KPolicy, Research, and External Affalrs 7 WORKING PAPERS World Development Report Office of the Vice President, Development Economics, and Country Economics Department The World Bank September 1990 WPS 507 Background paper for the 1990 World Development Repoin Poverty in Poland, Hungary, and Yugoslavia ion the Years of Crisis, 1978-87 Branko Milanovic The deep economic crisis in Eastern Europe in the 1980s substan- tially increased the number of people living below the poverty line. Before the crisis, most of the poor lived in rural areas. Now most of the poor (as many as 70 percent in Poland) live in cities. The Policy, Resarch, and External Affairs Complex distributes PRE Working Papers to disseninate thefindingp of wozk in progpss and to encounrge the exchange of ideas amnong Bank staff and all others intersted in developrnent issfes. These papers cary the names of the authors, efilect only their viws, and should be used ad cited accordingly. The findings, intrpretations, and conclusions are the authors' own. They should not be auributed to the World Bank, its Board of Dirwectos. its mnanaganeent, or any of itt member counries. - Policy, Research, and External Affairs| Iord Development Report A WPS 507 This paper- a product of the Socialist Economics Reform Unit, Country Economics Department, and the first such Bank study done for Eastem Europe - was prepared as a background paper for the World Development Repon on poverty. It is part of a larger PRE effort to understand the factors that contribute to poveny. Copies of t-he paper are available free from the World Bank, 1818 H Street NW, Washington DC 20433. Please contact Angelica Bretana, room N6-025, extension 37176 (30 pages with tables). The deep economic crisis in Eastern Europe The distribution of poverty changed in all between 1978 and 1987 greatly affected average three countries. Urban poverty became domi- incomes and increased the proportion of people nant, as the economic condition of state sector living below the poverty line. workers - manual and nonmanual - deterio- rated much more than that of agricultural and The situation deteriorated most sharply in mixed (rural-urban) households. Before the Poland, where declining incomes caused the crisis, most of the poor lived in rural areas. Now percentage of poor people to increase from less most of the poor (as many as 70 percent in than 10 percent of the population (before the Poland) live in cities. crisis) to more than 20 percent. In Yugoslavia the proportion of poor people increased from 17 Increases in poverty are explained entirely percent to 25 percent. In Hungary poverty by declining incomes; overall, income distribu- remained at about the same level as before the tion did not change and in some cases even crisis (less than 15 percent). "improved." But when such redistribution did occur, it was insufficient to offset the impact of declining incomes. The PRE Working Paper Series disseminates dhe Findings of work under way in thc Bank's Policy, Research. and Extemal AffairsComplex. Anobjcctiveofthcscrics is to gel these findings out quickly, even ifpresentations are less than fullypolished. The findings, interprctations, and conclusions in these papers do not nccessarily represent official Bank policy. Produced by thc IRE Disscmination Centcr Poverty in Poland, Hungary, and Yugoslavia in the Years of Crisis, 1978-87 by Branko Milanovic Table of Contents I. Macroeconomic Environment I II. Impact of the Crisis on Population Incomes 5 III. Comparison of Subsisten*e Minimum Lines 7 IV. The Size and Composition ot Poverty, 1978-87 11 V. The ("Canging Composition of Poverty 16 VI. Poverty, Income, and Redistribution 18 VII. Conclusions 21 Appendix 1- Poverty Lines 23 Appendix 2 - Sources of Data for Household Incomes 24 References 26 End Notes 28 POVERTY IN EASTERN EUROPE IN THE YEARS OF CRISIS: POLAND, HUNGARY ANO YUGOSLAVIA Branko Milanovic *. Macroeconomic Environment The purpose of this paper is to study a subject that has so far received little attention: poverty in Eastern Europe in the last decade. While decline In growth rates, debt problems, worsening economic efficiency, or widening technological gap with respect to the West, have been the subject of numerous recent studies,2 knowledge of how the crisis affected population's standard of living, and in particular what was Its Impact on poverty, Is fairly scant. Most of It is of anecdotal or circumstantial character. Our aim is, therefore, to chart evolution in poverty in the last ten years using headcount index of poverty, to see how composition of the poor (in terms of social groups or between urban and rural poor) has changed, and to try to relate changes in poverty to some macroeconomic Indicators. Before we proceed with analysis of poverty per se, it Is necessary to situate these developments in their macroeconomic context. In the three East European countries considered here, Poland, Hungary and Yugoslavia, economic crisis started at about the same time, In the late 1970's. The origin of the crisis lay in low efflciency in the use of capital, low level of motivation and productivity of the labor force, and pervasive interference of the state in economic decision-making. The underlying unfavorable economic trends were masked in the 1970's thanks to a substantial inflow of foreign capital. As the period of relatively easy and cheap borrowing drew to a close, the low quality of economic performance became more apparent. The second oil shock in 1979 and Increase In real Interest rates in the early 1980's (particularly since the countries tried to postpone the inevitable by increasing their variable rate short-term debt) were final blows for the economies already operating under the strain. By 2 1981, Poland, Hungary and Yugoslavia were In deep crisis. It is not very difficult to pinpoint exactly to the year of the o, et of the crisis. For Poland, It Is 1978, when Polish GDP reached its highest value (which It would regain only 10 years later) and when the economy started on Its fast downward plunge that would reduce gross domestic product by 20 percent in four years, thus recording the sharpest decline In post-War Europe. For Hungary the first full crisis year was 1979 when the growth rate decelerated significantly; since then It has failed to recover to levels It routinely reached during the 1970's. For Yugoslavia, the first crisis year was 1980: the beginning of the crisis was postponed by a year mostly due to a surge of short-term borrowing In 1979, which allowed the country to finance a current account deficit equal to 6 percent of GDP. Table 1 shows, in a cursory form, the outlines of the crisis. The first three lines detail the severity of stagnation. Outputs per capita, after growing between 5 and 7 percent p.a. throughout the seventies, Increased between -0.6 and less than 2 percent p.a. in the 1980's.4 Per capita consumption was practically stagnant In Poland and Yugoslavia, while In Hungary It grew by 2.6 percent p.a.--one-third slower than In the 1970's. Real wages took most of the brunt of adjustment. In Poland, after registering almost 6 percent average annual growth in the 1970's, real wages declined in the next period, on average, by 3.7 percent p.a., ending in 1987 some 30 percent below the peak year of 1978. The same was true for Yugoslavia where the entire gain of the 1970's was wiped out In the next decade. Hungary alone succeeded in preserving real wage level constant. 3 Table I - Some Macroeconomic Indicators of Crisis Poland Hunsary Yugoslavia I II I II I II 1.GDP per capita +7.1 -0.6 +5.9 +1.9 +5.1 +0.9 2.Consumption p.c. +6.5 -0.2 +3.7 +2.6 +4.5 +0.8 3.Real wages +5.6 -3.7 +2.9 +0.1 +2.1 -2.2 4-CA/GDP -5.0 -3.5 -4.8 -2.5 -1.7 +0.2 5.Investment/GDP 33.6 25.3 31.6 25.6 32.1 24.1 _______________________ Notes: For GDP, personal consumption and real wages average yearly per capita growth rate for the period Gross fixed investment and current account surplus (CA) as percentage of GDP (period average). Real wages In socialized sector of the economy. Period I = 1970 to 1978 (except for Yugoslavia 1970-79). Period II = 1979 to 1987 (except for Yugoslavia 1980-87). Sources: For all data except current account Statistical Yearbooks of the three countries. For the current account: International Financial Statistics (various Issues), Bilten Narodne Banke Jugoslavije (various issues), and Poland: Recent Economic Developments, International Monetary Fund, August 1987. Lines 4 and 5 Illustrate the adjustment pattern of the 1980's. Substantial current account deficits that the countries were able to run in the previous decade were no longer feasible as voluntary lending dried up. This compelled Hungary and Pola.nd to reduce the deficit to about half of Its previous amount (in terms of GDP),5 while In the case of Yugoslavia, the current account moved Into surplus. The squeeze on domestic demand thus originated from two directions: real output practically ceased to grow while resource transfer from abroad either decreased or became negative. The last line In Table 1 shows that all three countries attempted to protect personal consumption by severe cuts In fixed Investment. The share of gross fixed Investment In GDP decreased by 8 percentage points In Poland and Yugoslavia and by 6 points in Hungary. The decline In current consumption was thus arrested at some Indeterminate cost In terms of future output and consumption. In addition, cut In Investments probably contributed to the widening of the technological gap between these countries and the West (e.g. Brada (1989)). 4 In what way can we expect, on an a priori basis, that effects of recession on poverty may differ in socialist from those In capitalist economie'l There may be two such differences. First, recession In capitalist countries Is generally associated with decreasing income from ownership as profits and dividends decline faster than wages and/or become negative. Total income of those with relatively high share of capital Income decreases faster than Income of other groups and some of them may even (measured by ann-ual Income alone) become poor. In socialist economies, where the share of capital income ir. total household income Is much less (as profits are originally not received by Individuals but are retained by enterprises or transfered to the state) functional Income distribution at the household level need not change very much and the effects on size Income distribution may be absent. Second, socialist recession was not, in the period studied here, accompanied by Increased unemployment. In Poland and Hungary unemployment was virtually non-existant while in Yugoslavia rate of unemployment (as percentage of labor force) remained more or less unchanged. Wage blll was reduced through uniform cuts In all wages. This was probably Il. part the result of a strong role of workers' councils or trade unions at the enterprise level and in part was caused by government policies (including subsidization of loss-makers) designed to prevent Increases In unemployment. The effect of uniform wage cuts on poverty was probably less severe than if the same reduction In the wage bill was achieved through layoffs (even if in capitalist economies unemployment benefits may sustain some of the unemployed above the poverty line). The two effects (low share of capital Income and absence of Increase In unemployment) suggest that the Increase in poverty due to recession would, In a typical socialist economy and under ceteris paribus conditions, be less than In a capitalist twin. Growth slowdown and the about-turn In the resource transfer In the three East European countries were, however, of such a magnitude that they were reflected at the level of households in an almost uniform decline in real Incomes and increase in poverty. In the next section we shall explore how the economic crisis has affected average consumption levels and real Incomes of 5 different social groups, and will then proceed to a comparison of poverty lines, and extent and the composition of poverty in the three countries. II. Impact of the Crisis on Population Incomes The analysis of poverty is conducted using 1978 as the benchmark year. The choice of the benchmark year Is made relatively simple by the fact that 1978 or 1979 (in the case of Yugoslavia) was the last year of relative prosperity, displaying most of the features of the decade of the 1970's. The period of the 1980's represents for Yugoslavia and Hungary a homogeneous period of stagnation and crisis. The situation Is slightly different for Poland where we can distinguish two subperiods. The first Is the subperiod 1978-82 during which real GDP contracted by one-fifth; the second is the subperiod of recovery when GDP grew on average by 5 percent annually. Polish current account deficit decreased in that subperiod from a very high level of almost 6 percent of GDP in 1979-82 to cnly 1.6 percent. The Impact of recovery on domestic demand and p4;sor.al Incomes was therefore less than the relatively high rate of growth of GDP alone would suggest. Table 2 illustrates tne effects of crisis on some of the variables most closely related with populr on welfare. Table 2. Crisis and Its Effects on Population Incomes (levels in 1987 with 1978=100) Poland Hunfarv Yugoslavia Domestic demand per capitaa/ 87.9 103.6 85.9 Consumption per capita 98.0 126.1 100.5 Real wagesb/ 71.1 101.0 72.0 Real pensions 83.2 120.4 88.1 Real income of workers' HHC/ 67.8 103.9 63.0 Real income of farmers' HH 81.5 101.0 102.5 Real income of mixed HH 84.5 -- 81.8 Real income of pensioners' HHs 82.6 118.2 -- Notes: All data are on per capita basis. Data on real Incomes of different types of households are obtained from household surveys. HH = household. a/ For Yugoslavia, excludas changes in stocks. b/ Real wages In the socialized sector. c/ For Yugoslavia, includes all non-agricultural households, that Is, workers and pensioners. ____________________________ 6 In 1987 real per capita domestic demand in Poland and Yugoslavia was respectively 12 and 14 percent less than at the beginning of the crisis. while in Hungary It was practically the same. Contraction of consumption was less sharp since authorities tried to protect population's living standard by reducing more substantially other types of domestic demand, and in particular fixed investments. The attempt to moderate decline In living standards was most evident In Pcland. Thus while GDP reached Its peak In 1978, real per capita consumption continued to grow until 1980 (precip1tating a drop In the share of gross fixed investment In GDP from 34 to 28 percent). The ef.ect of the crisis on the population's standard of living Is best illustrated by changes In their Incomes. Real wages in Poland and Yugoslavia were reduced by about 30 percent. The impact on real pensions was not as severe, although in Yugoslavia and Poland they decreased respectively by 12 and 17 percent. In Hungary, however, real pensions at the end of the period were some 20 percent higher than in the beginning. We can already note one Interesting feature of the crisis: pensions and, as we shall see, agricultural incomes were less affected than wages. Consequently, the ratio between average pension and average wage went up in each of the three countries: in Poland It increased from 45 to 53 percent; In Yugoslavia, from 58 to more than 70 percent, and in Hungary, from 48 to 57 percent. Rising ratio of pensions to wages took place in conditions of increasing share of pensioners in total population, due both to aging population (as in Hungary), and more liberal pension policies (as In Yugoslavia and Poland). In an Interesting twist, more liberal pension policies were linked with economic reform. For example, in Poland, significant Increase in the number of pensioners (more than 1.3 million) took place in the period 1982-8', after the mandatory retirement age was lowered by five years. The decision was motivated by the fear that widespread unemployment would follow Introduction of market-oriented reforms In 1982 and 1983. In Yugoslavia, pensions kept Increasing compared to wages, because of a legal stipulation that guaranteed (until 1983 when It was rescinded) constant real v.iue of all old-age pensions.7 7 Household survey data show the deteriorating (relative and absolute) position of workers' households. In Poland, for example, per capita income of workers' households was a third less than in 1978; per capita Income of farmers', mixed .nd pensioners' households decreased by less than 20 percent, with mixed households doing somewhat better than the other two groups. In Yugoslavia, in 1987 per capita Income of non-agricultural households was 37 percent less than In 1978. Since these households include pensioners as well, whose relative position has deteriorated less than that of workers, real incomes of the latter group must have been reduced by even more. Real Incomes of farmers' households increased slightly due principally to favorable terms of trade experienced In the early 1980's. In Hungary real per capita Income of workers' and farmers' households was practically stagnant. III. Comparison of Subsistence Hinimum Lines Our objective Is primarily to analyze trends In poverty within each country. However, it is certainly helpful If the level of development, type of the social system and macroeconomic performance of the countries studied are relatively similar, both because It enables us to make some Inter-country comparisons, and to draw general lessons about the changes In poverty In conditions of economic crisis. Inter-country comparisons of income inequality or poverty a-e fraught with a number of difficulties; yet It can be argued ta.At a comparison between Poland, Hungary and Yugoslavia makes more sense than a comparison of each of these countries with say, Britain or West Germany. In that sense homogeneity of the sample Is very important. Comparisons can be valid only If poverty lines that are used in each case are close. We would expect this as similar levels of development , composition of the population (with still significant size of agricultural labor force), and social aystem (e.g. prevalence of state ownership outside agriculture; widespread redistribution and subsidies; egalitarian wage structure in the socialized sector etc.) would tend to Indicate that subsistence minima cannot be too far apart. 8 In calculations of poverty Incidence for each country we have used an officially calculated subsistence (or social) minimum (see Appendix 1 for the description). Official simply means that the minima were calculated by official institutions (generally attached to ministries), ar.d were used as poverty lines In broad social debates. In neither of the three countries, however, did the calculated minimum perform any real function, that Is was not used as an eligibility criterion for social aid. The social minima are constant In real terms: we arp thus using an absolute standard which, particularly In conditions of stagnant Incomes, seems reasonable. It Is evident from Table 3 that the level of what is officially considered to be subsistence minimum (poverty line) In Poland, Hungary and Yugoslavia Is quite similar. Table 3. Comparison between Official Poverty Lines (four-member urban household; per capita; 1985) za/ aPP bJ X. of averaze waze / Poland 52.5 85.4 38.6 Hungary 52.7 125.6 45.0 Yugoslavia 51.3 110.5 34.1 a/ Converted at the official exchange rate. Monthly amount. b/ In dollars of International consumption purchasing power parity. From United Nations Statistical Commission and Economic Commission for Europe (1988, p.24 and Annex Table 3). c/ Average wage In socialized sector. In nominal dollar terms poverty lines are almost identical. In terms of purchasing power, Hungarian subsistence minimum is the highest reflecting relatively low forint prices of consumer goods. In terms of average wage, poverty lines seem high compared to their West European equivalents. ..tkinson (1991) reports that the value of per capita poverty lir.e for a married couple with two children ranges from only 12.5 percent of average production worker wage in West Germany to 33.7 percent In Sweden. There seem to be two possible explanations for relatively high (in terms of average wage) poverty lines In Eastern Europe. First, a less developed country would tend to have a higher poverty line In terms of average wage, simply because poverty line will normally Increase less than In proportion to average Income (or wage). 9 Kilpatrick (t973) concludes, on the basis of surveys, that the elasticity of the poverty line with respect to income Is about 0.6. Similarly, Smith and Welch (1989, p.524) assume an elasticity of 0.5. Using surveys, Hagenaars and van Praag (1982, pp. 151-2) obtain an elasticity of 0.54 If Income distribution does not charnge (elasticity is greater with a more unequal distribution). Finally, for Yugoslav republic of Slovenht Stanovnik (1989) estimates, on the basis of surveys, Income elasticity of poverty line for urban households at 0.57. If we plot a variable poverty line per capita/average wage against an indicator of the level of development (real per capita consumption at International prices) for both market and socialist economies we observe (see Figure 1) that among the latter on!y Hungary stands out with a poverty line which seems too high for its level of Income. Poland and Yugoslavia are almost exactly on the regression line. The elastic ty of the poverty line with respect to income (or wage; here treated interchangeably) Is about 0.56, which Is consistent with earlier findings.9 The second argument often mentioned to explain relatively high level of poverty lines In socialist economies (which, as we just saw, can apply only to Hungary) Is that poverty lines In socialist countries represent simply accounting devices which are devoid of any policy role. Households below the poverty line are not entitled to any benefits. Consequently, "statistical" poverty lines can be set at relatively generous levels. There may be some merit In this view as poverty lines which do carry policy Implications (e.g. West German line denoted Dl In Figure 1, and UK poverty line) are rather on the low side. German example Is particularly partinent since an estimated poverty line not related to policy is 2.5 times higher than the official line which triggers access to welfare benefits. It is quite likely that, to the extent that Poland, Hungary and Yugoslavia move closer to market economies and a welfare system appears, official poverty lines will be lowered, if for nothing else then because it would be financially Impossible to give Income support to say, one-fourth of the population. 10 Fi c'iie I T iPor rr e xs ;~SF + + D2 POV'ERTY LINE AS -5 OF W'AGE GR+ IRL + us UK + 3 0 3.5 4-0 4 5 5-0 5-5 PER CAPiTA CONSUMPTION AT INTERNAT.PRICES Notes; Poverty line is a per capita poverty line for four-member urban household. Both axis are in logs. 1 1 IV. The Size and Composition of Poverty, 1978-87 In the eighties, the extent of poverty significantli increased ln Poland and Yugoslavia, and remained about the same In Hungary. Table 4 shows the overall percentage of people estimated to live below the official poverty line In the three countries.10 In Poland, the share of the poor increased from less than 10 percent of the population (before the crisis) to about 22 percent In 1987. In Yugoslavia, poverty rate was some 17 percent In 1978; by mid-1980's It stabilized at 25 percent. In Hungary, poverty rate oscillated throughout the period between 14 and 16 percent, without any apparent tendency to go up or down. For comparative purposes poverty Incidence based on official poverty lines for several OECD countries Is shown In Table 5. These estimates (except for Japan) are based on the same poverty lines discussed In the previous section. Table 4. Estimated Poverty Rates for Total Population Year Poland Yugoslavia Hunzarv 1978 9.2 17.5 15.4 1979 9.7 1980 11.1 13.8 1981 13.9 1982 19.8 14.8 1983 23.7 12.8 16.7 1984 21.9 21.5 1985 19.1 25.7 15.7 1986 17.3 25.1 1987 22.7 24.8 13.8 Source: See Appendices 1 and 2. Table 5. Poverty Incidence in Several OECD Countries Greece (1982) 22.7 Ireland (1987) 21.2 United States (1985) 14.0 Sweden (1985) 8.6 West Germany (1983) 6.1 United Kingdom (1985) 4.5 Finland (1985) 2.9 Japan (1984) 1.0 Sources: Atkinson (1991) for all countries except Japan (Larin (1989)). Data for West Germany based on the higher poverty line (D2 in Figure 1). Table 6: POVERTY RATES BY DIFFFRENT SOCIAL GROUPS (In % of total population of a given group) Years Farmers Mixed Households Worke5s Pensioners P YU HI P YU P VU H P 1978 14.9 41.8 9.5 17.3 6.4 8.6 20.8 21.1 1979 16.7 12.8 6.1 17.1 1980 17.2 10.6 7.8 23.7 17.2 1981 16.4 11.4 11.4 29.2 1982 20.9 15.8 17.3 35.7 13.3 1983 29.7 26.6 13.4 13.2 19.1 9.5 49.0 18.0 1984 25.1 27.2 12.9 27.9 19.0 16.6 39.3 1985 19.5 38.6 13.9 11.3 31.0 17.3 20.0 17.7 32.4 14.2 1986 19.2 44.7 9.4 29.9 17.0 MO.o 25.4 1987 21.4 44,6 10.7 12.6 26.7 25.2 20.0 17.3 27.6 10.5 1Includes mixed households. 2Includes pensioners. Source: See Appendices 1 and 2. P- Poland; H= Hungary; YU- Yugoslavia. 13 The breakdown of the total poverty rate between different social groups Is presented In Table 6. Following Polish classification, we have used four social groups: farmers', mixed, workers', and pensioners' households. Mixed households are those where at least one member works full-time In agriculture (generally, cultivating his own land) while another Is employed In Industry or services. Workers' households include both blue- and white-collar workers. Definitions of other types of households are self-evident. 11 Consider first Poland. Poverty rate among farmers Increased from about 15 percent to some 20 percent. The peak was achieved in 1983, following a 10 percentage point decllne In terms of trade between agriculture and industry. Poverty rate among workers Increased almost fourfold. At the outset of the crisis, workers had the lowest poverty rate among the four groups: only 6 to 7 percent of people belonging to workers' households were classified as poor. In 1987, one out of four persons living in workers' households was poor. Increase In the poverty rate mirrors a 30 percent decline In workers' real wages and per capita real Income (see Section 2). Already by 1983 average per capita Income of workers' households was lower than that of mixed and farmers' households. The tendency continued so that by 1987 farmers' and mixed households had average per capita incomes respectively 9 and 3 percent higher than workers (see Table 7). Table 7. Relative Per Capita Income of Different Social Groups (per capita workers' Income In each country = 100) a/ 1978 1983 1987 Poland Mixed 81.8 100.4 102.6 Farmers 90.8 107.0 109.2 Pensioners 76.6 78.3 93.4 Yugoslavia Mixed 70.1 86.8 91.1 Farmers 50.1 82.2 81.6 Hungary Pensioners 84.4 93.3 96.4 a/ Workers households In Poland; non-agricultural In Yugoslavia; active in Hungary. The latter Include both workers and farmers. 14 Throughout the period incidence of poverty (in Poland) among pensioners was the greatest. However, there are two distinct phases here. After a period of rapid Increase In the extent of poverty, which by 1983 left one out of two persons in pensioners households poor, poverty rates started gradually to decline and were by 1986-7 at the level slightly over 25 percent. This relative Improvement coincided (as we saw before) with an Increase In the number of pensioners: percentage of pensioners in total population Increased from 11 percent In 1978 to almost 18 percent at the end of the period. High numbers of new pensioners whose pensions were higher than those of previous retirees (which failed to keep pace with Inflation in the early 1980's), resulted in an increase of the average real pension, leading, in Its wake, to a drop in the poverty rate. Finally, poverty rates for mixed households increased the least, or even stayed constant. By almost any indicator, mixed households "weathered" the crisis better than others: their per capita income declined between 1978 and 1987 some 3 points less than that of farmers' and 15 percent less than workers' households. A relatively consistent picture of who was most affected by the crisis in Poland thus emerges: among workers' (blue- and white-collar) poverty rates increased by some 15 points; for farmers and pensioners, Incidence of poverty went up by respectively 5 and 7 points, while for mixed households It remained unchanged (see Table 8). Table 8. Who was most Affected by the Crisis? (approximate percentage point change in poverty Incidence between 1978-9 and 1986-7) Poland Yugoslavia Hmunarv Farmers + 5 +3 3- Mixed 0 +11 -- Workers +15 +11 -- Pensioners + 7 -- -11 Total +12 + 8 -1 This pattern Is not very different for Yugoslavia. Poverty rate for non-agricultural population (including both workers and 1 5 pensioners) doubled, Increasing from less than 10 percent to about 20 percent. Despite this, poverty Incidence among non-agricultural population still remains lower than among mixed households and farmers. Average per capita income for non-agricultural households is also higher than that of the other two groups even If the difference has narrowed (see Table 7). As for farmers' households, their poverty rate at the end of the period remains at about the same high level of 40 percent as In 1978 (following a relatively sharp reduction between 1978 and 1984 when agricultural terms of trade Improved). Main resemblance In the pattern of increased poverty between Poland and Yugoslavia concerns the position of the urban employed. In both countries they were the most severely affected. In Poland, their Incomes declined to such extent that poverty rate among them overtook that of farmers' and mixed households. An important difference between the two countries concerns the position of mixed households: In Poland they were the least affected by the crisis, while In Yugoslavia they fared as badly as workers', and much worse than farmers' households. As already mentioned, real incomes of households In Hungary deteriorated much less than In Poland or Yugoslavia. Poverty incidence among workers is between 17 and 18 percent, higher than among either pensioners or farmers. In effect, these two groups recorded decreases In poverty rates from about 14 percent In 1985 to less than 11 percent two years later. It Is also Interesting that poverty among pensioners steadily declined. While in the beginning of the period poverty rates for Hungarian and Polish pensioners were about the same (slightly in excess of 20 percent), by the end of the period Polish rate was 27 and Hungarian only 10 percent. It is only in terms of poverty rates existing among workers that Hungarian situation resembles that of Poland and Yugoslavia. Deterioration In the condltlon of workers' households took place in circumstances of an unchanged distribution. The Gini coefficients for workers' (non-agricultural) households In Poland (Yugoslavia) have remained remarkably stable.1 In 1981 and 1982 when real wages and real per capita Income in workers' households In Poland took a deep plunge, dispersion of Incomes shrunk (to an 16 extremely low Gini coefficient of 0.18). Other than In these two years the Gini coefficient has remained In the narrow range between 0.21 and 0.22. This Is also true for Yugoslavia where the Gini coefficient varied between 0.3 and 0.32. The same conclusion emerges if we look at the distribution of wages in the socialized sector. Increase In the percentage of the poor people among workers' households thus occured almost solely because of a uniform and across the board cut in real Incomes. It appears that everyone was In effect "pushed" by a given percentage below his earlier wage (income) level, leaving the overall shape of the distribution curve unchanged. This is not an unusual reaction to crisis, particularly In societies with strong egalitarian tradition and enterprises run by workers' councils or trade unions. It Is beyond the scope of this work to try to ascertain whether such a uniform decline had some deleterious impact on productivity, by, for example, reducing absolute income differences, and further discouraging (already scarce) innovation, risk-taking or outstanding work effort. Under this scenario continued uravnilovka might have unwittingly contributed to the deepening of the crisis. 13 V. The Changing Composition of Poverty Perhaps as Important as increased poverty was its changirg composition. Instead of being predominantly rural as it was at the end of the 1970's, poverty became an urban phenomenon. The three countries moved from a traditional distribution of poverty toward the one characteristic for fast Industrializing countries, that Is countries ex-periencing rapid change in the structure of production and population. However, to compound their woes, Increased urban poverty occured not because of a rapid migration to cities, but because of a gradual Impoverishment of the already existing urban labor. Its social significance was therefore entirely different. While Increased urban poverty can be politically accommodated if It Is the result of people migrating from villages, looking for better jobs, and accepting a temporary decline In their standard of living in anticipation of an early Improvement, a descent into poverty of people who are already 17 living In cities Is socially much more destabilizing. To realize this we need only focus on what It In reality means, and how It differs from urban poverty Induced by rapid structural change. Descent Into poverty for the already established urban population Is associated with inability to procure replacement for the worn-out consumer durables (TV sets, washing machines etc) that suddenly become out of reach for an ordinary household. Increased rents or electricity bills sharply compress the affordable standard of living below the accustomed level. This represents a process of reversal, In terms of standard and culture of llving, to a level that household might have had some 20 years earlier. Social Implications of such a decline are probably much greater than social implications of a temporary deterioration In one's standard of living following migration to city. There are at least two elements which render the first deprivation (subjectively) much worse: (1) absence of expectation that situation would Improve, a feeling pervading most of Eastern Europe In the 1980's, and which, by definition, cannot exist In a rapidly industrializing developing society; and (2) this was not a voluntarily accepted temporary decline In the standard of living as among migrants. Table 9. The Composition of the Poor in 000; (in percent) Rural poor Urban ooor Total Poland, 1978 1760 (52.0) 1627 (48.0) 3386 (100) Poland, 1987 2441 (29.3) 5879 (70.7) 8321 (100) Yugoslavia, 1978 2592 (75.6) 837 (24.4) 3429 (100) Yugoslavia, 1987 2707 (52.9) 2411 (47.1) 5118 (100) Note: Data In brackets give the percentage breakdown of the total number of the poor. Rural poor are farmers' and mixed households. Urban poor are workers and pensioners for Poland, and non-agricultural households for Yugoslavia. The changing composition of poverty Is Illustrated In Table 9. Before the crisis, the rural poor accounted for 52 percent of total poor In Poland, and three-quarters In Yugoslavia. About 10 years later, urban poor represent 70 percent of all the poor In 18 Poland, and almost one-half In Yugoslavia. The marked shift toward urban poverty Is not to any significant extent the product of the changing structure of population: urban population In Poland and Yugoslavia increased between 1978 and 1987 by respectively 15 and 24 percent, while the number of the urban poor Lncreased 3.6 times in Poland, and 2.9 times in Yugoslavia.14 VI. Poverty, Income and Redistribution One of the objectives of a study of poverty Is also to relate changes In the Incidence of poverty to macroeconomic variables and to disentargle the influence of Income changes from the Influence of redistribution. Regularities of this kind, if established and found statistically robust, allow us to make conclusions about the impact of various mac.roeconomic measures on poverty. In this section we shall try to do this on the example of Poland and 15 Yugoslavia (see Table 10) . In order to expand the time series and make our conclusions stronger we pool data for different social groups (workers, farmecs and pensioners) and estimate only one equation per country. The simplest and most natural candidates for explanatory variables are (1) average real per capita Income of the group, and (2) the Gini coefficient for the group's Income. We can expect that the first variable be negatively, and the second, positively, related to poverty. 19 Table 10. Determinants of Poverty Rates Period constant income distrib. RSE term term term (F) Poland 1978-87 391.53 -52.23 9.105 0.777 4.44 (0.00) (0.00) (0.04) (51.5) Ytugoslavia 1978-87 293.39 -42.98 46.405 0.890 4.05 (0.016) (0.0) (0.014) (45.6) Notes: Equations pool together data for different social groups for the period 1978-87. Equations are of the semi-log form: POOR = constant + B1 log (income) + B2 log (distribution), where POOR = percentage of the poor for a particular social group and year (workers, farmers and pensloners for Poland; non-agricultural and agricultural households for Yugoslavia), while Income and distribution terms are corresponding values of the average per capita Income and the Gini coefficient for that group and that year. Data in brackets below regression coefficients show the level of significance at wlhich the null hypothesis Is rejected. Number of observations Is 30 for Poland and 12 for Yugoslavia. ---------------------------------------- Elasticity of poverty rates with respect to Income Is about 0.52 for Poland, and 0.43 for Yugoslavia. This means that In order to rec wce poverty rate by one percentage point real per capita inco. needs to increase, on average and without any change In distribution, by about 2 percent in Poland and 2.5 percent In Yugoslavia. The Implication Is that more people In Poland than In Yugoslavia are "bunched" slightly below the poverty line. Not unexpectedly, In light of fairly egalitarian distribution of Income In Poland, change In inequality has less of an effect on poverty In Poland than in Yugoslavia: for poverty rates to go down by 1 point, Inequality of distribution (measured by the Gini coefficient) must decrease by 2 percent In Yugoslavia and by more than 10 percent in Poland. The trade-off bet.ween growth and redistribution as the two alternative ways to reduce poverty, can be also considered in a slightly different way. The change In poverty rates between 1978 and 1987 can be decomposed Into its three components: changed overall level of income, changed pattern of distribution, and changed demographic composition of the population. The last 20 effect means that even if Incomes and distribution remain the same, poverty rates could go up or down If there is a change in the composition of the population, due to different population growth rates between rural and urban households, or due to migration of the population from one area (social group) to another. Table 11 shows that the Income effect has been by far the most significant determinant of changes in poverty rates. Implied Increase In poverty due to the decline in Income has, In effect, been greater than the actual increase In poverty as redistribution (in Poland), and redistribution plus favorable demographic changes (in Yugoslavia) combined to mitigate the impact of falling incomes. The extent to which declining incomes have caused an Increase in poverty is best understood when one considers that with an overall Income equal to that before the crisis, population living below the poverty line In 1987 would amount to only 6 percent In Poland (instead of the actual 22.7 percent), and 12.7 percent in Yugoslavia (instead of 24.5 percent). Among all social grox,ps, the only one where each of the three effects contributed to an increase In poverty was the group of non-agricultural households in Yugoslavia. Negative demographic effect is explicable by higher pcpulation growth rates among urban population In less developed regions. 21 Table 11. Decomposition of Change in Poverty between 1978 and 1987 (in percentage points) Income Distribution Demographic Total effect effect effect effect Poland Workers +18.7 -2.4 +2.5 +18.8 Mixed + 9.8 -5.9 -0.2 + 3.1 Farmers +15.6 -7.7 -0.2 + 6.5 Pensioners +19.0 -11.3 +1.2 + 6.8 Total +16.7 -4.2 +1.5 +13.5 Yugoslavia Agricultural +18.3 -15.4 0 +1.0 Mixed +14.4 -3.1 -1.3 +9.4 Non-agricultural +9.1 +0.2 +1.9 +11.4 Total +11.8 -3.3 -0.7 +7.3 Notes: Income effect calculated by assuming 1978 level of country-wide real income and everything else a- It Is In 1987. Demographic effect calculated by assuming the 1978 composition of population (between social groups and sizes of households) and everything else as It is in 1987. Distribution effect calculated as residual. a/ The difference between the total effect and the sum of the three Individual effects Is due to an Interaction term. _______________________________ VII. Conclusions Deep economic crisis that characterized Eastern Europe In the 1980's had a strong Impact on average Incomes of the population and on the percentage of people living below the poverty line. The situation deteriorated most sharply In Poland, where the percentage of the poor In total population Increased from less than 10 percent before the crisis to more than 20 percent by the late 1980's. In Yugoslavia, Increase was somewhat less (from 17 to 25 percent). In Hungary overall poverty rate remained constant. Perhaps as important was the change In composition of poverty: while before the crisis most of the poor lived In rural areas, now most of the poor (as many as 70 percent in Poland) live In cities. Social group most affected by the economic decline are urban, socialized sector workers (inclusive of white-collar workers). Their real wages, or real per capita income of their 22 households, declined by more than 30 percent in Poland and Yugoslavia. Reduction in Income was accompanied by an unchanged pattern of distribution as socialized sector wages were generally cut uniformly across the board. 23 APPENDIX 1 - POVERTY LINES For Hungary and Yugoslavia the social minima are In effect subsistence minima that "render possible merely to satisfy very modest necessities conventionally considered to be essential to ensu-e continuous living" (Hungarian Statistical Yearbook 1987, p. 328). Both Hungary and Yugoslavia also publish data on the social minimum which Is some 20 percent higher than the subsistence minimum. In the analysis for Poland we are usinig the social minimum where some 10 percent of resources are supposed to have been left at the discretion of the household, i.e. to be used to satisfy non-essential needs. Polish social minimum, however. Is less, in real terms, than the subsistence minimum in Yugoslavia and Hungary (see Table 3 In the text). Consumption baskets are based on actual consumption patterns (e.g. in Yugoslavia, the structure of consumption of the lowest Income decile) corrected with some norms reflecting "healthy" consumption. All the three minima assume that the household rents a state-owned or cooperative apartment (with rents substantially below market levels). However, since most of the population either lives in own housing (necessitating only low direct maintenance expenditures and being charged mortgages at vastly negative interest rate), or In state or cooperative apartments, underestimate of the poverty line due to housing costs Is probably small. In Poland, social minimum Is calculated quarterly since 1980 by the Institute of Labor and Social Affairs. The basket of goods has been kept constant since 1983. In order to preserve an absolute standard of measurement we have extended back to 1978 the real value of the social minimum for 1983. In Hungary, social and subsistence minima are calculated since 1982 by the Central Statistical Office. For the period before 1982, it was assumed that the ratio of the minimum to the average monthly per capita Income was the same as In 1982. Since real per capita income In 1978 and 1980 was about the same as In 1982, this Is tantamount to using a constant absolute level. For Yugoslavia, social minimum Is calculated by the the Federal Secretarlat of Labor, Health and Social Policy since the 24 early 1980's. We had access only to 1987 and 1986 data, while the values for other years were obtained by deflating 1987 data by the retail price index. In order to obtain the subsistence minimum, two categories (clothing and consumer durables) out of seven categories of commodities of which the social minimum Is composed were excluded. Finally, it should be noted that the total percentage of rpeople In poverty is obtained as the sum of people DelonIging to different social (or regional) groups who are poor, where to each group we apply a different poverty line (minimum). For example, In Poland there are four such minima: for workers', mixed, farmers' and pensioners' households; In Hungary, there are separate minima for each type of household (active and non- active) and for each household size; In Yugoslavia there are eight minima (for each republic or province). Consequently, the use of the national minimum to the combined set of data for all households will not In general yield the same result as the one presented here. APPENDIX 2 - SOURCES OF DATA FOR HOUSEHOLD INCOMES Polish results are based on annual household surveys conducted by the Central Statistical Office, and covering between 8500, in the early years, and more than 20,000 households later. Households are divided Into four social groups (workers, mixed, farmers and pensioners), and 7 to 8 Income groups ranked by per capita household Income. Surveys are representative of about 90 percent of the population (not Included are Army, police and private non-agricultural sector). Yugoslav data come from five-year household surveys (conducted in 1978 and 1983). Data for other years are calculated from smaller-sample annual surveys (6000 households vs. 19,000 In five-year surveys). All households are divided Into three social groups (non-agricultural, mixed, and farmers), and between 10 and 11 Income groups (ranked by per capita household income). Surveys are conducted by the Federal Statistical Office. Hungarian data come from biannual household surveys conducted 25 by the Central Statistical Office. For 1985 and 1987, It Is possible to break overall data Into three social groups (workers, farmers and pensioners); for earlier years only results for total population and pensioners' louseholds could be calculated. Number of income groups varies from 8 to 12 (also ranked by per capita household Income). I1There are no elements that would inidicate that the surveys have in general become less reliable and/or display a systematic bias throughout the period under study. The surveys, however, do have several flaws. For example, Polish samples have a response rate of only about 70 percent. Director of the Department of Household Budgets in Polish Central Statistical Office points out that rejections are the highest among the poor and the well-off households (Mr. Kordos's personal communication to the author). Incidence of poverty (even In relative terms) may therefore be somewhat understated. In the case of Yugoslavia, surveys Include as income items those which do not belong there (e.g. revenues from the sale of assets). Since this item is shown together with Income from the lease of assets It Is Impossible to deduct It from Income. Income Is thus overestimated, and poverty Incidence biased downward. Income data in the surveys In all three countries fail to adequately reflect interest Income from financial assets. When real return on these assets (including foreign exchange) was slightly negative or nil, the bias was not very significant. In the 1980's, however, the real rate of return on foreign exchange was almost continuously positive and In some years it was very high. On the other hand, real return on domestic deposits, particularly In Poland and Yugoslavia which experienced high Inflation, swung violently: In some years It was strongly negative while in others It was around zero. Due to the lack of data on the distribution of financial assets by income groups It is impossible to adjust income for this element. Consequent underreporting of Income creates an upward bias In poverty statistics (even If the bias may not be high since poor households typically have few financial assets). 26 References Atkinson, A.B., "Poverty, Econonic Performance, and Income Transfer Policy in OECD Countries", The World Bank Economic Review, forthcoming, 1991. Bergson, Abram, "Income ineq'iality under Soviet Socialism", Journal of Ecor.omic Literature, Septemoer 1984, pp. 1052-1099. Brada, Josef, "Technological Progress and Factor Utilization", Economica, November 1989, pp. 433-448. Buse, Adolf, "The Cyclical Behavior of the Size Distribution of Income In Canada: 1947-78", Canadian Journal of Economics, May 1982, pp. 189-204. Flakierski, Henryk, Economic Reform and Income Distribution. A Case Study of Hungary and Poland, New York:M.E. Sharpe, 1986. Gomulka, Stanislaw, "Soviet Equilibrium Technological Gap and the Post-1975 Productivity Slowdown", Economics of Planning, No. 1-2, 1988, pp. 1-17. Gomulka, Stanislaw and Jacek Rostowski, "International Comparison of Material Intensity", Journal of Comparative Economics, December 1988, pp. 475-501. Hagenaars, Aldi and Bernard van Praag, "A Synthesis of Poverty Line Definitions", Review of Income and Wealth, September 1982, pp. 345-59. Kilpatrick, R.W., "The Income Elasticity of the Poverty Line", Review of Economics and Statistics, 1973, pp. 327-332. Knirsch, Peter, "The Limits to Growth In Socialism", Survey, Spring 1984, pp. 70-82. Larin, Kathryn, "Country Comparative Study: Industrialized Countries", Background paper for the 1990 World Development Report, October 1989, mimeo. Morrisson, Christian, 'Income Distribution in East European and Western Countries", Journal of Comparative Economics, 8, 1984, pp. 121-38. Milanovic, Branko, "Poverty in Poland in the Years of Crisis, 1978-87", Background paper for the 1990 World Development Report, November 1989, mimeo. Nolan, Brian, "Macroeconomic Conditions and the Size 27 Distribution of Income: Evidence from the United Kingdom", Journal of Post Keynesian Economics, Winter 1988-89, pp. 196-221. OECD Department of Economics and Statistics, Purchasing Power Parities and Real Expenditures, Paris: OECD, 1987. Okrasa, W. "Redistribution and the Two Dimensions of Inequality: An East-West Comnarlson", European Economic Review, 32, 1988, pp. 633-643. Posarac, Aleksandra, "Poverty in Yugoslavia 1978-87", Background paper for the 1990 World Development Report, November 1989, mlmeo. Smith, James and Finis R. Welch, "Black Economic Progress After Myrdal", Journal of Economic Literature, June 1989, pp. 519-564. Socha, Mieczyslaw W. "Wages and Incentives Problems", In Roger A. Clarke (ed.), Poland: The Economy in the 1980's, London:Longman, 1989, p.45. Stanovnik, Tine, "The Direct Measurement of Welfare Levels In Slovene Households", Economic Analysis and Workers' Management, No.1, 1989, pp. 43-54. Szalai, Julia, "Poverty In Hungary during the Period of Economic Crisis", Background paper for the 1990 World Development Report, November 1989, mimeo. Unlted Nations Statistical Commission and Economic Commission for Europe, International Comparison of Gross Domestic Product in Europe 1985, New York: United Nations, 1988. Winiecki, Jan, "Are Soviet-Type Economies Entering an Era of Long-term Decline?", Soviet Studies, July 1986, pp. 325-348. World Bank, The World Bank Atlas 1989, Washington D.C.: The World Bank, 1989. Zloch, Iliana, Debt Problems of Eastern Europe, New York: Cambridge University Press, 1987. 28 1 The discussion is based ir. part on papers prepared for the 1990 World Development Report by Milanovic (1989), Posarac (1989) and Szalal (1989) dealing respectively with Poland, Yugoslavia and Hungary. I gratefully acknowledge comments made by B. Balassa, R. Kanbur, A. Posarac, T. Stanovnik, and I. Topinska as well as by three anonymous referees. I am alone rasponsible for views and interpretation expressed here, as well as any errors that may remain. 2 To quote only a few which deal with Eastern Europe as a whole: Gomulka (1988), Gomulka and Rostowski (1988), Winiecki (1986), Zloch (1987), Knirsch (1984). 3 Almost all studies are concerned with income distribution, and not poverty: e.g. Flakierski (1986), Morrisson (1984), Okrasa (1988), Bergson (1984). 4 In Hungary and Yugoslavia the highest annual growth rate of GDP achieved since the onset of the crisis (respectively 3.4 and 3.6 percent) is less than the lowest growth rate In the period 1970-78. S In the case of Poland, current account deficit measured as percentage of GDP was underestimated In the 1970's. This Is due to an overvalued exchange rate of the zloty which made the dollar GDP appear higher than It was In reality. Reduction In the external Inflow of resources In the 1980's was thus, at least for Poland, greater than suggested by the numbers. 6 Unemployment Is almost always found to have a negative effect on the Income share of the bottom quintile or decile of the population (see Nolan (1988-89, p.205) and Buse (1982, p.203)). 7 The share of pensions in GDP Increased In all three countries. In 1987, it was the highest in Hungary (7.4 percent), little below 6 percent in Poland, and 5.6 percent In Yugoslavia. 29 8 According to the World Bank (1989) data, Poland's GDP per capita In 1988 was $18SO, Hungary's $2460 and Yugoslavia's $2680. 9 The estimated regression is log(-j-) = f(C), where P/W = poverty line as percentage of average wage in the country; C = real per capita consumption at international prices In 1985, obtained from United Nations Statistical Commission and Economic Commission for Europe (1988, Table 5.11, p.29) except for the US which comes from OECD (1987, Table 8, ICP classification). (P/W) data for market economies from Atkinson (1991). The estimated regression Is log (P/W) = 5.1645 - 0.4416 log C (0.001) (0.082) with R =0.221 and SE=0.364. Assuming that both W and C measure the level of inca.e and writing out the last relation we can easily see that elasticity of the poverty line with respect to income Is equal to 1-0.4416=0.5584. 10 A household whose Income divided by the number of consumption units is less than the minimum (per consumption unit) Is considered to be poor. Classified as poor are all members of this household. This approach was used for Poland and Yugoslavia. Hungarian data give subsistence minima for different sizes and types of households, so that a direct comparison of household income with the minimum allows one to place the household above or below the poverty threshold. For more details see Appendices 1 and 2. 11 If a household Includes both a worker and a pensioner It Is Included In one or another group depending on the dominant source of Income. This Is the Gini coefficient of persons living In workers' households when households are ranked by their per capita Income. 30 13 A Polish economist writes: "The crisis generated a new wave of uravnllovka (levelling) In a situation where engineering knowledge is needed to stimulate technological progress, and the qualifications of lawyers and economists are Indispensable to put a new programme of reforms Into practice. Continuance of this pay policy produces frustration among Intelligentsia and declining Interest In academic studies" (Socha (1989, p. 55)). 14 In effect, migrational movements were not partlcularly strong In either country. In Poland, urban population grew at an average rate of 1.4 percent (vs. 0.8 percent for the country as a whole). In Yugoslavia, urban population increased by 2.3 percent p.a. (vs. 0.6 percent for the whole country) It Is quite possible that Increased urban poverty discouraged potential migrants. is Data series for Hungary Is not sufficiently long to allow for regression estimation. 16 When calculating the income effect we assume that the country-wide income Is the same as in 1978. Average incomes of different social groups (or regions) are adjusted linearly--this implies that relative Incomes of these groups (i.e. compared to the overall average) are the same as In 1987, and can be different from what they were In 1978. PRE Working Paper Series Contact vA Author Do for gaper WPS483 An Evaluation of the Main Elements Refik Erzan August 1990 G. llogon in the Leading Proposals to Phase Paula Holmes 33732 Out the Multi-Fibre Arrangement WPS484 Stock Markets, Growth, and Policy Ross Levine August 1990 R. Levine 39175 WPS485 Do Labor Market Distortions Cause Ram6n Lopez August 1990 R. Luz Overvaluation and Rigidity of the Luis Riveros 34303 Real Exchange Rate? WPS486 A RMSM-X Model for Turkey Luc Everaert August 1990 S. Aggarwal Fernando Garcia-Pinto 39176 Jaume Ventura WPS487 Industrial Organization Implications Timothy Condon August 1990 S. Fallon of QR Trade Regimes: Evidence Jaime de Melo 38009 and Welfare Costs WPS488 Prepaid Financing of Primary Health Per Eklund August 1990 K. Brown Care in Guinea-Bissau: An Knut Stavem 35073 Assessment of 18 Village Health Posts WPS489 Health Insurance in Zaire Donald S. Shepard August 1990 K. Brown Taryn Vian 35073 Eckhard F. Kleinau WPS490 The Coordinated Reform of Tariffs Pradeep Mitra August 1990 A. Bhalla and Domestic Indirect Taxes 37699 WPS491 How Well Do India's Social Service Nirmala Murthy August 1990 E. Madrona Programs Serve the Poor? Indira Hirway 37483 P. R. Panchmukhi J. K. Satia WPS492 Automotive Air Pollution: Issues and Asif Faiz August 1990 P. Cook Options for Developing Countries Kumares Sinha 33462 Michael Walsh Amiy Varma WPS493 Tax Reform in Malawi Zmarak Shalizi August 1990 A. Bhalla Wayne Thirsk 37699 WPS494 Alleviating Transitory Food Crisis Victor Lavy August 1990 A. Murphy in Africa: International Altruism 33750 and Trade WPS495 The Changing Role of the State: Arturo Israel August 1990 Z. Kranzer Institutional Dimensions 37494 PRE Working Paper Series Contact ~~~~~~~~~~AMD An h o for paper WPS496 Issues in Evaluating Tax and Robert Conrad August 1990 A. Bhalla Payment Arrangements for Publicly Zmarak Shalizi 37699 Owned Minerals Janet Syme WPS497 The Measurement of Budgetary Carlos Elbirt August 1990 T. Gean Operations in Highly Distorted 34247 Economies: The Case of Angola WPS498 The Build, Operate, and Transfer Mark Augenblick August 1990 D. Schein ('BOT) Approach to Infrastructure B. Scott Custer, Jr. 70291 Projects in Developing Countries WPS499 Taxing Foreign Income in Capital- Chad Leechor September 1990 A. Bhalla Importing Countries: Thailand's Jack M. Mintz 37699 Perspective WPSSOO Projecting Fertility for All Countries Eduard Bos September 1990 V. Altfeld Rodolfo A. Bulatao 31091 WPS501 Tax Systems in the Reforming Cheryl W. Gray September 1990 L. Lockyear Socialist Economies of Europe 36969 WPS502 Patents and Pharmaceutical Drugs: Julio Nogues September 1990 M. T. Sanchez Understanding the Pressures on 33731 Developing Countries WPS503 Household Production, Time John Dagsvik September 1990 M. Abundo Allocation, and Welfare in Peru Rolf Aaberge 36820 WPS504 Applying Tax Policy Models in Henrik Dahl September 1990 A. Bhalla Country Economic Work: Pradeep Mitra 37699 Bangladesh, China, and India WPS505 Creating the Reform-Resistant Arye L. Hillman September 1990 CECSE Staff Dependent Economy: The CMEA Adi Schnytzer 37176 International Trading Relationship WPS506 Changes in Food Consumption Merlinda D. Ingoo September 1990 A. Daruwala Patterns in the Republic of Korea 33713 WPS507 Poverty in Poland, Hungary, and Branko Milanovic September 1990 A. Bretana Yugoslavia in the Years of Crisis, 37176 1978-87