José F. Alonso, Office of Research, Radio Martí[1],
Ricardo A. Donate-Armada, The Wyatt Company, and
Armando M. Lago, Ecosometrics Incorporated

DRAFT FOR COMMENTS. Any comments or inquiry, please fax them to: Armando Lago, Ecosometrics. (Work) 301-652-2640. (Fax) 301-907-8952

Section II (continued)
Cost Model of Health Care

A simple cost model of the Cuban health care system is presented in this section. The model structure is parametric, this means that costs are computed by simple multiplication of wages times work force in each occupational category, with other parameters for food, medicines, supplies, repairs and depreciation imputed from the Cuban data and from the experience of other developing countries.

Cost data from the Cuban health system is hard to come by and in the few cases when published it is published at high levels of aggregation, so that no information appears on cost elements or cost components. For many years, the 1979 hospital operating cost figure of 19.81 pesos per bed per day has been used for analytical purposes in several Cuban medical journals[88], cost figures verified by other investigators. Alemán and his associates estimated hospital operating costs per bed per day of 19.80 pesos in the period 1981-86 and 19.25 pesos per bed per day in 1986.[89] These cost figures have been updated in a more recent article, quoted earlier, in which Osvaldo Castro[90] estimated hospital operating costs of 25.49 pesos per bed per day in 1990, 25.46 pesos per bed-day in 1989, 26.44 pesos per bed-day in 1987, 25.21 pesos per bed-day in 1985, 22.97 pesos per bed-day in 1983 and 20.28 pesos per bed-day in 1981. In his article, O. Castro presents slightly detailed costs by type of care: hospitals and ambulatory care services, and within these categories a further separation into two cost elements: labor and all other costs.

While certainly an improvement over the previous cost analysis, O. Castro's methodology still suffers from two problems. One problem concerns the fact that his cost figures are much smaller than the Ministry of Public Health operating budget (excluding capital investments). Thus, O. Castro's cost data excludes overhead functions such as research and development and the general administrative functions performed at the national level by the Ministry personnel in La Habana. This problem in Castro's cost analysis is confirmed by the health-related cost data presented in the Anuario Estadístico de Cuba and in the Ministry of Public Health's Informes Anuales. The Informes Anuales present data on total employees working in the health sector and the Anuarios present the average salaries paid to them. Multiplication of these two items results in personnel costs which exceed the labor costs presented by Castro. A second problem concerns the level of aggregation in Castro's cost figures and its inability to distinguish between: intermediate costs( i.e. the costs of X-rays, laboratories, physiotherapy etc.), final costs (i.e. outpatient and inpatient hospital services in Medicine, Surgery, Pediatrics and Obstetrics-Gynecology, among others) and overhead costs ( administration, housekeeping, maintenance and utilities). Because of these problems, it was decided to build upon Castro's methodology by further separating into components and detailing his cost structure using data from Cuba's manning standards.

Calibrating Personnel Costs. The first step in the development of separate disjoined cost figures is the calibration of the personnel costs. To that effect the number of persons in the health work force (274,544 persons in 1987, and 290,799 persons in 1989) were collected from the Informes Anuales[91] and multiplied times the annual average salaries for the health sector (i.e. 2,256 pesos in 1987, and 2,341 pesos in 1988) from the Anuarios Estadísticos[92]. This is the same procedure used above to evaluate the reasonability of Castro's cost data. The personnel costs estimated by this procedure were: 619.37 million pesos in 1987 and 680.76 million pesos in 1989. Next, through contacts with recently arrived doctors from Cuba, salary ranges were ascertained for a variety of health personnel such as doctors, dentists, pharmacists, nurses and nurse aids, technicians (such as X-ray technicians, laboratory analysts dental technicians, pharmacy aides etc.), unskilled workers, workers in services (such as those in cleaning and food etc.), clerical and managerial staff, and other unspecified university graduates. These salary ranges were used to calibrate the personnel cost components, insuring that by selection of salaries within the given salary ranges the personnel costs estimated would approximate the personnel costs estimated using the aggregate average salary in the health sector. This calibration and approximation is presented in Table 9. The calibrated wages and salaries resulted in an estimate of personnel costs of 760.83 million pesos in 1992, salary costs corresponding to a total 310,726 health sector personnel for that year.

Allocating Health Personnel by Service Type. The Cuban government does not publish any data on the allocation of personnel by staff categories between hospitals, polyclinics, research centers, and Ministry of Public Health overhead. In view of this lack of pertinent data, manning standards at hospitals and polyclinics were used to allocate staff between the different services and to calibrate the personnel costs in order to add them to the comparable levels shown by Osvaldo Castro, as referenced above. The following allocation rules based on manning standards were followed:

Family Doctors:

the number of family doctors published in the Ministry of Public Health's Informes Anuales and an identical number of nurses (one nurse per doctor) were allocated to this program.

Polyclinics:

the number of doctors assigned to polyclinics was estimated assuming 4.5 consultations per hour (that is 8,100 consultations per doctor per year). The Cuban manning standard for polyclinics is five consultations per hour[93], but only half of the polyclinics seem to achieve this rate[94], which is lower than the 7.5 rates of consultation per hour achieved elsewhere in the Caribbean.[95] Nurses were allocated to polyclinics assuming rates of 2.85 consultations per nurse-hour, which is lower than the productivity rate of 3.135 consultations per nurse-hour achieved in 1980, according to statistics published by PAHO[96]. Deviations from the 1980 productivity rate were necessary to calibrate the costs of polyclinics. Dentists and pharmacists were imputed to polyclinics at the rate of one each per polyclinic facility. Allied- health technicians (x-rays, lab assistants etc.) were assigned at the rate of 35 technicians per facility, an average between large and small polyclinics. The staffing rates per polyclinic used for allocating the other personnel were developed from interviews with recently arrived doctors and included: five day-laborers ("peons" in Spanish), five service workers, 1.5 managers, and nine clerical workers per polyclinic.

Hospitals:

doctors were assigned to hospitals at the rate of 0.30 doctors per bed, rate which characterizes the Hermanos Almejeiras hospital (excluding the interns)[97]. Nurses were allocated at the rate of 0.62 nurses per hospital bed, again the source for this rate are the staffing standards of the Hermanos Almejeiras hospital, which we are forced to use for the lack of better data. Pharmacists and dentists were assigned as one per hospital. Technicians were assigned as 0.28 per hospital bed, in accordance with the previously presented data on hospital international staffing patterns. Day-laborers and hospital service workers were allocated at the combined rate of one per bed, clerical workers as 0.21 per bed, while managers were assigned as six per hospital facility. All these rates are within the international experience presented earlier, and their use allow us to approximate Osvaldo Castro's hospital cost figures.

Dental Clinics:

the remainder of the dentists and all the dental assistants were assigned to the dental clinics, which also included two day-laborers, two service workers, one manager and four clerical workers per clinic.

Overseas Medical Program:

the 1989 and 1992 allocations to this program come from Julie M.

Feinsilver[98] and include: 1,500 doctors, 1,500 nurses and

1,000 technicians in 1987-89, and 1,000 doctors and 1,000 nurses in 1991-92.

Research Institutes:

doctors were assigned to the research institutes on the basis of 0.66 doctors per bed, which is the rate for the Hermanos Almejeiras hospital counting interns and residents. All the rest of the personnel were assigned using the same rates as for hospitals, with the exception of other university-trained professionals. Twenty five percent of all the other University-trained professional working in the health sector were assigned to the research institutes.

Health Tourism:

staffing rates for the Cira Garcia hospital (44 beds) were assumed to be identical to the Hermanos Almejeiras hospital, which appears to be the jewel of the Cuban hospital system.

Other Health Facilities:

these facilities include the "balnearios" (mineral health spas), hogares maternos etc. amounting to 1008 institutions with 4,574 beds in 1989. The following manpower allocations were imputed: doctors at the rate of 0.1 per bed, nurses at the rate of 0.62 per bed, day laborers as 0.4 per bed, service workers as 0.60 per bed, managers and clerical workers as one per institution respectively.

Ministry of Public Health Overhead:

All the remainder personnel were allocated to the Ministry.

Application of these allocation rules result in the personnel costs presented in Table 10a which are in correspondence with the scant 1989 cost information presented by Osvaldo Castro, as referenced above. For example O. Castro estimates 1989 hospital costs as 310.48 million pesos, which closely correspond to the 1989 total hospital cost figure of 310.39 million pesos estimated in Table 10a using the manning standards. In addition O. Castro's 1989 ambulatory cost estimate of 190.04 million pesos is identical to the sum of the costs of the polyclinics, family doctors program, and dental clinics estimated in Table 10a using the manning standards. No similar comparison is possible for 1992 (see

Table 10b) because O. Castro's cost estimates do not include this latter year.

Allocation of Non-Personnel costs. There are also scant data available for the analysis of non-personnel costs. Willy de Geyndt[99] recorded the following components of hospital costs in the early seventies: wages and salaries (50%), food (10%), drugs, pharmaceutical and curative materials (25%), and hospital general administrative overhead (15%). Pedro Alemán and his associates[100] estimated the cost of drugs, pharmaceutical and curative materials to comprise 17.3%- 24.6 % of the costs of maternity and gynecological hospitals in 1981-82. With drugs and pharmaceutical comprising the bulk (82.4%-82.9%) of these expenses on drugs and materials. Alemán and his research group also reported that indirect costs, as a percent of total costs, at the Mariana Grajales gineco-obstetric provincial hospital in Santa Clara, amounted to 61.7 % in 1981 and 70.3% in 1982[101]. Furthermore, these researchers quote the cost analysis conducted using V. Y. Shilinskas's[102] costing methodology, which estimated that personnel costs were 56.2 percent of total health costs, with food accounting for 10.3 percent, 8.4 percent for drugs and materials, plus 4.9 percent for tools and equipment, among other basic costs in the overall health budget. UNICEF's[103] family doctor program plan specifies 1989 investment costs of doctors' offices of 35,000 pesos for construction and 10,400 pesos for equipment per office, and current costs of 6,528 pesos for covering doctor's and nurse's salaries plus supplies. By 1989[104] according to UNICEF the current costs of the family doctors program (including supplies and pharmaceutical) were 38.96 million pesos, that is 4.3 % of the Ministry of Public Health's budget, but even this UNICEF cost figure appears on the low side given the high proportion of family doctors out of the total stock of physicians.

To separate the hospital costs between in-patient and out-patient services, the unit cost relationship between them is assummed to be for 4:1, that is, in-patient cost per bed-day which are four times greater than the cost of out-patient visits.[105] The international experience on health costs that can be used to support our cost imputations. The costs of drugs and pharmaceuticals in developed countries range from 8.3% of total health costs in the United States to 21.0% in France. In between are Switzerland (8.5%), Sweden (9.0%), Canada (10.4%), United Kingdom (12.7%), Australia (13.3%), West Germany (17.9%),and Italy (20.0%)[106] The experience other Central American countries has been reported by Phillip Musgrove[107], who presents data on the composition of public expenditures in health in several countries. According to Musgrove, 1984 expenditures on materials, drugs and supplies as percentages of total health costs range from 29.4% in Nicaragua to 16.1% in Panama. In between were Costa Rica at 19.9%, Guatemala at 21.9% and Honduras at 25.6%. Comparable figures for health expenditures on machinery and equipment, as percentages of public health costs, were: 0.66% in Panama, 0.85% in Honduras, 1.17% in Costa Rica, 1.41% in Guatemala and at the peak was Nicaragua with 2.22% of public expenditure in health devoted to machinery and equipment. As the reader can note the Cuban cost estimates are in the general ballpark of the Central American experience.

Using other international hospital costs standards to estimate the missing cost elements, the total capital costs for hospitals have been estimated to range from three to five times (three is used as the capital costs multiple in this paper) the annual recurrent costs, while the total capital costs of polyclinics is estimated as twice their annual recurrent costs. For the Ministry of Public Health's general administration and overhead functions the capital costs are imputed as 1.5 times the corresponding annual recurrent costs. The capital costs of equipment are set at 30 percent of total capital costs (or 40 percent of the costs of the buildings) in accordance with international hospital costing conventions[108]. Capital costs data from developed countries show the capital costs of ambulatory services as comprising 2.3%-3.5% of total ambulatory costs in the United Kingdom and Australia respectively; whereas in hospital services the capital costs range from 7.5% of total costs in the United States to 15.6% in Switzerland, but in most of the developed nations (i.e. Sweden, West Germany and Australia), the capital costs of hospitals hover around 10.0% of total hospital costs.[109] Maintenance costs are imputed as two percent of the costs of the building per se plus seven percent of the equipment costs. The maintenance costs of buildings hovers between 7.0% and 7.5% of total health expenditures in several countries, such as in the Netherlands, United Kingdom, Australia, Sweden and Switzerland[110]. Annual depreciation costs are imputed by estimating payments for interest and capital amortization using capital recovery factors[111] of 10% annually during depreciation periods of 30 years for buildings and 15 years for equipment.

Projecting Real Price Inflation in Health Services. Since health price indexes generally rise faster than the GDP price deflator and other indexes of general price inflation, adjustments are needed to estimate the expected real rate of inflation in the health sector. Because of the excess supply of both professional and non-professional health personnel in the Cuban health system, the rate of price inflation for personnel costs can safely be projected at the same rate as the general rate of price inflation in Cuba. The same is true of the rate of price inflation in food costs which can be expected to correspond closely to the general rate of price inflation. No adjustments are therefore needed to correct for deviations in food and personnel prices from the general rate of price inflation.

However, price inflation adjustments are needed for two components: medicines and equipment, whose rate of inflation can be expected to exceed the national average of price inflation. From health cost data presented by P. Musgrove[112], the annual rates of price inflation for Costa Rica's health expenditures on medicines and supplies and on equipment was estimated as exceeding the overall rate of inflation in the medical sector by 9.4% for the period 1980-83. Similarly derived figures for other Central American countries were 9.6% annually for medicines and supplies in Guatemala in 1980-84, and 4.2% for medicines and supplies in Honduras from 1980 to 1984. In the United States, the excess rate of price inflation (in excess of the overall rate of price inflation measured by the GDP price deflator) for medicines was estimated as 4.5% in the period 1980-92, while electrical equipment prices grew 1.5% faster than the GDP price deflator in the period 1974-1981.[113] Based on the above estimates, the high inflation rates characteristic of Costa Rica were used for projecting the Cuban price inflation in medicines and equipment. No other price inflation adjustments are deemed necessary. Weighing the 1992 cost components presented in Table 10b by the 9.4% annual excess rate of price inflation in medicines and equipment results in the real price inflation rates for each of the health services presented in Table 11.

Structure of the Health Cost Model. The structure of the parametric health cost model used for projecting the costs of the existing Cuban health system is as follows:

(4) Ct = (1+ Oo) (1+ po)t {[[Sigma]]i [(CCit) (1+pci)t (1+Oi)] [[[Sigma]]j(Sijt)(POPjt)]+

+ [(KCit) (1+ pki)t ] [[[Sigma]]j (Sijt) (POPj)]}

whose terms are defined below . The health cost model used for projecting the costs of the health system reform options is almost identical to (4) except for the substitution of the overhead ceiling variable OCot for the product (1+O0) (1+po)t. The overhead ceiling is added to the other health services costs estimated in equation (4). The following variable definitions are used:

Ct = Total health costs in constant 1992 pesos during year t.

Oo =overall general and administrative expense overhead rate,which includes the expenses of the Ministry of Public Health, the research institutes, the overseas medical program, health tourism and other health facilities.

OCot =overhead costs ceiling during time t, which includes the overall general and administrative expenses , such as the expenses of the Ministry of Public Health, the research institutes, the overseas medical program, health tourism, and other health facilities.

CCit = current unit costs in 1992 dollars of health service type i during time t.

pci =annual real rate of price inflation, in excess of the overall rate of price

inflation, of current costs type c of health service type i.

Oi = overhead rate representative of general and administrative expenses of health service type i.

Sijt =number health service units of type i delivered to population cohort type j during time t.

POPjt =population in cohort type j during time t.

KCit = capital unit costs in 1992 dollars of health service type i during time t.

pki = annual real rate of price inflation, in excess of the overall rate of price inflation, of capital costs type k of health service type i.

The cost model formulated above is of the "pay-as-you-go" type; it is not estimated through actuarial methods, and thus results in intergenerational transfers which are usually corrected for in costing systems based on actuarial estimation methods. Perhaps the actuarial model will be developed in the second approximation health model contemplated for the future.

The structure of the cost model used in the cost projections is also presented in Table 12. The cost model was calibrated using 1992 costs, 1990 hospital utilization rates and 1992 ambulatory activity levels. All the costs are expressed in 1992 pesos. A projection of costs (in constant 1992 pesos) for the current unreformed and wasteful health system is presented in Tables 13 and 14 for year t+0 (1995). The projections use 1992 usage rates of ambulatory facilities and 1990 rates of hospital use, that is, before the recent crisis affecting the availability of medicines at hospitals. The projections use the population projections developed by Ricardo A. Donate-Armada[114] and the income elasticities presented earlier.

ANALYSIS OF HEALTH REFORM OPTIONS.

This section analyzes several options for reforming the Cuban health system so as to improve its efficiency and its financing system, but without imposing an undue burden on the employers and without reducing the quality of care available to the Cuban population. The reforms focus on reducing unnecessary waste without affecting the basic quality of health care in Cuba. First, the costs and demands of the current system are projected for the system as it is organized now, that is, with all the inefficiencies noted above. Two other general options are also analyzed which include private sector participation in both hospital and ambulatory services, including HMOs and private insurance systems. The first option undertakes the reform of the system so as to cut its level of waste and introduces health financing via payroll deductions (without co-payment provisions) from employers and employees, with the state financing only the heath care expenses of the unemployed and the physically- and mentally- handicapped. This first reform option has many elements similar to the health care financing system of Costa Rica. A second option also includes the cut in wasteful practices but adds a health financing system with co-payment features similar to the one in Chile. The next paragraphs describe and provide background on the reform options.

Policies to reduce waste, inefficiencies and over-investment. The inherent waste in the Cuban health system was documented in an earlier section. This waste included extremely high rates of use of ambulatory services (i.e. doctor visits), especially for infants and pregnant women. Also high staffing rates of non-health care personnel were observed at hospitals, whose utilization rates had decreased from the high levels of utilization of the early eighties (80%) to the current levels in the low seventies. To all these inefficiencies the family doctors program must be added, which appears to be unneeded, and should only be maintained in unserved rural areas. Finally, the excessive overheads of the Ministry of Public Health must be cut down to levels commensurable with the experience of other countries in Western Europe and North America.

The Cuban standards for the number of ambulatory health visits exceed those of the United States for two main cohort groups: infants with less than one year of age and pregnant women. In the case of the infants the Cuban standard was reduced from 16 annual visits to the U.S. standard of 8, thereby reducing ambulatory visits in half. The U.S. standard of visits for pregnant women of 12 visits was also adopted, leading to a 7.7% decrease in ambulatory visits for women in the 15-44 years old cohort. Policies to reduce waste in hospitals include increasing the utilization of hospital beds from the low 72% percent of the nineties to the 80% utilization rate of the early eighties and, in addition, reducing the current high rate of administrative and housekeeping staff per bed of 1.24 workers to the lower rate of 0.40 staff per bed of both Jamaica and the Dominican Republic.[115] The number of family doctors is frozen at the 1993 level of doctors in rural and mountain sectors plus half the number of family doctors in agricultural cooperatives.

The final waste reduction adjustment concerns the Ministry of Public Health overhead, which, at the current 27% (of all other current costs) level, exceeds by at factor of two or three all the other overhead costs referred to in the literature. The adjusted overhead expenses proposed kept intact the costs of the research institutes, the other health facilities, and the health tourism facilities, but a cut in half of the Ministry of Public Health's general and administrative expenses is proposed. The current overhead expenses are so large that, even with this draconian cut, the modified overhead expense rate hovers around 12.5% of the 1992 level of current expenses, a level larger than in all the countries researched, with the exception of the tiny Caribbean island of Dominica. The unit cost model impacts of these changes are presented in Table 12, while cost projections of the reformed system are presented in Tables 13 and 14 for two health financing system scenarios: a financing system financed entirely through taxes on wages (without co-payment features), and a second financing system with co-payment. Costs are smaller with co-payment because of the depressing effect of co-payment on demand for health care; since the effect of co-payment is to reduce unneeded visits to health care facilities (both ambulatory and hospital usage). The reader should note that expenses in doctors, nurses and allied health personnel remain unaltered, with the exception of the elimination of the unneeded family doctors, that is, basic health services remain unchanged under these reform proposals.

Health Care Financing Options without Co-Payment. The design of the financing option without co-payment follows the one in place in Costa Rica. In Costa Rica health and maternity benefits are financed through mandatory payroll deductions, where as percent of salaries, the employee pays 5.5 % of salary, the employer pays 9.25 % of all salaries paid by his firm, with the state contributing the remainder 1.25 % of wages and salaries. The total cost of health and maternity benefits in Costa Rica by the early eighties amounted to 16.0 percent of salaries and wages[116]. The health financing costs had been increasing in Costa Rica. According to Mesa-Lago[117], by 1979 health and maternity benefits were 11.0% of the total wages and salaries paid in Costa Rica, with salaried workers contributing 4.0 % of salaries, employers contributing 5.0 % of salaries and the residual 2.0% was contributed by the state. Focusing only on the relationship between employee vs. employer contribution, in 1979 employees paid 44.44 % of the combined employee/employer contributions, while after 1983 employees paid a lesser proportion: 37.29% of the combined employee/employer contributions to maternity and health benefits.

The high rates of payroll taxes used to finance health benefits in Costa Rica (14.75% of wages counting employer and employees contributions) are excessive. When applied to Cuban salaries and wages these rates are more than enough to finance an increase of 50% above the current Ministry of Public Health budget. In addition, the division of the burden between employer and employee is very unequal. The analytical task at hand is to design a payroll-based health financing system, without co-payment provisions, that would enable the financing of health for all the persons employed with their families. A 50%-50% split in payroll taxes is contemplated, with the employer responsible for the premiums (payroll taxes) to finance the employee's health expenses, and the employee roughly responsible for the premiums (payroll taxes)to cover his family's expenses.

Because of the large number of unemployed and handicapped persons resulting from the economic collapse of Cuba, the incipient private employer sector cannot afford to finance their expenses, which will be the responsibility of the state and which will be financed out of general tax revenues. Using the employment estimates presented in our earlier paper published by La Sociedad Económica[118], which were 3.998 million persons in year t+5, 4.195 million persons in year t+10, 4.757 million persons in year t+15 in the CBI scenario and 5.921 million persons in year t+15 in the NAFTA scenario, it was estimated that the Cuban government will be responsible for financing out of general tax revenues 20.10% of the population in years t+5, 20.26% in year t+10, and 13.03% of the population in year t+15 in the CBI scenario. The cohort groups financed by the state will include the unemployed and in addition the handicapped, which are estimated as 0.75% of the population following the 1980 survey of the handicapped conducted by the Ministry of Public Health.[119]

The final simulation of this modified Costa Rican system with no co-payment options is presented in Table 13, which shows that the combination of the reduction of waste and having the state become responsible for financing the unemployed and the handicapped results in payroll taxes of 3.84% - 3.99% of wages for the employers, employees and pensioners separately, a very competitive rate for financing health costs without adversely affecting the international competitiveness of Cuban exports.

Health Care Financing Options with Co-Payment . Co-payment has been a regular feature of health insurance programs in highly developed countries. The co-payment feature consist of the beneficiary sharing the cost of the health service--whether hospital or ambulatory -- with the insurance company or with the state as appropriate. Co-payment rates, measured in terms of the percentage of total hospital costs paid from general public revenues, vary among developed countries[120], ranging from 54% in the United States, 79% in West Germany and in excess of 90% in Canada (91%), France (92%), U. K. (99%) and Sweden (100%) in 1980. This variance extends to underdeveloped countries, with experiences in co-payment rates as varied as those in the Dominican Republic (97.3%-98.5% in 1986), Honduras (94.7%-96.5% in 1985) and Jamaica (92.5%-97.7% in 1986-87), and on the other extreme Bolivia (38.4%-64.0% in 1986-88).[121] .

In between these co-payment rates is the co-payment system of Chile, a country of interest to us because of its strong free market and free enterprise orientation. Chile's health system, which is devoid of employer health financing costs, has several tiers of service.: the state finances 100% of the costs of the National Health Service System (SNSS), which include municipal hospitals and clinics, as well as immunization and public health functions) which serve mostly unskilled and domestic workers, small farmers, the poor and indigent; and a health insurance system, called the Preferred-Provider System (SPP), administered by the National Health Fund (FONASA), which serves the rest of the population and is financed through a 6% mandatory payroll deduction plus a copayment system administered through vouchers bought at public outlets, such as banks, clinics, public health offices and other institutions.

The co-payment rates in the Chilean system vary by the quality of the freely-elected health service provided : the co-payment for the most basic service covering service by the general practitioners is 50%, a second tier of services has a co-payment rate of 33%, with the most expensive services, covering services from experienced specialists, having a 25% co-payment rate. Most persons covered by this system experience a co-payment rate which ranges from 50% to 33% of the costs of service. Private doctors and private hospitals participate in the system and negotiate rates with the government agency administering the system. In addition, the employee may contract with a private health insurance company approved by the state or with officially-approved health maintenance organizations, called Instituciones de Salud Previsional (ISAPREs), and remit his 6% payroll deduction contribution to any of these other plans. Private providers accounted for 66% of the value of health services provided in Chile in 1980, up from 53% in 1969.[122] The relative participation of Chileans in these systems in 1983 were: 28.0% were served by SPP, 4.2% were served by the ISAPREs' pre-payment plans (similar to our HMOs), with the rest 67.8% served by SNSS.[123]; that is, 67.8 % of the Chilean population were exempt from paying for health care. The state ends up paying 60% of the costs of serving both SNSS and SPP patients, with workers' mandatory payroll contributions accounting for the rest, or 40% of the public health system costs[124]. This translates into the fact that FONASA experiences a surplus in its services to the tax-paying SPP customers and that the surplus is used to finance partially services to the unskilled workers, the poor and indigent served by SNSS.

While the 6% Chilean payroll tax and the co-payment feature are generally adequate to finance the Cuban health sector, a modification of this system is proposed because of the lack of burden sharing (i.e. tax sharing) with the employers in the Chilean scheme. The analysis focuses on a payroll tax of half the Chilean rates (close to 3% of payroll for each -employer and employees-), thus getting rid of the pro-employer bias of the Chilean health system. Health users, except the unemployed and the handicapped, are assumed to pay 20% of the medical fees charged as a result of use of health facilities, a co-payment feature more advantageous to them than in the current Chilean health financing system. As in the first option, the health expenditures of the unemployed and the handicapped are assumed by the state, which should help in minimizing the regressive features of user fees and co-payments, since the price elasticities for health care expenses are more elastic for the lower income groups.[125] Using the mid-point elasticity formula, at price elasticity rate of -0.2, the imposition of 20% user fees will depress usage rates 33% for the population subject to user fees and by 26.0%-29.0% for the total population. The lower rate for the general population reflects the fact that 13.03% - 20.26% (the lower figure is for year t+15) of the population are unemployed and handicapped persons not subject to paying user fees.

The results of the simulation of the costs of a health financing system with co-payment are presented in Table 14. The costs with prepayment options are smaller due to the elimination of some perhaps unneeded wasteful usage of medical facilities. Under this health financing system employers pay health insurance taxes in the range of 2.70% - 2.80% of the payroll wages, again very competitive tax rates for competing in international export markets. Employees also pay less under co-payment provisions, but as the result of lower usage of medical services. The simulation of these health financing options shows that it is possible to cut wasteful practices and expenses in the Cuban health system without affecting the basic quality of medical care in Cuba.

Summarizing the discussion above, the government financing share of the two health financing options, which range from 13.03% to 20.26% of the health costs, while high in comparison to most countries, is still within the experience of countries like Switzerland (22.0%), Sweden (15.0%) and Japan (14.0%), that is the government financing share is reasonable. The employer payroll rates of 2.70% to 3.99% are in the lower third tier of the countries surveyed by the U.S. Social Security Administration[126] and in closed correspondence to those of Singapore (3.0%), Austria (3.0%), Canada (3.45%), Uruguay (4.0%), Korea (4.0%) and Belgium (4.0%) while higher than those of Spain (2.20%) and the United States (1.45%). This employer payroll rate is altogether reasonable and will not jeopardize the cost position of Cuban firms facing international competition.

Privatization Strategies within the Health Reform Options. Both health reform options--with and without co-payment--include strong privatization components. Private participation in both ambulatory care and in hospital care would be actively promoted.[127] The financing of health expenditures on the services provided by the private health sector would be financed in a fashion similar to the scheme adopted in Chile; that is, the health insurance financed through payroll taxes would be responsible for financing a "basic" service at fees equivalent to the cost of the public sector hospitals and polyclinics, while the excess over the "basic service fees" will be the responsibility of the individual health system user. This is similar to the way the Chilean system operates, where the individual pays sometimes 75% of the fees for some services provided by the private health sector, with the payroll tax-financed health insurance paying for the rest. In addition, individuals and firms should have the option of contracting directly with health maintenance organizations and opting out of the state health insurance system.

While it is unrealistic to expect that the private participation in the health sector can match the private participation in the days of pre-revolutionary Cuba in the short projection period (15 years) analyzed in this study, rapid movement to privatization should be expected as a result of the excess supply of doctors available now, excess supply that can only find employment in the private sector or self-employment by opening private practice offices. In 1954, approximately 34% of the hospital beds of Cuba were in private hospitals and health maintenance organization, while close to 50% of the doctors were providing services in the private sector[128]. The goal for privatization activities in Cuba would be to achieve half of these pre-revolutionary privatization rates by year t+15 of the projection period.

Go to Section III & IV