Institutional influences on economic policy in Turkey : a three industry comparison

PSD Occasional Paper No. 3 July 1995 Institutional Influences on Economic 3 Policy in Turkey A Three-Industry Comparison Jesse Biddle, Department of Sociology, The American University; and Vedat Milor, Department of Sociology, Brown University Rfi! The World Bank Private Sector Development Department  I  Private Sector Development Department Occasional Paper No. 3 Institutional Influences on Economic Policy in Turkey: A Three-Industry Comparison Jesse Biddle, Department of Sociology,The American University; and Vedat Milor, Department of Sociology, Brown University July 1995 The World Bank Private Sector Development Department   CONTENTS Abstract .............................................................. v I. Introduction ........................................................1 A. Theoretical Orientations and Hypotheses ................................................ 2 B. Summary of Main Findings ....................................................... 5 H. Industrial Policy and State-Business Relations In Turkey . . 9 A. Historical Overview. 9 B. Main Elements of the Incentive Regime .1 mi. Selectivity and Monitoring in the Incentive Regime . .15 A. Application Process for Incentive Regime .16 1. Transparency and Predictability .17 B. Effectiveness of Incentive Regime ..18 1. Initial Screening of Applicants .19 2. Monitoring and Discipline of Certificate Holders .20 3. Performance of Certificate Holders .27 IV. Bureaucratic Insulation and the Incentive Regime. 3 1 Box. An Example of Bureaucratic Insulation: TSKB .36 V. Industry Policy Networks and the Incentive Regime .41 A. Rent-Oriented Networks and the Auto Industry .42 B. Rent Seeking cum Growth-Oriented Network: The Glass Industry .48 C. Growth-Oriented Network in the Readywear Clothing Industry ..52 1. Private Sector Economic Governance .53 VI. Policy Recommendations .57 References .............................................................. 61 Appendix I: Selectivity of Incentive Regime: Traditional Criteria ..................... 65 Appendix II: Three Industries: An Overview ...................................................... 75 Appendix mII: Field Research Methods .............................................................. 79 iii   Abstract Why is it that seemingly similar instruments of industrial promotion produces vastly dissimilar economic outcomes across countries? Why do such instruments result in rent seeking in some countries, and promote productivity in others? This paper attempts to provide an answer by examining whether institutional variables, namely the nature of policy networks, the characteristics of incentive and accountability mechanisms, and the degree of bureaucratic insulation have an impact on the effectiveness of state intervention, where effectiveness is associated wlith the degree to which incentives are used for their stated purposes. More specifically, the paper addresses two questions: i) what are the institutional properties of the Turkish industrial incentive (subsidy) regime and how these properties condition the attainment of stated policy goals, and, ii) do these properties exhibit significant differences between the ready-wear clothing, automotive and glass industries? The overall finding of the paper is that certain institutional features of the Turkish incentive system exacerbate tendencies towards rent seeking. In Turkey, the bureaucracy is not well insulated from political pressures; incentive policies are hence discretionary and erratic. Moreover, state officials are generally unable or unwilling to monitor recipient performance for compliance; hence enforcement is quite lax. Such tendencies were exacerbated by a restructuring of the public administration in the 1980s, which while proved helpful to overcome a statist bureaucratic culture, also unduly politicized the economic bureaucracy. Nevertheless, the organizational structure of industry and the nature of the relation between business and the state can still have an independent effect of the effectiveness of policy. This was evident in the ready-wear industry, where the delegation of the monitoring function to an industry association has worked effectively. The mediation of the industry association between subsidy recipients and the state generated mechanisms of self-regulation whereby discipline was imposed by the association as a collective enforcement mechanism of the members. These findings have important implications for policy making. First, the findings question the notion, prevalent in some institutionalist literature, that it is easier to monitor the behavior of incentive beneficiaries in concentrated industries. Project findings would suggest that this argument is premised on the prior existence of an insulated and competent bureaucratic corps. When this institutional precondition is lacking, the decentralization of these tasks to self-regulating bodies is a possible solution. This solution is more likely to work in competitive rather than concentrated industries. v   1 Introduction1 In Turkey, as well as in many other newly industrializing countries, economic policy makers in the post world war II period used industrial policies as a means of promoting the industrial sector. A wide range of protectionist policies inspired by an import substitution industrialization strategy were prominent instruments of Turkish industrial policy through the 1970's. From the 1980's onward, however, given the context of a gradually liberalizing trade regime, the selective use of incentives and subsidies has played an important role. Until the summer of 1994, when the Turkish authorities agreed to streamline the incentive regime in line with the European Community Priorities, these incentives and subsidies, such as exemptions from customs duties, various tax breaks and subsidized credit, have been targeted for a number of purposes. Major objectives included not only industrial targeting but also promotion of investment in less developed regions and of exporting activities as well. There is a vigorous debate with respect to whether or not an industrial policy premised on selective use of incentives and subsidies as seen in Turkey may constitute a coherent means of promoting economic growth and international competitiveness. Although Turkey has been reforming the incentive regime in the last year and a half, an inquiry into the institutional dimensions of past industrial policy is a worthwhile task to undertake in order both to inform policy and to contribute to the ongoing debate. In fact, the contours of this debate have recently centered on alternative interpretations of the spectacular economic performance of so-called East Asian "developmental states," such as South Korea. At question is whether the well-established presence in such nations of selective use of incentives and subsidies are among the causes of their superior economic performance or are spuriously or negatively associated. While more orthodox analyses stress the openness and macro economic stability of these economies, there has been an explosion of "institutional" analyses arguing that East Asian economic performance is partially attributable to strategic government industrial policies which helped create a dynamic comparative advantage.2 I The authors wish to thank Izak Atiyas, Ben Ross Schneider and Ziya Onis for their helpful comments on an earlier draft of this paper. In addition, research assistance was provided by Hakan Batur while research assistance as well as quantitative data analysis was provided by Erkan Erdil. 2 See Balassa (1981), Lal (1983), and Krueger (1980) for empirical arguments relating East Asian economic performance to expectations of the neo-classical paradigm and theory of comparative advantage. The "institutionalist" critique is best articulated by Wade (1990). See also Amsden (1989), Deyo, ed. (1987) and Onis (1991). Recently, a World Bank study endorsed some of the tenets of the institutionalist critique, particularly those relating to bureaucratic insularity and state- 1  2 Institutional Influences on Economic Policy in Turkey Economic modelling can demonstrate that under certain imperfect market conditions the use of trade barriers, incentives and subsidies in a selective industrial policy may lead to optimal solutions. Strategic trade policy, for example, may help increase national market share in concentrated international markets (Krugman, ed., 1986). However, a serious difficulty arises as the availability of the rents created by industrial policies generates incentives for business actors, politicians and bureaucrats to rent-seek or, as put by Bhagwati (1982), to engage in "directly unproductive activities." Of course, the nature of the policy regime is related to the degree of rent-seeking. In particular, an import substitution industrialization (ISI) strategy is likely to be subject to higher degrees of rent-seeking than an outwardly-oriented policy whose aim is to promote externally competitive industries. This is because the type of policy instruments associated with ISI policies, such as quotas and tariffs, not only enhance government discretionism over resource allocation but also, by reducing competition, encourage the formation of a narrow, protectionist business culture bent on asking special favors and resisting modifications to the system. At the same time, even an export-oriented policy regime is not necessarily free from rent-seeking. This is because, even after a liberalization of the foreign trade regime governments can make use of a multitude of other direct inducements, such as preferential credits, discretionary tax breaks, custom duty exemptions, etc., which engender demands for favorable treatment. Thus, irrespective of the nature of the policy regime, as the rents created by industrial policy interventions are the product of decisions made in specific institutional contexts in which rent-seeking political pressure may be influential there is an unambiguous potential for an initially coherent export-oriented industrial policy to be diverted over time toward narrow, private ends. Hence, our purpose in this paper is less to address the contribution of Turkish incentive and subsidy policies to that nation's economic performance during the liberal export-led stage of development in the 1980's as to address the equally important issue of the coherence and effectiveness with which the state delivers such policies. There is no doubt that there has been a decrease in the overall degree of rent-seeking in the economy after the about turn in economic policy in 1980 when compared to the previous protectionist ISI policy era. This does not mean though that the policy process itself was particularly conducive to the attainment of stated goals. Thus, we ask, what are the institutional conditions i/fluencing whether Turkish strategic industrial policies remain focused on publicly-oriented goals (e.g., growth or export competitiveness) as opposed to becoming focused on narrow, rent-oriented ones? This question is inspired by the observation that, irrespective of which side one takes in the debate over the merits of selective industrial policies, all sides agree and comparative empirical evidence demonstrates that such policies can be easily abused when private economic actors, economic bureaucrats and politicians are able to channel such policies toward narrow, private ends. The pernicious effects of such rent-seeking on economic performance are well-established.3 business networks. See The East Asian Miracle: Economic Growth and Public Policy (1993) and related bank policy papers especially those by Campos (1993) and Stiglitz (1993). 3 On rent-seeking with respect to quotas see the seminal article by Krueger (1974). A generalization of the argument is offered by Bhagwati (1982). See also Buchanan, Tollison and Tullock, eds. (1980) and Colander, ed. (1984) for various perspectives on rent-seeking and associated "new neoclassical political economy."  Introduction 3 A. Theoretical Orientations and Hypotheses It is important to clarify what is meant by key terms such "institutional foundations" and "effective delivery of economic policy" as well as to be precise as to hypothesized relationships. In the first place, the state-business relationship in the context of an incentive and subsidy program may be viewed as a form of "contract" which carries reciprocal obligations for both parties.4 Insofar as information asymmetries characterize this state-business relationship, the "moral hazard" exists that incentives and subsidies will become give-aways as business recipients may be able to use them with scant reference to their contractual commitments. This hazard becomes particularly acute if it is known that the state does not monitor recipient use of incentives and subsidies or that the state is unwilling to enforce agreed-upon performance standards of the contract (i.e., to discipline program abusers).5 If it is further known that there is no initial screening process based on technical criteria or, worse still, if there is in place a selection process based on political criteria, then this encourages an "adverse selection" process whereby less competent and sincere investors may crowd out the more competent and sincere ones. Finally, to the extent that the state is commonly known to allow investors to renegotiate the original terms of the incentive and subsidy contract this ex-atile reduces the incentives for the investor to fulfill contractual obligations. As Bolton notes, "the result is that renegotiation induces undersupply of effort" and "sometimes the agent gets a positive rent from the contract" (1990: p. 306. Emphasis in original). Thus, for the purposes of this paper, the effective delivery of economic policy may be defined as the extent of enforcement of the "contracts" associated with incentive and subsidy programs.' Such enforcement relates to whether or not agent performance matches commitments as well as to whether the original time frame has integrity. Enforcement may be accomplished by the state's economic bureaucracy or, as our research discovered to be the case in the readywear clothing industry in Turkey (discussed below in Section V.), it may be de-centralized through placement in an intermediate association (industry-based ones in the case of readywear clothing). Our first hypothesis, therefore, is that effective economic policy is positively related to the presence of a technical initial screening process, monitoring of incentive and subsidy programs and imposition of disciplinary consequences (e.g., penalties) on beneficiaries who fail to :neet mutually agreed-upon performance standards. While we presume that beneficiaries will generally shirk when able, it is important to note that contractual provisions may be partially immune to shirking problems in the presence of "community enforcement" social norms (Kandori, 1992). We return to the discussion of the importance of social norms in the discussion of policy networks below. 4 Thus, a principle-agent relationship exists between the state and business. On agency theory see Alchian and Demsetz (1972), Fama (1980) and Jensen and Meckling (1976). Non-technical overviews are provided by Moe (1984) and Perrow (1990). 5 Amsden (1989) discusses the importance in Korea of state monitoring and enforcement of performance expectations for business use of incentives and subsidies. 6 Note that, in line with the earlier discussion, this definition is neutral with respect to the relationship between effective policies and their contribution to economic performance. Indeed, following this definition, one can imagine highly effective policies which negatively influence economic growth and international competitiveness.  4 Institutional Influences on Economic Policy in Turkey This focus on the monitoring and disciplinary capability of the economic bureaucracy, however, may be putting the cart before the horse as it presupposes the existence of sufficient bureaucratic will to use industrial policies for public-oriented goals as opposed to for more narrow, rent-oriented ones. In this regard, a vigorous political economy literature emphasizes the importance of allowing economric bureaucrats to enjoy a relative insulation from direct interest- based pressures from business, labor or other groups in society as well as from pork barrel political pressures.7 Comparative research findings show that the relative presence or absence of such bureaucratic insulation is largely a function of institutional characteristics of policy-making bureaucracies. In particular, bureaucratic insulation in economic policy arenas is positively related to the presence of a small and centralized economic bureaucracy, endowed with well-defined authorities for certain policy areas, and characterized by "Weberian" traits such as stable career paths and meritocratic recruitment and promotion criteria. Such institutional characteristics allow bureaucrats to develop a shared identity and purpose which facilitates unified decision-making and deters corruption (Rueschemeyer and Evans, 1985). Thus, our second hypothesis is that, effective economic policy is positively related to the degree to which economic bureaucrats enjoy an insulation from interest-based and political pressures. While insulation from political and interest-based pressures helps bureaucrats deliver publicly-oriented policies, too much bureaucratic isolation from real-world issues may hinder policy conceptualization and slow down policy reformulation given changing conditions or policies which were poorly conceived in the first place. Thus, we draw on a final literature focused on how institutionalized state-business relationships--e.g., "policy networks"--influence the effectiveness of policy formulation and adjustment.! By networks we refer not only to formal consultations between bureaucrats and business persons or business associations but also to the presence and density of informal ad hoc meetings and communications. The relationship of networks to the effectiveness of policy delivery is complex as it depends on the nature of the network. We characterize as "growth-oriented" those networks in which there is a two-way information flow, established norms of reciprocity, honesty and trust, and, lastly, transparency in information-sharing and decision-making among network members. The two-way information flow may both enhance initial policy design as well as allow for more rapid and flexible policy adjustment by helping to ensure that information with respect to "changes in world markets, new trends in technologies, perverse effects of regulations domestically and abroad are all communicated to the bureaucracy by private sector agents" (Campos, 1993: p. 28). Established norms may help reduce transactions costs and may facilitate the resolution of collective action problems confronting network members in the presence of prisoner's dilemma- type situations where in the absence of such norms inefficient outcomes are likely to prevail (Arrow, 1974; Kandori, 1992; Putnam, 1993). And finally, transparency is important for it raises the cost, to both members of the state and the business community, of private rent-seeking through rendering such actions more easily observed by other network members. In this sense, the term transparency does not necessarily entail widespread public access to knowledge. But it 7 For recent examples, see Zysman (1983), Shapiro and Taylor (1990), Geddes (1990) and Schneider (1991). 8 The literature on networks is large and varied. See, for example, Katzenstein (1977), Wilks and Wright, eds. (1987), Haggard (1990) and Granovetter (1985).  Introduction 5 does suggest that actual network members are aware of information and decision-making processes. In contrast to growth-oriented networks, we characterize as "rent-oriented" those networks in which there is a two-way information flow and the presence, perhaps, of various norms such as trust, but in which the level of transparency is sufficiently low that some network members successfully develop narrow rent-oriented ties. A lack of transparency in this regard is particularly harmful to economic growth because "since the fruits of growth occur in the future, growth creates enormous uncertainty over how its fruits are going to be distributed" and therefore, given a lack of transparency, "this creates incentives for various agents to preempt their competitors by bribing pubic officials" (Campos, 1993: p. 28-29). Our final hypothesis therefore is that, the state's ability to deliver effective economic policies is positively related to the presence of growth-oriented policy networks and negatively related to the presence of rent- oriented networks linking economic bureaucrats and beneficiaries of incentive and subsidy programs. B. Summary of Main Findings While our findings were not wholly disparaging as regards the effectiveness of incentive and subsidy policy in Turkey, the main one is that the institutional foundations for industrial policy in Turkey render the incentive and subsidy regime quite ineffective. In particular, the enforcement of incentive and subsidy contracts is quite lax as we found evidence that contract renegotiation to extend or improve access to benefits is frequent, easy and regularly accomplished by business beneficiaries irrespective of whether or not they had met agreed-upon performance expectations. However, there were several variations to this general pattern of lax enforcement. First, enforcement for export incentive contracts was more effective than for investment incentive contracts. And second, an industry-based difference in terms of enforcement was observed with respect to the readywear clothing industry as intermediate associations imposed performance expectations upon their members in exchange for access to incentives and subsidies. With respect to our hypotheses, our findings strongly support the first one as the lack of enforcement of incentive contracts is associated with the fact that Turkish economic bureaucrats are generally unable or unwilling to monitor incentive and subsidy programs or to discipline the violation of performance expectations by business persons. Furthermore, little or no initial negative selectivity was demonstrated as to the particular recipients of the program as almost no screening--based on technical as opposed to political criteria--takes place as to the competence of the investor or to the feasibility of the proposed activities. Worse, still, we found evidence of strong incentives operating at the institutional level for economic bureaucrats to actually avoid generating information with respect to recipient performance. Given the lower transaction costs involved, however, we did confirm that monitoring capacity has been higher with respect to export, as opposed to investment, incentives and subsidies. And monitoring has been higher in the readywear clothing industry due to the fact that this task has been assumed by industry associations rather than being left in the hands of the economic bureaucracy. Our second hypothesis, that bureaucratic insulation leads to more effective policy, was also supported as we found that insulation and the associated shared identity and purpose among economic bureaucrats, while already low, was further undermined in the 1980's due to both an influx of "contract personnel" who bypassed established career advancement procedures and  6 Institutional Influences on Economic Policy in Turkey frequent political intervention in the affairs of bureaucrats running the incentive program. The use of contract personnel may have helped overcome a statist bureaucratic culture perceived to be hostile to contemplated market-oriented reform efforts. However, it also encouraged the "politicization" of economic bureaucratic functions thereby not only introducing particularistic political agendas into economic policy arenas but also leading to a shortening of the economic policy planning horizon. And finally, in terms of our hypothesis regarding policy networks, while rent-oriented networks characterized the auto and glass industries, we found an emerging growth-oriented network in the readywear clothing industry. In general, however, rent-oriented networks appear to be the norm with respect to the Turkish incentive regime, particularly for concentrated (auto) or monopolistic (glass) industries. Such networks help promote the preferences first of politicians (both government and opposition parties) and their bureaucratic political appointees and second the preferences of big business actors. Small business actors in such industries appear largely left out of the network loop as they don't possess the necessary connections. Career bureaucrats in turn appear to be caught in the middle between the powerful business and political actors. The non-transparency and associated unpredictability of the incentive and subsidy regime nurtures these rent-oriented networks as those business groups who are well-positioned attempt to minimize risk by privately seeking privileged connections with top decision-makers. The very success of rent-seeking, however, as it leads to regular and often arbitrary bureaucratic interventions in the everyday implementation of the incentive and subsidy regime, further exacerbates the lack of transparency and unpredictability. In the readywear clothing industry the growth-oriented policy network was characterized by a high degree of transparency in its operation. In particular, the industry associations which mediated between the state and the beneficiaries of incentive program established rules governing the distribution of incentives and subsidies which a) linked member access to benefits to their quantifiable export performance and b) made transparent and trustworthy the process through which members documented their performance. Such transparency in the rules reduced the uncertainties surrounding the future distribution of incentives and subsidies and this was instrumental in freeing members to engage in growth-oriented competition over exports as opposed to competing with each other in lobbying and rent-seeking. Thus, such "private sector economic governance" allowed for more effective economic policy than would have been possible otherwise. Such findings clearly fall in line with our hypotheses regarding the institutional foundations for effective economic policy. However, the essentially symbiotic relationship between the decline of bureaucratic insulation and the nourishment of rent-oriented policy networks suggests a partial revision of our theoretical framework to emphasize the strong interaction between these two institutional influences on the effectiveness of economic policy. Of course, the overall validity of the findings remain contestable given that our study only considered a single country case. And while the inter-industry comparisons allowed for the teasing out of nuances associated with the network hypothesis, the hypotheses concerning bureaucratic insulation, and monitoring and performance, were examined only in light of a single economic bureaucracy. Thus, further comparative research would facilitate a more thorough examination of the considered questions. To be most effective, such research should be constructed on the principle of maximizing the variation in the dependent variable--effectiveness of delivery of economic policy--as this provides  Introduction 7 the most promising methodological strategy. Cases of apparently effectively delivered policy (e.g., South Korea), cases of apparently ineffective policy (e.g., The Philippines) and cases which exhibit inter-industry or inter-sectoral variation in the effectiveness of policy (e.g., Jamaica) would be particularly useful. The remainder of the paper is organized as follows. Section II. provides an historical review of Turkish industrial policy and state-business relations. Section m. examines the effectiveness of the incentive and subsidy regime. This section further details the monitoring and disciplinary capacity of the economic bureaucracy and provides an evaluation of the performance of incentive regime beneficiaries. Section IV. discusses the relative lack of insulation in the economic bureaucracy particularly in those divisions associated with the disbursement of incentives and subsidies. Section V. considers the nature of policy networks in Turkey and explores examples of both rent-oriented and growth-oriented industry-level networks. The concluding Section VI. offers policy recommendations.  8 Institutional Influences on Economic Policy in Turkey  2 Industrial Policy And State-Business Relations In Turkey A. Historical Overview Unlike some developing countries in Asia, Latin America and Africa which emerged from colonial rule with relatively weak state capacities, the Turkish Republic which came into existence in 1923 inherited a strong "dirigiste" state tradition from the Ottoman Empire. While the existence of a centralized state tradition differentiated Turkey from the bulk of once colonized countries, the young Turkish Republic resembled other independent nations which came into existence after the second world war in the sense that a local entrepreneurial class did not exist at the time of independence. At the same time, the founders of the new Republic were fully aware of the fact that the economic and military superiority of Western nations (which had militarily defeated the Ottomans) primarily rested on technological prowess which in turn was a function of the existence of a dynamic and innovative business class. Consequently, in an attempt to replicate the Western trajectory, the founders of the new nation decided to foster capital accumulation and create a "national business class" which was supposed to perform the entrepreneurial functions necessary for development. As a result, the bureaucratic elite which ruled the country during the 1923 to 1950 single party period assumed the task of transforming the economic structure through direct participation in capital formation either via the medium of State Economic Enterprises (SEEs), which were established in the "etatiste" period of the 1930's, or via a number of public-private joint ventures in many industries. During the same period, top bureaucrats in Turkey clearly formed a distinct, well paid and highly esteemed status group. The adoption of an import substitutions cum protectionist development strategy in Turkey coincided with the transition to a multi-party democracy (1946) and the unseating of the Republican People Party Government in 1950 by the right of center Democratic Party. Although the Democratic party Government only lasted for 10 years until it was overthrown by a military coup, various governments in Turkey which ruled between 1950 and 1980 did not alter the basic parameters of economic policy. That is to say, apart from the fact that the public sector typically accounted for more than a half of fixed capital formation, the state indirectly intervened in the operation of the market mechanism via heavy tariffs and quotas, an overvalued exchange rate, price controls and generous investment subsidies. At the same time, however, civil servants in Turkey began to lose their privileged status. The demise of civil servants was especially 9  10 Institutional Influences on Economic Policy in Turkey accelerated in the 1970's when the country was mostly ruled by coalition governments which were heavily engaged in unrestrained patronage and nepotism. The resulting reshuffling of bureaucrats in an arbitrary fashion involved all ranks, politicized the bureaucracy and undermrined the "esprit de corps." Following a protracted crisis of the import substitution cum protectionist development regime in the late 1970's, a major transformation of the economy was initiated in 1980 in the immediate aftermath of the third--and hopefully last--military coup. This new strategy, which gained momentum after the return to democracy and the triumph of the neo-liberal Motherland Party Government in the polls, was inspired by an outward-oriented development vision. The economic strategy aimed to remove price controls, privatize the bulk of the SEEs, eliminate major distortions except export promotion measures, and decentralize the economic decision-making process by empowering local governments. In the meantime, in an attempt to circumvent what it perceived as bureaucratic resistance to its minimalist state vision, the Motherland Party Government hired highly paid "contract personnel" to staff the upper ranks of key economic agencies and also created new agencies and autonomous units within the existing agencies to bypass the traditional policy institutions such as the State Planning Organization. On one hand, such administrative reforms gave a freer hand in policy making to the Motherland Party whose leaders relied on their close allies in the bureaucracy, the typically U.S.-educated "princes", in the design and implementation of reforms. On the other hand, because strategic policy powers were centralized and decisions hastily taken by a small number of technocrats allied to the Prime Minister without much consultation with private actors or other public officials, administrative reform measures further added to the overwhelming uncertainty regarding the "rules of the game" with regard to the established norms and procedures in economic policy making. Consequently, such actions led to increasing fragmentation and balkanization of the economic policy apparatus and high turnover of the top personnel who often opted for private industry. Effective policy design and implementation has been crippled not only by the absence of coordination among state economic agencies but also by the existence of private sector cleavages. Most notable is the fact that Turkish business in general and industrialists in particular are devoid of a peak all-encompassing umbrella organization able to aggregate various interests, attempt to shape policy, and function as a conflict-resolution mechanism between business firms. The absence of such an organization is partly due to the fact that as private business in Turkey has developed under the explicit guidance of the state, the political elite, at least until the early 1970's, was reluctant to consent to the formation of independent organizations. Instead, the Union of Chamber and Commodity Exchanges of Turkey, which dates back to 1950, was organized at the behest of the government with the stated task of undertaking certain quasi-legal functions--such as responsibilities with regard to foreign exchange allocation and export processing--and the unstated task of supporting the then reigning Democratic Party Government. The politicization of the Union reached extreme proportions in the 1960's and 1970's, i.e., the heyday of protectionist economic policy, when the reigning right-of-center Justice Party would prepare the candidate lists for elections to the Board of Directors of the Union as well as affiliated chambers. This situation and the resulting nepotism also led to the disaffection of large industrialists who were often bypassed in foreign exchange allocation decisions in favor of merchants and small-scale businessmen (Oncu, 1980). Consequently, the large industrialists representing major conglomerates, which are active in various branches of industry as well as banking and commercial activities, successfully pressured the military government in 1971 to take foreign  Industrial Pollicy and State-Business Relations in Turkev 11 exchange allocation out of the hand of the Union and place it under the authority of various Ministries. In addition, on August 2nd, 1971, eighty six big industrialists founded the first voluntary business association in Turkey--The Turkish Industrialists' and Businessmens' Association (TUSIAD). Although TUSIAD's membership and activities increased dramatically over the years, a number of factors cripple this organization's capacity to represent effectively big business, let alone the whole business community. This is because, for one thing, TUSIAD is very much perceived as a "club of the rich" and resented by other members of the business community. The Union of Chambers, for instance, does not welcome this Association's prominence in economic affairs and resists the adoption of certain policy initiatives proposed by TUSIAD such as the formation of a formal concertation mechanism between the state, business and labor, modelled after the French example of an Economnic and Social Council. Similarly, anti-TUSIAD feelings have also been widespread among the state elite which often has denounced the Association's economic policy proposals on the grounds that TUSIAD was not minding its own business. In addition, frequent friction between conglomerates which are members of this association, together with the existence of cleavages between large exporters and domestic market-oriented industrialists with regard to which policies TUSIAD should support, often interferes with the collective action capacity of this organization. Last, but not least, as is reported by S. Gulfidan (1993), who undertook a multi-faceted study of this organization, the most powerful members of TUSIAD, who are well represented in the Executive Committee, often found it expedient to bypass their own association in favor of personal contacts with top bureaucrats and politicians in order to find solutions to the economic problems which concern them. Consequently, despite the pleas by TUSIAD's leadership to establish tightly institutionalized and transparent channels of contact between government and business circles, the particularistic efforts on the part of TUSIAD members breed public cynicism conceming the sincerity of such proposals. In the meantime, following the about tum to a market-oriented economic policy in the 1980's, there has been a proliferation of various industry associations at the subsectoral level. These independent associations are generally established by the initiative of leading producers in the relevant industry and provide effective channels for their members to influence economic policy of local and national govemrnments. In addition, these associations often attempt to function as transmission belts between industry and government and, when necessary, they also serve as a conflict-resolution mechanism between firms in the given industry. Of course, the efficacy and developmental role of these associations varies. At worst, they can approximate what M. Olson calls a rent-seeking "distributionist coalition" which attempts to circumvent the market in order to fix prices and act in a cartel-like fashion to weaken competitors. On the other hand, under certain circumstances which will be specified later when contrasting the Association of Automotive Manufacturer's and the Readywear Manufacturers' Association, these organizations can accelerate the creative potential of the market mechanism by generating relevant information and facilitating the adjustment of their members to the requisites of an outward-oriented economy. B. Main Elements of the Incentive Regime During the time period which is the focal point of this study, i.e., the 1980's and onwards, these industry associations were also active in the pursuit of seeking maximum advantage for their members from various government-backed incentive and subsidy programs. In fact, even after the  12 Institutional Influences on Economic Policy in Turkey radical turn in economic policy, the Turkish state continued to provide formal incentive programs for private business, very much like in the pre-1980 period. There was even a procedural continuity in the sense that two formal incentive programs--the investment and export incentive regimes--continued to be officially announced each year. However, even though export incentives had been in operation since the early 1960's, until the early 1980's these incentives were very limited and could be regarded as an attempt to compensate for the distortions provided by an overvalued exchange rate. In the post-1980 period, on the other hand, not only were these incentives widely used, but also new instruments were added to the list of pre-existing investment and export incentives. Although the ostensible aim of the incentive program consisted of easing the burdens of adjustment, some researchers detected a quasi-political rationale aimed at rendering the reforms politically palatable by creating direct stakes for them in the ranks of domestic market-oriented industrialists and by broadening the scope of the beneficiaries. The existence of such a political rationale is hard to deny but, on the downside, the fiscal costs of the incentive programs amounted to 2.2 percent of the GNP for investment incentives and about 2.4 percent of the GNP for export incentives in 1988 (Arslan, 1993). In addition, throughout 1980's the incentives tended to become more widespread rather than more restricted primarily not only because modifications to the system were resisted by the beneficiaries but also because the maximization of the discretionary powers of the state proved expedient to engender "policy generated" rents which could be selectively allocated to reward friends and punish foes. The main elements of the incentive programs which correspond to the period under study are quite hard to pin down since these elements have kept changing. To give an example, the "exemptions from financial tax and stamp duty" on investments in priority development regions were introduced in 1984 and were still in operation in 1992; whereas the foreign exchange allocation scheme was only in operation between 1981 and 1989. This said, Table 1 below sums up the main elements of the export-oriented phase of economic development even though some changes have occurred during this period. As may be gleaned from this table, the tools through which exports and investments were subsidized have been numerous. The existence of these tools gave a great deal of discretionary power to officials, most specifically in those in an autonomous Directorate in the State Planning Organization in 1991 and in the Treasury until 1994, whose members were in charge of approving the "investment encouragement" and "export encouragement" certificates. The certificates effect the government budget in three ways: through direct cash outlays; through foregone tax and tariff revenue; and through the lending of funds to financial intermediaries at rates lower than the government's own borrowing rate.  Industrial Pollicy and State-Business Relations in Turkey 13 Table 1: Main Elements of the Incentive Program Instruments Investment Incentives Export Incentives Tax incentives Customs duty exemptions Customs duty exemptions Investment tax reductions Corporate income tax reductions Taxes, duties, and credit charge exemptions Financial transaction tax and VAT exemptions Deferment of tax payments (Finance Fund) Building and construction tax exemptions Value-added tax refunds Additional employment incentives Exemptions from social security, compulsory savings, and housing fund premium contributions Foreign exchange allocations Trade incentives Foreign exchange allocations Temporary import permits Credit Preferential credit (fund based) Preferential credit (Exim Bank) Direct subsidy Energy incentives Energy subsidies Land allocations Transportation premiums Investment subsidies Export tax rebates Source: Various Decrees published in the Official Newspapers.  14 Institutional Influences on Economic Policy in Turkey  3 Selectivity And Monitoring In The Incentive Regime In this section we consider the process through which business applicants obtain access to the Turkish incentive regime as well as the effectiveness of incentive policy. With respect to effectiveness we focus upon a) the nature of initial screening and selectivity as regards applicants, b) the monitoring and discipline of incentive beneficiaries, and c) the performance of these beneficiaries. Our principal goals are to evaluate the effectiveness of the incentive and subsidy program as well as our hypothesis regarding the importance of the state's initial screening, monitoring and disciplinary capacity on effectiveness. The effective delivery of economic policy, it may be recalled, was defined as the enforcement of the "contracts" associated with incentive and subsidy programs. Our findings, reviewed below, are that the incentives and subsidies are not delivered effectively as contracts are routinely violated by business recipients without consequence and as renegotiation of contracts by recipients is easy and common. Moreover, such ineffectively delivered policy is, as hypothesized, related to a) the state's lack of an initial screening of applicants, b) the presence of political, rather than technical, selectivity criteria influencing levels of access to program benefits, c) the state's inability and unwillingness to monitor the performance of beneficiaries and d) the lack of disciplinary consequences associated with contract violation or unwarranted renegotiation on the part of business beneficiaries. Let us note that our findings go against the commnon criticism that the Turkish incentive regime has been hampered principally by an overall absence of selectivity by economic activity (Guvemli, 1992). Instead, we argue that the problem is not a lack of selectivity but is rather the particular nature and dimensions of the existing forns of selectivity. While few applicants are initially turned away, thus suggesting that demand for certificates drives their overall pattern of distribution, considerable selectivity is nonetheless exhibited as a result of the widespread practice of ad hoc negotiations between business applicants and the bureaucratic and political elite over the particular termns and duration of incentive contracts. Such negotiations are not always linked to transparent and technical criteria for evaluating the merits of a particular application but rather often reflect the political access and clout of parties to the negotiations. Such political, as opposed to technical, selectivity, as discussed in Section IV. and Section V. below, may be attributed largely to the relative lack of insulation in the economic bureaucracy and to the prevalence of rent-oriented state-business policy networks in Turkey. 15  16 Institutional Influences on Economic Policy in Turkey A. Application Process for Incentive Regime As noted earlier, the existence of both investment and export incentives in Turkey has been rooted in the government's objective to influence the patterns and rate of industrial development in accordance with the priorities of five-year development plans. That is why in 1968 a specific bureau, the Directorate of Incentives and Implementation (TUD), was established within the State Planning Organization and entrusted with the task of evaluating applications for receiving various incentives and subsidies. The approval of these applications and the reception of "certificates of encouragement" entitles applicants, at least in theory, to receive benefits such as cash grants and subsidized credit as well as certain tax breaks and custom duty exemptions. The application process itself involves the submission of forms detailing planned investment and/or export related activities and, at least in theory, the state claims to make a "case by case analysis of each investment application accompanied by a feasibility study and financial projection" (Ministry of Industry and Trade, 1991: p. 32). It is important to note that there can be considerable variation by applicant with respect to the types and levels of incentives or subsidies granted. While formal approval of an application is a necessary step, it is not always a sufficient one to gain access to all of the promised benefits. This is the case as certain benefits, such as preferential credits, are disbursed by the state-owned development banks and, given the fact that there commonly has been a gap between the actual resources and the certificate backed claims on these resources, many beneficiaries of encouragement certificates have ended up only partially enjoying the benefits to which they were theoretically and legally entitled. The State Planning Organization's records indicate that from 1968, when the TUD was established, until the end of 1980, almost 5,000 investment encouragement certificates were issued. This number skyrocketed during the next decade, between 1981 and 1991, when 27,000 certificates were issued. The corresponding numbers for export incentive certificates is harder to compile since until 1980 various export incentives, such as subsidized credits, import duty exemptions and foreign exchange retention, were administered by different ministries. In 1980, however, the administration of export incentives was given to the TUD of the State Planning Organization. And in 1991 responsibility for the administration of both export and investment incentives was entrusted to the Treasury by the wholesale transfer of the TUD to this agency. Since neither the State Planning Organization nor the Treasury has kept records of total applicants and acceptance and rejections rates, it is not possible to quantitatively document to what degree the massive increase in the number of encouragement certificates issued throughout the 1980's represents a decline in the rigor and selectivity of the application process as practiced in the previous period. It may well be that given the about turn in industrial policy in 1980, the number of applicants dramatically increased. This said, it is clear that as far as business is concerned obtaining an encouragement certificate has not only been easier in the 1980's as compared to the earlier import substitution period but has become almost an automatic process. On one hand, prior to 1980 state officials who evaluated applications for certificates also considered them from the vantage point of a "fit" with the priorities of five-year development plans and the yearly investment programs which were prepared on the basis of development plans. Thus, although these were only indicative for the private sector, the development plan's priorities weighed on the decisions concerning which projects would enjoy privileged access to the state's largesse. In the post-1980 period, on the  Selectivity and Monitoring in the Incentive Regime 17 other hand, as our interviews both in the public and private sectors has indicated, the conformity between the plan's objectives and the proposed investments was no longer a criterion which affected decisions related to the issuance of certificates. On the contrary, decision-makers in the TUD took pride in the fact that the system in now both simple and rapid (certificates are issued within two to three days once applications are submitted) and, while regional priorities are still relevant, there is no discrimination on the basis of the industry to which the proposed investment would be directed. These claims are in accord with an internal World Bank Memorandum (1991) which found that "while incentives are not automatic, we have not heard reports of any exporters being turned down." The formal procedure to -obtain encouragement certificates is quite straightforward. Obtaining a certificate entails, in the case of export incentives, the undertaking of a quantifiable export commitment on the basis of which exporters become eligible for various subsidies. Similarly, investors undertake a host of commitments regarding various targets such as employment levels, fixed investments, and so on, the realization of which, at least in principle, the granting of incentives depends. Investment incentives are identical for both Turkish and foreign investors even if the administration of the system is divided, i.e., domestic projects fall under the purview of the Incentives Department and projects with a foreign investment component fall under the responsibility of the Foreign Investment Department, both of which are separate units of the Treasury. A. 1. Transparency and Predictability The Turkish incentive regime, at least on a formal basis, is quite transparent. Every year the TUD publishes two separate booklets--"Export Encouragement Procedures" and "Investment Encouragement Procedures"--which lay out the legal framework and available benefits. It then sends these booklets to various business organizations. These documents are also published in the Official Gazette. Furthermore, prior to the publication of these documents Treasury officials canvass the opinions of the Chambers of Trade and Industry and attempt to address the concerns of various private actors. In practice, however, the apparent transparency of the incentive regime is compromised for two reasons. First, the principles of the incentive regime laid out in the two documents change frequently throughout the year based on the publication of various decrees which alter the basic parameters of the regime in an ad hoc fashion. The sheer abundance of these decrees, which are issued by the Council of Ministers and also published in the Official Gazette, introduces considerable uncertainty into business life. Additionally, the frequent changes generate incentives for business lobbying efforts to focus not only on influencing the conceptualization of the basic properties of the incentive regime but also to focus at the later implementation stage in order to curry special favors. At times, the particularistic granting of such favors reaches dramatic proportions and creates public scandals when covered in the daily press. Bugra (1994: p. 152-53) reports on two egregious examples of such changes designed to do special favors for particular investors: One of these cases had to do with the decrease from twenty-five to fifteen cents of the payments made to a particular fund by the exporters of dried figs. The decision for this change was taken at the end of the export season for this commodity, and the only person likely to benefit from it was a particular exporter who still had a certain  18 Institutional Influences on Economic Policy in Turkey amount of this export commodity to be shipped abroad. This exporter happened to be a close friend of the Minister of Finance the future father-in-law of Turgut Ozal's son. The second similar case involved the neglect to erase a particular commodity, plastic slippers, from the list of commodities whose export premiums were reduced. Hence, for a period of about two weeks, exporters of plastic slippers received premiums which could often well exceed the cost of production of the commodity. Although the incident received a lot of press coverage, the lucky exporter for whose benefit the scheme was designed could not be discovered. And second, the transparency of the incentive regime is compromised due to the fact of ambiguities in the classifications and definitions of the benefits to which business is entitled in exchange for carrying out the promised export and investment related activities. To be specific, investment certificates were given for a number of activities which may overlap.9 What is the distinction between "expansion" and "modernization" investments for instance? How does one know whether a particular investment falls into one or the other? This is indeed an important decision since there are different benefits attached to these classifications. But even the definition of these benefits have not been devoid of ambiguities. Suppose.a particular benefit stipulates that when firms undertake a "completion" investment, they can import all raw materials without custom duties, but for unprocessed materials they will pay only reduced duties. But, as asked rhetorically by a Treasury official, "how clear is the distinction between "unprocessed" and "raw" materials? Does oil fall in the former or the latter category?" The frequency of changes in the list of commodities benefitting from different types of export promotion measures and also the ad hoc changes in the percentage of premiums and tax rebates attached to investment and export benefits create incentives to sharpen the skills of businessmen in what Bhagwati (1982) calls "directly unproductive activities" thereby crowding out in due process more dynamic elements of the private sector who refuse to play this game. Equally pernicious is the fact that, to the extent that legal and definitional ambiguities enhance the scope of discretionism available to decision makers, the transaction costs of collecting information to discover the "real" intention of policy-makers accord undeniable advantages to large conglomerates--as opposed to small and medium-sized industrialists--which can afford to hire personnel who can conduct business in Ankara and understand the often scholastic details of the official decrees and declarations. B. Effectiveness of Incentive Regime Next we turn to the crucial issue of the effectiveness of the incentive and subsidy program. As noted, our approach is to treat the granting of an investment or export certificate by the state and its acceptance by a business person as a form of contract which carries reciprocal obligations for the two parties. That is, the state commits to grant various incentives and subsidies to a particular business person in exchange for a commitment by that business person to undertake specific forms of business activity. The investment and export certificates themselves are not very detailed documents and consist of about one page each which spells out the type of benefits to be provided (which vary 9 For a listing of these different activities refer to Table A4 in Appendix I.  Selectivity and Monitoring in the Incentive Regime 19 from recipient to recipient) to the business person by the state. However, when applying for such benefits, business persons fill out detailed forms specifying various actions that they commit themselves to undertake. In the case of exports such actions consist of a commitment to export at a certain price a certain number of items and/or volume of goods. In the case of investments, a business plan is attached detailing the level of fixed investments, investments in machinery, new employment creation, and so on. Legally speaking, these commitments bind the certificate holders since in the case of both export and investment incentives they sign a letter, endorsed by a Notary Public, which indicates that in case of non-realization of the targets, except for reasons outside the control of the certificate holders (these reasons are also specified), or in the case of diversion of benefits to areas other than the proposed activity, certificate holders agree to pay back all of the subsidies they received to the Turkish Development Bank together with accrued interest and penalties.'" Hence, the legal basis of the Turkish incentive system is quite clear as to the nature of the contract, as to what constitutes its non-realization on the part of business persons, and as to the specification of different contingencies and subsequent measures which will apply in case of non- compliance. The problem then lies not with the contract per se but with its enforcement. When so conceptualized it becomes quite natural to inquire as to how effectively the state initially screens applicants, monitors their performance with respect to the "contract" and imposes disciplinary consequences following poor performance. B.1. Initial Screening of Applicants A first focus is with respect to if there exists an initial screening process, based on transparent and technical criteria, for investment and export certificate applicants. Insofar as little such screening takes place certain applicants will be successful in their applications despite, perhaps, their having too little commitment, preparation or competence to undertake the proposed project. One should, nevertheless, not jump to the conclusion that rational incentive regimes require screening at all costs. One could surmise that if a state lacks the requisite autonomy and capacity to anchor applicant screening on transparent and technical criteria, such as performance and competence, it may well be better to act like a generous traffic light giving a green signal to all. This, at least, would prevent bureaucrats from intervening in the incentive application process in an arbitrary and subjective and therefore unpredictable manner. While there is some merit in this perspective, insofar as an inability or unwillingness to screen applicants is publicly known or, worse still, as we contend is the case in the Turkish incentive regime, if there is a screening process based not on transparent and technical criteria but rather on political criteria, this nourishes the growth of a virulent "adverse selection" process. Such adverse selection means both that underqualified and insincere applicants are not deterred from applying and that there is generated an incentive for business applicants to sharpen their skills in directly unproductive activities. While the type of quantitative data one would want to fully investigate the nature of initial screening, such as the applicant rejection rates by industry, by year and by reason for rejection, are 10 The letter itself follows a standard form whose format is given in the booklet the Incentive Regime, published each year by the Treasury, separately for investment and export incentives (see, for instance Yatirimlari Tesvik Mavzuati, 1993, Appendix 4: p. 99).  20 Institutional Influences on Economic Policy in Turkey not available, certain comments may nonetheless be made. In the first place, there is little or no negative selectivity in initial screening as the general understanding among both business actors and economic bureaucrats is that incentive and export certificates are easily obtained. Interviews with SPO and Treasury officials, for example, clarified that as soon as exporters register their intention to export, incentive certificates could be obtained within a matter of weeks or even days. The private beneficiaries of the certificates, on the other hand, have verified these claims, but complained that, although approval was almost automatic, the whole process could sometimes be fairly slow. Thus, at least for larger investors, it was common to seek the assistance of an influential broker, such as a government minister or an ex-high level official, in order to speed up the process (Author's interviews, 1994). While at first glance the lack of rejected applicants suggests an overall lack of selectivity, it would be a mistake to reach this conclusion as the real selectivity at the initial stages of the application process concerns the wide variability in the types and generosity of incentives and subsidies promised by the state. Circumstantial evidence based on our interviews with various business persons indicates that political influence on the decisions concerning the actual beneficiaries of the incentive regime occurs at two stages. First, influence may be exerted with respect to definitions of the types and levels of benefits an application will carry. Not surprisingly, this is particularly common with respect to contemplated investment or export related activities which are of considerable magnitude. Thus, to take a recent example involving the auto industry, considerable bargaining and haggling over the type and level - of benefits characterized the encouragement certificate approved for the joint venture between Toyota and the Sabanci Holding in the auto industry. However, even more modest applicants, because of the politicized nature of the process, have strong incentives to engage in lobbying and to enlist tactics such as the exaggeration of the importance of their proposed project. Thus, according to a high-ranking Treasury official: Finns routinely prepare three business plans for their investments. First, they prepare the real one for themselves. Second, they prepare an only slightly inflated one for their overseas partners who they know will spend at least some effort evaluating it. Andfinally, for us they prepare a most inflated plan as they calculate this may help them receive benefits and as they know it cannot be reviewed adequately. And second, and perhaps most importantly, political influence has commronly been exerted during the immediate stage following the approval of a certificate by the Treasury when various certificate holders vie with each other to actually obtain the promised benefits, especially ones associated with access to cash grants or subsidized credit. At this stage, particularly given that at various times some of the funds--such as the Resource Utilization Fund--may not be adequate to cover all existing claims on them, access to benefits becomes partly a matter of having political connections and clout. To repeat our main point in this regard, significant selectivity does occur in the application process of the Turkish incentive regime. However, such selectivity is not based on transparent and technical criteria but instead is based on subjective political criteria. B.2. Monitoring of Performance Besides the issues of initial screening and selectivity, a second important focus is on whether or not, and in what fashions, the state monitors the subsequent behavior or "performance" of  Selectivity and Monitoring in the Incentive Regime 21 incentive or export certificate recipients. In particular, insofar as the certificate represents a form of contract in which specific commitments have been made by the business recipient, it is important to inquire as to the capacity and willingness of the state to monitor business compliance with these commitments. Such monitoring is particularly germane given that it is frequent that business recipients desire to extend the life of the original certificate. Certainly, one can imagine many contingencies in which such an extension is sensible as even with good faith efforts progress may fall short of earlier anticipations. However, insofar as the benefits provided by the state in the certificate, such as access to various incentives and subsidies, are costly, it raises the crucial question as to whether or not an extension--what we refer to as a renegotiation of the contract--is warranted. Put otherwise, easy renegotiation of the benefits associated with investment and export certificates, in the absence of any ability by the state to determiine whether or not the business recipient is in compliance or is even attempting to comply with the original contract, allows for the possibility of state subsidization of continuing noncompliance. In this regard, statistical data provided by the Treasury reveals extensive renegotiation of incentive contracts. In particular, statistical records are kept on the differing statuses--"closed," "open," "discarded" and "penalized"--of investment and export certificates subsequent to their being issued. At issuance certificates carry anticipated completion dates. While there is variation in such completion dates depending on the contemplated activities, a "normal" investment certificate contract lasts about three to five years while an export certificate contract lasts a bit shorter at about one to three years. Table 2 below provides details on the status of investment certificates while Table 3 refers to the status of export certificates. Table 2: The Effectiveness of Monitoring Investment Incentive Certifications, 1980-1 994: The Percentage Distribution of Closed, Discarded and Open Investment Incentive Certificates Year Readywear Glass Automobile Total 1980 Closed 75 38 19 Discarded 38 31 Open 25 24 50 1981 Closed 35 25 30 15 Discarded 30 25 30 60 Open 35 50 40 25 1982 Closed 44 75 63 28 Discarded 22 31 Open 34 25 37 41 1983 Closed 54 100 76 43 Discarded 15 6 18 Open 31 18 39  22 Institutional Influences on Econonic Policy in Turkey Table 2: The Effectiveness of Monitoring Investment Incentive Certifications, 1980-1994: The Percentage Distribution of Closed, Discarded and Open Investment Incentive Certificates Year Readywear Glass Automobile Total 1984 Closed 53 100 57 52 Discarded 18 9 16 Open 29 34 32 1985 Closed 57 92 36 58 Discarded 24 45 21 Open 19 8 19 21 1986 Closed 59 60 59 53 Discarded 8 20 14 12 Open 33 20 27 35 1987 Closed 47 74 58 45 Discarded 9 21 5 8 Open 44 5 37 47 1988 Closed 53 68 53 41 Discarded 6 12 9 7 Open 41 20 38 52 1989 Closed 43 37 31 33 Discarded 12 3 7 Open 45 63 66 60 1990 Closed 11 63 28 20 Discarded 16 4 5 Open 73 37 68 75 1991 Closed 12 10 18 16 Discarded 22 5 6 Open 66 90 77 78 1992 Closed 3 5 11 8 Discarded 2 Open 97 95 89 90 1993 Closed 1 5 11 8 Discarded 0.5 Open 99 96 96 98  Selectivity and Monitoring in the Incentive Regime 23 Table 2: The Effectiveness of Monitoring Investment Incentive Certifications, 1980-1994: The Percentage Distribution of Closed, Discarded and Open Investment Incentive Certificates Year Readywear Glass Automobile Total 1994 Closed Discarded Open 100 100 100 100 Total Closed 33 44 37 29 Discarded 10 6 5 15 Open 57 50 68 56 The figures display the percentages of certificates which have been "closed," discarded," or retained "open" as of June 1994. Source: Treasury and SPO. To help clarify the meaning of the different statuses of these certificates let us note that an "open" certificate is one in which the business recipient retains access to incentives and subsidies. A "closed" certificate is one in which access to incentives and subsidies has been terminated because the recipient informs the state that the contract (i.e., investments or exports) has been fulfilled. A "discarded" certificate is one in which in which access to incentives and subsidies has been terminated because the recipient has acknowledged inability to fulfill their end of the contract for some reason (we may speculate that these generally relate to matters beyond the control of the recipient). A "penalized" certificate is one in which the state terminated the contract after it was discovered that the business recipient was not undertaking committed activities despite efforts to draw on benefits. The explanation of the "open" status for older certificates warrants further detail. A certificate that remains "open" for a number of years after the "normal" time frame for the expiration of the contract may imply a number of things: that the beneficiary has completed the planned investment or export activity; has begun but failed to finish the planned activity; or has not yet begun the planned activity. But in all cases it is clear that the beneficiary did not choose to close the certificate thinking that it was currently useful or may prove useful in the future and that the state has not chosen to terminate the beneficiary's access to benefits. Thus, while it is true that one cannot deduce from a certificate which has remained "open" past the "normal" time period that planned activities have not been or will not be completed, one may say that the holders of incentive certificates have been successful in renegotiating the terms of the original contract with respect to its duration. Such renegotiation of the certificate contract is a serious matter. In the first place, when a certificate remains open beyond the length of the original contract, beneficiaries may continue to access benefits irrespective of whether or not they have complied with their original commitments. Moreover, it is not uncommon for the benefits available through a certificate to actually be  24 Institutional Influences on Economic Policy in Turkev increased at the renegotiation stage. As noted in the theoretical discussion in Section I, easy and frequent recourse to such renegotiation has the consequence of inducing beneficiaries not to put forth full effort to fulfill their commitments. Thus, as a particularly dramatic example, it may be noted that direct subsidies for development of hotels were available through the Resource Utilization Fund disbursed by the Turkish Development Bank and that massive abuses have been reported in the press where individuals were continuing to access such subsidies with "open" investment certificates even though ground had never been broken for the proposed developments. In the second place, renegotiation is a serious matter as even in the case of beneficiaries who have complied with the original planned investments and exports, their continued access to an "open" certificate means that they may continue to receive benefits despite having met the original expectations of the certificate "contract." The absurdity of this situation from the point of view of the state and its priorities is clear. If the incentive certificate was a "carrot" to encourage either new investments or exports or to shift the sectoral composition of investments or exports then it is irrational to continue subsidizing beneficiaries after they have done what was desired. At the same time, the state's investment or export priorities may change over time and thus certificates which remain "open" over extended periods of time may actually foster situations where the state is subsidizing at cross-purposes with itself Overall, Table 2 reveals a grim picture which is indicative of a profound lack of monitoring by the state and, therefore, of an open invitation to business persons to renegotiate the terms of their contracts. Look, for instance, at Readywear Clothing for the years 1980 and 1981. As of 1994, i.e. many years after the "normal" three to five year completion date of the investments, 25 percent of certificates for 1980 and 35 percent of certificates for 1981 still remain open. And since 30 percent of certificates were discarded for 1981, the percentage of closed investments-- that is, those for which the beneficiary has unambiguously held up their end of the contract--was a meager 35 percent. A strong inference which may be drawn from the table is that, since firms were not penalized for noncompliance with the initial contract, it follows that "renegotiation" of the initial terms was a relatively automatic process at the discretion of business groups. This makes the diversion of incentives to other areas highly possible. For example, a firm which is granted "custom duty exemption" for the importation of a certain raw material to be used in the production of good X in an underdeveloped region, can, without penalty, either import this raw material without a custom duty to be resold in the internal market or can undertake its promised activity in a more desirable location, instead of in the underdeveloped region. When the latter happened, it is clear that an important objective of Turkish planning throughout the years, i.e. the objective of moving industry from developed into less developed regions via attractive tax and financial incentives, could not be attained. In practice, businessmen do not think they are bound to comply with the terms of the agreement. Indeed, in interviews businessmen acknowledged the feasibility of getting an attractive investment certificate stipulating that the investment would be made in an underdeveloped region and then diverting the investment to a more attractive area since they are aware of the fact that there was no penalty attached to such practices. The Directorate of Incentive Certificates officials, in return, have rationalized their apathy in monitoring performance by the fact that they have not had sufficient personnel with appropriate experience (Author's interviews, 1994).  Selectivity and Monitoring in the Incentive Regime 25 In contrast to investment incentives, the monitoring of export incentives, as can be seen from Table 3 below, was more successful. One straightforward reason for this is that it was easier to formulate unambiguous performance standards given that an externally validated measure of success exists in the case of exports. In addition, the cost of monitoring exports incentives was lower as, an interviewee in the Treasury put it, it is easier to monitor performance "by sitting in one's desk." This is because one can track performance by obtaining two documents: the custom form proving that the commodity in question had been exported and the document from the Central Bank verifying that the counterpart in foreign exchange was sent. Of course, as happened with export rebates, it was possible to manipulate the process, overstate the quantity shipped, and then transfer the foreign currency from different sources. It is not easy to quantify the overall extent of such abuse as has been done with export tax rebates (Rodrik, 1988). The Treasury officials to whom we spoke thought that most blatant abuses were engineered by holding companies which own offshore foreign trade companies and, when they import machinery and raw materials through them, they overprice in order to benefit from the corporate tax deduction. It is possible to draw on similar benefits by swelling the value of goods exported, and therefore the table below may have a bias towards exaggerating the monitoring capacity. Another important difference between export and investment incentives has been the state's recourse to "penalties" when it was proven that a firm obtained an export certificate with no intent to export. Table 3: The Effectiveness of Monitoring Export Incentive Certificates, 1985-1994: The Percentage Distribution of Closed, Discarded, Open and Penalized Export Incentive Certificates Year Readywear Glass Automobile Total 1985 Closed 84 76 85 81 Discarded 6 9 8 Open 2 6 Penalized 8 24 6 5 1986 Closed 85 63 88 76 Discarded 10 25 7 9 Open 3 1 12 Penalized 2 13 4 3 1987 Closed 89 100 89 78 Discarded 5 5 11 Open 1 2 4 Penalized 5 4 7 1988 Closed 82 69 86 74 Discarded 6 13 6 10 Open 10 18 7 14 Penalized 2 1 2  26 Institutional Influences on Economic Policy in Turkey Table 3: The Effectiveness of Monitoring Export Incentive Certificates, 1985-1994: The Percentage Distribution of Closed, Discarded, Open and Penalized Export Incentive Certificates Year Readywear Glass Automobile Total 1989 Closed 76 68 74 73 Discarded 6 22 5 7 Open 16 10 19 18 Penalized 2 2 2 1990 Closed 71 65 63 67 Discarded 4 3 1 6 Open 24 29 36 26 Penalized 1 3 1 1991 Closed 58 41 56 54 Discarded 3 2 5 Open 38 59 42 40 Penalized I 1992 Closed 31 36 32 31 Discarded 3 3 4 Open 66 64 65 65 Penalized .2 .3 .4 1993 Closed 6 15 4 5 Discarded I 1 Open 93 85 95 94 Penalized 1994 Closed 9.2 0.1 Discarded 0.2 2 0.3 Open 99 100 100 100 Penalized Total Closed 49 55 52 47 Discarded 4 8 3 5 Open 46 55 52 46 Penalized I 8 1 2 The figures display the percentage of certificates which have been closed, discarded or retained "open" as of June 1994. Source: Treasury and SPO.  Selectivity and Monitoring in the Incentive Regime 27 Table 4: Performance of Incentive Regime Beneficiaries: Fixed Investments, 1983-1990* (figures in percentages) 1983 1984 1985 1986 1987 1988 1989 1990 Readywear Committed 199 333 184 503 362 37 97 138 Realized 108 237 154 442 287 19 43 7 Performance gap 46 29 16 12 21 49 56 95 Glass Comrnmitted 171 16 131 14 110 138 33 19 Realized 171 16 131 13 110 137 17 10 Performance gap 0 0 0 7 0 1 48 47 Auto Conunitted 146 239 3 50 65 2 206 370 Realized 111 92 2 43 43 1 15 26 Perfonnance gap 24 62 33 14 34 50 93 93 * The "committed" figures represents the percentage increase in fixed investments promised by new certificate holders with respect to industry-vide achievements in the base (i.e., the previous) year. The "realized" figure represents the percentage of fixed investments achieved as of 1994 with respect to achievements in the base year. The "peiformance gap" figure represents the percentage of non-realized fixed investments vis a vis commitments as of 1994. Source: Treasurv and State Institute of Statistics, Annual Manufactunng Industry Statistics. B.3. Performance of Certificate Holders The discussion above clarified that the monitoring and disciplinary capacity of the state with respect to the performance of incentive and export encouragement certificates beneficiaries is weak. In turn, arguments were offered as to the likelihood of negative consequences flowing from this lack of monitoring and discipline. It is at this point, therefore, that the absence of information regarding the operation of the Turkish incentive regime becomes most problematic. In particular, to most persuasively document the costs associated with the lack of monitoring it would be necessary to have access to measures of the specific initial commitments made by recipients and measures of the over-time "performance gap" between initial commitments and actual investment and export performance. Most of the raw data for such measures are not just unavailable but, to the best of our knowledge, simply do not exist. However, we have  28 Institutional Influences on Economic Policy in Turkey constructed an admittedly rough measure of the performance of investment and export certificate holders. Table 4 and Table 5 attempt to measure the size of the "performance gap," that is the difference between the initial commitments made by applicants and subsequent performance, for specific components of the "contract" with the state. In particular, for investment certificate contracts we measure the realization of a most important component--new fixed investment--and for export certificate contracts we measure the realization of exports. Table 5: Performance of Incentive Regime Beneficiaries: Exports, 1985-1991* (figures in percentages) 1985 1986 1987 1988 1989 1990 1991 Readywear Committed 33 87 54 70 83 75 84 Realized 33 86 53 63 79 70 63 Performance gap 0 1 2 10 5 7 8 Glass Comrnmitted 10 9 2 70 18 21 7 Realized 10 9 2 63 9 10 3 Performance gap 0 0 0 10 50 52 57 Auto Committed 45 72 104 105 105 87 146 Realized 45 72 95 99 86 81 99 Performance gap 0 0 9 6 18 7 32 * The "committed" figures represents the percentage increase in exports promised by new certificate holders *with respect to industry-wvide achievements in the base (i.e., the previous) year. The "realized" figure represents the percentage of exports achieved as of 1994 wsith respect to achievements in the base year. The "performance gap" figure represents the percentage of non- realized exports vis a Nis commitments as of 1994. Source: Treasury and State Institute of Statistics. Annual Manufacturing Industry Statistics. As our measures are complex it is important to clarify their construction. Based on figures from Treasury and the State Institute of Statistics, we constructed approximate measures of a) initial commitments of applicants by industry with regard to new fixed investments and exports, b) 1994 realization rates of these initial commitments and c) the 1994 performance gap between initial commitments and realization by industry. While all three measures are expressed as percentage ratios (see below), the figures upon which they are based were nominal Turkish Lira.  Selectivity and Monitoring in the Incentive Regime 29 The measure of initial commitment is the percentage increase promised by new certificate holders in a given industry for a particular year in terms of the base (previous) year's industry- wide actual achievement with respect to fixed investments or exports. Thus, for example, a figure of a 200 percent initial commitment for new fixed investments in a given industry signifies that the promises contained in the contracts of certificate holders for a given year are to double the actual record of new fixed investments of all firms in the industry from the base (previous) year. The measure of realization is based on using certificate holders who had "closed" certificates by 1994 as a proxy. Recall that closed certificates indicate that access to incentives and subsidies has been terminated because the recipient informs the state that the contract has been fulfilled. Thus, for instance, a 100 percent realization figure for fixed investment in a particular industry signifies that the particular certificate holders from a given year, say 1986, had as of 1994 reached 100 percent (i.e., had equaled) of the base year's (1985) level of new fixed investment by all firms in the industry. The measure of the performance gap is the simplest to understand as it is simply a percentage measure of non-realization to initial commitment. Thus, in the case of a 200 percent initial commitment with a 100 percent realization rate the corresponding performance gap is 50 percent. Most simply understood, this performance gap figure is a measure of the extent to which the initial promises by certificate holders are exaggerated." Several general conclusions may be drawn from these tables with respect to the performance of certificate holders in these three industries. A initial point is that as these figures were calculated without adjusting for inflation, and as there was significant inflation in some years, the performance gap figures are understated. More substantively, in the first place the Glass industry shows a unique pattern in that as a general rule the performance gap was minimal irrespective of whether investment or export performance was being measured. Although this pattern is challenged for the most recent years measured in the tables (1989-91), it should be noted that as these measures are based on 1994 figures there is likely to be an overstatement of the performance gap for the recent years. This is because currently "open" certificates with a 4 to 6 year "contract" which may well "close" on time are being measured in these tables as non- realization. See Section V below for an analysis of the reasons for the glass industry's superior performance. Second, the performance gap for readywear clothing, with the exception of 1983, is generally smaller than that for autos. Some reasons for this superior performance is provided in Section V below. Third, there is a clear tendency for the size of the performance gap to be narrower for exports than for new fixed investments. This is clearly the case for Readywear Clothing and Autos. This finding is in line with our expectations regarding the relative difficulty of monitoring investment certificate comrnitments as opposed to export certificate commitments. 11 As these measures are rough indicators it is worthwhile to briefly note that the likely direction of bias is to understate the performance gap. This is for several reasons. First, the measure of initial commitments is based only on commitments made by certificate holders whose status was either "closed" or "open." Thus, the initial commitments made by applicants whose certificates had been "discarded" or "penalized" are not counted even though many of these applicants may have benefitted from certain incentives or subsidies prior to having their access terminated. Second, the measure of realization, as it is based on the self-reports of holders who have "closed" their certificate, may be unreliable given a likely tendency for the overstatement of actual accomplishments. Again, such bias would serve to understate the size of the performance gap. Partially mitigating these two biases, however, is the fact that the measure of non-realization, that is holders who still had "open" certificates in 1994, may miss some partial compliance with initial commitments. And finally, it may be noted that as "contracts" are of varying duration and may extend up to 5 or 6 years the figures for the most recent years (1989-91) may understate eventual realization and thus overstate the size of the performance gap.  30 Institutional Influences on Economic Policy in Turkey And finally, particularly disturbing is the tendency for massive exaggeration of anticipated performance by investment certificate recipients. The fact of such exaggeration underscores the claims we made above regarding the lack of any initial negative screening of applicants and lack of subsequent monitoring and discipline procedures. That is, the size of the performance gap indicates that an adverse selection process may be well-engaged as it is quite plausible that many initially insincere and incompetent applicants successfully received incentives and subsidies. Moreover, that there are no consequences for non-compliance with the original investment incentive contract further diminishes the incentive for beneficiaries to meet stated goals.  4 Bureaucratic Insulation And The Incentive Regime The evidence in the preceding section underscored the fact of and the immediate mechanisms of the ineffective delivery process for the incentive and subsidy process. However, it does not yet explain why such an ineffective delivery process developed and is allowed to continue. For this we turn in this section to an examination of the institutional character of the economic bureaucracy. In the following section we turn to an examination of the policy networks connecting bureaucrats and business actors. Comparative evidence, it may be recalled, concerning the institutional preconditions of effective economic reform programs underlines the importance of a relatively insulated economic bureaucracy. Such insulation shelters bureaucrats from the immediacy of both "pork-barrel" political pressures and interest-based pressures from business, labor or other private groups. Relatively insulated economic bureaucracies tend to be small and centralized, endowed with well- defined policy authorities and characterized by stable career paths with meritocratic recruitment and promotion criteria (Rueschemeyer and Evans, 1985). In this regard, our second hypothesis was that the state's ability to deliver effective economic policies is positively related to the degree to which economic bureaucrats enjoy an insulation from interest-based and political pressures. Thus, the main finding discussed below, that there is a particular lack of insulation in the TUD section of the Turkish economic bureaucracy, supports this factor as being partially responsible for the above-documented ineffectiveness of the incentive and subsidy policies. In general, however, as the discussion below (based on both interviews and secondary sources) illustrates, the lack of insulation in the TUJD is symptomatic of more widespread problems affecting the Turkish Civil Service today. Most specifically, there are inadequate safeguards for civil servants from political pressures. Moreover, due to the lack of strong private sector associations capable of articulating coherent policy preferences to bureaucrats, these bureaucrats therefore end up becoming overly sensitive to the particular demands of well-positioned firms (Onis and Webb, 1994). Of course, such problems are not unique to Turkey and, indeed, Turkey may be better positioned to deal with them than some other developing nations. This is because prior to attempting to create an insular bureaucracy many developing countries of today, particularly those with a colonial history, have been faced first with the challenge of building and legitimizing the authority of the national state. Without such a state a stable basis for individual property rights could not be institutionalized (North, 1992, Huntington and Nelson, 1976). Yet this has not been the case in Turkey since the Turkish political system with its Ottoman legacy of strong central 31  32 Institutional Influences on Economic Policy in Turkey political authority stood out as a rare, if not unique, example of a modernizing country with a distinct state tradition (Hale, 1976). Furthermore, since in Turkey the bureaucratic elite played a leading role in the administrative modernization in the nineteenth century and economic modernization in the first half of the twentieth century, it is not surprising to find that high-level bureaucrats formed a distinct group united by a sense of shared identity in the sense that they often perceived themselves as the "sole formulator and the guardian of the long-term interest of the community" (Heper, 1989: p. 461). Moreover, empirical studies conducted during the 1960's and 1970's, i.e., during the heyday of protectionist industrial policy, invariably pointed to the fact that the bureaucratic elite in Turkey aspired to a substantive role in the development of the nation as opposed to being content with a Western-style more technocratic role. (Heper, 1976; Bozkurt, 1980). As has been emphasized by Turkish researchers, the generalist background of the top bureaucrats, combined with their lingering elitist attitude, inevitably brought them into a long-term conflict with the political elite. Consequently, from 1950, following the transition to a multi-party democracy, until 1980, various governments, with differing degrees of success, attempted to transform bureaucrats into loyal servants of politicians (Bugra, 1994; Heper, 1989). In the meantime, the political elite in Turkey, irrespective of political party affiliations, deeply distrusted a spontaneous and pluralistic development of interest group associations in the country. As a result, only a limited legitimacy was accorded to pluralistic forms of interest representation and professional associations were conceived of as appendages of the state rather then as autonomous bodies able to articulate their members' preferences and become involved in economic decision- making processes (Bianchi, 1984). Consequently, even after the transition to a multi-party democracy, the political elite in Turkey made most of the critical decisions concerning the economy "virtually without consulting the traditional civil bureaucracy, the Parliament and the political parties as well as interest groups" (Heper, 1991: p. 173). Since both voluntary associations and public professional organizations were controlled by rules which prohibited them from "engaging in politics," with politics being defined in a vague fashion, the state-imposed distrust of pluralist interest intermediation generated particularistic and clientalistic relations between the political elite and business interests (Kalaycoglu, 1991). From the point of view of the political elite these particularistic relations were expedient since they rendered business persons "individuals" and hence more vulnerable to government discretionism. That is, if business persons strayed from their pre-ordained role of carrying out activities as dictated by government, they could be punished by selective measures aimed at them (Heper, 1991). Indeed, governments acted on this threat through means such as lowering the tariff for goods produced by a non-favored business group, by passing retroactive laws which had a devastating effect on a particular business activity or by not honoring a contract concluded with a business person who fell out of favor (Bugra, 1994). So, the question may be asked, why did business persons, despite these vulnerabilities, not cease individual-level contacts and instead attempt to exercise influence over the government through a professional association? The short answer is that those business persons who were powerful enough to strike individual relations with top politicians and bureaucrats benefitted from these relations as well (Kalaycoglu, 1991). Earlier we provided several examples which portrayed certain dimensions of this rent-seeking process whereby certain business persons, without any  Bureaucratic Insulation and the Incentive Regime 33 competitive test in the market, were able to make quick fortunes. The problem, however, is that even though individual business persons often benefitted from such rent-seeking, thanks to privileged personal ties to decision-makers, it hindered the creation of a competitive and growth- conducive business environment not only because of the well-known opportunity costs of rent- seeking, such as foregone innovations, but also because the extent of particularism in state- business relations paralyzed the bureaucracy. As an interviewee put it to another researcher (Bugra, 1994: p. 164): He who explainis his case to the Prime Minister or the Minister conicerned solves his problem. You go to Isin Celebi (one of the Mfinisters responsible for the economy), you crv oni his shoulder and he says 'O.K., I'll finid youi thle niecessary funds.' Then the Central Bank says that the fiunds are not available, the AMinister orders the transfer of funds from onie budget to the other. T7his leadv to initerferences at all levels of the bureaucratic process. And, of couirse, you are very happy because your problem is solved. You tell others wvhat a nice, utnderstantdinig person the Minister is, anid how niicely he has solved your problem. Buit itle instilutions cannot fznction under these circutmstances, the State Planiniiitg Orgcaniization7 canns7ot fuintction, the Treasury anid the Central Bank cannot funtiction . Institutionalization becomes impossible. Bureaucratic insulation in Turkey was not just hampered by the extent of particularism in state-business relations but, in addition, politicians increasingly influenced bureaucratic recruitment and promotion patterns (Heper, 1985). In turn, this gradually rendered top bureaucrats overly sensitive to signals emanating from the political realm (Author's interviews, 1994). Furthermore, as it was relayed to us by a top bureaucrat who is now a high level executive in the private sector, the sensitivity to political signais correlates with the nature of the agency in question and the educational background of the incumbents. Most specifically, an agency like the Central Bank is somewhat less subject to political interference. save for pressures from the Treasury, as the Central Bank is not generally a source of rent creation (although some subsidized credit is disbursed) and as its high level staff are specialists in a technical skill which is highly sought after. Thus, the specter of being fired by politicians, while not absent, does not always threaten these officials who are, in fact, rendering a public service by accepting significantly lower salaries than they could obtain in the private sector. In contrast, in an agency which is potentially a source of rent creation, such as the Directorate of Incentives and Implementation (TUD), where most incumbents are generalists by background and who can be fired by politicians or transferred to less desirable jobs, it is not realistic to expect that bureaucrats can enjoy much insulation from political pressures. While opportunities for graft and rent-seeking are inbuilt features of any incentive and subsidy system, to the extent that the civil servants who administer the system do enjoy legal guarantees against arbitrary political influence, then they are more likely to withstand the pressures from politicians who want to rely on the incentive and subsidy systems to reward political allies and punish foes. When looked at from this vantage point, the decision after the about-turn to a market- friendly economic policy in 1980 to centralize the administration of the incentive and subsidy regime in the State Planning Organization (SPO) was an apt one. This is because the meritocratic basis of recruitment in this agency, combined with the educational background of its personnel, rendered the SPO a highly "competent" agency which was relatively well-protected against non-  34 Institutional Influences on Economic Policy in Turkey legal political interference. Table 6 compares the educational status of the SPO personnel in the 1980's with that of the economic bureaucracy (i.e., the Ministry of Finance, the Ministry of Industry and Trade, and the Treasury) as a whole. Table 7, on the other hand, demonstrates that the educational background of the SPO personnel reflects professional fields and engineering backgrounds rather than, as is the case in many other state agencies, only the social sciences. Finally, Table 8 provides evidence concerning the extent of meritocratic recruitment into the agency (successful applicants must past both a written and oral examination). Table 6: Distribution of Personnel in the Economic Bureaucracy According to Educational Status in 1984, 1992 Elementauy Secondarv High School University (percent) (percent) (percent) (percent) 1984 State Planning Organization 5 8 22 65 Total Economic Bureaucracy 17 25 40 18 1992 State Planning Organization 14 11 19 56 Total Economic Bureaucracy 15 12 36 37 Source: State Personnel Questionnaire: 1984, 1992. Table 7: The Percentage Distribution of Personnel According to Occupations in the State Planning Organization, 1960-1993 1960 1970 1980 1985 1990 1991 1992 1993 Engineering 9 15 21 31 35 27 27 28 Economics 66 21 29 28 22 22 24 25 Law 7 4 3 5 5 5 4 4 Management 5 8 10 12 11 12 12 13 Other 13 52 37 24 27 34 32 29 Source: State Planning Organization.  Bureaucratic Insulation and the Incentive Regime 35 Table 8: The Results of Entrance Exams to the State Planning Organization: 1960-1992 1960 1971 1973 1974 1974 1975 1978 1980 1985 1988 1992 (Jan.) (Sep.) Applicants 42 377 15 120 261 133 183 467 2220 2399 3327 Passing written 26 36 10 11 16 22 22 119 169 164 175 exam (percent) 62 10 67 9 6 17 12 25 8 7 5 Passing oral exam 18 21 10 6 5 7 12 73 36 66 40 (percent of total) 43 6 67 5 2 5 7 16 2 3 1 (percent of those 69 58 100 55 31 32 55 61 21 40 23 passing written exam) Source: State Planning Organization. At the same time, however, the insulation of the SPO from political pressures during the 1980's was compromised by the recruitment of new personnel--so-called "contract personnel"-- into the agency through a separate process which bypassed the normal written and oral exams. More specifically, with the enactment of Act 3046 in 1984 it was made possible to bypass certain restrictions of the Public Personnel Law No. 657 regarding the hiring and laying off of the state personnel, and also to determine salaries by the Council of Ministers. After the passage of this legal Act, overall within the Turkish administration "the conversion of managers to contract status has proceeded fairly rapidly: in 1984 there were 674 such administrators, in 1985 3491, and in 1986 26,578 or about 10 percent of the administrative corps" (Waterbury, 1988: p. 16). In the SPO, on the other hand, by looking at the directory which was published in 1990 to commemorate the 30th anniversary one can see that the percentage of contract personnel was much higher as, of the total 520 non-administrative personnel (i.e., including assistant experts, experts, mid-level managers, and top managers), only 281, i.e. about 54 percent, were recruited from the regular channels through written and oral exams, whereas the rest were contract personnel. It is even more interesting to see that, of the 157 civil servants who were employed in the bureaus of the SPO which deal with the actual implementation of the incentive and subsidy regime, i.e. the TUD, fully 69 percent have been recruited on the contract status.  36 Institutional Inftuences on Economic Policy in Turkev An Example of Bureaucratic Insulation: The Turkish Industrial Development Bank An example of relatively hgih institutional insulation with subsequent positive influences on the effectiveness of policv deliverv is illustrated in the casc of the Turkish Industrial Development Bank (TSKB). Founded in 1950 with the support of the World Bank, the Central Bank and leading commercial banks in Turkey, TSKB is among the oldest development banks in the world. TSKB's main activity is the provision of long-term loans for Turkish entrepreneurs. Moreover, to facilitate such long- term financing, TSKB further provides project and technical assistance relating to privatization, restructuring, investment, production. marketing, planning, management and finance. Insulation in the case of TSKB is enhanced by several factors. First, and most importantly, is that TSKCB's ownership structure has allowed it to retain some distance from political pressures. Originally, majortWv owvnership of TSKB wvas private and represcnted a diverse cross-section of commnercial banks and private holdings in Turkey. MNore recently, the majoritv owner has become the Is-Bank. While technically, the Is-Bank is state-owned, its ownership is complicated in that a major share is controlled by a private pension fund of Is-Bank employees. This has meant that the Is-Bank has retained considerable managerial independence from direct political pressures from the state. This stands in strong contrast to several of Turkey's other main state-owned development banks--the Turkish Development Bank and the EximBank--which are majoritv-owxnied by the Treasury and thus more directly subject to political pressures. Additionally, recruitment and promotion practices within TSKB are less subject to political interference and reflect more rcliance on meritocratic norms. This is partially a result of the ownership structure Nvith resultant relative lack of political pressures but is equally due to the strong sense of shared mission within TSKB which stems from its unique position in Turkey's financial system as well as from the institutional connection with the international banking community,| especially the International Finance Corporation, which strengthens such norms. The relatively high insulation of the TSKB has allowed it to develop effective procedures for extending loans and technical assistance. Such procedures involve technical, as opposed to political, initial screening of applicants and subsequent monitoring and disciplinary measures which dimiinish the room for institutional discretionism motivated by political pressures. Thus, for example, to obtain access to long-term concessionarv financing an applicant must not only provide the investment or export certificate provided from the TUD but must a) demonstrate their owvn commitment to the project by up- front providing 50 percent of the equity financing for the investment and b) demonstrate the project's capability of sustaining a financial rate of return of at least 15 percent. The latter demonstration is largely independent of the loan applicant and involves a detailed review process in which three separate TSKB employees--an economist, a financial analyst and an engineer--assess the project's potential. Subsequent monitoring is intensive and involves field visits to investment sites. For instance, all imports of significant machinery are verified by TSKB personnel through visits to the location of the linvestments. More generally, once a ycar there are two reports generated--an "Investmnent Follow-Up Studv" and a "Normal Follow-Up Study"--which evaluate the fit between the original intent of the loan and actual implementation.  Bureaucratic Insulation and the Incentive Regime 37 Table A: TSKB Loans in Arrears (figures in percentage of total value of loans) Year Textiles Glass Autos 1982 20.8 0.4 2.5 1983 17.9 0.5 1.4 1984 15.9 0.3 1.5 1985 20.9 0.2 0.2 1986 25.0 0 0 1987 28.6 0 0 1988 29.5 0 3.7 1989 48.4 0.4 0.4 1990 43.9 0 2.0 1991 40.5 0 1.5 1992 26.7 0 2.5 1993 27.9 0 2.2 Source: TSKB As seen in Tables A & B, with respect to the three industries of particular interest for this study, the effectiveness of TSKB's screening, monitoring and disciplinary procedures may be noted. Table A demonstrates a relatively low rate of arrears for loans which is a comment on the rigor of the application process. The recovery rates for the loans in both the glass and auto industries are particularly high. In the case of the readywear clothing industry the rate has been lower. On one hand, this may be accounted for in that the higher number of recipients made monitoring more difficult. On the other hand, there is reason to believe that there was some diminishment of the insulation of the TSKB, particularly during the latter half of the 1980's, as there were strong political pressures to advance credit to small business persons who were an integral part of the coalitional basis of the ruling ANAP government. And as may be seen in Table B monitoring and subsequent discipline of poor performing loans is a regular occurrence in the TSKB as "withdrawals" reflect those loans which passed initial screening process but, as became clear at a later date, were no longer meeting TSKB's rigorous standards. Table B: TSKB Loan Approvals and Withdrawals: 1983-87* Approvals Withdrawals Withdrawal Rate (number) (number) (percentage) Textiles 173 38 22 Auto 27 7 26 Glass 10 4 40 All industries 683 174 25 Source: TSKB * The figures for withdrawals include an undetermined nunber of loans which were originally approved prior to 1983. At the same time, the figures for approvals count an underdetermined number of loans which were withdrawn after 1987.  38 Institutional Influences on Economic Policy in Turkey One should not, however, overstate the difference in legal status between contract personnel and other bureaucrats in terms of its impact on bureaucratic incentives. That is, irrespective of legal status, all top bureaucrats are vulnerable to political pressures. When asked about these pressures, all six of our interviewees who occupy Director or Vice President positions in the Treasury (in charge of the incentive regime) acknowledged the existence of such pressures. Even in the case of non-contract personnel such impositions from above to cater to particular interests carry some weight since politicians often threaten to downgrade their status and "exile" them to less developed regions. When asked to what extent politicians affect their careers and promotional opportunities as opposed to the extent to which experience, competence and expertise play roles, these bureaucrats repeated to us what was found in a similar, albeit more comprehensive, survey conducted 20 years ago--that subjective factors such as "good personality" and "good relationships" with important individuals were considered as important as "expertise" in promotion decisions (Heper, et. al., 1980). In contrast, in a nation such as Korea, politicians' ability to affect bureaucratic careers is limited and expertise emerges as the single most important criteria in promotion decisions (Schneider, 1993). We also asked the officials in the TUD how they attempt to protect themselves against ad hoc pressures and to remain focused on the objective tasks of their job. They were asked to rank order possible protection mechanisms in terms of the following options: protection through appeal to the media; protection through legal means; protection through one's expert status; and protection through key relationships. Only one official selected the option of protection through legal means as most important (another official found this option wholly irrelevant). In any case, 80 percent of these officials pointed to their expert status as the best recourse when attempting to resist top-down demands for favoritism. In this regard, all of these officials have impressive education backgrounds but nonetheless, in the end, neither the legal system nor the mass media can effectively protect them and prevent abuses. In contrast to both politicians and big business persons they often do not benefit from the particularistic relations between the state and business and seem to be squeezed in between these other powerful actors. Finally, if one discusses the general effects of administrative reforms in the 1980's, as opposed to focusing on the effectiveness of the incentive and subsidy regime as is done here, it is possible to discern positive aspects brought about by Act 3046 in terms of bypassing the rigid procedures of the Code of Public Personnel, enacted in 1970. One rationale behind the changes brought about in the mid- 1980's was to actually expand the application of merit principle by bypassing salary limitations stipulated by the Code in order to draw top talent into public service. Another rationale, on the other hand, was to create a new cadre of reform-minded bureaucratic officials to carry out staff or line duties as a means to overcome a statist bureaucratic culture perceived to be hostile to contemplated market-oriented reforms. Overall, there is no doubt that these reforms helped to de-bureaucratize the administration and enabled the ruling Motherland Party government to have a freer hand in economic policy making. And when de- bureaucratization was combined with a substantive reduction in the amount of rents distributed by a public agency, as has been the case in the Central Bank in the post-1980 period due to a significant reduction of preferential credits disbursed by this agency, as well as being combined with increased exposure to external financial agencies concerned about rent-seeking, the overall effect of the administrative reform seemed to be positive (Atiyas and Ersel, 1994).  Bureaucratic Insulation and the Incentive Regime 39 Yet, as has been suggested earlier, in an agency which is par excellence a source of rent creation, such as the TUD the use of contract personnel with considerable job insecurity rendered the agency particularly vulnerable to political pressures. These pressures, in turn, often take the form of requests to grant certain incentives and subsidies to favored business persons and/or not to monitor and enforce penalties in the case of non-compliance with mutually agreed goals. When looked at in from this angle, it becomes possible to understand why it is that TUD officials do not collect data related to the actualization of investments which could subsequently then help them enforce the incentive contracts. In other words, it is less the absence of administrative capacity then the incentive structure within which state officials are situated that is responsible for the reluctance to impose performance standards in line with the contracts. In this context, when asked to account for the failure to follow-up the implementation of the incentive regime one high level official in the TUD stated that such follow-up is like playing "Russian roulette" since: If I tell my people to collect data on the actualization of investments then what do I do with the data? Once I have the documentation on the implementation of incentives, then I have to do something with it. But this will pose a true dilemma for me. If I don't do anything about the situation then I am violating the legal system which stipulates that I should punish the violators, But if I take steps to punish the violators then this is like playing with fire. They are powerful individuals who have access to top politicians and will find a way, sooner or later, to unseat me. So, I would rather not collect data in order not to disturb anybody.  40 Institutional Influences on Economic Policy in Turkey  5 Industry Policy Networks And The Incentive Regime In the previous section we discussed the manner in which the relative lack of insulation of Turkish Civil Servants, particularly those associated with economic policy delivery and the TUD (home for incentives and subsidies), from political pressures undermines the effectiveness of econormic policy. At the same time, we noted that the general absence of strong private sector associations able to articulate policy preferences fostered a situation in which economic bureaucrats end up becoming overly sensitive to the particular preferences of well-positioned economic actors. It is to a fuller examination of the nature of such state-business relations we now turn. We will examine the nature of the policy networks connecting economic actors with economic bureaucrats in three separate industries--autos, glass and readywear clothing--arnd we will investigate how the different policy networks influence the effectiveness of economic policy. It may be recalled that by policy networks we refer not only to formal consultations between bureaucrats and business persons or business associations but also to the presence and density of informal ad hoc meetings and communications. Growth-oriented networks are ones in which there is a two-way information flow, established norms of reciprocity, honesty and trust, and transparency in information-sharing and decision-making among network members. Alternately, "rent-oriented" networks are those in which there is a two-way information flow and trust but in which the level of transparency is sufficiently low that some network members successfully develop particularistic rent-oriented ties. While two-way information flows and certain normative expectations, albeit the nature of the information exchanged and the character of expectations may vary, are common features of all networks whether growth-oriented or rent-oriented, it is especially the lack of transparency which is particularly harmful to economic growth. This is because "since the fruits of growth occur in the future, growth creates enormous uncertainty over how its fruits are going to be distributed" and, therefore, given a lack of transparency, "this creates incentives for various agents to preempt their competitors by bribing public officials" (Campos, 1993: p. 28-29). In contrast, when rent- sharing rules are transparent to members of a network and rules cannot be easily bent to favor some parties, members of a network are less likely to engage in wasteful lobbying activities since there is little to gain from having privileged access to decision-makers. Our hypothesis regarding policy networks was, simply, that, the delivery of effective economic policies is positively related to the presence of growth-oriented policy networks and negatively related to the presence of rent-oriented networks linking economic bureaucrats and beneficiaries of incentive and subsidy programs. While our comparative evidence only covers 41  42 Institutional Influences on Economic Policy in Turkey three industries, and thus the strength of generalizations is limited, our findings are a) the rent- oriented network which characterizes the auto industry severely hampers policy effectiveness, b) that despite opportunities to develop rent-seeking ties the glass industry has achieved a pro- competition orientation as a result of both industry-based "objective" technological needs and "subjective" factors relating to managerial strategies and c) the readywear clothing industry developed a growth-oriented network in the mid-1980's which enhanced the effectiveness of incentive and subsidy policy. The enhancement of policy effectiveness in the readywear clothing industry is not due to any change in characteristics of the Turkish bureaucracy. Instead, effectiveness was enhanced due to the decentralization of screening, monitoring and disciplinary functions away from the TUD and into collective associations representing the readywear industry. Such associations helped to ensure that incentive and subsidy contracts were enforced with more regularity and consistency than is the case in the TUD. A. Rent-Oriented Networks and the Auto Industry Earlier we characterized as "rent-oriented" those networks in which there is a two-way communications flow and the presence of various established norms but in which the level of transparency is sufficiently low that some network members successfully develop particularistic and rent-oriented ties. When looked at from this angle the nature of state-business ties in both the auto and glass industries in Turkey approximate rent-seeking features. One common feature in these industries is that even if a formal business organization exists, as does in the auto industry (i.e., the so-called Association of Automotive Manufacturers or OSD), the major players--Sisecam AS in glass and a handful of vehicle manufacturers in auto-- bypass their own association and prefer to deal directly with decision-makers regarding the benefits they may derive from the incentive regime. Moreover, in the case of OSD, which was established in 1974 and has generally been presided over by a member of the KOC Holding, its major role historically has been cartel-like in efforts to fix prices on intermediate goods needed for the industry. More recently, the functioning of this organization has been crippled by a rivalry between old members of the OSD who control a captive domestic market and new entrants such as Sabanci (with Toyota), Dogus (with GM) and potential new entrants such as Hyundai, Honda and Mazda. Under these circumstances, it becomes difficult for OSD to formulate any serious positive and long term vision for the future of the industry. Instead, its function approximates what Olson calls a distributionist coalition in two regards (Olson, 1982). It campaigns for high tariffs and a cap on imports--about one-fourth of the passenger vehicles are now imported--and it lobbies for ending the issuance of licenses to multinationals (often with a domestic partner as in the above-noted joint venture) to domestically manufacture passenger vehicles. It is quite significant that the major shareholders of the two holdings which control the OSD, i.e. the KOC Holding and Oyak-Renault, do not, in contrast to what may be observed in the readywear clothing industry, assume official positions in this organization. According to the general manager of this organization, who is an ex-academician, this fact thwarts the efforts of OSD to convince decision-makers that the organization, as opposed to individual auto  Industry Policy Networks and the Incentive Regime 43 manufacturers, should be treated as the main interlocutor of the state (Author's interview, 1994). 12 In the end, thanks to their privileged access to top bureaucrats and politicians when seeking special favors vis a vis incentives, and subsidies from the state, industry leaders prefer to bypass their own organization particularly when they think the stakes involved are high. The more significant interaction patterns--networking--between industry leaders in the auto and glass industries and government decision-makers are quite non-transparent. The contacts are very frequent and the Treasury, since it is in charge of the incentive regime, is the preferred target of access among all economic bureaucracies. In fact as can be seen from Tables 9A-C below, for all of the three industries the Treasury is the bureaucratic agency which is contacted most frequently. However, these contacts occur on a more frequent and casual basis for the auto and glass industries than in readywear clothing. In our interviews with high ranking bureaucrats in charge of industrial subsidies they too admitted that these contacts are very frequent and further noted that they occur most regularly at the firm level as opposed to through the intermediation of a representative business association. In fact, 80 percent of these bureaucrats, when asked to rank their responses between five options concerning the most frequent level of contact with business, selected the "firm level" option first as opposed to other options representing more institutionalized, i.e. collective organization mediated, levels of contact. Hence, it is clear that in Turkey particularistic relations prevail. This is in contrast to more institutionalized forms of interaction between big business and bureaucrats as seen in countries such as Korea where professional associations mediate individual demands for subsidies (Byung-Kwon Cha, et. al., 1989). This situation has implications concerning both the nature and properties of the information which is exchanged in these personalized and multifaceted networks and also concerning the nature of trust and reciprocity which undergird them. When such relations are struck on an individual basis chances for the convergence of initial preferences towards a common, growth- oriented goal are often precluded. This is because the emergence of such a goal entails the collective involvement of business in the conceptualization stage of the incentive regime. Firm- level involvement, however, is often oriented toward individual manipulation at the implementation phase of the incentive delivery process, i.e. rent-seeking pure and simple. An organization such as OSD generates a great deal of information concerning various aspects of the industry and regularly sends this information to major players and top bureaucrats. But, presumably because owners of auto firms do not consent to have this organization mediate their access to the government, top officials do not perceive this information to be very relevant for making decisions related to the industry. Instead, when such information is needed, state officials turn to heads of individual firns whom they know well. The provision of such information to state officials who lack adequate capacity to collect data themselves and who are often riddled by inter-agency rivalries which make it difficult to enlist each other's help, impart a certain legitimacy to auto makers' demand for protection and individual favoritism. 12 In this regard the following anecdote is quite revealing and worth mentioning. When we conducted our interviews in May and June, 1994, the auto industry in Turkey was in the middle of a "crisis" due to a shrinkage in the absorption capacity of the internal market. The general manager of OSD invited a high-level official in Treasury to Istanbul, where OSD is located, to inform him about the dimensions of the crisis and lobby for the protection of the national industry. He also called the head of a major car manufacturer which is a member of the OSD to enlist his support as well and to request him to meet with the high- level bureaucrat. He did not succeed on either count.  44 Institutional Influences on Economic Policy in Turkey Table 9: Business Networking with State Concerning Incentive Policies (figures in percentages) A. Bureaucracv Contacted Most Regularly SPO Treasury Min of Min of Exim Other No (N) Ind/Trd Fin Bank answer Auto 0 53 11 0 0 26 11 (19) Glass 25 63 0 0 0 0 13 (8) Ready wear 11 61 0 11 17 0 0 (18) All 8 58 4 4 8 11 7 (45) B. Frequency of Contacts with Important Bureaucracies Once/ Once/ Several Once/ No (N) week month times/yr year Answer Auto 26 26 26 11 11 (19) Glass 50 0 38 0 13 (8) Ready wear 6 28 61 6 0 (18) All 22 22 42 4 7 (45) C. Formality of Contacts with Important Bureaucracies Very Very No formal Formal Casual casual answer (N) Auto 5 21 53 11 (19) Glass 0 38 38 0 25 (8) Readv wear 6 44 44 0 6 (18) All 4 33 47 4 11 (45) Source: Author's Interviews. There can be little doubt but that the quite long-standing relations between a handful of manufacturers in the two industries and top economic bureaucrats creates a certain degree of trust based on reciprocal obligations. This is not easy to measure. Table 10 below indicates that auto executives hold a particularly positive evaluation of bureaucrat's responsiveness to their concerns  Industry Policy Networks and the Incentive Regime 45 but this does not fully reveal the nature of the "reciprocity" in question. To complicate matters even further, business persons in auto and glass do not solely contact top bureaucrats but, when the issue at stake is very important and the meeting with bureaucrats fail to address their concerns, they directly contact the Minister in charge. This was not only revealed to us in interviews but is captured in the survey data presented in Table 11 below in which some of the executives in both the auto and glass industries, but none in readywear clothing, acknowledged that contacting a Minister was the second most preferred means of expressing dissatisfaction. Table 10: Business Views on State Responsiveness to Business Concerns (figures in percentages) Very responsive Responsive Unresponsive Very unresponsive (N) Auto 5 84 5 5 (19) Glass 0 25 75 0 (8) Ready wear 6 28 61 6 (18) All 4 51 40 4 (45) Source: Author's Inteniews. Table 11: Business Avenues for Expressing Dissatisfaction with State Incentive Policies (figures in percentages) Minister Bureaucrat Business Assn. Media No Ans. (N) A. First choice Auto 0 32 58 0 11 (19) Glass 0 100 0 0 0 (8) Ready wear 0 44 44 0 11 (18) All 0 49 42 0 8 (45) B. Second choice Auto 11 1 11 0 68 (19) Glass 38 0 13 13 25 (8) Ready wear 0 22 33 6 39 (18) All 11 13 22 4 49 (45) Source: Author's interviews.  46 Institutional Influences on Economic Policy in Turkey The precision with which business persons network with various levels in the economic bureaucracy to achieve their ends was expressed by a quite high level official in a major auto company as follows (Author's interview, 1994): It is important to know the level at which you should try; to solve problems. We do not always want lo conitact the Minister or the Prime Minister, even if these are friends. If we do this all the time bureaucrats will be angry at us and, even if they comply with orders, they may obstruct us at the earliest opportunity. Instead, we start with the middle level bureaucrat before we even talk to the top guy. If need arises to resolve things at a higher level, before contacting govemnment members, we call some Congressmen whom we know well since there are many, both in the governing coalition anad the oppositioni. Sometimes they are effective in solving our problems. So, what are the reciprocal obligations which underlie state-business relations in rent- seeking networks? The principle of reciprocity at work does not include, as has been the case in Korea, the expectation that improved business performance--based on a medium to long-term time frame and measured by tangible indicators--is to be associated with access to state subsidies (Amsden, 1989). Clearly, however, this is a quite delicate issue which we tried to investigate less through surveys than through confidential interviews. According to this anecdotal evidence, in exchange for firms' demands for subsidies, political elites frequently ask for highly personalized services such as support for their political party or even support for family members (which may take the direct form of a payment or an indirect formn such as employing a family member in the firm). Bureaucratic elites, in their turn, are said to be generally content with the provision of information which they can incorporate in reports and rarely demand personalized services in exchange for favors provided to business persons. According to our business interviewees this is partly due to the subordinate position of the bureaucrat in the decision-making hierarchy (such subordination, according to one business person has accelerated since the early 1970's) and partly due to the lingering effects of an elite educational system and elite bureaucratic tradition which instills strong notions of morality in bureaucrats. There is, however, little solace in the fact of alleged relative absence of sheer corruption since the major structural characteristic of networks linking decision-makers to business persons in the glass and especially in the auto industry is the absence of transparency. That is, since many deals are made on a personal basis, and the level of benefits accorded to different firms remains at the discretion of decision-makers, rival firms have difficulty in finding out about each other's success in receiving benefits. Thus, business persons in auto and glass do not pay too much attention to the specifics of any particular year's incentive regime, although this is published in a quite transparent document, as this has little bearing on many of the major outcomes of interest to them (Author's interviews, 1994). Hence, as suggested earlier, large conglomerates shy away from becoming involved in the initial conceptualization stages of decision-making processes concerning the incentive regime as this would subsequently bind them to certain rules, and instead attempt to later manipulate policy at the implementation stage. The lack of transparency then is both a cause and a consequence of the fragmented and particularistic networks between political and bureaucratic elites and business persons. In other words, the nature of the system creates an uncertain environment in which business groups attempt to minimize risk by seeking privileged connections with top decision-makers and then rely  Industry Policy Networks and the Incentive Regime 47 on these connections to insure themselves against unexpected turns in economic policy. To the extent that business in Turkey is literally dependent on hundreds of decisions which affect prices, costs of inputs, access to preferential credits, and so on, incentives will continue to be generated for business persons to seek allies in the state as an insurance mechanism. At the same time, however, a vicious circle is created in that the resulting frequency of personal contacts between business persons and top politicians leads to arbitrary interventions by the politicians into bureaucratic affairs thereby causing ad hoc changes in the everyday implementation of the incentive program at a dizzying speed. Consequently, any "successful" rent-seeking effort on the part of one business actor may easily lead to efforts by rivals to accomplish the same or to better the effort through their own rent-seeking. In this system, while it is possible to score partial victories, the nature of the competition generated is unable to deal with industry-level collective problems, such as restructuring to take advantage of changing world market conditions, the resolution of which would be in the interest of all major parties. In all of these regards, the auto industry approximates the characteristics of a rent-seeking network. Major firms in the industry, such as OYAK-Renault and Otosan, hold a number of incentive certificates, albeit, information about the content of these certificates and the nature of performance standards is not available. Renault officials, for example, told us that they were presently holding 5 different investment incentive certificates and 6 or 7 different export investment certificates, but did not want to talk about the details of, or answer questions about performance standards for, these certificates. Since auto manufacturers in Turkey in the 1980's have not undertaken major new investment projects, one can surmise that the majority of the investment certificates relate to expansion, completion and quality improvement projects (see Table 5). One exception with respect to new investments, however, is the new joint venture between the Sabanci Holding and the Toyota Motor Corporation. This joint venture aims to increase passenger vehicle capacity by 100,000, presumably not solely for the domestic market. Nobody knows the details of the agreement between the state and ToyotaSA. The whole agreement is clouded in a veil of secrecy which breeds distrust among the few firms in the auto industry. The press reported that the Sabanci Holding and Toyota were given a secret deal by the state which granted them public land in an industrial area. At the same time, although the investment was made in a developed region, the state conferred upon ToyotaSA the benefits of an investment in a less developed region. In return, we learned through interviews that state officials were able to exact promises with respect to some performance standards in exchange for the incentives granted. But in the absence of public knowledge, and given the low enforcement capacity of the economic bureaucracy, it is doubtful that compliance can be assured. In the absence of transparency among network members, the major players in the auto industry assume that their rivals have managed to obtain maximum benefits with minimum performance obligations. This assumption gives them every incentive to engage in badmouthing their rivals and engaging in costly and inefficient rent-seeking. An anecdote reveals the severity of the situation. Despite numerous attempts, one author of this paper had failed to obtain an appointment from a high level manager of ToyotaSA. Finally, an intermediary was found, and the firm official accepted to be interviewed by the World Bank consultant. In the interview, the author was told that despite his credentials and the presumed World Bank connection, he could have been an imposter, namely a spy for the rival business holdings. The author was also told that  48 Institutional Influences on Economic Policy in Turkey ToyotaSA officials had had bad experiences with people who represented themselves as "World Bank consultants" in the past, and therefore they were now overly cautious. One wonders how these firms which suspect each other's intentions so much can co-exist under the banner of the same association, the OSD. B. Rent-Seeking cum Growth-Oriented Network: The Glass Industry Our original expectation was to find a rent-seeking network in the glass industry, Mar excellence, because the glass industry, which dates back to 1935 in Turkey, is virtually monopolized by a single firm--"Turkiye Sise ve Cam Fabrikalari A.S, or SiseCam. Composed of 12 glass, 8 glass integrated and 13 trading, distribution and financial enterprises, SiseCam is a vertically integrated enterprise which manufactures its own major inputs, and dominates 81 percent of industry activity."3 Other companies operate on the edges of the market and shy away from direct rivalry with SiseCam. Indeed, in 1971, when Anadolu Cam Sanayi A.S., producing flat glass and bottles, attempted to operate outside of the SiseCam group, the company was forced to quickly throw in the towel--that is, to be bought by SiseCam--as a result of clever maneuvering by SiseCam officials. The details of this interesting incident in Turkish business history remains to be published.'4 Our surprising finding is that despite its unchallenged hegemony in the domestic market SiseCam officials have put a higher premium on improved business performance through cost cutting and conquering external markets than on sheer rent-seeking. This does not mean that rent-seeking efforts do not exist. In fact, as we suggested earlier in the context of the auto industry, the networking patterns between industry leaders in auto and glass, unlike in the readywear clothing industry are quite non-transparent. Particularly given the absence of an intermediate industrv association, SiseCam officials commonly strike direct and particularistic relations with economic bureaucrats in charge of the incentive regime. Currently, the firm holds more than 40 export and investment incentive certificates and top officials spend considerable time in Ankara attempting to obtain these certificates in November and December, the last two months of the prevailing incentive regime. This is because, in the absence of formalized consultations and involvement in the decision-making phase of the preparation of the "new" annual incentive regime, industry officials attempt to insure themselves against unexpected turns in the incentive regime by hoarding, so to speak, as many incentive certificates as possible. At the same time, however, SiseCam officials, unlike in the auto industry, have not rested on their laurels by indefinitely taking advantage of a captive internal market while simultaneously lobbying for protection against imports. The glass industry in Turkey today is not highly protected (see Appendix II), is certainly protected less than the other two industries we studied, and, to that extent that firm officials (unsuccessfully) lobbied in the past against glass imports, their complaints were not directed against imports per se but against dumping practices systematically employed by certain nations. Moreover, as may be seen in Tables 4 and 5, the performance gap has been quite low in the glass industry. In other words, even in the absence of 13 The most expensive component of glass production is soda ash with a 56 percent cost input and a 17 percent raw material input. 14 The economnic historian, Professor Zafer Toprak,has written the history of the SiseCam group by relying on coiwpany sources. To date, company officials have not accepted the publication of this manuscript.  Industry Policy Networks and the Incentive Regime 49 effective monitoring by the state, SiseCam officials have mostly fullfilled their side of the incentive contracts and avoided exaggerated initial promises as a means of gaining access to more lucrative state incentives or subsidies. Besides the tendency to uphold their end of incentive contracts, the pro-competition orientation in the glass industry may be detected from the fact that the industry has experienced an export boom in recent years. After the about turn in economic policy in the 1980s, in fact, the ratio of exports to total sales remained around 20-22 percent, reaching a record high of 42 percent in 1984 and falling to 14 percent in 1993 (largely due to the overvaluation of the Turkish Lira). Hence, unlike the auto industry, which opposed the government's initiative to open up the economy, SiseCam officials did not concentrate their political efforts on seeking high tariffs and a cap on imports. Instead, they harnessed their energies toward speeding up technological innovations and finding new export outlets."5 Even more revealing in terms of understanding the pro-competition stance of SiseCam officials is that the firm has aggressively sought export outlets abroad, not only after the turn in economic policy, but prior to it as well. In fact, during the 1970's, the heyday of the protectionist ISI years, the ratio of exports to total sales varied between 7 and 16 percent, a not insignificant proportion amounting to $70 million in 1980. According to firm officials, these numbers do not by themselves reveal the cause-effect realtionship between export performance and firm behavior as the positive externalities generated by the exports spread well beyond the increased availability of hard currency. More difficult to quantify but equally tangible was the fact that the export orientation prevented the type of complacency which afflicts industrial monopolies as competition in export markets forced SiseCam officials to exert much self-discipline, avoid the waste of human capital, strive for cost-cutting and technological innovation, and operate at full capacity. When one looks at employment figures in the firm, what is most significant is that despite steady growth in production the firm became leaner by cutting employment from 19,000 in 1989 to 12,000 in 1993. The key question to ask in this context is, why is it that prior to 1981 in the absence of pressures from state officials, and despite holding a hegemonic position in the internal market, SiseCam officials voluntarily sought to engage in export competition? Our findings are that both objective and subjective factors played a role in propelling decision-makers at SiseCam to adopt such a long-term, pro-competition focus. In terms of objective factors, given that the glass industry requires a continous production process and advanced float technology has increasing costs, full capacity utilization is obligatory. At the same time, since the domestic market cannot absorb the total production of the Turkish glass industry, which is currently above 1,000,000 tons per year, it becomes necessary to export the surplus. But this is only a partial explanation for SiseCam's success in the export market. In the auto industry, for instance, technological considerations and the existing scale economies should have dictated similar arrangements but managers at the outset opted for sub-optimum scales designed to satisfy the domestic market. Subjective factors, therefore, such as initial choices made by entrepreneurs and managers, are also important in structuring SiseCam's industrial evolution. In the auto industry since these 15 SiseCam A.S. is quite unique in Turkey in the sense that it spends approximately 27 percent of its budget on R & D.  50 Institutional Influences on Economic Policy in Turkey initial choices were in favor of an inward-orientation certain policy options, such as the pursuit of competitive advantages in the marketplace, were foreclosed and stakes were generated for seeking profits through rent-seeking. In contrast, in the glass industry, which is as much a product of the ISI years as the auto industry, a different initial strategic vision, one which incorporated the objective of conquering external markets through sheer excellence in the design and manufacturing of glass products, made it possible to harness energies toward productive ends as opposed to rent-seeking. Later this vision became institutionalized and thus part and parcel of the everday routines of the company management and this helped enable SiseCam to become a presence in the markets of advanced industrialized countries."6 As important as the initial "vision" to the pro-competition focus are the issues of continuity of upper-level management and the relative absence of personnel turnover in the middle levels. Indeed, between 1953 and 1981 SiseCam was managed by the same individual, Dr. Sahap Kocatopcu (who is a graduate of the elite Galatasaray high school and holds a Ph.D. from MIT), and since then it has seen only two Presidents, both of them coming from within the ranks of the IsBank, the largest commercial bank in Turkey and the majority shareholder in SiseCam. Hence, from the perspective of mechanisms of corporate control, a commercial bank is the main agent of control of SiseCam. Historicaily speaking, the leading role of IsBank in the glass industry dates back to the First Five Year Industrial Plan of 1934, which assigned the duty of initiating a glass industry in Turkey to IsBank.'7 To date, the example of SiseCam, which is reminiscent of corporate governance systems in Japan and Germany, remains much the exception in Turkey. The general rule is that the diversified conglomerates, or holdings--which are commonly active in the auto industry--also own a commercial bank to ease financial constraints. But as far as the strategic managment of the bank is concerned, it is subjected to the priorities of other, industrial and commercial, firms, which make up the diversified conglomerate, rather than vice versa. One drawback of the subordinate position of the bank in the holding hierarchy is that it creates a moral hazard problem by creating the perception in the eyes of the managers in the conglomerate that even in the case of failure they will be bailed out by soft loans. The case of SiseCam, however, is different as due to debt financing and equity participation (which allows IsBank officials to retain a majority in the SiseCam board) the IsBank is well-positioned to exert control. The next question then becomes one of understanding how and in which direction the IsBank exerts control. The literature on corporate governance in Germany and Japan usually draws a positive picture, noting that with the close ties to industry forged through the years, banks are well-positioned to decrease information asymmetries, provide adequate finance with minimum intervention during period of normal profitability, and become quite active during periods of poor performance.'8 To answer the question conclusively about IsBank's influence on 16 SiseCam is now considered a leading glass company with 1.3 percent of world glass production and 4 percent of EC glass production. 17 Turkey is the first non-communist c.: .;try which undertook a (rudimentary) indicative planning exercise. Later, the Second Five Year Industrial Plan was thwarted due to the onset of the Second World War. Following the military coup of 1960 in Turkey, a State Planning Organization was formed and Turkey undertook a more comprehensive planning exercise which still plays a role in economic management. For more details, see Vedat Milor, Forging Industrial Policy in Turkey and France: Bringing Planners Back In, University of Wisconsin Press, forthcoming. 18 See Izak Atiyas, "Restructuring Programs in Transitional Economies," in Vedat Milor, ed., Changing Political Economies: Privatization in Post-Communist and Reforming Communist States. Boulder and London: Lynne Rienner Publishes, 1994.  Industry Policy Networks and the Incentive Regime 51 SiseCam requires more systematic analysis than we were able to conduct. However, on the basis of our interviews with both company and bank officials, the following educated guesses may be ventured. First, direct intervention by IsBank in SiseCam affairs has been limited primarily due to the fact that the company has never been in financial distress over an extended period of time. Moreover, according to Dr. Sahap Kocatopcu, the President of the company for 30 years, IsBank managers agreed at the outset not to become involved in the day-to-day affairs of SiseCam provided the company could remain afloat. At the same time, however, SiseCam managers acknowledge that without the cushions provided by the bank, sometimes through direct loans but also through loan guarantees, it would have been much more difficult to emerge as the vertically integrated, muti-divisional firm which SiseCam eventually became. Here one can make the observation that the typical big business corporation in Turkey, like in Latin America and East Asia, is a horizontally integrated, multi-sectoral diversified conglomerate. One reason for the emergence of this type of firm as the dominant form of enterprise in the post-WWII period is that in an attempt to take advantage of state subsidies which were disbursed in an indiscriminating fashion--with the exception of some of the East Asian economies which had more selective industrial policies--business enterprises found it lucrative to diversify into as many fields as possible. Consequently, breadth was preferred over depth and, due to the poor degree of specialization and highly leveraged nature of some of the business ventures, many diversified conglomerates came to oppose trade liberalization policies. SiseCam, in contrast, approximates more closely the Western experience studied by Chandler and Williamson in the sense that the firm grew over the years not artificially by solely taking advantage of the largesse of the public purse but by attempting to reduce transaction costs. The presence of IsBank facillitated this process by opening up some options to SiseCam while foreclosing others. Complacency was foreclosed since the managers of SiseCam knew that in the case of poor business performance the bank could interfere by becoming active in screening and replacing managers. The bank also checked the tendency of managers to overspend but did not force SiseCam to do business only with itself On the contrary, SiseCam managers, especially in the 1980's after the ending of financial repression in the country, have often tapped international financial resources with or without the approval of IsBank. Perhaps most important of all, the presence of IsBank made it possible for SiseCam officials to be more insulated from political pressures. As the Treasury owns 41 percent of IsBank it is indirectly a shareholder in SiseCam. Hence, SiseCam is potentially subject to pressures emmanating from this politicized entity. In the early 1990's, for example, these pressures were deeply felt when SiseCam managers desired to close down the loss-making glassware plant in Sinop. Annecdotal evidence suggests that managers successfully warded off political pressures to keep the plant open by pointing to IsBank as the agency with which politicians had to settle their dispute. IsBank, in turn, was able to withstand these pressures partially due to the prescience of its own management but also because its own incentive structure makes it contrary to its own interest to cave in to such pressures. IsBank is primarily a private bank in which the Pension Fund of its employees is the largest shareholder (at 37 percent). Hence, IsBank is under direct pressure to monitor the profitability of its industrial participation since poor performance translates into shrinking pensions for its employees.  52 Institutional Influences on Economic Policy in Turkey The problem is that the substantial Treasury share in IsBank can be a potential source of conflict between the sometimes incompatible goals of profitability versus maximization of employment. In the 1980s, for example, Treasury has twisted the arm of IsBank to head lending consortia to rescue certain distressed companies in which IsBank eventually became a shareholder. An increasing proliferation of IsBank participation in unprofitable ventures then creates the possibility that the bank has or will pressure SiseCam to cross-subsidize its losses. We have no proof that this has happened but our hunch is that it may happen. Moreover, SiseCam managers, while acknowledging the overwhelmingly positive contribution of IsBank to their evolution, are now eager to reduce their dependence on the bank. It is therefore possible that in a changing external environment characterized by the impending integration of Turkey into the EC market the past conguence of preferences between IsBank and SiseCam managers may soon be replaced by more complex game-theoretic interaction patterns. In the meantime, one can conclude that, in the context of a weak state and despite the presence of a monopolistic market structure, the existence of a dense network linking glass industry leaders to bank officials may create a governance structure that is pro-competition and export-oriented overall. C. Growth-Oriented Network in the Readywear Clothing Industry The case of readywear is interesting in that the network in question was created from scratch. During the import substitutionist years the myriad of small-scale manufacturers in the industry which competed in the domestic marketplace were bereft of a collective organization to advance common goals. Given their small scale they did not have the requisite political and economic clout to be able to enjoy the benefits of personalized contacts with decision-makers. Things did not automatically improve with the about-turn in development strategy after 1980 either as the most attractive rents attached to exporting, i.e. the export tax rebates, were skimmned off by large export houses affiliated with various Holdings. Moreover, the EEC and U.S. quotas affecting the industry were also monopolized by large manufacturers (Onis, 1992). The Readywear Manufacturers' Association, the GSD, which was actually founded in the late 1970's by small scale industrialists, then became quite active in the 1980's in the pursuit of a more fair system to distribute quotas, one which would not discriminate against small business persons. Moreover, in 1987, 97 small-scale industrialist members of the GSD formed what has since its inception become the largest export house in the country, the GSD AS (Readywear Manufacturers' Association--Joint Stock Company). The GSD AS was established with the purpose of enabling small business persons to partake of benefits of the incentive regime which individually they could not have accessed given that these benefits were made available primarily to large-scale exporters (Onis, 1992). Interestingly, even after the ending of the export tax rebates, which had been the most lucrative subsidy, GSD AS continued to flourish by increasing exports. In contrast, many Holding-affiliated export houses foundered after considerable abuse of the export tax rebate system took place and international pressures from the EEC and the World Bank forced the government in 1988 to abolish the rebates and curb some other benefits (Ilkin, 1991). Most specifically, members of GSD AS exported U.S. $690 million worth of readywear clothing in 1992, i.e. a fifth of total readywear exports, as opposed to a meager sum of $218 million in 1987, the first year of this organization's existence (GSD AS, 1993). Moreover, this phenomenal growth is not due to an  Industry Policy Networks and the Incentive Regime 53 increase in membership since GSD AS froze membership at 100.'9 Nor may this increase in growth be attributed to increases in export subsidies as not only did the export tax rebate program end but export subsidies have generally diminished during the life of this organization. Hence, ceteris paribus, organizational practices and the type of network spawned by these practices seem to have made a positive difference in the governance structure of this industry. 1. Private Sector Economic Governance Our main finding in this context related to the distinguishing feature of the type of growth- oriented network in readywear clothing is that, it enabled the participants to reap the benefits of cooperation without forsaking the advantages of competition. Cooperation enabled the small business actors who had previously been left out of the particularistic network loop to become influential at the policy-making stage of the incentive regime. In addition, thanks to cooperation, small-scale industrialists were able to derive tangible benefits such as access to export quotas and preferential credits, access to information about market conditions and the macro environment, and so on. In the meantime, potential abuse of cooperation (such as free-riding) was limited and a competitive environment maintained due to the following factors. First, access to incentives, which was mediated by the leadership of the organization, was linked to performance, unlike as has been the case has with the TUD in the Treasury. Second, thanks to the existence of transparent, credible rules governing the distribution of incentives, uncertainties regarding the distribution of future growth were reduced and incentives were created to channel energies toward competition in the marketplace rather than competition to lobby state officials. Last but not least, members were not allowed to free-ride by receiving the information provided by others, not only because they knew that in this case they would be put out of the information circle and lose advantages associated with it, but also because certain organizational arrangements explained below made it very difficult to engage in free-ridership. As a result of all these factors, transaction costs of monitoring individual business persons in the use of incentives have declined given that a form of self-enforcing monitoring was created upon the discovery that improved performance in the competitive marketplace is the quickest road to keep oneself in the virtuous loop. How has all of this happened? Our study indicated that success depended, at least initially, on the choice of certain entrepreneurs. Thus, unlike as was the case in the OSD, these entrepreneurs directly took part in the affairs of the GSD and, after the GSD's export house was formed, clearly delegated authority to a professional manager who had been a high level officer in the Treasury (the previous head of the Center to Promote Exports) This manager's actions subsequently allowed for the institutionalization of a certain principle of "reciprocity" between the state and the private beneficiaries of the incentive regime. This reciprocity was based on mutual obligations in the sense that the state sanctioned the GSD, conferred some benefits to members of the GSD AS and allowed these organizations to participate in the decision-making process in exchange for their demonstrated willingness to monitor and discipline those members who abused the incentive system. As such, it became possible to institutionalize the initial organizational choices over time and to reduce the transaction costs of incentive monitoring by constructing a growth-oriented network in an industry dominated by a myriad of small players. 19 The reason for this number is that the Turkish Company Law stipulates that when membership exceeds 100 then the company has to float shares in the stock market.  54 Institutional Influences on Economic Policy in Turkey Of course, the obvious question to ask, how has successful monitoring occurred? We have not found a case of any member who was put out of the circle due to abuse and certainly the discovery and documentation of such a case would have been helpful in building credibility for our argument. But does the lack of such an example imply the absence of monitoring? Not necessarily if monitoring has become self-enforcing over time. There are, indeed, good reasons to believe that this may have been the case. For one thing, members derive real benefits from membership to the GSD and to the privileged export club of the GSD. One such benefit is that, as discussed in the previous section, small business persons need a peak association to have access to decision-makers. The archives of GSD contain some documents which illustrate the dimensions and consequences of this access. That is, several letters have been sent to decision- makers, including top bureaucrats and government ministers, with suggestions to reformi the incentive system in the case of readywear clothing. These demands are not geared to particularistic concerns, rather they represent certain collective aims geared to facilitate long-term transformation and restructuring of the industry. Besides, since this process--unlike the case in the other two industries--is transparent, prisoner's dilemma-type situations which plague actors enmeshed in the rent-seeking networks are avoided. Another benefit of membership is that GSD is very powerful in the quasi-public organization which distributes quotas for the U.S. and the EEC (this organization is called ITKIB, the Exporter's Union for Readywear and Textile Manufacturers). Membership in GSD AS also delivers many collective goods. One such good is access to preferential EximBank credits for exporters for which many small business persons may not be qualified due to their limited size.20 GSD AS also addresses liquidity problems of the members and provides working capital, a very crucial service in a country where due to certain rigidities and distortions in the financial markets small business in industry rarely gain access to banking services. Another collective good generated by the GSD AS is access to crucial information concerning market conditions and the macro environment. This is because the SPO no longer delivers information useful to small business and the contacts between the leaders of this organization and decision-makers enables business persons to find out about the macro environment and the future intent of the government on the macro front, a crucial and scarce resource in a country where frequently changing and unpredictable decisions impact business activity to a great degree. Given the type of benefits derived from membership in the readywear clothing associations it becomes easier to understand why it was that flagrant abuse did not occur. The member firms were in a repeated relationship and at the same time they belonged to a powerful club which froze membership; a firm which cheated could be replaced by another one since there were many which were waiting at the door. To maintain a good public image has always been very important to the readywear associations to legitimize the fact that this was an industry which greatly benefitted from incentive certificates. Thus, a member which harmed this image could have been ostracized. Hence, the resulting peer pressure partly explained the continuing good performance and the internalization of monitoring functions. At the same time, and this is equally as important, the leadership did have direct instruments to induce growth-oriented behavior. Most specifically, the leadership mediated between the TUD and individual business persons, and--most crucially--it linked rewards to good performance. It is certainly possible that abuse could have occurred in the 20 Currently there is a U.S. 5 million dollar per year requirement to gain access to these credits.  Industry Policy Networks and the Incentive Regime 55 sense that some members may have inflated their achievement or exaggerated their needs to get more resources. It is also possible that some may have tried to conceal crucial information necessary for the association to design policy, such as information about their marketing outlets, and instead may have attempted to steal information provided by others. Yet certain organizational features made these outcomes highly unlikely. For one thing, since decisions were made collectively by often-changing boards of directors who were entrepreneurs themselves, those who inflated their achievements would have a hard time convincing their peers who were equally knowledgeable about conditions in the industry. As to the question of free-ridership in information, with respect to matters such as export markets for example, the leadership tacitly acknowledged that this was a serious potential problem. In a context like Korea, the type of "trust" which is necessary for the continued two-way flow of information has been created over the years thanks to repeated relationships between the parties, to the willingness of business persons to relegate decision-making authority to their professional association, and to the insularity of top bureaucrats from political and social pressures. In Turkey, on the other hand, as the professional manager of the GSD AS acknowledged, trust had to be created from scratch (Author's interviews, 1994). Hence, the organization had to construct a system in which members could believe that giving crucial information was a safe practice and no member including the ones in the board were empowered to free-ride on this information. Such a system was constructed by taking the management consultant company MacKenzie to emulate in the sense that no single employee would have access to classified files of more than a handful of member companies while, at the same time, employee morale and loyalty to the company would be kept high by attractive salaries, good working conditions and promotion opportunities based on merit. Additionally, the organization adopted the maxims of "transparency," "solidarity," "efficiency" and "respectability" as operational principles to guide relationships among members and between the association and the state (these principles are also laid out in the annual reports). Our interviews with several entrepreneur-members and employees of this association, as well as with the state officials who have dealt with them, confirmed that everyday practices did not greatly deviate from these principles. Can this success in export performance continue without generous subsidies? The breakthrough in export performance for GSD AS members, as we have indicated, came after some of these benefits were rescinded. As of today, as can be seen from Tables 12 A, B & C below, GSD members are the least sanguine in the three industries surveyed about the usefulness of subsidies. It is also interesting that most of these entrepreneurs do not question the general logic of incentives, which they think may be useful for manufacturing (Table 12). This is symptomatic of the lingering effect of an etatiste ideology among the business persons who do not necessarily ask for subsidies in their own industry. It is not possible to say for how long this attitude, which is a legacy of the past, will continue, but the fact that certain beneficiaries of the incentive regime perceive the subsidies to their industry as not particularly useful nor, perhaps, necessary, is a promising one. That is, contrary to conventional wisdom, easy access to rents may not always translate into a protectionist attitude for the infinite prolongation of these rents.  56 Institutional Influences on Econonic Policy in Turkey Table 12: Business Attitudes toward 1994 State Incentive Policies (figures in percentages) A. Utility of Incentive Policies in Generai Very Very useful Useful Useless useless (N) Auto 100 0 0 0 (19) Glass 75 13 13 0 (8) Ready wear 61 33 6 0 (18) All 80 16 4 0 (45) B. Utility of Incentive Policies for Own Industry Very useful Very No Useful Useless useless answer (N) Auto 21 63 11 5 0 (19) Glass 0 63 38 0 0 (8) Ready wear 0 28 56 11 6 (18) All 8 49 33 7 2 (45) C. State's Management of Incentive Policies in Own Industry Very Very No effective Effective Ineffective ineffective answer (N) Auto 5 47 42 5 0 (19) Glass 0 25 63 1 3 0 (8) Readywear 5 0 74 11 6 (18) All 4 24 60 9 2 (45) Source: Author's interviews.  6 Policy Recommendations The findings of this research report are that the institutional arrangements upon which the Turkish incentive and subsidy regime rest prevent the effective functioning of this regime. Given an absence of bureaucratic capacity, and given the job insecurity of the bureaucrats who administer the system, various forms of incentives, such as tax rebates, preferential credits and grants, end up becoming "give aways," pure and simple. The political logic driving this system is that these generous incentives help expand and consolidate the coalitional basis of the government. In the early 1980's such operation of the incentive regime helped create stakes among internal-market oriented industrialists in an export-led development strategy. As an export-led strategy has largely been achieved, it is difficult to find a continuing political rationale for the indefinite subsidization of industry. This subsidization is especially ineffective in highly concentrated industries, such as the auto industry, where leading industrialists are able to strike particularistic relations with policy makers geared around rent-seeking. There are two exceptions to the general pattern of ineffectiveness in the incentive regime. First, export incentives fare much better than investment incentives in terms of attaining stated objectives. This is because in the case of exports there is a clear, externally validated measure of success. It is also easier to monitor private sector use of export incentives given lower transaction costs in comparison to investment incentives. Second, there are cross-industry differences in the effectiveness of the incentive scheme. The policy is more successful in competitive industries, such as the readywear clothing industry, where a business organization mediates between subsidy recipients and state officials. Business organizations in concentrated industries in Turkey, such as the OSD in the auto industry, are a drag on efficiency and growth as they are distrustful of competition both domestic and foreign. As such they approximate what Olson refers to as "distributional coalitions" particularly insofar as they are dominated by their oligopilistic clientele. In industries with competitive market structures, on the other hand, different forms of business associations may emerge. And their influence may be positive for growth when they enable their members, which are small business persons, to reap the benefits of cooperation without forsaking the advantages of competition. To the extent that such organizations can emerge from the bottom up in competitive industries aimed at exports, the effective implementation of the incentive regime will benefit from decentralization. This can be accomplished through "private sector economic governance" whereby the state can decentralize important elements of an effective incentive disbursement process, such as initial screening of applicants and monitoring of performance, in exchange for the 57  58 Institutional Influences on Economic Policy in Turkey commitment by industry association officials to monitor and enforce performance expectations of beneficiaries with respect to growth and exports. In the case of concentrated industries, on the other hand, it is doubtful whether decentralization alone can improve effectiveness. And, given that it is not easy to eliminate the existing rent-seeking networks via administrative fiat, or legal changes (such as anti-trust laws), the ending of incentives in their present form would be the best policy. The evidence presented in this report supports the continuing subsidization of small business investments aimed at exports, especially when the decentralization of some elements of the incentive process is possible. However, the effectiveness of the incentive regime would be greatly enhanced through legal and bureaucratic reforms designed to help prevent the formation of rent- seeking networks The problem is that such networks are made possible as the existing bureaucratic structure does not sufficiently insulate economic policy makers. More specifically, as recruitment, promotion and other aspects of the bureaucratic career have become politicized, as opposed to being oriented around stable and meritocratic principles, economic bureaucrats have become subject to pressures from politicians and well-positioned business persons to manipulate the incentive regime for narrow, private ends. This is particularly the case for officials of the TUJD, the office which administers the incentive regime. As a consequence of this lack of bureaucratic insulation the effectiveness of the incentive regime is compromised. There is little initial screening of applicants according to technical criteria and instead access to incentives and subsidies is frequently determined by political criteria. There is only limited monitoring of the performance of incentive regime beneficiaries and little disciplinary action taken with respect to abuses of the program. Such problems further encourage an "adverse selection" mechanism as many competent investors are crowded out of the application process if they are not in the network loop. Thus, only a combination of legal safeguards and significant changes in the bureaucratic recruitment, promotion and career pattems can alter the existent perverse incentive structure facing state officials. In the event that the political will to undertake such reforms is lacking, and if the incentive system is not going to be eliminated altogether in the near future, then three complementary steps should be taken. First, the system in its present form lacks transparency, and the involvement of private sector associations in the initial design stage of the programs will be helpful in making the system more transparent. Second, a plethora of government regulations and frequent, decree-based, changes in these regulations often create confusion and uncertainty as to government intentions. This encourages private sector persons to engage in wasteful lobbying activity. To end this situation a committee involving representatives from both public and private sectors should be formed in order to simplify regulations and set up basic rules as to when and under what circumstances exemptions may be made from clearly stated general principles of the incentive regime. Finally, until the banking system is privatized, the granting of cash and other benefits to incentive regime beneficiaries should be removed from state-owned development banks and entrusted to private banks. As the Turkish Eximbank, albeit owned by the Treasury, works together with private commercial banks and does not directly hand out money, it should be considered a private bank. Finally, one can raise the question of the interaction between global forces and domestic structures. International pressures to liberalize the Turkish economy and the evolving process of integration to the EEC may deter some members of protectionist rent-seeking networks from  Policy Recommendations 59 continuing to operate on the basis of very short-term horizons. Thus, the auto industry for instance, grudgingly accepted the 1991 reform of the trade regime even though it substantially lowered the tariffs for the industry. As Turkey now faces the challenge of custom unity with the EEC, it is possible that new reforms will be implemented in the near future despite the possible resistance by rent-seeking interests. At the same time, however, it would be erroneous to overlook the capacities of social and political forces which may attempt to delay--if not thwart-- further economic liberalization. Experience elsewhere suggests that, especially if these forces choose to ally with workers who enjoy a wage premium thanks to employment in a protected industry oriented to the domestic market, government may be forced to adopt populistic economic policies and invent more subtle protectionist trade policy measures. For example, recent events in the summer of 1994 in Turkey brought auto industry leaders and trade union officials together to denounce the government's alleged "auto-industry intentions." Thus, unless institutional reforms aimed at the promotion of developmental networks as a means to alter the power balance between them and protectionist forces are undertaken, the expected benefits from globalization may not materialize in the near future.  60 Institutional Influences on Economic Policy in Turkey  References Alchian, A.A. and H. Demsetz. 1973. "The Property Rights Paradigm," Journal of Economic History, 33, March. Amsden, Alice. 1989. Asia's Next Giant. Oxford: Oxford University Press. Arrow, K. 1974. The Limits of Organization. New York: Norton. Arslan, I. 1993. "A Review of Turkish Incentive Systems," World Bank (December 10). Atiyas, I. and H. Ersel. 1994. "The Impact of Financial Reform: The Turkish Experience," in G. Caprio, Jr., I. Atiyas and J. Hanson, eds., Financial Reform: Theory and Experience. Cambridge: Cambridge University Press. Atiyas, I. 1994. "Restructuring Programs in Transitional Economies," in Vedat Milor, ed., Changing Political Economies: Privatization in Post-Communist and Reforming Communist States. Boulder and London: Lynne Rienner Publishes. Balassa, B. 1981. "The Process of Industrial Development and Alternative Development Strategies," in The Newly Industrializing Countries in the World Economy. New York: Pergamon. Bianchi, R. 1984. Interest Groups and Political Development in Turkev. Princeton: Princeton University Press. Bhagwati, Jagdish N. 1982. "Directly Unproductive, Profit-Seeking (DUP) Activities," Journal of Political Economy, 90, October. Bolton, P. 1990. "Renegotiation and the Dynamics of Contract Design," European Economic Review, 34: 303-310. Bozkurt, 0. 1980. Functionaries: A Sociological Perspective in Turkish Public Bureaucracv (in Turkish). Turkiye ve Ortadogu Amme Idaresi Enstitusu. Buchanan, James M., R. Tollison and G. Tullock (eds.). 1980. Toward a Theorv of the Rent-Seeking Society. College Station: Texas A & M University Press. Bugra, A. 1994. State and Business in Modem Turkev. New York: State University of New York Press. Byung-Kwon, Cha, et. al. 1989. "The Political Economy of Fiscal Policy in the Republic of Korea," in M. Urrutia, et. al., eds., The Political Economy of Fiscal Policy. Tokyo: The United Nations University. Campos, Ed. "Insulation Mechanisms and Public Sector--Private Sector Relations," Part 1 of The Institutional Foundations of High Speed Growth in the High Performing Asian Economies. The World Bank. Colander, David C., ed. 1984. Neoclassical Political Economy: The Analvsis of Rent-Seeking and DUP Activities. Cambridge, MA: Balinger. Corbett, J. and C., Mayer. 1993. "Financial Reform in Eastern Europe: Progress with the Wrong Model," Oxford Review of Economic Policy, Vol. 18, No. 4. Demsetz, Howard. 1967. "Toward a Theory of Property Rights," American Economic Review, 57, May. 61  62 Institutional Influences on Economic Policy in Turkey Deyo, Frederick C., ed. 1987. The Political Economy of the New Asian Industrialism. Ithaca, NY: Comell University Press. Duruiz, L. and N. Yenturk. 1994. "The Competitive Strength of the Turkish Readywear Clothing Industry" (in Turkish). Istanbul: GSD. Fama, E. 1980. "Agency Problems and the Theory of the Firm," Journal of Political Economy, 88: 288-305. Furubotn, E. and S. Pejovich. 1972. "Property Rights and Economic Theory: A Survey of Recent Literature," Journal of Economic Literature. Geddes, Barbara. 1990. "Building State Autonomy in Brazil 1930-1964," Comparative Politics, 22. Guvemli, 0. 1992. "Developments in Investmnent Promotion in Turkey" (in Turkish). Paper presented at the Izmir Economic Congress, Izmir. Granovetter, Mark. 1985. "Economic Action and Social Structure: The Problem of Embeddedness," American Journal of Sociology, 91, #3, November. GSDAS. 1993. Activity Report, 1/1/1993 -- 12/31/1993 (in Turkish). Gulfidan, Sebnem. 1993. Big Business and the State in Turkev: The Case of TUSIAD. Istanbul: Bogazici University Press. Haggard. S. 1990. Pathways from the Peripherv. Ithaca: Cornell University Press. Haggard, Stephen and Chung-In Moon. 1990. "Institutions and Economic Policy," World Politics, 2, January. Hale, W. 1981. The Political and Economic Development of Modem Turkey. London: Croom Helm. Heper, M. 1976. "The Recalcitrance of the Turkish Public Bureaucracy to Bourgeois Politics: A Multi- Factor Political Stratification Analysis," The Middle East Journal, 30 (Autumn). Heper, M. 1985. The State Tradition in Turkey. London: The Eothen Press. Heper, M. 1989. "Motherland Party Govemments and Bureaucracy in Turkey, 1983-1988," Govemance: An Intemational Joumal of Policy and Administration, Vol. 2, No. 4 (Oct.). Heper, M. 1991. "Interest Group Politics in post-1980 Turkey: Lingering Monism," in M. Heper, ed., Strong State and Economic Interest Groups: The Post-1980 Turkish Experience. Berlin and New York: Walter de Gruyter. Heper, M., Et. Al.. 1980. "The Role of Bureaucracy and Regime Types," Administration and Society, Vol. 12, No. 2 (August). Huntington, S. and J. Nelson. 1976. No Easy Choices: Political Participation in Developing Countries. Cambridge: Cambridge University Press. 1kin, S. 1991. "Exporters: Favored Dependency?" in M. Heper, ed., Strong State and Economic Interest Groups: The Post-1980 Turkish Experience. Berlin and New York: Walter de Gruyter. Jensen, M. and W. Meckling. 1976. "Theory of the Firn: Managerial Behavior, Agency Costs and Ownership Structure," Joumal of Financial Economics, 3: 306-360. Johnson, Chalmers. 1982. MM and the Japanese Miracle. Stanford, CA: Stanford University Press. Kalaycioglu, E. 1991. "Commercial Groups: Love-Hate Relationship with the State," in M. Heper, ed., Strong State and Economic Interest Groups: The Post-1980 Turkish Experience. Berlin and New York: Walter de Gruyter. Kandori, M. 1992. "Social Norms and Community Enforcement," Review of Economic Studies, 59: 63-80.  References 63 Katzenstein, Peter. (ed.). 1977. Between Power and Plentv: Foreign Economic Policies of Advanced Industrial States. Madison: University of Wisconsin Press. Katzenstein, Peter. 1985. Small States in World Markets: Industrial Policy in Europe. Ithaca, NY: Cornell University Press. Krueger, Anne 0. 1974. "The Political Economy of the Rent-Seeking Society," American Economic Review, 64, June: 291-303. Krueger, Anne. 0. 1980. "Trade Policy as an Input to Development," Amerir n Economic Review, 70 (2): 288-92. Krugman, P., ed. 1986. Strategic Trade Policy and the New International Econ. w- Cambridge: MIT. Lal, D. 1983. The Poverty of Development Economics. London, IEA, Hobart Paperback 16. Milor, Vedat. Forzing Industrial Policy in Turkey and France: Bringing Planners Back In, University of Wisconsin Press. Ministry of Industry and Trade. 1991. Turkey: Economic and Industrial Report. Anakara, Turkey: Ministry of Industry and Trade. Moe, Terry M. 1984. "The New Economics of Organization," American Journal of Political Science, 28, November. North, Douglas M. 1974. "Government and the Cost of Exchange in History," Journal of Economic History, 44, June. North, Douglas M. 1992. Institutions and Economic Performance. Cambridge: Cambridge University Press. Olson, M. 1982. The Rise and Decline of Nations. New Haven: Yale University Press. Oncu, A. 1980. "Chambers of Industry in Turkey: An Inquiry into State-Industry Relations as a Distributive Domain," in E. Ozbodun and A. Ulusan, The Political Economv of Income Distribution in Turkey. New York: Holmes and Meier Publishers. Onis, Ziya. 1991. "The Logic of the Developmental State," Comparative Politics, (October). Onis, Z. 1992. "Organization of Export-Oriented Industrialization: The Turkish Foreign Trade Companies in Comparative Perspective," in T. Nas and M. Odekan, eds., The Economics and Politics of Turkish Liberalization. Bethlehem, PA: Lehigh University Press. Onis, Z. and S. Webb. 1994. "Turkey: Democratization and Adjustment from Above," in S. Haggard and S. Webb, eds., Voting for Reform: Democracy. Political Liberalization and Economic Adjustment. Oxford: Oxford University Press. OSD. 1994. The 1994 Economic Crisis and its Impact on the Turkish Auto Industry (in Turkish). Istanbul: OSD. Perrow, Charles. 1990. "Economic Theories of Organization," in Zukin and DiMaggio, eds., Structures of Capital. Cambridge: Cambridge University Press. Putnam, R. 1993. Making Democracy Work. Princeton: Princeton University Press. Rodrik, D. 1988. "How Much of the Export Boom is Fictitious?," Topume ve Bilim. Istanbul. Rueschemeyer, Dietrich and Peter Evans. 1985. "The State and Economic Transformation: Toward an Analysis of the Conditions Underlying Effective Intervention," in Evans, Peter B., Dietrich Rueschemeyer and Theda Skocpol, eds. 1985. Bringing the State Back In. New York: Cambridge University Press. Schneider, Ben Ross. 1993. "The Career Connection," Comparative Politics, April.  64 Institutional Influences on Economic Policy in Turkey Shapiro, H. and L. Taylor. 1990. "The State and Industrial Strategy," World Development, 18. Stiglitz, Joseph E. 1993. "Some Lessons from the Asian Miracle," Mimeo, The World Bank. Wade, Robert. 1990. Governing the Market: Economic Theory and the Role of Government in Taiwan's Industrialization. Princeton, NJ: Princeton University Press Waterbury, John. 1988. "Coalition-Building, Export-Led Growth and the Public Sector in Turkey." Paper prepared for "Dvnamics of a Mixed Economy: The Turkish Case," Middle East Studies Association Annual Meeting, Los Angeles, November 2-6. Wilks, S. and M. Wright (eds.). 1987. Comparative Government-Industry Relations. Oxford: Clarendon Press. World Bank. 1993. The East Asian Miracle: Economic Growvth and Public Policy. Washington: The World Bank. Yasar, E. "Turkish Textile and Readywear Clothing Sector (in Turkish). Istanbul: ITKIB Publication No. 0041. Zysman, J. 1983. Governments, Markets and Growth. Ithaca, NY: Cornell University Press.  Appendix I Selectivity Of Incentive Regime: Traditional Criteria Statistics on actual investments receiving subsidies, particularly their inter-sectoral and regional distribution, are not available since there is no published data on realization. Available statistics which can be compiled from the State Planning Organization's annual activity reports are based on the number of certificates granted to investors, rather than realized projects. A. Magnitude of Incentive Program The fiscal costs of the incentive programs in the 1980's amounted to 4 to 5 percent of the GNP in a given year. According to a rough estimate of the cost of the main investment incentives for instance, these costs, calculated as a percent of GNP, amounted to 2.2 in 1988, 2.1 in 1989, 1.8 in 1990 and 1.6 in 1991. The corresponding figures for the cost of export incentives are, respectively, 2.4, 2.3, 2.1 and 2 (Arslan, 1993). Our own calculations which are given in the two tables below and which are based on Treasury data also indicate that the cost has declined over time relative to GNP but is still significant. Also note that the calculation of the fiscal costs of export and investment incentives takes into account two major tax incentives (customs duty exemption and investment allowance) and direct subsidies such as cash grants from the resource utilization fund but excludes some cash grants which are hard to track such as the transportation premium (navlun primi in Turkish) and preferential credits as well. Thus, the fiscal costs of the export and investment incentive programs illustrated in Table A-1 and Table A-2 below are likely to underestimate the real costs of the program. On the other hand, the implicit assumption in the calculation of such fiscal costs is that, without the incentives and subsidies, the proposed activities would have taken place anyway. Based on the costs calculated, tax incentives are slightly more important than direct grants. In 1991 and 1992, for instance, tax incentives made up 53 and 64 percent of the total costs, respectively. B. Sectoral Distribution of Incentives The sectoral distribution of investment encouragement certificates from 1980 to 1992 is given in Table A-3 below. Table A-4, on the other hand, disaggregates the general category of investment incentives by calculating the type of "activity" which was supposed to be promoted. A statistical analysis of the data in Table A-3 shows a preference for manufacturing over other sectors in the 1980's. Within manufacturing there was a significant preference for Textiles, and a mildly significant preference for Food and Beverages. Other interesting points that can be drawn from the data include the sharp jump in subsidies for Transportation Vehicles in 1981. Table A-4 indicates that new investment was favored over other investments, and the significant increase in subsidies for Leasing in 1992 is noteworthy. 65  66 Institutional Influences on Economic Policy in Turkey The sectoral distribution of the export incentive certificates, on the other hand, is calculated on the basis of certificates given for the Foreign Exchange Allocations and Customs Duty Exempilons which are the most important export incentives. The TUD approves these foreign exchange allocations for export incentive certificate holder to import duty free raw materials, packing goods and intermediate goods used for their exports. As such, this is the most highly sought after tax incentive. In addition, the TUD started issuing global and/or project based export incentive certificates in 1990. The unique nature of this new incentive program is that, contra.y to the other incentives which are targeted to the export activity per se, this program aims to reward the producer for its contribution to the value added exported. The idea here is inspired by the Korean system of export incentives which encourages exports through the rationalization of production and new capacity creation as opposed to drawing on existing capacity made possible by a suppression of internal demand. A statistical analysis of Table A-5 indicates a preference for Manufacturing over other sectors, and an overwhelming preference for Textiles among the Manufacturing sector. Also noteworthy is the support for Iron and Steel in 1980 and Electrical Machinery in 1981. These results are further confirmed by Table A-6 for the years 1990-92 which indicates a strong preference for Manufacturing and also a strong, if not as overwhelming, preference for supporting investments in Textile and Clothing geared to exports. C. Regional Distribution of Incentives Finally, it is important to note that the incentive system in Turkey aims to discriminate in favor of investments in "priority regions" (i.e., most of the Eastern half of the country), and at the expense of "developed regions" (the area around Istanbul and the eastern end of the sea of Marmara as well as around Ankara and Ionia). From the investor's point of view it is much more attractive to have investment classified as in a "priority region" since benefits are consideiably higher. Whether the TUD has the requisite monitoring capacity to ensure the undertaking of the proposed investment in the "priority region" is, of course, another matter. And this issue will be taken up in the next section. On the basis of these two tables, it is interesting to note that there has been a significant jump in 1990 in the number of certificates given for investments in priority regions. As to the sectoral allocation of these certificates between 1980 and 1992, Manufacturing displays a steady increase (between 22 percent in 1982 to 68 percent in 1992) while Agriculture and Services, which started out collecting the bulk of certificates in 1982 (27 percent and 47 percent respectively), display a steady decrease with the exception of 1990 when 59 percent of the certificates were issued for Agriculture. Within Manufacturing on the other hand, Food and Beverages seems to be the leading industry, followed by Textiles and Clothing and, as a distant third, Clay and Cement products.  Appendices 67 Table A-I: Total Export Incentives, 1982-1992 Year Nominal Mil. TL Real Mil. TL 1000 S 1982 86,716 4,144,4417 538,809 1983 148,990 5,559,229 665,045 1984 329,060 8,384,654 901,905 1985 332,908 5,985,348 642,258 1986 435,913 6,144,859 651,209 1987 660,927 6,689,143 772,391 1988 1,016,695 6,398,311 715,599 1989 1,127,216 4,299,357 531,510 1990 1,287,952 3,284,180 493,919 1991 1,886,482 3,153,129 452,410 1992 2,739,943 2,739,943 397,813 TOTAL 10,052,802 56,782,570 6,762,868 Table A-2: Total Investment Incentives, 1985-1992 Year Nominal Mil. TL Real Mil. TL* 1000 S 1985 555,281 9,983,388 1,071,268 1986 820,969 11,572,811 1,226,443 1987 1,189,426 12,038,003 1,390,020 1988 2,144,240 13,494,229 1,509,220 1989 2,852,185 10,878,627 1,344,875 1990 2,493,288 6,357,695 956,155 1991 3,890,717 6,503,075 933,059 1992 6,308,846 6,308,846 915,983 TOTAL 20,290,951 77,136,673 9,347,023 Source: Undersecretariat of Treasury and Foreign Trade. Calculated by Wholesale Price Index 1992 = 100. The value of S per investment incentives calculalted from Treasury data for export incentives.  Table A-3: Sectoral Breakdown of Invetainent Incentive Cerfificdles, 1980-1992 (%0 Sectors 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 Agriculture 18% 15% 4% 16% 9% 5% 3% 7% 4% 14% 45% 8% 2% Minining 3% 2% 4% 3% 5% 5% 6% 6% 6% 4% 2% 4% 4% Manufacturing 69% 32% 46% 41% 51% 52% 54% 55% 56% 48% 39% 57% 64% 0 Food& Beverages n.a. 7% 6% 5% 8% 11% 6% 8% 7% 8% 11% 10% 8% Textile & Clothing n.a. 2% 16% 8% 13% 17% 14% 13% 13% 15% 10% 14% 22% Forestry Products n.a. 1 % 1 % 1 % 1 % 1 % 10% 2% 2% 2% 2% 2% 1 % V: co Paper n.a. 0% 1 % 0% 0% 1 % 0% 1 % 1 % 0% 0% 1 % 1 % Leather& Leather Products n.a. 1% 1% 1% 1% 2% 1% 2% 2% 1% 1% 2% 1% CD Rubber n.a. 0% 0% 0% 1% 1 % 1 % 1 % 2% 2% 2% 3% 2% O Chemicals n.a. 1 % 1 % 3% 4% 3% 2% 2% 2% 2% 1 % 3% 2% Glassware n.a. 0% 1 % 0% 0% 1 % 1 % 1 % 1 % 2% 0% 1 % 1% O 0 Iron e Steel n.a. 0% 1 % 2% 2% 2% 1 % 1 % 1 % 1 % 1 % 1 % 2% O Non-Ferrous Metals n.a. 0% 0% 0% 1 % 0% 0% 1 % 1 % 0% 0% 1 % 1 % 0 Transpoft Vehkies n.a. 14% 12% 4% 4% 1 % 3% 2% 3% 4% 2% 4% 5% 0 Metal Goods n.a. 1 % 1 % 2% 2% 3% 3% 2% 3% 2% 2% 3% 3% o Measunng Devices n.a. 0% 0% 1 % 1 % 0% 1% 1 % 1 % 1 % 0% 0% 1 % Machinery n.a. 1 % 3% 1 % 1 % 2% 1 % 1 % 1 % 2% 1 % 2% 2% Electilcal Machinery n.a. 0% 1 % 1 % 1% 1 % 1 % 1 % 1 % 1 % 1 % 1 % 2% H Electronks n.a. 0% 1% 1 % 1% I% 1% 1% 1% 1% 1% 1% 1% % Cement n.a. 0% 0% 1 % 1 % 2% 4% 5% 4% 1 % 1 % 2% 4% :7 Clay e Cement Products n.a. 1 % 1 % 2% 2% 4% 8% 9% 8% 2% 2% 2% 2% Building n.a. 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 1 % 1 % Ceramics n.a. 0% 1 % 0% 0% 0% 0% 0% 0% 0% 0% 1 % 1 % Others n.a. 1 % 0% 2% 3% 3% 2% 2% 2% 2% 1 % 2% 3% Energy 0% 0% 0% 1% 1% 1% 1% 1% 1% 1% 0% 1% 1% Services 10% 51% 46% 39% 35% 36% 37% 32% 33% 33% 14% 31% 29% Source: Undersecretariat of Treasury and Foreign Trade. General Directorate of Incentives and Implementation Activity Report 1983, 1984, 1985,1986,1987, 1988, 1989,1990,1991,1992.  Table A-4: Breakdown of Investment Incentive Certificates According to Their Types, 1981-1992 (%) Investment Type 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 New investment 92% 84% 71% 66% 63% 64% 66% 65% 74% 87% 72% 52% Expansion 5% 97% 13% 16% 20% 17% 15% 13% 9% 5% 16% 23% Completion 1% 2% 3% 4% 3% 2% 3% 2% 2% 1% 3% 3% Renewals 1% 2% 4% 5% 3% 2% 2% 2% 1% 4% 1% 2% Quality Improvement 0% 1% 2% 2% 2% 2% 1 % 1% 0% 0% 0% 1% Elimination of Bottlenecks 1% 2% 3% 3% 2% 4% 3% 4% 2% 1% 1% 2% Modernization 0% 1% 2% 3% 5% 7% 9% 11% 10% 4% 4% 9% Integration of Facilities 0% 1% 2% 1% 2% 3% 1% 2% 2% 1% 1% 2% Removal 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 1% 0% Leasing 0% 0% 0% 0% 0% 0% 1% 0% 0% 1% 1% 6% Restoration 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% Source: Undersecretariat of Treasury and Foreign Trade. General Directorate oF Incentives and Implementation Activity Report 1983, 1984, 1985, 1986, 1987, 1988, 1989, 1990, 1991, 1992. .2~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~E 0E C'  Table A-5: Sectoral Breakdown of Export Incentive Certificates for Duty Free Imports, 19800%1992 (%) Sectors 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 Agdriculture 0% 0% 0% 0% 0% 0% 0% 0% 3% 10% 1% 1% 0% Mining 0% 0% 0% 0% 0% 0% 0% 0% 0% 1% 0% 0% 0% Manufacturing 100% 100% 100% 10% 100% 99% 99% 99% 97% 99% 99% 99% 99% 0 Food & Beverages 3% 4% 5% 6% 10% 10% 13% 12% 10% 6% 9% 12% 14% Textile & Clothing 44% 51% 66% 71% 70% 61% 59% 46% 58% 67% 69% 66% 59% Forestry Products 0% 0% 0% 1% 0% 0% 0% 1% 1% 1% 1% 0% 1% Paper 0% 0% 1% 1% 0% 0% 1% 1% 0% 1% 0% 0% 0% Leather & Leather Products 7% 7% 3% 4% 5% 3% 4% 7% 7% 5% 4% 3% 4% C Rubber 6% 5% 2% 1% 2% 1% 1% 1% 1% 1% 1% 1% 1% o Chemicals 5% 6% 7% 5% 3% 6% 9% 13% 6% 2% 3% 3% 3% Glassware 0% 1% 0% 0% 0% 0% 1% 1% 0% 1% 1 % 1% 1 % C 0 Iron & Steel 16% 8% 4% 1% 1% 7% 3% 2% 0% 1% 2% 2% 2% o 0 NonO%Ferrous Metals 0% 1% 3% 1% 1% 1% 1% 2% 0% 1% 1% 0% 1% Transport Vehicles 1% 1% 1% 1% 1% 1% 2% 2% 1% 1% 1% 2% 3% C) Metal Goods 3% 0% 0% 1% 1% 2% 1% 2% 3% 1% 0% 1% 2% o Measunng Devices 0% 0% 0% 0% 0% 0% 0% 1% 1% 1% 0% 0% 0% - Machinery 2% 2% 1% 1% 1% 1% 1% 1% 1% 2% 1% 1% 2% 5 Electrical Machinery 1% 6% 3% 2% 2% 2% 2% 2% 2% 1% 1% 1% 1% H Electronics 4% 3% 2% 1% 1% 2% 1% 1% 1% 1% 1% 2% 3% Cement 0% 0% 1% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% Clay & Cement Products 0% 1% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% Ceramica 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 1% Building 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% Others 6% 3% 0% 1% 1% 2% 2% 4% 3% 6% 6% 2% 1% Energy 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% Services 1% 1% 0% 0% 0% 0% 0% 0% 1% 0% 0% 0% 0% Source: Undersecretariat of Treasury and Foreign Trade. General Directorate of Incentives and Implementation Activity Report 1983, 1984, 1985, 1988, 1987, 1988, 1989, 1990. 1991, 1992.  Appendices 71 Table A-6: Sectoral Breakdown of Global andlor Project Based Export Incentive Certilcates, 19900%1992 (%) Sectors 1990 1991 1992 Agriculture 1% 2% 1% Mining 1% 0% 1% Manufacturing 98% 98% 98% Food & Beverages 5% 7% 8% Textile & Clothing 36% 36% 37% Forestry Products 1% 1% 1% Paper 1% 1% 1% Leather & Leather Products 11% 12% 10% Rubber 2% 2% 2% Chemicals 7% 7% 6% Glassware 1% 1% 1% Iron & Steel 6% 6% 10% NonO%Ferrous Metals 2% 2% 1% Transport Vehicles 3% 5% 5% Metal Goods 1% 2% 4% Measuring Devices 0% 0% 0% Machinery 2% 2% 3% Electrical Machinery 3% 5% 4% Electronics 3% 3% 3% Cement 0% 0% 1% Clay & Cement Products 1% 0% 1% Budiding 0% 0% 0% Ceramics 1% 1% 1% Others 11% 6% 4% Energy 0% 0% 0% Services 0% 0% 0% Source: Undersecretariat of Treasury and Foreign Trade. General Directorate of Incentives and Implementation Activity Report, 1990, 1991, 1992  Table A-7: Regional Breakdown of Investment Incentive Certificates, 1980-1992 (%) Regions 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 Marmara 42% 39% 48% 51% 45% 40% 39% 38% 37% 29% 14% 30% 48% Central Anatolia 17% 26% 16% 14% 12% 15% 16% 14% 14% 12% 6% 12% 13% Aegean 11% 6% 8% 10% 15% 16% 15% 17% 19% 15% 4% 10% 17% Mediterranean 12% 14% 13% 11% 10% 10% 10% 11% 11% 11% 5% 10% 9% Black Sea 11% 7% 6% 6% 8% 9% 10% 10% 9% 14% 12% 10% 7% East Anatolia 4% 3% 3% 3% 6% 5% 5% 4% 4% 6% 22% 11% 3% South East Anatolia 3% 5% 6% 5% 4% 5% 5% 4% 5% 12% 36% 15% 2% 0 Multiregion* - - - - - - - 2% 1 % 1 % 0% 2% 1 % 0 ^ Investment located in more than one region. 00 Source: Undersecretariat of Treasury and Foreign Trade. General Directorate of Incentives and Implementation Activity Report 1983, 1984, 1985, 1986,. 1987, 1988, 1989, 1990, 1991, 1992. 0 .,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~C ~1<  Table A-8: Sectoral Breakdown of Investment Incentive Certificates for Priority Development Regions, 1980-1992 (%) Sectors 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 Agriculture 27% 37% 16% 6% 3% 12% 15% 39% 59% 19% 1% Minining 4% 7% 9% 12% 9% 5% 7% 4% 1% 3% 9% Manufacturing 22% 30% 36% 36% 36% 47% 58% 59% .4.31 58% 68% Food & Beverages 14% 16% 10% 14% 10% 13% 11% 14% 13% 13% 14% Textile & Clothing 0% 5% 10% 4% 5% 7% 9% 7% 6% 17% 14% Forestry Products 1% 0% 1% 2% 1% 4% 5% 4% 2% 3% 5% Paper 0% 0% 1% 1% 1% 0% 1% 0% 0% 0% 0% Leather & Leather Products 0% 0% 2% 2% 1 % 1 % 1 % 1 % 1 % 3% 2% Rubber 0% 0% 0% 1% 1% 1% 3% 3% 2% 4% 4% Chemicals 1% 1% 1% 0% 1% 1% 1% 1% 1% 1% 1% Glassware 0% 0% 1% 0% 1% 0% 0% 0% 0% 0% 1% Iron & Steel 0% 0% 2% 1% 1% 2% 0% 0% 0% 1% 3% Non-Ferrous Metals 10% 0% 0% 0% 1% 1% 0% 0% 0% 1% 1% Transport Vehicles 1% 0% 0% 0% 1% 0% 1% 0% 0% 1% 4% Metal Goods 1% 1% 1% 1% 3% 1% 2% 1% 1% 4% 5% Measuring Devices 0% 0% 0% 0% 0% 0% 1% 0% 0% 0% 1% Machinery 1% 0% 1% 1% 1% 1% 1% 2% 1% 3% 1% Electrical Machinery 1% 0% 0% 0% 0% 0% 1% 1% 0% 0% 1% Electronics 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% Cement 0% 1% 1% 1% 4% 5% 4% 1% 0% 1% 2% Clay & Cement Products 2% 4% 3% 7% 19% 15% 16% 4% 2% 4% 3% Building 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 1% Ceramics 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 1% Others 0% 1% 2% 1 % 1 % 2% 2% 2% 1 % 2% 1 % Energey 0% 0% 1 % 1 % 0% 0% 1 % 0% 0% 1 % 1 % Services 47% 26% 37% 42% 41% 24% 18% 16% 8% 19% 22% Source: Undersecretariat of Treasury and Foreign Trade. General Directorate of Incentives and Implementation Activity Report 1983, 1984, 1985,1986, 1987, 1988, 1989, 1990, 1991, 1992. -e *T3  74 Institutional Influences on Economic Policy in Turkey  Appendix 2 Three Industries: An Overview In terms of both structure and performance the three industries we have chosen for examination, i.e., readywear clothing, autos and glass, display a great deal of variation. To start with concentration ratios, calculated as the top four firm's percentage share in terms of total industry sales revenue, readywear clothing is an industry which is becoming increasingly competitive as the concentration ratio declined from 23.5 in 1985 to 12.62 in 1986 and to 6.75 by 1989, the last year for which data is available. Tn contrast, both glass and autos are highly concentrated industries with the 1989 concentration ratios being 61.87 in the former and 53.17 in the latter.2' All of these industries are quite significant in terms of their contribution to employment and to total output in the manufacturing sector. Readywear clothing's contribution to employment far belies its modest contribution to manufacturing output. Regarding the latter, although its share in total output in manufacturing (calculated on the basis of 1990 constant prices) is increasing--from .5 percent in 1980, i.e. the beginning of liberalization, to 1.26 percent in 1985 and to 3.36 percent in 1990--it remains small. However, in 1985 the share of the readywear and textile industri3s in all manufacturing employment was fully 31 percent (Yasar, 1990 p. 72). And while figures for employment in readywear clothing alone are not available, an educated guess on the basis of our interviews suggests that at least half of this employment, which would amount to roughly 15 percent of all employment in the manufacturing sector, was created in the readywear clothing industry. The reason the exact employment figures are unavailable relates to the very high number of firms which do not register workers as a strategy to avoid paying social security taxes (Author's interview, 1994). But the fact remains that is a very dynamic industry which not only continues in the early 1990's to create new employment but is also characterized by a high level of new entrants and, to a lesser degree, exits. The auto and glass industries are also important in terms of employment generation and their contribution to manufacturing output. The auto industry is particularly important as its share in the 1990 total manufacturing output stood at 6.03 percent, while the corresponding share of glass was 1.26 percent. Overall, the 18 firms active in the auto industry employ 155 to 165 thousand workers, according to 1992 figures (OSD, 1994). Another 130 to 160 thousand are employed in related areas such as dealerships, spare parts and services. In the glass industry, on the other hand, only some 15 thousand workers are employed, more than two-thirds of whom 21 Concentration ratio statistics are from the State Institute of Statistics. Calculations are made according to the International Standard Classification with classification codes of 3222 for readywear clothing, 3620 for glass and 3843 for autos. It should be noted that the code for autos also includes the spare parts industry. Thus the given ratio actually understates the level of concentration in this industry. 75  76 Institutional Influences on Economic Policy in Turkey work for the dominant producer, SiseCam AS (SiseCam AS, Annual Reports). Hence this industry is the most capital-intensive of the three. The glass industry in Turkey manufactures a variety of products, such as flat glass, glassware and bottles. The SiseCam Group dominates more than 80 percent of the market in all major items and other companies operate only at the edge of the market. This market dominance is no accident as SiseCam AS has aggressively sought to defend its quasi-monopoly status. Thus, for example, in 1971, Anadolu Cam Sanayi AS, producing flat glass and bottles, attempted to operate outside of the SiseCam Group but as a result of a series of political maneuvers combined with price cutting measures Sisecam AS forced this budding rival to give up and eventually to be bought off the group.22 Today, with about 20 plants ranging from flat glass to soda ash products- -which is the major input in this industry--SiseCam AS can be considered a typical example of a vertically integrated, multi-divisional holding company. It is owned by one of the three largest conglomerates in the Turkey, the IsBank Group, which not only owns a number of other industrial firms but also the largest commercial bank in the country.23 As opposed to glass which is a monopolistic industry, autos is an oligopolistic one. There were about 229 thousand passenger vehicles sold in Turkey in 1991, 317 thousand in 1992 and 440 thousand in 1993. Of these, approximately one-sixth in 1991, one-fifth in 1992 and one- fourth in 1993 were imported (OSD, 1994). Koc Holding, the largest of the multi-sectoral conglomerates in Turkey, dominates the domestic production in passenger cars by supplying more than half of the cars sold. The manufacturing of autos is very important for the Koc Holding overall since the share of autos in its total corporate profits in 1993 stood at about 55 percent (Author's interview, 1994). The second largest player in passenger cars is OYAK-Renault, a joint venture between the Army Pension Fund and the French company, which controls about one-fifth of cars sold in the domestic market. But entry to the auto industry is not prohibited as recently Sabanci Holding entered into a partnership with Toyota to manufacture 100 thousand Corollas per year. Besides their industrial structure, the performance of these three industries has also been very different in the 1980's and into the early 1990's. Despite the recent liberalization measures all of these industries are still protected. According to a study by SPO the effective protection rates amounted to 44.56 and 31.89 for readywear clothing, 26.11 and 19.01 for glass and 38.65 and 33.77 for autos in 1993 and 1994, respectively (SPO, mimeo). Nonetheless, export performance and growth figures have been very impressive for readywear clothing. Exports in this industry, which amounted to less than U.S. $500 million in 1982, skyrocketed in subsequent years and exceeded $4 billion--almost a third of total Turkish export revenues--by 1992. In terms of the rate of increase in export performance between 1980 and 1991, Turkey outperformed other developing countries and became the world leader, Indonesia being second and Thailand being third (Duruiz and Yenturk, 1994). The bulk of these exports are to the European Community. As in readywear clothing, both the glass and auto industries have amply benefitted from various incentives in the last decade. Throughout the second half of the 1980's and into the 22 Information from Author's interviews (1994). An official history of the Sisecam group has been written by Prof. Zafer Toprak but has not yet been released to the public. 23 IsBank is a semi-public institution with an independent management team. Its shares are distributed as follows: IsBank Pension Fund, 37 percent; Treasury, 41 percent; and individual and corporate shareholders, 22 percent.  Appendices 77 1990's, the ratio of exports to total sales in the glass industry has been around 22 percent and more than 200,000 tons of flat glass are exported each year to Middle East, Gulf and North African countries. Overall, Sisecam AS in 1993 earned U.S. $355 million in export revenue and this company's share in total world production in glass is an impressive 1.3 percent (Sisecam AS, 1992 Annual Report). The auto industry, on the other hand, has been in a deficit situation in terms of its trade balance throughout the 1980's and into the early 1990's. Due to the poor quality and high costs, no passenger cars are able to be exported (other than those which have been sent to CM[EA markets via barter arrangements) while imports of spare parts and passenger vehicles amounted to U.S. $41 billion in 1991 (Third Izmir Economy Congress, 1991). Hence, although the readywear producers in particular and to a lesser extent SiseCam AS have adapted themselves to the requisites of an outward-oriented economy, auto producers have been lagging far behind.  78 Institutional Influences on Econonic Policy in Turkey  Appendix 3 Field Research Methods The original field research upon which this report is based was conducted by the authors in June and July, 1994, in Ankara and Istanbul, Turkey. As referenced in the text when appropriate, considerable secondary document collection was accomplished which added considerably to our factual understanding of the Turkish incentive regime. Moreover, during this period extensive, in- depth and confidential interviews were conducted with major figures in the Turkish business community as well as with government officials in Treasury and in the State Planning Organization. Finally, a semi-structured questionnaire was administered to business members of the three industries of interest and to government officials involved in implementing and monitoring the incentive regime. With respect to interviews in the business community, on one hand, high-level managers in leading businesses were interviewed, with a particular focus on managers in the three industries of interest--glass, autos and readywear clothing. On the other hand, high-level representatives of industry-level and other "umbrella" (e.g., TUSIAD) business collective organizations were interviewed as well. To the extent possible, we followed a combined "positional" and "reputational" method for determining who to interview. While some individuals were interviewed solely on the basis of their holding an important position, such as the President of an industry-association for example, we also identified important business persons by asking business people we interviewed who they believed were "the most important and influential members of the business community" both for the business community in general as well as for their particular industry. With respect to the questionnaire, for the business community we adopted different methods for disbursing it for the three different industries. In the glass industry there is only one major firm--SiseCam AS--and hence our method was to distribute the questionnaire directly to high- level managers. Thus, while only 8 questionnaires were completed, this constitutes a quite representative sample of the decision-makers in the industry. In the auto industry, in contrast to glass, there are a small number of major firms. In this industry our method was to request of those high-level managers we interviewed in person to both complete the questionnaire themselves as well as to distribute it to other managers in their firm. Our 19 responses ended up being most representative of the largest player in the auto industry, the KOC Holding, as well as its major rival, the Sabanci Holding. For the readywear clothing industry the market situation is completely different as this industry is highly competitive with numerous small, medium and large firms. Our method for distributing our questionnaire was to seek the cooperation of the industry- association, the GSD, in distributing it to their members. Due principally to our own time constraints, the relatively small sample of 18 returned questionnaires must be viewed to be less representative than the samples for either glass or autos. We can report, however, the readywear clothing sample is most r -esentative of the larger and more powerful industry members. 79  80 Institutional Influences on Economic Policy in Turkey We interviewed in person a number of government officials from Treasury and SPO who also filled out our questionnaire. Obviously, we cannot be specific as to whom we interviewed. However, in terms of their formal positions we can say that the officials we interviewed in Treasury, the home at the time for the incentive regime, represented all of the major offices involved with implementation and monitoring of both investment and export incentives.