79 PSP Discussion Paper Series 2084 1 Social Sectors in Transition: The Case of Romania K. Subbarao Kalpana Mehra November 1995 Poverty and Social Policy Department Human Capital Development and Operations Policy The World Bank This Booklet of Abstracts contains short summaries of recent PSP Discussion Papers; copies of specific papers K may be requested from Patricia G. Sanchez via All-in-One. The views expressed in the papers are those of the authors and do not necessarily represent the official policy of the Bank. Rather, the papers reflect work in progress. They are intended to make lessons emerging from the current work program available to operational staff quickly and easily, as well as to stimulate discussion and comment. They also serve as the building blocks for subsequent policy and best practice papers. Abstract Access of the poor to social services is an important component of the poverty reduction strategy. Although many countries in transition have had reasonably well functioning health and education systems prior to the transition, considerable deterioration has occurred in these services following the transition to a market economy. In this context, it is important to know how the poor have fared relative to the nonpoor, in their access to social services. In this paper, the data from the Integrated Household Survey (IHS), 1994 and the available secondary information are used. Since the information on health and education in the survey data is not extensive, the scope of this paper is very limited. The main findings of the analysis are: public health care system is inadequately serving the needs of the poor especially in rural areas; and likewise the access of the poor to higher levels of education is limited. i Contents Abstract .......................................................... i I. Introduction .......................................................... 1 IH. Access to Health Care during the Transition ..................................................... 1 III. The Poor's Access to Education .......................................................... 11 IV. Conclusion ........................................................... 14 List of Tables: Table 1: Dispensaries and Hospital Beds, 1989-1993 ........................... ....................3 Table 2: Availability of Hospital Beds and the Incidence of TB ............. .....................4 Table 3: Probit Estimates of Children Being Enrolled in Schools ............ .................... 13 List of Figures: Figure 1: Sample Population 15-59 years of Age Reporting Sickness in Last 30 Days (Total Sample) ....................................................6 Figure 2: Sample Population 15-59 years of Age Reporting Sickness in Last 30 Days (Rural Sample) ....................................................6 Figure 3: Sample Population 15-59 years of Age Reporting Sickness in Last 30 Days (Urban Sample) ............................................................7 Figure 4: Of those Sick, Percent Seeking Care from a Private Doctor .......... ................7 Figure 5: Of those Sick, Percent Seeking Care from Public Providers (Total Sample) ....... 9 Figure 6: Distribution of Total and Health Expenditures, Total ............. .....................9 Figure 7: Distribution of Total and Health Expenditures, Rural ............. ..................... 10 Figure 8: Distribution of Total and Health Expenditures, Urban ............ ..................... 10 Figure 9: Level of School Completed of the Sample Population over 7 Years of Age ........ 12 Figure 10: Sample Population 7 Years and Older Currently in School by Age Group ... 12 Figure 11: Distribution of Subsidies for Education ................................................. 13 ii SOCIAL SECTORS IN TRANSMON: THE CASE OF ROMANIA I. Introduction 1.1 Ensuring access for the poor to basic social services-primary health care and primary and secondary education-is important to alleviate the immediate consequences of poverty. The need for efficient and assured delivery of social services is even more imperative during the period of transition when the poor are especially vulnerable to shocks from short-term, transition-induced pressures (such as a contraction of publicly provided social services). 1.2 Even prior to the transition (in 1990), Romania's health indicators were the lowest in Europe. The management of the health sector was such as to limit the sector's ability to deliver services to the population in general (World Bank, 1992)1. Both before and after the transition. health care services are financed exclusively by the central government, and delivered by the local health authorities. 1.3 In contrast to the health sector, the overall education system appears better organized and efficient. Enrollments are high and drop-out rates are low. However, educational administration, like the health administration, is very hierarchical, and under strict central control. The Ministry of Education regulates practically every aspect of the system-admissions, scholarships, examinations, specializations, curriculum design, etc. 1.4 This paper is very limited in its scope: it focuses exclusively on the access of the poor to social services in the wake of transition to a market economy. The paper is based primarily on the data generated by the Integrated Household Survey (IHS), 1994; recent secondary information at the district level has also been used. The IHS data has better coverage on the access to the health sector than the education sector; accordingly, our analysis of the former is more detailed than that of the latter. Access to health care is analyzed in the next Section, and access to education is analyzed in Section III. Conclusions and policy recommendations are stated briefly in Section IV. II. Access to Health Care during the Transition 2.1 Since the transition, the quantity and quality of social service provision have shown signs of deterioration. Over the period 1989-93, there has been a general deterioration in the supply of health services in the nural areas. Thus, while the number of dispensaries increased in urban Rornania by 19.2 percent, it decreased by 1.4 percent in mrual Romania (Table 1). Whereas the improvement in urban areas has been uniform, the deterioration in rural health care delivery has not been uniform across the country; some rural districts suffered more than others. The number of dispensaries in urban Romania increased in 40 out of 41 districts; the increase in 25 districts was IWorld Bank. 'Romania: Human Resources and the Transition to a Market Economy." A World Bank Country Study. over 20 percent. In rural areas, the number of dispensaries decreased in 20 out of 41 districts (Table 1). More an half of the districts registering a decrease in the number of dispensaries belonged to the South West and North West regions of Romania. 2.2 There has also been a dramatic contraction in the number of hospital beds. The decrease was much higher in rural (28.5 percent) than in urban areas (10.7 percent) (Table 1). In some districts, the decrease in beds in rual areas was over 50 percent. Even when corrected for differences in population size across districts, the decline in hospital bed availability is evident-40 of the 41 districts showed a decline in availability of hospital beds. (Table 2). 2.3 While the availability of dispensaries and beds decreased, the incidence of disease has increased. A particularly worrisome aspect is that the estimated number of cases of TB per 1,000,000 has sharply increased in most districts-only 2 districts out of the 41 show a decline in the estimated number of cases. (Table 2). 2 fTable 1: Dispensaries and Hospital Beds, 1989-1993 Number of Disuensaries Number of Hospital Beds Rural Urbani Rural Urban District 1989 1993 %change 1989 1993 % change 1989 1993 % change 1989 1993 % change Total 2918 2878 -1.4 2583 3078 19.2 28,219 177,932 -28.5 20,172 158,910 -10.7 Alba 73 72 -1.4 64 77 20.3 220 105 -52.3 3,711 3,261 -12.1 Arad 72 72 0.0 62 74 19.4 656 421 -35.8 4,124 3,699 -10.3 Arges 110 107 -2.7 50 65 30.0 1,714 1,203 -29.8 3,622 3,170 -12.5 Bacau 82 81 -1.2 70 94 34.3 425 294 -30.8 4,612 3,912 -15.2 Bihor 101 98 -3.0 77 99 28.6 810 645 -20.4 6,325 5,271 -16.7 Bistrita-Nasaud 56 56 0.0 23 38 65.2 180 100 -44.4 2,021 1,794 -11.2 Botosani 68 68 0.0 25 33 32.0 740 560 -24.3 3,727 3,414 -8.4 Brasov 48 51 6.3 105 117 11.4 356 330 -7.3 5,435 4,443 -18.3 Braila 44 41 -6.8 40 57 42.5 530 180 -66.0 2,810 2,741 -2.5 Buzau 89 86 -3.4 41 54 31.7 1,669 1,065 -36.2 2,406 2,299 -4.4 Caras-Severin 74 74 0.0 65 76 16.9 205 132 -35.6 3,481 2,904 -16.6 Calarasi 57 52 -8.8 21 31 47.6 884 405 -54.2 1,468 1,569 6.9 Cluj 80 80 0.0 119 148 24.4 433 285 -34.2 8,672 7,489 -13.6 Costanta 54 56 3.7 98 102 4.1 120 255 112.5 5,178 4,927 4.8 Covsna 38 38 0.0 23 25 8.7 121 110 -9.1 2,572 2,305 -10.4 Dimbovita 82 83 1.2 47 57 21.3 1,180 963 -18.4 2,902 2,694 -7.2 DoIj 99 100 1.0 78 77 -1.3 1,717 1,166 -32.1 4,958 4,400 -11.3 Galati 57 56 -1.8 90 106 17.8 600 490 -18.3 4,128 3,686 -10.7 Giurgiu 52 48 -7.7 18 25 38.9 1,236 975 -21.1 495 465 -6.1 Gorj 72 70 -2.8 41 55 34.1 856 480 -43.9 2,550 2,786 9.3 Harghila 49 49 0.0 62 69 11.3 690 645 -6.5 2,749 2,502 -9.0 Hunedoara 69 64 -7.2 69 99 43.5 1,231 860 -30.1 5,326 4,544 -14.7 lalomita 50 50 0.0 25 29 16.0 125 80 -36.0 1,397 1,214 -13.1 lasi 85 85 0.0 97 113 16.5 1,682 1,321 -21.5 7,770 7,229 -7.0 Maramures 67 66 -1.5 68 88 29.4 274 215 -21.5 4,740 4,266 -10.0 Mehedinti 63 62 -1.6 40 52 30.0 205 240 17.1 2,690 2,250 -16.4 Mures 95 93 -2.1 70 73 4.3 435 155 -64.4 6,359 5,461 -14.1 Neamt 77 76 -1.3 40 57 42.5 490 395 -19.4 3,960 3,265 -17.6 Olt 99 94 -5.1 35 52 48.6 589 220 -62.6 3,201 2,911 -9.1 Prahova 102 102 0.0 116 141 21.6 630 415 -34.1 6,449 5,635 -12.6 Salu-Mare 57 57 0.0 42 50 -19.0 350 220 -37.1 2,433 2,053 -15.6 Salaj 57 56 -1.8 23 29 26.1 190 11( -42.1 2,078 1,749 -15.8 Sibiu 57 53 -7.0 57 68 19.3 144 0 -100.0 4,233 3,906 -7.7 Suceava 99 99 0.0 65 85 30.8 380 275 -27.6 4,285 4,147 -3.2 Teleorman 85 84 -1.2 34 39 14.7 707 585 -17.3 2,956 2,694 -8.9 Timis 82 82 0.0 103 119 15.5 1,645 1,380 -16.1 6,594 6,295 -4.5 Tulcea 44 44 0.0 33 44 33.3 0 - 2,016 1,879 -6.8 Vaslui 71 71 0.0 31 42 35.5 880 543 -38.3 3,244 2,966 -8.6 Vilcea 81 80 -1.2 37 48 29.7 498 570 14.5 2,752 2,507 -8.9 Vrancea 61 61 0.0 27 33 22.2 575 360 -37.4 2,143 1,882 -12.2 Munic Bucuresti 60 61 1.7 352 338 -4.0 1,847 1,419 -23.2 25,360 22,326 -12.0 3 Table 2: Availability of Hospital Beds and the Incidence of TB Hospital Beds per Incidence of TB per 1,000,000 people 1,000,000 people District 1989 1993 1989 1993 Total 8.9 7.9 2.0 2.8 Alba 9.2 8.2 1.6 1.9 Arad 9.4 8.5 2.4 2.7 Arges 7.9 6.4 1.6 1.8 Bacau 6.9 5.7 2.1 2.9 Bihor 10.8 9.3 1.3 1.7 Bistrita-Nasaud 6.7 5.8 1.3 1.6 Botosani 9.6 8.6 1.7 2.4 Brasov 8.3 7.4 1.5 2.2 Braila 8.3 7.4 1.4 2.4 Buzau 7.8 6.5 1.7 2.2 Caras-Severin 9.0 8.1 2.3 2.8 Calarasi 6.7 5.8 3.7 5.0 Cluj 12.3 10.7 1.4 1.8 Costanta 7.2 7.0 2.4 4.2 Covsna 11.3 10.3 0.9 0.9 Dimbovita 7.2 6.5 2.4 4.0 Dolj 8.6 7.3 2.1 3.3 Galati 7.4 6.5 1.9 2.9 Giurgiu 5.3 4.7 2.3 3.4 Gorj 8.8 8.3 1.6 2.2 Harghita 9.5 9.0 0.7 1.0 Hunedoara 11.6 9.9 1.7 3.0 Ialomita 4.9 4.2 1.9 2.8 Iasi 11.7 10.5 2.4 3.6 Maramures 9.0 8.3 1.5 2.5 Mehedinti 8.8 7.5 3.7 4.6 Mures 10.9 9.2 1.4 2.4 Neamt 7.7 6.3 1.6 2.2 Olt 7.1 6.0 2.3 3.5 Prahova 8.1 6.9 1.8 2.1 Satu-Mare 6.7 5.7 2.4 3.1 Salaj 8.4 7.0 2.3 3.0 Sibiu 8.6 8.7 1.3 1.8 Suceava 6.7 6.3 1.7 2.1 Teleorman 7.3 6.8 2.2 3.6 Timis 11.4 11.1 2.8 4.5 Tulcea 7.3 7.0 3.0 2.6 Vaslui 8.8 7.6 2.0 2.9 Vilcea 7.6 7.0 1.8 2.5 Vrancea 6.9 5.7 1.8 2.4 Munic. Bucuresti 85.3 10.1 17.4 3.4 4 2.4 Given the decrease in physical availability and increase in the incidence of disease, it would be interesting to know how the poor have fared relative to the nonpoor in terms of access to health care. The data generated by the 1994 IHS are useful in shedding light on the incidence of morbidity and access to health care for the poor. First, the poorest households report a higher number of days of sickness, and a higher nunber of work days (productivity) lost due to sickness, than the richest 3 (Figure 1); such differences, however, are to be seen only in urban areas and not in rural areas (Figures 2 and 3). The proportion of the sick seeking private health care sharply increases with income class (Figure 4). It is not possible to disentangle the various factors behind this finding. It is possible that richer households may be opting for private care by choice, whereas poorer households seek private care because the public health care is not adequately reaching them. Also, a higher percentage of households seek private care in urban areas than in rural areas. 2 Poorest households are defined as those belonging to the bottom expenditure quintile. 3 Richest households are defined as those in the top expenditure quintile. 5 Fgure 1: Sample Poplatdon 15-59 years of Age feporting Sickness in Last 30 Days [Total Sanphle 25 20 15 10 poor"s 11 I111 IV Rkhest Averas IN-Days Sick 13Das*ot Source: Romarnla IHS R gure 2: Sampis Population 16-59 years of Age Reporting Sickness in Last 30 Days FtRwal Samplej 5 20 r., 15 * 10' r S~~~~~~~~~~ . ,. ..a.; '" Poorest II ll IV Richest Average | Days Sick Ei Days Lost Source: Romania IHS 6 Figure 3: Sanple Populthon 16-59 yews of Age Reportng Slcknies in Last 30 Days I Urban Sanpiel 25 20 15 10 Poorest II lii IV Richest Average IS Days Sick 0 Days Lost Souro: Romania IHS Figure 4: Of those Sick, Percent Seeing Care from a Private Doctor 10 8- a. 4 2 0 Poorest II IIl IV Richest Average Urban Rural 7 Soure: Romania IHS 2.5 Public health care is heavily subsidized in Romania, as in most other economies in transition. The inequality in the distribution of public subsidies on health cannot be directly measured owing to an inadequate data base, but can be indirectly assessed. Hospitals in Romania are better equipped, and receive as much as 89 percent of total public spending on health, whereas dispensaries receive only 11 percent. Available estimates suggest that a hospital bed per day costs twenty times more than a bed in a dispensary. The distribution of access by facility-dispensary, public clinic and hospital-for per capita expenditure quintiles is shown in Figure 5. The sick in the poorest households seek care largely from the (relatively ill-equipped) dispensaries, whereas the richest households receive a much higher proportion of care from public hospitals and clinics. About 25 percent of the sick in the rural households seek care at public clinics and hospitals as compared to over 60 percent of those sick in urban households. To the extent that over 90 percent of total public spending is on public clinics and hospitals, the observed pattern of utilization of facilities does point to the significant inequality in the distribution of public subsidies on health care. The subsidy is accruing disproportionately to the richest quintile. 2.6 How much do the rich and the poor spend on their health care, relative to their total expenditures? Figure 6 shows the distribution of household-level health expenditures, with total expenditures by deciles. Both distributions are similar, implying that the poor's spending on health is proportional to their share in total expenditure. Rural-urban differences are negligible. 8 Figure 5: Of Those Sick, Percent Seeking Care from Public Providers (Total Sample) 100 40 20 0 ~~~~~~~~~~~~~~~~~~S I 04-5 o 0 0 Dispensary 0 Public Clinic U Hospital Figure 6: Distibution of Total and Health Expenditures. Total 5100 U0 - / i f 1--_ ~~~~~~~~~~~~Health Expenditues| ZO 20_ Source: Romania IHS 0 20 40 60 80 100 Cumulative share of population 9 Figure 7: Distribudon of Total and Heat Expnditures, Ruwa 100 a 80 - - o 60 - . , |oTatal Expenditures -_______ 40______ - __._/__ ________ I HeaKth Expenditures 40) X2 20' - t; O - X { t Source: Romania IHS 0 0 20 40 60 80 100 Cumulative share of population Figure 8: Disribution of Tota and Health Expenditures, Urban 1 100 g 80 1 x a I 80 - -Total Expenditumes ____ _ 4_ _ / | ----Health Expenditures 4 40 - IS 20 Souroe: Romania IHS o 20 40 60 80 100 Cumulative share of populatfon - 0 2.7 The main finding is that the public health care system is inadequately serving the needs of the poorest groups especially in rural areas, and that public subsidies on health care are disproportionately accruing to the rich. mll. The Poor's Access to Education 3.1 Unfortunately, information on education is thin in the household survey. Even secondary data on expenditures by levels and with a mral/urban breakdown are unavailable. The World Bank Country Study (1992) identified the primary issue to be one of reforming the system such that it becomes responsive to the needs of a market economy. The Country Study noted that reforms were needed in five areas: pedagogy, curriculum, textbooks, examinations, and finance and management. The Household Survey does not shed light on any of the above areas of reform. The data permit us to answer a very limited question of inequality of access to different levels of education. 3.2 The proportion of those who have had no schooling is highest amongst the poorest deciles. The proportion with higher grades of education completed increases with the level of expenditure. (Figure 9). On average, in urban areas, the percentage of people who have completed higher grades is twice as large as in the rural areas. Moreover, a higher percentage of older children are in school in the richest two deciles than in the poorest decile (Figure 10). These findings are indicative of inequality of access to higher levels of education. 3.3 In general, public subsidies on primary and secondary education are more equally distributed than for post-secondary and higher grades of education (Figure 11). In this respect, Romania's position is similar to other countries in transition such as Poland and Vietnam. 3.4 In order to identify the household characteristics of families with children (of school age) who are more likely to be currently in school (the dependent variable), a probit equation was estimated. Table 3 presents results of the probit analysis. The coefficient for the per capita expenditure of the household is positive, suggesting that the probability of a child being in school increases with households' per capita expenditure; the coefficient is highly significant. Moreover, the higher the level of educational achievement of the head of the household, the higher the probability the child is in school. This coefficient is also highly significant. However, the coefficient of female headship is positive and statistically significant. This suggests that while female-headed households are disadvantaged with respect to receiving cash transfer4, children in female-headed households are not disadvantaged so far as school enrollments are concerned. The urban dumnmy variable is highly significant implying that the probability of a child located in urban areas attending school is high. In order to learn about possible gender bias in schooling, the sex of the child was also entered in the probit equation, but this was not statistically significant. Thus, household expenditure (income), parental education and urban residence are the dominant factors enhancing the probability of a child being in school. Children belonging to the poorest and least- educated heads of households are clearly at a disadvantage. 4Subbarao, Kalanidhi and Kalpana Mehra. 1995. Social Assistance and the Poor in Romania. 11 Rgure 9: Level of School Completed of the Sample Population over 7 Years of Age 100% 80% 60% 40% 20% Poorest II III IV Richest Average Urban Rural | None U Primary Q Secondary I I Secondary II, Higher Source: Romania IHS Figure 10: Sanple PopuLadon 7 Yeas and Older Currently in School by Age Group 100 80 - 80 40- 20- 0 Poorest II II IV Richest Average Urban Rural M17 - 11 yrs 0 12 - 15 yrs *16- 20yrs 021 - 25 yrsl 12 Source: Romania IHS Figure 11: Distribuiion of Subsidies for Education 100 No Schoot 60 * a i --> ,_-'/ , / i ............. --- ---- Secondary I _-s . - Secondary II _y3 .A -ExpenditExpendiuure c; .X 20 0 0 20 40 60 80 100 Cumuladve Share of Popuiatdon Table 3: Probit Estimates of Children Being Enrolled in Schools Asymptotic Marginal effect at Coefficient t-value mean of x Constant 0.172 2.97 0.098 Per Capita Consumption Expendiumre 0.000 6.16 0.000 Head's schooling (level) 0.133 13.53 0.101 Sex of head of household (f=1; m=0) 0.118 2.32 0.059 Rural/urban residence (urban=1; rural=0) 0.205 5.90 0.112 Region Southwest 0.032 0.68 0.016 Region Northwest 0.081 1.73 0.041 Region Northeast -0.011 -0.24 -0.005 Region Bucharest -0.259 -1.34 -0.129 Region Unclassified -0.063 -0.93 -0.032 Sex of child (female=0; male= 1) -0.008 -0.27 -0.004 N 11369 Log-likelihood -4030.9 Source: Romania IHS, 1994. 13 IV. Conclusions 4.1 In health, the major issue is rural-urban disparity in access to health care. The deterioration in health care provision in rural areas is adversely affecting the poor. Immediate steps are necessary to protect public expenditure on health in rural Romania. Moreover, the inequity in spending between public hospitals and dispensaries, and the generally regressive pattern of spending needs to be corrected. A comprehensive approach is required to bring about a pro-poor alignment of public expenditure in health, both regionally (rurallurban), and by levels of health care. 4.2 The distribution of health facility utilization shows important differences between the poor and the rich. Increasing modern public health facility utilization by poor households ought to be a priority in public policy. Considering the poor's inability to incur expenditure on private health care to the same extent as the rich, access to a publicly provided modern health facility is desirable to reduce the current substantive losses in work days due to sickness among the poor. 4.3 The issues in education are similar: the differences between the children of the rich and the poor, and rural-urban disparities. The inequality is particularly large at higher grades of education. The disadvantaged children belong to households in poverty and/or households with low parental educational attainments. A conscious effort needs to be made to increase access to education of children of poor households in both rural and urban areas. 14