a not-for-profit basis. For-profit hospitals and for-profit nursing homes are – even today – forbidden. With the exception of municipal public health agencies, public providers do not exist. The reforms have not altered this basic structure.1 The mix of public financing and private provision can be understood as a historical political compromise. In the 18th and 19th centuries, private not-for- profit organizations, many affiliated with a religion, dominated the provision of health services. Most municipalities fulfilled only a residual role in caring for their inhabitants. State interference in health care provision was met with great suspicion. The private not-for-profit sector (particulier initiatief) perceived state interference as an encroachment of its independent position, which it justified not only on historical but also on democratic grounds. In the private sector’s conception of democracy, private organizations had the constitutional right to organize health care and other (social) services for their own clientele. State intervention was permitted only if the private not-for-profit sector failed to fulfil its social responsibility. Gradually, however, this ideological belief eroded, not only as a result of the “depillarization” of Dutch society that started in the late 1950s and early 1960s (Lijphart, 1968), but also for practical reasons. Because of mounting financial problems, private not-for-profit organizations had gradually become dependent on financial support from the state for the provision of health services. Also, public support for a leading role of the state in health care (and other sectors of public life as well) increased. As an ultimate political compromise, the state was accorded overall political responsibility for the financial accessibility and quality of health care, while the private sector kept its leading role in the quotidian provision of these services. We will see in this book that the 2006 reform of health care and the 2015 reform of long-term care fully respected this historical compromise. 1 See Kroneman et al. (2016) for an extensive description of the Dutch health care system. The market reform in Dutch health care6 Health insurance originated in the private sector (Companje et al., 2011). The first sick funds date from the 19th century. It took until the beginning of the 20th century for the state to propose legislation to regulate health insurance. However, all efforts to introduce a statutory scheme failed in the period 1900–1940, not only because of the resistance of the funds to state intervention but also because of fierce opposition from the medical profession, which feared state control of its activities. Eventually, in 1941, the German occupier implemented a scheme of state-controlled health insurance for employees. The statutory scheme did not eliminate the funds, but it downgraded them to implementing agents under strict hierarchical control. The scheme was largely codified in 1964 in the Sick Fund Act (Ziekenfondswet), which covered about two thirds of the population. Persons not covered by the scheme could purchase substitute private health insurance, which also had a long history in Dutch health insurance (Schut, 1995; Maarse & Jeurissen, 2020). The traditional public–private mix in Dutch health care has certainly facilitated the reforms. For instance, there was no need to privatize hospitals (in 1997 the last public general hospital had been converted into a private not-for-profit entity) or implement a purchaser–provider split, as was necessary in the United Kingdom, the Nordic countries and the countries in central and eastern Europe. Furthermore, it was relatively easy to convert sick funds and private insurers into implementing agencies of the new health insurance scheme. In sum, the reform built on the traditional balance between the public and private sectors. The structure of Dutch health care on the eve of the reforms can be conceptualized as the result of a long historical development in which rivalling ideologies, conflicting interests, power relations and accidental events always heavily influenced the direction of health care policy-making. Disputes about issues such as regulations, policy goals and instruments or the balance between the state and the private sector were business as usual. Negotiating political compromises was the only way to agree on new legislation. State intervention in health care is, therefore, not the outcome of hierarchical intervention, but rather the outcome of a process of consultation and negotiation with the leading organizations representing the interests of health insurers and service providers. The impact of this political culture, which constitutes an important element of the so-called consensus democracy (Lijphart, 1999) and the neo-corporatist style of public policy-making in the Netherlands (Visser & Hemerijck, 1997; Andeweg et al., 2020), is clearly recognizable in the reforms of medical and long-term care. The government’s strategy has always been to gain the support of the lead organizations for its reforms by engaging them in the policy-making process in return for political influence. Consequently, the reforms were less radical than the government originally had in mind and took much more time Introduction to the reform of Dutch health care 7 than envisaged. The need for political compromise was also closely associated with the tradition of coalition government in Dutch politics. The reforms were feasible only after they sufficiently accommodated the wishes of the political parties constituting the government coalition of the day. Borrowing a term from Moran, one may conceptualize Dutch health care as a “health care state”, that is, a state in which public and private agents have deeply penetrated each other (Moran, 1999: 4). Parallel to the growing complexity of health care, the increased number of policy actors in health care and the close connection of health care with other sectors of public life, mutual dependency has significantly increased. It is no exaggeration to postulate that neither the state nor the private sector can effectively operate without cooperation. We will see in this study that the increased mutual dependency influenced the shape and results of the reforms. The deeply rooted principle of universal and equal access in Dutch health care also greatly influenced the reforms. The leading normative principle holds that each sick person or person with a handicap has access to state-of-the-art health services and that age, gender, income, social position, race or any other discriminating factor should not influence what type of care a person receives. One may speak of an egalitarian culture in health care. Violation of the principle of universal and equal access is considered inequitable. The impact of this principle is clearly recognizable in the reform of medical care and long-term care. To be politically and socially feasible, both reforms had to respect the historical legacy of universal and equal access. Universal access in the Dutch context also includes the absence of financial barriers to health care. There is widespread public support for the view that the financial burden of health care must be shared by means of a solidary financing model. People should pay according to their ability to pay for a broad package of health services, and the amount they pay should not be linked to pre-existing medical disorders (van der Aa et al., 2018). The principle of solidarity also helps to explain why many people in the Netherlands consider private payments for health care unfair and why in the past several arrangements for user charges in statutory health insurance were short-lived. The impact of the normative principle of solidarity upon the reform of medical care and long-term care can hardly be underestimated. 1.4 What is health care reform? There is no consistent and universally accepted definition of what constitutes health care reform. Policy-makers and policy analysts assign different meanings and connotations to the concept. It also happens that, for political reasons, The market reform in Dutch health care8 policy-makers sell incremental policy changes as reform (Saltman & Figueras, 1997: 2). In this study we define health care reform as a deliberate attempt to implement a major change in a country’s health care system. It is an orchestrated effort to bring about “system change”, reflecting a belief that the existing system is failing or will be unable to respond adequately to future changes in disease patterns or technological changes. The study of health care reform (Okma & Tenbensel, 2020) requires an investigation into the policy goals of the reform and the policy instruments necessary to achieve these goals. We will see that the policy goals of the market reform only pointed out a direction of change. Health system performance had to be improved, in particular in terms of quality and cost control, while guaranteeing universal access. The absence of quantified policy goals makes it difficult to measure the success of the reform. How much change in the desired direction is needed to qualify as successful reform? What is more, stakeholders and citizens may appreciate the results quite differently. What has been said about the formulation of policy goals for health care reform also applies to the formulation of the policy instruments. The main instrument of the reform of Dutch health care is institutional change through the introduction of regulated competition. However, this is a rather abstract concept that raises the question of how it has been worked out in concrete regulations and how it has been put into practice. 1.4.1 Paradigmatic shift Health care reform involves a major shift in the prevailing policy paradigm (policy framework, assumptive world) that, following Tuohy (2018: 8), consists of three intersecting elements: (a) a balance of influence among state, market and civil society; (b) a mix of instruments regulating the interaction between state, market and civil society; (c) a set of beliefs (assumptive world) about what is (causal), what works (instrumental) and what should be (normative).2 Health care reform rests on the assumption that the enhancement of system performance requires a paradigmatic shift; incremental policy changes or “piecemeal engineering” will fail to solve the system’s problems. The paradigmatic shift or policy reframing (Schön & Rein, 1994) underpinning health care reform is an important part of the study of health care reform. In the Dutch case, the advocates of the market reform postulated that the traditional system of strong state involvement and bureaucratic control was exhausted. They 2 Tuohy does not speak about beliefs but about “organizing principles” that legitimize “the resulting distribution of costs and benefits so long as those principles are observed” (p. 8). Introduction to the reform of Dutch health care 9 depicted Dutch health care as an inefficient, non-innovative and supply-driven system that left citizens little freedom of choice. In their view there was a strong need for moving towards a demand-led health system with more freedom of choice, more room for entrepreneurship and less detailed state control to achieve better system performance (Chapter 2). 1.4.2 Scope and pace In her study of health care reform in Canada, the Netherlands, the United Kingdom and the United States, Tuohy (2018) distinguishes between the scope and the pace of change. The scope of change varies from small to large and the pace of change from gradual to simultaneous. A combination of the two dimensions gives four strategic types of policy change (Fig. 1.1). Fig. 1.1 Scope and pace of policy change: four strategic types BLUEPRINT BIG-BANG INCREMENTAL MOSAIC Scope Pace gradual simultaneous large small Source: Tuohy (2018: 14). The reason that Tuohy speaks about strategic types is that she considers timing “an element of the strategic judgments policy-makers make in response to their reading of political circumstances”. In addition to decision-making on the scale of reform, “they also need to decide how quickly to enact the desired change” (Tuohy, 2018: 11). Her typology is useful for our analysis of the market reform in Dutch health care. We will see that this reform can be classified as a “blueprint” type of reform – large in scope, slow in pace. Its scope was rather comprehensive. It comprised major changes in the financing, purchasing and provision of medical care. As for the pace, the market reform took place gradually. Although it is beyond any doubt that the introduction of a new health insurance scheme in 2006 was an important step in the reform, it was nevertheless only the first step. Various market-making decisions to extend the scope of competition were taken later. The market reform in Dutch health care10 1.4.3 Implementation Each health care reform starts with some global ideas on the need for and direction of system change. These ideas are worked out in policy documents describing what is going wrong in health care, why reform is required to improve health system performance and how this should be achieved. Policy documents also include a formulation of the policy goals and instruments. Decision-making on health care reform is usually not a matter of only a few months. We will see in this study that the incubation of the reform took many years. In large part this was due to persistent political struggle over their shape. A window of opportunity was needed to accomplish a breakthrough (Kingdon, 1984). The study of the decision-making process and the content of reform is an important element of any study of health care reform. However, for a proper understanding, it is equally important to study the implementation process. Policy implementation is not a merely technical or linear process of carrying out the decisions taken by policy-makers. On the contrary, it is a crucial stage in all reforms. The success or failure of health care reform depends to a great extent on what citizens, providers, insurers, municipalities and other players make of it in practice. It is during implementation that capacity problems or other, sometimes unexpected, practical problems arise, that regulations appear ambiguous in their application to concrete cases, that players find loopholes in the legislation or start lawsuits to overturn decisions of implementing agencies and so on. Policy implementation may develop as the continuation of the political struggle over reform in another arena with other players. Successful policy implementation often requires policy adjustments and regulatory revision to resolve practical problems. While it is true that policy shapes implementation, it is equally true that policy implementation shapes policy. It is during policy implementation that reforms show their true face (Majone & Wildavsky, 1978). For this reason the focus of this study is on the implementation of the market reform. Furthermore, implementation research may reveal significant differences between the macro- and micro-worlds of health care reform. Steering the implementation of health care reform (macro level) is but one side of the reform. The other side is how players at the local level (micro level) deal with it. They may experience entirely different problems from the players at the macro level and perceive the effects of the reform differently. 1.5 Structure of the study To structure our study, we make use of a simple analytical model (Fig. 1.2). This model makes a distinction among primary functions that each care system must fulfil: (a) the provision of health care; (b) the financing of health care; and (c) the Introduction to the reform of Dutch health care 11 allocation of scarce resources to health care (purchasing). The market reform is conceptualized as a deliberate attempt to implement major changes in the way that these functions are fulfilled. All relationships between the building blocks in Fig. 1.2 are reciprocal. On one hand, the reform intends to bring about certain changes in health care financing, health care purchasing and health care provision. On the other hand, experiences with each of these changes may lead to adjustments in the reform. In a similar way, the effects of the reform may be reason for adjustments in the implementation of the reform or in the content of the reform itself. Fig. 1.2 Analytical model for the investigation of health care reform Health care financing Health care provision Health care purchasing Content of reform Policy-making process Freedom of choice Universal access Solidarity Quality of care Financial sustainability Goals/effectsImplementationReform The study starts in Chapter 2 with an analysis of the content of the reform plan (policy goals and instruments, policy paradigm), the policy-making process and the political struggle that preceded the adoption of the reform plan. Next follows an analysis of the implementation and the effects of the reform. Chapter 3 presents a study of the implementation strategy of insurers and the response of consumers to the reform of the health insurance market and the policy lessons that can be learned from it. Furthermore, attention will be paid to the impact of the reform on freedom of choice, risk solidarity and income solidarity. Both types of solidarity make up an important dimension of access to health care and health care financing. The practice of health care purchasing is the subject of analysis in Chapter 4. How do insurers shape purchasing? What kind of contracting models exist, and how are they used in contracting? Since purchasing has often been considered the heart of the reform, the question arises whether insurers have indeed been able to fulfil their assigned agency role. What policy lessons can be drawn from the experience with purchasing in Dutch health care? We will also touch on a hot political issue in the reform, namely the power balance between insurers and providers. Chapter 5 is devoted to an analysis of health care provision. To what extent has the reform truly altered the landscape of health The market reform in Dutch health care12 care provision? Do health care providers engage in competition with each other as assumed in the model of regulated competition? What has been the influence of the reform on the quality and accessibility of health care (particularly in terms of waiting time)? What are the policy lessons to be learned from it? An overriding issue in the political debate on the market reform has always been its impact on health care expenditures. The reform is assumed to contribute to effective cost control and financial sustainability by squeezing out technical and allocative inefficiencies. The big question, of course, is whether this ambitious policy goal has indeed been achieved. This question will be addressed in Chapter 6. The main topic in Chapter 7 is our assessment of the reforms, including the future of regulated competition. After an overview of the main findings in this study, we will reflect upon the role of the state and insurers in current discussions on regulated competition. Our conclusion is that the political debate tends to go in the direction of a more leading role for the state in combination with a less prominent role for insurers. The call for reassertion of the role of the state in health care is also motivated by experiences with the handling of the COVID-19 pandemic in 2020. Chapter 2 The market reform in Dutch health care: plan and strategy 2.1 Introduction The introduction of the Health Insurance Act on 1 January 2006 put an end to the traditional divide between the sick fund scheme, which covered about 63% of the population, and a mix of public and private substitutive schemes, which covered the remaining 37% (Maarse & Jeurissen, 2020). The new legislation integrated all schemes into a single mandatory basic scheme covering every legal resident1 in the Netherlands. The scheme is carried out by private health insurers (most of them acting not-for-profit), which compete for clients on the health insurance market. Each person has the legal right to switch to another insurer at the end of each year. Health insurers are obliged to accept every person. A complex system of regulations is in place to spur competition while upholding the public interests of health care. The policy objectives of the new system are to enhance freedom of choice, foster efficiency and quality of health care, reinforce solidarity in health care financing and safeguard financial sustainability (van de Ven & Schut, 2008). The enactment of the Health Insurance Act marked the end of a process originating in the 1980s. In 1987 the Dekker Committee – named after its chairman Wisse Dekker, who had been the chief executive officer (CEO) of Philips – published its report “Willingness to change”, which called for a fundamental restructuring of health care based on the principles of regulated, or managed, competition. In 1994, after two successive governments had failed to convert the Committee’s proposals into new legislation, the reform seemed politically dead. However, at the turn of the century, it was again put on the political agenda. After a process of intensive consultation, deliberation and 1 A person living abroad but paying payroll tax in the Netherlands is also covered (for example, a Netherlander living in Belgium but working in the Netherlands or a Belgian living in Belgium but working in the Netherlands). The market reform in Dutch health care14 negotiation, the government managed to build a political majority for its reform in the lower and upper chambers of parliament. With the reform came a new policy vocabulary. Examples of verbal innovation (van Twist, 1994) are terms such as “regulated competition” or “managed competition”, “consumers”, “providers”, “purchasing”, “supply-driven” health care and “demand-led” health care. Each of these terms reflected the impact of the neoliberal wave in public policy-making on the redesign of Dutch health care. Many health economists advocated regulated competition as a new policy paradigm to overcome structural deficiencies in Dutch health care (Enthoven & van de Ven, 2007). The introduction of the new health insurance scheme was an important yet intermediate step in the reform process. Now, competition had to be put into practice. It could work properly only if insurers and providers had sufficient room for entrepreneurship. As this room was still restricted in 2006, various market- making policy measures had to be taken to stimulate strategic purchasing by insurers and care innovation by providers. Almost all of these measures elicited new political struggle. Indeed, some proposals to complete the reform were turned down by the parliament or withdrawn by the government because of lack of political support. This chapter starts with a brief overview of the history of the market reform through the enactment of the new health insurance legislation in 2006. We will see that this process had anything but a linear structure. It actually consisted of three major policy rounds. The first round started in the late 1980s and ended with the temporary political death of the reform in the mid-1990s (section 2.2). The second round began in the late 1990s and ended with the enactment of the Health Insurance Act and some other market regulation (section 2.3). The next two sections present an analysis of the intellectual policy paradigm of regulated competition (section 2.4) and its concrete shape in Dutch health care (section 2.5). There follows an overview of the third round in the reform, during which policy measures to extend the scope of competition were taken (or turned down) (section 2.6). The last topic of discussion concerns a particular aspect of the Dutch version of regulated competition – namely, the practice of collective framework agreements on cost control and other policy issues signed by the government and the lead organizations of providers and insurers (section 2.7). The chapter ends with a brief summary and discussion of the reform (section 2.8). 2.2 First round: a failed attempt to reform The political decision to reform health care was taken during the formation of the coalition government of the Christian Democratic Appeal and the People’s Party The market reform in Dutch health care: plan and strategy 15 for Freedom and Democracy (VVD)2 in 1986. Concerns about surging health care expenditures and persistent inefficiencies in health care, in combination with increasing doubt about the effectiveness of regulatory state control, were primary motives for reform. However, the call for reform was not an isolated phenomenon; it formed part of a wider trend in public policy-making at the time. The economic crisis in the 1980s and escalating welfare expenditures had convinced many politicians of the need for a reform of the welfare state. Incremental policy adjustments such as wage freezes or lower social security benefits were considered insufficient to tackle the spending crisis (Helderman et al., 2005; Hemerijck et al., 2000). There was a need for institutional change. At the same time, the neoliberal wave in public policy-making started to make its mark. Many considered the so-called New Public Management approach an appealing alternative policy paradigm. What was needed was a transition from a supply-controlled system (“Moscow at the North Sea”) to a demand-led system with more freedom of choice for consumers and more room for entrepreneurship. In the phrase of advocates for the New Public Management, the “bureaucratic state” had to be transformed into a “managerial state” in which the state no longer steered by means of detailed bureaucratic regulations, but rather managed through incentives and disincentives, competition, contracting, outsourcing and performance-based payment models (Pollitt, 1993; Clarke & Newman, 1997). The call for regulated competition in health care did not come out of the blue. In the early 1980s, the umbrella organization of the sick funds had argued for more competition, freedom of choice and entrepreneurship in health care to stimulate efficiency. Gradually, the perceived need for reform gravitated to the concept of regulated competition (Tuohy, 2018; Schut, 1995). A window of opportunity for reform had opened (Helderman et al., 2005; Helderman & Stiller, 2014). The first step in the reform process was the establishment in 1986 of an ad hoc advisory committee consisting of seven independent experts and two advising officials from the Ministry of Health.3 (Known informally as the Dekker Committee, its full name was the Committee on the Structure and Financing of Health Care.) Its remit was to advise the government on the redesign of health care. In his installation speech the then State Secretary of Health hinted at regulated competition as a leading concept for reform. The Dekker Committee identified the lack of efficiency and flexibility as fundamental problems in Dutch health care. There were no powerful incentives for providers to replace expensive care with less expensive but equally effective health care. The fee-for-service 2 The VVD can be described as positioned on the right wing of the political spectrum in the Netherlands. 3 The name of the ministry responsible for health and other public sectors has often changed. Since 1994 the full name of the ministry has been the Ministry of Health, Welfare and Sport (Ministerie van Volksgezondheid, Welzijn en Sport). For convenience, the term “Ministry of Health” (MoH) will be used throughout this study. The market reform in Dutch health care16 payment system gave doctors little incentive to limit health care expenditures. Other manifestations of inefficiency were the oversupply of hospital beds, the bifurcated structure of health insurance (Box 2.1) and the fragmented structure of payment arrangements. Lack of patient-centredness was flagged as yet another fundamental problem. Health care was depicted as a supply-driven system, leaving the insured and patients little choice (CSFG, 1987). Box 2.1 Structure of health insurance before 2006 Prior to the 2006 reform, the Sick Fund Act (Ziekenfondswet) covered about two thirds of the population. The Act, which had come into force in 1964, was in fact little more than the codification of the Sickness Fund Decree introduced by the German occupier in 1941. The Act established compulsory coverage for workers with incomes under a state-predetermined income threshold. Older people and self-employed persons could join a sick fund on a voluntary basis. Employers and employees paid an equal share of income-related contributions. The rest of the population was excluded from coverage and had to rely on (voluntary) private health insurance. The market for private health insurance had a complex structure. About one quarter of the population had a truly private scheme, with risk-related premiums, medical screening upon entrance and various choices (for example, a voluntary deductible). Another 5% of the population was covered by a private scheme that was actually heavily regulated by the state. This scheme had been introduced in 1985 under the Access to Health Insurance Act (Wet op de Toegang tot Ziektekostenverzekeringen). It introduced a sick-fund-like safety net for persons who could not afford private insurance or were denied access because of their medical history. Enrollees had to pay a flat-rate premium. Because premium revenues did not cover all spending, people with private health insurance coverage had to pay an annual surcharge to make up the deficit. Another piece of legislation required private insurers to contribute to the sick fund scheme to compensate for the overrepresentation of older people in these funds. The bifurcated structure was increasingly considered problematic. One of the problems was the borderline between the sick fund scheme and private schemes. Another problem was complexity. Though the Access to Health Insurance Act had solved the access problem, it was generally considered a temporary solution. Sick funds were permitted to offer a private scheme to their subscribers who had to leave the fund because of crossing the income threshold, but they were obligated to establish a separate organizational entity for this. Last but not least, the bifurcated structure of health insurance caused fairness problems. For instance, persons with children could be confronted with a much higher insurance bill if they had to leave their fund. (Children were automatically co-insured in the sick fund scheme.) Individuals with a part-time job paid only a small contribution, even if their partner had high income. The middle-income categories had to bear the greatest share of health care funding. In sum, the structure of health insurance was considered non-solidary. Source: Maarse & Jeurissen, 2020. The market reform in Dutch health care: plan and strategy 17 The main recommendations of the Committee were as follows: • Introduction of an integrated system of health care financing by means of a single basic health insurance scheme covering the entire population and providing about 85% of the benefit package of statutory health insurance.4 Each individual would be obligated to purchase a basic scheme but would be free to choose an insurer. Insurers would have to accept all applicants (open enrolment) but would no longer be obligated to contract with each provider. The new scheme would be financed through a mix of social contributions set by the government and, to spur competition, a nominal (flat-rate) premium set by each insurer separately. • Creation of a flexible supply of care facilities with room for entrepreneur- ship. To increase flexibility, state hospital planning would have to be largely abolished, and hospitals would be responsible for all financial consequences of their investments. • Reform of health care funding by reimbursing health care functions instead of the provider organizations themselves in order to breach the monopoly of hospitals and other providers. • Elimination of the production incentive in the reimbursement scheme of doctors and other providers. The analysis and recommendations of the Dekker Committee were inspired by the concept of managed (or regulated) competition that had come from the United States of America (Enthoven, 1988; 1993). The concept was that competition with strict regulations to preserve public interests – efficiency, quality, solidarity and financial sustainability – could significantly improve the performance of health care. In retrospect, several observations can be made. First, the recommendations of the Dekker Committee provided only an alternative institutional structure to tackle the problem of persistent inefficiencies in Dutch health care and escalating costs. In other words, the focus of the report was on policy instruments, not on policy goals. Second, the recommendations were rather a hybrid. For instance, the Committee advocated the introduction of regulated competition and the termination of hospital planning to foster efficiency and flexibility, but it also recommended the acceleration of the hospital bed reduction programme that had been one of the government’s key regulatory tools to rein in health care expenditures since the late 1970s. Third, the Dekker report offered only a 4 Including the statutory insurance scheme of long-term care and social care. The market reform in Dutch health care18 blueprint for reform. It left many important but politically sensitive technical details unaddressed – for instance, the structure of risk equalization to compensate insurers for bad risks and revision of the reimbursement of hospitals and doctors. Somewhat paradoxically, its global approach did not deter the Committee from projecting that the reform could be implemented in just five years. The Dekker report reflects the redesign approach to reform. It started with an analysis of major failures in the Dutch health care system (problem analysis) and then outlined an alternative model that logically built upon the Committee’s problem analysis (problem resolution). Alternatives to regulated competition were not investigated. Nor did the Committee show much interest in designing a strategy to move from “here” to “there”. In fact, it underestimated the technical aspects of the reform and its political complexity. The report received a mixed welcome. While the government declared its acceptance of the recommendations as leading towards reform, there was also much scepticism from the very beginning. Doctors and provider organizations feared that competition would lead to commercialism and would eventually crowd out the prosocial spirit of the medical profession. Doctors also expressed concerns about the repercussions of the reform for their revenues. Unions and patient organizations were positive about the introduction of a national health insurance scheme but critical of its consequences for solidarity in health care financing. Sick funds cast doubt on the feasibility of a fair risk compensation system and warned against risk selection by insurers. Private insurers feared loss of business. Employers and private health insurers advocated a less prominent role for income- related contributions in health insurance. The Central Healthcare Tariffs Agency (Centraal Orgaan Tarieven Gezondheidszorg) predicted that competition would fail to control costs. In short, the reception for the Committee’s recommendations was only moderately positive (Okma, 1997; Tuohy, 2018). 2.2.1 From Dekker to Simons As noted, the Dekker report presented only an outline for reform. Now the government had to translate the outline into new legislation. It did so in a policy document titled “Change assured” (Verandering verzekerd) (MoH, 1988). The government’s plan matched the recommendations of the Dekker Committee, with one important exception: Whereas the Committee had opted for a big- bang, all-at-once implementation strategy, the government chose a step-by-step introduction strategy (a “blueprint” approach). In 1989 a crisis in the governing coalition not only delayed the legislative process but also resulted in a significant change of the political context: The Labour Party replaced the Liberal Party in the government coalition. To accommodate the The market reform in Dutch health care: plan and strategy 19 objections of the Labour Party to the privatization of health care financing, the new Deputy Minister of Health and member of the Labour Party, Hans Simons, proposed a few, but significant, modifications to the original Dekker Plan. His plan – known as the Simons Plan – was to make the new insurance scheme more equitable by extending the span of statutory service coverage (for example, dental care, physiotherapy and prescription medicines had to be covered) and lowering the fraction of the nominal or flat-rate premium (set by each insurer separately) in financing the new scheme (MoH, 1990; 1992). The revision of the Dekker Plan did not end political controversy over the reform but instead fuelled and deepened it. While the Labour Party had become a positive but critical supporter, the Christian Democrats, employer organizations and private health insurers adopted an ever more critical position. Employers and some higher-income groups became increasingly concerned about higher payroll taxation. Private insurers demanded the formal privatization of sick funds as a precondition for their support of the reform. Provider organizations and doctors protested against the functional, instead of provider-based, definition of health services and the envisaged agency role of insurers. Labour unions worried about the regressive effect of the flat-rate premium (Okma, 1997; Tuohy, 2018). Political opposition increased, and in the autumn of 1993 the government announced a pause in the reform process. After Simons failed to build a political majority for his stepwise introduction of the new health insurance legislation, he drew his conclusion and resigned. How to explain the failure? First, as a result of the deterioration of the economic situation, the relationship between government and the representative organizations of employers and workers had come under increasing strain. Second, the original support for reform had largely crumbled. The new policy paradigm of regulated competition clashed head-on with the institutionalized paradigm of state planning in health care. Gradually, the Simons Plan got stuck in a swamp of ideological differences, doubts, disbelief and vested interests. A parliamentary commission that investigated the failure of the reform concluded that the government had been unable to break through the “clay layer of vested interests”. It also cast doubt on the presumed overarching consensus on the need for and direction of the reform. In fact, the commission concluded that the reform had suffered from a lack of sense of urgency. Furthermore, it referred to the widespread profusion of negative power (the power to obstruct) and the lack of positive power (the power to get something done) (Willemsen Committee, 1994). 2.2.2 Incremental steps towards competition Despite the political failure of the Simons Plan, a number of policy measures were taken in the period 1991–1993, each of which took a small step towards The market reform in Dutch health care20 regulated competition (Helderman et al., 2005). In 1991 sick funds were permitted to set their own nominal premiums. (A uniform government-set nominal premium had been introduced in 1989.) In 1992 sick funds were permitted to operate nationwide to trigger competition. Until then, they had acted as regional monopolists. Two other changes were the introduction of maximum prices for health services, which enabled insurers and providers to negotiate lower rates, and the termination of the insurers’ obligation to contract some categories of providers. The obligation to contract hospitals remained. Another significant development was a start on risk equalization. At first, the structure of risk equalization was very simple (based on age, gender, geographic location), and the financial risk to the funds was capped at just 2.5% of their expenses (Douven, 2004). The direct impact of these policy changes should not be overstated. They did not significantly alter the health insurance landscape or the contracting of providers. Nevertheless, they helped to pave the road to the introduction of regulated competition. Another relevant development was the merger of the lead organizations of sick funds and private insurers in 1995. The merger under the name Health Insurers Netherlands (Zorgverzekeraars Nederland, ZN) signified a convergence of the interests of these organizations. Their intention was to increase their influence in the reform process by speaking with one voice. 2.3 Second round: the politics of the Health Insurance Act The government coalition consisting of the Labour Party, the right-wing liberal VVD and the left-wing liberal Democrats 66 (D66), which took office in 1994, decided to put the reform on hold. In its view the reform effort had only resulted in an unproductive ideological debate about the pros and cons of competition. The first priority of the new government was controlling health care costs. It continued the system of hospital budgeting (introduced in 1982) and capped the net real volume growth of health care at a modest 1.3% a year. In the government’s second term (1998–2002), health care reform reappeared on the political agenda. By the end of the 1990s, providers’ frustrations with the squeeze of budget controls imposed by the previous government and burgeoning patient waiting lists motivated the government to resume the political discussion of a major overhaul of the Dutch health system. An important court ruling in 1999 reinforced the need for reform. The court interpreted long waiting times as a breach of the legally established right of patients to health care. This ruling compelled the government to substantially increase the nation’s budget for health The market reform in Dutch health care: plan and strategy 21 care. Furthermore, two standing advisory agencies, the Social and Economic Council (Sociaal-Economische Raad) and the Council for Public Health and Health Care (Raad voor de Volksgezondheid en Zorg, RVZ), recommended, each in its own way, the introduction of a comprehensive national health insurance scheme with regulated competition. After many consultations and long internal debate, the government published a position document in 2001, titled “A question of demand” (Vraag aan bod), which reflected in several respects the earlier analysis in the Dekker report (MoH, 2001). The document identified two main problems. First, the strong supply orientation of the health care system was perceived as an important cause of inefficiency and lack of innovation. The second problem was the incoherent structure of health care financing. The co-existence of statutory and private health insurance had resulted in an opaque and unfair distribution of the financial burden in health care financing (lack of solidarity), with the middle class paying the greatest part of the national health care bill (RVZ, 2005). The bifurcated insurance structure also frustrated the realization of an efficient and innovative health care system. As an overall direction for reform, the government argued for a universal mandatory health insurance scheme in combination with regulated competition, just as the Dekker Committee had done 13 years earlier. The government’s sketch of the reform was more conceptual than concrete. Each stakeholder was given the opportunity to voice its opinion and make suggestions. The government’s strategy of consultation was intended not only to collect information and policy suggestions, but also to mobilize political support for its reform plan (Helderman & Stiller, 2014; Tuohy, 2018). The reform process stagnated again after an electoral earthquake in 2002. The coalition lost many seats in the parliament (the Labour Party’s representation was halved), and a new populist political party, named Pim Fortuyn List,5 entered the political stage with 26 of the 150 seats in parliament. A short, turbulent political period culminated in the fall of the government after only 11 months. The new, right-wing coalition government consisting of the Christian Democratic Appeal, VVD and D66 declared health care reform one of its priorities. Building on the analytical framework outlined in “A question of demand”, the new government sent its proposal for new health insurance legislation to the parliament in September 2004. The government started an intensive consultation with private health insurers. Their support was important to the success of the reform. Provider organizations were less intensively involved, and their attitudes ranged from indifference to moderate support. Hospitals and general practitioners were most interested in the consequences of the reform for their revenues. The reason for the involvement of patient organizations was largely to legitimate the 5 Pim Fortuyn, the charismatic leader of the party, was assassinated nine days before Election Day. The market reform in Dutch health care22 reform process. The employers’ associations and unions showed little support for the planned reconstruction of Dutch health care (Tuohy, 2018: 361–362). The lower chamber of parliament approved the Health Insurance Act in December 2004, and the upper chamber did so in June 2005. The Act came into force in 2006. The sick fund scheme and all other health insurance schemes were repealed. The enactment of new health insurance legislation marked the beginning of the end of a political debate that had lasted almost 20 years since the Dekker Committee had published its outline for reform. The long policy-making trajectory is probably the best illustration of the controversial nature of the reform. In the following sections, we will see that political controversy did not stop after 2006. 2.3.1 Strategic choices The reform built on several strategic choices. The first was to reform health care according to the principles of regulated competition. In the view of the leading policy-makers, competition in combination with strict regulations would promote efficiency and innovation without compromising the public interests of solidarity in health care financing, universal access, quality of care and financial sustainability. The steering model of hierarchical supply control (for example, hospital planning, hospital budgeting) no longer functioned well. Important elements of the new approach were consumer choice, room for entrepreneurship and, to discipline market behaviour, devolution of financial risks to insurers and providers. The choice of the model of regulated competition was supported not only by the true believers in the benefits of competition in health care, but also by the Christian Democrats in the new government coalition. Building upon their established concept of subsidiarity, the Christian Democrats argued that the financing and provision of health care had to be delegated to organizations with a social purpose. The role of the state had to be restricted to establish an effective regulatory framework to preserve the public values (also referred to as public constraints) of quality, affordability of health care and universal access (van de Gronden, 2004).6 The second strategic decision was the choice for the private model. The new health insurance legislation was construed as a private law arrangement to underscore the changes in the roles of the state and the private sector in health care. Insurers could operate either on a for-profit or a not-for-profit basis. In the new policy narrative, health care was “given back” to the private sector after many years of 6 A salient detail is that the Christian Democrats supported the model of regulated competition only in the lower chamber, while in the upper chamber they rejected it. The market reform in Dutch health care: plan and strategy 23 intensive state intervention. The private model was promoted as the best way to reinforce the agency roles of insurers, providers and consumers. With its choice for an arrangement under private law, the government followed a different route than that suggested by the Dekker Committee. The Committee’s preference for an arrangement under public law had also been advocated in the government’s report “A question of demand”: A public law would preserve the social nature of health insurance and accord the state a leading role. The alternative of a private model reflected the influence of the neoliberal wave in public policy-making and the political idea of subsidiarity and self-regulation. It also was essential for getting the (private) insurers on board. However, it was a much-disputed choice that eventually motivated the left-wing political parties to vote against the Health Insurance Act, even though they embraced the idea of a single scheme.7 The choice of the private route also raised questions about the compatibility of the new health insurance legislation with European competition regulation. As the new insurance scheme no longer fulfilled the conditions of social health insurance, the European Union’s third directive on non-life insurance was operative. Regulations were compatible with that directive only under the conditions of necessity and proportionality to preserve the “social good”. This condition raised questions about the compatibility of the new scheme with the directive. To check whether the new legislation was indeed “Euro proof”, the government contacted the Dutch Commissioner in the European Commission. The advice of the Commissioner was interpreted as positive. However, even today there are still questions about the reform’s compatibility with European legislation. The third strategic choice was to focus the reform on medical care, excluding long-term care. The integration of the new health insurance legislation with the Exceptional Medical Expenses Act – a universal statutory scheme covering long-term care, in force since 1968 – was postponed. The main reason for the focus on medical care was to reduce complexity. The introduction of a single comprehensive national health scheme covering both curative medicine and long-term care would have been too complicated and politically risky. Therefore, the government gave priority to the reform of health insurance legislation. 2.4 A new policy paradigm: regulated competition In Chapter 1 we pointed out that health care (or any reform) requires a shift in policy paradigm. A paradigmatic shift was needed to overcome the failure of the established system and improve health system performance. The new 7 In an arrangement under public law, individuals who meet the eligibility criteria are automatically insured. In an arrangement under private law, individuals are insured only if they take out a health insurance plan. The market reform in Dutch health care24 paradigm was found in the model of regulated competition. Based on insights from health economics theory, the model contended that competition would be the best institutional instrument to achieve an efficient and equitable system of health care, provided strict regulation preserved public values in health care, and, most notably, to provide an adequate system of risk equalization to uphold risk solidarity. Thus, competition could be reconciled with public values. Through its strong emphasis on efficiency, regulated competition also would help to control costs (Enthoven, 1988; 1993). The model is based on a set of assumptions that can be summarized as follows: 1. Freedom of choice enables consumers to make choices that best fit their preferences and, thus, increases the consumer surplus. 2. Freedom of choice elicits competition in the health insurance market, encouraging insurers to offer clients a health plan that is competitive in terms of costs and quality of care. 3. On a competitive market, active purchasing by insurers is a prerequisite to offering clients a health plan that is attractive in terms of quality and costs. 4. Active purchasing incentivizes providers to perform better in terms of costs and quality and, thus, spurs innovation and the growth of productivity. 5. The state must regulate health care to preserve public values (universal access, quality of care, fiscal sustainability and effective competition). Van de Ven et al. (2013) have pointed out that regulated competition works only if a set of preconditions is fulfilled. These preconditions include, among others, freedom of consumer choice in health insurance, full consumer information and market transparency, risk-bearing buyers (insurers) and sellers (providers), competitive markets, freedom to contract, effective risk equalization, effective quality supervision and guaranteed access to basic health care. In their view, Dutch health care fulfilled these conditions reasonably well. Paraphrasing Schultze, the model of regulated competition can be conceptualized as a deliberate attempt to make “public use of private interests” (Schultze, 1977). Its advocates argued that the extensive set of state regulatory control measures to steer Dutch health care had resulted in a supply-driven system that left consumers, insurers and providers little freedom of choice. Hospital planning (introduced in the early 1970s) and hospital budgeting (introduced in the early 1980s) had largely failed as policy instruments to foster efficiency and innovation. The same was true for the central regulation of tariffs. The resolution of these problems The market reform in Dutch health care: plan and strategy 25 required a transition from a hierarchical type of governance towards a devolved model of governance. Operational decisions on beds, budgets, tariffs, quality of care and many other issues had to be delegated to the level where risk-bearing insurers and providers interacted with each other. Insurers were accorded an agency role on behalf of their clients in negotiating contracts with providers. The state would carry only “system responsibility”, which meant that the state was responsible for the regulatory framework to preserve the public values and the organization of an effective supervisory system. To spur competition, the new regulatory framework had to offer insurers and providers enough leeway for entrepreneurship. And, last but not least, consumers required more freedom to make their own choices. 2.5 The shape of regulated competition in Dutch health care Fig. 2.1 presents a stylized picture of the financial flows in the new health insurance scheme. The figure shows that subscribers pay a flat-rate premium (nominal premium) to their insurer. To spur competition in the health insurance market, this premium is set by each insurer separately. The second financial flow refers to income-related (social) contributions set by the government. These contributions are paid by employers for their workers and by self-employed persons and pensioners, and they flow into the risk equalization fund. Furthermore, the state pays the premium for children under 18 years by means of a tax-financed contribution to the risk equalization fund. Finally, patients pay a mandatory deductible set by the government. The Health Insurance Act requires that the income-related contributions cover 50% of health care expenditures; the other 50% is covered by nominal premiums, taxation (state grant for children) and the revenues from the mandatory deductible (see Fig. 2.1).8 The policy paradigm of regulated competition offers only an abstract template for reform. It must be translated into concrete regulations to be effective. This section presents an overview of the most important regulations. Table  2.1 presents an overview of how these regulations are connected with the policy goals of health care reform. 2.5.1 Universal access To guarantee universal access, the Health Insurance Act introduced a single mandatory basic health insurance scheme covering all legal residents of the 8 In 2020 the share of income contributions in financing the Health Insurance Act was 49.4%, and the share of nominal premiums was 39.2%. The state grant for children and the mandatory deductible accounted for 5.3% and 6.1%, respectively (MoH, 2019b). The market reform in Dutch health care26 Netherlands. This scheme put an end to the traditional dual structure of health insurance, with sick funds and a complex mix of public and private insurers. The basic health insurance scheme must be distinguished from supplementary health plans, which pay for health services not covered by the basic health insurance scheme (for example, physiotherapy, dental care for adults, glasses). In contrast to the basic insurance scheme, supplementary health insurance is voluntary. It is a “pure private” arrangement offering insurers much freedom – for instance, regarding the benefit package, enrolment criteria and premium-setting. The Health Insurance Act does not regulate supplementary health insurance, with one exception: Insurers are forbidden to cancel a supplementary plan if consumers switch to another insurer for their basic health plan. With a view to promoting universal access, the Health Insurance Act covers a broad standard package of health services, including, among others, general practitioner care, specialist care, maternity care, hospital stays, geriatric rehabilitation, mental health care, prescription medicines, community nursing and care, medical aids and devices, and dental care for children until the age of 18 years. Some types of care, including physiotherapy and dental care for adults, are only partially covered. Insurers are permitted to require a referral or a pre-authorization for some types of care but are forbidden to offer consumers a health plan with a restricted benefit package (a ban on package differentiation). The Minister of Health decides on the content of the standard benefit package on the advice of the National Health Care Institute (Zorginstituut Nederland, ZiN). The main Fig. 2.1 Structure of regulated competition in Dutch health care Providers State Subscribers Employers Self-employed Pensioners Risk equalization fund Contribution for children <18 years Health care provision Incom e-rela ted co ntribu tion Nomin al pre mium + ma ndato ry ded uctibl e Insurers Health Insurance Act Health Care Market Regulation Act Health Care Allowance Act Health Care Institutions Accreditation Act Ca re al lo w an ce Payments to providers Risk-adjusted payments The market reform in Dutch health care: plan and strategy 27 decision criterion is whether the service is essential, effective, cost-effective and unaffordable for individuals (Kroneman et al., 2016). Finally, the Health Insurance Act obligates insurers to accept each applicant (open enrolment) and guarantees its subscribers access to all health services listed in the standard benefit package. 2.5.2 Freedom of choice Freedom of choice is spotlighted as a central element of the market reform, enabling consumers to make choices that best fit their individual preferences and spurring competition. The Health Insurance Act contains several provisions for freedom of choice: Consumers can choose their own insurer and switch to another insurer at the end of each year. They also have freedom of choice with regard to the type of health plan. They can opt for a benefit-in-kind plan (naturapolis), which obligates the insurer to guarantee access to adequate and timely care. Another option is to purchase a cost-reimbursement plan (restitutiepolis). A third option is to take out a managed care plan (budgetpolis). This plan can be described as a lower-priced benefit-in-kind plan with a limited network of preferred providers for planned care. In addition, the Health Insurance Act enables consumers to opt for a voluntary deductible, on top of the mandatory deductible, in exchange for a lower premium. The maximum voluntary deductible is regulated by the Table 2.1 Policy goals and instruments of the market reform Policy goal Instruments Universal access • Single mandated basic health insurance scheme • Standard benefit package • Open enrolment • Duty of care Freedom of choice • Free choice of health insurer with option to switch • Free choice of health plan with option to switch • Voluntary deductible • Free choice of doctor Solidarity • Provisions for risk solidarity: community rating and risk equalization • Provisions for income solidarity: income-related premiums and care allowance Quality of care • (Selective) contracting • Accreditation of providers Financial sustainability • Competition on the health insurance market • (Selective) contracting • Free pricing • Exposure to financial risk • Mandatory deductible • Yearly global budget constraint The market reform in Dutch health care28 government and currently set at €500. Finally, the Health Insurance Act respects the free choice of doctor, but subscribers with a managed care plan may be faced with some restrictions for planned care. Freedom of choice is not unlimited. The purchase of a basic health insurance plan is mandated so as to avoid free-rider behaviour, and the ban on package differentiation implies that consumers cannot choose a plan with limited coverage. The reason for these restrictions is to preserve solidarity. 2.5.3 Solidarity The new health insurance scheme would have been politically infeasible without strict regulations to preserve solidarity. A solidary system of health care financing is considered essential for upholding the policy goal of universal access. A distinction must be made between risk solidarity and income solidarity. Risk solidarity means that the premium for the health plan is not related to a person’s health condition; income solidarity means that the premium is related to income. Risk solidarity To achieve risk solidarity, the Health Insurance Act employs two important instruments: (1) the obligation to apply community rating and (2) ex ante risk equalization. The obligation to apply community rating means that insurers must charge their subscribers the same premiums for the same type of health plan (a ban on premium differentiation). However, premiums may differ by type of health plan. For instance, an insurer may charge a lower premium for a plan with restrictions than for a cost-reimbursement plan. Ex ante risk equalization is an essential financial instrument to preserve risk solidarity. Its purpose is to create a level playing field by compensating for all differences in the risk profile of the insurers’ population. Thus, insurers receive a risk-adjusted payment from the risk equalization fund for each insured person.9 In the ideal situation, then, premium differences between insurers express only differences in efficiency and purchasing strategies rather than differences in risk profile. The development of ex ante risk equalization started in the early 1990s in the sick fund scheme. Initially, risk equalization involved only age and sex as risk parameters. In 1992 these parameters were used for 20% of health care expenditures. As a result, sick funds were at risk for only 2.5% of their expenses. 9 The risk equalization fund also pays providers directly to cover the costs of teaching programmes and the availability of academic health facilities and facilities for specialized burn units, trauma care, emergency care and acute obstetric care. The market reform in Dutch health care: plan and strategy 29 Since then, the model has become ever more sophisticated, and in 2001 the percentage for which insurers were at risk had risen to 38% (Douven, 2004). The current model (2019) includes both general parameters and disease-related parameters. The general parameters are age, gender, source of income, number of persons in household, socioeconomic status and region. The disease-related parameters include, among others, pharmaceutical cost groups, diagnostic cost groups, physiotherapy cost groups and multiple-year high and low spending and medical devices cost groups. The number of risk classes adds up to more than 500. Despite its sophisticated structure, there are, nevertheless, categories of the insured with a predictable loss after compensation (see Chapter 3). Income solidarity To achieve income solidarity, the Health Insurance Act requires that employees, self-employed persons and pensioners pay an income-related contribution at a rate set annually by the government. Employers pay the social contribution for their employees. A second policy instrument for income solidarity is the care allowance (premium subsidy). This allowance compensates individuals and families for the steep rise in the average nominal premium in 2006.10 This raise would have had unacceptable financial consequences for low-income and middle-income individuals and families. To compensate for this, the reform was complemented with the Health Care Allowance Act (Wet op de zorgtoeslag). Subscribers meeting the income criteria receive a care allowance (premium subsidy) set by the government. The allowance is calculated as the difference between the standard premium and the norm premium set by the government. The standard premium is the average nominal premium that the Ministry of Health expects the insured will pay in a given year. The norm premium is set as a percentage of a person’s income. Under the current regulations, single-person households with an income higher than €31 000 and multiple-person households with a total income higher than €39 500 do not qualify for allowances. The total costs of the care allowance system were estimated at €5.2 billion in 2020 (MoH, 2019b). 2.5.4 Quality of care To enhance the quality of care, insurers should engage in active purchasing. The contracts with providers that they negotiate on behalf of their subscribers can include agreements on quality of care. The Health Insurance Act permits insurers to apply selective contracting. Furthermore, the Health Care Institutions 10 Former sick fund subscribers paid an average nominal premium of €384 in 2005. The introduction of the reform in 2006 meant that this premium tripled to an average of €1060. The market reform in Dutch health care30 Accreditation Act (Wet Toelating Zorginstellingen) requires providers to apply for accreditation. The most important requirements a provider must fulfil to receive accreditation relate to its governance structure, its operational management and the accessibility of emergency services. 2.5.5 Financial sustainability Several instruments are used to achieve the policy goal of financial sustainability. The first instrument is competition in health insurance to encourage insurers to offer subscribers competitive premiums. Other important instruments are selective contracting, lack of price controls and financial risk exposure. Exposure to financial risk is assumed to incentivize insurers to be competitive on prices and to act as efficiently as possible. The mandatory deductible has proved to be one of the most disputed elements of the reform. It was introduced in 2008 as an alternative to the no-claim arrangement that had failed as an instrument to avert the risk of moral hazard. The deductible is set by the government. It rose from €150 in 2008 to €385 in 2016. Since then, the government has refrained from further increases for political reasons. Children under 18, maternity care, general practitioner consultations and a few other services are exempt from the mandatory deductible.11 The mandatory deductible must be distinguished from the voluntary deductible (see above). Finally, the Health Care Market Regulation Act (Wet marktordening gezondheidszorg) enables the Minister of Health to set an annual ceiling on health care spending and to recoup excess spending. 2.5.6 Institutional structure The Health Care Market Regulation Act also regulates the relationship between the Minister of Health and the Dutch Healthcare Authority (Nederlandse Zorgautoriteit, NZa).12 The Healthcare Authority is a regulatory agency at arm’s length from the government. It is charged with regulatory and supervisory tasks, including the concrete description of reimbursement titles of health services covered by statutory health insurance, the determination of maximum prices of health services that are excluded from free pricing and oversight. Also, the Authority monitors the effects of regulated competition and advises the Minister of Health on its further development. The Health Care Market Regulation Act establishes a delicate balance between the roles and responsibilities of the Minister of Health and the Healthcare Authority. As a general principle, the Minister bears “system responsibility” and the Healthcare Authority has “case responsibility”. 11 Insurers are permitted to exempt specific health services from the mandatory deductible. 12 The Dutch Healthcare Authority is the successor of the Central Healthcare Tariffs Agency. The market reform in Dutch health care: plan and strategy 31 2.6 Third round: the politics of market-making The introduction of the Health Insurance Act was, by any standard, a landmark event in the reform process. Nevertheless, it was only the first step, because the scope of competition and entrepreneurship was still restricted. Several market- making decisions still had to be taken. Some of these decisions were politically controversial, and on some occasions the government was unable to build a political majority for its proposals. 2.6.1 Free pricing Market competition requires clearly defined services to enable negotiations between insurers and hospitals on prices and quality of care. Until 2005, this was hardly possible. Hospitals were still (largely) funded through a global, function-based budget13 on the basis of a set of parameters (adherence, capacity and production volume). As early as 1994, the Ministry of Health and the national representative associations of general hospitals, university hospitals, medical specialists and health insurers had started a concerted project to revise the funding of hospital care. The project resulted in a new model based upon Diagnosis Treatment Combinations (Diagnose behandel combinaties, DBCs). Each DBC covers the complete process of care, from the first consultation with a medical specialist through completion of treatment (Hasaart, 2011). The new model was introduced in 2005. However, at that time the room for free pricing (described as the B-segment in hospital funding) was still quite limited. It contained only 1246 DBCs of an estimated 30 000 DBCs and amounted to about 10% of the total budget for hospital care. These 1246 DBCs were routine procedures such as cataract surgery and total hip and knee replacements. For the rest, the system of function-based hospital budgeting remained in place. The scope of free pricing has been extended since 2006 in a number of successive steps. The tariffs of the DBCs in the A-segment (regulated prices) continued to be set by the Dutch Healthcare Authority (Table 2.2). Each expansion of the room for free pricing prompted a heated political dispute over its merits within the government coalition. As a compromise, the coalition parties agreed that decisions to extend the scope of free pricing could be considered only after careful evaluation by the Healthcare Authority of its effects thus far. There had to be empirical evidence that free pricing had, indeed, resulted in lower prices and that the price effect had not been cancelled out by a 13 The purpose of the function-based model was to link the hospital budget to its function in the provision of medical care. Budgets differed by specialty, and big hospitals were paid a higher fixed tariff than small hospitals on the assumption that big hospitals performed more complex procedures than small hospitals. The market reform in Dutch health care32 countervailing increase in the volume of care. A plan of the Minister of Health to extend the room for free pricing to 50% in 2011 failed because of resistance from the Ministry of Finance. The political debate on free pricing took place in the context of the financial crisis that had hit the Netherlands in 2009. In the government’s view, the crisis required a policy of austerity, from which health care could not be exempted. The Minister of Health resisted raising co-payments and reducing service packages. Instead, the Minister insisted on the extension of free pricing to curb expenditure growth. Recognizing that free pricing alone would fail, he simultaneously argued for an extension of the health policy agenda of care innovation to curb the high costs of overtreatment and undertreatment and eliminate the production incentive in health care provision. The Minister saw the needed innovation as the common responsibility of government, insurers and providers. Eventually, it was the Minister’s Liberal successor, Edith Schippers, who decided to push through. In her view the reform was “stuck in the middle” between a supply-driven and a demand-driven health care system and had largely failed to stimulate innovation. The many safety nets to restrict the insurers’ financial risk hardly incentivized insurers to negotiate low prices. In order to put an end to the “worst combination of both worlds”, the room for free pricing was expanded from 34% to 70% (MoH, 2011a). Hospital funding was further simplified by the introduction of about 4400 DOTs (DBCs on the way to transparency). The old system had proved very complicated and quite inefficient. Not only did a small number of DBCs cover the majority of cases and costs,14 but also the funding system failed to incentivize doctors to work towards integrated care. The revision of the funding model was intended to enhance efficiency and encourage the development of integrated care pathways. To assuage the concerns of hospitals about the financial impact of the revised funding model, a three-year safety net, by means of “shadow budgets”, was developed to protect them against budget shocks. 14 Four percent of DBC codes and 14% of diagnosis codes represented 80% of all cases, and 3% of DBC codes and 14% of diagnosis codes represented 80% of total costs (Krabbe-Alkemade, 2014: 41). Table 2.2 Size of A-segment and B-segment in hospital funding Year A-segment (regulated prices) B-segment (negotiated prices) 2005 90% 10% 2006 90% 10% 2007 90% 10% 2008 80% 20% 2009 66% 34% 2012 30% 70% The market reform in Dutch health care: plan and strategy 33 2.6.2 Termination of hospital planning and capital investment regulations An important element of the market reform was the termination of the Hospital Planning Act (Wet Ziekenhuisvoorzieningen) in 2008. This legislation, introduced in 1971, enabled the government to regulate the supply and regional distribution of hospital facilities (and residential facilities for long-term care). Hospital planning was considered necessary for cost control. Legislation required hospitals to apply for a licence (certificate of need) to start, build or reconstruct a hospital. Planning legislation also regulated the number of specialties of each hospital. In practice, hospital planning often entailed lengthy negotiations between hospitals and the Ministry of Health on the terms of the licence. After hospitals had acquired a licence, they could pass on the costs of rent and depreciation in their tariff of an inpatient day, with the result that they did not bear any financial risk on their capital investments. Neither did the banks as financiers of these investments. The argument for terminating hospital planning was its incompatibility with competition. State planning not only hindered new players from entering the market (see Chapter 5), but also restricted hospitals’ latitude for strategic decision-making. Under the new regulation, providers still need a state licence to operate, but the conditions have been considerably relaxed. The termination of hospital planning (with some exceptions) meant a significant extension of the hospitals’ discretionary space. However, now they also incur the financial risk of their investment decisions. The policy assumption was that the devolution of the financial risk would discipline hospitals in strategic decision-making and encourage their financiers to critically assess the financing of hospital investment plans. The termination of planning meant that the Minister of Health was no longer responsible for the survival of individual hospitals and other providers. “System responsibility” held that the Minister was responsible only for ensuring access to health care, not any specific provider institution. In line with this, the arrangement for government financial support in case of a looming bankruptcy was abolished in 2013. However, since access to health care may be at risk in circumstances beyond the insurers’ control, the Healthcare Authority issued a regulation in 2015 to ensure access to essential health services including ambulance care, emergency medical and psychiatric care, acute obstetric care and long-term care. The regulation enables the Authority to intervene with financial assistance if an insurer has convincingly demonstrated doing everything in its capacity to guarantee access to essential health services and no other solution is available. The market reform in Dutch health care34 2.6.3 Extension of coverage of the Health Insurance Act Mental health care had traditionally been covered by the Exceptional Medical Expenses Act. This changed in 2008 with the transfer of curative mental health care shorter than one year to the Health Insurance Act. As a consequence, insurers were charged with purchasing these services. In 2013 all mental health care shorter than three years as well as geriatric rehabilitation followed the same route. The coverage of the Health Insurance Act was further extended in 2015 with the transfer of community nursing and care as part of the reform of long-term care. Due to these transfers, the Health Insurance Act increasingly turned into the home base for services on the interface between long-term care and medical care. 2.6.4 Termination of ex post equalization and retrospective macro- compensation Earlier, we described risk equalization as an important policy instrument to level the playing field in health insurance. Insurers were compensated for bad risks by means of ex ante risk-adjusted capitation payments. Because of many imperfections in the model of risk equalization, the market reform was complemented with two temporary safety arrangements. The first was ex post equalization to mitigate the insurers’ financial risk. The effect of this arrangement was that their deficits vis-à-vis the ex ante equalization were partially compensated, and that surpluses were partially being skimmed off. The second arrangement compensated each insurer equally for the difference between health care expenditure and the government’s global annual budget estimate. A drawback of both arrangements was that they weakened the motivation of insurers to negotiate lower prices. For this reason, the Minister decided to accelerate the phase-out of these arrangements. Fig. 2.2 highlights how the reduction of ex post compensation caused a substantial increase in the insurers’ financial risk as of 2015. 2.6.5 The experiment with free pricing in dental care The Health Insurance Act covers dental care for children (<18 years). Coverage of dental care for adults is limited to certain surgical procedures, X-ray examinations and dentures. Consumers can take out a supplementary health insurance plan to cover the costs of regular dental care, but plans usually cover these costs only in part. In 2012 the government started a three-year experiment with free pricing in dental care. The policy goals of this experiment were cost control, better service quality and promotion of entrepreneurship. However, the experiment was terminated after one year because of complaints of significant price increases. The Healthcare Authority found that price increases in some subgroups of treatments The market reform in Dutch health care: plan and strategy 35 averaged 10.7% higher in 2012 than in 2011 (NZa, 2012). Also, the experiment with free pricing had adversely influenced the use of dental care. Researchers found that the proportion of treatment sessions involving preventive services had decreased by 3.4% among adults and by 5.3% for children and adolescents (Trescher et al., 2020). 2.6.6 Reimbursement of non-contracted care The Health Insurance Act permitted insurers to refuse full-cost reimbursement if a subscriber with a benefit-in-kind plan had consulted a non-contracted provider. However, the legislation did not mention a percentage of the costs that insurers had to reimburse to comply with their obligation to guarantee free choice of doctors. In a lawsuit the court judged that a low reimbursement percentage would severely restrict access to non-contracted providers and for this reason obligated insurers to reimburse 75–80% of the costs of non-contracted care. After consultations with the lead associations of health insurers and hospitals, the government submitted a proposal to the parliament to relieve insurers of this obligation. The goal of the proposal was to stimulate selective contracting and prevent consumers from frustrating selective contracting by consulting non- contracted providers. The lower chamber approved the revision of the Health Insurance Act, but the upper chamber turned it down in December 2014 after strong lobbying by providers, particularly those in private practice. The opponents reasoned that a lower percentage would severely restrict the legally established free choice of doctors. Fig. 2.2 Financial risk of health insurers, 2012–2018 0 10 20 30 40 50 60 70 80 90 100 2012 2013 2014 2015 2016 2017 2018 fully at risk Partially at risk full compensation Source: MoH. The market reform in Dutch health care36 2.6.7 For-profit health care For-profit health care and paying dividends to shareholders have traditionally been a contested topic in Dutch health care policy-making. While various providers are permitted to act for-profit (for example, providers of general practitioner care, community nursing, home care, dental care and patient transport), health legislation has always banned profit-making on specialist care (that is, hospital care). All Dutch hospitals are private, not-for-profit entities. They are obliged to re-invest any residual surplus in health care or use it as a financial buffer. In practice, however, the distinction between for-profit and not-for-profit hospitals is blurred, and there are several routes to circumvent the ban. The origins of the ban trace back to the period when the state guaranteed all costs of rent and depreciation of capital investments for health care facilities (Jeurissen, 2010). The ban was also ideologically motivated. Opponents of its lifting argued that hospitals would turn into money-driven organizations. The Liberal Minister of Health Hans Hoogervorst (2003–2007) – architect of the health insurance reform in 2006 – refused to lift the ban. Both on practical and political grounds, he considered it inopportune to burden the market reform with an ideological discussion about this controversial topic. Lifting the ban was something for later. The issue reappeared on the political agenda in 2010 and resulted in the adoption of a legislative proposal by the lower chamber in 2014 to lift the ban under “strict conditions”. The main argument for the proposal was to enable hospitals, as risk-bearing agents, to attract private capital. However, given the contested nature of the issue and the probability of a political defeat, the Minister of Health Schippers (2010–2017) decided to postpone a vote on the proposal in the upper chamber. In 2017 the government announced an investigation of the judicial, economic and practical consequences of the option of replacing the ban with a diversified model of paying dividends. Arguments for the investigation were the need to create a level playing field for provider organizations, their need to attract private capital and the illusion of a clear-cut dividing line between for-profit and not-for-profit medicine. However, in 2019 a new plan for permitting for-profit hospital care under strict conditions was withdrawn because of lack of political support (MoH, 2019a). Also controversial was the issue of how to deal with for-profit insurers. The Health Insurance Act allowed for-profit health insurers. The former sick funds had promised the government to abstain from paying dividends to shareholders for a 10-year period that ended in 2016. To prevent insurers from starting to pay dividends to shareholders, a left-wing coalition of members of parliament The market reform in Dutch health care: plan and strategy 37 took the initiative to permanently forbid this by law. Its legislative proposal was adopted by the lower chamber in 2017, but is still pending in the upper chamber. 2.7 New corporatism: collective framework agreements Advocates of regulated competition are, in principle, critical of top-down global budgets for health care. Although Enthoven concedes that regulated competition will not automatically hold spending to acceptable levels, even under ideal circumstances, and that global budgets would avoid some of the worst inefficiencies and disincentives, he nevertheless objects to the use of global budgets because they would make health care spending contingent on political choices. Global budgets are in his view at odds with the basic principles of regulated competition (Enthoven, 1993). Schut and van den Berg (2010) point to two other problems with global budgets: The threat that the government will skim off excess spending may encourage hospitals to overspend. Furthermore, spending constraints may deter new entrants. In spite of these objections, the market reform did not end the practice of yearly global budget constraints (Budgettair Kader Zorg), which were introduced in the mid-1990s as a new policy instrument to control expenditure growth. After an adjustment in 2012, the Health Care Market Regulation Act gave the Minister of Health the formal authority to recoup excess revenues if total spending surpassed the spending limit. Not surprisingly, providers lobbied against the use of this “last resort” instrument. They not only considered it a hierarchical policy instrument that fit poorly in the Dutch culture of consultation, but also complained frequently about the gap between the global budget and the growing demand for health care. However, this would change in 2013. In the wake of the financial crisis (the gross domestic product (GDP) had dropped by 3.4% in 2009), the government announced a substantial reduction of the benefit package of the Health Insurance Act as a cost-saving measure. Because of the controversial nature of this proposal, the Minister abstained from endorsing this plan and agreed instead on an alternative approach based on shared responsibility. This strategy consisted of a series of collective framework agreements (bestuurlijke hoofdlijnenakkoorden) with the lead organizations of providers and insurers that included not only a yearly ceiling on net expenditure growth but also an agenda for health care innovation such as efficient prescribing of medicines, substitution of primary care for hospital care and promotion of “the right care at the right place”. The agenda for innovation reflected the notion that budgetary measures to control expenditure growth would ultimately fail. This had been the lesson of the government’s policy of budgetary restraint in the 1990s. Table 2.3 presents an overview of the agreed annual growth percentages for the period 2012–2019. The market reform in Dutch health care38 The framework agreements signify the “corporatist style” of health care policy- making and echo the role of institutionalized shared responsibility in Dutch public policy-making. At the same time, the agreements also served the self- interest of each signatory. They provided the state with an alternative instrument to control expenditure growth and mitigated the financial risk of insurers, who would be exposed to greater risk because of the envisaged accelerated termination of ex post risk compensation. The interest of health care providers was to avert the government’s proposal to reduce the benefits package of the Health Insurance Act. Furthermore, they got something in return for signing the contract. Since the pressure to make collective agreements was very high, there is good reason to describe them as agreements made “under the shadow of hierarchy”. 2.8 Summary and discussion Health system reform is a deliberate attempt to make a major change in a country’s health care system. It is an orchestrated effort to implement “system change” by introducing new rules of the game. Under this definition, there is no question that the market reform in Dutch health care can be classified as a reform. The introduction of the universal mandatory health insurance scheme was certainly the most striking change. The new rules altered the roles and responsibilities of the state in health care, the relationship between insurers and providers and, last but not least, the role of consumers. The state abandoned a great deal of its direct control over the supply of health care facilities, insurers were incentivized to adopt active purchasing, the providers’ room for entrepreneurship was enlarged, and consumers were given more freedom of choice. The reform did not involve a change of policy objectives. Instead, regulated competition was intended as an alternative institutional arrangement to achieve the traditional goals of Dutch health care policy: solidarity in health care financing, universal access, quality of care and financial sustainability. Table 2.3 Overview of collective framework agreements on net expenditure growth (percentages) 2012 2013 2014 2015 2016 2017 2018 2019 MSC 2.5 2.5 1.5 1.0 1.0 1.0 1.0 0.8 HD 2.5 2.5 2.5 2.5 – – – – MHC – 2.5 1.5 1.0 1.0 1.0 1.0 1.3 GPC – 3.0 2.5 2.5 2.5 2.5 2.5 2.5 CN – – – – – – – 2.4 Abbreviations: MSC, medical specialist care (hospitals); HD, non-salaried hospital doctors; MHC, mental health care; GPC, general practitioner care; CN, community nursing. The market reform in Dutch health care: plan and strategy 39 In Tuohy’s classification of the scale and pace of policy change, the market reform fits in the quadrant of the “blueprint” approach (large scale, gradual pace) (Tuohy, 2018). The introduction of the new legislation put an end to the bifurcated health insurance landscape that had existed for many decades. However, it was only the first – major – step in the reform process. The termination of hospital planning came in 2008. Free pricing in hospital care was extended only gradually, and a temporary financial arrangement was in place to accommodate the transition from global budgets to performance-based hospital funding. Two temporary compensation arrangements had to ensure a disturbance-free transition from the old to the new situation. Lifting the ban on for-profit hospital care was postponed and never materialized. However, the stepwise introduction of regulated competition was only in part a strategy of averting risks and creating opportunities for policy learning and, if necessary, corrective action. Politics also influenced the pace of the reform process. Recurrent political controversies over the appropriateness and shape of regulated competition in health care were reason for the Minister’s complaint in 2011 that the reform had “stuck in the middle” and cause for her to opt for a major leap forward. The reform can be conceptualized as a rational attempt to resolve the persistent lack of efficiency and innovation in a supply-driven health care system. A new policy paradigm based on the principles of regulated competition provided the intellectual basis for the alternative design of health care. The old paradigm of hierarchical control and detailed regulations was considered exhausted and had to be replaced by a new design of the relationships among state, market and civil society that built not only upon the neoliberal wave in public policy-making but also upon the traditional Christian body of thought of subsidiarity and delegation. However, the reform was more than a rational exercise. Ideological controversies, vested interests, power relations and the persistent need for compromises also influenced its shape and pace. The 20-year period between the publication of the Dekker report and the enactment of the Health Insurance Act was no coincidence. Building a political majority for new legislation required political manoeuvring and a willingness to compromise. The government’s policy of extending the scope of free pricing was rife with political struggle. Safety nets were not only a vehicle to avoid system disturbances but also an instrument to overcome political resistance. A revision of the Health Insurance Act to absolve insurers from the obligation to pay 75–80% of the costs of uncontracted care was turned down in the upper chamber, and a renewed proposal to permit for- profit hospital care under strict conditions was withdrawn because of lack of political support. The model of DBCs to fund hospitals and the new model to The market reform in Dutch health care40 reimburse general practitioners were “negotiated products”. For-profit health care has remained controversial. The reform was also influenced by the historical context of Dutch health care. Close inspection demonstrates that the reform consists of a complex mix of change and continuity. The new health insurance scheme, the termination of hospital planning, the upgrading of the role of health insurers in contracting and the redefinition of the role of the state in health care point to major changes. The same is true for the introduction of a contractual relationship between insurers and providers, the devolution of financial risks to insurers and providers and the enhancement of consumers’ choices. At the same time, however, the reform also respected the legacy of Dutch health care. For instance, it would have been politically unacceptable without strict provisions to ensure risk solidarity and income solidarity in health insurance. The mandatory character of the basic health insurance scheme was rooted in the tradition of the Sick Fund Act and was widely accepted. The reform also built upon the traditional public–private mix in Dutch health care. The deeply entrenched informal rules of the game for health care policy-making affected the shape and pace of the reform. Reform decisions followed extensive consultations among stakeholders. The impact of the neo-corporatist style of health care policy-making is perhaps best visible in the collective framework agreements on cost control and other issues. This chapter presented an overview of the course and shape of the market reform in Dutch health care based on the principles of regulated competition. However, these principles were abstractions that gave only global direction to the reform. There was no single or linear route from principles to concrete regulations. Politics and the institutional context always influenced the shape and pace of the reform and accommodated it to the reality of Dutch health care. Hence, what we described in this chapter was the Dutch version of regulated competition in health care. The question is how the new regulations really changed the structure and performance of Dutch health care. This question is the topic of the research presented in the following chapters. Chapter 3 Health insurance reform in practice 3.1 Introduction The 2006 Health Insurance Act meant a significant change in the bifurcated structure of Dutch health insurance. By integrating the former sick fund scheme and all other health insurance schemes into a single mandatory basic health insurance scheme covering the entire population, the new legislation put an end to the traditional role of sick funds and private insurers. The private health insurance market was restricted to supplementary health insurance only. In Chapter 2 we have seen that the reform gave insurers a central role in health care. Because consumers have the freedom to switch to another insurer at the end of each year, insurers must compete on the health insurance market by offering consumers a competitive health plan in terms of costs and quality of care. To be successful in the health insurance market, active, or strategic, purchasing is a prerequisite. This chapter investigates how in practice insurers have shaped their new upgraded role in the health insurance market and how consumers have made use of their freedom of choice in health insurance. The next chapter will analyse active purchasing. The Health Insurance Act contains many provisions that insurers must take into account when acting in the health insurance market – the “regulation” in regulated competition. In particular, they are bound to the standard benefit package of health insurance (a ban on package differentiation); they must accept each applicant (open enrolment); and they must apply community rating (a ban on risk rating). Although these regulations restrict insurers’ actions, health insurance legislation still offers insurers ample room for developing their own implementation strategy (including a market strategy). Thus, the effects of the market reform will be influenced by the insurers’ implementation strategies. As spelled out in the previous chapters, the main policy goals of the reform were to achieve a health care system offering patients high quality of care that is accessible to each person (universal access), based upon solidarity and affordable The market reform in Dutch health care42 from a financial perspective (financial sustainability). The focus in this chapter is upon the goals of universal access and solidarity. The structure of the chapter is as follows. Section 3.2 presents an analysis of the implementation strategy of insurers. Which choices did they make, and what were the consequences of those choices for universal access? The section ends with a brief analysis of the financial performance of insurers. We will see that their financial results sparked a dispute over “high profits”, which had consequences for premium-setting. Section 3.3 presents an analysis of how consumers have reacted to the new health insurance legislation. How did they make use of their freedom of choice? Who benefited most from the new options? Another important topic is the problem of the uninsured and defaulters. Section 3.4 analyses the effects of the reform on solidarity. This analysis is continued in sections 3.5 and 3.6 with a brief discussion of public support for a solidary health insurance scheme. The chapter ends with a discussion of the results of regulated competition in health insurance and a few policy lessons from the Dutch experience with it (section 3.7). 3.2 Insurers’ implementation strategy The basic health insurance scheme is mainly implemented by private not-for- profit insurers. Only a few insurers, among them Achmea (one of the market leaders), operate on a for-profit basis. What strategies did the insurers follow in implementing the Health Insurance Act? 3.2.1 Mergers and takeovers The prospect of the reform of health insurance confronted insurers with a strategic issue: how to anticipate a more competitive market and how to deal with increased exposure to financial risk? Fig. 3.1 shows that the market reform sparked a period of consolidations in the health insurance market. Fig. 3.1 shows a gradual but significant decline in the number of insurers in the pre-reform period, from 83 in 1997 to 58 in 2005. This decline was due to mergers and takeovers that, to a great extent, were motivated by the prospect of reform (an anticipatory effect). The sudden drop from 58 insurers in 2005 to 33 in 2006 is a direct effect of the reform. In the pre-reform period, sick funds were required to set up a separate organizational entity to ensure a smooth transition to private health insurance for clients crossing the income threshold. The market reform made this organizational split no longer necessary, with the result that the total number of insurers dropped. Also, some smaller private insurers sold their businesses: The number of small insurers (<50 000 enrollees) fell from 19 in 2004 to 3 in 2007 (Vektis, 2006; 2008). Health insurance reform in practice 43 Since 2006 the number of insurers has further decreased, from 33 in 2006 to 24 in 2019. Two main reasons for consolidation, both in the pre-reform and post-reform periods, were assertion of market power and increased exposure to financial risk. Insurers considered merging an appropriate strategy to respond to the restructuring of the health insurance market. The current Dutch health insurance market is concentrated. In 2019 four insurance groups (Achmea, VGZ, CZ and Menzis) had a total market share of 85.9% (NZa, 2019a). Nevertheless, the market share of the “big four” has shrunk somewhat. For instance, in 2014 their market share totalled more than 90%. Although insurers operate nationwide, their activities often concentrate in certain regions. Measured by the Herfindahl-Hirschman Index (HHI), the insurance market has always been concentrated in most regions, although presently somewhat less than a few years ago. At the national level, the HHI dropped from 2119 in 2013 to 2035 in 2019 (a market is considered concentrated if the HHI is more than 2000) (NZa, 2019a). These figures suggest that the health insurance market has gradually become somewhat more competitive. An important condition for competition is the threat of new entrants. Until 2018 there were no new entrants. Two initiatives in 2014 and 2016 failed because of high capital requirements and the lengthy and complex admission procedure (insurers need a licence to operate from the Central Bank of the Netherlands (DNB)). In 2018 a new insurer entered the market via a takeover. A second Fig. 3.1 Decline in the number of health insurers, 1997–2019 0 10 20 30 40 50 60 70 80 90 19 97 19 98 19 99 20 00 20 01 20 02 20 03 20 04 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 20 14 20 15 20 16 20 17 20 18 20 19 Nu m be ro fi ns ur er s Source: Vektis, 2019a. The market reform in Dutch health care44 entrant followed in 2019.1 The conclusion is that, so far, the Dutch health insurance market has remained rather closed. The concentrated health insurance market has always been an object of criticism. The general complaint is that insurers have turned into powerful players that are able to determine the terms of contracts in health care purchasing. The alleged power position of insurers is also reason for politicians who are to the left of the political spectrum to downplay the role of insurance in health care. Some political parties even argue for the abolition of insurers. In Chapter 7, we will discuss this criticism further. 3.2.2 Health plans The Health Insurance Act establishes a standard benefit package, but it leaves insurers free to introduce several basic health plans (polissen) on the condition that they offer their clients at least one plan without a voluntary deductible. The Dutch Healthcare Authority is in charge of checking whether health plans meet the regulatory requirements. There are three types of health plans: (1) A benefit-in-kind plan (naturapolis) obligates an insurer to guarantee its clients timely access to care. This type of plan was also a defining characteristic of the former sick fund scheme. (2) A cost-reimbursement plan (restitutiepolis), the most common type of health plan in the former private health insurance, guarantees the reimbursement of the costs of medical care up to the “normal” market price level. Insurers also offer (3) a combi-plan, which is a combination of a benefit-in-kind plan and a cost- reimbursement plan. The total number of health plans offered by all insurers has dropped from 71 in 2015 to 59 in 2018. In 2019 consumers could choose from 33 benefit-in-kind plans, 21 reimbursement plans and 5 combi-plans. Benefit-in-kind plans were the most popular: 75% of consumers took out such a plan. The market shares of cost-reimbursement plans and combi-plans were much smaller – 19% and 6%, respectively (NZa, 2019a). Although formally different in structure, the differences between these plans are limited because reimbursement plans function in practice as benefit-in-kind plans.2 With a cost-reimbursement plan, in most cases insurers directly pay the medical care bills of their clients. However, consumers with a reimbursement plan have somewhat greater freedom of choice because their insurer must also fully pay for non-contracted care.3 The reimbursement percentage of non-contracted 1 IptiQ (belonging to Swiss Re) took over the labels Promovendum, National Academic and Besured from insurer VGZ in 2018. However, Promovendum went back to VGZ in 2021. The other new entrant was Maltese EUCARE, which took over the label Aevitae. 2 In fact, many insured people do not know what kind of health plan they have chosen. 3 Note that they are only required to do so for health care covered by the Health Insurance Act. Health insurance reform in practice 45 care is only 75–80% for benefit-in-kind plans.4 In Chapter 2 we saw that a legislative proposal by the government to permit insurers to reimburse a lower percentage, meant to motivate patients to visit only contracted providers, failed in the upper chamber. Opponents argued that a lower percentage would severely restrict patients’ freedom of choice of doctors. In general, reimbursement plans are only slightly more expensive than benefit-in-kind plans. A new development is the introduction of benefit-in-kind plans with some restrictions. These managed care plans – also known as budget plans – restrict the number of providers that a subscriber can consult by reimbursing only 65–80% of the costs of medical treatment when consulting a non-contracted provider. The number of managed care plans has increased from 5 in 2011 to 14 in 2019, with a peak of 17 in 2015. The percentage of those insured by a budget plan has steadily grown from 1.1% in 2011 to 14.1% in 2019 (NZa, 2015; 2019a). The Healthcare Authority has always been critical of the large number of health plans. In practice, insurance groups frequently sell almost identical plans under different labels. For instance, in 2018 consumers had a choice of 59 plans, of which 45 plans were, with exception of the premium, almost identical (NZa, 2018a). In other words, the number of health plans gives a misleading picture of the choice options. What is more, in 2018, 9.8 million insured people had chosen a plan for which an identical but cheaper alternative plan was available (NZa, 2018a). The large number of health plans and their complexity (much small print) obstruct deliberate choice. For this reason the Minister has requested repeatedly that insurers exercise restraint regarding their number of health plans. 3.2.3 Transparency A recurrent issue is lack of transparency. Insurers are obligated to inform their current and potential customers in a timely manner and accurately about their health plan. However, in practice there have been frequent signs of lack of transparency: information was not available, incorrect or misleading (NZa, 2014a). In order to protect the interests of consumers, the Dutch Healthcare Authority issued a specific regulation on transparency in 2014. The regulation obligates insurers to assure the accuracy of their website information and to inform their customers about their purchasing policy. For instance, insurers must inform their customers correctly and in a timely manner about premiums, types of health plans, the providers they have contracted, the cost-reimbursement policy and pre-authorization procedures. The regulation was extended in 2019 to obligate insurers to inform consumers about volume ceilings in their contracts 4 In practice, in order to retain their clients, providers may cover the difference between the actual tariff and the reimbursed part of the tariff. The market reform in Dutch health care46 with providers (Chapter 4). According to the Healthcare Authority, transparency has greatly improved, but it needs further attention. 3.2.4 Open enrolment In order to guarantee consumers free choice of a health plan, health insurance legislation obligates insurers to accept each applicant for the basic health insurance scheme (open enrolment). The Healthcare Authority has repeatedly reported that insurers implement this obligation lawfully (NZa, 2014b). There are no indications of denying applicants access. Nevertheless, there is some evidence of creative practices that, although lawful, conflict with the intentions of the Health Insurance Act. For instance, insurers may put up administrative hurdles in the application procedure or focus their marketing on specific market segments (selective marketing). Other examples are the targeting of supplementary health plans to specific groups or the application of health questionnaires in supplementary health insurance (van Kleef et al., 2014; KPMG/ Plexus, 2014). The Healthcare Authority has urged insurers to stop these practices (NZa, 2014b). In its plan “In perfect health” (Kern-gezond), published in 2015, the national association of health insurers (ZN) promised to do so. Box 3.1 Pre-authorization The Health Insurance Act obligates insurers to cover all medical services listed in their standard benefit package, but it permits them to include a pre-authorization review for specific services. The medical adviser of the insurer decides whether a patient qualifies for coverage of that service. For example, reconstructive surgery, such as breast reconstruction and scar revision, is covered, but insurers may decide that they will pay for it only if the patient meets predetermined conditions. The professional community has formulated a list of conditions to standardize this decision-making, but in practice the list appears open to different interpretations. Some treatments are authorized only if performed by a specific type of specialist. Specific dental treatments, medical aids and devices, specialized mental health care and outpatient prescription medicines are other examples of service that often are subject to pre-authorization. The pre-authorization procedure may require a cost estimate from the provider. Pre-authorization is also used to counteract proliferation of the practice of consulting non-contracted providers (for instance, in community care). In some cases patients have filed lawsuits to overturn insurers’ decisions to deny coverage. An example is the denial of pre-authorization of a rehabilitative treatment. The court concluded that insurers should refrain from interfering in the patient–doctor relationship and that the professional’s opinion should decide the choice of treatment. The court said that withholding authorization is justified only if there is clear evidence of a doctor violating the medical guidelines set out by his or her professional society (Mous & de Jongh, 2017). Health insurance reform in practice 47 3.2.5 Premium-setting To spur competition, health insurance legislation leaves insurers free to set their nominal premium rates, but it requires them to apply a community rating. Thus, insurers must charge the same premium to their clients who have a similar plan. As spelled out in Chapter 2, legislation categorically prohibits risk rating by relating the premium directly or indirectly to personal characteristics, including age, gender and pre-existing medical conditions. Fig. 3.2 presents an overview of the rise of the average yearly nominal premium and mandatory deductible since 2008. In the period 2009–2019 the nominal premium increased from €1094 to €1453 (32.8%),5 and the mandatory deductible increased from €150 in 2008 to €385 in 2019 (156.7%). Since 2015, however, for political reasons the government has refrained from further raises. Fig. 3.2 indicates that the premium drop in 2014 corresponded to a substantial rise of the mandatory deductible in 2013 (from €220 to €350). Since 2015 insurers have used their financial reserves to restrict premium growth (see subsection 3.2.6 on financial performance). Fig. 3.2 Average nominal premium and mandatory deductible, 2008–2019 (euros) 0 200 400 600 800 1000 1200 1400 1600 Nominal premium Mandatory deductible 20 08 20 09 20 10 20 11 20 12 20 13 20 14 20 15 20 16 20 17 20 18 20 19 Source: Vektis, 2019b. Premium rates vary not only among insurers, but also among types of health plans. For instance, in 2019 the premium of a benefit-in-kind plan averaged €1397, the premium of a cost-reimbursement plan averaged €1433, and the 5 The premium raise has been influenced by the transfer of a great deal of mental health care, geriatric rehabilitation and community nursing and care from long-term care insurance to the Health Insurance Act (see subsection 2.6.3). The market reform in Dutch health care48 premium of a plan with restrictions averaged €1291. The average premium of cost-reimbursement plans has risen the fastest. In the period 2016–2019, it rose by 19.1%. For benefit-in-kind plans and plans with restrictions, the increases in the same period were 15.4% and 15.7%, respectively (NZa, 2019a). The spread between the lowest priced and highest priced premium has increased from €320 in 2015 to €438 in 2019 (NZa, 2019a). In 2019 the lowest priced plan was 32% cheaper than the highest priced plan. Given the standard benefit package, the spread cannot be attributed to major differences in coverage. This raises the question: is the spread due to differences in efficiency (administration, purchasing strategy, investments and so on) or due to deficiencies in the system of ex ante risk equalization favouring insurers with an overrepresentation of good risks? It seems unlikely that the spread can be solely explained by differences in efficiency (see section 3.4). The Healthcare Authority has frequently concluded that insurers respect the ban on premium differentiation. Yet, it has observed some activities that, though lawful, raise questions. For instance, an insurer may charge different premiums for almost identical health plans under different names, which reeks of premium differentiation. Discount percentages in collective health plans (for instance employer-based plans) is another practice that may be considered premium differentiation. The Health Insurance Act allows discounts, but it restricts the maximum discount to 10% (as of 2020, to 5%). The effect of collective health plans is that consumers actually pay different premiums for an identical health plan. For this reason, the option of collective plans has been disputed. Discounting of collective plans will be abolished in 2022. 3.2.6 Financial performance Fig. 3.3 demonstrates that insurers realized a surplus per insured person 18 years and older in the period 2010–2015, but incurred a deficit in the period 2016–2017.6 The deficit has a political background. The financial buffers and surpluses of insurers had sparked a critical discussion of insurers’ “high profits”.7 Opponents of the reform characterized insurers as money-driven organizations that considered patients a cost item. In reaction, insurers promised to raise their premiums less than needed to cover the expected rise of health expenditures. This restraint led to the negative results in 2016 and 2017. Another way of measuring financial performance (and market competition) is to look at the claim ratio, defined as the balance between revenues and 6 Insurers incurred a deficit in the first years of the reform: −€422 million in 2006 and −€217 million in 2007 (Vektis, 2008). 7 The surplus insurers made is also the result of the positive results of investments (DNB, 2017). Health insurance reform in practice 49 provider claims. Revenues include premium revenues and revenues from the risk equalization fund. Revenues from investments and administrative costs are not included. The average claim ratio in basic health insurance is high, varying in the period 2012–2018 from 93.2% in 2013 to 100.3% in 2016. These percentages indicate that insurers spend almost all their revenues paying providers’ claims. The average claim ratio in supplementary health insurance is significantly lower, varying over the same period between 82.5% in 2012 and 88.4% in 2016 (Vektis, 2016; 2019b). An indicator to measure the financial health of insurers is the solvency rate, which reflects the balance between total assets and claim turnover. The Central Bank of the Netherlands sets standards for the solvency rate. The required solvency rate was set at 8% of total claims in 2006 but was raised to an average of 17% in 2017. Fig. 3.4 gives an overview of the insurers’ annual average solvency rate since 2012. As the figure shows, insurers scored well on solvency (a score of 100 means that they meet the required rate). The transition from the European standards of Solvency I to the standards of Solvency II8 in 2015 meant that insurers had to meet higher standards. A consequence of this transition was that the average excess solvency ratio dropped from 241 in 2014 to 160 in 2015. The total financial buffer of all insurers amounted to €8.9 billion at the end of 2017 (NZa, 2019a), equalling approximately one fifth of the total health care expenditure covered by the Health Insurance Act. This buffer indicates that, on average, insurers are able to absorb financial setbacks. 8 Solvency II obligates insurers to have enough capital available so that the probability they can meet their financial obligations in the next 12 months is at least 99.5%. The Central Bank of the Netherlands has translated this obligation into concrete regulations. The bank is also in charge of supervising compliance to these regulations. Fig. 3.3 Average annual financial result per insured (≥18 years) in basic health insurance, 2010–2018 (euros) -60 -40 -20 0 20 40 60 80 100 2010 2011 2012 2013 2014 2015 2016 2017 2018 Sources: NZa, 2017a; 2018b; 2019a. The market reform in Dutch health care50 Fig. 3.4 Solvency rate of health insurers 0 50 100 150 200 250 300 So lv en cy ra te 2012 2013 2014 2015 2016 2017 2018 Solvency I Solvency II Sources: DNB, 2017; NZa, 2018b. The high solvency rates are inextricably bound to the choice for a private model in health insurance. The Health Insurance Act is construed as a private law arrangement (see subsection 2.3.1). The consequence of this choice is that insurers must abide by higher solvency requirements than they would have if health insurance had been based on a public model such as the Sick Fund Act in the past. Nevertheless, the question may be raised whether the solvency requirements are too stringent. This point was put forward by the Authority for Consumers and Markets (ACM) in a critical 2017 report on the capital requirements imposed by Solvency II, and interpreted and applied in the Netherlands by the Central Bank of the Netherlands. The Authority considered these requirements for insurers unnecessarily stringent. In its view, an insurer bankruptcy would not destabilize the health insurance system. The risk of financial contagion in case of liquidity problems was estimated as low. The Authority also pointed out the risk that Solvency II might deter new entrants because of high capital requirements and regulatory uncertainty (ACM, 2017a). 3.2.7 Administrative costs The 2006 reform had a positive effect on administrative efficiency. In 2005 sick funds spent 4% of their turnover on administrative costs, and private insurers spent 10.9% (Vektis, 2006). Due to the reform, total administrative costs, including commission costs, dropped to 4% in 2007 (Vektis, 2008). In 2019 administrative costs in basic health insurance accounted for only 2.7% of insurers’ total expenses (Vektis, 2019a). In supplementary health insurance, this percentage was significantly higher, at 11.2%, mainly due to the handling of individual claims. Health insurance reform in practice 51 A politically sensitive topic is the insurers’ spending on marketing and intermediaries. Critics of competition frame every euro spent on these targets as waste of public money. In their document, Kern-gezond, insurers promised to reduce marketing costs. They have kept their promise: In 2018 these costs had dropped to €2.18 per adult enrollee, one quarter less than in 2014. In 2018 the costs of marketing and commissioning accounted for only 0.6% of total costs (NZa, 2019a). The political sensitivity of marketing costs indicates an ambiguous attitude towards competition in health insurance. On one hand, insurers are required to compete, but, on the other hand, they should restrict their marketing costs to a bare minimum. 3.3 Consumer response As described in Chapter 2, freedom of choice is an essential element of the market reform. Freedom of choice is intended to enable consumers to makes choices that best fit their individual preferences and spur competition in the health insurance market. Under the new regime, they can switch to another insurer or another type of basic health plan at the end of each year. They may take out a health plan with restrictions, a plan with a voluntary deductible or a supplementary health plan. This section presents an overview of how consumers have made use of their opportunities. 3.3.1 Switching After 2006, when a surprising 18% or so of the population switched insurers, the yearly percentages of switchers have been much lower.9 Fig. 3.5 shows an overview of the percentage of switchers annually since 2008.10 It illustrates that, as of 2012, the percentage of switchers has fluctuated around 6–7%, with a peak of 7.2% in 2013. The peak in 2013 is probably associated with the substantial rise in the mandatory deductible in that year. Consumers may have sought compensation for the premium rise by switching to an insurer with a more attractive premium. As of 2019, 50% of all consumers had never switched since the introduction of the new legislation, 27% had switched once, and 4% had switched four times or more (Vektis, 2019b). The stability of the percentage of switchers in the period 2016–2018 is somewhat surprising, given the increased spread between the lowest priced and highest priced health plans. 9 In the pre-reform era switching rates varied between 2% and 4% a year. 10 The percentage measures the percentage of consumers who switched to another insurer. If persons younger than 18 years of age are not included, the percentage of switchers in 2019 drops to 6.6%. Consumers may also switch to another insurer that belongs to the same insurance group. Excluding these switchers, the switching rate further drops to 6.2% (Vektis, 2019a). The market reform in Dutch health care52 Fig. 3.5 Percentage who switched insurance provider, 2008–2019 Pe rc en ta ge 0 1 2 3 4 5 6 7 8 20 08 20 09 20 10 20 11 20 12 20 13 20 14 20 15 20 16 20 17 20 18 20 19 Source: Vektis, 2019b. Persons from 18 to 39 years of age, persons with more education and persons perceiving their health as good switch relatively more frequently than older persons, persons with less education and persons perceiving their health as poor (de Jong et al., 2015).11 Thus, the former categories of persons have benefited more, relatively, from increased freedom of choice. Much empirical research has been done on motives for switching and for not switching. Although the relative weight of motives varies over the years, a lower premium, dissatisfaction with the coverage of supplementary health insurance, the expectation of needing medical care and the decision to take out a collective plan were the most important motives for switching. The most important motives for not switching were loyalty and satisfaction with the current insurer’s coverage and service (NZa, 2016a; Holst et al., 2019). However, research also has identified obstacles to switching. A potential switcher could abstain from switching “for fear of non-acceptance in basic health insurance”. This obstacle indicates an information problem, because insurers must accept each person. Fear of non-acceptance in supplementary health insurance was another reason to forgo switching, even though insurers promised to accept without conditions every applicant who was choosing a plan similar to their previous supplementary plan.12 In their 2017 study of consumer behaviour in health insurance, the Healthcare Authority and the Authority for Consumers and Markets (ACM, 2017b) 11 These categories of persons may overlap. 12 The Health Insurance Act permits insurers to include exemption clauses in their health plans with regard to the mandatory deductible. For instance, they may exempt someone who is insured and has a chronic illness from the co-payment for specific medicines or for consulting a preferred provider. In 2019, 19 insurers had exemption clauses in their health plans. The effect of these clauses on switching behaviour is unknown. Health insurance reform in practice 53 concluded that many consumers could not find their way in the health insurance marketplace, despite the availability of comparison sites on the Internet (Box 3.2). Consumers often found information confusing, and there were limits to the amount of information consumers could grasp. Collecting and processing information proved costly and difficult for them. Knowing that the devil is always in the detail, consumers had, therefore, abstained from switching (NZa, 2018a). 3.3.2 Other choices The new health insurance legislation offers consumers the option to choose a voluntary deductible on top of the mandatory deductible. The percentage of consumers who used this option rose from 5.9% in 2010 to 13.1% in 2019 (Table 3.1). These percentages suggest that the great majority of Dutch consumers are rather risk averse, although the trend points to more risk-taking behaviour. Almost 75% of consumers with a voluntary deductible have opted for the highest possible deductible of €500 (Vektis, 2019b). Research indicates that low-income people are less likely to select a health plan with a voluntary deductible than high-income people. This finding implies that low-income people benefit less from the option of a voluntary deductible than high-income people. There is also evidence that voluntary deductibles elicit adverse selection. Croes et al. (2019) found that persons with higher previous and future health care costs are substantially less likely to take up or keep a €500 deductible. However, they also reported that the propensity to take up a €500 deductible among low-cost individuals was only 3.5%. This low percentage, in their view, points to “consumer inertia” (Croes et al., 2019). We question this Box 3.2 Comparison sites for health insurance In Dutch health care there is no public marketplace to support consumers in choosing a health plan. There are several commercial websites, however. These sites promise to put the best choice on top, given the client’s specific preferences. Insurers pay the website company a tariff if a consumer purchases a health plan via its website. Table 3.1 Consumer response to choice options Choice option 2010 2012 2014 2016 2018 2019 % voluntary deductible 5.9 7.0 10.7 12 12.6 13.1 % collective health plan 64.3 67.6 70.5 67.5 66.6 65.3 % supplementary plan 89 88 84.5 84.3 83.6 83.7 % plan with restrictions 1.1 (2011) 1.7 4.4 8.6 13.4 14.1 Source: Vektis, 2019b. The market reform in Dutch health care54 terminology because of the implicit implication that any consumer not making the economically optimal choice would apparently be suffering from “inertia”. We consider this a narrow perspective on making choices. There are good reasons for not making other choices, such as high search costs, satisfaction and loyalty. Although collective health plans have always been popular in Dutch health care, the percentage of those insured with such a plan has slightly dropped since their peak in 2014. Insurers are permitted to give collective plans a discount of up to 10%. Consequently, those insured with a collective plan pay a lower premium (on average –2.7% in 2018) than those insured individually (NZa, 2018b). Employer-based collective plans are the most popular. Collective plans are criticized because they offer their clients little more than a lower premium that charges to people without such a plan must cover. In other words, they erode the ban on premium differentiation. Because of this problem and doubts about the claimed extra value of collective plans, the government has decided to put an end to the practice of discounts in 2022. Interestingly, a lot of municipalities have availed themselves of the option of a collective basic and supplementary health plan for low-income people in their communities in order to cover some extras for these people (for example, the mandatory deductible). Municipalities use their state grant for social support to subsidize the premium of these plans. Supplementary health plans cover the costs of services that are not listed in the benefits package of the Health Insurance Act. The great majority of people have such a plan. However, the percentage of people with a supplementary plan has steadily declined, from 89% in 2010 to 83.7% in 2019.13 The market for supplementary health insurance is diverse. There are plans with limited coverage (“bronze plans”) and plans with broad coverage (“golden plans”). Most people purchase a plan to cover the costs of paramedical care (for example, physiotherapy) and dental care (the basic health insurance scheme covers most forms of dental care for children but only a limited number of interventions for adults). 3.3.3 Rational consumers? Freedom of choice is not only a policy instrument to enable consumers to make their own choices in health care and to give insurers an incentive to develop health plans that are optimally attuned to consumer preferences. It also serves as an instrument for cost control. Managed competition assumes that rational consumers select the health plan that best meets their preferences. Aware of the switchers’ high sensitivity to the premium rate, insurers will be motivated to 13 Total revenues in supplementary health insurance dropped from €4.7 billion in 2012 to €4.5 billion in 2018. Health insurance reform in practice 55 keep their premiums as low as possible by strategic purchasing, economizing on administrative costs and so on. However, the assumption of the well-informed rational consumer is flawed. For instance, the Healthcare Authority and the Authority for Consumers and Markets found that in 2018 about 72% of consumers had chosen a basic plan for which a lower-priced alternative plan with identical conditions was available. The potential welfare gain of switching (switching gain) that consumers passed up in 2018 was estimated at somewhere between €905 million and €1.17 billion. These figures correspond to 5% and 6.5%, respectively, of total premium revenues in that year (NZa, 2018a). However, a study of the Netherlands Bureau for Economic Policy Analysis (Centraal Planbureau, CPB) found much lower switching gains. It estimated that the actual average annual switching gain between 2007 and 2015 was €40 million. Had all consumers switched to one of the lowest priced plans, the gain could have been ten times higher (Douven et al., 2017). We are sceptical about these research findings. If all consumers had acted as rational consumers and taken out the lowest priced alternative health plan with identical conditions, insurers would certainly have been forced to raise their premiums to cover their health expenses. There is no a priori reason to assume that more rational choices in health insurance would translate into lower expenditures (see section 3.4 on risk solidarity). The findings on consumer behaviour confirm the “paradox of choice” described by Schwartz (2004). The multitude of choice options and the complexity of health plans, with much small print, mean that consumers grapple with an information overload. It is a good thing that the Healthcare Authority requires insurers to be transparent, but greater transparency is not the magic bullet that solves the problem of choice. 3.3.4 The uninsured and defaulters A person who fails to purchase a basic health plan violates the law and is by implication uninsured. To discourage free-rider behaviour, the Health Insurance Act stipulates the sanctioning of uninsured persons. The percentage of those uninsured dropped from 0.17% in 2014 to 0.14% in 2018 (NZa, 2019a). This remarkably low percentage indicates that it is possible to combine a formal private health insurance scheme with a very high insurance coverage rate. This high rate cannot be attributed only to the mandatory character of the basic health insurance scheme and the regulations on sanctioning. It is also the result of collective efforts of insurers and municipalities to track and trace uninsured persons at an early stage. As discussed in subsection 3.3.2, many municipalities offer persons on social security the option of a collective plan for basic and The market reform in Dutch health care56 supplementary health insurance, which may also reimburse the mandatory deductible. What, furthermore, may play a role is the risk-averse nature of Dutch society and the broad public support for a solidarity-based model of health care financing (see section 3.6). A relatively more pressing problem is defaulting. Defaulters are insured persons who have failed to pay their nominal premium for a period of at least six consecutive months. Specific regulation is in place to track and sanction defaulters. The percentage of defaulters dropped from almost 2% in 2015 to 1.3% in 2018 (NZa, 2019a). Insurers and municipalities have undertaken initiatives to minimize the number of defaulters. 3.4 Risk selection and risk solidarity The Health Insurance Act contains various regulations to prevent risk selection and preserve risk solidarity. Insurers must accept every applicant (open enrolment) and apply community rating. Risk rating is forbidden. Other instruments to promote solidarity are the mandatory structure of basic health insurance, the standard benefit package and a sophisticated system of ex ante risk equalization to compensate insurers for differences in the risk profiles of their insured (Chapter 2). Is there evidence of risk selection in Dutch health insurance? An answer to this question requires a conceptual framework of risk selection. Van de Ven et al. (2017) define the concept as “actions (other than risk rating per health plan) by consumers and insurers with the goal and/or the effect that the cross-subsidies as intended by the regulator are not fully achieved” (p. 168). As indicated by this definition, risk selection may take different forms (van Kleef et al., 2013). It can be insurer-induced or consumer-induced or the result of a combination of both. An example of insurer-induced risk selection (cream skimming) is selective marketing. For their part, consumers engage in intended risk selection by choosing a health plan with preferred providers or a plan with a voluntary deductible in exchange for a lower premium. Risk selection may also be unintended. Table 3.2 demonstrates that, despite the sophisticated structure of the Dutch system of risk equalization, there is both undercompensation and overcompensation. Take, for instance, the category of people with the worst score on health. Had there been no risk equalization, insurers would have had an average loss of €670 on this category of patients. Risk compensation reduces the loss but only by 75%. In its market scans, the Dutch Healthcare Authority has repeatedly concluded that insurers comply with the formal prohibition of risk selection (NZa, 2014b; 2019b). Nevertheless, it has found some signs of indirect risk selection by insurers – for instance, by targeting their marketing to better-educated people or using a medical questionnaire in supplementary health insurance. The formal Health insurance reform in practice 57 prohibition on risk selection is not watertight. It applies to health insurers but not to intermediary agencies which insurers call in to sell their health plans for them. Another loophole in the legislation is that the ban on premium differentiation applies only at the health plan level. Insurers are obligated to charge their clients the same premium for the same plan, but they are free to charge differing premiums for health plans that are almost identical. Collective contracts also cause premium differentiation. Van de Ven et al. (2017) found some empirical evidence of risk selection, but they acknowledged that they were unable to measure its true level. This was because, for instance, the positive and negative effects of risk selection could have cancelled each other out or because risk selection had failed. Using the aforementioned definition of risk selection, the Healthcare Authority also investigated the presence of risk selection. Using data from 2010–2013, it found indications of risk selection in 7 of 74 health plans. These plans had a disproportional number of low-risk individuals due to a selective inflow and outflow of the insured. The Authority also detected 20 health plans with minor signs of selective inflow and outflow. In its view these results did not necessarily point to intended risk selection. The results could also be an unintended effect of insurer and consumer actions (NZa, 2016b). Another finding was that insurer groups had taken “a fair share” of high-risk and low-risk individuals at the group level. Table 3.2 Average under- and overcompensation per person in year t for selected groups based on information from year t–1, using the Dutch risk equalization formula of 2014 Selected groups based on information from year t–1 Undercompensation in year t Reduction of undercompensation compared with no equalization Worst score on physical health −€670 −75% Contact with medical specialist in the last 12 months −€326 −75% Use of physiotherapy in the last 12 months −€328 −71% At least one chronic condition −€331 −80% Selected groups based on information from year t–1 Overcompensation in year t Reduction of overcompensation compared with no equalization No chronic condition +€152 −66% Best score on physical health +€291 −71% No health care utilization in the last 12 months +€298 −75% Highest education levels +€142 −61% Source: van de Ven et al., 2017. The market reform in Dutch health care58 Voluntary deductibles and health plans with restrictive conditions, though legal, may also cause risk selection. The trouble with voluntary deductibles is twofold. First, assuming that they do not affect the demand for health care, the effect is that persons abstaining from a deductible pay a higher nominal premium. Thus, they cross-subsidize the lower premiums of persons who have chosen a voluntary deductible. Second, recent research has indicated that a voluntary deductible induces self-selection. Plans with restrictive conditions in exchange for a lower premium have an even stronger effect. Consumers who cannot afford such a plan – for example, because they face comparatively high costs due to chronic illness – cross-subsidize consumers with managed care plans (Croes et al., 2019). It is for this reason that these plans and voluntary deductibles have been criticized. A study by Equalis (2019) showed that current risk equalization is not perfect. The bars in Fig. 3.6 represent the net result per insured for a selected number of insurers. Insurers can realize a positive result by greater efficiency but also by an overrepresentation of “good risks”. It is difficult to assume that the big differences in net result are caused by greater efficiency only. The difference between newcomer iptiQ and ONVZ is €180. The researchers confirmed the finding of the Healthcare Authority that the result may significantly differ between the brands of the same insurer group, even though they make use of the same purchasing organization. This finding suggests that the net results are associated with the health characteristics of the insured in a health plan. Within one insurer group the researchers even found a difference of €120 per insured. Fig. 3.6 Net result per insured for selected insurers in 2019 (euros) Avero ONVZ Nationale Nederlanden OHRA IZA Zorg & Zekerheid Zilveren Kruis CZ IZZ Menzis De Friesland VGZ IPTIQ Amersfoortse Interpolis DSW FBTO Anderzorg 50 100 150 200 250 Source: Equalis, 2019. Health insurance reform in practice 59 We conclude that the provisions in health insurance legislation to preserve risk solidarity and avert risk selection are not watertight. Nevertheless, there are no hard indications of widespread risk selection. The remarkable differences in net result per insured indicate that the system of ex ante risk adjustment has not worked as well as hoped. The increased premium spread between the lowest priced and highest priced plans also points to this (see section 3.2). However, to put these conclusions in perspective, perfect risk equalization does not exist, and comparative research indicates that the Dutch system of risk equalization scores rather well in comparison with risk equalization in some other countries (van de Ven et al., 2013). 3.5 Income solidarity The reform of health insurance in 2006 fundamentally altered the structure of health care financing. The new legislation required those 18 years and older to pay an income-related contribution set by the Minister of Health as well as a flat nominal premium set by each health insurer separately (Chapter 2). Because the average nominal annual premium jumped from €384 in 2005 to over €1000 in 2006, additional measures were necessary to ensure income solidarity and make the reform politically acceptable. For this reason, the government introduced the Health Care Allowance Act in 2005 to compensate people in lower-income categories by means of an income-related care allowance (premium subsidy). Fig. 3.7 Number of recipients of a care allowance (in millions) 0 1 2 3 4 5 6 20102009200820072006 2011 2012 2013 2014 2015 2016 2017 2018 single person two-person family Source: MoH, 2019b: 147. Fig. 3.7 shows that the number of recipients of an allowance steadily increased in the period from 2006 to 2011. In 2014 the government announced that it would focus the allowance programme on the lowest income categories, with the result that fewer households could apply for an allowance. In 2018 about one quarter of all subscribers received an allowance. In the same year, the maximum The market reform in Dutch health care60 allowance for a single-person household was set at 70% of the standard premium and for a two-person family at 67% (the premium was estimated by the Minister of Health). The total costs of the allowance programme in 2018 were estimated at about €5.2 billion (MoH, 2019b). The care allowance system has always been controversial because of high administrative costs and the problem of “pumping money around”. A simple alternative would have been to raise the fraction of social contributions and reduce the fraction of nominal premiums in health care financing. As yet, this alternative has not been politically feasible. It even almost caused a political crisis in 2012 because of opposition of the Liberals in the government coalition. As described in Chapter 2, an important goal of the integration of all insurance arrangements into a single mandatory scheme was to make health care financing more solidary. To what extent has this goal been achieved? The answer to this question requires a concept of “fair share”. That political question is beyond the scope of this study. Nevertheless, a few observations can be made. First, every person has access to health insurance. Second, the system of allowances limits the share of income that people with a low income pay for health insurance. Third, the state pays for children under 18. Fourth, the total contribution per household to health care averaged €5500 in 2020 (including the contribution to long-term care), but persons with a low income pay a much lower amount and persons with a high income a much higher amount (MoH, 2020). 3.6 Declining public support for solidarity? The great majority of the population supports a solidarity-based health care system. Nevertheless, there are concerns that solidarity may come under strain in the future. Escalating health care expenditures and increasing individualization may gradually lessen the willingness of people to share the financial burden of health care (Jeurissen & Sanders, 2007). Another factor is the increased body of knowledge about the impact of lifestyle on health and illness. The Council for Public Health and Health Care (RVZ) recommended in a report the option of co-payments or differentiated health insurance premiums on the basis of health- related behaviour. Quoting a famous statement of Rawls, the Council concluded that “the veil of ignorance” had dropped and that there was now plenty of scientific evidence about the impact of lifestyle on health. Given this evidence, the Council argued in 2005 it would be fair to hold people more responsible for their own health. Solidarity could not be sustained without self-responsibility (RVZ, 2005). The Council’s recommendations proved politically controversial, not only on empirical grounds but also on moral grounds, and they were not adopted in the 2006 Health Insurance Act (prohibition of premium differentiation). Health insurance reform in practice 61 The risk of crumbling solidarity has always been an important argument of the government for the need for effective cost control. But has public support for solidarity declined? Research suggests that this is not the case. The solidarity monitor of the Netherlands Institute for Health Services Research (NIVEL) indicates that public support for solidarity was still widespread in 2017. Almost three quarters of survey respondents said that they were willing to pay for the health care of other people (Fig. 3.8). Fig. 3.8 Willingness to pay for the health care of other people and expected solidarity of other people (percentages) 0 10 20 30 40 50 60 70 80 2013 2015 2017 2019 willingness to pay for others expected solidarity Source: Holst et al., 2020. The NIVEL monitor on solidarity indicates that people with higher education are the most solidary and people with lower education the least solidary. Remarkably, the willingness to pay for others is lower among people with a health risk. These people also expect other people to be less willing to pay for their health care. In order words, they believe that public support for solidarity will decline in the future. 3.7 Discussion and policy lessons The introduction of the 2006 Health Insurance Act constituted a major overhaul of the Dutch health insurance landscape. After many years of political debate, the reform put an end to the traditional dividing line between the sick fund scheme covering about two thirds of the population and the labyrinth of private and public schemes covering the rest. The Health Insurance Act integrated all schemes into a single mandatory basic health insurance scheme. The new scheme was initially shaped as a scheme under public law. With this choice the government built upon the tradition of the sick fund scheme. However, The market reform in Dutch health care62 this line of thought was left out of the legislative proposal sent to the parliament in 2004. The reason for this change was political: to keep private insurers on board. Yet, the question may be raised whether the shift from public to private was not mainly a semantic issue. The basic health insurance scheme is formally a private scheme but with many strict conditions. Consumers are obligated to take out a basic health plan, and the legislation includes several regulations to protect risk and income solidarity. Also, insurers are obligated to accept every person (open enrolment). For this reason, one may actually consider it a quasi-public scheme. However, this is only one side of the story, because the choice for a private scheme also meant that insurers had to meet higher solvency requirements than would have been necessary under a public system with sick funds. The high solvency rate of insurers implies that they have built up substantial financial buffers. The new health insurance scheme has become an institutionalized part of Dutch health care, and there is no good reason to expect a return to the old bifurcated system. This does not mean that criticism is absent. The mandatory deductible is probably the most contested element of the reform of health insurance. Many people consider this instrument unfair (a “fine on disease”). The protest against the deductible does not come as a great surprise. Co-payment arrangements have always been disputed in Dutch health care, and such arrangements introduced in the 1980s and 1990s were short-lived. The resistance to the mandatory deductible has led the government to refrain from further increases since 2016. Left-wing political parties promise their voters to fight for its complete abolition. The current health insurance system functions reasonably well in practice. The reform did not create shocks in the health insurance market. Universal access to health insurance is guaranteed, not only in theory but also in practice. The percent- age of people without insurance has remained very low. The average switching rate of 6% keeps insurers on the alert with respect to their premium rates. Nevertheless, there are some problems. Despite its sophisticated structure, the risk equalization mechanism is imperfect. Although legislation bans risk selection, insurers may exploit loopholes to circumvent the ban. Targeting supplementary plans at specific groups and subtle forms of selective marketing are examples of these loopholes. An important policy lesson is that the basic and supplementary health insurance, although formally separate schemes, interact in practice. Insurers may use supplementary health insurance (a truly private scheme) as an instrument to influence consumers’ choices in basic health insurance. Consumers have the legal right to switch to another insurer or take out an alternative health plan at the end of each year. Research has shown that persons with more education and persons perceiving their health to be good have benefited relatively more from the choice options than older persons, persons with less education and persons perceiving their health to be poor. Another Health insurance reform in practice 63 problem regards the assumption of rational consumers making well-informed choices. The policy lesson from this study is that the assumption of the rational consumer is flawed. Many consumers feel that they are not well informed, despite successive attempts of the Healthcare Authority to press insurers for maximum transparency. The Authority’s regulation on transparency does not really solve the problem of information asymmetry and what Schwartz (2004) called the paradox of choice: information overload hinders a deliberate choice. The absence of a public marketplace offering consumers comparative information on health plans is a missing link in the current system. Private comparison websites may bias choices. Since 2012 the percentage of switchers has fluctuated between 6% and 7%. An unanswered question is how to interpret these percentages. Empirical research has shown that the premium is an important determinant of switching behaviour. Many switchers do so to benefit from a lower premium. This result implies that insurers are keen for costs to be competitive – precisely what the advocates of regulated competition had in mind. But it also creates a dilemma for insurers in purchasing. If the premium rate is the important determinant of switching, then why invest in quality of care – the more so because of the lack of transparency of health care quality? This issue will return in the next chapter about purchasing. Empirical research demonstrates that a select group of consumers (young, highly educated and in good health) has benefited the most from the reform. There is also evidence of unintended cross-subsidization. Consumers without a collective health plan cross-subsidize consumers with a collective plan. Equally, consumers who cannot afford to take out a managed care plan cross-subsidize consumers with a managed care plan. Against this background it is understandable that the government wants to put an end to collective health plans. There are voices calling for a ban on managed care plans as well because they undermine risk solidarity. A complex system is in place to preserve income solidarity. Persons in the lowest and middle-income categories are compensated through an income-related allowance. This is a cumbersome system that, in the view of its critics, boils down to “pumping money around”. The question is how it will develop in the future, all the more so in the context of serious questions about the practicability of allowance systems in general.14 Is it not the role of the tax office to tax people instead of distributing allowances? One thing is certain: There is no easy and widely accepted answer to the question of how to divide the financial burden of health insurance across income categories. 14 This is not a problem that is exclusive to health insurance. Welfare benefit regulations have become so complicated that many recipients simply do not understand them. The ombudsperson has repeatedly criticized the state for its wrong assumption that all people can understand all regulations. More transparency is fine but has limitations. There is a need for simple regulations and a readily accessible and competent help desk to inform people.

Chapter 4 Purchasing health care in practice 4.1 Introduction In their 2005 study of purchasing, Figueras and his colleagues described strategic purchasing as a promising instrument for improving health systems performance. They defined strategic purchasing as a process that “aims to increase health systems’ performance through effective allocation of financial resources to providers, which involves three sets of explicit decisions: which interventions should be purchased in response to population needs and wishes, taking into account national health priorities and evidence on cost-effectiveness; how they should be purchased, including contractual mechanisms and payment systems; and from whom, in light of relative levels of quality and efficiency of providers” (Figueras et al., 2005: 45). Strategic purchasing, also referred to as active purchasing, starts with a needs assessment by the purchaser. How will the demand for health care develop? What kind of care and how much care must be purchased and for whom (Øvretveit, 2009)? Which priorities should be set and which new developments should be encouraged? How to deal with new medical technologies? The next step is contracting with providers. Contracting plays an instrumental role in purchasing and can be described as the process of making concrete agreements on the provision and payment of health care that are usually formalized in written documents (contracts) between the purchaser and the provider. Nowadays, many countries experiment with strategic purchasing (hereafter purchasing) in health care. However, purchasing is anything but a uniform policy instrument. A recent study of purchasing in 10 European countries found important differences with regard to the assessment of population health needs in purchasing, the regulatory framework for purchasing, the involvement of citizens, the development of effective purchaser organizations and the inclusion of cost-effective contracting (Klasa et al., 2018). The market reform in Dutch health care66 Purchasing constitutes a crucial element of regulated competition and involves a transition from a hierarchical governance model towards a devolved model of governance (Greer et al., 2020). In the Dutch version of regulated competition, purchasing is delegated to competing insurers who should act as prudent purchasers of medical care on behalf of their clients. Insurers are assumed to be capable of making informed decisions on the price and quality of care. For this reason they fulfil an agency role in purchasing. In a competitive multiple payer system, the theory further states that consumers who are satisfied with the purchasing strategy of their insurer will have no reason to switch to another insurer, while customers believing that another insurer performs better may switch. Their exit option (Hirschman, 1970) will serve as an incentive to insurers to purchase strategically. Thus, purchasing is closely connected with competition in health insurance. Purchasing rests upon a number of policy assumptions: • Insurers are better informed than the state, consumers and patients about the performance of providers and, for this reason, better equipped to obtain value for money. • Insurers are more capable than patients of resisting the pressure of providers for inappropriate care and inflationary price increases, as the market compels them to limit premium increases. • The delegation of purchasing helps to depoliticize the allocation of scarce resources. The state is more subject to political pressure than insurers. • The delegation of purchasing stimulates variation and innovation, because insurers will develop initiatives, each in their own way, to perform this task. • The delegation of purchasing is a prerequisite for freedom of choice. For example, while some consumers choose a health plan with preselected providers, others prefer a plan guaranteeing them access to a broad range of providers. A former Minister of Health described purchasing as “the most powerful instrument insurers possess to foster efficiency, quality and cost control” (Schippers, 2017a). However, purchasing is no panacea. Insurers are expected to develop a purchasing strategy that will enhance the efficiency, quality and accessibility of health care. This poses complex challenges. For instance, how to shape an effective purchasing strategy in the context of uncertainty about the costs and quality of health care and low institutional trust (Maarse & Jeurissen, 2019)? Purchasing health care in practice 67 How to balance selective contracting and patients’ freedom of choice? How to reconcile the interest of the healthy premium payer with the interest of patients demanding the best care available, irrespective of its costs? This chapter offers an overview of how insurers have dealt with purchasing. It consists of three main parts. The first part discusses some general aspects of purchasing, including the regulatory framework for purchasing (section 4.2), models of purchasing (section 4.3), the evolution of purchasing (section 4.4) and the structure of contracts (section 4.5). The second part consists of a number of case studies: purchasing specialist medical care (hospital care) (section 4.6), purchasing mental health care (section 4.7), purchasing prescription medicines (section 4.8), purchasing general practitioner care (section 4.9) and purchasing other types of primary care (section 4.10). Together, these services accounted for 86.4% of the expenses covered by the Health Insurance Act in 2020 (ZiN, 2020). The power relationship between the purchaser and the provider is analysed in the third part (section 4.11). The chapter ends with a discussion of our findings and the formulation of some policy lessons (section 4.12). 4.2 Regulatory context The Health Insurance Act includes a regulatory framework for purchasing. This section offers a brief description of the important regulations. Duty of care (zorgplicht) The Health Insurance Act obligates insurers to guarantee their clients who have benefit-in-kind policies access to all types of necessary health care. The duty is formulated as an open norm that the Dutch Healthcare Authority has worked out in a set of detailed policy rules (beleidsregels) (NZa, 2017b). These rules prescribe, for example, that insurers purchase a sufficient volume of appropriate care that is accessible at reasonable geographical distance and without undue delay. Selective contracting The Health Insurance Act permits selective contracting, with the goals of spurring competition, enhancing quality, reinforcing insurers’ negotiating power and channelling patients to a limited number of providers. Without the threat of selective contracting, providers will feel little pressure to sign a contract that does not meet their demands. The market reform in Dutch health care68 Free pricing Health insurance legislation permits insurers to negotiate prices with providers for the great majority of health services and to take measures to restrict total spending (budget ceilings, volume ceilings and so on). Prohibition of collective contracting Competition law bans collective purchasing. The basic principle is that each insurer contracts with each provider separately and, conversely, each provider contracts with each insurer separately. Consequently, the contract an insurer signs may be different for each provider, and the contract a provider signs may be different for each insurer. However, collective purchasing is not completely prohibited. The Authority for Consumers and Markets permits contractors to engage in collective purchasing on the condition that there is convincing evidence of its positive effect on the quality and efficiency of health care. The test is that consumers must benefit from collective purchasing. Prohibition of vertical integration Health insurers are forbidden to run provider organizations of their own. Transparency In order to settle recurring conflicts on procedural aspects of purchasing, the Healthcare Authority issued a formal policy rule on transparency that obligates insurers to publish annually their purchasing plan and purchasing procedure. The policy rule was the successor to the Good Contracting Practices that the Authority had developed in concerted action with the stakeholders (NZa, 2014d). That code was replaced because it did not function well in practice. State responsibility A basic principle of the market reform is that the state abstains from direct involvement in the purchasing process. Purchasing is delegated to insurers. Legislation permits state intervention only if the government considers the accessibility or quality of health care in a specific situation to be at serious risk (for instance, in case of a looming bankruptcy). Policy documents describe the role of the state as bearing “system responsibility” for health care by providing a regulatory framework for purchasing and instituting effective oversight mechanisms. The Healthcare Authority, for its part, bears “case responsibility”. The 2006 Health Care Market Regulation Act also accords the Minister the formal competence Purchasing health care in practice 69 to give the Healthcare Authority binding instructions (for instance, on the total budget available for health care in a given year) and to recoup budget overruns. In addition to this regulatory framework, the Minister may start policy initiatives to influence the purchasing process. Examples are the programmes Appropriate Care (2013), The Right Care at the Right Place (2018) and Care Evaluation and Appropriate Use (2019). Each programme was launched in concerted action with the stakeholders in health care. Each of these programmes envisioned how the organization and provision of health care should develop in future. 4.3 Models of purchasing In the Dutch version of regulated competition, insurers purchase health services on behalf of their clients. They can make use of three basic models for this. The first model is the public tender model. Providers are invited to submit an offer for the provision of a set of specified services. The offer contains a price for the provision of these services plus a list of additional conditions. Next, the insurer in its role of purchaser decides on the basis of the price offered or the best price–quality combination according to a predefined score list. Insurers use this strategy for purchasing various services such as medical auxiliaries, generic outpatient prescription medicines (section 4.8), laboratory services and so on. Some insurers also experiment with a public tender in purchasing specialist care. An example is the tender for specialist cataract care or the treatment of drug addicts organized by the insurer Zilveren Kruis. The second model is the negotiated-contract model, whereby an insurer negotiates a contract for the provision of medical care for its clients with each provider. The negotiations start with the insurer informing providers about its purchasing strategy and requesting providers to submit offers for the provision of care. There is no public tender, and the contract with a provider may contain specific conditions (for example, waiting time). Contracts vary between providers, for instance, on price. This strategy is by far the most common in contracting with hospitals and provider organizations in mental health care. In the third model, the standard-contract model, the insurer determines a standard contract with conditions and a price for service provision. All providers who are qualified to provide these services and who accept the contractual conditions receive a contract. Insurers use this model in contracting general practitioners, physiotherapists and other single-practice providers. Many providers experience the standard contract as a dictate, because it offers virtually no room for negotiating. The market reform in Dutch health care70 4.4 Evolution of purchasing The purchasing of health care has proved to be an evolutionary process. Purchasing in the early stage differed in many respects from what it is nowadays. This does not come as a surprise. Initially, insurers had little experience with purchasing and had to learn their new role in the context of a lack of fundamental information. Valid and reliable information on costs and quality of care hardly existed. Furthermore, until 2008, only 10% of the hospital budget was open for price negotiation (see Table 2.2). Gradually, however, the context of purchasing changed. Currently, insurers possess more information on the cost and quality of care. Nevertheless, the quality of this information still has serious shortcomings for many types of care (for example, mental health care). As a consequence, quality information often fails as a steering instrument in purchasing. Other examples of a changing context are the stepwise extension of free pricing, the expansion of the benefits package of the Health Insurance Act, modifications in the funding model of hospitals and other providers, the increased exposure of insurers and provider organizations to financial risk, the introduction of new medical technologies, and changes in the landscape of health care (for example, mergers, new entrants). Insurers have also gradually come to understand that a simple price-cutting strategy is doomed to fail because providers may cancel out its effect on revenues by increasing the volume of care. There is a need for a purchasing strategy with a focus on efficiency and innovation and with powerful incentives to terminate the costly practices of both overtreatment and undertreatment. In this changed context, insurers have started rethinking their purchasing strategies. Recently (2021), the Dutch Healthcare Authority even requested them to do so. Nowadays, most insurers publish sizeable publicly available documents providing information on their purchasing strategy in each sector in health care. These so-called purchasing documents describe how they intend to follow up on and steer new developments in the provision of medical care and how they seek to connect with national innovation programmes such as The Right Care at the Right Place and Care Evaluation and Appropriate Use (see Box 4.1 for examples). The list of topics in Box 4.1 makes clear that the policy agenda in purchasing has become more comprehensive than it used to be. Purchasing should involve more than just a struggle about financial issues and contractual conditions. Insurers see their policy documents as a guiding instrument for health care innovation and lower costs. At the same time, they have come to learn that they should not overstate their capacity as guidance and that care innovation requires a strategy of cooperation with providers and patient organizations. An integrative style of purchasing based on cooperation and a win-win model may yield better results Purchasing health care in practice 71 than a competitive style of purchasing in which the gain of one is the loss of the other. The obvious question is to what extent the purchasing documents give direction to the practice of purchasing in the real world. After all, the proof of the pudding is in the eating. How do insurers’ intentions compare with their negotiating style in practice and their real steering capacity? To what extent do their intentions contrast with the daily reality of purchasing that has traditionally been dominated by financial concerns? More research is required to answer this question. A glimpse of an answer was recently given by the Netherlands Court of Audit (Algemene Rekenkamer, 2020). In its evaluation of the programme Appropriate Care, the goal of which was to save costs by avoiding unnecessary or ineffective patient care, the Court concluded that the results had fallen short of expectations. It identified several factors to explain this disappointing result. One of these factors was that insurers had lacked accurate steering information to guide purchasing; data on medical practice were old and incomplete (Algemene Rekenkamer, 2020). What has hardly got off the ground yet is integrated purchasing. Now that insurers are responsible for the purchasing of all medical care (and long-term care as well), they have, in principle, excellent opportunities for integrated purchasing – for instance, by contracting provider networks or by using bundled payments as instrument to facilitate the creation of comprehensive care pathways for specific Box 4.1 Concise list of topics in purchasing documents Each year insurers publish a set of purchasing documents to inform providers about their priorities, requirements and the organization of the purchasing process. The documents cover a wide range of topics. There are separate documents on hospital care, mental health care, general practitioner and multidisciplinary care, pharmaceutical care, maternity care, elderly care, care for specific patient groups, laboratory services and so on. A common characteristic of these documents is that each insurer declares that its priorities are the purchase of value-driven health care, the measurement of health outcomes and the elimination of groundless practice variation. Examples of other priorities are the concentration of specific complex medical interventions in expert centres, the establishment of regional provider networks, the promotion of appropriate care, the restructuring of hospitals in a changing health care landscape, the organization of health care for specific patient categories and specific treatments, the tackling of long waiting lists, the shift of health care, where possible, to primary care centres, the stimulation of e-health and shared decision-making. Strategy documents also inform providers about selective contracting, the use of multi-year contracts to facilitate the innovation of health care, the use of budget or volume ceilings and price lists, the requirements that providers must meet to qualify for a contract, the organization of the purchasing process and the non-reimbursement of treatments not backed by evidence. The market reform in Dutch health care72 patient categories. With some limited exceptions (for example, multidisciplinary primary care), they have hardly seized this opportunity yet. Only recently have some insurers taken initiatives to work on integrated purchasing and to experiment with shared-savings contracts. 4.5 Contracts Contracts between insurers and providers are complex documents consisting of many pages of formal agreements, rights and obligations on a great variety of issues. In her analysis of contracts with hospitals, Ruwaard (2018) distinguishes six important characteristics of contracts. They may be different in each contract: • Property rights allocation: Contracts include provisions in the event that the governance structure of a hospital changes – for instance, because of a merger or de-merger or a modification of the relationship between the hospital and medical specialists working in the hospital. Insurers may deny provider organizations a contract if they fail to meet the good health care governance code. • Contract duration: Contracts contain a provision on duration. In the early stage of contracting, contract duration was mostly one year. However, there is a trend towards multi-year contracts. • Incentive alignment: Contracts include provisions on reimbursement, volume and quality of care. They may also require referrals for non- emergency care and prior authorization of some specific treatments. As for quality, contracts require that the provision of health care accord with the state of the art in medicine or the professional standards of medical societies. Contracts may further include accreditation requirements and provisions specifying maximum waiting times. A new development is to include agreements on performance rewards and shared savings arrangements. • Contract completeness: Contracts contain provisions for adaptation to deal with unforeseen situations such as a substantial underestimation or overestimation of the expected volume of patients. • Risk allocation: Contracts include provisions for dealing with the financial risks of the contractors. Contracts may obligate providers to continue the treatment of patients after the lump sum has been spent or the volume ceiling has been reached (doorleverplicht). Purchasing health care in practice 73 • Protective measures: Contracts include provisions for monitoring hospital performance and settling disputes. 4.6 Purchasing specialist medical care In the remainder of this chapter, we will offer a descriptive analysis of the purchasing of specific health services. The first topic is purchasing specialist medical care (that is, largely hospital care). A broad variety of providers offer specialist health care, including university hospitals, general hospitals, specialized hospitals and, to a limited degree, independent treatment centres (ITCs) that focus upon specific, mostly routine treatments and specific categories of patients (see Chapter 5 for details). Most specialists work in hospitals. About 50% of them are self-employed; the rest of the specialists are salaried (NVZ, 2018). In 2020 specialist medical care accounted for 51.6% of total expenses covered by the Health Insurance Act (ZiN, 2020). This percentage explains why insurers consider the purchasing of this type of medical care big business. 4.6.1 Funding model The funding model of specialist medical care plays an important role in purchasing. Changes in this model may have great repercussions for the purchasing process. As discussed in Chapter 2, the market reform required a change in the hospital funding system. Before the reform hospitals received a function-based global budget annually. Since this funding model was deemed incompatible with competition, a group of experts from the Ministry of Health, hospitals, medical specialists and insurers developed a new activity-based model that was introduced in 2005 for about 10% of specialist care. The new funding model made use of DBCs and, as of 2012, DOTs to pay for specialist care. The theory was that each insurer would negotiate with each hospital on the price of DBCs or DOTs. The introduction of the new funding model involved high administrative costs. The transition to the DOT system was meant to reduce complexity, but it appeared to be a complex operation in itself. It was complemented by a temporary safety net employing “shadow budgets” to achieve a smooth transition from the old to the new situation and to avert budget shocks. Another significant change took place in 2012, when the government extended the scope of free pricing from 34% to 70% of the hospital budget (for the rest the Healthcare Authority prescribed the prices). Furthermore, to incentivize insurers to invest in purchasing, it chose an accelerated phasing out of ex post compensation. By lifting these arrangements, the government hoped that insurers would become keener on costs (see section 2.6). A final revision of the funding model with major implications for purchasing took place in 2015. Until then, The market reform in Dutch health care74 insurers had paid self-employed specialists directly. This arrangement ended in 2015. Ever since, insurers and hospitals have negotiated a total integrated budget for medical care. Each hospital bears responsibility for the remuneration of its physicians. 4.6.2 Purchasing and cost control From the very beginning, financial issues have played a prominent role in purchasing. Insurers have always seen keeping health expenditures in check as one of their priorities because higher expenditures would require premium increases and undermine their competitiveness in the health insurance market. They also felt committed to the framework agreements on cost control. For their part, hospital administrators and medical doctors have often expressed their frustration with what they called the insurers’ myopic interest in financial issues. In their view insurers paid lip service to quality issues or only showed interest in quality projects with cost-saving potential (Zuiderent-Jerak et al., 2011). In practice, purchasing often boiled down to a negotiating game on the level of cost increases in the upcoming year. Insurers negotiate different types of contracts, for instance, fee schedules, fee schedules with a volume ceiling, all-inclusive, or complete, contracts, lump- sum contracts, input contracts to guarantee the availability of essential services, performance-related contracts and so on. In its analysis of insurer–hospital contracts (year 2008) the Dutch Healthcare Authority found that insurers make frequent use of volume ceilings and budget ceilings to control health care expenditures. Volume-free contracts are only used at small scale. This practice points to the return of global budgets, albeit with a new look (NZa, 2019g). A new development is the use of multi-year contracts to offer hospitals budget certainty for a longer period of time, the elimination of the production incentive and the reduction of administrative costs. Multi-year contracts are also used as transformation instrument: a hospital is guaranteed a multi-year budget to facilitate its repositioning in the changing hospital landscape. In 2019 the insurer CZ and the Zuyderland hospital even concluded a 10-year contract, including the provision that the hospital will cut back its expenses in successive steps by reducing its volume of care by 5% over 10 years. Multi-year contracts may become the standard in purchasing specialist care. A recurrent issue in purchasing is how to deal with an underestimation of demand in contracts. Insurers handle this problem in different ways. While some of them include a provision for extra budget in the contract, other insurers obligate hospitals to continue the treatment of patients and bear the financial burden of “overproduction”. As can be expected, hospitals feel unhappy with Purchasing health care in practice 75 this contractual provision. Some of them contacted the news media to pressure an insurer for extra budget, arguing that the insurer had violated its duty of care and restricted patients’ freedom of choice in an unacceptable way. 4.6.3 Purchasing and quality of care In the model of regulated competition, insurers are assumed to sign agreements with providers on the quality of care. The optimistic idea is that the quality of care can be measured by means of outcome indicators, and insurers can use this information in purchasing, for instance, to reward hospitals for high quality and sanction hospitals for underperforming. In the early stage of purchasing, quality of care did not play a prominent role because information on quality was, for the most part, missing or patchy. Moreover, there were endless discussions about the validity and reliability of quality data. Another hurdle in steering quality was duty of care. Even with serious doubts about the quality of care, insurers could not refrain from contracting with a hospital. What made the problem even more complicated was that their clients would not accept the denial of the contract. In 2010 a remarkable event took place: One of the four big insurer groups (CZ) announced that they would not contract with six (later four) hospitals for breast cancer surgery because they did not meet the insurer’s quality standards (a volume threshold). The announcement attracted much public attention, and the medical community responded furiously, arguing that it was not the task of an insurer to set its own quality standards. In their view setting quality standards had to remain the responsibility of the medical profession. Nowadays, contracts typically require hospitals to comply with the quality standards of the medical societies (for example, volume standards), but insurers may set higher standards. Insurers have realized, however, that they cannot be in the driving seat of quality improvement. What they can do is encourage the medical profession to innovate and to make financial resources available to put these new ideas into practice (Box 4.2). As to whether attempting to manage quality of care gives an insurer a competitive advantage, the opinions of insurance executives and staff members differ. The existence of a difference was suggested in a study with four large insurers and one small insurer of the potential for managing quality of care. The researchers conducted semi-structured interviews with the CEOs of these insurers and also organized two focus group sessions, one with staff members directly engaged in purchasing (purchasers) and one with marketing staff (marketers). They found diverging views on incentives to manage quality. Purchasers had little faith that quality management would improve an insurer’s competitiveness, The market reform in Dutch health care76 whereas marketers and CEOs perceived various competitive advantages (Stolper et al., 2019). 4.6.4 Selective contracting The Health Insurance Act permits selective contracting in purchasing. The purpose of this instrument is to reinforce the negotiating power of insurers. Selective contracting takes several forms. An insurer may choose not to contract a provider or not to contract for all health products of a provider. There are different motives for selective contracting. Insurers may use the instrument to concentrate specific Box 4.2 Experiments with new purchasing models Menzis (another of the “big four” insurer groups) has signed innovative contracts with several hospitals and ITCs on the treatment of knee arthrosis and hip arthrosis. The contract comprises all specialist care from 120 days before surgery to one year after surgery. The insurer and providers also make agreements about the price per episode and quality of care. The purpose of this new type of contracting is to stimulate value-based care and avoid unnecessary care. Similar contracts are underway for maternity care, breast cancer and other types of care (NZa, 2018c). Zilveren Kruis is experimenting with models of best value procurement for purchasing cataract surgery. In this model providers are asked to specify the terms of the contract based upon their own expertise. The insurer refrains from setting minimum standards or process requirements. The insurer’s choice of providers rests upon objective and quantified evidence of past performance and not on the lowest price (Dohmen & van Raaij, 2019). Insurers CZ and VGZ signed a five-year contract with two hospitals to transform them into innovative centres for appropriate and patient-centred care (van Leersum et al., 2019). The contract guaranteed both hospitals a stable budget, regardless of patient volume and burden of claims. The contract included flanking investments to make the transformation plans named Dream Scenario (Bernhoven hospital) and Quality as Medicine (Rivas Care Group) come true, as well as an arrangement for shared savings. A recent evaluation found that both hospitals had successfully implemented a great deal of their transformation. Efficiency has improved, and expenditure growth was less than in comparison hospitals. Quality of care was high, but there was no evidence that it had improved. The financial results of both hospitals lagged behind the results of the comparison hospitals (van Dulmen et al., 2020; Douven et al., 2020). VGZ developed the programme Appropriate Care to redesign the organization of health care. The objective of the programme is to provide better care for lower costs. For this purpose, the insurer set up a network of so-called alliance hospitals, mental health care providers and home-care organizations to work on continuous care innovation. Two concrete examples are permitting general practitioners more time for patient consultations and promoting the potential of telemedicine in cardiac care. An explicit purpose of the programme is to eliminate the production incentive. Purchasing health care in practice 77 complex interventions in centres of expertise or to refrain from contracting provider organizations that fail to meet the standards of good governance and quality of care. Another option is to use selective contracting to create a preferred provider network for planned care. Finally, the instrument may be used to force providers to terminate certain practices (for example, overproduction). Stadhouders (2019) has demonstrated that selective contracting plays out differently for hospitals and ITCs. The average contract index, which measures the percentage of health services provided by hospitals and ITCs contracted by nine insurers in 2015–2016, was 0.88 for hospitals and 0.53 for ITCs (p. 105).1 The probable explanation for this result is that insurers tend to give priority to contracting hospitals first because of large-scale displacement of specialist care; ITCs might put the financial position of hospitals under strain, which insurers cannot afford. The effectiveness of selective contracting varies. It is a valuable instrument to concentrate the provision of specific medical interventions in centres of expertise and to stop the uncontrolled proliferation of new providers. For other purposes, however, its effectiveness should not be overstated (Halbersma et al., 2012). Selective contracting is effective only under the condition of realistic alternatives. Another precondition for effectiveness is that insurers must be able to channel their customers to contracted providers. However, the “channelling” capacity of insurers is questionable (Bes et al., 2017). Selective contracting also poses a dilemma for insurers, because it may restrict the patients’ freedom of choice. Insurers’ fear of damaging their reputation with clients has also proven a barrier to selective contracting, although more so in the early stage of contracting than currently. Boonen and Schut (2011) described this dilemma as the credible commitment problem. Finally, the use of collective contracting is influenced by the power balance between insurer and hospital. A big hospital has more market power than a small hospital in a geographical market with multiple options. 4.6.5 Collective contracting In the model of regulated competition, each insurer contracts with each hospital separately and, conversely, each hospital contracts with each insurer separately.2 1 The index measures the average percentage of health services contracted or not contracted by health insurers. The index runs from 0 (not contracted for any type of service) to 1 (contracted for all types of services). The index is weighted for the market share of each insurer, because it makes a lot of difference whether a service is contracted by an insurer with a large or small market share. Selective contracting does not necessarily mean that a hospital can no longer deliver non-contracted services. If insurer A decides not to contract for a specific service, insurer B may decide otherwise. Furthermore, insurers must pay the hospital tariff on non-contracted services for clients with an indemnity health plan. 2 Small insurers are permitted to coordinate purchasing. Multizorg VRZ is an organization purchasing medical care on behalf of its members. The market reform in Dutch health care78 Consequently, one finds much variation in the terms of contracts (Ruwaard, 2018) and prices. Separate contracts lead to higher administrative costs and may also hinder the coordination of medical care. To address the second problem in particular, the Authority for Consumers and Markets has issued guidelines on the acceptability of coordination. Although coordination is at odds with competition, the Authority permits coordination on the condition that contractors convincingly demonstrate that coordination enhances efficiency and quality of care. However, it maintains its strict ban on price coordination (ACM, 2014; 2019). The dividing line between competition and collective contracting has always been a source of uncertainty as to what is permitted and what is forbidden. An example is the Authority’s critique of the insurers’ plan to concentrate complex emergency care in a restricted number of hospitals. It found the plan not well founded and warned of spillover effects (cascade effects) of concentration to other areas of specialist care. The Authority reasoned that concentration could impair the overall quality of specialist care, push up prices and restrict freedom of choice. Concentration was justified only on the condition that evidence-based quality standards are met (ACM, 2014). However, in purchasing, things are not black and white. In practice, the regional market leader in health insurance usually takes the lead in contracting, after which the hospital demands that other insurers join the contract (excepting for prices). Another option is that two regional market leaders negotiate a contract on behalf of all insurers. This practice means that the representation model that was common in pre-reform times has never really disappeared. Collective purchasing is also used in purchasing high-cost medicines and contracting with expert centres such as the new Princess Máxima Center for Pediatric Oncology. Furthermore, the Minister of Health recently charged the two largest regional insurers with the purchase of ambulance services on behalf of all insurers. Another new development is the initiative launched in 2018, in which two regional market leaders will take the lead in contracting for acute health care services, including general practitioner posts, ambulance care, acute mental health care, acute pharmaceutical care, acute medical specialist care including trauma care, acute obstetric care and acute long-term care. The purpose of this initiative is to foster uniformity in purchasing (ZN, 2019). 4.7 Purchasing mental health care Mental health care is a sizeable sector in Dutch health care; in 2019 it accounted for 8.8%3 of total health care expenses covered by the Health Insurance Act 3 This percentage is exclusive of the costs of long-term mental health care, which is covered by the statutory scheme for long-term care. Purchasing health care in practice 79 (ZiN, 2020). It is delivered by 570 provider organizations and some 4000 self- employed providers. The market is quite concentrated: 36 organizations take about 70% of the total budget for mental health care (Algemene Rekenkamer, 2017). Until 2008 all mental health care was covered by the Exceptional Medical Expenses Act. This changed in 2008 with the transfer of mental health care of a duration of up to one year to the benefit package of the Health Insurance Act (section 2.6). As of 2015 the Health Insurance Act covers all mental health care up to three years’ duration.4 Mental health care has been subjected to many other revisions since 2008. Coverage has been restricted several times (for example, the deletion of psychoanalysis in 2010 and of treatment of adjustment disorders in 2012), and co-payments were introduced (2012) and then partially withdrawn (2013). In 2014 mental health care was split into basic and specialist mental health care. In contrast with specialist care, basic care serves patients with mild disorders. Treatments are limited to a maximum of 12 (Westra et al., 2016). The funding of mental health care also underwent major revision. Until 2008 mental health care organizations received an annual budget. The transfer of mental health care to the Health Insurance Act went along with the phasing in of a new funding model based on DBCs. Since 2013 all providers are in theory funded by means of DBCs, for which the Healthcare Authority sets maximum prices. In practice, however, the old budgeting model has remained more or less in place. On average 95% of the budgets of the mental health care provider organizations are based upon lump-sum agreements. The reason for this gap between theory and practice is the lack of valid information to inform management. Only self-employed providers and new entrants are funded by means of DBCs (Algemene Rekenkamer, 2017). Purchasing mental health care is complex. While the utilization of specialist mental care has dropped, utilization of basic mental health care has significantly increased. Also, providers seized the opportunity to extend the supply of basic mental health care. Increased demand for this type of care has contributed to the pressing problem of providers not accepting new patients and long waiting lists for treatment of certain disorders. Personnel shortages have aggravated this problem. Another cause of patient stops and long waiting lists is insurers’ tight purchasing strategy of paying lump sums; there are numerous examples of providers requesting additional funding (NZa, 2019f ). An indication of the insurers’ tight strategy is that they did not contract to the full extent of the global budget for mental health care (see Chapter 6). 4 Since 2015 municipalities have been charged with the purchasing of youth care and sheltered living. As of the same year, long-term mental health care has been covered by the Long-term Care Act that came into force in 2015. The market reform in Dutch health care80 Just as in hospital care, the lack of information on quality of care hinders the purchasing of mental health care (Frank & McGuire, 2000). Despite efforts to measure the quality of mental care by means of routine outcome measures, insurers still have little insight into the effectiveness of mental health care. Another big challenge is how to organize value-driven mental health care. The  funding model is also a source of problems. One may question the appropriateness of DBCs for funding mental health care. There is also evidence of a production incentive in the funding model of self-employed providers (see Chapter 6). Insurers have used the split between basic and specialized mental health care to save costs by allocating a substantial amount of money to basic mental health care. This purchasing strategy has been to the detriment of specialist care. A recent study found a serious mismatch between the demand for mental health care and the purchasing strategy of most insurers (van Os, 2020). Recently, several insurers announced a revision of their purchasing strategy: A greater part of the budget will be reserved for the treatment of patients with complex mental problems. Some examples of other revisions are contracts without a budget ceiling for crisis-related mental health care, multi-year contracts and experiments with public tenders. Non-contracted care is on the rise in mental health care. The share of the total budget that insurers spent on non-contracted care rose from 4.4% in 2014 to 6.3% in 2016 (NZa, 2019f ). Insurers may not contract with all providers (selective contracting), and providers may decide not to sign a contract with every insurer. Frequently mentioned rationales of both provider organizations and self-employed providers to opt for non-contracted status are discontent about tariffs, lump sums, high administrative costs and other contractual provisions. Important reasons that insurers prefer selective contracting are lack of quality, mismatch of demand and supply and sufficient care in the region contracted (NZa, 2019f ). 4.8 Purchasing prescription medicines In 2019 prescription medicines accounted for 10.4% of total expenditures covered by the Health Insurance Act (ZiN, 2020). Purchasing outpatient prescription medicines differs from purchasing medicines prescribed in inpatient hospital care. The costs of the latter are part of the contract between insurer and hospital. Special arrangements exist for new high-cost medicines. Purchasing health care in practice 81 4.8.1 Outpatient prescription medicines Insurers contract with pharmacies for service delivery and the costs of prescription medicines. The Healthcare Authority has defined a list of 13 services that insurers may purchase from pharmacies. Examples are delivery of a first-time prescription with a check on the appropriateness of the prescription and potential interference with medicines already used by the patient, advice on how to take the medicine and providing information about potential side-effects, a periodic evaluation of the medicines used by patients with chronic disease and pharmaceutical counselling (Kroneman et al., 2016). The Authority sets maximum tariffs for these services. Insurers and pharmacies may negotiate a block contract or, to reduce transaction costs, an all-inclusive tariff for service delivery. Contracts with pharmacies also include agreements on the price of medicines. As of 2004 (thus, already before the 2006 reform), insurers were permitted to reimburse only the costs of the lowest priced generic medicine or a set of the lowest priced generic medicines with the same relevant chemical substance. Patients using a different formulation must either co-pay the price difference or pay the full price of the other formulation. This so-called preference policy (preferentiebeleid) has been highly successful in terms of cost control (see Chapter 6). However, doctors and pharmacists are critical because they perceive the policy as an infringement on their professional expertise. Also, the need to switch to another lower-priced medicine may confuse some patients. Recently, shortages of certain prescription medicines have been linked to the preference policy. 4.8.2 Innovative high-cost medicines Over the last two decades, the pharmaceutical industry has introduced a multitude of innovative high-cost medicines, and many others are in the pipeline. Examples are TNF-alpha inhibitors, immunotherapy for oncological conditions and new medicines for metabolic diseases and blood clotting. The fraction of these medicines in total hospital expenditures rose from 6.8% in 2012 to 8.2% in 2016 (NZa, 2019c). Innovative medicines are heralded as breakthroughs, but their costs threaten the accessibility and affordability of health care in the future. Former Minister Schippers (2010–2017) spoke in this context about the growing imbalance between patients’ interests and the interests of shareholders of international pharmaceutical companies (MoH, 2016). In her view every patient has the right to innovative medicines at reasonable costs. In principle, hospitals purchase high-cost medicines and insurers reimburse the costs of these medicines. For some of these medicines, the contract between insurer and hospital may include a spending ceiling. Full-cost reimbursement rests upon a voluntary agreement between hospitals and insurers. Until 2015 The market reform in Dutch health care82 insurers were given ex post compensation for the financial risk of these medicines. The ex post compensation of the costs of oncolytics ended in 2016. Consequently, insurers are now financially at risk for most high-cost medicines. Insurers follow several strategies to control their expenses for coverage of medicines. The most common strategy is negotiating a discount, in particular for high- cost medicines and medicines with full competition (NZa, 2019c). However, insurers face an information problem if hospitals declare the net price of these medicines confidential or are contractually forced to do so by the manufacturer. To respect confidentiality, hospitals and insurers may use a trusted third party as an intermediary vehicle. Insurers and hospitals may also make agreements on the appropriate utilization of high-cost medicines (for instance, which patients qualify for these medicines, start and stop criteria, appropriate dosage and so on). Contracts usually require hospitals to report to insurers on a regular basis on the utilization of high-cost medicines. A new strategy is collective contracting. The Authority for Consumers and Markets has permitted insurers, under specific conditions, to engage in collective contracting for high-cost medicines. A new development is the direct involvement of the Ministry of Health in negotiating confidential financial arrangements with the pharmaceutical industry.5 The expected savings of this strategy were reported to be €132 million in 2017 (MoH, 2018a). The direct involvement of the government in purchasing is understandable, but it is also remarkable from the perspective of regulated competition, because the model of regulated competition does not provide for a direct role of the state in purchasing. 4.9 Purchasing general practitioner care General practitioners play a central role in Dutch primary health care. In principle, each person needs to register with a general practitioner. Patients need a referral from a general practitioner to consult a medical specialist. The market reform has reinforced the general practitioners’ role because now all patients need referrals, whereas in the pre-reform era only sick fund patients needed them. The fraction of general practitioner care in health care expenditures covered by the Health Insurance Act amounted to 8.2% in 2019 (ZiN, 2020). Understanding the purchasing of general practitioner care requires a brief description of the funding model. We have seen in Chapter 2 that the reform made a new uniform model necessary to establish a common and level playing field. The national associations of general practitioners and health insurers 5 The outbreak of the COVID-19 pandemic in 2020 has intensified the direct involvement of the Ministry of Health in purchasing, not only with regard to vaccines but also with regard to necessary equipment and protective gear. Purchasing health care in practice 83 negotiated a model that not only regulated tariffs, but also included agreements on the quality and transparency of general practitioner care. After heated debate, general practitioners accepted the so-called Vogelaar agreement (Vogelaar akkoord) as a “minimum compromise”. The new model consists of three segments (NZa, 2019d). The first segment covers basic family health care. Insurers pay general practitioners a capitation fee per registered person and a fee for each consultation. The segment also includes a tariff for support personnel. The Healthcare Authority sets maximum tariffs for each item. The second segment regulates the payment of multidisciplinary primary care for patients with diabetes mellitus type 2 (DM2), chronic obstructive pulmonary disease (COPD), vascular risk management (VRM) and asthma. The tariffs for these services are freely negotiable. Insurers also pay for the organization and infrastructure of other multidisciplinary health services. The third segment includes a great variety of performance-related payments and payments for modernization and innovation. Tariffs are freely negotiable. Insurers may pay additionally for extra services and facilities not covered in these segments. Examples are care for vulnerable people, the promotion of smaller practices to create more time for patient visits, substitution projects and initiatives to ensure the reach of general practitioner care in some regions. In the course of time, the funding model has undergone numerous adjustments (ZiN, 2018). Insurers use a mix of standard and negotiated contracts for general practitioners’ health services and products. The latter type of contract is negotiated with local or regional groups of general practitioners. In most cases they adopt the maximum tariffs. Selective contracting does not exist. Insurers usually voluntarily join the contract of the regional market leader (silent coordination). However, in some regions insurers with more or less equal market share hold to their own strategies, with the result that general practitioners and networks must sign multiple contracts with differing tariffs, conditions and durations. For instance, while some contracts for multidisciplinary cardiovascular management include a separate tariff for smoking cessation programmes, other contracts do not include such a tariff. While some contracts specify an inclusion ceiling (maximum number of patients), other contracts do not. Also, some insurers experiment with multi-year lump-sum-like contracts to reduce transaction costs (NZa, 2019d). General practitioners also conclude framework agreements on expenditure growth and other issues. For instance, in 2018 they signed an agreement to ensure the accessibility and effectiveness of general practitioner care in the context of demographic and social changes. This requires a restructuring of the regional organization and infrastructure of primary care, among other ways by investments in information and communications technology (ICT) and e-prevention, the provision of “the right care at the right place”, the decrease of workload and the The market reform in Dutch health care84 creation of more time for patients. General practitioners complain that insurers fail to live up to their promises in the framework agreement, with the result that the global budget for general practitioner care has been underspent. 4.10 Contracting other primary care providers Insurers sign standard contracts with other providers of primary care, including, among others, physiotherapists, psychotherapists, dieticians and dentists. These caregivers complain that they have no option but to sign the contract. The Royal Dutch Society for Physiotherapy reported that only three of every five physiotherapy practices contracted with all insurers in 2019, compared with four of every five in 2018. The percentage of practices opting not to sign contracts with all insurers rose from 14% in 2018 to 34% in 2019. Low tariffs and discontent with the many regulations they must follow were the main reasons for not signing (KNGF, 2019). Non-contracting also occurs in community nursing.6,7 The percentage of clients receiving community nursing from non-contracted providers increased from 3.6% in 2016 to 5.6% in 2018, and the fraction of the costs of non-contracted care in total costs rose from 4.3% to 9% over the same period (van Gerwen et al., 2019). Insurers deny small providers a contract if they fail to meet governance standards or the insurers’ conditions of efficiency, needs assessments and palliative care. Other reasons of insurers for non-contracting are lack of integrity and suspicion of fraud. For their part, providers may refrain from signing a contract because of discontent with tariffs, the frequent need for mid-term contract adjustments and the administrative burden of registration. Experiments have been started to lower the administrative burden by means of hour-tariffs, day-tariffs, month- budgets, preferred provider agreements and other arrangements (NZa, 2019e). Non-contracting is considered a serious problem – in the words of van de Ven (2018), even “a time bomb” that may undermine the effective control of health care costs. In this respect, it is striking that the costs of non-contracted care per client and per month in 2016 were about twice as high as the costs of contracted care (Puijk et al., 2017). Insurers are reconsidering their administrative procedures to resolve this problem, but provider organizations have filed a lawsuit to block the new procedures. The government has announced measures to tackle the problem of non-contracted providers and fraud by tightening the criteria for market entrance and governance. 6 Insurers have been charged with the purchase of community nursing since 2015 (see Chapter 7). 7 In 2018 the Minister of Health signed a framework agreement on community nursing for the period from 2019 to 2022 with the national associations of providers, insurers, municipalities and patients. The annual net growth rate of expenditures was set at 2.4%. Purchasing health care in practice 85 4.11 Purchasing and power General practitioners and other providers in private practice (physiotherapists, psychologists, pharmacists and so on) have frequently complained about the dominant position of insurers. They argue that they have no option other than ticking the box. In 2015 general practitioners expressed their frustration in the manifesto “Need for redirection” (Het roer moet om). In this document, which drew much attention and was signed by some 8000 general practitioners, they fulminated against the new “product thinking” in health care, in which general practitioners were treated as “market vendors”; against increased interference of insurers in medical practice (for instance, the insurers’ preference policy in pharmaceutical care); and against the steep growth of the administrative burden due to detailed procedures and accounting requirements. Another complaint was that the Authority for Consumers and Markets had silenced the professional organizations through its refusal to permit collective negotiation of contracts. What the writers of the manifesto in fact urged was a shift away from the contract model in health care and a return to the old professional model based on trust and expertise. The practice of “fake negotiations” had to be terminated, they contended. The manifesto resulted in an agreement with insurers to improve their mutual relationship.8 Although there is certainly an element of truth in this critique, the alleged weakness of the position of general practitioners should not be overstated. Insurers have a duty of care, and in practice they contract with all providers. Usually, they accept the maximum tariffs for capitation and consultation. While it is true that general practitioners often object to the tariffs, it is also a matter of fact that the new fee schedule has not done them financial harm (see Chapter 6). Finally, it is important to stress the traditionally strong position of general practitioners in Dutch health care. Patients need a referral for consulting a medical specialist, and any attempt to restructure the local or regional health care system is doomed to fail if general practitioners withhold their cooperation. The negotiation power of other providers with private practices, such as physiotherapists, psycho- therapists or dieticians, is clearly weaker than the power of general practitioners. They, too, are forbidden to engage in collective negotiations on tariffs. However, unlike general practitioners, they do not act as a “spider in the web” in local or regional networks. What has been said about general practitioners applies even more to hospitals. While the reform has reinforced the market power of insurers, their relative power should not be overestimated (Greer et al., 2020). The threat of selective 8 In 2019 a new version of the manifesto, with a broader scope than the first, came out. The basic complaint was that the provision of health care was completely stuck due to fragmentation, lack of capacity and competition. The document repeated the demand for collective negotiation of contracts. The market reform in Dutch health care86 contracting is rather blunt. Many hospitals play a central role in the regional organization of health care and, therefore, cannot be dismissed. Furthermore, hospital mergers have reinforced the negotiating power of tertiary care hospitals, and university hospitals are known as power centres. In short, an insurer’s market share is a poor indicator of its market power (Loozen, 2015; Loozen et al., 2016). The situation is different for ITCs. While some centres have established positions in specialist care, many other centres are frustrated about their subordinated role in purchasing. They feel that they are “the last in the row”. 4.12 Discussion and policy lessons This chapter described purchasing in Dutch health care. Purchasing medical care is delegated to insurers who, on behalf of their customers, negotiate contracts with providers on the costs and quality of health care. Purchasing is not a goal in itself, but it is an instrument to enhance system performance. It marks the advance of contractual relationships in health care. The Dutch experience teaches some important policy lessons. The first lesson is the absence of a one-size-fits-all model for purchasing. There are not only alternative models for purchasing; its success also depends upon what insurers make of it. The development of an effective purchasing strategy is a precondition for success. The regulatory framework must leave the purchaser sufficient discretionary power to shape its purchasing strategy. Formally, insurers hold considerable discretionary power under Dutch law. Secondly, it would be naive to expect immediate and sweeping results from purchasing. The slow uptake, marked by trial and error, did not come as a surprise for the simple reason that insurers had to learn their new role. In the early stage, they lacked a clear purchasing strategy. Perhaps of even greater importance was the absence of valid and reliable information on costs and quality and, initially, the limited scope of free pricing. What is true for cost information also applies to quality information. Nowadays, insurers have more information on what providers do and on practice variation, but valid and reliable information for quality management has remained scarce. The third lesson connects with the previous lesson: Purchasing has proven an evolutionary process. It takes time to materialize. Purchasing in 2019 looks different from purchasing in 2006. Currently, the purchasing agenda comprises many more items than in the past. Although cost control has remained an awkward issue in budget negotiations, one can observe a gradual transition from a competitive style of purchasing towards a more integrated style, with multi-year contracts, value-based contracts, contracts on a controlled shift of hospital-based care to off-hospital facilities and so on. Purchasing health care in practice 87 However, the new role of co-directing health care innovation that insurers seek to assume is no easy role. Are they really able to do so? This is an intriguing question against the background of the low institutional trust in the new health insurance scheme (see Chapter 7). Insurers struggle with the existential problem that the “coalition” of doctor and patient is much stronger than that between insurer and consumer. Also, consumers may hop from one insurer to another, and they show more interest in an insurer’s premium than in the merits of its purchasing strategy. An additional problem is that insurers are primarily perceived as money-driven financial agents. In short, insurers struggle with a legitimacy problem in purchasing. The fourth lesson is about incentives. Purchasing is doomed to fail without powerful incentives that give insurers an interest in investing in purchasing. Until 2012 temporary financial arrangements limited their financial interest to invest in purchasing, as surpluses were partially skimmed off and deficits partially compensated for. The decision to accelerate the termination of these arrangements meant a major change in purchasing. Now, being at risk for nearly all expenditures, insurers had a strong interest in cost control. Equally important is the need for an effective ex ante risk equalization scheme to prevent insurers from seeking loopholes in the prohibition of risk selection instead of investing in purchasing. The fifth lesson bears upon the prohibition of collective contracting in purchasing. We have seen that the practical effect of this prohibition should not be overstated. The representation model in purchasing has never disappeared and, in practice and increasingly so, insurers often voluntarily join the contract negotiated by the regional market leader. The representation model is also making a comeback in purchasing emergency care. The ban on collective contracting has been criticized as a source of confusion and higher administrative costs. However, the ban has not only costs but also yields benefits. Innovation benefits from different flowers flowering. It is a process of trial and error. Insurers and providers may learn from each other’s experience. Collective contracting deprives insurers of the possibility to create a distinct profile for themselves in care innovation. The experience with selective purchasing also offers some lessons. Selective contracting is a precondition for effective purchasing. Insurers have used this instrument to concentrate the provision of specific medical services in centres of expertise. Selective contracting by tenders is also practised in purchasing medicines, laboratory services and other medical goods. Contract denial concentrates in some specific sectors, including community nursing and outpatient specialist care provided by ITCs. For the rest, its role should not be overstated. Market conditions, the absence of information on costs and quality, the legally established duty of care, the power balance in the insurer–provider relationship and the The market reform in Dutch health care88 complexity of channelling patients to contracted providers explain the limited use of selective contracting. Selective contracting also creates a dilemma for insurers. Not only may it constrain patients’ freedom of choice and cause negative publicity (sometimes intentionally stirred up by providers), but it also worsens the credible commitment problem, even more so because insurers struggle with the problem of low public trust (see Chapter 7). Low public trust undermines the insurers’ power (Greer et al., 2020). The final lesson relates to the role of the state in purchasing. In the Dutch version of regulated competition, the state is formally held at a distance from purchasing. The state is said to bear only “system responsibility”. It determines the playing field and the rules of the game and is permitted to intervene in specific cases only if it considers access or quality of care at risk. Purchasing is a matter of the interplay between insurers and providers. In practice, however, system responsibility appears to be an ambiguous concept. Health care has always been a politically sensitive sector of public policy, and the call for state intervention in concrete cases has never ceased. The standard reaction of ministers of health has always been that intervention was beyond their formal competence. However, the picture of a minister who is not in the lead and unable to intervene feels uncomfortable. The concept of system responsibility simply does not work in practice, and even more so in a situation of a looming bankruptcy of a large provider (see the next chapter). Therefore, political initiatives to redefine system responsibility and reinforce its position do not come as a surprise. We will discuss this topic further in the next chapter and the final chapter. Chapter 5 Health care provision in regulated competition 5.1 Introduction The heart of any health care system is high-quality care for patients. The regulation, organization and financing of health care are only institutional vehicles for the provision of health care. This is so in the model of regulated competition. The introduction of competition in the health insurance market and the purchasing role of insurers are preconditions for the policy goals of the market reform: to organize the health care system so that it provides value-based health care that is accessible to each person, based upon solidarity and affordable from a financial perspective (financially sustainable). Competition in health insurance, together with active purchasing, is assumed to trigger competition in health care provision, which in turn incentivizes providers to provide value-driven health care. This chapter investigates the evidence for the validity of this essential policy assumption in the model of regulated competition (see section 2.4). Our focus is on three main issues. First, to what extent has the market reform indeed triggered competition in the provision of health care? Second, has regulated competition contributed to the accessibility of health care? Third, what is the evidence for the impact of the reform on quality of care? The impact of the market reform on health care costs is the topic of analysis in the next chapter. Investigating the effects of the market reform on the provider market is a precarious exercise for the simple reason that observed changes in provision are influenced by many factors in addition to the reform itself. These factors include, among others, demographic changes, advances in medical technology, the emergence of new concepts for the organization of care, the rise of the assertive patient and numerous policy changes (for example, spending cuts, extension or reduction of the service coverage of the Health Insurance Act or changes in the co-payment regime). These confounders and the absence of a control group make it difficult to disentangle the specific influence of the market reform on the provision of medical care. Furthermore, as we have seen in Chapter 2, the The market reform in Dutch health care90 market reform itself consists of a complex mix of interventions, the effects of which cannot be clearly isolated. In short, a complex causality problem makes it difficult to draw firm conclusions on the effects of the reform on health care provision. This chapter has the following structure: It starts with an investigation of the impact of the reform on the structure of the provider market. An important research finding in Chapter 3 on health insurance was that the market reform resulted in a substantial decline in the number of insurers. Did a similar decline take place in health care provision (section 5.2)? The analysis of the market structure effects continues with an overview of some hospital bankruptcies, two of which caused a great deal of public controversy (section 5.3). The next three topics of analysis concern the competition–price relationship (section 5.4), the competition–quality relationship (section 5.5) and the consequences of the reform on access to health care (section 5.6). Another topic is the impact of the reform on the financial performance of providers and hospital investments (section 5.7). The chapter ends with a brief review and some policy lessons (section 5.8). 5.2 Concentration and deconcentration In Chapter 3 we found that the market reform was followed by a significant drop in the number of insurers and that there have been hardly any new entrants since. As a consequence, the health insurance market is described as concentrated. The question now is: Can a similar process of concentration be observed in health care provision? We will see that the answer to this question is both positive and negative. The number of hospitals has significantly decreased (concentration), but the reform has simultaneously elicited a rapid increase in the number of ITCs (deconcentration). Deconcentration can also be observed in other fields of health care provision. To what extent can these contrasting developments be ascribed to the market reform? 5.2.1 Concentration Fig. 5.1 presents an overview of the numbers of general hospitals and university hospitals since 2000. The figure shows a steadily declining trend (−27%).1 The number of hospital locations is larger than the number of hospitals. For instance, 71 hospitals had 130 locations in 2016.2 1 In addition, there are 22 categorical specialist centres, including the centres for renal dialysis, radiotherapy, epilepsy and audiology. Many of these centres belong to hospitals. There are also categorical hospitals such as the Netherlands Cancer Institute (Antoni van Leeuwenhoek Hospital), the Princess Máxima Center for Pediatric Oncology and the Rotterdam Eye Hospital. 2 However, the effect of concentration on travel time is limited. With a few exceptions, in 2019 more than 99% of the population could reach the nearest hospital by car within 30 minutes (www.volksgezondheidenzorg. info). Health care provision in regulated competition 91 Fig. 5.1 Number of general hospitals and university hospitals, 2000–2018 Nu m be ro fh os pi ta ls 0 10 20 30 40 50 60 70 80 90 100 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 general hospitals university hospitals Source: The State of Public Health and Health Care website (www.staatvenz.nl). At first glance the declining trend in the number of hospitals may suggest that the market reform has propelled hospital mergers. However, this answer is largely wrong for the simple reason that the decline in the number of hospitals had already been going on for many years. In the period 1970–2000 the number halved, mainly as a result of mergers and takeovers (Meloen et al., 2000; den Hartog & Janssen, 2014). Roos et al. (2018) counted 1o6 hospital mergers between 1978 and 2015, of which there were 33 since 2000. Nowadays, patients in rural areas have limited choice if they want to visit a nearby hospital. The concentration of health care provision can also be observed in other sectors. For instance, between 1998 and 2004 the number of stand-alone residential homes dropped from 599 to 222, and the number of home-care providers from 107 to 55 (Fabbricotti, 2007). How to explain this concentration trend? Before the turn of the century, the provision of hospital care was fragmented. Many hospitals had a religious affiliation and had a small number of beds and specialties. Most of these hospitals considered merging or takeovers as the only way to survive in a health care landscape that was changing rapidly due to the advance of new technologies, the increasing subspecialization in medical care and the government’s bed reduction programmes. The ongoing “depillarization” in society (Chapter 1) meant that traditional religion-based obstacles to merging had gradually disappeared. The new function-based hospital funding model (introduced in the late 1980s) also stimulated merging because large hospitals were paid higher tariffs than small hospitals. The rationale for this arrangement was that large hospitals had a heavier caseload than small hospitals. This “merger bonus” made merging financially attractive (Hasaart, 2011). Hospital executives frequently mention cost reduction and quality enhancement as arguments for their mergers. Important quality-related arguments are better The market reform in Dutch health care92 opportunities for subspecialization and access to expensive equipment, the 24/7 presence of specialists and meeting volume standards (den Hartog & Janssen, 2014). Merging also reinforces the employer status of hospitals and extends the possibilities for learning from best practices, multidisciplinary consultations and task reallocation (Batterink et al., 2016). Cost reduction is assumed to be achieved by economies of scale. What can be said about the influence of the market reform on merging? A survey in 2012 among executives from all sectors of Dutch health care who had been engaged in mergers shows that the market reform has proved to be an additional motive for merging (see Table 5.1). Referring to the concentration of the health insurance market, they mentioned the merger as an effective strategy to reinforce their market position and bargaining power in an ever more competitive market. We conclude that the influence of the market reform on the concentration of hospital care should not be overstated. Concentration already started long before the reform became operative. However, the reform has shed more light on mergers than was common in the past. Critics argue that competition in health care requires a strict antitrust policy and question3 the merger approval policy of the Authority for Consumers and Markets, which in the Netherlands is responsible for overseeing competition law (Loozen, 2015). They consider this policy too lenient and a risk for competition. In their view some approvals have been doubtful or based on flawed information (Varkevisser & Schut, 2019). This antitrust-inspired criticism is not undisputed. The Health and Youth Care Inspectorate (IGJ) supports the concentration trend from a quality perspective. While recognizing the need for a geographically well-distributed supply of hospital care, it considers concentration the best guarantee of quality of care. Insurers, too, have frequently favoured hospital mergers. They see concentration as an effective strategy to fight the oversupply of hospital care and enhance quality of care. The counterargument that mergers could lead to higher costs is brushed aside by insurers with the argument that they are quite able to discipline hospitals as regards the costs of medical care. 3 These critical questions also apply to the approval of mergers between health insurers. Table 5.1 Main categories of motives for mergers (multiple answers) Quality of health care provision 69% Market/bargaining position 63% Efficiency 46% Financial reasons 28% Pressure from internal and external stakeholders 12% Source: Roos & Postma, 2016. Health care provision in regulated competition 93 5.2.2 Deconcentration Since the turn of the century, the drop in the number of general hospitals has been paralleled by a spectacular increase in the number of ITCs (zelfstandige behandelcentra). According to Kruse et al. (2019), 68% of ITCs are physician- owned. Most of these centres focus on a specific patient group, a specific specialty or a specific treatment, but there are also centres with a broader scope. Their activities involve mainly plannable, high-volume routine interventions and diagnostics. Insurers have contracted with many ITCs, but general hospitals are contracted at a larger scale. Stadhouders (2019) found that in 2015 the contracting index of ITCs and hospitals were 0.53 and 0.88, respectively. ITCs may also carry out activities not covered by the Health Insurance Act (for example, many forms of cosmetic surgery). In Dutch health care, organizations providing only non-reimbursable care are known as private clinics. Fig. 5.2 Rise in the number of independent treatment centres, 2000–2016 Nu m be ro fI TC s 0 100 200 300 400 2000 2002 2004 2006 2008 2010 2012 2014 2016 Source: Kruse, 2018. Fig. 5.2 highlights the increase in the number of ITCs since 2000 that were contracted by at least one insurer.4 Until the late 1990s, the government hindered the entrance of new providers with the argument that they were redundant. This policy changed in 1998 when ITCs were permitted, provided that their activity contributed to the resolution of the waiting list problem. The regulation was further relaxed in 2003 and finally terminated in 2006. Ever since, hospitals and ITCs must simply comply with the conditions listed in the Health Care Institutions Accreditation Act. Fig. 5.2 shows that ITCs were not a new phenomenon at the time of the introduction of regulated competition in 2006. However, the number of centres rapidly climbed afterwards to reach a peak in 2012. The data suggest a 4 The growth pattern fits an international trend, although with some delay. In the United States, the number of Medicare-certified independent ambulatory surgery centres doubled between 1991 and 2001, but recently this growth has slowed. In the United Kingdom, the number of ITCs peaked in the mid-1990s and has declined since (Kruse, 2021). The market reform in Dutch health care94 close relationship between the rise of ITCs and the introduction of regulated competition in health care.5 Some hospitals set up or participated in ITCs to strengthen their position in the marketplace, while at the same time specialists seized the opportunity to free themselves, in part, from hospital bureaucracy. Since 2012 the growth has reversed into a decline due to mergers and liquidations. Although the spectacular increase in the number of ITCs suggests deconcentration, in fact the market appears rather concentrated. Kruse et al. (2018) calculated that the market share in terms of total turnover of the four biggest ITCs was 32% in 2015, while the market share of the four biggest hospitals was 15%. The rapid growth in the number of ITCs does not mean that they have been able to take a substantial part of the total expenditures in specialist care. So far, their share in total expenditures has remained small, namely 3.4% in 2014 and 3.8% in 2016 (Vektis, 2018). This low percentage suggests that the impact of the deconcentration trend should not be overstated. Apparently, hospitals have managed to protect their traditional market share. The collective framework agreements on expenditure growth also played an important role in this respect. The low growth percentages (see Table 2.3) limited the room for expansion for the simple reason that insurers could not afford to put hospitals under extra financial pressure. As spelled out in the previous chapter, many ITCs felt that they were the last in the row. However, the picture varies by specialty. For instance, the ITCs’ share of ophthalmology in total turnover rose from 15.7% in 2014 to 18.4% in 2016 and their share of dermatology rose from 15.7% to 18.2% over the same period (Vektis, 2018). Kruse et al. (2019) calculated that the market share of ITCs in cataract surgery rose from 19.3% in 2013 to 24.1% in 2015. The deconcentration of the market for health care provision can also be observed in other fields of health care. For instance, the numbers of providers of community nursing and community care and of private-practice providers of mental health care have also significantly increased.6 This development suggests that the market reform has been accompanied by a growth in the number of providers in those areas where the investments needed to start a new provider unit are relatively low (low asset specificity). Later in this chapter we will see that this development has caused some problems with respect to the quality and costs of health care. In 5 Regulated competition is not the only factor that propelled the increase in the number of ITCs. The rapid advancement of medical technology also contributed to this increase. Nowadays, a great number of interventions that formerly required a hospital environment can be carried out in an ambulatory setting. 6 A similar development is seen in long-term care. The reform of long-term care has offered new opportunities for commercial providers to enter the market. The number of for-profit nursing homes, although still small, is rising. For-profit nursing homes are smaller than their non-profit counterparts. Whereas for-profit homes have 20 clients on average per location, this number is 64 for not-for-profit homes. Approximately 4.0% of the total nursing home client population lives in for-profit homes. The majority of for-profit facilities are affiliated with a chain, and most are owned by private individuals. For-profit facilities are frequently located in affluent regions (CTTO, 2020). Health care provision in regulated competition 95 Chapter 4 we saw that insurers are reconsidering their administrative procedures to address these problems (section 4.10). The resolution of these problems has also been a reason for a revision of the Health Care Institutions Accreditation Act. The new legislation, which will come into effect in 2022, includes stricter rules on governance and quality of care for licencing. 5.3 Hospital bankruptcies Hospital bankruptcies are not a recent phenomenon in Dutch health care. In the past also, hospitals went bankrupt, but the impact on access to health care was minimal because of their small size. Their activities were smoothly taken over by neighbouring hospitals. However, there are also examples of hospitals on the brink of bankruptcy that were rescued by state financial support. Under the present legislation, such bailouts are at odds with competition law, which permits state support only under specific conditions. Since the introduction of regulated competition, four hospitals have had to terminate their activities, suggesting that bankruptcies are no longer exceptional. The bankruptcies of the Ruwaard van Putten hospital in 2013 and the Sionsberg in 2014 did not attract much national attention because their activities were assumed by neighbouring hospitals. In contrast, the bankruptcies in 2018 of the MC Slotervaart and the MC IJsselmeer hospitals drew a lot of public attention. All bankruptcies followed years of mounting financial problems, concerns about the quality of care and dropping patient numbers (see Box 5.1). Four bankruptcies in such a short period of time may suggest that bankruptcies in health care have become an accepted outcome of competition. However, this was not the case. The bankruptcies of the MC Slotervaart and the MC IJsselmeer hospitals prompted a fierce debate on the pros and cons of competition. Left- wing politicians seized the opportunity to denounce competition in medical care, arguing that hospitals perform an essential public function and for that reason should never go bankrupt. They depicted a bankrupt hospital as an excrescence of competition. Not surprisingly, a great deal of the political debate focused on the role of the incumbent Minister of Health and his “system responsibility” for health care. In both cases the Minister’s role appeared to be rather detached. The Minister even declared that he was rather surprised by the rapid bankruptcy of both hospitals. He said that he understood the public’s criticism regarding the closure of both hospitals, but nevertheless he emphasized that it was not a public task to rescue weak hospitals, saying, “the ultimate goal is good patient care”. In an interview he reasserted his system responsibility for health care and did not see it as his task “to watch a stack of stones”. However, the parliament called for a The market reform in Dutch health care96 more active approach in future cases and for additional policy instruments to intervene in such instances. Insurers and hospitals were obligated to inform the Minister in an early stage so that, if necessary, appropriate measures could be taken. The Minister should no longer be empty-handed in situations like these. In the aftermath of the bankruptcies of the MC Slotervaart and the MC IJsselmeer hospitals, the government established a commission to evaluate what had happened. The commission did not condemn the bankruptcies but underscored the need for cooperation in case of a looming bankruptcy. Hospital bankruptcies Box 5.1 Bankruptcies of two hospitals in 2018 In October 2018 the Netherlands was startled by the simultaneous bankruptcies of two small hospitals, the MC Slotervaart hospital in Amsterdam and the MC IJsselmeer hospital in Lelystad. These hospitals were two separate holdings with the same ultimate beneficiary owners. They shared a turbulent history. The MC Slotervaart was originally a public hospital, established in 1975 by the city of Amsterdam and converted into a private not-for-profit hospital in 1997. The hospital always struggled with financial problems, culminating in 2006 when it almost went bankrupt. A private company was willing to settle the debt and took over the hospital. In 2013 the MC Group became the new owner. The MC IJsselmeer hospital was the result of a merger in 1999. The hospital was taken over by the MC Group in 2009, at which point it was close to bankruptcy. Due to intervention by the Minister of Health, the hospital was declared a “system hospital” that was needed for continuity of care in the region. It was then rescued by public financial support. A characteristic common to both hospitals is the dominant role of a single insurer. Zilveren Kruis has a market share of 55–60% in Flevoland province (the location of the MC IJsselmeer hospital) and 60% in Amsterdam. Consequently, the insurer played a leading role in the bankruptcy of both hospitals. The strong position of Zilveren Kruis was reinforced by the relatively small size of both hospitals. They had an average annual turnover of €100–120 million, whereas general hospitals had an average turnover of €270 million in 2017. The insurer used its strong position to switch from a yearly lump-sum contract to a price times quantity (PxQ) contract with a budget ceiling and, compared with other hospitals, low tariffs. By 2016 the relationship between Zilveren Kruis and the two hospitals had become increasingly tense. The insurer’s view on the future role of the hospitals increasingly diverged from how the hospitals saw their role, and mutual distrust grew. Eventually, Zilveren Kruis decided in October 2018 to apply for a moratorium. A few days later both hospitals declared bankruptcy. The consequences of the two bankruptcies for patients were different. Two nearby hospitals assumed a great deal of the medical care from the IJsselmeer hospital, with the result that continuity of care for patients could be relatively easily guaranteed. The situation in Amsterdam was quite different, however. Even though the Dutch Healthcare Authority declared that Zilveren Kruis had abided by its duty of care, the transfer of patients to other hospitals was chaotic. Health care provision in regulated competition 97 should not happen overnight but must take place in a controlled manner to guarantee patients uninterrupted access to health care. In the commission’s view, competition regulation should not hinder cooperation between insurers and providers to do so (COFZ, 2020). The independent Dutch Safety Board (Onderzoeksraad voor Veiligheid, OVV) concluded in a critical report that patient safety had been put at risk in both bankruptcies. For this reason, the Board argued for a revision of the insurers’ duty of care to guarantee the rights of patients in case of a bankruptcy. The regulatory framework should also be revised to facilitate a controlled completion of bankruptcy proceedings, said the Board (OVV, 2019). 5.4 The competition–price relationship The model of regulated competition assumes competition between providers and, then, that competition will enhance the efficiency and quality of health care. Competitive markets will perform better in terms of efficiency and quality than non-competitive markets. The question is: To what extent has the reform spurred competition? This section focuses on competition as rivalry and price competition. Competition on quality is discussed in the following section. 5.4.1 Competition as rivalry Competition defined as rivalry is anything but new in Dutch health care. Doctors and provider organizations have always contested for market share or patients. A greater market share meant more prestige and more revenues, while loss of market share meant less prestige and lower revenues. Rivalry between hospitals also played an important role in the period of state hospital planning. Under the Hospital Planning Act, hospitals competed for state licences to extend their activities or to build a new hospital (or rebuild the old one). Given the government’s policy of controlling cost by regulating the supply of health care, licences were scarce. State planning triggered a kind of “political rivalry” for licences between hospitals, with the Minister of Health in the role of market superintendent. The diffusion of medical technology is another example of rivalry between hospitals. In their study of the diffusion of da Vinci robots in Dutch hospitals, Abrishami et al. (2014) concluded that rivalry had elicited a medical arms race. Doctors wanted to render the best care possible to their patients with prostate cancer and feared losing status if they did not have the new technology. Hospitals used the new technology as a marketing tool to attract patients. Rivalry has also encouraged hospitals to invest in the “service quality” of health care. Examples are the opening of outreach clinics, the introduction of evening The market reform in Dutch health care98 consultation hours and online reservations, the development of facilities for one-stop provision of care and integrated care pathways, the reduction in the number of multiple-bed rooms and many other innovations to make the hospital more patient-friendly (NVZ, 2017; 2018). 5.4.2 Price competition Has the reform triggered price competition in hospital care? The fair answer to this question is probably to some extent. Price is indeed one of the factors taken into consideration when insurers organize a public tender. However, as noted in the previous chapter, the role of this purchasing strategy has remained limited in health care; it is seen mostly in the provision of generic prescriptions. Price competition is also absent in the model of standard contracts. Price partly plays a role in negotiated contracts. Insurers may use comparative price information to exact low prices in negotiating price–volume contracts. As described in Box 5.1, Zilveren Kruis used its strong market power to negotiate low prices in its contracts with the MC Slotervaart and the MC IJsselmeer hospitals. There is also evidence of the reverse. Halbersma et al. (2011) found a significant positive effect of hospital market power on price–cost margins. Berden et al. (2019) used the market share of a hospital and the insurer’s willingness to pay for its services as two indicators of a hospital’s market power. Controlling for differences in the complexity of patients (caseload), they measured a positive although small impact on hospital revenues. In its study of the impact of mergers on the volume and prices of hospital care, the Authority for Consumers and Markets found that prices in merged hospitals had risen faster than those in comparison hospitals. However, there was no evidence for systematic differences in volume growth between the two categories of hospitals (ACM, 2017b). Finally, Roos et al. (2018) found evidence of heterogeneous price effects across health insurers, hospital products and hospital locations. These effects were related to the degree of substitution between hospitals that could vary across products and with the relative bargaining capability of hospitals and insurers and the premerger price–cost margins of different products delivered by these hospitals. There is also some empirical evidence that providers of mental health care with market power were able to negotiate higher prices than providers with less market power (Brouns et al., 2020). It is important to note that competition on price has lost much of its relevance now that insurers and providers increasingly negotiate budget ceilings and lump- sum contracts (Chapter 4). The strategy of providers aims to achieve budget certainty for a certain period, and the aim of insurers is to control costs. In this situation prices typically do not reflect the true costs of medical services because they are calculated so that the product of prices and volume equals the negotiated Health care provision in regulated competition 99 lump sum (see Chapter 6 for more information). This administrative procedure is required because hospitals are paid per DBC/DOT. The consequences of this practice are astounding. Taking data from one big insurer, Douven et al. (2019) found that, for about half of hospital products, the highest and lowest contract prices across hospitals differed by a factor of three or more.7 Taking data from one hospital, they also found that 27% of contract prices were at least 20% higher or 20% lower than the average contract price in the market.8 In a really competitive market, one would not expect such big price differences. 5.4.3 Coopetition The debate on the role of competition in hospital care diverts attention from another important aspect of the relationship between hospitals. Hospitals not only compete but also cooperate with each other. The combined strategy of competition and cooperation (“sleeping with the enemy”) is known as coopetition. There is much evidence of coopetition in Dutch health care and that it is on the rise. In a report on cooperative relationships between hospitals, the consultancy group KPMG identified various types of coopetition, such as operating a common centre for diagnostics and treatments, concentrating some medical treatments in one hospital and other treatments in another hospital, establishing regional centres of expertise and making agreements on interhospital referrals (KPMG, 2018). Another form of coopetition is sharing medical staff. Using the claims data of hospitals, Westra (2017) found that roughly 20% of all medical specialists were affiliated with other hospitals, that sharing specialists occurred in all hospitals and that the rate of sharing specialists had increased between 2013 and 2015 in about half of all specialties. The personal motives of specialists for working in more than one hospital were predominantly facilitating subspecialization, adhering to volume requirements, improving quality of care, working in diverse settings, fulfilling 24/7 shift duties and financial benefits. The main motivations of hospital administrators to agree to sharing specialists were adherence to volume requirements, protection of market share, monopolization of the market and raising of barriers to market entrants. The presence of coopetition in specialist care indicates that the dichotomy between competition and cooperation fails to recognize the complex dynamics of specialist medical care. Hospitals not only compete with each other but simultaneously seek cooperative relationships with their competitors. The current emphasis on 7 The contract prices in the study included all hospital products that were priced by at least one hospital below the sum of the maximum deductibles (mandatory and voluntary) for consumers. The insurer had made these prices public because they were considered relevant for consumers. 8 Only freely negotiable prices were included in the data set. Prices regulated by the Dutch Healthcare Authority were excluded. The market reform in Dutch health care100 the need for regional cooperation in the programme The Right Care at the Right Place signifies that the rhetoric of competition has lost much of its justification as enhancing the efficiency and quality of health care. The pendulum seems to be swinging in the opposite direction. 5.4.4 Price competition in social support Strictly speaking, social welfare falls beyond the scope of regulated competition and market reform. Nevertheless, there are two reasons to briefly touch upon the experience with competition in this sector. First, since the reform of long-term care, social welfare services have proved to be one of the most competitive areas in the Dutch system. Second, the experience with competition in social welfare services influences the current political debate on the future of the market reform in Dutch health care (Chapter 7). The Social Support Act (Wet maatschappelijke ondersteuning), introduced in 2007 and revised in 2015, decentralized the purchasing of social support services (for example, household services, local public transport and personal guidance) to municipalities. The new law was part of the reform of Dutch long-term care (Maarse & Jeurissen, 2016). Municipalities received a state grant for carrying out their task but with a significant expenditure cut that was motivated by the assumption that decentralization would foster efficiency. Being exposed to financial risk, municipalities in their role as single payers have frequently made use of public tenders by inviting all interested providers or a preselected group of providers to submit a proposal with information about price and nonmonetary issues (for example, the level of schooling of their employees). The municipalities then contracted with at least two providers based on either the price only or the best price–quality combination. This is the public procurement model. Some municipalities instead chose the open-house model, in which they determined the catalogue of services to be contracted, a set of quality standards that a provider had to meet and the price of each service. Providers who accepted these conditions and were able to meet them according to the municipality qualified for a standard contract. Although use of the open-house model is on the rise, the public procurement model is still the most widely used model. A new development is the introduction of budget ceilings in contracts (PPRC, 2020). A politically sensitive issue in the public procurement model is whether European competition law requires municipalities to organize an open procurement procedure. This procedure has come under attack for being cumbersome and expensive. Critics have complained of “overreached competition” with no added value. The Minister of Health has brought forward an initiative to absolve municipalities from the requirement to organize a costly open procedure (Uenk, 2019). Health care provision in regulated competition 101 The market for social support services is quite competitive. The market share of initial market leaders has significantly shrunk, while other providers, including new entrants, have managed to gain substantial market share (van Eijkel et al., 2017). Municipalities have used their role as purchasers to extract lower prices for household services. After an increase in 2009, the average price of these services slightly decreased in the period from 2010 to 2013 (van Eijkel et al., 2017). In the view of critics, their purchasing strategy has triggered a race to the bottom, resulting in the bankruptcy of several large provider organizations, with thousands of care workers losing their jobs. In response to frequent complaints about low prices, the government issued a regulation to force municipalities to pay fair prices. However, the implementation of this new regulation has appeared to be troublesome in practice. Another source of complaints is administrative costs. Provider organizations have frequently criticized the high administrative costs of public procurement procedures and the lack of uniformity in municipal accounting procedures. Providers who are active in more than one municipality can be confronted with different accounting procedures. 5.5 The competition–quality relationship Over the last decades the attention to quality of care has increased enormously. Quality management to improve clinical quality of health care, by means of guidelines, quality measurement and health technology assessment, has evolved as a new policy path in Dutch health care (Box 5.2). What is the evidence for a positive impact of the market reform on the clinical quality of Dutch health care? While advocates of competition assert that competition will have a positive effect on clinical quality, opponents contend that the emphasis on containing costs will ultimately lead to skimping on quality. The international evidence on the impact of competition on the quality of care is mixed. While some studies find a positive impact (for example, Gaynor et al., 2013), other studies find no impact or even a negative impact. In his review of the competition–quality relationship, Gaynor (2006) suggests that competition may have a positive effect on quality in non-price competition markets (providers compete only on quality). However, if providers compete on both prices and quality, as is typical in the Netherlands, the impact of competition on quality is uncertain because health care providers have two options. They may invest in quality, hoping that this will ultimately be the winning strategy, or they can follow a strategy of cost saving by skimping on quality. Which strategy prevails depends on the relative weights that purchasers and providers attach to price and quality. The market reform in Dutch health care102 Unfortunately, research on the competition–quality relationship in the Netherlands after the reform is scarce. The Health and Youth Care Inspectorate (IGJ, 2018) has always expressed concerns about the quality of medical care in ITCs. The Inspectorate was particularly critical in the early days of these centres. Since then quality of care has made much progress, although to a lesser extent than expected. Although many ITCs meet the quality norms, “there are still (too) many centres failing to do so” (p. 5). The Inspectorate found structural deficiencies in the care of vulnerable groups, in surgical care, in compliance with patient rights and in control of the professional quality of caregivers. There are only a few quantitative studies on the competition–quality relationship. Using data on cataract surgery in the period 2006–2009, Heijink et al. (2013) found no clear relationship between prices and quality. Because of the small number of available quality indicators, the researchers warned against drawing firm conclusions on the competition–quality relationship. Another study on the competition–quality relationship was conducted by Croes et al. (2018). Their study included both hospitals and ITCs and covered the period 2008–2011. Competition was measured as the weighted average of a provider’s market share per micro-market (market power).9 Data on quality 9 Based on claims data. Box 5.2 Rise of quality management in Dutch health care The first initiatives to improve the quality of health care date from the late 1970s. In the course of time many guidelines have been introduced. For instance, the Dutch College of General Practitioners published its first guideline in 1989 (on diabetes mellitus type 2) and celebrated the issuance of its hundredth guideline in 2012. Quality measurement and public reporting were a spearhead of the Innovating Quality Programme, launched by the Ministry of Health in 2002. The key message of the programme was transparency. Quality should not be taken for granted but rather should be made visible as much as possible by means of outcome indicators (van de Bovenkamp et al., 2014). In its early stage, quality management lacked effective coordination. The result was a patchwork of initiatives. After failed attempts to coordinate, the government created the National Health Care Institute (ZiN) to fill the coordination gap. Its mission was “to reinforce, in an integrated way, the quality, safety, transparency, efficiency and effectiveness of care, from a perspective that is recognizable to both clients and professional caregivers”. The Institute is a regulatory agency at arm’s length from the Ministry of Health. If stakeholders are unable to reach consensus on quality standards or quality indicators, the Institute can take a binding decision (enforcement power). Quality standards and quality indicators meeting the Institute’s quality framework are recorded in a public register. Health care provision in regulated competition 103 were taken from the Dutch Health Care Transparency Programme set up by the Health and Youth Care Inspectorate. The researchers found a significant negative relationship between market share and quality scores for cataract and bladder tumour care. This result meant that providers of these treatments had better quality scores in competitive markets than providers in concentrated markets. A similar relationship was not found for treatment of adenoids and tonsils. The researchers suggest that this difference may be because the procedures are less complex. Another finding was the absence of a relationship between price and quality scores: providers with higher quality scores did not charge higher prices. The researchers mentioned that a limitation of their study was that quality had not been measured by outcome indicators but rather by structure and process indicators. In their study of the quality of cataract surgery, Kruse et al. did not find differences in clinical quality (measured by patient-reported outcomes) between hospitals and ITCs. However, the researchers reported that ITCs performed somewhat better on patient satisfaction than hospitals (Kruse et al., 2019). A study commissioned by the Authority for Consumers and Markets found no evidence for the claim by hospitals that mergers were necessary to uphold and foster quality of care. Comparing the quality of care of merger and non-merger hospitals, the researchers found, with a few exceptions, no significant effect of merging upon the quality of care. Some significant effects found were positive, while other effects were negative. The research was based on three types of quality indicators: disease-specific outcome indicators, waiting times and hospital standardized mortality rates. The researchers also pointed to some adverse effects of mergers on quality. The new organizations tended to be sluggish and uncommunicative, with great distance between specialists, between specialists and patients and between specialists and management. Another problem was that mergers were associated with culture clashes and temporary uncertainty, which diverted attention and resources from patient care to organizational issues for a long period. Whether the newly created organizations would ultimately perform better on clinical quality remains uncertain (Batterink et al., 2016). To test the claim of the critics of the market reform that competition may lead to skimping on quality, Roos et al. studied the impact of market concentration on the quality of non-acute hip replacements. Quality was measured by the 90-day unplanned readmission rate. Using data for the period 2003–2007, the researchers found no evidence of a negative effect of price competition on clinical quality. There was even evidence of a temporary positive effect. The “null effect” was robust for different definitions of market size and measuring quality by the 30-day unplanned readmission rate (Roos et al., 2018). The market reform in Dutch health care104 This brief overview leads us to conclude that the evidence of a positive effect of competition on quality is patchy and in part contradictory. Studies focus on specific interventions with relatively simple outcome indicators. The results cannot be generalized to other types of care. In sum, there is little evidence of a positive or a negative impact of competition on quality of care. Instead, our assumption is that the improvement of quality of health care must be largely ascribed to initiatives of the medical profession to get better outcomes. The current trend in thinking is that enhancing the quality of Dutch health care requires cooperation instead of competition. Currently, there are several cooperating networks claiming that they have achieved some remarkable results. For instance, Govaert (2017) found that participation in the Dutch Surgical Colorectal Audit had resulted in a 20% decline of mortality among patients with colon cancer in a period of only three years and a 29% decline of the number of complications. Another remarkable finding was the decline of treatment costs by 9%. Another study reported that cooperating heart centres had benefited from paying more attention to the preoperative treatment of patients, multidisciplinary cooperation and the use of supplementary checklists in the operating theatre. Mortality within 12o days after bypass surgery had dropped from 2.3% to 1%, mortality after one year dropped from 3.1% to 2%, and the percentage of re-operations dropped from 5% to 3.2% (van Veghel, 2019). However, the increased emphasis on quality is not costless. Quality management is associated with high administrative costs. These costs have recently been studied by Zegers et al. (2020). They used a mix of observational methods, including participative observation, a survey and semi-structured interviews with health professionals in two academic and one teaching hospital in the Netherlands. Professionals were found to spend 52.3 minutes a day on quality registrations. The average number of quality measures was 91, with 1380 underlying variables. Only 36% of these measures were perceived as useful. Respondents said that filling out redundant paperwork made their work less enjoyable. 5.6 Access to medical care What has been the impact of the market reform on access to medical care? In Chapters 2 and 3 we saw that the regulatory framework includes several provisions to preserve universal access to health care: (1) the broad standard package of health services; (2) the ban on risk selection and risk rating; and (3) the care allowance to uphold income solidarity. This section discusses the impact of the reform on access from two perspectives. First, attention is paid to the impact of the mandatory deductible on access. The second topic of discussion is the impact of the reform on waiting time. Health care provision in regulated competition 105 5.6.1 Mandatory deductible The introduction of the mandatory deductible and the successive increases of the deductible from €150 in 2008 to €385 in 2016 (Chapter 3) have always been a much-disputed element of the reform. In 2020 the deductible accounted for 6.3% of health care expenses covered by health insurance (ZiN, 2020). Critics argue that the deductible may provoke care avoidance because patients cannot afford to pay the deductible. The critique fits into a historical pattern. In the 1980s and 1990s, the government had introduced several co-payment regimes to make patients aware of the high costs of medical care and encourage them to refrain from medical care that might not be necessary. Health care was no free lunch! However, each regime was considered unfair and revoked after only a short period of time. The mandatory deductible is still in place, but, for political reasons, it has not been raised since 2016. Some political parties promise their voters to reduce the mandatory deductible or abolish it altogether. What is the evidence for the claim that the deductible has reduced access to care? Unfortunately, research on this question is limited. The results of surveys on this topic are difficult to interpret because of doubts on the reliability of the answers given. An exception is a study by van Esch et al. (2017). They studied care avoidance by investigating whether patients with a referral to a medical specialist had indeed visited the specialist. They did so by linking recorded referrals from general practitioners to medical specialists and claims from medical specialists to health insurers. Non-compliance was measured as the absence of a claim after referral. Fig. 5.3 demonstrates what they found. Fig. 5.3 Relationship between the mandatory deductible and non-compliance with referrals Percentage 15 17 19 21 23 25 27 29 Eu ro s 100 150 200 250 300 350 400 Year 2008 2009 2010 2011 2012 2013* mandatory deductible non-compliance rate *Data for the first half of the year. Source: van Esch et al., 2017. The market reform in Dutch health care106 Fig. 5.3 shows a moderately positive association between non-compliance and the amount of the mandatory deductible. The increase of the non-compliance rate (line) in 2011 and 2012 corresponds to the increase of the mandatory deductible (bars) in these years. This result suggests that there is indeed some evidence that a group of patients has refrained from medical care because they had to pay the mandatory deductible. However, the considerable increase of the deductible in 2013, from €220 to €350, was not followed by a substantial increase in the non-compliance rate. An explanation of this finding may be the presence of a ceiling effect: The non-compliance rate does not further increase beyond a natural limit. The researchers did not investigate the reasons for non-compliance, but they emphasized that refraining from specialist care could be caused by many more factors than just the mandatory deductible.10 A remarkable but problematic research finding was that the increase in non-compliance had been greatest in children under 18. Because medical care for children under 18 is exempted from the deductible, the finding suggests that there is still a lot of misunderstanding about the mandatory deductible. 5.6.2 Waiting time for appointments Another dimension of access is waiting time. In Dutch health care waiting time norms are known as “Treek norms”, named after a rural estate where they were developed by health care providers and insurers in 2000. The norm for the maximum acceptable waiting time for non-emergency diagnostics and polyclinic visits is four weeks and for treatment, seven weeks. Fig. 5.4 illustrates that the trend in waiting times for diagnostics, polyclinic services and treatments in hospitals is rising again. The vertical axis represents the percentage of hospital departments above the waiting time norms. Fig. 5.4 shows that the percentage of hospital departments with excess waiting times dropped in the period 2010–2013. The great majority of hospitals managed to comply with the waiting time norms. However, as of 2014 the percentage failing to meet these norms began to rise. In its monitor of specialist care, the Dutch Healthcare Authority reported a few overruns of the waiting time targets in polyclinic care in 2014 (NZa, 2014c). Since then the situation in several specialties has worsened, in particular in polyclinic care. In 2016 the percentage of hospital departments with excessive waiting times varied from 77% for gastroenterology to 34% for neurosurgery (NZa, 2017c). 10 International comparative research of the Commonwealth Fund reveals a socioeconomic gradient in skipping medical care. The study found that, in the Netherlands, 12% of lower-income persons said that they had problems paying their bills for medical care in the previous year, compared with 3% of higher-income persons (Commonwealth Fund, 2020). Health care provision in regulated competition 107 The increase in waiting times since 2013 cannot be attributed to a single factor. Two explanatory factors are the shortage of medical staff and nursing personnel and the rising demand for health care. The tight restrictions on net annual expenditure growth in the framework agreements (Chapter 2) can be mentioned as a third factor. The impact of the market reform on waiting times is unclear. One would expect a positive effect because contracts with insurers include provisions requiring hospitals to meet the waiting time norms (see Chapter 4). However, the trend since 2013 casts doubts on the effectiveness of these provisions. There is also some circumstantial information that contracts may cause longer waiting lists. This effect occurs, mostly at the end of the year, if a hospital reaches the contracted volume or budget ceiling for that year and postpones the treatment of new patients unless the insurer is willing to raise the volume or budget ceiling. It even happens that only the patients of a specific insurer must wait longer for health care. Unfortunately, we have no insight into the magnitude of this problem. 5.7 Financial management and capital investments Due to the reform, the interest of banks, insurers and other financial agents in the financial performance of hospitals has significantly increased. A survey among hospital administrators found that insurers and banks are closely involved in plans for merging. Banks played a decisive role in 40% of the consolidations in 2014, compared with 9% in 2013 (KPMG, 2015). The interest of banks in hospitals’ financial performance is closely related to the increased exposure of hospitals to financial risk. Under the regime of the Hospital Planning Act, hospitals had to apply for a state certificate of need for major construction works. The government used this instrument to regulate the Fig. 5.4 Percentage of hospital departments above waiting time norms, 2010–2018 0 5 10 15 20 25 30 35 201820172016201520142013201220112010 DiagnosticsTreatmentPolyclinic services Pe rc en ta ge Source: OECD/World Health Organization, 2019. The market reform in Dutch health care108 capacity of hospitals. The certificate guaranteed that hospitals cover the costs of rent and depreciation of their capital investments and, consequently, kept the banks’ financial risk to a minimum. The regulation of capital investments changed in 2008. The termination of the Hospital Planning Act meant that hospitals from then on could freely decide about their investments. However, they also had to bear the financial risk of these investments.11 Thus, decentralization of strategic decision-making went along with devolution of financial risks. Not surprisingly, hospitals reacted to their increased risk exposure by making a critical assessment of their investment and business plans. Hospitals’ increased risk exposure meant that banks also were confronted with a higher risk. As a reaction to this development, they increased the surcharge for risk exposure and gradually shortened the depreciation period from 40 to 25 years, typically, with the result that the annual funding costs of the investments themselves increased. This development was reinforced by the introduction of a stricter international regulatory framework on bank capital adequacy, stress testing and market liquidity (Basel III). The new requirements, which also apply to insurers, trickled down to the micro level where hospitals and financiers decide about investments (Janssen, 2017). Fig. 5.5 shows that devolution of financial risks has clearly had a downward effect upon hospital investments in major construction works. In effect, since the price of these investments has gone up, the downward slope is actually an overestimate with regard to the infrastructure volumes of these investments. Fig. 5.5 Investments in major hospital construction works, 2010–2017 (billions of euros) 0 0.5 1 1.5 2 2.5 2010 2011 2012 2013 2014 2015 2016 2017 Source: NVZ, 2018. The new regime for capital investments is not the only explanation for the declining trend in major construction works. The changing context of hospital care, such as the shortening of the length of stay and the ongoing substitution 11 Tariffs that hospitals negotiate with insurers include a component for covering the costs of rent and depreciation. Health care provision in regulated competition 109 of ambulatory care for inpatient care, has also made a revision of capital investment plans necessary. According to the Health Care Sector Guarantee Fund (Waarborgfonds voor de Zorgsector, WFZ),12 hospitals struggle with an overcapacity problem (WFZ, 2017). Fig. 5.6 Financial reserves of hospitals as a percentage of total turnover, 2007–2018 0 5 10 15 20 25 30 35 40 45 50 20 08 20 09 20 10 20 11 20 12 20 13 20 14 20 15 20 16 20 17 20 18 20 07 Best performers Average Pe rc en t Worst performers Source: Own calculation. Fig. 5.6 shows that – as Chapter 3 has shown for insurers as well – hospitals’ financial reserves have significantly increased since 2007. Hospitals have been able to build up their financial buffers in response to their increased exposure to financial risks. Fig. 5.6 also indicates an increased spread between the best and worst performers. This spread corresponds to the warning of the Health Care Sector Guarantee Fund that averages may conceal great differences in financial performance among individual providers. While some hospitals perform quite well financially, others struggle with serious financial problems. 5.8 Discussion and policy lessons This chapter has analysed the consequences of the market reform for the provision of health care. The deregulation of accreditation was followed by a remarkable growth of the number of ITCs, or “focus clinics”, which focus their activity on specific patient groups, specific treatments or specific diseases. These centres can 12 Hospitals (and other provider organizations) can apply for membership in the Health Care Sector Guarantee Fund. This Fund was created by the government and the lead organizations of health care providers in 1999. The goals of the Fund are to foster the continuity of capital investments in health care, restrict rent percentages and stimulate good financial management. Nowadays, the Fund guarantees about 50% of long-term capital investments in health care. Guarantees require a critical examination of a hospital’s financial situation. Most hospitals are members of the Fund (WFZ, 2017). The market reform in Dutch health care110 be considered, at least to some extent, a “winner” in the reform. Also, in other areas of health care, the reform has induced an increase in the number of small providers. The lesson of the market reform in Dutch health care is that relative low investment costs attract new providers to enter the market and benefit from competition. However, in actual monetary figures, the impact of this development should not be overstated. With the exception of ophthalmology and dermatology, the market share of ITCs has remained small. Hospitals have kept their strong position in the provision of specialist medical care. Since 2006 four hospitals have gone bankrupt. This number suggests that bankruptcies have become part of the game and are accepted as an outcome of competition. The Dutch experience shows this is not true. Particularly, the bankruptcies of the MC Slotervaart and the MC IJsselmeer hospitals aroused much public anger. Was it acceptable that hospitals went bankrupt and that this could happen without any intervention by public authorities, in particular the Minister of Health? Does “system responsibility” imply an empty-handed Minister in cases like these? These questions clearly resonate in political debates on how to reshape health care (see Chapter 7). Advocates of regulated competition argue that competitive markets yield better efficiency and quality than non-competitive markets. This argument presupposes that the reform will trigger fierce provider competition on price. Our analysis suggests that there is reason to question this assumption. The policy lesson drawn from the Dutch experience with competition is that practice differs from theory. The role of price competition in current Dutch health care should not be overstated. Competition exists only if insurers organize a public tender with price as the main discriminating factor, and this is not the case in standard contracts. In non-standard contracts a sought-for lump-sum amount dominates the outcome and spurs substantial artificial heterogeneity in hospital prices. There is, nevertheless, some evidence of the influence of market power on prices in negotiated contracts, but the picture that providers seek to outbid each other is wrong. Prices have lost much relevance because of the frequent use of lump- sum contracts. Further, the picture of fierce competition ignores the role of the widespread (and increasing) phenomenon of coopetition in hospital care. Advocates of regulated competition postulate a positive effect of competition on quality of care. We found little evidence of such an effect. The experience with competition in Dutch health care suggests that the assumption of a positive impact on quality of care is highly questionable. The current debate actually goes in the opposite direction: policy-makers and professional communities emphasize the need for cooperation to enhance quality of care. Several expert networks have been set up for this purpose, and it is only here that there is some evidence of promising results. Health care provision in regulated competition 111 The market reform had major implications for financial management and capital investments. The termination of state hospital planning and the devolution of financial risks have made hospitals more risk aware. The price of capital has increased substantially. The critical assessment of major construction plans helps to explain the decline of investments in construction works in the period 2010–2017. Another effect of the devolution of financial risks was that financial agents (banks) found themselves exposed to increased financial risk. They reacted by raising rent percentages and shortening the depreciation period, as a consequence of which the costs of investments have grown. Whether this negative effect has been compensated for by greater efficiency, as assumed by the advocates of regulated competition, is an important question that deserves more research. On average, hospitals – like insurance companies – have been able to reinforce their financial performance. The financial reserves as percentage of total turnover have increased in the period 2007–2018. However, these percentages are averages, and the spread around the average has widened in the course of time. Competition has motivated hospitals to develop a strategy of avoiding uncertainty by building substantial financial buffers.

Chapter 6 Impact of the market reform on health care expenditures 6.1 Introduction Rising health care expenditures have always been a matter of concern in Dutch health care policy-making, but became a “headache dossier” in the 1970s after the first and second oil crises. In order to curb expenditure growth, the government developed various cost control instruments, including hospital planning (with reduction in the number of beds) and hospital budgeting. The budget instrument that put an end to the open-ended funding of hospitals proved particularly effective, but after two decades its drawbacks had also become apparent (Hasaart, 2011). The main problems were lack of powerful incentives to invest in innovation, inflexibility and lagging productivity growth. In the 1990s the government introduced a new policy instrument: Net annual expenditure growth was restricted to 1.3%. Concerns about expenditure growth and the effectiveness of cost control instruments were important reasons for the government to establish the Dekker Committee in the late 1980s. In its request for advice, the State Secretary of Health explicitly hinted at the development of a new strategic approach, as now old-style policy instruments had proved largely ineffective. The report “Willingness to change” put the problem of cost control and financial sustainability central in its problem analysis. In the Committee’s view, Dutch health care suffered from a fundamental lack of efficiency and flexibility. All its proposals were intended as a new approach to cost control. As spelled out in section 2.2, the reform as proposed by the Dekker Committee got stuck in the early 1990s. However, after an intermezzo of several years, the government again put the market reform on the political agenda. And once again the need for effective cost control proved a major reason for a fundamental overhaul of the health care system. What made reform even more urgent was the waiting list crisis in the late 1990s. A court ruling that long waiting lists were unlawful compelled the government to substantially increase the health The market reform in Dutch health care114 care budget. Hospitals and other provider organizations were paid extra if they managed to push up their production. This policy measure actually put an end to the strategy of curbing expenditure growth by means of strict budgets. In order to reopen the debate on reform, the government published a policy document titled “A question of demand”, which offered a global sketch for restructuring health care. The new policy paradigm was that regulated competition would make health care more efficient and, along the way, restrict expenditure growth (Chapter 2). The report mentioned several reasons for promoting effective cost control. Escalating expenditures would put universal access to health care at risk in the future and erode public support for a fair (solidary) system of health care financing. Another risk was “crowding out”: uncontrolled health care expenditure growth would constrain the space for other necessary public expenditures (for instance, education, housing and public safety) and would also limit the room for private spending. Last but not least, uncontrolled growth would impair the competitiveness of the Dutch economy (Taskforce Beheersing Zorguitgaven, 2012). In sum, the government presented its sketch of the market reform as a new approach to effective cost control in order to assure the financial sustainability of health care (sections 1.5 and 2.4). Many of the opponents of this proposal saw it exactly the other way and feared cost increases, such as experienced in the United States, as a result of the market reform. The central question in this chapter is: To what extent has the market reform been an effective instrument for cost control? This question refers to a controversial issue in the scientific and political debate on the cost-saving effect of regulated competition. The evidence from empirical research is mixed (Stadhouders, 2019). Some studies of the effects of pro-competition reforms found a cost-saving effect. For instance, Melnick and Zwanziger (1995) concluded that competition in California had resulted in a 12% cost decrease in high competition areas. In the United States, states that relied on non-competitive strategies proved less successful in cost control. However, other US studies found no such effect (Merrill & McLaughlin, 1986). A recent study on hospital competition in the United Kingdom reached a similar conclusion. While the researchers measured positive effects of competition with fixed prices on quality of care, they could not confirm a cost-saving effect on expenditures (Gaynor et al., 2013). Two methodological problems hinder the research into the cost-saving effect of the market reform. The first problem is the absence of a control group: We do not know the change in health care spending if there had been no reform. The second problem is that the financial impact of the reform cannot be well isolated from the financial consequences of other policy measures taken during the reform, for instance modifications in the standard benefit package, expenditure cuts and the introduction of new payment models or co-payments. Consequently, we Impact of the market reform on health care expenditures 115 must be cautious in drawing firm conclusions, the more so because empirical research into the cost-saving effect of the market reform in Dutch health care has been scarce so far. The structure of this chapter is as follows: After a brief sketch of the international picture (section  6.2), there is an analysis of the growth of total health care expenditures (section  6.3) and expenditure growth in a number of sectors (section 6.4). Next follows an analysis of cost control of outpatient prescription medicines (section 6.5) and the impact of the collective framework agreements in constraining expenditure growth (section 6.6). The following sections discuss a few specific issues: the impact of the production incentive (section 6.7) and of co-payments on health care spending (section  6.8) and the problem of administrative costs (section 6.9). The chapter ends with a discussion and some policy lessons (section 6.10). 6.2 The international picture Fig. 6.1 compares total public health care spending in 2018 in seven countries of the European continent, measured as a percentage of GDP. The figure indicates that Dutch health care performed relatively well in 2018. Fig. 6.1 Health care spending as a percentage of gross domestic product in selected countries, 2018 or nearest year Pe rc en ta ge 0 2 4 6 8 10 12 14 Belgium France Germany Sweden Switzerland United Kingdom Netherlands Source: OECD/World Health Organization, 2019. The market reform in Dutch health care116 Fig. 6.2 Annual growth of per capita health care spending (real terms) in selected countries, 2008–2018 Pe rc en ta ge 0 1 2 3 4 5 6 Belgium France Germany Netherlands Sweden Switzerland United Kingdom2000-2008 2008-2013 2013-2018 Sources: OECD, 2014; OECD/World Health Organization, 2019. Fig. 6.2 compares the growth of per capita health care spending in three consecutive periods. The figure shows that the Netherlands had the highest average growth in the period 2000–2008. However, this picture radically changed in the next two periods. The growth of health care spending in the Netherlands was relatively low in both subsequent periods and lowest in the period 2013–2018. Particularly, the low growth rate in the period 2013–2018 is remarkable. In section 6.6 we will argue that this pattern is closely associated with the introduction of the collective framework agreements on expenditure growth and the abolition of the safety nets in the risk equalization for health care insurers. 6.3 Rise of total health care expenditure Fig. 6.3 shows the growth of per capita health care expenditure (international definition) from 2000 to 2018. In this period total health care expenditures rose from €34.8 billion to €76.9 billion, and per capita expenditures rose from €2187 to €4462. Fig. 6.3 demonstrates the effect of the spending programmes that the government implemented to shorten waiting times in the early 2000s. Expenditures rose by 11.5% in 2002 and 11.6% in 2003. The introduction of the reform in 2006 was followed by a period of more moderate growth. Since 2009 expenditure growth has been relatively low, and as of 2013 the growth of the GDP even outstripped health care expenditure growth (Fig. 6.4). The remarkable difference in percentages between GDP growth and health expenditure growth in 2009 is an effect of the fall of the GDP (−3.7%) in that year due to the financial crisis across Europe. Impact of the market reform on health care expenditures 117 Fig. 6.3 Growth of total per capita health care expenditures, absolute (blue bars) and percentages (red line), 2000–2018 Pe rc en ta ge 0 3 6 9 12 Eu ro s 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 20 00 20 01 20 02 20 03 20 04 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 20 14 20 15 20 16 20 17 20 18 Source: Statistics Netherlands (https://www.cbs.nl/nl-nl/cijfers). Fig. 6.4 Growth of health care expenditure and gross domestic product, 2009–2018 Pe rc en ta ge -6 -4 -2 0 2 4 6 8 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 health care GDP Source: MoH, 2019b. Figures 6.3 and 6.4 give only a global picture of expenditure growth in Dutch health care. They give no sign of the cost-saving effect of the market reform. For this reason, it is necessary to investigate expenditure growth in greater detail. This will be done in the next sections. 6.4 Expenditure growth by sector Table 6.1 gives a comparative overview of growth, in four successive time intervals, of total health care expenditure and expenditures in the sectors of hospital care, general practitioner care and mental health care. The table demonstrates that the growth pattern in each sector is to a great extent congruent with the growth pattern of total health care expenditures (Fig. 6.3). The market reform in Dutch health care118 Still, a few preliminary observations can be made. First, the rate of expenditure growth in hospital care outstripped the rate of total expenditure growth in each period. This result will be discussed further in section 6.6. The second observation relates to expenditure growth in general practitioner care. The percentages show the comparatively largest expenditure growth rate in this sector since 2010. This pattern can be interpreted as an effect of the government’s policy to substitute primary health care for hospital care (Table 2.3). As of 2010 the growth in mental health care spending has been less than spending growth in the other sectors. 6.4.1 Hospital care The sharp expenditure growth after the turn of the century is the result of the government’s spending programme to shorten waiting times. The decline of the average growth percentage in 2005–2010 suggests that it went back to normal (regression to the mean). Yet, it was a period with some fierce conflicts. Govern- ment and self-employed specialists were in conflict over specialists’ revenues after the introduction of the reform. The origin of the conflict was that the government, for fear of escalating costs, did not permit self-employed specialists and insurers to negotiate on tariffs in the free-pricing segment (B-segment) of hospital care. As an alternative, the government negotiated a norm hour tariff for specialist care. The tariff that specialists were paid for a specific treatment was calculated as the hour tariff divided by the normative time for that treatment. This alternative model failed in practice due to flawed estimations of norm times.1 To correct 1 Until 2005 self-employed specialists were reimbursed on a fee-for-service basis, but with a lump-sum ceiling (budget constraint) per hospital. As of 2005 this model was continued for the A-segment of hospital care (regulated prices), while specialists were paid per DBC in the B-segment (free prices). The flaw in the new reimbursement model was that the norm time per DBC had been overestimated. Since 2008 specialists have been completely reimbursed per DBC, but in 2010–2011 with a discount factor to correct for overspending. In 2012 there was another revision of the reimbursement model. Ever since, self-employed specialists have been reimbursed for each health care product (DOT) multiplied by a proportionality factor (Krabbe-Alkemade, 2014). Table 6.1 Average nominal percentage growth of total health care spending and spending in three selected sectors 2000–2005 2005–2010 2010–2015 2015–2018 Total expenditures 9.3 6.0 2.3 3.0 Hospital care 10.2 7.6 2.8 3.1 General practitioner care 7.8 6.9 4.7 4.9 Mental health care 12.4 7.0 2.5* 1.7 * This percentage applies to the period 2010–2014. The year 2015 has been excluded because of the transfer of youth care, including specialist mental health care, to the Social Support Act. This transfer was part of the reform of long-term care (see Chapter 7). Source: Statistics Netherlands (own calculations). Impact of the market reform on health care expenditures 119 for the resulting overspending, the Minister decided to skim off the specialists’ revenues by tariff cuts in 2010 and 2011. The introduction of a free-pricing segment in the hospital funding model also sparked a conflict between the government and hospitals. While hospitals claimed that the hospital budget constraint did not apply to the B-segment, the government argued that this segment was not exempted from the budget constraint. Both conflicts were important reasons for the stakeholders to negotiate a multi-year framework agreement on expenditure growth and other policy issues. Most remarkable in Table 6.1 is the significant drop of expenditure growth in the periods 2010–2015 and 2015–2018. This drop is likely to be closely associated with the aforementioned framework agreements on health care spending and the abolition of the safety nets for health insurance companies. In section 6.6 the effect of these agreements on spending will be further investigated. Unfortunately, detailed empirical research on the effect of competition on hospital spending is scarce. An exception is a study by Ikkersheim and Koolman (2013), who contended that price increases in the free-pricing segment from 2006 to 2009 had been lower than price increases in the segment with regulated prices (segment A) (see Table 2.2 for more detailed information). This finding led them to conclude that competition had the expected effect on prices. The Healthcare Authority (NZa, 2013) also reported larger price increases in the A-segment than in the B-segment in the period 2006–2010. It even found some price decreases in the B-segment.2 However, what eventually counts in cost control is the product of price and volume, because higher volume can compensate for lower prices. This sheds a different light on the results. For instance, Ikkersheim and Koolman (2013) reported higher volume growth in the B-segment than in the A-segment. The government’s task force on cost control underscored this observation by pointing to a substantial volume growth of relatively simple routine interventions. For instance, from 2006 to 2008 total expenditure on the 10% lowest priced DBCs had risen by 47.9%, whereas total expenditure on the 10% highest priced DBCs had risen by only 2.2% (Taskforce Beheersing Zorguitgaven, 2012). In its analysis of the effects of the market reform from 2009 to 2013, the Dutch Healthcare Authority reported that an average price effect of −1.6% had been offset by a 2.1% increase in costs per patient. The Authority suggested four potential explanations for the cost increase per patient: (a) changes in the demand for health care (for example, more multimorbidity because of the ageing of the population); (b) more treatments per patient; (c) more expensive treatments per patient; and (d) earlier active treatment of patients (NZa, 2013). 2 The distinction between A-segment and B-segment has disappeared since 2012. The market reform in Dutch health care120 With the exception of the first explanation, each of these suggests the risk of supply-induced demand. Besides, upcoding – defined by Simborg (1981) as “a deliberate and systematic shift in a hospital’s reported case mix in order to improve reimbursement” – might also be in play. Upcoding in Dutch health care has been studied by Hasaart (2011). After correcting for variations in case mix, Hasaart found remarkable differences in the degree to which hospitals had opted for more expensive treatments. She interpreted this empirical result as a “soft” indication of upcoding. In sum, we conclude that the reform constrained the increase in prices in the B-segment of hospital care from 2005 to 2012. However, its effect on total expenditure was more than offset by volume growth in the B-segment and higher prices in the A-segment. This effect has motivated insurers to negotiate volume ceilings and lump-sum payments. In section 6.7 we will see that these cost control instruments have coincided with a substantial volume drop in hospital care. 6.4.2 General practitioner care In Chapter 4 we saw that the market reform required the development of a uniform funding model of general practitioner care. The new model was the result of hard negotiations between the national associations of general practitioners and insurers. It consisted of three segments, including a capitation tariff for each registered patient; a long tariff list for consultations, support personnel and multidisciplinary care; and a set of performance-related indicators (see Chapter 4). General practitioners have benefited from the new scheme, because they are now paid for out-of-office hours and multidisciplinary care. There is also some evidence that patients ages 65 years and older with private health insurance started consulting their general practitioners more frequently than in the pre- reform era because consultations were free of charge (private health plans often required patients to pay or co-pay for a consultation). The volume of long visits, recurrent receipts for prescription medicines and some diagnostic services named “modernization and innovation” also increased. In terms of effect on total expenditure, volume increase outweighed the lower tariffs (van Dijk et al., 2013). An unexpected expenditure growth in 2006 and 2007 (due to flawed assumptions in the new funding model) was reason for the Minister of Health to revise the model of funding general practitioner care and skim off excess revenues seen in previous years. The national association was furious about this revision. Table 6.1 demonstrates that spending on general practitioner care rose faster than spending on hospital care in the periods 2010–2015 and 2015–2018. As spelled out in section 6.4, this result is not surprising given the emphasis on substitution Impact of the market reform on health care expenditures 121 of primary care for hospital care. The framework agreements included higher growth rates for general practitioner care than for hospital care. 6.4.3 Mental health care Until 2014 expenditure growth in mental health care was largely similar to the expenditure pattern found in hospital care (see Table 6.1). The sector benefited from the government’s spending programmes to reduce waiting lists after the turn of the century. From 2000 to 2005 expenditure growth in mental health care even outstripped expenditure growth in hospital care (12.4% versus 10.2%). As pointed out earlier, since 2010 expenditure growth in mental health care has been less than in hospital care and general practitioner care. This growth pattern is likely to be associated with the government’s policy to substitute general mental health care for some specialized care. There is some evidence that the average costs of general mental health care were lower (-€2132) than the costs of specialized mental health care (van Mens et al., 2018). However, this result must be interpreted with caution because of the absence of randomization of patients between the group with specialized care and the group with basic health care. Expenditure growth in mental health care may also have been constrained by the introduction of a separate co-payment regime in 2012. In section 6.8, however, we will see that there are some good reasons not to overstate this spending effect. A final explanatory factor is the purchasing policy of insurers (see section 6.6). 6.5 Expenditure on outpatient prescription medicines Expenditures on outpatient prescription medicines include the costs of medicines, fees for delivery services of pharmacies and value-added tax. Total expenditure rose from €4.4 billion in 2003 to €4.6 billion in 2017. In the same period annual per capita expenditure dropped from €269 to €267. Fig. 6.5 shows the per capita annual changes in expenditure. Several factors help to explain the fluctuating pattern. The significant drop in expenditures in 2012 and 2013 is a statistical effect. In 2012 hospitals were made responsible for expenditures on a number of expensive new medicines (TNF-alpha inhibitors, growth hormones, oncolytics and immunosuppressants). This change explains a great deal of the sharp decline in patients’ expenditures and the strong negative price effect in 2012 shown in Fig. 6.6. In the same year the patents on several widely used medicines ran out. Changes in the standard benefit package of health insurance, for instance, the lifting of the age restriction for contraceptives in 2008 and of limits on coverage of sleeping pills and tranquillizers in 2009 influenced expenditures as well. The market reform in Dutch health care122 The fluctuating expenditure pattern contrasts with the rise in the number of prescriptions and the defined daily dose (DDD) per prescription. For instance, from 2009 to 2011 expenditure climbed by 2.5%, whereas the number of prescriptions rose by 16.4%. Fig. 6.6 shows that in 2004 and since 2008 price decreases have partially counteracted the effect of volume growth on expenditures. Fig. 6.6 Price effects and volume effects of outpatient prescription medicines Eu ro s (m ill io ns ) -650 -450 -250 -50 150 350 550 20 04 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 20 14 20 15 20 16 20 17 price volume Note: The sharp price decrease in 2012 and 2013 are to a great extent the statistical effect of the transfer of high-cost innovative medicines to the hospital budget. Source: National Health Care Institute (www.gipdatabank.nl/servicepagina/gipeilingen). Fig. 6.5 Per capita expenditures on outpatient prescription medicines, absolute and percentage change, 2004–2017 Percentage -15 -10 -5 0 5 10 15 Eu ro s 0 50 100 150 200 250 300 350 20 04 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 20 14 20 15 20 16 20 17 per capita HCE % change Abbreviation: HCE, health care expenditure. Source: National Health Care Institute (www.gipdatabank.nl/servicepagina/gipeilingen). Impact of the market reform on health care expenditures 123 With the exception of 2012 and 2013, the price decreases can, to a great extent, be ascribed to the introduction of the preference strategy of insurers in 2004. This strategy can be considered a specific variant of selective contracting. Insurers were permitted to reimburse only the costs of the lowest priced generic medicines with the same effective chemical substance. In 2012 there followed an extension of the scope of the preference policy. Nowadays, insurers also directly contract with producers of generics, with the agreement that they are paid a portion of the profit made. In financial terms, the insurers’ preference strategy has proved successful. Average costs per DDD of three generic medicines dropped from €0.35 in 2004 to €0.24 in 2008, and €0.08 in the period from 2008 to 2013. Meanwhile, the average costs per DDD of non-preference medicines remained stable until 2008 but climbed to €0.60 in 2013 (ZiN, 2014). 6.6 Effectiveness of global spending constraints and increase of insurer risk Since the mid-1990s the government has constrained expenditure growth by means of a spending ceiling, or global budget (Budgettair Kader Zorg). This instrument has been its most important policy instrument for cost control. The fiscal rules held the Minister of Health responsible for budget overruns and required him or her to compensate for these by expenditure cuts or other cost-saving measures (Schakel, 2020). Fig. 6.7 presents an overview of annual overspending and underspending in the period 2000–2019. Fig. 6.7 Overspending and underspending as percentage of the global budget 20 00 20 01 20 02 20 03 20 04 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 20 14 20 15 20 16 20 17 20 18 20 19 Overspending +€14 billion Underspending –€14.5 billion 5 4 3 2 1 0 -1 -2 -3 -4 -5 Pe rc en ta ge Source: MoH (personal communication). Fig. 6.7 shows that spending ceilings did not seem to be very effective until 2012. With the exception of 2007, health care expenditure exceeded the ceiling every year. However, since 2013 the picture has reversed. Insurers felt committed The market reform in Dutch health care124 to the agreements on expenditure growth in the framework agreements and used them as reference points in their purchasing strategy (Chapter 4). The Netherlands Court of Audit concluded, on the basis of this result, that the framework agreements had proved an effective instrument to control health care expenditures (Algemene Rekenkamer, 2016). In our view this conclusion needs qualification because the accelerated phasing out of the arrangements for ex post risk equalization (discussed in Chapter 2) also played an important role. Knowing that they no longer would receive compensation for deficits, insurers were keen on cost control in negotiating contracts with health care providers. This effect was precisely what the accelerated phasing out of ex post risk equalization was intended to accomplish. Fig. 6.8 Overspending and underspending by type of health care Pe rc en ta ge -2 -1.5 -1 -0.5 0 0.5 1 1.5 2 2.5 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 hospital care mental health care GP care medicines Abbreviation: GP, general practitioner. Source: MoH (personal communication). Fig. 6.8 gives more detailed insight into the effectiveness of spending constraints by differentiating among types of health care. For the entire period, we found overspending in hospital care and significant underspending in medicines.3 The pattern varies for the other types of care, in particular for mental health care. This result sheds an interesting light on the effectiveness of the framework agreements. Underspending in outpatient prescription medicines, general practitioner care (2015) and mental health care (2014, 2016 and 2018) compensated for overspending in hospital care. An explanation of this result may be the market 3 There are no framework agreements on medicines. The bar shows the difference between the state’s global budget for medicines and actual spending. Impact of the market reform on health care expenditures 125 power of providers and insurers in contract negotiations: hospitals are able to exert more market power than other providers. 6.7 Production incentive? An often-heard criticism of activity-based funding models is the built-in production incentive. Activity-based funding enables providers to increase their revenues by raising the volume of care. The costs of their rent-seeking behaviour are shifted to the third-party payer. The problem of information asymmetry between doctor and patient means that many patients are unable to weigh the pros and cons of additional treatments. The assumed production incentive in a competition-based health care system has been the reason for the Dutch Healthcare Authority to call for a redesign of the funding of health care. In the Authority’s view, there is much evidence of such an incentive, and doing more does not necessarily mean doing better (Welch et al., 2011). Funding should encourage providers to render appropriate care and should pay for good outcomes instead of for volume. The Authority mentions multi-year contracts and bundled payments for care pathways as two instruments to eliminate the production incentive in health care funding (NZa, 2020b). The production incentive is an old problem in health care, including in Dutch health care. The purpose of fixed hospital budgets in the 1980s was actually an attempt to abolish the open-ended model of hospital funding that gave hospitals a financial interest in maintaining a high production level. The introduction of a yearly budget ceiling was meant to break the production incentive. However, it did so only partially because of the hospitals’ interest in avoiding underproduction, which entailed the risk of a downward budget adjustment in subsequent years. These examples make clear that a production incentive is certainly not an exclusive characteristic of a competition-based health care system. Nevertheless, the criticism of a built-in productive incentive is no surprise, because competition puts a premium on growth. This problem was one of the reasons that the former Central Healthcare Tariffs Agency criticized the report of the Dekker Committee. The Agency warned of the counterproductive effect of the model of regulated competition on health care expenses. The model would simply not reduce the growth of health care expenditures, as the Committee believed (section 2.2). The production incentive also played an important role in the political discussion on the extension of free pricing in hospital care after the Health Insurance Act became operative. As spelled out in subsection 2.6.1, the then Minister of Health held onto the extension of the scope of free pricing but set a policy agenda directed at care innovation to avoid the high costs of overtreatment (and undertreatment) The market reform in Dutch health care126 and to eliminate the production incentive in health care provision. What he called for was, in essence, the provision of appropriate care. Recent studies have, indeed, shown some evidence of a production incentive, specifically in those sectors that saw new entrants and a high penetration of non- contracted care. Douven et al. (2015) studied the effect of a new reimbursement schedule in mental health care (introduced in 2008), in which the reimbursement corresponds to treatment duration. Once the provider has passed a treatment duration threshold, the fee is more or less flat until the next threshold is reached. The schedule was used only for paying self-employed providers; psychiatric hospitals received a global budget. While self-employed providers had a financial interest in shifting duration time to over the next threshold, psychiatric hospitals did not. The researchers compared treatment duration of both types of providers and found that self-employed workers had shifted duration time to over the next threshold more often than salaried workers. The difference corresponded to a cost increase of approximately 7% to 9%.4 Another example is a study by van Dijk et al. (2013), which investigated the impact of the new reimbursement schedule introduced in 2006 on the provision of general practitioner care in 2005–2007. The new schedule consists of a mix of capitation and fee-for-service payments (see Chapter 4 for more information). Using a difference-in-difference approach, the researchers found that former sick fund patients had a 5.3% higher rate of physician-initiated contacts than former privately insured patients. They see the explanation of this supplier-induced effect in the extension of the reimbursement schedule with fee-for-service payments (in the old schedule general practitioners received only a capitation payment for each registered patient).5 There is also some evidence of a production incentive in community nursing. The introduction of competition in 2015 was followed by an increase in the number of small providers. Research journalists of the platform Follow the Money found several examples of high profits and self-enrichment among non- contracted providers. They detected a similar problem in the provision of social support services and youth care6 at the local level. These findings can certainly be considered excrescences of competition that influence the current political debate on the future of the market reform. This dark side of competition has led the government to modify the Health Care Institutions Accreditation Act by 4 As of 2014 the new activity-based schedule (by means of DBCs) also has been used for paying psychiatric hospitals. 5 Another finding of the study was that the abolition of cost-sharing led to an increase of patient-initiated utilization of general practitioner care by former privately insured patients ages 65 and older. 6 As of 2015 the purchase of youth care has been decentralized to municipalities on the assumption that municipalities can carry out this task much more efficiently than the state. Municipalities receive a state grant for youth care and bear the financial risk of their new task. Impact of the market reform on health care expenditures 127 introducing stricter criteria on quality and governance. Insurers, too, are taking measures to resolve the problem, for instance by means of selective contracting or administrative measures. However, there are counterforces too. Opponents argue that these measures hollow out the patients’ right of free choice. A recent study of the growth of spending on hospital care sheds a different light on the presence of a production incentive. The study made use of a new method to separate price effects and volume effects. It reported that the average volume growth in hospital care was substantially lower (1.1%) from 2013 to 2017 than in the period from 1998 to 2012 (3%, with spikes as high as 6%). Over the period from 2013 to 2017, the number of inpatient days dropped by 9.1%; the number of hospital admissions dropped by 19%; and the number of patient admissions dropped by 13.6%. These decreases took place despite the ageing of the population and the increased number of patients with multimorbidity. What made the study interesting, however, was that volume growth in hospital care from 2013 to 2017 could be explained entirely by the increased use of new expensive medicines (Chapter 4). Researchers measured a volume growth of 79%. The number of patients treated with these medicines doubled, from 126 000 in 2013 to 259 000 in 2017. In the same period, the price of these medicines fell by 17.6%, and the average costs per patient decreased from €11 300 to €8200 (Klijs & van Hilten, 2020). These results suggest that volume growth in hospital care in the research period was technology-driven and unlikely to be the effect of a production incentive – the more so because of the heavy financial burden of these medicines upon the hospitals’ budget and the risk of displacement effects in hospital care (Chapter 4). It would have been unacceptable for doctors to deny patients access to the new treatment options. 6.8 Effect of co-payments on health care spending From the very beginning, the Health Insurance Act has included a system of co-payments in the financing of health care as a way to discourage the utilization of unnecessary medical care (moral hazard). Until 2008 a no-claim system was in place. Those insured who were ages 18 years and older had to pay an extra premium of €255 on top of their nominal premium, which they were paid back in proportion to their health care consumption. The system proved a failure. As of 2008 health insurance legislation includes a mandatory deductible that has risen from €150 in 2008 to €385 in 2016. For political reasons, it has not been elevated further since 2016 (see Chapter 3). Patients under 18 years of age are exempted from the mandatory deductible as is general practitioner care, maternity care (including deliveries), community nursing and multidisciplinary care (“chain care”). Fig. 6.9 gives an overview of the size of the mandatory deductible and its fraction in health care spending covered by the Health Insurance Act. The market reform in Dutch health care128 Fig. 6.9 Amount of mandatory deductible and percentage of health care financing 0 1 2 3 4 5 6 7 8 9 0 50 100 150 200 250 300 350 400 450 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 mandatory deductible % health care financing Eu ro s Percentage Source: Own calculation. Although the fraction of the mandatory deductible in the financing of the Health Insurance Act is relatively small, it has always been heavily contested. While opponents consider it a financial obstacle that most harms patients with a low income and patients with chronic diseases, advocates defend it as a necessary instrument to discourage moral hazard. Without such a deductible, the user would falsely experience health care as costless. A former Minister of Health complained in this respect about the phenomenon that “too many people act as if health care were free. If you cancel the deductible, the nominal premium will rise every year by approximately €284. This is the true story” (Schippers, 2017b). The Bureau for Economic Policy Analysis (the leading economic advisory agency for the Dutch government) has estimated the cost-saving effect of the mandatory deductible at €0.8 billion. Abolition of the mandatory deductible would increase total public health care expenditures by €3.7 billion (later corrected to €4.7 billion). However, the estimation is “soft” because of the absence of a control group. It rests upon expert guesses of price elasticities. Unfortunately, this method is problematic because price elasticities are influenced by the point of departure,7 the type and structure of health care provision and the response of providers (Remmerswaal et al., 2015). What also complicates the estimation of the cost-saving effect is that providers may compensate for the effect of lower demand by starting extra activities (known as the “fill effect”) or that consumers may opt for another type of care exempted from the co-payment (known as the “waterbed effect” or “balloon effect”). In addition to the mandatory deductible, patients must co-pay for some specific health services. Empirical research concentrates on the effect of these co-payments 7 The demand effect of an increase in a co-payment from €5 to €10 is different from an increase from €150 to €300. Impact of the market reform on health care expenditures 129 on health care demand. For instance, Lambregts and van Vliet (2018) found that the introduction of a co-payment of €200 for specialized ambulatory mental health care for adults in 2012 was followed by a drop of 35% in the demand for this type of care. The drop was greatest for the short-duration treatment of vague complaints. Providers responded to this drop by raising treatment duration to compensate for the loss of revenues. This is another example of the fill effect. Other research found that the introduction of a €200 co-payment was not only associated with a drop in demand, in particular in the lower-income categories, but also followed by an increased number of crisis interventions or compulsory admissions, causing extra costs at a later time (Ravesteijn et al., 2017). This is an example of a waterbed effect. 6.9 Administrative costs Administrative costs is a recurring topic in the debate about the market reform and health care governance in general. Critics argue that the reform has been accompanied by a significant increase in administrative costs, taking money away from caring for patients and causing a lot of “frustration with administrative burden and insurance hassle” (Larjow, 2018). The measurement of administrative costs is complicated because of the ambiguity of the concept. How to demarcate administrative costs from other costs? How much should an organization spend on administrative tasks (overhead)? Is there a norm for this? Research shows disagreement among experts on how to answer this question. A related problem is that administrative costs are incurred not only at the national level in health care, but also at lower levels (for example, the regional level) and in particular at the micro level of health care provision. It is particularly at the micro level where administrative costs cause frustration. Care workers experience the administrative burden as a time-consuming activity that reduces time for patient care (Hagenaars, 2021). High administrative costs constituted a reason for general practitioners to urge a shift from the contract model in health care to the professional model based on trust. Given the ambiguity and multilevel structure of the concept of administrative costs, data on its size must be interpreted with caution. Reliable data do not exist. A rough estimate of administrative costs of hospitals in the Netherlands is that they accounted for 18–20% of total hospital costs in the period 2005– 2017 (calculated from Dutch Hospital Data). An international comparison of hospitals’ administrative costs (including Canada, England, France, Germany, Netherlands, Scotland, United States of America and Wales) found that the Netherlands ranked second highest in administrative costs (Himmelstein et al., 2014). However, these figures do not include the time that professionals The market reform in Dutch health care130 themselves spend on such activities, and neither do they include the expenses of the different regulators and administrative bodies.8 Comparative international research of the Commonwealth Fund shows that the perceived administrative burden, measured by the time spent on administrative issues and on claiming payments, is highest among general practitioners in the Netherlands (Fig. 6.10). There are good reasons to argue that the market reform has pushed up admin- istrative costs. Contract negotiations with multiple insurers, complex regulations, procurement procedures, activity-based funding models and recurrent revisions of these models, complex accounting procedures, risk reduction, supervision and the detection of inappropriate care or fraud are often mentioned as factors pushing up administrative costs. However, it would be a mistake to see the reform as the sole cause of higher administrative costs. The rise of these costs is also closely associated with the increased emphasis on quality of care. The enhancement of quality of care requires extensive guidelines (standards) for good practice. Quality of care has become something that needs to be checked by outcome measurement. A great deal of the quality indicators currently used have been produced by professional and client organizations. In other words, the time-consuming bureaucratization of health care is not the result of the 8 The problem of high administrative costs is also present in long-term care. A recent survey found that in 2019 health care workers spent on average 35% of their working time on administrative tasks. The highest percentage was found in mental health care (40%) and the lowest percentage in care for people with a disability (33%). The researchers also measured an increase in administrative costs from 25% in 2016 to 31% in 2018 (Hanekamp et al., 2019). Fig. 6.10 Percentage of time spent by general practitioners on administrative issues/claiming payments Ca na da Ne w Ze ala nd Au str ali a Un ite d K ing do m Sw ed en Sw itz erl an d Ge rm an y Un ite d S tat es Ne the rla nd s No rw ay 70 60 50 40 30 20 10 0 Source: Commonwealth Fund, 2015. Impact of the market reform on health care expenditures 131 market reform only, but rather of broader developments in health care (Zegers et al., 2020). 6.10 Discussion and policy lessons This chapter presented an overview of the impact of the market reform on health care expenditure. What lessons can be drawn from the Dutch experience? Unfortunately, there is no simple answer to this question. Two reasons are the absence of a control group and that the effects of the reform cannot be well isolated from the financial impact of other policy measures taken during the reform. However, since 2012 cost increases in the Netherlands are substantially lower than those in other countries. Such data do not include the strong improvement in the balance sheets of both health care insurers and providers. In recent years health care growth was even slower than GDP growth, a historic anomaly. Most sectors witnessed structural underspending (hospital care was an exception). What was behind this slow growth? Was it “price” or was it “volume”? In this chapter we found two examples of a price effect. Before 2012, price increases in the free-pricing segment of hospital care (B-segment) were lower than in the regulated-prices segment (A-segment). However, higher volumes compensated for the cost-saving effect of lower prices. A study by a task force on cost control found that expenditures on the 10% lowest priced DBCs had risen much faster (47.9%) than expenditures on the 10% highest priced DBCs (2.2%). These results suggest that the fill effect cancelled out (in part) the price effect of competition on total expenditures. The second example relates to generic outpatient prescription medicines. The cost-saving effect of price competition was much stronger than the volume effect on expenditures. Our conclusion is that the preference policy of insurers in purchasing outpatient prescription medicines has so far proven to be one of the most effective instruments in controlling health care spending. By successfully negotiating lower prices, insurers were able to absorb a great deal of the cost-raising effect of volume growth. However, as said above, the effect of price competition on expenditure growth cannot be overstated. The very moderate expenditure growth is, in our view, at least to a great extent the result of two other factors. First, the framework agreements among government, insurers and providers have proved very effective from a cost control perspective. Insurers have used these frameworks as points of reference in negotiating contracts with providers. The frequent use of volume ceilings or lump sums was an effective instrument to constrain costs and counteract a great deal of the production incentive in health care. This slowing of volume growth turned out to be the most prominent determinant behind the slowing growth of expenses from 2012 onwards. The second factor, and the reason why insurers were so keen to put more financial pressure on providers, was the The market reform in Dutch health care132 accelerated termination of ex post compensation mechanisms (subsection 2.6.4), which gave insurers a strong incentive to keep their expenses in check. Price competition in the insurance market reinforced this incentive. Nevertheless, it is important to note that strict budget constraints are not risk free because they may be set artificially low and miss long-term credibility. The ultimate test of their effectiveness also depends on whether they incentivize insurers and providers to render value-based health care and prevent long waiting lists and provider inertia. If budget constraints fail to improve allocative and technical efficiency by a dedicated focus on appropriate care and instead mainly result in expenditure cuts by salami slicing, their cost-saving effect will eventually turn out to be only temporary, and a new “cost explosion” will be imminent. More empirical research on the cost impact of the reform is necessary. In particular, more research is needed to gain insight into the cost-effectiveness of specific innovations. The ultimate success of the reform does not depend on its impact on total expenditure, but rather on the extent to which insurers and providers agree on innovative programmes that simultaneously save lives and costs. Chapter 7 Critical assessment of the market reform and its future 7.1 Introduction The purpose of this study is to gain insight into the theory and practice of the market reform in Dutch health care. The main policy goals of the reform were to achieve a health care system offering high-quality care to patients that is accessible to each person (universal access), based upon solidarity and affordable from a financial perspective (financial sustainability). This chapter starts with a brief summary of the market reform in the Netherlands (section 7.2). This overview will be followed by two important observations. The first observation is that the regulated competition analysed in this book is the Dutch version of regulated competition (section 7.3). Secondly, we will argue that the term “market reform” is actually a misleading concept to depict the reform of Dutch health care (section 7.4). The critical role of insurers is the topic of discussion in section 7.5. Section 7.6 calls for a reassertion of the role of the state. The chapter ends with some observations on the impact of the COVID-19 pandemic on the discussion of the future of the market reform (section 7.7) and a few final conclusions (section 7.8). In our view two main issues stand out in the political discussion of its future: the position of insurers and the call for reasserting the governance role of the state in health care. These issues are closely connected. 7.2 Overview of the market reform This study addresses the introduction of regulated competition in Dutch health care. This reform – also referred to as market reform – became operative in 2006 with the introduction of the Health Insurance Act. The model of regulated competition rests upon the assumption that it is possible, through strict regulations, to reconcile the logic of the market with the public interests in universal access, quality and financial sustainability in health care. Respect for The market reform in Dutch health care134 this normative legacy of the past was an absolute precondition for the political feasibility of the reform. The enactment of the Health Insurance Act in 2006, after almost 20 years of political debate, put an end to the traditional divide in the Dutch health insurance landscape by integrating the sick fund scheme and other (mainly private) health insurance schemes into a single mandatory basic scheme covering the entire population. The new scheme had to be carried out by private insurers competing for clients on the health insurance market. Other central elements of reform were: freedom of choice, open enrolment, a comprehensive standard benefit package, a mandatory deductible (€385 since 2016), risk solidarity supported by a complex system of risk equalization, and income solidarity by compensating low-income subscribers. The scheme is financed by nominal premiums set, to spur competition, by each individual insurer, uniform income-related premiums set by the state and a state grant to cover the premium for children. The introduction of the new health insurance scheme constituted the first major step in the reform process. It established an alternative institutional structure with new relationships between state, providers, insurers and the population, who in the economic jargon of the reform are referred to as consumers. Since then, various other market-making decisions followed to complete the reform, such as the gradual extension of the room for free pricing, the termination of a great deal of government oversight of hospital planning and the devolution of financial risk to insurers and providers. The state was put at a distance from the daily practice of health care. Policy documents described its new role in terms of “system responsibility”. A great deal of regulatory and oversight activity was delegated to the Dutch Healthcare Authority, which formally constituted an independent agency at arm’s length from the state. The Health Care Market Regulation Act, which came into force in 2006, established a delicate balance between the formal role and responsibility of the state and the Healthcare Authority. In 2012 the market reform was complemented with a series of collective framework agreements on net annual expenditure growth and other policy issues. These agreements, signed by the Minister of Health and the top organizations of providers and insurers, have no legal basis and can be understood as informal instruments on top of the formal regulatory framework for competition. They reflect the corporatist tradition in the governance of Dutch health care and controlling expenditure growth. Chapter 3 presented an overview of the consequences of the reform for the health insurance market. An immediate effect of the integration of social and private health insurance into a single scheme was a wave of mergers and takeovers in health insurance. The number of insurers dropped from 58 in 2005 to 33 in 2006, and to 24 in 2019. Four insurance groups have a total market share of Critical assessment of the market reform and its future 135 85.9%. To date, only two new insurers have entered the Dutch market by means of takeovers. In 2018 consumers had a choice of 59 health plans. However, this figure is somewhat misleading because 45 plans turned out to be more or less identical. A new development is the increased popularity of managed care plans (“budget plans”), which give policyholders a premium reduction in exchange for a restricted choice of providers. In 2019, 14.1% of the insured had such a plan, compared with only 1.1% in 2011. In addition, more people have chosen a voluntary deductible, while the numbers with supplementary coverage have declined. As a result, a larger portion (but still a minority) of the Dutch population faces more financial risk than in the pre-reform period. After a surprising 18% of the population switched to another insurer in 2006, the switching rate dropped to about 6–7% a year since 2012. Research has indicated that, so far, persons ages 18–39 years, persons with higher education and persons perceiving their health to be good have benefited the most from their enhanced freedom of choice. The new health insurance scheme includes various regulations to avert risk selection and preserve risk solidarity: Insurers must accept each applicant, risk rating is forbidden and a sophisticated system of risk equalization compensates insurers for differences in their risk profiles. Notwithstanding these regulations, there are indications of some risk selection. Flaws in risk equalization, voluntary deductibles and health plans with restrictive conditions may cause risk selection. So far, regulated competition has not proved to be the ultimate solution to prevent future erosion of solidarity and risk selection. Chapter 4 analysed the practice of purchasing. The ultimate success of the reform largely depends on what insurers make of their agency role on behalf of their clients. There is no one-size-fits-all approach to purchasing. Insurers make use of different models: the public tender model (for example, for outpatient prescription medicines); the standard-contract model (for example, with general practitioners); and the negotiated-contract model (for example, with hospitals). Purchasing health care has proved to be an evolving process. In the beginning insurers not only had little experience with purchasing, but also grappled with a fundamental information problem regarding the costs and quality of care. The validity of much quality information is still questionable. Insurers have learned not to overstate their influence via purchasing as well as the effectiveness of the instrument of selective contracting. In practice, purchasing largely boils down to a mixed motive game in which insurers and providers have a common interest to strike a deal. Nevertheless, contract negotiations on costs and volume (ceilings) can be hard and evoke much frustration on the providers’ side. Purchasing is gradually moving away from a competitive style of negotiation with a one-sided focus on costs towards a more integrative model of purchasing with multi-year The market reform in Dutch health care136 contracts and attention to care innovation. An important lesson from the Dutch experience is that purchasing is doomed to fail without the powerful incentives of financial risk-bearing to give insurers a strategic interest in purchasing. Partly as a result of demands from regulators, insurers have built strong balance sheets to cope with financial risk. Chapter 5 offered an analysis of the influence of the reform on the provision of health care. A remarkable development was the explosive increase in the number of ITCs, which focus their activities on specific patient groups, specific treatments or specific diseases. In some specialties (ophthalmology and dermatology), ITCs have acquired substantial market share. Still, the share of ITCs in total health care expenditure has remained limited, at 3.8% in 2016. This percentage indicates that hospitals have maintained their dominant position in the provision of specialist medical care. Since 2006 four hospitals have gone bankrupt. Two bankruptcies in 2018 caused much public anger and prompted a political debate on the social acceptability of hospital bankruptcies and the system responsibility of the state in such cases. The role of price competition in purchasing should not be overstated. Practice differs from theory. Price competition exists only if insurers organize a public tender with price as one of the discriminating criteria. When negotiating price– volume contracts, insurers may exploit their power to extract prices under the average market price. If insurers and providers negotiate lump-sum or multi-year contracts, prices are largely artificial. We found no clear evidence for a positive impact of competition on the clinical quality of health care. The devolution of financial risk has made hospitals more risk aware. This development helps to explain the decline in investments in major construction works. Hospitals have managed to reinforce their financial performance and build substantial buffers to cope with financial uncertainty, but the spread around the average performance appears to be considerable. Chapter 6 addressed the question of whether competition has been an effective instrument for cost control. Expenditure growth has been remarkably moderate since 2012 and, in contrast with previous years, total expenditures have not overrun the global budget for health but seen persistent underspending. This result is likely to be the effect of the framework agreements on annual net expenditure growth and insurers’ increased exposure to risk. This largely boils down to lower volume growth. Thus, the effect of price competition on expenditure growth should not be overstated, with the exception of its role in the purchase of outpatient prescription medicines (the so-called preference policy of insurers). Critical assessment of the market reform and its future 137 7.3 The Dutch version of regulated competition Regulated competition can be conceptualized as an alternative institutional model for a state-directed type of health care system. The model rests upon the assumption that competition, provided it is well regulated to achieve fair competition and preserve public values in health care, offers the best institutional guarantee for an efficient allocation of scarce resources (Enthoven, 1993). However, maximum value for money is realized only if a set of preconditions regarding freedom of choice, consumer information, transparency, risk-bearing buyers and sellers, quality supervision and some other conditions are met (van de Ven et al., 2013). Regulated competition is an abstract and reductionist model that policy-makers must convert into concrete regulations. How the model really works is contingent on the “translation process” and on how the state, consumers, insurers and providers deal with the regulations in practice. With its focus on institutional issues, the model is also rather silent about substantive policy issues. For instance, policy-makers must determine the concrete list of health services covered or exempted from coverage. The model does not prescribe a single national health insurance scheme, nor how much a nation should spend on health care, nor how to shape the balance between risk solidarity and freedom of choice. Other examples of “open ends” in the model are the scope of the duty of care of insurers, the role of the state if the continuity of care is at risk, the number of parameters in risk equalization, the fraction of co-payments in health care financing and the design of an implementation strategy. One key central message here is that this book has examined the Dutch version of regulated competition. This version cannot be well understood without taking the political and historical context of Dutch health care into consideration. In each stage of the reform, political compromises were needed to adapt the reform to the preferences of stakeholders and to win a political majority. Although the reform entailed many institutional changes, it also drew upon the foundations of the pre-reform system that had been institutionalized over a long period. If the reform had not respected the principles of risk solidarity and income solidarity, it would not have been politically feasible. The practice of collective framework agreements to control health care expenditures built upon the institutionalized corporatist pattern of shared responsibility. The devolution of purchasing and financial risks to insurers and providers rested not only upon the principles of regulated competition, but also upon the traditional Christian body of thought of subsidiarity, according to which the financing and provision of health care had to be delegated as much as possible to organizations with a social purpose. Actually, the reform was well suited to the traditional public–private mix in Dutch health care. The market reform in Dutch health care138 This key point contains an important policy lesson: Be cautious when drawing lessons from the experience with health reforms in other countries. Each reform must somehow take into account the legacy of the past, and each reform involves political compromises to get necessary support. What works in Dutch health care may not work in another country, and what does not work in Dutch health care may work in another country. 7.4 Market or quasi-market? As spelled out in Chapter 1, it is common to frame the reform of Dutch medical care as a market reform. In this book we have adopted this terminology. Now that we have more insight into the concrete shape of the reform and how the reform plays out in practice, the question arises whether it is indeed correct to speak about “market reform”. Could it be that the terminology is misleading? For an answer to this question, it is helpful to make a distinction between the inner and outer circles of health care. In the inner circle, doctors and nurses care for patients, consumers seek health insurance and insurers pay for health care. This is the core of any health care system. The outer circle comprises all other activities: purchase of materials and technical equipment for medical care, ICT, relations with financial agents, construction activities, public security, cleaning services and so on. Competition in this outer circle is standard. Criticism about competition in health care most of the time is not directed at competition in the outer circle, with one important exception: the market behaviour of pharmaceutical companies that gives priority to shareholder value over broad accessibility of their medicines. However, what about the role of competition in the inner circle? For health insurance, there is good reason to consider the new system as a market system. Insurers compete for clients. The premium they must pay depends on the health plan chosen. Although four insurance groups dominate the market, the market is competitive. Insurers are aware of the sensitivity of consumers to increased premiums. Consumers can choose from a variety of health plans. At the same time, the health insurance market is heavily regulated. The purchase of a basic health plan is mandated; insurers must accept each applicant; regulation bans risk selection and risk-related premiums; the benefit package is uniform for all health plans; a sophisticated equalization scheme is in place to compensate insurers for poor risks; low-income individuals can apply for a care allowance to pay their premium; insurers have a duty of care; and so on. None of these regulations to protect consumer rights are common in car insurance, liability insurance or many other private insurance schemes. In sum, the Dutch health insurance market is indeed a market – however, a market with quite a peculiar structure to reconcile private interests with public interests. Critical assessment of the market reform and its future 139 The situation is different in health care provision. Theoretically, patients have much freedom of choice, but the practical role of this freedom should not be overstated. Patients usually visit the nearest hospital. Most people do not switch to another general practitioner. Quality information has no or hardly any influence on the choices that patients make. Due to the extensive coverage of health insurance, patients are not interested in prices. Insurers contract with the great majority of providers by means of a negotiated contract or a standard contract. The scope of selective contracting appears quite restricted in practice. Public tenders with price competition exist only in some specific areas, typically when no patients are directly involved, such as in pharmaceutical care. Collective contracting is on the rise. Some services, including ambulance services, have been exempted from competition. For-profit hospital care is forbidden. Billing patients extra for private treatment or treatment by a specific doctor is forbidden also. In sum, we find a big discrepancy between the theory and practice of regulated competition. In the Dutch version of this model, competition that goes beyond the normal rivalry between providers in the inner circle exists mainly in health insurance. The big discrepancy between the theory and practice of regulated competition leads us to conclude that the terms “market reform” and “competition” are actually misleading labels. In many respects Dutch health care resembles a quasi-market at best. A question that comes up is: Why is it that the term competition can still be frequently heard in discussions about Dutch health care? Why do opponents use this misleading terminology? Why is competition an ideal scapegoat in discussions about health care? The reason is largely political. Competition appears to be an appealing container concept for almost everything that is considered wrong in health care. It is an effective verbal instrument in the ideological struggle about the shape of Dutch health care. 7.5 The critical role of insurers Purchasing was described in Chapter 4 as the cornerstone of the market reform. The reform positioned insurers as a countervailing power at the meso-level in health care, where insurers and providers interact with each other. The upgrading of the insurers’ role rested upon several policy assumptions – first, that insurers are better informed than the state about the performance of providers and, second, that delegation of purchasing is an effective instrument to depoliticize the allocation of scarce resources. The state is more subject to political pressure than insurers; delegation of purchasing is a precondition for variety and innovation and for freedom of choice. The market reform in Dutch health care140 If purchasing is the cornerstone of the reform, the success of the reform is contingent on what insurers make of it in practice. We have seen that there are several reasons to understand purchasing as a complex activity and that it would be naive to expect sweeping results from it. Lack of timely, accessible, independent and accurate quality information, the structure of regional markets and insurers’ fear of a damaged reputation have been complicating factors from the very beginning. The steering capacity of insurers should not be overstated. Various initiatives to reorganize the regional landscape or encourage the cooperation of providers in regional networks were started but have failed in practice. Also, purchasing confronts insurers with a fundamental dilemma. They are expected to keep costs in check but also to encourage care innovation and quality of care. Will investments in care innovation pay off if switchers are mainly looking for lower premiums and patients want the best care available irrespective of costs? Insurers can be best conceptualized as “split” organizations with two missions that cannot be easily reconciled. These divergent missions are also visible in the gap between the beliefs of top insurance company administrators and the experience of its front-line purchasers. While the purchasers said that they have little real-world potential to create a distinct profile of themselves on quality, top administrators had an opposite opinion (Stolper et al., 2019). The agency role of insurers has been criticized from the very beginning of the reform. Critics consider it a serious flaw to charge private insurers with the allocation of billions of euros of public money, even though they are bound by strict regulations, such as the standard benefit package and the duty of care. This is an important reason to restore the role of the state (see section 7.6). Another criticism is that insurers are assumed to act as profit-driven juggernauts imposing squeezing contracts on providers. This criticism neglects the not-for-profit status of most insurers. Furthermore, we have seen in section 4.11 that there is good reason not to overstate the purchasing power of insurers and the effectiveness of the instrument of selective contracting in particular. A third main criticism is that purchasing by contracts hollows out the professional autonomy of health care providers. 7.5.1 From the contract model back to the professional model? Tuohy (2003) conceptualizes the move to regulated competition as the transition from the professional model to the contract model in health care. In the professional model, providers are in the lead. Doctors are trusted as expert agents who possess discretionary knowledge and a sheltered position that is based on qualifying credentials. They show greater commitment to doing good than to economic gain and to quality rather than to the economic efficiency of work (Freidson, 2001:  127). Accountability is organized internally by peer group Critical assessment of the market reform and its future 141 control. The role of insurers (or the state) is restricted to that of third-party payer. In the contract model, the relationship between doctors and insurers has a different structure. Insurers no longer simply pay the bill but instead contract with providers for the provision of health care on behalf of their clients (agency role). Contracts regulate the relationship between payer and provider. Termination or denial of a contract is a real possibility. The medical profession has always contested the contract model. Doctors and nurses perceive it as an evisceration of their professional autonomy that affects the doctor–patient relationship. The “logic of the market” is seen as antithetical to “the logic of health care” (Mol, 2008), because it replaces the trust-based relationship between doctor and patient with a contract-based type of relationship (Freidson, 2001). Providers with private practices experience contracting as a farce that leaves them no choice but to tick the box. Other complaints are the costly information and accountability obligations that may vary per insurer. Each insurer has its own pre-authorization rules for specific medical interventions. The contract model goes hand in hand with the bureaucratization of health care. In their manifesto “Need for redirection”, published in 2015, general practitioners urged less bureaucratic control and a level playing field in contracting. In essence, the manifesto called for trust in the general practitioner as a professional. The manifesto drew much public attention. Similar critical comments are voiced with respect to the purchasing role of municipalities in long-term care. Provider organizations express deep concerns about the municipalities’ implementation strategy, their lack of expertise and exclusive interest in cost saving. Critics fear skimping on quality and a race to the bottom. That purchasers are keen on costs is beyond doubt. However, this is one of their intended roles. A former Minister of Health (2010–2017) said about insurers, “In every system you need a police officer who looks at what it costs and what it yields and in our system health insurers play this role” (Schippers, 2017b). That the medical profession does not appreciate this crucial role is no surprise. It has always constituted a guild system accustomed to running its own affairs and averse to interference from the outside. How the relationship between insurers and health professionals will evolve in the future is difficult to predict. Experiments with an integrative style of purchasing as an alternative to a competitive style could help to restore mutual trust. Health professionals must understand that the old times will not return and that health care must undergo a radical change to meet future challenges. They should also realize that a take-back-control strategy of the state may not be in their interest. Health care innovation flourishes with bottom-up initiatives. For instance, one may wonder whether the strategy of alliance hospitals in narrow networks or a 10-year contract with a hospital with the provision that the hospital The market reform in Dutch health care142 will cut its expenses in successive steps would ever have been possible under a centralist regime. Top-down innovation orchestrated by the state runs the risk of bureaucratization. The basic principles of lawfulness and legal certainty will inevitably restrict variation and experimentation. Innovation would also run the risk of being compromised with political considerations. Another option would be a model in which the state formulates the national health agenda but leaves it largely to the meso-level in health care to fill in this agenda with innovative action. 7.5.2 Low institutional trust The agency role of insurers is also put under pressure by low institutional trust. Trust can be conceptualized as the expectation of consumers that their insurer will avoid choices that, while attractive to themselves, would hurt the interests of their insured. Trust is meaningless without specifying the object of trust. Who or what is trusted or distrusted? In this respect, it is necessary to distinguish between trust in one’s own insurer (specific trust) and trust in health insurance (institutional trust). Specific trust may differ from institutional trust. An insured individual may trust his or her insurer and yet distrust health insurance as a system. Trust is a matter of perception (Maarse & Jeurissen, 2019). Advocates of regulated competition implicitly assume that consumers consider their insurer to be their trustworthy partner. This assumption seems doubtful. An early indication of a potential trust problem was the finding that the percentage of respondents willing to follow their insurer’s recommendation on preferred providers had declined from 50% in 2005 to 25% in 2007. Insurers faced a “credible commitment problem” (Boonen & Schut, 2011). The Authority for Consumers and Markets also found indications of an institutional trust problem. Asked whether insurers serve the interest of their customers, 38% of the respondents in a survey expressed a low level of trust, whereas only 17% said they had a high level of trust, and the remaining 45% were neutral. Another finding was that 39% of the respondents who had switched to another health insurer in 2017 said they had little trust in health insurers. The Authority also reported that the percentage of “low trust” respondents had increased from 32% in 2014 to 40% in 2015 (ACM, 2017c). When asked for their reasons for a low level of trust, respondents said that insurers had raised their premiums even though they had made a significant surplus in previous years. They also believed that insurers could restrict the patients’ freedom of choice and that insurers did not always reimburse the full costs of medical treatment (for example, the costs of consulting a non-contracted provider). In sum, they believed that insurers did not serve their interests. A recent study by Groenewegen et al. found remarkable differences in trust in care providers and health insurers (see Fig. 7.1). While 80–90% of the respondents Critical assessment of the market reform and its future 143 said that they trusted health care providers, only some 30% considered insurers to be trustworthy. Institutional trust had slightly decreased since 2012 (the first year of data collection). Fig. 7.1 Percentage of people who trusted in providers and insurers 100 90 80 70 60 50 40 30 20 10 0 2006 2008 2010 2012 2014 2016 General practitioners Medical specialists Physiotherapists Pharmacists — Dentists Insurers Source: Groenewegen et al., 2019. Providers, too, show little trust in health insurers. In 2016 only 6% of general practitioners said that they trusted insurers. The scores of medical specialists, physiotherapists, pharmacists and dentists were 5%, 7%, 2% and 6%, respectively. Conversely, the scores of insurers’ trust in providers were much higher: 74% for general practitioners, 86% for medical specialists, 53% for physiotherapists, 66% for pharmacists and 56% for dentists (Groenewegen et al., 2019).1 Low institutional trust is a systemic risk factor. Eventually, no system can function properly without wide public support. Low institutional trust also fuels the political debate on the pros and cons of regulated competition and the need for fundamental redirection. Whatever their strategy, insurers will always grapple with their fate – the “coalition” between doctors and patients appears much stronger than the “coalition” between insurers and the insured. 1 Interestingly, specific trust scores were much better. One study found that trust in one’s own insurer averaged at 7.3% compared with 5.9% for trust in insurers in general. Another study reported that 58% of the respondents said that they trusted their own insurer, while only 28% said that they trusted health insurers in general. The difference between specific and institutional trust suggests that low institutional trust is based more on global impressions than on personal experience with one’s own insurer. In this respect, it is worth recalling that satisfaction with one’s own insurer has always been mentioned as an important argument for non-switching. The market reform in Dutch health care144 7.6 Reasserting the steering role of the state? How will the Dutch health care system develop in the future? Will regulated competition survive, and which factors will influence its further development? In answering these questions, it is helpful to distinguish between incremental adjustments intended to repair flaws and changes that may fundamentally alter the structure of health care. A number of remedial adjustments are underway. A great number of them are directed at the sectors of community nursing, ambulant mental health care and home care. In each of these sectors, many small providers are active. There are not only concerns about the quality of health services and the internal governance structure of some providers, but also indications of outright fraud. Some providers have made excessive profits. The government is determined to tackle these problems with new legislation to assure that fraudulent providers are denied access (by means of tighter licencing criteria) and that health care money is indeed spent on health care (MoH, 2018b). However, the current political debate on the future of Dutch health care goes beyond the need for remedial adjustments to optimize the current system. The system itself has become a topic of discussion. Two lines of thought dominate this discussion: the need for cooperation and the need for a more directive role of the state in health care. More market seems no longer a serious option. Possibly, Dutch health care is approaching the end of the policy cycle of the market reform. There is a growing belief that cooperation in regional provider networks and care pathways will yield better quality and lower costs. Cooperation has replaced competition as the buzz word in health care and is the leading concept in the programme The Right Care at the Right Place (De Juiste Zorg op de Juiste Plek) that was launched in 2018 by a task force with a broad membership under the chairpersonship of the Ministry of Health. The key question is how to encourage cooperation and prevent it from getting stuck in the swamp of private interests and mutual distrust. In this respect, one should not forget that coordination is not per se at odds with competition. The approach of the Authority for Consumers and Markets has always been to permit cooperation on the condition that it is in the interests of consumers. A second development is the emergent disbelief in the merits of competition for the production of public goods and services. The support for the neoliberal policy paradigm of regulated competition as a leading principle is waning in the current political climate. “System responsibility” for health care should go far beyond the narrow interpretation of this concept in the model of regulated competition. What is needed, the argument goes, is a reassertion of the role of the state in health care. Voices on the left side of the political spectrum have called for Critical assessment of the market reform and its future 145 the introduction of a new national health insurance scheme – dubbed National Health Fund. The new fund is proposed as a truly public scheme, without private insurers, covering the entire population. It is considered a serious system flaw to entrust private insurers with the allocation of billions of euros, even though they are bound by many restrictions such as the standard benefit package and duty of care. Other radical proposals are abolition of the mandatory deductible and replacing the nominal premium with largely income-related contributions. Realization of these proposals would be tantamount to restoration of the former sick fund scheme. The government has announced that it will send a consultation paper to the parliament with a rough sketch of how the current system should be adapted to cope effectively with the ageing population and the looming problem of labour shortages. In an interview with a Dutch newspaper, the Prime Minister declared that “there is a need for more central coordination and direction, often also with the visible hand of the state. The current organization of health care does not work anymore” (EW, 2020). The call for a reassertion of the state role in health care and the waning belief in competition are not isolated phenomena but instead fit a broader trend in Dutch public policy-making. For instance, the government recently decided to concede Dutch Railways the use of the main railway network through a private procedure for the next 15 years. To create stability and foster quality, it deliberately abstained from public procurement. Another example is the recent approval of a merger between two postal services despite a negative decision of the Authority for Consumers and Markets, which feared the creation of a monopoly in postal services. The responsible minister spoke of a “unique case” in a “unique market” as motivation for her decision. Similar developments have taken place in health care. An example is the recent decision to exempt ambulance transport from competition. Ambulance services are considered a public service of general interest. In the new legislation, the Minister will assign one provider per region the exclusive right and duty to provide ambulance services. No new entrants will be permitted. The plan also involves a return to the classic representation model in health insurance: The two largest insurers in the region are requested to contract for ambulance services on behalf of all insurers, but tariffs will be set by the Dutch Healthcare Authority. The current Minister of Health formulated the need for a more directive role of the state in health care in an interview as follows: “The belief in the market as a problem solver is waning … Competition in health care is overstretched … Competition in health care has become more important than cooperation. Freedom of choice has become so absolute that we do not see the price we have to pay for it. This has led to a fragmented health care system and loss of quality The market reform in Dutch health care146 as well” (Algemeen Dagblad, 2019). Later in the interview, he postulated that without more cooperation health care will become unsustainable, leaving only two options: deep cuts in the standard benefit package of health insurance and long-term care or raising co-payments. Both alternatives inevitably erode solidarity in health care. The recognition that regulated competition may not yield an optimal allocation of scarce resources is also the thread in a recent report of the Council for Public Health and Society (Raad voor Volksgezondheid en Samenleving, RVS) about the organization of acute care in some regions. The Council advocated more regional cooperation and, if necessary, central direction to ensure universal access to acute care. Concretely, it recommended moving decision-making about the organization of acute care from the negotiating table of providers and insurers to a regional decision-making centre in which all regional stakeholders participate under the chairpersonship of an independent person appointed by the Minister of Health. The “commissioner of acute care” should be authorized to inform the Minister if access to acute care is at risk. This direct line gives the Minister leverage to intervene (RVS, 2020). The need for a direct line is also what the Minister wanted in the aforementioned interview: “I want to know who in the region I can address” if regional cooperation stagnates (Algemeen Dagblad, 2020). How the call for system change will evolve remains to be seen. There are strong countervailing forces and political discussions on reform that run the risk of paralysing decision-making for a long time. Opponents of a “return to the past” warn against high expectations of a new reform. Was not the lesson of the pre- reform period that a state-dominated health care system led to public failures (bureaucratization and lack of dynamic efficiency and innovation). Despite this scepticism, the pendulum in health care reform is likely to swing back from regulated competition to a more central direction. The big question is how much? Another question is how much the political discussion on reform will be influenced by the COVID-19 pandemic. 7.7 The COVID-19 pandemic and regulated competition The COVID-19 pandemic (see Box 7.1) had several consequences for the institutional structure of public health policy-making. One of the most striking consequences was the centralization of policy-making. The Prime Minister, the Minister of Health and the Minister of Justice and Security took the lead in managing the crisis. The Prime Minister and the Minister of Health held weekly presentations (broadcast on TV) to inform the public of the policy measures taken to confine the spread of the virus and, in a later stage, the road map to terminate the lockdown. Every week the parliament discussed the government’s strategy. Initially, these discussions were mainly informational, but in the course Critical assessment of the market reform and its future 147 of time they became more critical and political. While some parties asked for caution in terminating the lockdown, other parties stressed economic interests. Persistent issues in the beginning of the crisis were the shortage, distribution and effectiveness of face masks and the lack of testing capacity. The “macro-reality” at government level frequently differed from the “micro-reality” of providers. For instance, while the government argued that the problem of a shortage of face masks had been solved, some care providers told the media that they still struggled with shortages. While the government spoke about a lack of testing capacity, some laboratories reported that their capacity had been unused. Box 7.1 The COVID-19 pandemic in the Netherlands The first COVID-19 patient (“patient zero”) in the Netherlands was confirmed at the end of February 2020. Fig. 7.2 shows that the number of weekly hospital admissions of COVID-19 patients increased exponentially in March 2020. In April and May, there followed a rapid decline in the number of hospital admissions. The effect of the second wave on hospital admissions became manifest in September. Fig. 7.2 Number of weekly hospital admissions of COVID-19 patients 0 100 200 300 400 500 600 700 28 /0 2/ 20 7/ 03 /2 0 15 /0 3/ 20 23 /0 3/ 20 31 /0 3/ 20 8/ 04 /2 0 16 /0 4/ 20 24 /0 4/ 20 2/ 05 /2 0 10 /0 5/ 20 18 /0 5/ 20 26 /0 5/ 20 3/ 06 /2 0 11 /0 6/ 20 19 /0 6/ 20 27 /0 6/ 20 5/ 07 /2 0 13 /0 7/ 20 21 /0 7/ 20 29 /0 7/ 20 6/ 08 /2 0 14 /0 8/ 20 22 /0 8/ 20 30 /0 8/ 20 7/ 09 /2 0 15 /0 9/ 20 23 /0 9/ 20 1/ 10 /2 0 9/ 10 /2 0 17 /1 0/ 20 25 /1 0/ 20 2/ 11 /2 0 10 /1 1/ 20 18 /1 1/ 20 26 /1 1/ 20 4/ 12 /2 0 12 /1 2/ 20 20 /1 2/ 20 28 /1 2/ 20 5/ 01 /2 1 13 /0 1/ 21 Source: National Institute for Public Health and the Environment (https://www.rivm.nl/coronavirus- covid-19/grafieken). The growth of hospital admissions was accompanied by an equally exponential increase in patients in intensive care (IC) units. The IC capacity of hospitals was radically scaled up to treat all patients. The severe triage scenario, in which doctors in emergency departments would be compelled to decide who will live and die because of the shortage of IC capacity, was only narrowly avoided. In December 2020 the registered number of COVID patients totalled slightly more than 4% of the population, and the registered death toll, measured as a percentage of registered COVID patients, was about 1.5%. However, data from Statistics Netherlands on excess mortality suggested a 50% higher all-causes death toll in the weeks of the pandemic than in those weeks in other years. The market reform in Dutch health care148 Another sign of centralization was the creation of national policy centres to coordinate the approach of the crisis. For instance, a national centre was set up to coordinate the distribution of COVID patients among the country’s hospitals. Other centres were created for the coordination of the distribution of the newly bought respiratory equipment among hospitals and the purchase and distribution of protective equipment. The government also made use of the military’s expertise in logistics operations. A striking impact of the pandemic has been the greater role of experts in policy-making. From the very beginning, the government heavily relied on the input of public health experts. The Outbreak Management Team under the chairpersonship of the Director of the National Centre for Infectious Diseases Control (which is part of the National Institute for Public Health and the Environment), consisting of physicians, epidemiologists, virologists and other public health experts, advised the government weekly on the course of the pandemic and the measures that could be taken to confine the spread of the virus. Of course, the government was responsible for final policy-making, but it frequently stressed the expert-driven nature of its decisions. The paradox of the pandemic is that public health measures with unprecedented consequences for people and social life were largely expert-driven and non-political. Only after The pandemic also caused a silent public health crisis. The treatment of non-COVID patients was largely postponed. The Dutch Healthcare Authority estimated that the number of hospital referrals by general practitioners dropped by some 360 000 and the number of referrals without follow-up increased by 290 000 (NZa, 2020a). Many patients avoided medical care for fear of infection. In the early stage of the pandemic, the government underestimated its size and its impact on public health. This radically changed in mid-March when the government announced an “intelligent lockdown” to stop the spread of the virus and to avert the looming overload of IC departments. The lockdown included the temporary closure of many places where people gather (for example, schools, bars, restaurants and sport institutes) and a ban on public activities, including sports matches, concerts and the like. Workers were requested to work from home as much as possible. People were also asked to social distance (“one-and-a-half-metre society”), wash their hands regularly and stay at home as much as possible. However, it was not forbidden to go outdoors, as was the case in some other European countries. Fig. 7.2 shows that the lockdown proved successful. Because the number of hospital admissions dropped at a high speed, the lockdown was terminated in a stepwise fashion as of the end of May 2020. However, in September the first signs of a second wave became manifest. After two failed attempts to reverse the rising trend of infections, the government decided to implement a complete lockdown in December. This lockdown was more radical than the first and included the closure of all shops for a five-week period. Also new was the obligation to wear face masks in public spaces. In January 2021 there followed a much-disputed measure: a curfew from 9:00 PM to 4:30 AM. Critical assessment of the market reform and its future 149 some period of time were critical voices heard and pleas made for involving other than public health experts in the advisory process. The pandemic has revealed a structural weakness in the Dutch health care system. The shortage of IC capacity did not come as a surprise. After all, competition is assumed to squeeze out all alleged overcapacity, which is discredited as waste. The conclusion must be that regulated competition works well only under stable public health conditions.2 The COVID-19 pandemic has put the model of regulated competition out of action. None of the contracts with hospitals had reckoned with a dramatic surge of IC patients and an equally dramatic decline in the treatment of non-COVID patients. Insurers were almost completely out of the picture during the early months of the crisis. They confined themselves to facilitating the handling of the crisis by cash advances. Insurers followed the same strategy for care workers who could not see patients anymore because of the lockdown and for that reason had no income. Thus, insurers actually paid for never-provided care. The national associations of hospitals and insurers also negotiated a deal on how to cover the costs of the pandemic and to avoid bankruptcies. Because of stagnation of the yearly contract process, they also decided to extend the contract duration by one year. At the request of the Minister of Health, the Dutch Healthcare Authority has issued a temporary policy regulation to enable the reimbursement of not- provided health care to ensure the continuity of providers of long-term care (NZa, 2020c). Another decision is that insurers may share a great deal of their costs by means of ex post compensation to avert budgetary shocks due to the COVID- 19 crisis (Baltesen, 2020). The Central Bank of the Netherlands has permitted insurers to use their “excess solvency” to cover the costs of the pandemic. It also announced that it would not require immediate recovery measures if insurers did not meet the required solvency rate (DNB, 2020). 7.7.1 Will the pandemic influence the future of regulated competition? Traditionally, Dutch health care has had a decentralized structure. Health insurance and health care provision are delegated to formally separate and independent organizations. A hierarchical relationship between the state and these organizations does not exist and, even more important, has never existed. The origins of this structure are rooted in the dominant role of the Christian 2 The question might be raised whether Dutch hospitals would have been better off in a state-controlled health care system. We have some doubts in this respect. In the period preceding the market reform in 2006, one of the government’s most important policy instruments to curb expenditure growth was a reduction of hospital bed capacity. For many years it has also stressed the need for restructuring the health care landscape through replacing expensive hospital care with less expensive but equally effective primary care. In other words, we presume that the regular capacity of hospitals to treat COVID-19 patients would also have fallen short had the Dutch health care system still been heavily state-controlled. The market reform in Dutch health care150 body of thought of subsidiarity, which held that the financing and provision of health care had to put as much as possible in the hands of organizations with a social purpose. The decentralized structure is not the result of the market reform. Instead, the reform built on it. The pandemic has clearly revealed an important weakness of a decentralized health care system. The government had no formal authority to issue orders. For example, the public health agencies that play a crucial role in testing and tracking and tracing fall under the jurisdiction of local government.3 At each stage, policy-making required a lot of consultation, sometimes at the expense of decisiveness in the implementation of government measures to resolve the pandemic. According to the Prime Minister, the crisis had made very manifest the need for more central coordination and state direction in health care. “With eight thousand ‘know-all’ general practitioners, a hundred hospitals, eight academic centres and 70 public health agencies, we have a world-famous health care sector. Yet we must draw lessons from what has happened” (EW, 2020). The Minister of Health was clearer, depicting the crisis as a big plea for less competition and more central coordination. In other words, a decentralized structure does not work in times of pandemic. Critics of the market reform have seized the opportunity to blame the reform for the lack of treatment capacity in hospitals. In this respect they refer to the bankruptcies of hospitals and the reduction of IC capacity in the name of efficiency. They argue that these events would never have happened in a state- led health care system. Whether the COVID-19 pandemic will result in incremental or more drastic adjustments in the current system is still an open question. However, it has provided critics of regulated competition with extra ammunition to urge a more directive state role in health care. Dutch health care is heading into an interesting period. 7.8 Conclusions The policy goals of the reform were to achieve a health care system with freedom of choice that offers high-quality care to patients and that is accessible to each person (universal access), based upon solidarity and affordable from a financial perspective (financial sustainability). Has the reform been a success? We finish this study with the following conclusions. 3 Many municipalities participate in regional public health agencies. Critical assessment of the market reform and its future 151 Freedom of choice • The reform has increased freedom of choice. Consumers are free to choose their health insurer and type of health plan. They can switch at the end of each year. They can also opt for a voluntary deductible. • A select group of consumers (young, high education, good health) have benefited most from freedom of choice. • Selective contracting (which is seldom applied) has not restricted freedom of choice. Universal access • The Dutch health care system scores well in terms of universal access to health insurance. The Health Insurance Act guarantees every person access to health insurance. The percentage of the uninsured is very low. In international comparisons Dutch health care scores high on the access dimension (OECD, 2020). Although the mandatory deductible has been criticized as a financial barrier to health care, it is relatively low in comparison with those in other countries (OECD, 2020). • The duty of care works well, obligating insurers to guarantee their clients access to all types of necessary health care covered by the Health Insurance Act. • The rapid growth of the number of ITCs has enhanced access to health care. However, with a few exceptions, in particular ophthalmology and dermatology, the market share of these centres has remained small. • The market reform has not resolved the problem of long waiting times in some specialties. The proliferation of ITCs (a direct effect of the reform) may help to resolve this problem in the future. • There is some evidence that the introduction and increase of the mandatory deductible (since 2016) has had a negative influence on access to health care. Solidarity • The Dutch health care system scores well in terms of risk solidarity. The ban on risk rating (experience rating) works well. The market reform in Dutch health care152 • The increased popularity of budget plans and voluntary deductibles signifies some adverse risk selection and undermines risk solidarity. • The Dutch health care system scores well in terms of income solidarity. The premiums are affordable, but they require substantial subsidies (care allowance) for people with lower income. The political debate on a fair distribution of the financial burden of health care has never stopped. • The risk equalization system to correct for differences in the insurers’ risk profile is sophisticated but imperfect. However, a perfect system is illusionary. Quality of care • There is no hard and consistent evidence of a positive or negative impact of the reform on quality of care. • The introduction and development of value-based purchasing may help to improve the quality of care. • Providers argue that cooperation instead of competition is the best route to quality improvement. There is some evidence for this claim. Financial sustainability • As of 2012 expenditure growth has been very moderate, both in percentage and in comparison with other countries. This is largely due to the collective framework agreements on expenditure growth and the abolition of the safety nets in risk equalization. The impact of price competition is limited, with one exception: the preference policy of insurers in purchasing outpatient prescription medicines. • The risk of the framework agreements is that low growth percentages during a certain period of time will be followed by a new cost explosion (as happened after the turn of the century). • High administrative costs is a serious problem, but a great deal of the administrative burden is related to other processes, in particular initiatives in quality improvement. • Insurers have been required to build up substantial reserves and have been able to do so. Their buffers are actually much higher than required. Critical assessment of the market reform and its future 153 • Hospitals, too, have built up substantial financial buffers, although there is much spread among hospitals. • With a few exceptions, there is no evidence of substantial rent-seeking behaviour by payer and provider organizations. Nevertheless, revision of legislation is underway to tackle rent-seeking behaviour by tightening licencing criteria. The experience with health care reform in the Netherlands is a good illustration of health care reform as a process of ups and downs. The rhetoric of reform has in some respects been remote from the hard realities of daily practice. High expectations and frustrations have gone hand in hand. Regulatory adjustments are certainly needed to correct for failures. Competition is not a one-size-fits-all model for health care. There is an argument for reinforcing the role of the state in some respects. However, a new large-scale reform directed at restoring the role of the state in health care runs the risk of public failures and, consequently, new frustrations. What is most needed is a strong focus on substantive issues to achieve value-driven health care instead of a renewed focus on institutional reform. The main lesson from the COVID-19 pandemic is that competition works only under normal public health conditions. It fails to work in the context of a worldwide and persistent pandemic. Given its strong emphasis on efficiency, competition fails to build up a reserve capacity that can be quickly mobilized. Building up such a capacity is a public problem that must be addressed by the state.

References Abrishami P, Boer A, Horstman K (2014). Understanding the adoption dynamics of medical innovations: affordances of the da Vinci robot in the Netherlands. Social Science & Medicine, 117:125–133. ACM (Autoriteit Consument & Markt [Authority for Consumers and Markets]) (2014). ACM wijst zorgverzekeraars op mededingings risico bij uitvoering gezamenlijke plannen spoedeisende zorg. The Hague (in Dutch). ACM (2017a). Toetredings- en groeidrempels op de zorgverzekeringsmarkt. The Hague (in Dutch). ACM (2017b). Prijs- en volume-effecten van ziekenhuisfusies. Onderzoek naar effecten van ziekenhuis fusies 2007-2014. The Hague (in Dutch). ACM (2017c). Premiestijging leidt niet tot meer overstappers. The Hague (in Dutch). ACM (2019). ACM biedt ruimte voor samenwerking bij juiste zorg op juiste plek. The Hague (in Dutch). Algemeen Dagblad [newspaper] (2019). Marktwerking in de zorg is doorgeslagen [interview with the Minister of Health]. 1 March (in Dutch). Algemeen Dagblad (2020). Zorg moet fundamenteel anders volgens ‘coronaminister’ Hugo de Jonge: ‘Regie terug bij overheid’ [interview with the Minister of Health]. 13 June (in Dutch). Algemene Rekenkamer [Court of Audit] (2016). Zorgakkoorden: uitgavenbeheersing in de zorg deel 4. The Hague (in Dutch). Algemene Rekenkamer (2017). Bekostiging van de geestelijke gezondheidszorg. The Hague (in Dutch). Algemene Rekenkamer (2020). Verzekerd van zinnige zorg. The Hague (in Dutch). Andeweg R, Irwin G, Louwerse T (2020). Governance and politics of the Netherlands. London: Red Globe Press. Arrow K (1963). Uncertainty and the welfare economics of medical care. The American Economic Review, 53(5):941–973. Baltesen F (2020). Zorgverzekeraars mogen volgend jaar 85 procent verevenen. Skipr [news website], 22 September (in Dutch). Batterink M, Reitsma J, Bakker L, Pomp M, Plu M (2016). Ziekenhuisfusies en kwaliteit van zorg. Barneveld: Significant (in Dutch). Berden C, Croes R, Kemp R, Mikkers M, van der Noll R, Shestalova V, Svitak J (2019). Hospital competition in the Netherlands: an empirical investigation. Tilburg (TILEC Discussion Paper, No. 2019-008). Bes R, Curfs E, Groenewegen P, de Jong J (2017). Selective contracting and channelling patients to preferred providers: a scoping review. Health Policy, 121(5):504–514. Boonen LH, Schut FT (2011). Preferred providers and the credible commitment problem in health insurance: first experiences with the implementation of managed competition in the Dutch health care system. Health Economics, Policy and Law, 6(2):219–235. The market reform in Dutch health care156 Brouns CH, Douven R, Kemp R (2020). Prices and market power in mental health care: evidence from a major policy change in the Netherlands. The Hague: Netherlands Bureau for Economic Policy Analysis (CPB). Clarke J, Newman J (1997). The managerial state. London: SAGE. COFZ (Commissie Onderzoek Faillissementen Ziekenhuizen) (2020). De aangekondigde ondergang: onderzoek naar de faillissementen van het MC Slotervaart en de MC IJsselmeerziekenhuizen. The Hague (in Dutch). Commonwealth Fund (2015). 2015 Commonwealth Fund international survey of primary care physicians in 10 nations. Washington, DC. Commonwealth Fund (2020). 2020 Commonwealth Fund international health policy survey. Washington, DC. Companje K-P, Hendriks R, Veraghtert K, Widdershoven B (2011). Two centuries of solidarity: German, Belgian and Dutch social health insurance 1770-2008. Amsterdam: Amsterdam University Press. Croes R, Mikkers M, Krabbe-Alkemade Y (2018). Competition and quality indicators in the health care sector: empirical evidence from the Dutch hospital sector. The European Journal of Health Economics, 19(1):5–19. Croes R, Schut F, Varkevisser M (2019). Chapter 6: Is adverse selection effectively mitigated by consumer inertia? Empirical evidence from the Dutch health insurance market. In: Croes R, editor. Market power in hospital markets and selection in health insurance markets. Rotterdam: Erasmus University Rotterdam; 151–173. CSFG (Commissie Structuur en Financiering van de Gezondheidszorg) (1987). Bereidheid tot verandering [Willingness to change]. The Hague (in Dutch). CTTO (Commissie Toekomst zorg thuiswonende ouderen) (2020). Oud en zelfstandig in 2030: een reisadvies. The Hague (in Dutch). de Jong J, Brabers A, Bouwhuis S, Bomhof M, Friele R (2015). Het functioneren van de zorgverzekeringsmarkt: een kennissynthese. Utrecht: NIVEL (in Dutch). de Swaan A (1988). In care of the state. Oxford: Oxford University Press. den Hartog M, Janssen R (2014). Ontwikkeling van de marktstructuur van Nederlandse ziekenhuizen 1978 tot 2013. Tijdschrift voor Gezondheidswetenschappen, 92(8):334–341 (in Dutch). DNB (De Nederlandsche Bank [Dutch Central Bank]) (2017). Visie op de toekomst van de Nederlandse zorgverzekeraars. Amsterdam (in Dutch). DNB (2020). DNB geeft zorgverzekeraars tijdens coronacrisis ruimte voor liquiditeitssteun aan zorgaanbieders [press release]. 20 April (in Dutch). Dohmen P, van Raaij E (2019). A new approach to preferred provider selection in health care. Health Policy, 123(3):300–305. Douven R (2004). Risk adjustment in the Netherlands: an analysis of insurers’ health care expenditures. The Hague (CPB Discussion Paper, No. 39). Douven R, Remmerswaal M, Mosca I (2015). Unintended effects of reimbursement schedules in mental health care. Journal of Health Economics, 42:139–150. Douven R, Katona K, Schut F, Shestalova V (2017). Switching gains and health plan price elasticities: 20 years of managed competition reforms in the Netherlands. The Hague: CPB. Douven R, Burger M, Schut F (2019). Does managed competition constrain hospitals’ contract prices? Evidence from the Netherlands. Health Economics, Policy and Law, 1–14. Douven R, Diepstraten M, Kopányi-Peuker A, van Dulmen S, Stadhouders N, Wackers E, Jeurissen P, Wijnker G, Mikkers M (2020). Evaluatie programma’s Beatrixziekenhuis en Bernhoven. The Hague: CPB (in Dutch). Emanuel E (2020). Which country has the world’s best health care? New York: PublicAffairs. Enthoven A (1988). Theory and practice of managed competition in health care finance. Amsterdam/ New York: North Holland. References 157 Enthoven A (1993). The history and principles of managed competition. Health Affairs, 12(1):24– 48. Enthoven A, van de Ven W (2007). Going Dutch – managed-competition health insurance in the Netherlands. The New England Journal of Medicine, 357(24):2421–2423. Equalis (2019). Premieverschillen tussen zorgverzekeraars nemen toe. Utrecht (in Dutch). EW (Elsevier Weekblad) (2020). Premier Rutte wil gezondheidszorg centraliseren [interview with the Prime Minister]. 17 December (in Dutch). Fabbricotti IN (2007). Zorgen voor Zorgketens [Taking care of integrated care] [thesis]. Rotterdam: Institute of Healthcare Policy and Management (in Dutch; summary in English). Figueras J, Robinson R, Jakubowski E, editors (2005). Purchasing to improve health systems performance. Maidenhead: Open University Press. Frank RG, McGuire TG (2000). Chapter 16: Economics and mental health. In: Culyer AJ, Newhouse JP, editors. Handbook of health economics, vol. 1B. Amsterdam: North Holland; 893–954. Freidson E (2001). Professionalism: the third logic. Cambridge, United Kingdom: Polity Press. Gaynor M (2006). What do we know about competition and quality in health care markets? Cambridge, Massachusetts: National Bureau of Economic Research (Working Paper 12301). Gaynor M, Moreno-Serra R, Propper C (2013). Death by market power: reform, competition, and patient outcomes in the National Health Service. American Economic Journal: Economic Policy, 5(4):134–166. Govaert J (2017). Value-based healthcare in colorectal cancer surgery: improving quality and reducing costs [thesis]. Leiden: Leiden University. Greer S, Klasa K, van Ginneken E (2020). Power and purchasing: why strategic purchasing fails. The Milbank Quarterly, 98(3):975–1020. Groenewegen P, Hansen J, de Jong J (2019). Trust in times of health reform. Health Policy, 123(3):281–287. Hagenaars L (2021). Reducing paper and sugar for fiscally sustainable healthcare systems: studies on administrative costs and junk food tax policies [thesis]. Nijmegen: Radboud University. Halbersma R, Mikkers M, Motchenkova E, Seinen I (2011). Market structure and hospital–insurer bargaining in the Netherlands. The European Journal of Health Economics, 12:589–603. Halbersma R, van Manen J, Sauter W (2012). Voldoen de verzekeraars in hun rol als motor van het zorgstelsel? Utrecht: Dutch Healthcare Authority (NZa), Research Paper 2012-3 (in Dutch). Hanekamp M, Heesbeen S, van der Helm I, Valks R (2019). Administratieve belasting langdurige zorg 2019: uitkomsten onderzoek Berenschot. Utrecht: Berenschot (in Dutch). Hasaart F (2011). Incentives in the diagnosis treatment combination payment system for specialist medical care [thesis]. Maastricht: Maastricht University. Heijink R, Mosca I, Westert G (2013). Effects of regulated competition on key outcomes of care: cataract surgeries in the Netherlands. Health Policy, 113(1–2):142–150. Helderman J-K, Stiller S (2014). The importance of order and complements: a new way to understand the Dutch and German health insurance reforms. Journal of Health Politics, Policy and Law, 39(4):811–840. Helderman J-K, Schut F, van der Grinten T, van de Ven W (2005). Market-oriented health care reforms and policy learning in the Netherlands. Journal of Health Politics, Policy and Law, 30(1–2):189–209. Hemerijck A, Unger B, Visser  J (2000). Chapter 5: How small countries negotiate change: twenty-five years of policy adjustment in Austria, the Netherlands, and Belgium. In: Scharpf F, Schmidt V, editors. Welfare and work in the open economy, vol. 2: diverse responses to common challenges in twelve countries. Oxford: Oxford University Press; 175–263. Himmelstein DU, Jun M, Busse R, Chevreul K, Geissler A, Jeurissen P, Thomson S, Vinet M-A, Woolhandler S (2014). A comparison of hospital administrative costs in eight nations: US costs exceed all others by far. Health Affairs, 33(9):1586–1594. The market reform in Dutch health care158 Hirschman A (1970). Exit, voice, and loyalty. Cambridge, Massachusetts: Harvard University Press. Holst L, Brabers A, de Jong JD (2019). 8% van de verzekerden geeft aan te zijn overgestapt van zorgverzekeraar in 2019. Utrecht: NIVEL (in Dutch). Holst L, Brabers A, de Jong J (2020). Infographic. Barometer Solidariteit - update met cijfers uit 2019. Utrecht: NIVEL (in Dutch). Houwaart ES (1991). De hygiënisten: artsen, staat & volksgezondheid in Nederland 1840-1890. Groningen: Historische Uitgeverij (in Dutch). IGJ (Inspectie Gezondheidszorg en Jeugd [Health and Youth Care Inspectorate]) (2018). Het resultaat telt 2016. Indicatoren particuliere klinieken: scores, bevindingen en acties. Utrecht (in Dutch). Ikkersheim DE, Koolman X (2013). Chapter 4: The first effects of Dutch health care reform on hospitals. In: Ikkersheim DE, editor. The Dutch health system reform: creating value. Amsterdam: Free University; 50–63. Janssen R (2017). Het onzekere voor het zekere: tweebenig besturen in de zorg [Uncertain times: ambidextrous management in healthcare]. Rotterdam: Erasmus University Rotterdam (English, 56 ff.). Jeurissen P (2010). For-profit hospitals: a comparative and longitudinal study of the for-profit hospital sector in four western countries [thesis]. Rotterdam: Erasmus University Rotterdam. Jeurissen P, Sanders F (2007). Chapter 3: Solidarity: who cares? In: Steenbeek OW, van der Lecq SG, editors. Cost and benefits of collective pension systems. Berlin: Springer; 33–48. Kingdon J (1984). Agendas, alternatives, and public policies. Boston: HarperCollins Publishers. Klasa K, Greer S, van Ginneken E (2018). Strategic purchasing in practice: comparing ten European countries. Health Policy, 122(5):457–472. Klijs B, van Hilten O (2020). Prijs- en volumeontwikkelingen in de medisch specialistische zorg in 2013-2017. The Hague: Statistics Netherlands (CBS) Statistische Trends (in Dutch). KNGF (Koninklijk Nederlands Genootschap voor Fysiotherapie [Royal Dutch Society for Physiotherapy]) (2019). Contracteergraad fysiotherapie daalt [press release]. In Dutch. KPMG (2015). Invloed banken op fusies ziekenhuizen neemt toe. In Dutch. KPMG (2018). Wie doet het met wie? In Dutch. KPMG/Plexus (2014). Evaluatie Zorgverzekeringswet. Amstelveen (in Dutch). Krabbe-Alkemade Y (2014). The impact of market competition and patient classification on Dutch hospital behavior [dissertation]. Delft: Eburon. Kroneman M, Boerma W, van den Berg M, Groenewegen P, de Jong J, van Ginneken E (2016). Netherlands: health system review. Health systems in transition, vol. 18(2). Copenhagen: World Health Organization (acting as the host organization for, and secretariat of, the European Observatory on Health Systems and Policies) (https://www.euro.who.int/__data/ assets/pdf_file/0016/314404/HIT_Netherlands.pdf, accessed 23 August 2021). Kruse F (2018). Chapter 14: Focusklinieken en betaalbaarheid. In: Jeurissen PPT, Maarse JAM, Tanke MAC, editors. Betaalbare zorg. The Hague: SDU; 259–276 (in Dutch). Kruse F (2021). Healthcare provision: open for business? [thesis]. Nijmegen: Radboud University. Kruse F, Spierings E, Adang E, Jeurissen P (2018). Marktconcentratie is ook punt van zorg bij zelfstandige behandelcentra. Economisch Statistische Berichten, 103(4766):453–455 (in Dutch). Kruse F, Groenewoud S, Atsma F, van der Galiën O, Adang E, Jeurissen P (2019). Do independent treatment centers offer more value than general hospitals? The case of cataract care. Health Services Research, 54(6):1357–1365. Lambregts T, van Vliet R (2018). The impact of copayments on mental healthcare utilization: a natural experiment. The European Journal of Health Economics, 19:775–784. Larjow E (2018). Administrative costs in health care – a scoping review. Health Policy, 122(11):1240–1248. References 159 Lijphart A (1968). The politics of accommodation: pluralism and democracy in the Netherlands. Berkeley: University of California Press. Lijphart A (1999). Patterns of democracy: government forms and performance in thirty-six countries. New Haven: Yale University Press. Loozen E (2015). Public healthcare interests require strict competition enforcement. Health Policy, 119(7):882–888. Loozen E, Varkevisser M, Schut F (2016). Goede zorginkoop vergt gezonde machtsverhoudingen. Het belang van markt- en mededingingstoezicht binnen het Nederlandse zorgstelsel. Rotterdam: Instituut Beleid & Management Gezondheidszorg (in Dutch). Maarse J, Jeurissen P (2016). The policy and politics of the 2015 long-term care reform in the Netherlands. Health Policy, 120(3):241–245. Maarse J, Jeurissen P (2019). Low institutional trust in health insurers in Dutch health care. Health Policy, 123(3):288–292. Maarse J, Jeurissen P (2020). Chapter 11: Private health insurance in the Netherlands. In: Thomson S, Sagan A, Mossialos E, editors. Private health insurance: history, politics and performance. Cambridge/Copenhagen: Cambridge University Press/World Health Organization (acting as the host organization for, and secretariat of, the European Observatory on Health Systems and Policies); 349–376. Majone G, Wildavsky A (1978). Implementation as evolution. In: Freeman HE, editor. Policy studies review annual, vol. 2. Beverly Hills: Sage Publications; 103–117. Melnick G, Zwanziger J (1995). State health care expenditures under competition and regulation, 1980 through 1991. American Journal of Public Health, 85(10):1391–1396. Meloen J, Groenewegen P, Hingstman L (2000). De toekomst van het algemene ziekenhuis. Utrecht: NIVEL (in Dutch). Merrill J, McLaughlin C (1986). Competition versus regulation: some empirical evidence. Journal of Health Politics, Policy and Law, 10(4):613–623. MoH (Ministry of Health) (1988). Verandering verzekerd [Change assured]. The Hague (in Dutch). MoH (1990). Werken aan zorgvernieuwing. The Hague (in Dutch). MoH (1992). Weloverwogen verder. The Hague (in Dutch). MoH (2001). Vraag aan bod [A question of demand]. The Hague (in Dutch). MoH (2011a). Zorg die loont. The Hague (in Dutch). MoH (2016). Visie op geneesmiddelen: nieuwe geneesmiddelen snel bij de patiënt tegen aanvaardbare kosten [policy memorandum]. The Hague (in Dutch). MoH (2018a). Voortgangsbrief financiële arrangementen 2018. 2 October. The Hague (in Dutch). MoH (2018b). Rechtmatige zorg: aanpak van fouten en fraude. The Hague (in Dutch). MoH (2019a). Intrekking voorstel van wet vergroten investeringsmogelijkheden in medisch- specialistische zorg. The Hague (in Dutch). MoH (2019b). Rijksbegroting Ministerie van Volksgezondheid, Welzijn en Sport 2020. The Hague (in Dutch). MoH (2020). Rijksbegroting Ministerie van Volksgezondheid, Welzijn en Sport 2021. The Hague (in Dutch). Mol A (2008). The logic of care: health and the problem of patient choice. Abingdon: Routledge. Moran M (1999). Governing the health care state: a comparative study of the United Kingdom, the United States and Germany. Manchester: Manchester University Press. Mous K, de Jongh L (2017). Machtigingsvereiste zorgverzekeraars moet op de schop. Zorgvisie [news website], 6 January (in Dutch). NVZ (Nederlandse Vereniging van Ziekenhuizen [Dutch Hospitals Association]) (2017). Ziekenhuiszorg in cijfers. Brancherapport: algemene ziekenhuizen 2017. Utrecht (in Dutch). NVZ (2018). Ziekenhuiszorg in cijfers. Brancherapport: algemene ziekenhuizen 2018. Utrecht (in Dutch). The market reform in Dutch health care160 NZa (Nederlandse Zorgautoriteit [Dutch Healthcare Authority]) (2012). Marktscan mondzorg. Utrecht (in Dutch). NZa (2013). Marktscan medisch specialistische zorg: weergave van de markt 2009-2013. Utrecht (in Dutch). NZa (2014a). Monitor transparantie voor consumenten. Utrecht (in Dutch). NZa (2014b). Rechtmatige uitvoering Zorgverzekeringswet 2013. Utrecht (in Dutch). NZa (2014c). Samenvattend rapport: rechtmatige uitvoering Zorgverzekeringswet 2013. Utrecht (in Dutch). NZa (2014d). Good Contracting Practices. Zorgverzekeraars en vrije beroepsbeoefenaren. Utrecht (in Dutch). NZa (2015). Marktscan Zorgverzekeringsmarkt 2015. Utrecht (in Dutch). NZa (2016a). Marktscan Zorgverzekeringsmarkt 2016. Utrecht (in Dutch). NZa (2016b). Kwantitatief onderzoek naar risicoselectie en risicosolidariteit op de zorgverzekeringsmarkt. Utrecht (in Dutch). NZa (2017a). Marktscan zorgverzekeringsmarkt 2017. Utrecht (in Dutch). NZa (2017b). Beleidsregel toezichtkader zorgplicht zorgverzekeraars Zvw. Utrecht (in Dutch). NZa (2017c). Marktscan medisch-specialistische zorg 2016. Utrecht (in Dutch). NZa (2018a). Beter kiezen op de polismarkt. Utrecht: NZa/ACM (in Dutch). NZa (2018b). Marktscan zorgverzekeringsmarkt 2018. Utrecht (in Dutch). NZa (2018c). Advies bekostiging medisch-specialistische zorg: belonen van zorg die waarde toevoegt. Utrecht (in Dutch). NZa (2019a). Monitor zorgverzekeringen 2019. Utrecht (in Dutch). NZa (2019b). Samenvattend rapport Zorgverzekeringswet 2017/2018. Utrecht (in Dutch). NZa (2019c). Monitor geneesmiddelen in de medisch-specialistische zorg. Utrecht (in Dutch). NZa (2019d). Monitor contractafspraken huisartsenzorg en multidisciplinaire zorg 2019. Utrecht (in Dutch). NZa (2019e). Monitor contractering wijkverpleging 2019. Utrecht (in Dutch). NZa (2019f ). Monitor contractering GGZ 2019. Utrecht (in Dutch). NZa (2019g). Monitor inhoud contracten medisch-specialistische zorg 2018. Utrecht (in Dutch). NZa (2020a). Analyse van de gevolgen van de coronacrisis voor de reguliere zorg. Utrecht (in Dutch). NZa (2020b). Stand van de zorg: 15 jaar Zorgverzekeringswet. Utrecht (in Dutch). NZa (2020c). Beleidsregel continuïteitsbijdrage en meerkosten in verband met de uitbraak van het Sars-CoV-2 virus. Utrecht (in Dutch). OECD (2014). Health at a glance: Europe 2014. Paris: OECD Publishing. OECD (2020). Health at a glance: Europe 2020. Paris: OECD Publishing. OECD/World Health Organization (acting as the host organization for, and secretariat of, the European Observatory on Health Systems and Policies) (2019). The Netherlands: country health profile 2019. State of health in the EU (series). Paris: OECD Publishing (https://doi. org/10.1787/9ac45ee0-en, accessed 23 August 2021). Okma G (1997). Studies on Dutch health politics, policies and law [thesis]. Utrecht: Utrecht University. Okma K, Tenbensel T, editors (2020). Health reforms across the world. New Jersey: World Scientific. Øvretveit J (2009). Does improving quality save money? A review of evidence of which improvements to quality reduce costs to health service providers. London: The Health Foundation. OVV (Onderzoeksraad voor Veiligheid [Dutch Safety Board]) (2019). Faillissement MC Slotervaart en MC Ijsselmeer-ziekenhuizen: risico’s voor patiëntveiligheid. The Hague (in Dutch). Pollitt C (1993). Managerialism and the public services: cuts or cultural change in the 1990s? Oxford: Blackwell Business. PPRC (Public Procurement Research Centre) (2020). Monitor gemeentelijke zorginkoop 2019. Lunteren: PPRC/MoH (in Dutch). References 161 Puijk L, Baas N, Meijer R (2017). (Niet-)gecontracteerde zorg in de wijkverpleging. Hilversum: Arteria Consulting (in Dutch). Ravesteijn B, Schachar E, Beekman A, Janssen R, Jeurissen P (2017). Association of cost sharing with mental health care use, involuntary commitment, and acute care. JAMA Psychiatry, 74(9):932–939. Remmerswaal M, Douven R, Besseling P (2015). Toelichting op de effecten van eigen betalingen in de zorg. The Hague: CPB (in Dutch). Roos A, Postma J (2016). Why healthcare providers merge. Health Economics, Policy and Law, 11(2):121–140. Roos A, van Doorslaer E, O’Donnell O, Schut F, Varkevisser M (2018). Chapter 3: Does price competition damage healthcare quality? In Roos A, editor. Mergers and competition in the Dutch healthcare sector. Rotterdam: Erasmus University Rotterdam; 68–103. Ruwaard S (2018). Purchasing healthcare: beyond getting the financial incentives right [dissertation]. Tilburg: Tilburg University. RVS (Raad voor Volksgezondheid en Samenleving [Council for Public Health and Society]) (2020). Van deelbelangen naar gedeeld belang. The Hague (in Dutch). RVZ (Raad voor de Volksgezondheid en Zorg [Council for Public Health and Health Care]) (2005). Houdbare solidariteit in de gezondheidszorg. Zoetermeer (in Dutch). Saltman R, Figueras J (1997). European health care reform: analysis of current strategies. Copenhagen: World Health Organization Regional Office for Europe (https://apps.who.int/ iris/handle/10665/107267, accessed 23 August 2021). Schakel C (2020). Fiscal rules for health, a guide and not a jailer? [dissertation]. Nijmegen: Radboud University. Schippers E (2017a). Naar twintig jaar Zorgverzekeringswet. Economisch Statistische Berichten, 102(4745):16–19 (in Dutch). Schippers E (2017b). Interview in Vrij Nederland (in Dutch). Schön D, Rein M (1994). Frame reflection: toward the resolution of intractable policy controversies. New York: Basic Books. Schultze C (1977). The public use of private interest. Washington, DC: The Brookings Institution. Schut F (1995). Competition in the Dutch health care sector [dissertation]. Rotterdam: Erasmus University Rotterdam. Schut F, van den Berg B (2010). Sustainability of comprehensive universal long-term care insurance in the Netherlands. Social Policy & Administration, 44(4):411–435. Schwartz B (2004). The paradox of choice: why more is less. New York: Harper Perennial. Simborg DW (1981). DRG creep: a new hospital-acquired disease. The New England Journal of Medicine, 304(26):1602–1604. Stadhouders N (2019). Effective healthcare cost containment policies [thesis]. Nijmegen: Radboud University. Starr P (1982). The social transformation of American medicine. New York: Basic Books. Stolper K, Boonen L, Schut F, Varkevisser M (2019). Managed competition in the Netherlands: Do insurers have incentives to steer on quality? Health Policy, 123(3):293–299. Taskforce Beheersing Zorguitgaven (2012). Naar beter betaalbare zorg. The Hague (in Dutch). Thomson S, Busse R, Crivelli L, van de Ven W, van de Voorde C (2013). Statutory health insurance competition in Europe: a four-country comparison. Health Policy, 109(3):209–225. Trescher A-L, Listl S, van der Galien O, ADVOCATE Consortium, Gabel F, Kalmus O (2020). Once bitten, twice shy? Lessons learned from an experiment to liberalize price regulations for dental care. The European Journal of Health Economics, 21:425–436. Tuohy C (2003). Agency, contract, and governance: shifting shapes of accountability in the health care arena. Journal of Health Politics, Policy and Law, 28(2–3):195–216. Tuohy C (2018). Remaking policy: scale, pace, and political strategy in health care reform. Toronto: University of Toronto Press. The market reform in Dutch health care162 Uenk N (2019). Commissioning of social care services [dissertation]. Utrecht: University of Utrecht. van de Bovenkamp HM, de Mul M, Quartz J, Weggelaar-Jansen A, Bal R (2014). Institutional layering in governing healthcare quality. Public Administration, 92(1):208–223. van de Gronden J (2004). Zorg tussen lidstaat en interne markt: zorgverzekeringen, EG-recht en particulier initiatief. Deventer: Kluwer (in Dutch). van de Ven W (2018). Onbeheersbare kostenstijging dreigt in de zorg. TPEDigitaal, 12(1):28–38 (in Dutch). van de Ven W, Schut F (2008). Universal mandatory health insurance in the Netherlands: a model for the United States? Health Affairs, 27(3):771–781. van de Ven W, Beck K, Buchner F, Schokkaert E, Schut F, Shmueli A, Wasem J (2013). Preconditions for efficiency and affordability in competitive healthcare markets: are they fulfilled in Belgium, Germany, Israel, the Netherlands and Switzerland? Health Policy, 109(3):226–245. van de Ven W, van Vliet RCJA, van Kleef RC (2017). How can the regulator show evidence of (no) risk selection in health insurance markets? Conceptual framework and empirical evidence. The European Journal of Health Economics, 18:167–180. van der Aa MJ, Klosse S, Paulus A, Evers SMAA, Maarse JAM (2018). Solidarity in insuring financial risks of illness: a comparison of the impact of Dutch policy reforms in health insurance and disability insurance since the 1980s. Journal of Comparative Policy Analysis: Research and Practice, 21(2):199–215. van Dijk C, van den Berg B, Verheij R, Spreeuwenberg P, Groenewegen P, de Bakker D (2013). Moral hazard and supplier-induced demand: empirical evidence in general practice. Health Economics, 22(3):340–352. van Dulmen S, Stadhouders N, Westert G, Wackers E, Jeurissen P (2020). Op weg naar hoge kwaliteit en lage kosten in de medisch specialistische zorg Evaluatie veranderprogramma’s Rivas Zorggroep en Bernhoven. Nijmegen: Radboudumc (in Dutch). van Eijkel R, van der Torre A, Kattenberg M, Eggink E (2017). De markt voor huishoudelijke hulp: Gemeentelijk inkoopbeleid en gebruik door burgers onder de Wmo 2007. The Hague: CPB/Netherlands Institute for Social Research (SCP) (in Dutch). van Esch T, Brabers A, van Dijk C, Gusdorf L, Groenewegen P, de Jong J (2017). Increased cost sharing and changes in noncompliance with specialty referrals in the Netherlands. Health Policy, 121(1):180–188. van Gerwen L, Huijsmans K, Luijk R (2019). Ontwikkeling (niet-)gecontracteerde wijkverpleging 2016-2018. Zeist: Vektis (in Dutch). van Kleef R, van de Ven W, van Vliet R (2013). Risk selection in a regulated health insurance market: a review of the concept, possibilities and effects. Expert Review of Pharmacoeconomics & Outcomes Research, 13(6):743–752. van Kleef R, Schut F, van de Ven W (2014). Evaluatie zorgstelsel en risicoverevening. Acht jaar na invoering Zorgverzekeringswet: succes verzekerd? Rotterdam: Instituut Beleid & Management Gezondheidszorg (in Dutch). van Leersum N, Bennemeer P, Otten M, Visser S, Klink A, Kremer J (2019). Cure for increasing health care costs: the Bernhoven case as driver of new standards of appropriate care. Health Policy, 123(3):306–311. van Mens K, Lokkerbol J, Janssen R, van Orden M, Kloos M, Tiemens B (2018). A cost-effectiveness analysis to evaluate a system change in mental healthcare in the Netherlands for patients with depression or anxiety. Administration and Policy in Mental Health, 45:530–537. van Os J (2020). Verzekeraars verschillen dramatisch bij zorginkoop ggz. Zorgvisie [news website], 7 July (in Dutch). van Twist M (1994). Verbale vernieuwing. The Hague: VUGA (in Dutch). van Veghel D (2019). Improving cardiovascular disease management by implementing value-based healthcare principles [dissertation]. Eindhoven: Technical University Eindhoven. References 163 Varkevisser M, Schut F (2019). Fusietoetsing in de zorg: een terug- en vooruitblik. Tijdschrift voor Toezicht (doi:10.5553/TvT/187987052019010001003) (in Dutch). Vektis (2006). Zorgmonitor – Jaarboek 2006. Zeist (in Dutch). Vektis (2008). Zorgmonitor – Jaarboek 2008. Zeist (in Dutch). Vektis (2016). Jaarcijfers zorgverzekeraars en zorgfinanciering 2015. Zeist (in Dutch). Vektis (2018). Aandeel zelfstandige behandelcentra blijft toenemen. Zeist (in Dutch). Vektis (2019a). Jaarcijfers zorgverzekeraars en zorgfinanciering 2019. Zeist (in Dutch). Vektis (2019b). Verzekerden in beeld 2019. Zeist (in Dutch). Vickers G (1965). The art of judgment: a study of policy making. London: Chapman & Hall. Visser J, Hemerijck A (1997). A Dutch miracle: job growth, welfare reform and corporatism in the Netherlands. Amsterdam: Amsterdam University Press. Welch HG, Schwartz L, Woloshin S (2011). Overdiagnosed: making people sick in the pursuit of health. Boston: Beacon Press. Westra D (2017). Healthcare’s competition conundrum [thesis]. Cooperative inter-organizational strategies in competitive healthcare markets. Maastricht: Maastricht University. Westra D, Wilbers G, Angeli F (2016). Stuck in the middle? A perspective on ongoing pro- competitive reforms in Dutch mental health care. Health Policy, 120(4):345–349. WFZ (Waarborgfonds voor de Zorgsector [Health Care Sector Guarantee Fund]) (2017). Jaarverslag 2017. Utrecht (in Dutch). Willemsen Committee (1994). Onderzoek besluitvorming volksgezondheid. The Hague (in Dutch). Zegers M, Veenstra GL, Gerritsen G, Verhage R, van der Hoeven JG, Welker GA (2020). Perceived burden due to registrations for quality monitoring and improvement in hospitals: a mixed methods study. International Journal of Health Policy and Management. Doi: 10.34172/ ijhpm.2020.96. ZiN (Zorginstituut Nederland [National Health Care Institute]) (2014). GIPeilingen 2013. Diemen (in Dutch). ZiN (2018). Rapport verdiepende analyses zorglasten Zvw 2015. Diemen (in Dutch). ZiN (2020). Zorgcijfers monitor: Zorgverzekeringswet en Wet langdurige zorg 4e kwartaal 2020. Diemen (in Dutch). ZN (Zorgverzekeraars Nederland [Health Insurers Netherlands]) (2015). Actieplan Kern-gezond [In perfect health: action plan]. Zeist (in Dutch). ZN (2019). Zorgverzekeraars in de regio. Zeist (in Dutch). Zuiderent-Jerak T, Grit K, van der Grinten T (2011). Markets and public values in healthcare. Wetenschappelijke Raad voor het Regeringsbeleid. The Hague (in Dutch). TH E M A R K ET R EFO R M IN D U TCH H EA LTH CA R E The market reform in Dutch health care Results, lessons and prospects 55 In 2006, the Netherlands embarked upon an ambitious reform of the Dutch health care system based upon the principles of regulated competition. Some 15 years later, it is an appropriate time to find out how this ‘market reform’ has worked out, and what the experience has been like for those involved in putting it into practice. The authors of this important new study review the reforms and their impact to date and ask whether the reforms merit being counted as a success. Did they alter the relationship between state, insurers, providers and patients? Has there been evidence of problems that market-based systems are often associated with, such as high administrative costs, restricted access to health care, rent-seeking, skimming and adverse selection? Whilst addressing these questions and suggesting possible answers, the authors also examine what can be learned from the Dutch experience with competition in health care and what changes might be expected in the near future in the Netherlands and more broadly, especially considering the context of the COVID-19 pandemic. The editors Patrick Jeurissen is professor of Sustainable Health Care at the Radboud University, Nijmegen Hans Maarse was professor of Health Policy Analysis at Maastricht University, Maastricht Health Policy Series Series No. 55 www.eurohealthobservatory.who.int Edited by: Patrick Jeurissen Hans Maarse O bs er va to ry St ud ie s Se ri es 55 E dited by P atrick Jeurissen and H ans M aarse R E S U LTS , LE S S O N S A N D P R O S P E CTS OBS_55_A4_Dutch health care.qxp_WHO_A4 21/10/2021 14:33 Page 1