Poverty, Social Services, and Safety Nets in Vietnam

Discussion Papers present results of country analysis or research that are circulated to encourage discussion and comment within the development community. To present these results with the least possible delay, the typescript of this paper has not been prepared in accordance with the procedures appropriate to formal printed texts, and the World Bank accepts no responsibility for errors. Some sources cited in this paper may be informal documents that are not readily available. The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s) and should not be attributed in any manner to the World Bank, to its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent. The World Bank does not guarantee the accuracy of the data included in this publication and accepts no responsibility whatsoever for any consequence of their use. The boundaries, colors, denominations, and other information shown on any map in this volume do not imply on the part of the World Bank Group any judgment on the legal status of any territory or the endorsement or acceptance of such boundaries

Poverty, Social Services, And Safety Nets In Vietnam

Nicholas Prescott

The World Bank

Washington, D.C.

Copyright © 1997

The International Bank for Reconstruction

and Development/THE WORLD BANK

1818 H Street, N.W

Washington, D.C 20433, U.S.A

All rights reserved

Manufactured in the United States of America

First printing October 1997

Discussion Papers present results of country analysis or research that are circulated to encourage discussion andcomment within the development community To present these results with the least possible delay, the typescript

of this paper has not been prepared in accordance with the procedures appropriate to formal printed texts, and theWorld Bank accepts no responsibility for errors Some sources cited in this paper may be informal documents thatare not readily available

The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s) andshould not be attributed in any manner to the World Bank, to its affiliated organizations, or to members of itsBoard of Executive Directors or the countries they represent The World Bank does not guarantee the accuracy ofthe data included in this publication and accepts no responsibility whatsoever for any consequence of their use.The boundaries, colors, denominations, and other information shown on any map in this volume do not imply onthe part of the World Bank Group any judgment on the legal status of any territory or the endorsement or

acceptance of such boundaries

The material in this publication is copyrighted Requests for permission to reproduce portions of it should be sent

to the Office of the Publisher at the address shown in the copyright notice above The World Bank encouragesdissemination of its work and will normally give permission promptly and, when the reproduction is for

noncommercial purposes, without asking a fee Permission to copy portions for classroom use is granted throughthe Copyright Clearance Center, Inc., Suite 910, 222 Rosewood Drive, Danvers, Massachusetts 01923, U.S.A

The complete backlist of publications from the World Bank is shown in the annual Index of Publications , which

contains an alphabetical title list with full ordering information The latest edition is available free of charge fromthe Distribution Unit, Office of the Publisher, The World Bank, 1818 H Street, N.W., Washington, D.C 20433,U.S.A., or from Publications, The World Bank, 66, avenue d'Iena, 75116 Paris, France

ISSN: 0259−210X

Nicholas Prescott is senior economist in the World Bank's East Asia and Pacific Region

Library of Congress Cataloging−in−Publication Data

1 Human services—Vietnam 2 Human services—Vietnam—Finance.

3 Public welfare—Vietnam 4 Poor—Services for—Vietnam

5 Vietnam—Social policy I Title II Series: World Bank

Who Benefits from Public Spending on Education? link

Targeting Public Expenditures on Health link

Annexes

1 Trends in School Enrollment, 1987−1993 link

2 Net Enrollment Rates by Income Quintile, 1993 link

3 Public and Private Financing of Education in Vietnam, 1993 link

4 Per Capita Subsidies for Education, 1993 link

5 Distribution of Subsidies for Education, 1993 link

6 Private Cost of Public Schooling, 1993 link

7 Affordability of Public Schooling, 1993 link

8 Trends in Utilization of Health Services, 1987−1993 link

9 Health Service Contact Rates by Provider and Quintile, 1993 link

10 Public and Private Financing of Health Services, 1993 link

11 Per Capita Subsidies for Health, 1993 link

12 Distribution of Subsidies for Health, 1993 link

13 Private Costs of Public Health Services, 1993 link

14 Affordability of Public Health Services, 1993 link

Tables

1 Expenditure on Pensions and Social Relief, 1994 link

2 Distribution of Government Transfer Payments by Quintile,

1993

link

Forewo rd

The decade of the 1990s marks Vietnam's transition to sustained and rapid growth following the economic

reforms initiated in 1989 Broad−based growth will generate new and diverse income−earning opportunities forthe poor in Vietnam, but some will be unable to take full advantage of them because of illiteracy, lack of skills,ill−health and malnutrition Ensuring access for the poor to basic social services—especially primary education,basic health care and family planning is doubly essential It alleviates the immediate consequences of poverty andattacks one of its principal causes Greater investment in human capital will ensure that the poor both gain fromand contribute to growth Not all the poor will benefit from these policies It may take a long time for some, such

as those living in remote regions, to participate fully and the old and disabled may never be able to do so Evenamong those who do benefit, some of the poor will remain acutely vulnerable to adverse shocks from short−termstress and natural calamities These groups need to be protected by a system of targeted transfers and safety nets.Vietnam has made impressive progress in providing widespread access to basic social services for a country withsuch a low level of income It has also developed an extensive system of social transfers and safety nets

Altogether public spending on these programs absorbs around one−third of the government's discretionary currentexpenditure But in the late 1980s and early 1990s the quantity and quality of social service provision showedsigns of deterioration Secondary school enrollments declined, and utilization of public sector health services alsofell These developments occurred in parallel with major changes in the provision and financing of social services

As part of Vietnam's ambitious program of structural reform, user fees were introduced for publicly−providededucation and health services and private sector provision was liberalized in both sectors in 1989 This paperexamines the changing role of the public sector in financing and provision of social services and safety nets, andassesses its efficiency in targeting the poor in the wake of these important policy reforms Much of the analysisdraws on distributional data generated by the Vietnam Living Standards Survey of households and rural

communes carded out in 1992−93

Introduction

Broad−based economic growth will generate new and diverse income−earning opportunities for the poor inVietnam, but some will be unable to take full advantage of them because of illiteracy, lack of skills, ill−health andmalnutrition Ensuring access for the poor to basic social services—especially primary education, basic healthcare and family planning is doubly essential It alleviates the immediate consequences of poverty and attacks one

of its principal causes Greater investment in human capital will ensure that the poor both gain from and

contribute to growth Not all the poor will benefit from these policies It may take a long time for some, such asthose living in remote regions, to participate fully and the old and disabled may never be able to do so Evenamong those who do benefit, some of the poor will remain acutely vulnerable to adverse shocks from short−termstress and natural calamities These groups need to be protected by a system of targeted transfers and safety nets

Vietnam has made impressive progress in providing widespread access to basic social services for a country withsuch a low level of income It has also developed an extensive system of social transfers and safety nets

Altogether public spending on these programs absorbs around one−third of the government's discretionary currentexpenditure But since the late 1980s the quantity and quality of social service provision has shown signs ofdeterioration Secondary school enrollments have declined sharply, and utilization of health services has alsofallen These developments have occurred in parallel with major changes in the provision and financing of socialservices As part of Vietnam's ambitious program of structural reform, user fees were introduced for

publicly−provided education and health services and private sector provision was liberalized in both sectors in1989

This paper focuses on the access of the poor to social services which has emerged in the wake of these importantreforms and its implications for the changing role of the public sector in ensuring that adequate access is

consolidated and sustained It begins with an assessment of performance and policy options in education and then

turns to the health sector where the impact of reform appears to have been more far−reaching Finally the paperexamines the large program of social transfers and safety nets and discusses its effectiveness and efficiency intargeting public resources to the poor Much of the paper is based on distributional data generated by the VietnamLiving Standards Survey of

households and rural communes carried out in 1992−93 Annexes 1 and 2 document the empirical results

underlying the analysis

Education

Falling School Enrollments

Vietnam has made impressive progress in expanding access to education during the last three decades It hasestablished a comprehensive network of educational institutions throughout the country and laid the foundationfor universal primary education by placing a primary school in every commune As a result, Vietnam has

achieved high levels of literacy and school enrollment relative to its per capita income level—but not relative toother East Asian countries To consolidate and sustain this progress in expanding educational opportunities, acloser watch is needed over the poor, especially over their ability to complete primary education, to receiveeducation of acceptable quality and to have equitable access to secondary education

However, there is evidence that the impressive gains achieved during the last 30 years are under serious threat Amajor deterioration in both schooling quantity and quality indicators has taken place during the last decade This

is evident in the marked decline in school enrollments which has occurred, most dramatically in secondary

schools, since the late 1980s (see Figure 1)

Figure 1:

Trends in School Enrollment, (in millions)

Enrollments in lower secondary schools have dropped sharply, falling by around 20 percent from a peak of 3.29million in 1987 to 2.71 million in 1990

And at senior secondary level, enrollments fell even more sharply by almost 50 percent from 0.93 million in 1987

to only 0.52 million in 1991 Since the secondary school age population was increasing during this period, thesedeclines in absolute student numbers represent an even sharper drop in secondary school enrollment rates

Secondary school enrollments have risen again somewhat in the early 1990s but have not yet recovered theirformer levels Similar declines have taken place at all levels of post secondary education Including studentsenrolled in technical vocational and secondary vocational schools together with higher education (universities andcolleges), overall post secondary enrollments fell by about 20 percent between 1987 and 1992 Only primaryeducation, which experienced a slight dip in enrollments between 1987 and 1989, appears to have escaped theoverall picture of quantitative decline There is also evidence of lower female enrollment at all levels of

schooling, especially at the secondary and tertiary levels

The explanation for this decline must lie in some change in the determinants of school enrollment Whetherchildren enroll in school is influenced by a multitude of factors reflecting family background, expected returns toeducation and the costs of access to school in the community Faster economic growth means more opportunitiesfor higher paying work and greater incentives to invest in human capital Those expected future returns, however,have to be balanced against the present costs of schooling, that is the out−of−pocket expenditures on educationincurred by families for public or private schooling The full private costs of education to the family includes notonly the schooling costs that are formally passed on in terms of official fees, but also the hidden costs of

unofficial parental contributions, learning materials, uniforms and transportation, plus the opportunity costs oftime associated with school attendance When costs are too high, the poor are less likely to continue in school.The private costs of schooling may well have risen enough to keep more children out of school A nationwidesystem of official tuition fees for public schools was introduced in September 1989 Parents are also expected topay contributions to parent teacher associations and to bear the cost of textbooks, clothing and food At the sametime, more young people may be choosing to take advantage of the job opportunities created by the restructuringand growth of the Vietnamese economy which have raised the opportunity cost of schooling These changes in thefactors determining the private costs of schooling are particularly likely to have affected poor families and theymay have borne the brunt of the decline in aggregate enrollments which Vietnam has experienced

Access By The Poor

Averaging across all income groups about 78 percent of children age 6 to 10, the official primary school agegroup, were enrolled in primary schools in 1993 Children begin to drop out of school in large numbers beyondthis age

group and only 36 percent of those age 11 to 14 were enrolled in lower secondary schools; however, a muchhigher proportion of these lower secondary school age children (69 percent) were attending some kind of school,indicating a considerable amount of overage enrollment in primary schools due to delayed entry and grade

repetition More children left school in the upper secondary age group, 15 to 17 years, leaving only 11 percent ofthose children enrolled in upper secondary schools; again a higher proportion (26 percent) were still attendingschool, suggesting substantial overage enrollment in lower secondary schools Very few youths—2 percent ofthose aged 18 to 24—were enrolled in any kind of post secondary education, broadly defined to include technicaland secondary vocational schools together with universities and colleges

How large is the enrollment gap between the poor and the better off? The aggregate enrollment rates mask largedifferences between income groups, with the exception of primary education where a fairly high enrollment rateimplies that most of the poor have access (see Figure 2)

Figure 2:

Net Enrollment Rates by Income Quintile, 1993 (% of target group enrolled in target level)

At primary level in 1993, the net enrollment rate among the poorest quintile was 68 percent, or about 10 percentlower than the rate in the top quintile But at junior secondary level the gap widens considerably with a threefolddifference between the poorest and the richest quintiles; only 19 percent of the poorest children are enrolled inlower secondary schools At upper secondary level the gap widens still further to a 15−fold difference; less than 2percent of the

poorest 15 to 17 years old are in upper secondary school The differentials become even wider at post secondarylevel, where no youths aged 18 to 24 in the poorest quintile are enrolled in any kind of post secondary educationcompared to 7 percent in the richest quintile

Changing Role Of The Public Sector

An important goal of the public sector in Vietnam's emerging market economy is to complement rather thansubstitute for the private sector by focusing on priority areas of government involvement where private marketscannot perform efficiently and equitably Basic education—especially primary but also lower secondary

levels—is a priority area for government involvement because it provides broad benefits to society as a whole andwould tend to be undersupplied, especially to the poor, without government subsidies to lower the costs of access

so that it is affordable Yet in Vietnam, despite public intervention in subsidizing basic education, net enrollmentrates among the poor still lag behind those achieved by the better off Indeed the quality of schooling received bythe poor may lag considerably more The scope of the government's involvement in basic education is constrained

by the fact that it spreads its resources thinly across all levels of the education system—including the higher levels

of education which are more costly per student, can serve fewer people and which are largely used by the betteroff This section puts into perspective the scope and nature of public sector involvement in education relative tothe role of the private sector

Provision Of Education

Up to now the Government has continued to assume virtually the entire responsibility for providing education.Since 1989 private schools have been tolerated although not actively encouraged However, private schools enrollonly a limited number of students Moreover, non−public schools are not entirely private Some of them aremixed schools whose operating costs are still subsidized by the Government Overall, the purely public schoolsenroll about 98 percent of all primary school students; this proportion is the same across the income distribution

On average, the public sector also enrolls 98 percent of all lower secondary students At upper secondary level theoverall public sector share declines slightly to 94 percent, falling from 100 percent of enrollments among thepoorest quintile to 93 percent among the richest At post secondary level the public sector enrolls nearly allstudents

Financing Of Education

While it dominates the provision of school places, the public sector actually finances much less than this

proportion of all schooling expenditures State budget subsidies for education and training are allocated broadlyaccording to the distribution of administrative responsibilities across the main tiers of government The centralgovernment budget for the Ministry of Education and Training subsidizes the higher education institutions

(universities and colleges) which it administers directly It also finances a variety of targeted education programs(a subset of the 28 national programs) which are

implemented directly by local governments on behalf of the center Local government budgets subsidize the lowerlevels of schooling, with the provincial tier being responsible for secondary schools together with post−secondarytechnical and vocational training, and the district governments responsible for subsidizing the operation of

primary schools Under these arrangements, the central government budget finances only one−quarter of the statebudget for education, while the remaining three−quarters are spent by the local governments Overall state budgetexpenditures for education and training amounted to 2,700 billion dong in 1993, or about 10 percent of

discretionary current expenditures (excluding interest payments) The largest share of the state budget—35

percent—is allocated to primary schools Another 29 percent of the state budget is spent on post−secondaryeducation Most of the remaining one−third of the budget goes to secondary schools and the targeted nationalprograms—14 percent to lower secondary schools, 5 percent to upper secondary schools and 10 percent to thetargeted programs

Expenditures by the private sector have now emerged as an important complement to state budget outlays at all

levels of education Estimates based on reported household expenditures from the VLSS suggest that total privatespending on education amounted to about 2,050 billion dong in 1993 Almost all of this—nearly 2,000 billiondong—was spent by students enrolled in public schools Spending on textbooks—an essential input to productiveschooling—absorbed around 480 billion of private outlays on public schooling Payment of official fees to publicschools cost students another 270 billion dong while informal parental contributions amounted to a further 190billion dong Private expenditures on public school uniforms totaled 290 billion dong and other school−relatedexpenditures (transport, food and lodging and some other expenses) added another 750 billion dong to the privatecosts of public schooling

Putting these public and private expenditures together (averaging across the 1992 and 1993 budgets to estimatepublic expenditures for the 1992−1993 school year) suggests that the state budget finances only 51 percent ofoverall education expenditures (see Figure 3) This aggregate figure masks considerable variation across levels.Private outlays turn out to be larger than public subsidies for all levels of schooling except for post−secondaryeducation Thus the state budget finances 48 percent of public primary schooling, but only 32 percent and 28percent of lower and upper secondary schooling respectively Only at post−secondary level does the role of thestate become dominant with the budget paying for 78 percent of total expenditures These figures highlight the

diminished role that the public sector now plays in the financing, as distinct from physical provision, of education.

On the one hand this means that the education sector has been successful in mobilizing a considerable volume ofprivate resources to finance schooling On the other hand it means that a variety of prices already play an

important role in rationing access to public schooling—the prices of official fees, unofficial contributions,

textbooks, uniforms, transport etc This factor is

especially likely to influence access by students from poor families and may limit the scope for further costrecovery to finance expanded access and better quality of education

Figure 3:

Public and Private Financing of Education in Vietnam, 1993 (in billions of dong)

Who Benefits From Public Spending On Education?

The differentials in enrollment rates suggests the need for spending more resources on education programs whichyield higher social returns and can benefit the poor more effectively: the key priority is closing the enrollment gapand raising the quality of basic education This can be achieved by directing more public spending to these

programs, either by spending more on the education sector or by redirecting the allocation of resources amongprograms within the education budget Public expenditure choices clearly play an important role in determiningthe effectiveness with which the education sector as a whole reaches the poor, and how efficiently it does so Inorder to target the poor effectively, the allocation of public expenditures needs to give priority to subsidizing thelower levels of education which the poor are more likely to use extensively But targeting the poor efficiently alsorequires that they use a large share of the subsidized programs so as to minimize leakages to the better off; thismeans encouraging the better off to switch out of the public sector altogether

Assessing how well public spending on education is targeted to the poor requires a profile of who uses publiclyprovided education, together with measures of the in−kind subsidy received by these users This analysis uses theVLSS data to generate the distribution of public school enrollments by per capita consumption, together withper−student subsidies estimated from the public

finance data The main elements of the utilization picture are already clear In general, school enrollment rates risewith the level of per capita consumption while public sector shares are fairly constant, so public school enrollmentrates also tend to be higher among the better off than the poor This enrollment gap widens considerably at higher

levels of education At the same time per student subsidies rise sharply: from around 80 thousand dong for

primary schools, 100 thousand at lower secondary, 170 thousand at upper secondary and 2,470 thousand at postsecondary schools Thus the small part of the population who are able to gain access to higher education, amongwhom the better off are over−represented, tend to receive a disproportionate share of the education budget

Another way of viewing the tradeoff is that subsidizing one better off student in post secondary education costs 30poor students who could be enrolled in primary school

Putting the per student subsidies together with the per capita public school enrollment rates generates the

distribution of per capita subsidies averaged across all persons in each quintile group (see Figure 4) Most of theeducation subsidies benefiting the poor are delivered through the primary school program In 1993 the per capitasubsidy for primary education averaged 10.9 thousand dong The subsidy per capita was highest for the poorergroups, averaging 11.9 thousand dong in the poorest quintile and falling off to 8.9 thousand dong in the topquintile The key factor driving this pro−poor bias of primary education is variations in age composition acrossquintiles, which more than offset the lower enrollment rates among the poor Since primary education is targeted

at young children, it turns out to be a particularly effective mechanism for channeling resources to poor

households who tend to be younger and have more children Primary school age children constitute nearly twice

as high a share of the population in the poorest quintile (18 percent) than in the top quintile (11 percent)

Figure 4:

Per Capita Subsidies for Education, 1993 (in thousands of dong)

The pro−poor bias of public spending on primary schools reverses at the post−primary levels of education,

becoming more unequal the higher the level of education Thus the subsidy received by the poorest quintileamounts to only 2 thousand dong for lower secondary education, falling off to 0.3 thousand at upper secondarylevel and dropping to zero at the post−secondary levels In contrast, per capita education subsidies accruing to therichest quintile are three times as high at lower secondary level, 13 times higher for upper secondary schools andhighest of a11—29 thousand dong—at the post−secondary levels Overall, the pro−rich bias of public spending onpost−primary education offsets the pro−poor bias of primary education Aggregating across all programs,

education subsidies become larger for the better off groups, rising to 38 thousand dong among the richest 20percent of the population The per capita subsidy was more than three times as high for the richest quintile as forthe poorest, which averaged 12 thousand dong

A useful way of summarizing the efficiency of targeting education subsidies to the poor is in terms of the

percentage shares they receive of the total subsidy A weak criterion for judging targeting efficiency is whetherthe poor receive a larger share of the education subsidy than their share of national consumption: if so, the

distribution of the subsidy is more progressive than the underlying distribution of personal consumption, and thesize of the per capita subsidy relative to per capita consumption is larger for the poor than the better off A

stronger test is whether the poor receive a larger share of the subsidy than their share of the national population:this means that the absolute size of the per capita subsidy is larger for the poor These alternative targeting criteriaare illustrated in Figure 5 which presents Lorenz curves for the distribution of the education subsidy comparedwith the distribution of personal consumption in 1993 The horizontal axis shows the cumulative percentage of thepopulation ranked by per capita consumption The vertical axis shows the cumulative percentages of the subsidyand national consumption received by these population groups

The Lorenz curve for the subsidy on all education programs shows that it is more evenly distributed than personalconsumption because it lies above the consumption distribution curve But it also shows that the overall educationsubsidy is only weakly pro−poor because it still lies below the diagonal 45 degree line indicating equal shares ofthe total subsidy The only line above the diagonal represents the strongly pro−poor primary education program:poorer individuals receive a larger share of the primary school subsidy than their share of the overall population.The lower secondary education subsidy is only weakly pro−poor while public spending at higher levels of

schooling is not at all pro−poor Not only do public subsidies on upper and post−secondary education favor thebetter off in absolute terms (their Lorenz curves all lie below the 45 degree diagonal), they are distributed moreunequally than the underlying distribution of personal consumption: the poor receive a smaller share of thesehigher level subsidies than their share of personal consumption

Figure 5:

Distribution of Subsidies For Education, 1993 (cumulative percentage)

Policy Instruments

The design of policy instruments to expand educational opportunities for the poor depends critically on

understanding the determinants of school enrollment Parents (or students) can be expected to invest in education

as long as the return from an additional year in school is greater than the private costs associated with that

additional year This means that public policy instruments aimed at raising enrollments among the poor must seek

to lower the costs and raise the benefits of education Here the private costs of schooling refer to all

out−of−pocket expenditures including fees, books and other schooling expenses, together with transport costs,and the opportunity costs of time spent in school rather than working Most of these costs are directly determined

by public expenditure policies—the volume of recurrent subsidies determines how much parents are asked to pay,and investment expenditures determine the density of school placement The benefit side is influenced by thequality of schooling which also depends directly on how much the government spends on subsidizing schoolinginputs such as trained teachers, textbooks and other classroom supplies

Increasing Availability

Vietnam has already succeeded in lowering and equalizing distance to primary schools across all income groups

by placing primary schools within every commune All communes sampled in the VLSS rural community surveyreported having at least one primary school Nearly the same level of coverage has been achieved with lowersecondary schools: 87 percent of rural children of lower secondary school age live in communes which have alower secondary school and this proportion is fairly constant across income groups For those children who didnot have lower secondary schools in their commune, distances average 4.7 km But distances facing the poorest

quintile are much higher, averaging 7.9 km compared to 3 km for the top quintile The density of school

placement drops off sharply at upper secondary level Only 10 percent of upper secondary school age youths havenearby access to school in their commune At this level availability tends to be concentrated among the better off;

15 percent of youths in the richest quintile have nearby access while only 8 percent of the poorest do Distancesfor those without access in the commune averages 7.5 km, but distances tend to be higher for youths with lowerincomes, averaging 8.7 km in the poorest quintile Overall then, these indicators suggest that distance is not themain deterrent to primary school enrollment by the poor However, at junior secondary level there remains asignificant number of poor children who face long distances which could easily cause them to drop out

Improving Affordability

As suggested by the large volume of out−of−pocket expenditures on education, families incur nontrivial costseven to enroll their children in subsidized public schools The average private cost of attending school varieswidely across the different levels of education For each child enrolled in a public primary school, families pay onaverage about 83 thousand dong per year despite the fact that official fees are zero Outlays nearly triple to 200thousand dong at lower secondary level, double again to 430 thousand in upper secondary schools, and nearlydouble again to around 790 thousand at post−secondary school The composition of these costs also varies withthe level of education (see Figure 6) Thus in public primary schools the average student pays 4 thousand dong onfees, 11 thousand for PTA contributions, 12 thousand on uniforms and another 24 thousand for

books—accounting for some three−quarters of total schooling costs At higher levels of education travel andsubsistence costs become more important components of private costs

Figure 6:

Private Cost of Public Schooling, 1993 (in thousands of dong)

Private costs also vary across income groups Typically poor students spend less than the better off at all levels ofschooling Thus, in public primary schools students from the poorest quintile pay 35 thousand dong while therichest students pay around 215 thousand Similarly, poor students in lower secondary schools pay about 74thousand dong, compared to around 360 thousand for those in the richest quintile To some extent these variations

reflect differences in the level of official fees and mandatory unofficial PTA contributions But the major factor isvariations in semi−discretionary expenditure on textbooks, uniforms, subsistence and other items Certainly, there

is no obvious pattern of exempting the poor from fees or other costs On average 83 percent of all students inpublic primary schools report paying no fees, which is fairly consistent with the official policy of charging zerofees for primary education, but this proportion does not vary much across the income distribution At the sametime, only 13 percent of public primary students report paying zero PTA contributions and this proportion is notmuch higher for the poor Almost everybody in primary school pays for books, whether poor or better off At thelower secondary level, very few students pay nothing for fees (14 percent), or PTA contributions (13 percent) orbooks (less than one percent), and these ratios are not significantly higher for the poorest quintile

The fact that students incur nontrivial costs to attend public schools suggests that the prices faced by some poorfamilies are so high that they are an unaffordable barrier to raising enrollment rates Indeed, according to theVOLS rural community survey, 40 percent of the school age population live in areas where excessive costs arecited as the main reason for non−attendance at primary schools; the same ratio is 52 percent for lower secondaryschools A rough measure of the affordability of education compares the cost of education relative to families'nonfood consumption For the poorest families—in the bottom 40 percent of the population—whose total

consumption expenditure is not enough to obtain even a minimally acceptable level of food consumption, thisrepresents an appropriate measure of how difficult it may be to finance education costs given what they arewilling to displace in order to spend on nonfood items The average cost/nonfood ratio rises sharply with the level

of education from 14 percent for primary education to 33 percent for lower secondary education, and 72 percentfor upper secondary education; costs at post−secondary levels exceed average nonfood expenditures Usingquintile−specific costs—which are much lower among the poorer groups—shows that the burden of paying foreducation is higher for the poor than for the better off (see Figure 7)

Figure 7:

Affordability of Public Schooling, 1993 (cost/nonfood expenditure ratio in percent)

For primary schooling, costs per student are 22 percent of nonfood consumption, nearly twice as high as for therichest quintile (12 percent) At lower secondary level the cost/nonfood ratio is 45 percent, more than twice asgreat as for the rich Upper secondary schooling costs as much as the entire nonfood budget of the poorest

quintile, three times as much as the relative costs facing the richest quintile

These cost ratios suggest that raising enrollment rates among the poor— even at primary but especially at lowersecondary levels of education—may be difficult to achieve without pricing reforms designed to lower the privatecosts of schooling for poor families.

A first step could be to eliminate official fees but, since these comprise only a small part of total costs, furtherprice reductions may need to be considered A second step would be to increase schooling subsidies to reduce theneed for parental PTA contributions A third step would be to provide free textbooks instead of charging for them

as present policy requires These pricing reforms would not have to be implemented nationwide Instead theycould be delivered through geographic price discrimination by targeting only schools in areas which have a highincidence of poverty

Improving Quality

Considerable scope exists to improve the quality of education available to the poor An obvious indicator of thequality gap is the wide variation across income levels in the amount of private expenditures on public schooling,much of which purchases quality−enhancing inputs such as textbooks and contributions for school maintenanceand teacher incentives

If average expenditures are taken as a crude index of the average quality of schooling, it would be unaffordablefor poor families The average costs of primary schooling would absorb more than 50 percent of per capita

nonfood expenditure by the poorest quintile, while the average cost of lower secondary schools would more thanexhaust it (see Figure 7) A more direct indicator of the quality gap is the availability of textbooks in schools.Data from the VLSS rural community survey indicate that more than 40% of the poorest 20% of the target schoolage groups live in communes where both primary and lower secondary schools reported having either no

textbooks or not enough; corresponding ratios for rural children in the richest quintile are around 20%

Health Services

Declining Utilization

Vietnam has developed a vast infrastructure of health facilities, achieving a density of service provision at alllevels—including hospitals and health centers— that far exceeds availability in most developing countries Forexample, Vietnam's ratio of commune health centers per million population is around 170, compared with 32 inIndonesia, 63 in China and 141 in Thailand Similarly, Vietnam's hospital bed ratio of one per 389 persons

compares favorably with one per 465 in China, 665 in Thailand and 1,743 in Indonesia However, beginning withreunification and accelerating during the 1980s, the health sector has come under pressure and some performanceindicators have deteriorated sharply

The symptoms of decline are seen in a marked reduction in the utilization of health services According to

Ministry of Health service statistics, the number of outpatient consultations has fallen in half since the late 1980s,from an annual rate of around 2.1 visits per capita in 1987 to 0.9 per capita in 1993 (see Figure 8) The inpatientadmission rate also dropped sharply during this period, falling from about 105 per 1,000 persons in 1987 to 68 per1,000 in 1990; since then inpatient utilization appears to have recovered but not to its former level While thereliability of these reported trends is in some doubt, given weaknesses in the official reporting system and itsinability to capture utilization of private sector services, they certainly suggest a picture of reversal in Vietnam'spast gains in delivering health services A key policy concern in the face of such a major deterioration is whetherthe poor have suffered disproportionately

Figure 8:

Trends in Utilization of Health Services, 1987−1993

Little is known about the causes underlying the apparent decline in service utilization One factor may be thedeterioration in quality of government health services resulting from the compression of public expenditures inthe late 1980s A large proportion of health facilities have become dilapidated to the point of being unusable forwant of equipment and medical supplies At the same time salaries of health personnel have declined in real termsleading to low morale and productivity Another factor may be an increase in the costs of access to health

services A system of official user fees was introduced for government health services, except for communehealth centers, in 1989 Private sector provision of curative services and pharmaceutical sales was liberalized atthe same time, allowing users to shift out of the public sector and take advantage of better quality, but morecostly, private providers

Access By The Poor

Overall utilization rates—broadly expressed in terms of the frequency of medical care contacts with public andprivate providers and self−medication—are somewhat lower among the poor In 1993, the utilization rate

averaged about 3.3 contacts per person per year for the whole population, ranging from 2.9 among the poorest 20percent to 3.4 in the richest quintile This indicates relatively broad

access to medical care of some kind for all income groups Very few people—only 2 percent of those who weresick in the last month—report seeking no treatment for their illness This figure is slightly higher for the poorestquintile—4 percent— than for the richest (0.7 percent) However the composition of utilization varies sharplybetween the poor and the better off

The most striking feature of the utilization pattern is the heavy reliance on self−medication instead of formalmedical care providers (see Figure 9)

Figure 9:

Health Service Contact Rates by Provider and Quintile, 1993 (percent of persons reporting ill last month)

On average, about two−thirds of those who are sick resort to self−treatment This proportion is significantlyhigher among the poor Around 70 percent of the poorest quintile choose self−medication when sick compared to

55 percent among the richest Another striking fact is that among the minority who choose treatment from formalservice providers, the private sector is more important than the public sector Overall, the proportion of the sickwho use private providers averages 19 percent compared to 15 percent for the public sector This is true whetherrich or poor Among the poorest quintile, more of the sick obtain treatment from the private sector −15

percent—than from the public sector (12 percent) These stylized facts lead to a third: the role of public sectorproviders is much less important in providing services to the poor than the better off While only 12 percent of the

sick poor use any kind of public sector provider, nearly twice as many—20 percent of the richest quintile—do so.Utilization patterns within the public and private sectors also vary considerably across income groups Amongpublic sector providers, commune health centers are the most important source of treatment for the poorest 20percent although they actually take care of only 6 percent of the poor when they are sick As incomes rise, peoplesubstitute away from commune health centers and rely more heavily on higher−quality hospital care Thus, only 3percent of the sick poor use hospital outpatient services compared to 12 percent in the richest quintile The

frequency of hospital inpatient utilization is much lower but also rises with income, from 2 percent to 3 percent.Looking at the private sector, paramedics are the main type of provider used by the sick poor (10 percent) whilethese are much less significant among the rich (6 percent) Instead as incomes rise the better off substitute awayfrom paramedics to higher−quality private doctors Only 4 percent of the sick poor use doctors, compared with 18percent of the richest quintile Summarizing these differential patterns in utilization, the key fact is that the poorhave much less access to high−quality providers—defined as public hospitals and private doctors—than the betteroff Overall, there is a threefold quality gap between the poor and better−off: the probability of seeing a

high−quality provider when sick increases from less than 10 percent among the poorest 20 percent of the

population to more than one−third in the richest quintile.

Changing Role Of The Public Sector

The priority function of the public sector in health is to provide preventive health care which produces widespreadbenefits to society as a whole and to ensure adequate access to basic health services for the poor who are lesslikely to be able to take advantage of services provided by the private sector By liberalizing the private sector andimposing user fees for government health services, Vietnam has taken important steps towards mobilizing privatesector resources so as to allow the public sector to focus on these priority areas of government involvement Thissection looks at the relative roles of the public and private sectors which have emerged since the introduction ofthese policy reforms in 1989

Provision Of Health Services

The public sector no longer monopolizes the delivery of health care, except for hospitals which remain whollyowned and operated by the government Hospital inpatient care is still 100 percent provided by the public sector.But the public sector is no longer the main provider of outpatient services, which are delivered by a wide range ofproviders including not only public hospitals and commune health centers but also private doctors and

paramedics, many of whom are public employees operating private practices Almost two−thirds of outpatientconsultations are now delivered by the private sector And drugs for self−medication without any formal

consultation, which

accounts for the bulk of all treatment contacts when people are ill, are obtained exclusively from the privatesector Overall then, the role of the public sector in providing health care has now become relatively minor—lessthan 20 percent of all medical treatment involves contact with any public sector provider

Financing Of Health Services

In terms of overall financing the public sector assumes an even smaller role than the proportion of medical carecontacts which it treats State budget expenditures for health are distributed between the central and local

governments in line with their administrative functions The central government budget for health subsidizescentral and branch hospitals, which are administered by the Ministry of Health and other ministries respectively,together with preventive health care, medical training and research Local government budgets subsidize lowerlevels of health care, with provincial governments responsible for provincial hospitals and the district level

subsidizing district hospitals and, in some cases, commune health centers Thus the central budget financesaround one−quarter of the state budget for health, while the rest is spent by the local governments Overall statebudget expenditures on health in 1993 amounted to 1,442 billion dong, or some 20 thousand dong per

capita—about half as much as the government spent on education No official breakdowns of budgetary

expenditure are available for local governments but estimates suggest that almost all of this budget—90

percent—is spent on curative care in hospitals Only a small amount, about 3 percent, is reported to be spent onpreventive health care An even smaller amount of the budget, 2 percent, is estimated to be spent on communehealth centers This reflects the fact that only one−third of commune health centers are subsidized by the statebudget, in which case the subsidy is for staff salaries, while the majority are self−financed at the commune levelwhich is not part of the state budget

Expenditures by the private sector are by far the largest source of financing for health care Estimates from theVLSS indicate that total household payments for medical care amounted to some 7,500 billion dong in 1993.Around 3,000 billion dong was spent on medical care contacts in which treatment was provided by the publicsector—twice as much as the total amount of public subsidies for health Only a small proportion of this amountwas spent on payment of fees to public providers Rather, most of it—2,800 billion dong—paid for drugs

associated with, but often not provided as part of, treatment in public facilities; the state budget allocates onlyabout 60 billion dong for drugs which are intended to be provided free of charge Even more—4,300 billiondong—was spent on drugs by people who sought treatment from private providers or who resorted to

self−medication These figures suggest that drug utilization dominates health expenditure in Vietnam, yet thepublic sector plays little direct role in providing or financing it

Taking the public and private sectors together, it is clear that the public sector role in financing health care overall

is small—the state budget paid for 16

percent of all health expenditures in 1993 (see Figure 10) Even for important public services, private paymentdominates Thus, public subsidies contribute only one−third of all spending associated with public hospitals, andless than 10 percent of all spending associated with utilization of commune health centers These estimates showthat the financing reforms introduced in 1989 have had a far−reaching effect, transforming the provision of healthcare into a largely private market This transition presents both an opportunity and a threat to the poor On the onehand, successful mobilization of private resources has enabled a much higher level of access to health care to besustained than would have been possible with continued reliance on the severely constrained level of budgetaryfinancing On the other hand, the complementary role of the public sector remains thinly spread across all levels

of curative and preventive health care, rather than narrowly targeted at priority areas of government involvement.This means that the poor face lower quality and higher prices for basic health care at public facilities than theywould if public resources were better targeted

Targeting Public Expenditures On Health

Reaching the poor effectively means giving priority in allocating public resources to those health programs whichthe poor are more likely to use extensively Targeting public expenditures efficiently requires that the poor alsouse a large fraction of the subsidized health programs This section assesses how well public spending on health istargeted to the poor using the profile of health service utilization by income group generated from the VLSS,together with estimates of the subsidy for different kinds of health service provided by the public sector Most butnot all public expenditures can be allocated to different income groups in this way—notably, the benefits ofpreventive health services cannot be assigned to individuals The main features of the utilization profile arealready obvious Commune health centers are used more extensively by the poor, while higher level hospitalservices are used more by the better off Public subsidies are also larger for the higher level services—rising from

3 thousand dong per visit at commune health centers, to 33 thousand dong per outpatient visit and 118 thousanddong per inpatient stay in hospitals In fact these average

Figure 10:

Public and Private Financing of Health (in billions of dong)

figures for hospitals mask a much steeper gradient in the unit subsidy across different levels of the hospital

system For example, an inpatient stay at a district hospital costs only 55 thousand dong, compared with 137thousand dong at provincial hospitals and 414 thousand dong at central level While the better off are more likely

to use the more expensive hospitals located in provincial capitals this cannot be observed directly in the VLSSutilization data

Merging the estimated subsidies together with the utilization rates gives the distribution of per capita subsidiesaveraged across all persons in each quintile (see Figure 11) Not surprisingly since hospitals dominate public

expenditure, most of the health subsidies accruing to the poor are delivered through the hospital system,

especially through inpatient care In 1993, the per capita subsidy for hospital inpatient care averaged 11 thousanddong The subsidy was lowest among the poor, averaging 7 thousand dong in the poorest quintile and doubling to

14 thousand dong among the richest Per capita subsidies for hospital outpatient care increase more sharply asincome rises, averaging 3 thousand dong for the poorest quintile and more than quadrupling to 14 thousand dongamong the richest Of course these figures may greatly underestimate the pro−rich bias of hospital spending sincethey do not take into account the likelihood that the poor are more likely than the better off to use cheaper districthospitals Moreover, this bias is only slightly offset by the pro−poor distribution of state budget spending oncommune health centers Per capita subsidies at this level of the health system are very low in absolute terms butare higher for the poor, averaging 0.4 thousand dong in the bottom quintile and falling off to 0.2 thousand for thetop quintile Aggregating across all health sector programs, health subsidies are larger for the better off, reaching

28 thousand dong per capita for the richest quintile or more than double the amount spent on the poor, whichaveraged 11 thousand dong In other words public subsidies for health reach the poor less effectively than theybenefit the better off This simply reflects the fact that most of the budget is allocated to public

programs—hospitals—which are not efficiently targeted to the poor

The efficiency of targeting health subsidies to the poor can be evaluated in terms of the percentage shares theyreceive of the total health subsidy Subsidies are strongly pro−poor if the poor receive a larger share of the

subsidy than their share of the national population: in this case per capita subsidies are higher for the poor

Subsidies are only weakly pro−poor if they receive a larger share of the subsidy than their share of nationalconsumption: in this case the ratio of per capita subsidies to personal consumption is higher for the poor Thesetargeting criteria are applied in Figure 12 which compares the cumulative distribution of health subsidies andpersonal consumption

Figure 11:

Per Capita Subsidies for Health, 1993 (in thouands of dong)

Figure 12:

Distribution of Subsidies for Health, 1993 (cumulative percentage)

Only the very small budgetary subsidy for commune health centers is strongly pro−poor: its distribution liesabove the 45 degree line representing equal shares—with almost 50 percent of spending accruing to the poorer 40percent of the population The distributions of hospital spending for inpatients and outpatients are only weaklypro−poor: 30 percent and 23 percent of these respectively go to the poorer 40 percent of the population who haveonly 18 percent of aggregate consumption Taken together, all health spending does appear to be at least weaklypro−poor, although this conclusion is almost certainly exaggerated by the inability to differentiate the levels ofhospital used by different income groups using the VLSS data

Policy Instruments

Better targeting of public spending to improve access of the poor to appropriate health care depends on three mainpolicy instruments—reducing barriers to utilization imposed by distance and the user costs associated with servicefees and drugs, and improving the quality of care available for those who do gain access This section examinesthe status of these policy instruments as they affect the poor, drawing on data from the VLSS

Lowering Distance

Vietnam has already achieved remarkably widespread coverage with basic health facilities—94 percent of ruralresidents responding to the VLSS community survey reported having a commune health center available in theircommune Such coverage implies good access by the poor, and indeed there is almost no difference in reportedavailability between the poor and the better off However, there may be a need for targeted investments to

improve availability in the minority of rural areas where no commune health centers are available The poor who

do not now have nearby access have to travel on average about 8 kilometers to reach a health center, which takesabout two hours—about twice as much as rural people from the richest quintile who do not have immediateaccess Not surprisingly, access to hospitals is more limited and tends to improve as incomes increase Only 6percent of the poorest 20 percent in rural areas live in communes with a hospital available, compared with 12percent of people in the richest quintile Distances to the nearest hospital for those without average more than 11kilometers for the poorest quintile

Reducing User Costs

The large amount of out−of−pocket expenditures on health care shows that families pay significant user costseven to use subsidized public facilities User costs vary widely across different levels of the health system

Expenditures on fees and drugs average 31 thousand dong per visit to a commune health center despite the factthat official fees are zero Average outlays more than double to 77 thousand for a hospital outpatient visit, andnearly triple to 210 thousand dong per hospital inpatient stay Typically fees account for only a small proportion,around 10 percent, of out−of−pocket expenditures—most of which go for drugs

User costs also vary across income groups (see Figure 13) Typically the poor spend less than the better off at alllevels of the health system At commune health centers, users from the poorest quintile pay around 13 thousanddong, while the richest patients pay some 53 thousand dong Similarly, poor outpatients pay 43 thousand dong athospitals, compared to about 100 thousand dong for those in the richest quintile The private costs of a hospitalinpatient stay average 74 thousand for the poor, rising to 210 thousand for the rich Fees at commune healthcenters are low for everybody, averaging 0.2 thousand dong, so at this level the variations across income groupsreflect differences in discretionary spending on drugs In hospitals, fee payments vary more, but this probablyreflects

differences in the level of hospital used by the different income groups In any case, because fees are a smallproportion of total user costs the main source of variation is again private spending on drugs which are typicallynot subsidized by the hospital system

In general there does not seem to be a systematic effort to protect the poor by exempting them from paying forfees and drugs, except to a limited extent at commune health centers While most people do not pay the low feescharged at commune health centers, this ratio is higher for the poor (91 percent) than for the better off (68

percent) On the other hand most people do have to pay the much higher cost of drugs, with only 16 percent of thepoor paying nothing for drugs compared to none of the better off users Within the hospital system, a significantproportion of users still pay nothing for fees but most people pay for drugs—and these ratios are not much

different whether patients are poor or rich

Figure 13:

Private Costs Of Public Health Services, 1993 (in thousands of dong)

The magnitude of out−of−pocket costs facing users of publicly provided health services suggests that they may be

a barrier to access by the poor A rough test of affordability is to compare the expected cost of utilization with thelevel of nonfood consumption expenditure (see Figure 14) On average for Vietnam, this cost/nonfood ratio riseswith the level of service, from 3 percent at commune health centers, to 6 percent for hospital outpatients and 19percent for hospital inpatients More importantly, even using quintile−specific costs—which are lower

for the poor because they cannot afford to pay as much for drugs—the burden of paying for health care is higherfor the poor than the better off A visit to the commune health center costs the poor the equivalent of 8 percent oftheir nonfood consumption, more than twice as high as the for the richest quintile (3 percent) Hospital outpatientvisits are much more expensive for the poor, costing around 26 percent of their nonfood consumption—more thanfour times as much as for the rich For inpatient admissions, the cost/nonfood ratio for the poor becomes 45percent, more than twice as great as the relative costs facing the richest quintile

Figure 14:

Affordability of Public Health Services, 1993 (user charges as percent of per capita nonfood expenditures)

If these cost ratios are a deterrent to utilization then improving access for the poor may require selective pricereductions for priority services, compensated by increased subsidies from the budget Implementing this strategywould call for a pricing policy that consciously differentiates prices by the income class of users In some cases,especially for basic care delivered by commune health centers and district hospitals, it may be appropriate tolower existing charges for fees and drugs to more affordable levels so as to encourage greater participation by thepoor In other cases, such as high level hospital care, selective price increases targeted at better off users of publicservices may be appropriate to reduce their absorption of public subsidies In this way, public subsidies could bebetter targeted instead of distributed indiscriminately A policy of selective price discrimination could be

implemented in several ways: targeting poor individuals, by strengthening the present system of certification forprice exemptions;

targeting poor districts, identified by household survey data; self−selection, by charging lower prices at facilitiesmore likely to be used by the poor; and targeting indicators of income status, such as charging higher prices tobetter off civil servants with health insurance

Another way of improving affordability is for the public sector to lower the cost of drugs which are the majorexpenditure faced by use of the health care system This could be done by selling cost−effective essential drugsthrough the health system instead of leaving patients with little choice except the private sector to purchase drugs.While the private sector is filling an important need, private drug distribution appears to be largely unregulatedand highly priced Improving the public drug supply system to rationalize drug management is a high priority

Improving Quality

Considerable scope exists to improve the quality of publicly−provided health services available to the poor—byspending more on better trained staff and an adequate supply of affordable drugs An indirect indicator of thequality gap is the large difference in private expenditures on health care across income groups for the same level

of service The better off pay more and buy better quality, especially for drugs Another indicator is the quality of

medical attention received by the poor who use government facilities The poor are more likely to use communehealth centers when sick, but the probability of being treated there by a trained doctor is less than 10 percent Bycontrast, the better off are far more likely to get treatment in hospitals where the likelihood of consulting a traineddoctor is over 90 percent This situation reflects the fact that financing of commune health centers falls outside thestate budget except in the minority of cases—around 30 percent—where district governments have enough

resources to subsidize salaries Recognizing the poor quality of health manpower available at the periphery inpoorer areas, the Government has recently issued a decision (No 58) to bring salaries of commune health workers

on to the state budget With a planned staffing norm in the range of 3−5 health workers per commune facility, theincremental costs are estimated at some 200 billion dong per year Under anticipated budget constraints,

implementation is expected to be phased in gradually, with incremental budgetary spending of around 40 billionper year

Transfers And Safety Nets

Structure Of Social Protection

Vietnam has a very extensive program of social protection funded directly through the government budget andadministered jointly by the Ministry of Labor, Invalids and Social Affairs (MOLISA) and the Vietnam

Confederation of Labor (VCL) In 1992 outlays on social protection—classified as ''pensions and social relief' inthe government budget—amounted to 2,370 billion dong or 15 percent of

relief" in the government budget—amounted to 2,370 billion dong or 15 percent of discretionary current

expenditures (i.e net of interest payments)—almost as much as education and health combined (10 percent and 7percent respectively) This budget has more than doubled to 5,074 billion dong in 1994 (see Table 1) Theseoutlays are spent on three distinct programs which together cover a large segment of the population: (a) socialsecurity for public sector employees—including civil servants and employees of state enterprises; (b) allowancesfor some 400,000 handicapped war veterans and one million martyrs' families; and (c) social relief measurestargeted at victims of sudden natural disasters, starvation and "social evils" By far the largest share of the

budget—82%—is spent on pensions and disability payments for government workers Another 7% finances warvictims, leaving 11% of the budget for social relief

Table 1:

Expenditure on Pensions and Social Relief, 1994

Budget(billions)

Recipients(millions)

Expenditure perrecipient(thousands)

The social security program is long−standing, dating back to 1947 Originally social security was restricted to

public sector employees, but in 1993 Decree 43 CP expanded coverage to include private enterprises, joint

ventures and workers in Economic Processing Zones Most—about two−thirds—of social security spending paysfor pensions, with the rest allocated to disability payments In theory, these benefits are funded by a 15 percentpayroll tax, of which 5 percent is retained by VCL to finance the short−term benefits which it administers

(sickness, maternity and employment injury) and 10 percent goes to the Ministry of Finance to pay for the

long−term benefits administered by MOLISA (old−age pensions, disability and survivors' benefits) In practice,however, weak revenue collection from state enterprises has left the Ministry of Finance with an effective

contribution rate of only around 5 percent of the eligible payroll This represents a large shortfall from the

contribution rate actually needed to fully fund existing pension obligations, which is estimated at 13 percent Thusmore than half of the public pension program is subsidized out of general revenues collected from the entirepopulation even though it covers only a part of the population Recognizing

plans to restructure the financing of social security spending by shifting it gradually off the state budget to anewly created and independent social security institution, the Vietnam Social Security Organization These planswere announced in Government Decision No 43, issued on 22 June 1993 and were scheduled to be implementedearly in 1994 Under these new arrangements state budget subsidies for social security are expected to decline, butwill nevertheless is expected to remain relatively high at some 30% of total funding

The social relief program protects those who are not insured by the formal social security program and comprises

three main elements which are largely financed out of local government budgets First, regular relief targeted atthree groups—the lonely elderly (such as widows), orphans and the handicapped Eligible beneficiaries areidentified by local authorities based on targeting criteria set out in Decree 167 issued in 1994, with numbersdepending on the local budget available Up to 140,000 recipients are eligible for a monthly subsidy of 24,000dong in 1994 The second element is emergency relief targeted at victims of natural calamities, which often resultfrom typhoons in the central area, flooding due to heavy rains in mountainous areas and periodic river floods inthe delta areas In 1993 some 2 million persons received funds under the emergency relief program Beginning in

1994 every province budgets in the range of 2−10 billion dong for emergency relief A third element of theprogram is temporary starvation relief , targeted at the very poor who suffer periodic "between crops" starvation.This affects 3−5 million people per year A new element addressing "social evils" has recently been added to thesocial relief program, focusing on AIDS−related prostitution and drug abuse, with funding of some 20 billiondong and 50 billion dong respectively in 1994

Who Benefits From Social Protection?

The large amount of the state budget claimed by expenditures on social protection imposes a significant constraint

on how much of the budget can be allocated to other sectors Thus it is important to examine how effectively andefficiently it reaches the poor relative to expenditure programs in other sectors such as education and health.Some insight into the targeting of government transfer payments can be obtained from the VLSS (Table 2)

Table 2:

Distribution of Government Transfer Payments by Quintile, 1993

Beneficiaries (in millions)/a

Pensions & disability 1.30 1.53 2.23 2.25 2.39 9.71 2.94 6.77

Other social subsidies 1.79 1.35 1.40 1.21 1.32 7.07 1.67 5.40

All transfers 2.86 2.70 3.17 3.22 3.16 15.10 3.87 11.23

Beneficiaries as % of population

Transfers per Beneficiary (in thousands of dong)

Total Transfers (in billion dong)

Pensions & disability 158 263 440 538 885 2,284 1,006 1,278

Share of Total Transfers (%)

Transfers per Capita (thousands of dong)

Transfers as % of per Capita Consumption Net of Transfers

relief) Table 2 gives the results Overall, these transfers are estimated to benefit some 15 million people, or 21percent of the population, living in households which received one or both kinds of transfers These include anestimated 10 million

beneficiaries of spending on pensions and disability, and another 7 million beneficiaries of other social subsidypayments (some of whom overlap and receive both kinds of transfer) The aggregate value of government

transfers is estimated at some 2,500 billion dong—quite consistent with outlays reported in the state

budget—most of which goes to spending on pensions and disability (2,300 billion dong)

Transfers per beneficiary are much larger for pensions and disability, averaging about 235 thousand dong per yearcompared to 35 thousand dong for other social subsidies The amount of the transfer per beneficiary also risessharply across income groups Pension and disability payments average only 158 thousand dong per year perbeneficiary in the poorest 20 percent of the population, increasing threefold to 370 thousand dong per beneficiaryfrom the richest quintile The same is true for other social subsidies, which rise from 18 thousand in the bottomquintile to 50 thousand in the top quintile The proportion of people in each quintile who benefit from the pensionand disability payments also increases with income, rising from 9 percent of the poorest to 17 percent of therichest This pattern is reversed for the other social subsidies—13 percent of the poorest and 9 percent of therichest are beneficiaries of these payments

This profile of who benefits from government spending on social transfers suggests that overall it is not pro−poor.Developing a stronger social safety net for the rural poor implies the need for better targeting of these

expenditures The poorest 20 percent of the population benefits from only 7 percent of national expenditures onpension and disability payments, and 13 percent of spending on other social subsidies—or 8 percent of combinedexpenditure on both programs Only their share of spending on "other social subsidies" exceeds the 8 percent ofnational consumption which accrues to the poorest quintile Instead most government expenditures on socialtransfers—38 percent—goes to the richest 20 percent of the population This reflects the fact that most of it—over80%—is spent on pension and disability payments, which pays more per beneficiary, most of whom are

concentrated in the higher income groups As such the transfer programs are comparable to the education sectorwhich also allocates 38 percent of spending to beneficiaries in the richest quintile; health appears to be lesspro−rich, with 29 percent of public subsidies absorbed by the richest group Improving the poverty alleviationeffect of the large expenditures on social protection will require a significant reallocation of the budget frompensions to social relief

Education Data

Tables

1 Trends in School Enrollments, 1987−93 link

2 Target School age Groups by Quintile, 1992−93 link

3 School Enrollments by Quintile, 1992−93 link

4 School Enrollment Rates by Expenditure Quintile, 1993 link

5 Public and Private Sector Shares of School Enrollment, 1992−93 link

6 Average Out of Pocket Expenditure on Fees Per Student, 1993 link

7 Average Out of Pocket Expenditure on PTA Contributions Per

Student, 1993

link

8 Average Out of Pocket Expenditures on Uniforms Per Student,

14 Aggregate Private Expenditure on Education, 1993 link

15 Aggregate Public Expenditure on Education, 1992−93 link

16 Sources and Uses of Education Expenditure, 1992−93 link

17 Public and Private Shares of Education Financing, 1992−93 link

18 Public Subsidies and Cost Recovery for Public Education,

21 Efficiency of Targeting Subsidies for Education, 1992−93 link

22 Lorenz Distributions of Public Subsidies for Education,

24 Official Fee Structure for Grades VI through XII link

25 Official Criteria for Fee Discounts in Grades VI through XII link

26 Official Criteria for Fee Discounts in Post Secondary Schools link

27 Proportions of Students Who Pay Zero in Public Schools, 1993 link

28 Average Nonzero Expenditures for Students in Public Schools,

1993

link

29 Affordability of Public Schooling, 1993 link

30 Average Price of Schooling in Rural Areas link

31 Access to Scholarships for Public Schooling link

32 Access to Public Schools While Enrolled, 1993 link

33 Quality of Schooling in Rural Areas link

34 Most Important Reasons for not Attending School in Rural

Target School age Groups by Quintille, 1992−93

Memo:

Population