Colombia - Social security review

13 Leport No. 6540-CO Colombia: Social Security Review September r 3, 19e7 Humani Resources Division Country Department Il! Ldtrn Ainerica and the (aribbean Regional Offk t' FOR OFFICIAL USE ONLY Document of the World Bank This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise he disclosed without World Bank authorization.  FOR OMCIAL USE ONLY ACRONYMS USED ATEP Occupational 1,isk Program BCII Central Morigage Bank CAJANAL Fund for Public Sector Emp]oyees CAPRECOM Fundl for Communication Workert.- CAPRESUB Banking Superintendency Soc-ial Insurazice Fuund CORPOANONIMAS uiund for lIncorporated Companip es CPI Cotnsumer Price Index DANE Nat ionra.l Deparlitnmlr. of Statist.:ics DNP National Department of Planninlg EAP Economically Active Population EGM Health-Maternity F'rograni FNII National Hospital Fund FONADE National Development Fund ICBF Colombian Inst,itute of Child WetTare IFI Institute of Indus'rial Developmvent ILO Internat.ional Lab)or Organizalion INC National Cancer Institute INS National Ilistitute of' Health INSFOPAL Institute for Municipal Development ISS Institute of' Social Insurance lvM Pension Prog-ram MLSS Ministry or' Labor and Social Secur.ity MOH Ministry of Health NtIS National Healt.h System PHC Primary Heal th Care PREALC Employmenit Program for Latin America and thfe Caribbean SENA National Training Service TAN National Savings Certificates This document has a restricted distribution and may be used by recipients only in the performance of their officil duties. Its contents may not otherwise be disclosed without World Bank authorization.  COLOMBIA: SOCIAL SECURITY REVIEW Table of Contents Executive Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . i I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 II. Historical Evolution and Current Organization of Social Security . 1 A. History .I.. . . . . . . . . . . . . . . . . . . . . . . . I B. Organization . . . . . . . . . . . . . . . . . . . . . . . . . 2 III. Sources of Social Security Finacin . . . . . . . . . . . . . . 5 A. Private Sector Contributions . . . . . . . . . . . . . . . . . 5 M. Public Sector Contributions . . . . . . . . . . . . . . . . . 6 C. Revenue Distribution by Source . . . . . . . . . . . . . . . . 7 D. Evasion and Payment Delays . . . . . . . . . . . . . . . . . . 7 E. Impact of Contributions on Income Distribution and Employment 8 F. Financing Issues . . . . . . . . . . . . . . . . . . . . . . . 9 IV. Uses of Social Security Funds . . . . . . . . . . . . . . . . . . 9 A. Benefits to Insured . . . . . . . . . . . . . . . . . . . . . 9 B. Administration . . . . . . . . . . . . . . . . . . . . . . . . 20 C. Investment . . . . . . . . . . . . . . . . . . . . . . . . . . 22 D. Issues Relating to Uses of Social Security Funds . . . . . . . 25 V. Social Security-s Financial Situation . . . . . . . . . . . . . . . 25 A. Financial Equilibrium . . . . . . . . . . . . . . . . . . . . 25 B. Actuarial Equilibrium . . . . . . . . . . . . . . . . . . . . 28 C. The Cost of Social Security and of its Expansion . . . . . . . 30 D. Financial and Actuarial Equilibria Issues . . . . . . . . . . 31 VI. Summary of Issues . . . . . . . . . . . . . . . . . . . . . . . . 31 VII. Recommendations. . . . . . . . . . . . . . . . . . . . 33 A. Changes Requiring Administrative Action . . . . . . . . . . . 33 B. Changes Needing Legislative Action . . . . . . . . . . . . . . 37 This report was prepared by W. De Geyndt (PHND3) and C. Mesa-Lago (Consultant).  COLOMBIA: SOCIAI. SECURITY REVIEW Table of Contents PagCe No. Annex I Sources Bibliographical References . . . . . . . . . . . . . . . . . 39 Statistical Data Prepared on Request . . . . . . . . . . . . 43 Annex 2 Statistical Tables .... . . . . . . . . . . . . . . . . . 44  LIST OF TABLES Tuble No. Summary of Evolution of Social Security Legislation by Risk 1 and Group Covered in Colombia: 1843-1986 Organizational Chart of Social Security in Colombia: 1986 2 Legal Contributions to Social Security by Program, Source 3 and Fund in Colombia: 1986 IPercentage Distribution of All Social Security System by 4 Revenue by Source In Colombia: 1965-1980 Percentage Distributionl of Revenue in Major Social Insurance 5 Funds by Source in Colombia: 1984 Percentage Distribution of ISS Active Tnsured and Covered 6 Enterprises by Enterprise Size in Colombia: 1983 Social lnsurance Coverage of the Population in Colombia by 7 All Funds: 1970-1985 Plercentage Distribution of Total Insured by Social Insurance Funds 8 in Colombia: 1953 Social Insurance Coverage oi the Population in Colombia by ISS and 9 Cajanal: 1964-1985 Fha,ily Allowance Coverage of the Population in Colombia: 10 1980-1984 Degree or Social Insurance Coverage by Type of Worker in Colombia: 11 1984 1)egree of Social Insurance (ISS) Coverage in Hlealth Care by 12 Population age--group in Colombia: 1984 Degree of Social Tnsurance Coverage by Economic Activity in 13 Colombia: 1973, 1978, 1980 Degree of' Social Insurance Coverage by Depait.ments in Colombia: 14 1984 Selected Differences in Benefit Entitlemeni Among Social 15 Insurance Funds/Occupational 'Groups in Colombia: 1986 Percentage Distribution of All Socia.' Security System Expenditures 16 by Main Item in Colombia:: 1965--1980 Estimated Plercentage Distribution of IS' Benefit Expenditure by 17 Program in Colombia: 1975-1985  Real Value of ISS Annual Pensions and Relation with Minimum Wage in 18 Colombia: 1970-1985 ISS Health Facilities and Efficiency: 1975-1984 19 ISS Health Efficiency by Departments in Colombia: 1982 20 Differen es Among Annual Average Pensions by Insured Groups in 21 Colombia: 1978 and 1982 Employment and Labor Costs in the ISS, Co'ombia: 1970-1985 22 Percentage Distribution of Annual Total Expenditures by Current 23 and Capital (Investment) Expenditures in ISS, Colombia: 1975-1985 Estimate of Real Value of ISS (IVM) Cumulative Financial 24 Investment and Yields: 1976-1984 ISS Physical Investment in Colombia: 1981-1987 25 Balance o2 ISS Income and Expenditures by Program in Colombia: 26 1975-1985 Actuarially Established and Effectively in Force Percentage 27 Contributions Over Wages for Pension Program in ISS, Colombia: 1967-1985 Balance of Income and Expendituies in Public Social Insurance Funds 28 in Colombia: 1984 Comparison of Quotient of Demographic Burden Among Social Security 29 Funds in Colombia: 1984 Percentage Distribution of the Population by Age Groups in Colombia:. 30 1970--2025 Secial Security Expenditures .. 31  COLOMBIAt SOCIAL SECURITY REVIEW Executive SummaUy 1. Colombia's social security systeam is one of the most fragmented and stratified in Latin America, has one of the lowest population coverages by social insurance, and is one of the most expensive anrd heavily subsidized by the state. Colombia lags in coverage level achieved by Latin Americar countries in a similar stage of ecoPomic development. Social insurance benefits reach "bout one sixth of the Colombianis. Among the 4.6 million people covered a subset of about 400,000 public sector employees enjoys exceptional benefits and entitlement conditions and contributes only :3% to 14% to the cost of their benefits with the state paying from 69% to 88%, Expans.ion of coverage should be a primary target but it would not be feasible with the current high costs per benefitiary and low efficiency levels. The consernsus in Colombia on the need to reform the social security system drastically is not. matched by a widespread agreement on how to solve the issues and on which policir- to formulate and imple.ment. The historical experience in Colombia and in otlher Latin American countries indicates that a strong political commitment is neede-d to cope with difficult social security problems. 2. This report focuses on The privatte and public sector social insurance funds which are based on employment and financed through employer/employee contributions and central government transfers. About 2.6% of GDP was spent in 1984 for health care and income maintenance by the two largest social insurance funds which cover 12. ol the population, or US$ 287 per insured. Thus, over one third of all the money spent in Colombia by public and private sources for health care and income maintenance was spent on one etight.h of the population. This is one of the highest expenditure ratios per insured in Latin America. With the present social insurance model and the current rate of expenditures, social insurance coverage of the entire population in Colombia would absorb 23% of GDP. Soc.'al Security issues 3. Low Coveragy. In 1985 only 16% of the toLal population had social insurance coverage ranking Colombia 13thi among all Latin American countries. Practically all the insured are salaried1 workers: 70% are coveredl by ISS (private sector), 7% by CATANAL (national civil service sector), and the remaining 23% by some 300 smallfer instittutions. 4. Inequal ities in ConLtributiotn Rates. The combined insured/employer wage contribution fot the largest social insurance fund (ISS for private sector employees) ranges from 150% to 20%, plus 4% for family allowances. T'he total average percenitage is the seventh highest in Latin America, while Colombia ranks 13th in population coverage. Contri but.ions range frorit 5%O to 20% among various other social insurance funrds. The lowest tates ale in the public sector, where the insured do not contrribute lo the pension or the health programs which are government finaneced. 5. Burden on Employers and Pavmient Evasion. The ptercentage contribution paid by the insured in Colombia is thie smallest in Latin America and the Caribbean, while the employer and the state finance from 59% t.o 88% of social  -- ii - security. Employer evasion of contributions and delays in remittance of collected monies amounted to about US$ 305 million in 1985. 6. Generous Benefits. Benefits for members belonging to social insurance funds, especially the public sector ones, are more generous and qualifying conditions are easier than in most Latin American coantries. The most influential groups in Colombian society (the judiciary, selected civil servants, congressinen, the armed forces, and strategic trades or professionals) have been able to obtain privileged social security ec.nditions and significant state subsidies. 7. Tneqqulities in Ben fit.s. Important, inequalities in pension benefits exist among funds. In 1982 the lowest average pension was paid in ISS while pensions were 66% higher in CAJANAL, 151% higher in the military fund, and 245% higher in CAPRESUB. The retirement age in the public sector fund is ten years less tharn in ISS and in some public funds the insured can retire with only 10 or 20 years of service regardless of age. The best health facilities are available to the military and the police and their dependents are fully covered. Health benefits for dependents of ISS insured are limited to maternity care and to care for chiildren under tbe age of' one. T-his is also the case in CAJANAL except that the eligibility age of children is reduced to six months. 8. Declinin_ Inve.stment and Incteasina Government Subsidies. Social insurance investment by ISS reached US$ 819 million in 1985. The percentage of [SS annual revenues allocaied to investment steadily declined from 27% in 1975 to G% in 1985, a decrease of two-thirds in constant pesos. ISS can only invest in state instruments adjusted annually for inflation. 9. High Administrative Cost. The overali cost of managing the Colombian socia.l security system stead-ily increased from 4.4% to 12.4% of total expend- itures in 19t65-1980. Out of 18 Latin American and Caribbean countries for which these data were available in 1980, Colombia r-anked among the five with highest administrative costs. The major administrative expense item is personnel. The ratio of' ]SS emplovet.s per 1,000 insured rose from 9.8 to 10.9 in 1970--1975 but, following thie reform of the mid-1970s, steadily declined thereafter and reached 7.4 in 1985. The ratio employees/insured in CAJANAL was 10.6 per 1,000 in 1985, similar to the one 1Sv' had ten years ago. 10. Deficient. Management. Systems. Weaknesses in mantagement systems, in addition to excessive staffing, are: lack of continuity in top management positions; no proper information system capable of providirig accurate, timely and useful data for planning andi control decisions; absence of individual accounts; deficiencies in the registration process, long (lelays in processing pension benefits; poor- inspection anid ineffectual legal enforcement permitting evasion and payment delays; lack of integration of the ISS accounting and budget divisions with thie ISS treasury division; poor and complex accounting techniques impeding an accurate estimation of administrative costs and medical care unit costs; and cumbersome triple external auditing. 11. Financial Imbalance. The total deficit of a.l1 the public sector funds for which information was available in 19834 was 9.2 billion pesos (US$ 91  - iii - million). The public sector fund deficit was offset by a small ISS surplus and by a substantial family allowances surplus. The public sector funds is the most troublesome financial component of the system especially CAJANAL which has the worst financial imbalance of all funds. The state subsidy to the system has gradually increased and, if the ptesent trend continues, such a burden will soon become initolersable. 12. Actuarial Imbalance. ISS has not. conducted formal actuarial reviews at least since 1980 and the IVM ald ATEP programs have been decapitalized due to transfers to JGM. The increase of the IVM contribution in 1985 fr.i 4.5% to 6.5% is only a temporary solution since in the near future the investment yield and part of the reserves will lwave to be used to meet expenses. The ATEF actuarial deficit in 1981 was estimated at 44% of the needed reserves. The EGM program uses the pay-as-vou-go system and it would show a deficit without the subsidies from other TSS programs. CAJANAL's pension program lacks an actuarial base and a financing method. Recommendations 13. There is a briad conisensus in Colombia hEat. Lhe social security system needs major reforms to make it more effective, efficient and equitable. The suggested recommendations are structured along the lines of the four major topic areas of the report and are divided into chaniges that require administrative action arid changes that. need legislative action. This differentiation is also indicative of the time horizon for implementing the proposed chaniges. Changes Requiring Administrative Action 14. Prganization. ta) The compilatiorn o-f all social security legislation should be undertalken at once with a view towards preparation of a wuiform legal code and the submission of proposals to Congress for legislat ive action. (b) Cooperation between MOH and ISS in health services delivery has improved in the last two or three years. Management, decisions with time-specific objectives and monitoring mechanisms are needed to aCcelerate the process of harmonizing MOH and 1SS policies, plans, programs an(d activities. 15. Financig. All of the lol lowing measures woul I enhance revenues: (a) use the entire compensation as a base for conitributions; (b) i ncrease the wage conIribuition cei I ing; tel increase the premium to be paid by the insured especially in public sector funds: (d) ad,just the professional risk classification and chargIe the apptopriate ATEP premium; te) aggressively pursue payment. evaders and( late payors: and (f) introduce copayments anid deductibles lor curative health services.  - iv - Some revenue enhancement strategies canl be implemented by administrative actions but changes in contributions are the domatin of the legislature. 16. Uses of Funds. (a) Top priority shouldl be givezi to accelerate the expansion of the ISS family health program to cover the insured's spouse and children below 18 years of age. This model shoulcd also be applied to CAJANAL and other public sector funds which now lack such services. (b) Changes in benefits and -ntitlement conditions for benefits may require legislative actiotn. Unt l uniformity is achieved in pensions, a meChanism should be jn! uiced to allow: the accumulation of time of service and contributio. ,nder various funds; the establishment of individual accounts: the ability to purchase supplementary insurance (private, publir or mixed) by income groups who have the capacity to pay for better benefits. (e) Management weaknesses should be corrected: ISS and CAJANAI. should continue to reduce administrative expenditures and personnel; the ISS accounting system must be reformed to estimate the administrative costs of IVM and ATEP properly and to standardize accounting procedures in the budget, accounting and treasury divisions; and an information system should be implemented to increase the quantity, quality and utility of statistical data. (d) ISS health services at the local level should be improved to reduce congestion at the tertiary level. A comparative study of unit costs in ISS' own and contracted services should be made in order to decide which are more efficient. The introduction of user fees to contain the use of &ambulatory visits, the prescription of drugs and other services should be given careful consideration. (e) Further research is needed to estimate the real value of ISS investment and its yield, as well as viable alternatives to the "constant value bonds". (g) The value of ISS investment, in physical plant shouldl be adjusted for inflation; execution of physical investment plans should be accelerated to avoid the shift of investment fmnds io recurrent. cost budgets; and maintenansce, remodelitng and reequipment shouJd have priority over construetion of new facilities. 17. Financial anid Actuarial Eauilibria (a) Within .SS, ATEP transfe:s to EGM should be eliminated, and adequate IVM/ATEP cost sharinig of administrative expeniditures, physical investment and debt service introduced. (b) Actuarial studies should be carried out inmediately for all three ISS programs and for CAJANAL.  v Changes Needing Legislative Action 18. Organi_.ation. (a) I.egislation should be enacted prohibiting the establishment of new social security funds and the formulation of special regulations or exceptions to established funds. (b) The legal basis should be set for the gradual unification of all public sector funds un0er CAJANAL. 19. Financing. (a) Contributionj rates among public funds and between public funds and ISS should be standardized. The total ISS percentage is relatively high and so is the employer's share compared to the insurtd's contribution. The IVM contribution should be increased to 7.5% as recommended by a 1982 ISS study; the EGM contribution should be set at the level of the family health program to facilitate expansion; and the ATEP contribution should be adjusted to risk. (b) As part of the standardization process, CAJANAL, CAPRESUB, and CAPRECOM workers should starL contributing for pensions, and armed forces members for health benefits, in order to reduce the state subsidies to these programs substanLially and free these resources for expansion of coverage to the lowest income groups. 20. Uses of Funds. (a) Entitlement conditions for beniefits should be standardized in the public sector. tb) Specific recomumend1ations for uniformity include: (i) elimination of all seniority pensions; (ii) raising retirement age to at least 60 years for both sexes; (iii) normalization of the wage base and the percentage to calculate pensions; (iv) introduction of a national maximum for pensions; (v) equalization in rates applied to wages to calculate common sickness pay leave: and ivi) elimination of personal and mortgage Loan programs. (c) Flexible alternatives to ISS inveslment in "constant value bonds" should be studied such as placing funds in the capital market or in the banking sysLtem or transfer investment admixiistration to an independent agency in order to reduce the state subsidy and possibly increase the reserves. The latter alternat:ive woul(d require easing investment constraints as previous attempts to use a private company were not successful.  - vi - 21. Financial and Actuarial Equilibria. (a) Insured and employer's contributions to TVM, ATEP, and EGM should be reviewed. (b) A Pension Fund should be created in CAJANAI. with the following featyres: contributions by insured, effective employer contributions, continued state support for' the time necessary to develop the needed reserves, and a stricter package of benefits and entitlement conditions. Once the Fund is solidly in place, it should gradually incorporate other national and departmental funds in the public sector.  COLOMBIA: SOCIAL SECURITY REVIEW 1. INTRODUCTION 1. Deefinitions. The ILO defines social security as all statutory programs which provide income maintenance and access to medical care, including private schemes, and which may or may not be related to employment. Social insurance is a technique to achieve the objective or the concept of social security. These definitions are not applied in their strictest sense in the present review. The activities of the Ministry of Health (MOH) and of the private medical care sector are excluded. MOH policies and programs were analyzed in detail in the Health Sector Review (Report No. 4141-0O) and summarized in the Staff Appraisal Report (Report No. 5532-CO) of Loan 2611-CO. Almost no data are available on the private medical care and medical insurance sector to carry out. the most minimal analysis. Therefore, this report focuses on the private and public sector social insurance funds which are based on employment. and are mainly financed through contributions (i.e. the Bismarckian model) and the term social security will be used to refer to the totality of these social insurance funds. 2. Structure of the Review. The four main chapters of the report cover: (a) the history and the present organization of Colombia's social security system; (b) the sources of funds; (c) the uses of funds; and (d) a diagnosis of its financial situation. Each chapter concludes with an identification of the issues which are then grouiped in a summary issues chapter. The final chapter suggests a set of recommendations which address the stated issues. II. HISTORICAL EVOLUTION AND CURRENT ORGANIZATION 0O" SOCIAL SECURITY A. History 3. As is typical in many countries in Latin America, social insurance evolved in Colombia in a gradual and fragmented manner, resulting in a multiplicity of institutions and a system stratified along occupational lines. Following a common regional pattern, the most powerful occupational groups in Colombia were the first to obtain coverage and receive more generous benefits while the least influential groups were the last to be covered (or still remain unprotected) and receive the most restricted package of benefits (19, 21, 55). 4. Table I summarizes the historical evolut.ion of the system. At the beginning of the 20th century oDly military officers, Supreme Court judges, selected civil servants and public teachers and/or their dependents were eligible for some pensions. The Labor Code introduced the principle of employer's responsibility for occupational risks affecting wage earners in industry and commerce while female wage earners in these two sectors were granted some protection during pregnancy and childbirth (1938); however,  -2- these rights were not realized until later. In 1925 the military and police, as well as their dependent families, became covered by a comprehensive package of pensions, health-maternity care and protection against occupational risks; a similar package was granted in 1943 -- 45 to Ministry of Communication employees (the CAPRECOM fund). 5. Two major groups of salaried workers began to be covered in the 1940s: (a) civil servants in 1946; the legislation on government employees paved the way for the creation of their fund (CAJANAL) with a national mandate but it also allowed the establishment of separate funds for civil servants of DepartAents and Municipalities thereby providing a legislative base for the current large number of funds; and (b) private sector white and blue-collar employees through the Institute of Social Insurance (ISS). Protection to private sector employees was provided gradually: it began with health-maternity care followed by the incorporation of occupational risks. Pension rights were added only in 1967 and therefore the maturation process is taking place now. In the 1970s the self-employed were given the option to affiliate with ISS on a voluntary basis. 6. The legislation providing protection to government employees spawned a large number of social insurance funds: Banking Superintendency - CAPRESUB (1946), policemen (1955), civil aviation (1961), congressmen - special regulations (1962), the attorney general and judiciary employees (1971), national comptroller employees (1976), civil servants in the armed forces (1977), and congressmen - independent fund (1985). Parallel to the creation of these national funds (or special regvlations within an established fund), numerous programs were established on behalf of employees of Department and Municipality administrations. 7. The family allowance program was introduced in 1957 for private sector wage earners and extended to public sector employees in 1963. A 1982 law extended coverage to all salaried employees,. Family allowance in the Colombian context means a fund which pays 60% of its benefits in cash to members who earn less than four minimum wages and 20% for health, education and recreation activities to all members. B. Organization 8. The social security "system" in Colombia consists of three benefit programs: (a) health services; (b) disability, old age and death payments (IVM) referred to as pensions; and (c) family allowances. There is no effective coordination among the close to 300 institutions which provide some form of social protection. Table 2 illustrates the organizational complexity of the "system". Pensions 9. Pensions is the least coordinated program. In otf'er LaTin American countries the establishment of a national pension institution precluded the creation of additional pension funds. In Colombia the  3 legislation creating the CAJANAI. f'urnd for civil servants actually encouraged the proliferation of funds in the publici sect.or, part'iculsariv at the departmental/municipal levels. sonie of these funids have as few as 30 insured anldCE are clearly not v i ab I e l h ('I T ure1nt. Ilegislation does not impede the establishnient of new national funids either and, at rtecently as 1985, congressmen separated from CAJANA]. (where they al ready enjoyed special regulations) managed to establish an independen I funLd. The situation in the private sector is considerabl y belt.ter since there are basi(allyv two funds: ISS and Civil Aviation. Hc'-ever, some sertors operate independently from ISS and have been ordered to affiliate with I1.S. Some sectors (e. o.i, oil) have not complied with the order (4(;). 10. There are at least 1945 pens jion funds in Colombia: 29 at the national level; 32 in Departrment.s arid 13J in Municipalities (a 1986 study, however, raises the latter fi gure to 170) (33) . Thiese data refer to funds registered at and providing information to the Ministry of Labor and Social Security (MLSS). Starting in 1975 and as recent ly as 1981 [lie MLSS conducted a survev of all these fun)ds and maintains fairly updated information on their principal legal F'eatullres and financial situation (62 - 65). However, not all fundis arc registered or provide information. Some funds 1 ike Civil Aviat ion refuse t o send infolmat ion. A l986 decree has made mandatory the registration of all pension funds with MLSS. 1I. There is ilo c-mpitat ion of' all the sociaL secUrity legislation in Colombia (the last was done in 1975 and was limited to 1SS) and lhe normative diversity results in a legal . lalvrinth. Some of ihe pension programs have an independent fund with its own sources of revenue, entitlement conditions and benefits. Other pyrograms do not. have funds blut. contract with and contribute to 18S or CAJANAL which provide the benefits. Still other programs' pensions are )paid thbrougih the state budget. Withinl CAJANAL there are about 70 special regulationis ior special groups. An illustrative example of' the existingr multiplicitv is banking where, in addition to CAPRESUB, thete are separate funds/programs for five public banks, while prival.e banks are coveted by 1SS. Departrments have often modified pension cegulationis although a 198(' decree prohibited this practice and rat.ified that, Congress is the on ly body autfior iz.ed to Iegislate on the subject. Some 77 municipal prograims do not have any regulations at all anid many simply do not function in pracitic. In 1968, a decree ordered tihe creation of a national commissior; which, in a period of one year, was expected to elaborate a .ltudy of alI pension lunds wilth (fhe g8oal or designing a policy of lepal uniformitv and eventual integration of aLl funds into ISS. The commission was never appointed (5aJ). A bill was prepared in 1986 to unify pensions in the public sector but did not make it through the legislative process t9). Health Services 12. Colombia's health sector- is divided into three major parts: (a) the public health subsector; (b) the social insurance subsector, including ISS for private employees, CAJANAL ror public employees, and some funds targeted for specific areas and groups such as the military, the police, and  - 4 - the oil industry; and (c) the private sector. The public health subsector is the largest health services provider and is legally responsible for about 75% of the Colombian population. The social insurance and the private subsecturs serve respectively 16% and 9% of the population. Only health services provided by the social insurance subsector are discussed here. Among all social insurance funds only ISS and CAJANAL have hospitals although some funds may have ambulatory care facilities. ISS owns 37 hospitals with 4,200 beds but CAJANAI, only has two small hospitals. Some funds contract for services with ISS and most social insurance funds, including ISS, contract with the public health and the private subsectors. Health services for the military and the police are not administered by their social insurance funds but through their own hospitals funded by the central government budget. 13. The 1969-1972 National Development Plan recommended the coordination and/or integration of all health services delivery (16). The National Health System was established in 1975 and one of ius explicit objectives was for MOH to coordinate all health service delivery. The Superintendency of Health lnsurance was created in 1979 to foster- coordination. The Development Plan "Change with Equity" (1983-1986) gave "maximum priority" to the integration of social security and the National Health System and to that effect a Technical Commission of Coordination and Integration was established in 1984 witb representatives from MOII, ISS, CAJANAL and other institutions. Thus for almost 20 years has there been an awareness and a desire to coordinate the delivery of health services, but it is only in the last two or three years that cooperation between MOH and 1ss has improved. There are now several examples of cooperation: decisions by ISS not to construct a hospital but to purchase inpatient services from MOH, joint health promotion campaigns, joint preventive care actions, expansion of ISS family health program in Uraba (Antioquia) through investment in a new MOH hospital. Family Allowances 14. The family allowance program is the least complex, but there are still more than 70 funds in Colombia with different regulations and benefits. Contributions are 4% of the payroll paid by the employer and are legally uniform. Many institutions provide family allowances directly to their employees, thus, increasing existing multiplicity and diversity. The Superintendenc; of Family Allowances supervises all funds, gathers information and is empowered to intervene and even eliminate those funds that do not abide by the law or practice irregularities. Organizational Issues 15. Colombia has remained isolated from the process of social security unification and legal uniformity that took place in the 1960s and 1970s in several pioneer countries in Latin America (e.g., Argentina, Brazil, Cuba, Peru, Uruguay). Its two major organizational issues are: (a) the multiplicity and diversity of public sector pension programs with the resultant urgent need to push the process of pension integration/uniformity  - 5 - vigorously; and (b) the need to accelerate the process of cooperation between MOH and ISS in the delivery of health services. III. SOURCES OF SOCIAL SECUR_IY FINANCING A. Private Sector Contributions 16. Social Security Institute (ISS) Health Services. Legal wage contributions for health and maternity benefits (EMG) are 7.0% of which the employer pays two thirds (Table 3). This percentage increases to 12% ir areas of the country where the family health program has been initiated which covers dependents of the insured. The employer pays an average fixed contribution of 1.5% for work accidents and occupational diseases (ATEP). The premium for ATEP should vary according to risk but the classification of risks has not been updated and is not in force. 17. ISS Pensions. The ISS total contribution for pensions (IVM) was initially (1967) fixed at 6%: 1.5% from the insured, 3% from the employer and 1.5% from the state. The latter never contributed and its share was legally abolished in 1971, hence the total legal contribution was reduced to 4.5% and remained unchanged until 1985 when it was increased to 6.5% with one third to be paid by the insured and two-thirds by the employer. 18. ISS Family Allowances. The percentage contribution of 4% for family allowances is standardized and entirely paid by the employer. It is calculated on the basic wage without any ceiling. 19. Base for Contributions. Legally the ISS contributions should be estimated over the total salary. In practice only the base salary excluding all fringe benefits is used, with the exception of a few large enterprises. A 1983 attempt to enlarge the tax base to include the entire salary was unsuccessful. Due to this practice, about 40% of all the insured fell in the one-minimum-wage category in 1981 and 75% did not exceed the two- minimum-wage category. Using the full salary as a base for contributions would greatly increase ISS revenue. Another source for increasing ISS revenue is a change in the wage contribution ceiling. Although the wage ceiling has been rising gradually it still favors the highest paid insured. No study has been undertaken to estimate how much additional income could be generated by the ISS by using the total salary as a base for contributions and by eliminating the wage ceiling. 20. Comparison with Other Countries. The combined overall ISS contribution, including family allowance, is either 19% or 24% depending on whether the ISS insured is in the family health program or not. An average contribution of 22% is used for a comparison with the other 19 countries of Latin America (55) but such an average excludes other contributions, e.g., 2% for ICBF (National Institute for Child Welfare) and 2% for SENA (National Training Service). Colombia has the seventh highest percentage contribution in Latin America but ranks 13th in total population coverage. Five of the  - 6 - six countries with higher percentage contribution than Colombia's have universal coverage of social security or are close to it (Argentina, Brazil, Chile, Costa Rica, and Uruguay') and the remaining country (Bolivia) has a population coverage almost twice as high as Colombia. Furthermore, three other countries have a lower percentage contribution than Colombia but a population coverage three to seven times higher (Cuba, Panama, and Venezuela). B. Public Sector Contributions 21. Health Services. Percentage wage contributions vary widely among public sector funds (Table 3). In CAJANAI. the total is 10% of which the insured pays half and the employer (which is the government) pays half. In the case of CAPRESUB (Banking) and CAPRECOM (Communications) the insured pays 5% and the employer/state pays the difference between the program's cost and the employee contribution. The entire cost of the health program is paid by the state for the military and police funds. 22. Pensions. Contributions to pension programs also show wide variation among funds. The active insured contributes only in the case of the military (8%) and the police (5%). The state/employer's percentage wage contribution is only fixed in the case of CAJANAL (3%). 23. State Subsidies. The total percentage wage contribution in the p,blic sector funds is considerably smaller than in ISS because all these funds receive the highest contribution from the state and, if this could be quantified in a standardized manner, the total percentage contribution in the public sector might be higher than in the private sector. The state contribution goes far beyond what the employer's contribution in the private sector is, hence it involves a subsidy. The state contribution/subsidy is always paid to CAPRESUB, CAPRECOM and the military and police funds. However in the case of CAJANAL, the legally ztipulated 8% for health services and pensions is not always fully paid >v the state-employer. The insured's total percentage cont.ribution in the public sector funds (except for the military) is smaller than that of the ISS insuu-d in the family health program. 24. Family Allowances. As in the case of the private sector, the percentage contribution for family allowances is standardized anal entirely paid by the employer, on the basic wage without any ceiling. 1 All these countries have pensioni programs which are about 20 years older than Colombia's (except Costa Rica) and life expectancies from six to eight years higher (except Brazil) hence their programs are more mature and expensive, all of which together with high population coverage justify a higher percentage contribution.  7- C. Revenue Distribution by Source 25. The percentage distribution of al] social--security system aetual revenue by source in Table 4 shows that the insured shart. increased slightly in 1970--I980 while the employer share increased considerably more anfl the state share declined sharplv. In 1980 the insured share was 16%, one fourth of the combined employer and state share of' 66%. A comparison of 17 Latin American and Caribbean countries for which these data were available in 1980 indicates that the share paid by the insured in Colombia was the smallest of all. 26. Colombia's social security financial structure also shows sigiificant inequalities among funds. Table 5 shows that in ISS the insured share is higher and the combined employer and( state share is lower t ban the national average shown in Table 4. Contributions by the insured decrease dramatically within the public sector while the employer and state share increase to 88%. The distributioni of revenue from the military is probably nore skewed than shown in Table 5 because it only includes data from the pension program. Health benefits for the military are entirely financed by the state and the insured's total share is likely to be the smallest of all. D. Evasioni and Payment Delays 27. Evasion and payment delays seem to be serious problems in Colombia. Within the ISS, major causes of these phentomena are: incomplete enrollment, deficient registration and lack of effective mechanisms for control and enforcement. Large and middle size enterprises are easy to control, but not small enterprises. ln Colombia 97% of the enterprises are small having less than 50 employees and they account for about half of the insured (Table 6). ISS authorities believe that there is a large number of small enterprises (e.g., small commerce, restaurants, repair shops, professional services) which are evading their obligations. In 1980-1981 the ISS conducted a campaign in the four largest cities to find and enroll these evadees. The campaign was particularly successfujl in CaJi but has not been repeated or extended to other cities. 28. Rough estimates of payment delays in ISS range from 8% to 10%. This figure cannot be disaggregated by time period since delinquent, accounts are not aged. However, in 1985, past dues reached 19 billioni pesos (tUS$135 million) or more than 20% of collected revenue in that year. The starting problem is deficient registration: there is no adequate check of the employer's address, assets, income, number of employees and payroll. The monthly bill is sent to the original address and for- the declared payroll. In 1980 the department of inspection was eliminated on charges of corruption and fraud, hence the initial employer's declaration cannot be corroborated nor is all updated checking possible. There is a lack of coordination and overlapping between the ISS divisions of accounting, budget, and the treasury, e.g., the accounting division gathers data on what should be collected and the budget division on what is actually paid; the treasury has elaborate regulations for control but the departmental administrators often ignore them. Convicted delinquent employers are expected to pay 2.5%  - 8 - interest monthly which is only slightly more thant the current. rate of inflation, and fines are not regulated. The Central ISS Juridical Office cannot prosecute delinquent employers; such power is granted to the Departments. But the latter have few incentives to take the initiative since the ISS has a fund to help defray budgetary deficits in departments. The departmental labor courts are overburdened with cases and arbitrarily select a fraction of the ISS pending claims for prosecution. 29. State evasion and payment delays in CAJANAL is a serious prot.lem also. In this case the evader or delinquent is easily identifiable but almost impossible to prosecute. Furthermore, public institutions often collect the emo]oyees' contributions but do not transfer them to CAJANAIL: 949 million pesos (US$ 1 million) were due for this reason alone in the first half of 1983. In other cases the state Treasury has not fixed the contribution which the public institution (e.g., departmental health services, private schools) ought to pay. Finally, the state often fails to pay fully the legally established contribution as an employer: in 1.985 the debt to CAJANAL for pensions alone was 2.4 billion pesos, equivalent to US$ 170 million (15, 73). E. Impact of Contributions on Income Distribution and Employment 30. Although there are no studies of the impact of social security financing on income distribution in Colombia, the above description and the experience of other Latin American countries permit one to make an educated guess. Regressive features of the Colombian system are: very low population coverage and of a relatively highi income group of the population: exclusion of neediest from coverage; very high wage contributions of which the insured pays a rather small proportion; high employer contributions; huge state subsidies to groups with relatively high income; and wage conttributioll ceilings. Due to the overall regressive structute of the tax system in Colombia and the relatively high income level of those who receive the bulk of the state subsidies, the latter probably have a regressive impact on distribution. The issue of the economic impact of the employer's contribution is unresolved in Latin America but the available literature and scarce empirical evidence suggest that in countries like Colombia the possibilicy of a forward transfer to the consuimer through price increases (or of stagnant employment due to the substitution of capital for labor) is higher than the possibility of a backward transfer to the insured worker (as a cut in his real salary). If the regressive impact. is via employment, lhe Colombian high contribution rate should accentuate that negative effect (this is particularly important in view of growing unemployment). On the other hand, if the regressive impact is via transfer to the consumer, the very small coverage of the population should also aggravate the regressive effect (55). Currently a research project is being conducted by DNP on the impact of taxes on distribution and employment.  - 9 - F. Financing Issues 31. Two pressing needs facing Colombia's social security system are to increase revenue and to eliminate inequalities in contributions. All of the following measures would enhance revenues: (a) use the entire salary as a base for contributions; (b) increase or eliminate the wage contribution ceiling; (c) have the insured pay a higher share of the premium especially ini public sector funds; (d) adjust the professional risk classification and charge the ATEP premium accordingly; and (e) aggressively pursue payment evaders a,ld late payors. A quick study can and should be done to calculate the impact of each one of these measures on revenue increases. The second issue deals with the inequity and the inequalities in the contribution rates by the insured especially in public sector funds. The very low or zero insured contribution in some public funds increases the burden on the central government budget. IV. USES OF SOCIAL SECUWITY FUNDS 32. Social security funds are used for three purposes: (a) to pay for benefits for the insured and their dependents; (b) to pay the administrative cost of the funds; and (c) to invest for future use. This chapter describes and analyzes these three uses of funds. The bulk of money is used to pay benefits and therefore more attenition will be paid to a description of who is insuredI and what benefits are provided. A. Benefits to Insured 1. Who is Insured? 33. A distinction needs to be madf- between legal coverage and statistical coverage. The former reflects the intent of existing laws and specifies who ought to be protected from the financial consequences of ill health, old age, disability and death. Statistical coverage is closer to reality showing who is insured andl thus able to receive the benefits prescribed by law. 34. Private Sector Legal Coverapg. ISS should cover all salaried workers in the private sector for old age, disability, death, common sickness, maternity ancl occupational risks. The self-employed and domestic servants can voluntarily join the ISS; unpaid family workers and the unemp]oyed are not eligible for coverage; pensioners are entitled to health care. Until the mid--1970s the ISS only covered the dependent children of' the insured below one year of age as well as the wife of the insured for maternity care. With the introduction of the family health program in 1975, the ISS has slowly extended dependent, c.overage to include children below 18 years of age and to provide full health care to the insured's wife.  - 10 - 35. Public Sector Legal Coveage. CAJANAL should cover all civil servants, who are not under specific funds, against the same risks as the ISS insured; pensioners are equally entitled to health care. Originally, dependent children below one year of age were entitled to care in CAJANAL but this right was later limited to infants less than six months old; the insured's wile is covered only for maternity. Dependents under CAPRESUB, CAPRECOM, the military and the police are fully covered for health care. 36. Family A_lowance Legal Coverage. All salaried workers in the private and public sectors should be covered by family allowances. Financial subsidies are paid to families who receive less than four minimum wages. All the insured and their dependent families are entitled to other benefits (health care, recreation, merchandise and food discounts) which vary accor(ling to the size and financial resources of the familv allowance fullnd. 37. Occupa4tional Risk Legal Cove.rge. All salaried workers are protected against the financial consequences of work accidents and occupational diseases (ATEP). Protection is provided through social insurance funds (e.g ISS) or, when that is not available, through the employer's direct responsib.ility. Employers can take out private insurance to protect themselves. 38. Statistical Coverage. Table 7 present.s estimates of statistical coverage on health care and pensions of the total population and the economically active population (EAP). Because of overestimates and reports of double coverage (particularly among dependents), the overall insured estimates in Table 7 are on the high side. The table shows that coverage of the total population for health care (although with different degrees of coverage among funds) doubled between 1970 and 1980 from 7.5% to 15%, but the next five years only added one percentage point. The principal reason for ihe rapid increase in coverage in the decade of the 1970s was the introduction of Llte fanmily hlealth program in !SS. Coverage of the RAF, is higher but las not expanded so rapidly. It. went from 22% to 30% between 1970 and 1980 and remained at that level for the next five years partially due to the negat.ive impact of the economic crisis on employment., particularly in the formal (salaried) sector. 39. The large majority of the insured are in ISS: 69.5% or about three million Colombians (Table 8). CAJANAL accounts for 6.9% of the insure(d or about 300,000 people. These two funds represent 76% of the total number of insured in Colombia and because of size and data availability this analysis focuses on these two funds. Disaggregated data on popnlat ion coverage by ISS and CAJANAL are presented in Table 9 for a 20 year perio(i. 40. Even though the two major social insurance funds were established 40 years ago, the populatioin coverage in Colombia is very low. In a comparison made among 20 Latin American countries in 1980, Colombia ranked 13th in total population coverage placing itself below lesser developed countries like Bolivia, Paraguay and Peru. In terms of EAP coverage,  - 11 - Colombia ranked 12th, below lesser--developed countries such as Guatemala and Peru (55). About the same time Colombia ranked ninth among Latin American countries on the' basis of per capita income (71a). 41. Private Sector Statistical Coverage. Total population coverage in ISS steadily increased from 3.7% to 10.2% in 1964--1978 (Table 9). In the next six years the percentage climbed only to 10.9%. (The 11.3% for 1985 is a preliminary figure). The pattern of expansion of active ar,d dependent insured is very similar to that of the total insured, however, dependents coverage sharply increased in 1975--1977 due to the introduction of the family health care program, butt slowed down after 1978. However, only 20% of the total insured were under the family health program (46) and they were mostly concentrated in five departments. Table 9 also shows that the ISS quotient of demographic burden (passive divided by active) increased 13 times in 1970-1985 as a result of the pension program maturing and a significant slowdown in the rate of increase of active insured. Still the ISS quotient is rather low in comparison with that of other major funds in Latin America. 42. Public Sector Statistical Coverage. CAJANAL total insured in relation to the total population increased from 0.5% to 1.1% in 1973-1985, while active insure(d in relation to EAP rose from 1.7% to 2.9% in the same period (Table 9). The ratio of dependents/active insured in 1984 in CAJANAL was less than one-fifth of that of 1SS, rteflecting a smaller dependent coverage in CAJANAL. On the other hand, the quotient of demographic burden in CAJANAL was more than three Limes that. of' ISS because of a lower retirement age and a longer time in retirement. 43. Family Allowance Coverae. Coverage for family allowances is lower than that for pensions and health care. In 1984 only 20% of the EAP or 9% of the total population was covered for family allowances (Table 10) compared with respectively 30% and 16% for pensions and health care benefits (Table 7). Only 37.8% of the salaried labor force was covered for family allowances compared with 51.6% for pensions and health care (Table 11) which may indicate that evasiorn is worse in family allowances than in social insurance. However-, the rate of expansion of family allowance coverage in the first half of the 1980s was three t.imes higher than the rate of expansion of social insurance coverage. It should be noted that there is a significant duplicat.ion in the populations covered by social insurance and family allowances but the number of children receiving family allowance benefits is higlier than children ent-itled to health care under social insuranice. Assessment ol_ Social Insurance Coverage 44. The low social security coverage and its slow rate of expansion since the late 1970s is largely determined by the composition of Colombia's labor force. The percentage of Colom)ian wage earners in the EAP is one of the lowest in Latin America (53.5% in 1983, the ninth among 13 countries for which these data were available) while the combined percentage of self- employed and inpaid family workers (42.5%) is among the highest in the  - 12 - region and has been rising lately.2 In 1980, Colombia ranked eighth in the size of its urban formal sector and had the fourth largest urban informal sector among 14 countries compared by PREALC (55). Low coverage is partly explained by two factors: (a) Colombia adopted the traditional Bismarckian social insurance model which basically covers wage earners in urban areas; and (b) poor law enforcement and evasion. 45. Only half of all wage earners in the country were covered by social insurance in 1984 (Table 11). Estimates of ISS coverage in 1981 for private wage earners ranged from 41.1% to 49.8%, thus smaller than overall coverage of all salaried workers (38, 41). The non-covered by social insurance in Colombia are the self-employed, domestic servants, unpaid family workers and the unemployed, as well as a good number of salaried workers who should be legally covered but who are not. The number of self- employed covered by ISS is very small and they are concentrated in the major cities. The total estimate is 7,000 or 0.6% of the total number of self- employed in the country. There are no figures on the even smaller number of domestic servants that may have joined the ISS voluntarily since 1984. 46. Coverage Differences by he and Sex. ISS health-care coverage is concentrated among the insured of productive age (15 - 59) and to a lesser extent among the elderly (Table 12). The perceentage of coverage in children below 14 is one-fourth that of the productive-age cohort. Data on the composition of the active insured by sex indicates a ratio of 2.2 males to 1 female (but exhibiting a declining trend) and reflects the salaried labor force composition in the nation (4). 47. Cover-age Differences by Economic AcLivities. Table 13 shows significant differences in coverage by major social insurance funds among economic activities. In 1980 the highest coverage occurred in public utilities (68%), followed by manufacturing (45%), government/personal services (39%), and transport (32%). The lowest coverage was in agriculture (5%), mining (11%), and construction (20%). Colombia's extreme differences in coverage among economic activities have been found only in very few Latin American countries (55, 56a). There is a low proportion of wage earners and a high proportion of self-employed in agriculture and commerce which may explain the low degree of coverage in these two activities. This is not the case in construction and transport, however, and it. may be that practices used by employers such as piece work and subcontracting may facilitate evasion of affiliation (4). 48. Geographical Coverage Aiong Deartments. Differences in coverage among departments are presented in Table 14. The more developed a department, the higher its social insurance coverage. The four most developed departments (Cundiriamarca, Valle, Antioquia, and Atlantico) have 2 The percentage distribution of Colombia's EAP by type of worker was not in the 1980 comparison of 13 Latin American countries; it was taken from ISS estimates and corresponds to 1983 (44) while data from the other countries were from 1980 (55).  - 13 the highest percentages of coverage: 12% to 24.7%. Conversely, four of the five least developed departments (Choco, Caqueta, Sucre and Cordoba) have the lowest percentages of coverage; 2.2% to 4.2%. Coverage in Atlantico (and Bogota -- 22% - not shown in the table) is about, nine times higher than in Choco. These extreme differentials in geographical coverage are found only in very few Latin American countries (55, 56a). There is an exception in the relationship between department development and social insurance coverage: departments where the family health program has been introduced and has reached the majority of the insured, exhibit a percentage of coverage higher than their corresponding level of development. This is the case of Cesar, Cauca, Norte Santander and Meta. Profiles of Insured and Non--Insured 49. Natio,ial Health Survey data has permitted the analysis of the relationship between other economic variables and social insurance coverage. Such analysis shows that (a) coverage by all funds increases with population size of communities; (b) household income is positively related with the number of insured in the household: the lowest income househoids do not have any insured whiile the highest income households have an average of four insured; (c) the illiteracy rate among all the insured i5 one-third of the rate of the nor-insured; and (d) home ownership and access to sanitary facilities is highly related with social insurance coverage (4). 50. The previous analysis demonstrates that the typical insured in Colombia is a male, in his productive age, with relatively high income, a wage earner in the most powerful enterprises or in government services, who lives in the most urbanized, developed departments/cities, and with high access to other social services such as education and housing. by contrast the non-insured is either a child or a female or a low-income male who is either self-employed, a domestic servant, an unpaid family worker, an unemployed or a wage earner in agriculture, mining, construction, commerce or personal services, who lives in the least. developed departments and least populated communities and who has poor access to education, decent housing and other social services. Thus, the neediest are excluded from social insurance. 51. Colombia lags in coverage level achieved by Latin American countries in a similar stage of economic development and in the ultimate goal of universal social securit.y. Expansion of coverage should be a primary target but it would not be feasible with the current high costs and low efficiency levels of the country's traditional model of social insurance. These issues are discussed in the following sections. 2. What Benefits Are Provided to the Insur-ed? 52. The preceding section defined who is insured and entitled to benefits under the various social insurance funds. The next. set of paragraphs describes the five types of benefits available, the size of each benefit category andI its evolutioni over time. The two major entitlements,  - 14 - viz. pensions and health care, are then described in more detail and an assessment of the benefits provided concludes this section. Yypes_of Benefits 53. Social insurance coverage in Colombia is quite low by Latin American standards but the small troportion of insured, particularly in public sector funds, enjoy more benefits and easier entitlement conditions than in most countries in the region. Table 15 summarizes the most significant dlifferences among social insurance funds in terms of benefit entitlement. 54. Penisions. All pension programs cover old-age, disability and death risks, but most funds in the public sector also provide seniority pensions. Ages of retirement in ISS are 60 for males and 55 for females after only ten years of contribution. This places Colombia among the five Latin American countries with the lowest combined age of retirement and years of contribution for old-age pensions. The retirement age in some countries is 65 for both sexes (including Guatemala which has a lower life expectancy than Colombia) or 65 for males and 60 for females (including Honduras, also vith a lower life expectancy) or 60 for both sexes (including Nicaragua and the Dominican Republic, both with lower life expectancies). Six countries set the same ages of retirement as Colombia but require from five to 20 more years of work. 55. ISS retirement conditions are the strictest in Colombia. An insured in CAJANAL and other public funds can retire at age 55 (females can do so at 50 if they had 15 years of service by 1985), and in some funds retirement is possible at 50 for both sexes or even at 45 for females. These conditions are among the most favorable in the Latin American region. 56. In some public funds and among certain occupations within CAJANAL, the insured receives a seniority pension and is allowed to retire with ten to 30 years of service at any age. Thus, an insured who begins work at age 15 can technically retire at, age 30 in three funds, at age 35 in eight funds, at age 40 in two funds, and at age 45 in one fund. These exceptional condit.ions are usually not. relat.ed to the strenuous or dangerous nature of the insured's work since they are applied, among others, to musicians, judges, journalists and public registrars. Few Latin American countries still have seniority pensions. Chile, a pioneer, abolished them in 1979. 57. A lump sum is paid as life insurance in aill Colombian funds in addition to pensions, but the number of months of salary paid varies from 12 to 48 (in case of natural death) and from 24 to 60 (in case of death by occupational risk). Most funds pay an indemnity (lump sum) to those insured or to their survivors when there is no right to pension. Finally, funeral aid is paid to the dependents of the insured. 58. Health Care Benefits. Health-maternity benefits are more limited than pensions. The worker receives full health care benefits but not the dependents. Benefits are only granted to the wife for maternity care and to  - 15 - the children under one year in ISS and under six months in CAJANAL. Benefits provided to the 20% of the ISS insured covered by the family health program are equivalent to those in the most advanced Latin American countries. The insured in some puiblic funds enjoy exceptionally good medical benefits. For instance, the military, CAPRESUB, and CAPRECOM offer orthod.ntia and cost of treatment abroad when not available domestically. 'l'he latt.er benefit was eliminated by ISS in 1986. Conditions for entitlement to health and maternity benefits (monetary and in services) are more liberal in ISS than in most Latin American countries except for the duration of paid maternity leave (56). Entitlement conditions for health benefits in some public funds are very generous, e.g., the percentage of salary paid for common sickness leave is 66% in ISS and CAJANAL but it increases to 95% and to 100% for the judiciary, the military, the Compt oller and other groups (Table 15). 59. Family_Allowances. Colombia is one of seven Latin American countries with a program of family allowances. The location of the enterprise det.er-mines in which fund those that are entitled to coverage must enroll. Funds in economically well developed cities, areas, or Departments, have more resources and provide better benefits. About 60% of fund revenue is paid in cash benefits to members who earn less than four minimum wages and 20% is paid in services: health (11%), education (4%), and training and recreation (5%). The wealthier funds own theatres and vacational centers which are out of reach for many of their members. The role of general merchandise and food stores with subsidized prices is important in financing the family allowance program; they generated 66% of income and accounted for about 54% of expenditures in 1984 (67, 83). 60. Unemployment Compensation. There is no typica] unemployment. compensation in Colombia. Only five Latin American countries - all more developed than Cotombia - have such a program. However, there is severance pay: a lump sum fixed according to salary and years of work. This benefit is usuallv paid by social insurance funds but increasingly it is paid directly by the employer. 61. Loans. Finally, in several public sector funds the insured has the right to request a variety of loans under preferential terms for personal consumption and housing. Distribution of Benefits by Type 62. The percentage distrioution of Colombia's social security system expenditures by major categories is based on ILO data and is shown in Table 16. Public funds expenditures are placed in a separate category and are not disaggregated by benefit program. Public funds have had the largest share of expenditures since 1978 and the lack of disaggregation obscures the interpretation of the expenditure distribution. Bearing in mind this caveat, Table 16 shows a declining trend in health expenditures, an increasing trend in pensions, and stable expenditu.res in family allowances. The share of pensions in Table 16 should be higher and its increase more  - 16 - pronounced because the bulk of public fund revenues is paid out for pensions. 63. The distribution of ISS expenditures cannot be presented in the same format as in Table 16. ISS administrative expenditures cannot be separated from other expenditures and there are also difficulties in estimating the cost of benefit expenditures accurately. Table 17 provides a rough estimate of the distribution of ISS expenditures by benefit program3 and shows a steady decline in the health-maternity share (from 79.1% to 53.7% in 1975--1985), a steady increase in the pension share (form 16.7% to 41.2%) and a stable occupational risk share (about 5%). If these trends continue, the shares of pensions and health benefits will reverse themselves in 1987, and pensions will become the largest expense item within ISS. Expenditures in public sector funds are larger for pensions than for health beniefits. CAJANAL spent 71% of its revenue on pensions and 20% on health in 1985; CAPRESUB spent 72% on monetary benefits (mostly pensions) and 18% on health in 1984; and in the same year CAPRECOM spernt 63^ on monetary benefits and 30% on health. Most expenditures of Department/Municipal funds are also for pensions and for severance pay (9, 14, 62, 72, 73). Adjustment of Pensions 64. Colombia has one of the best. annual pension adjustments of Latin America. A comparison of the real value of pensions of Colombia with those of six countries with m;e advanced socia.l security systems indicates that only in two (Costa Rica and Mexico) was the real value of pensions in 1980 compared to 1970 higher than in Colombia (55). 65. From 1972 to 1976 all pensions had a minimum equivalent to 90% of the minimum wage but all pensions were not adjusted annually to the cost of livitng. In 1976 the minimum pension was increased to 100% of the minimum wage and it became mandat.ory to readjust pensions on an annual basis. Currently pensions are adjusted on January 1st for the entire year according to the following formula: 50% of the difference between the value of the old and new minimum wage plus 50% of the consumer price index (CPI) increment, with an overall minimum pension increase of 15% over the previous pension. Most pensioners are also paid a thirtaenth month. Pension adjustment has a lag of two years: (a) t.he CPI used is December 31 of two years before the adjustment (i.e., pensions readjusted on January 1st, 1986 used the CPI of December 31, 1984); and (b) the "old" and "new" minimum wages used in the adjustment are respectively two and one years old (i.e., pensions readjusted on January 1, 1986, were based on t.he "old" minimum wage of 1984 and the 3 In ISS accounting, all costs of salaries and fringe benefits, supplies and equipment, investment and debt service are charged to the EGM program, hence artificially increasing its expenditures. On the ot.her hand, IVM and ATEP programs transfer to EGM a fixed percentage of their revenue to defray administrative costs. In order to estimate Table 17, IVM and ATEP transfers to EGM were charged as expenditures in each of these two programs and the combined value of such transfers was deducted from EGM expenditures.  - 17 - "new"o minimum wage of 1985, since the 1986 minimum was not in force yet) (12). 66. The real value of pensions in ISS is estimated in Table 18. Declines in seven years and gains in eight years resulted in a net gain of 65% over the 1970-85 period. About. 97% of all ISS pensioners receive the minimum pensiot. (wage). The average (nominal) pension is about 20% higher than the minimum wage andl exhibits a trend towards equalization with the minimum wage. 67. IL is becoming increasingly difficult to gain the right. to a pension and to collect it, particularly in public funds such as CAJANAL. The average length of time to process a pension in ISS is eight to twelve months and at times procedures may take two years or more. It takes one to two years on the average Lo process a pension at. CAJANAL and some cases have taken five years. These delays encourage the use of intermediaries and bribes. In addition, CAJANAL is behind in the payment of pensions and tens of thousands of court claims have been filed which are costly to the insured and the fund (1). 68. A number of legal projects and technical recommendations have been advanced in recent years to correct some of the flaws of the pension system: standardization of ent.itlement conditions; integration of public anl private funds to eliminate differences and to allow vesting and portability of benefits; increment of the age of retirement. to 60/55 in CAJANAL and 65 for both sexes in ISS; elimination of the right to pension by brothers and sist.ers; expedition of pension procedures; and stimulation of supplementary private pensions with premiums paid monthly by the insured (1, 2, 15). Only one partial change was made, i.e. an increase in the retirement age of females under CAJANAL to the same age as males. Provision of Health Care Benefits 69. The social insurance sub-sector of the Colombian health care system covers about 16% of the population, the Ministry of Health 75% and the private sector 9%. ISS owns 37 hospitals with 4,200 beds, contracts beds in 171 MOH and private sector hospit.als, and has a network of about 50 basic care centers (CABS). The military and the police have their own health care facilities to serve their members and dependents and little operational information is available on the system. CAJANAL has two small hospitals: one in Bogota and one in Tunja. No other social insurance funds have their own inpatient facilities. In general public funds, except the military and police, contract for inpatient care with ISS or the private sector. They usually have their own facilities for outpatient care staffed by their own health personnel. ISS is the largest provider of healt.h care under social insurance auspices and has a iarge physical infrastructure. Most data available relate to ISS. 70. Health care provided by private and public insurance funds is curative, hospital-based, technology intensive and delivered by physicians. The medical care orientation fits the social insurance clientele which is  - 18 - partly urban, literate, employed, and with low healthl risks, and partly old and retired with large medical needs. 71. The population served has high expectations from the benefits of medical science. Services are free because t.hey have been prep)aid through contributions. Physicians are salaried and face few constraints in t he use of resources . These fact.ors woul (I predict. excess i ve ut tIi zat i on of services. Yet. ISS reports about three physician visits per petson per year which is reasonable and well below developed country figures of more, than five visits. These visits include maternity care which accounts for above average outpatient. visils. [SS reports two beds per 1,000 insured which is on the low side for the kind of population served (Table 19). At lhe same time, it reports a low 70% hospital oc(c-upancy rate indicatinjg underutilization. Hospital occupancy percentages vary wi(deLy amonig Departments but reach lhe 80% level onily in the Bogola area (Table 20). Average lengths of hospital stay are within an acceptable range of five lo six days. The ratio of employees per hospital bed in ISS is close to 5 which is considered high by international niorms. Thus, available dlata indicat.e modlerate ut.ilization of services, generous staffing an(d a slightly underused installed capacity. 72. Underutilized hospital capacity andl medical resources occurs mostly at the local level whete hospital occupancy is lowest. The rate of occupancy increases a little at. Ihe secondary level (regional hospit.als) and becomes quite high at the tertiary level (national specialized hospitals). Hospital occupancy at, the local level is about one-half of that of the tertiary level. This is partly due lo the low quality of local level services, relatively easy geographic access to higher level facilities, cheap transportation and time off to seek medical care. These factors encourage the population to skip the local and regional levesls in search for care at the specialized facilities. The ISS emphasis on hospital building has changed in the last year. More resources are now being allocaLed to construct, equip and staff ambulatory care centers to reduce congestion at the top, reduce cost, and provide beLter services at the entry level '-16). 73. ISS contracts with MOH and the private sector about 30% of the needed hospital beds in some 171 hospitals. The cost. of these service-s in 1985 was 7.1 billion pesos or US$ 50 million. Contracted services are at the secondary and local levels in areas whern ISS does not have its own hospitals, or at the tertiary care level for high technology services which cannot be provided by ISS. There are no cost studies of cont.racted services and ISS does not have accurate data on the number of insured that use them (49). Fees paid by ISS to contracted services are set on the basis of 1SS' own health costs. However, the accounting difficulties to estimate rss costs properiy have been pointed out earlier. In addition, 15S officials acknowledge that unit costs are only available for specific hospitals and that there are doubts i, to the representativeness of the sample used i.61). 74. In the mid-1970s, at the time of the ISS reform, there was criticism of subcontract:ing services because of their high cost and low quality. Subsequently there was some reduction in subcontracting in 1977-  - 19 --- 1980 but it increased again thereafter (54). Currently, the ISS office of quality control reports that. it, is extremely difficult, to evaluate quality in contracted hospitals. The ISS is planning to check by sampling some specific contracted services such as x rays and high technology services. The new ISS policy of building basic care centers (41 are planned or under construction) shou]d reduce the need for sub-contracting at. the primary level. Before expanding this policy it is vital to estimate ISS unit costs accurately, to establish criteria for quality evaluation) and control of services, and to conduct a study of comparative costs between ISS and the public and private sectors. Assessment of Benefits 75. Benefits for members belonging to social insurance funds, especially the public sector ones, are more generous and entitlement conditions are easier than in most Latin American countries. The most influential pressure groups in Colombian society (the judiciary, selected civil servants, congressmen, the armed forces, and strategic trades or professionals) have been able to obtain privileged social security conditions and significant state subsidies. Even when these groups belong to a general fund they have successfully extracted special rules, e.g., there are 71 exceptions within CAJANAL, and ISS' own employees enjoy more liberal conditions than the rest of the ISS insured. 76. Important inequalities in pension benefits exist among funds (Table 21). In 1982 the lowest average pension was paid in ISS while pensions were 66% higher in CAJANAL, 151% higher in the military fund, and 245% higher in CAPRESUB. Pension differentials are not necessarily the exclusive result of higher incomes and contributions, but of the way the pension is calculated. ISS pensions are calculated as 75%4 of the average wages in the last two years, but as one moves up the organizational ladder the percentage gradually increases to 100% and the base becomes the highest salary or even the current salary at. the time the pension is granted. Furthermore, in ISS the basic wage is used for the calculation of pensions but in most public funds an enlarged base is used (although the contribution is imposed on the salary base), e.g., in CAJANAL, 14 fringe benefits are included in the base and in the Military all fringe benefits are included. 77. Health services benefits are generous for those covered by social insurance and they enjoy the best facilities and the highest per capita expenditures in the country. Within the social insurance sector, the best facilities (either direct or contracted) are available to the military and the police. 4 Before 1985 ISS pensions were calculated at 45% over the average salaries of the last two years (for the first 10 years of service); this was increased by 1.2% for each additional year of service with a maximum of 57%. Currently the addition is 2.5% per year of service above 10 until 20 for 75% of the average.  - 20 - 78. Social insurance benefits reach about one sixth of the Colombians. Among the 4.6 million people (overed a subset of about 400,O00( public sector employees enjoy exceptional beniefits and entitlement conditions and contributes only 3% lo 14% to thie cost of tiheir benefits with thie state paying frow 69% to 88%. 13. AdministraLtion Overall Administrative Costs 79. According to ILO data, the overall cost of' administrat ion of the Colombian social security system steadily increased from 4.4