Argentina - Crisis and poverty 2003 - a poverty assessment (Vol. 1 of 2) : Main Report

Report No. 26127-AR Argentina Crisis and Poverty 2003 A Poverty Assessment (In Two Volumes) Volume I: Main Report July 24, 2003 Poverty Reduction and Economic Management Latin America and the Caribbean Region Document of the World Bank   TABLEOF CONTENTS EXECUTIVE SUMMARY ......................................................................................................................... i THEECONOMIC CRISIS AND THE POOR ..................................................................................................... i THEGOVERNMENT RESPONSE .................................................................................................................. 11 REVERSING THE TRENDS -POLICIES FOR THE IMMEDIATEFUTURE ....................................................... i11 LONGER-TERM POVERTY REDUCTION MEASURES ................................................................................... V CHAPTER 1 THE EVOLVING CRISISAND THE POOR ................................................................. 1 BACKGROUND ........................................................................................................................................... 1 POVERTY TRENDS ..................................................................................................................................... 3 LABOR MARKETS ADJUSTMENT AND ....................................................................................................... 9 CHAPTER 2 HOW THE POORCOPEWITH A CRISIS .................................................................. 14 IMPACTOF THE CRISISINARGENTINEANFAMILIES - A SURVEY OF CSOS ........................................... 14 RESULTS OF A SPECIAL HOUSEHOLD SURVEY ........................................................................................ 18 COMPARISONS WITH OTHER COUNTRIES ............................................................................................... 25 CHAPTER 3 THE GOVERNMENTRESPONSETO THE CRISIS: SOCIAL SPENDING ..........27 BACKGROUND AND GENERALTRENDS .................................................................................................. 27 SOCIAL ASSISTANCEPROGRAMS ............................................................................................................ 29 EFFICIENCYANDEFFECTIVENESSTARGETED SOCIALSPENDING..................................................... OF 32 IMPACTOF THE CRISIS ONEDUCATION .................................................................................................. 35 IMPACTON THE HEALTH SECTOR ........................................................................................................... 37 IMPACTON AFFORDABILITY OF BASIC INFRASTRUCTURE SERVICES ...................................................... 40 CHAPTER 4 POLICY IMPLICATIONS AND THEFUTURE ......................................................... 45 SHORT-TERM ISSUESRECOVERY.............................................................................. FOR A PRO-POOR 45 LONGER-TERM POVERTYREDUCTION MEASURES ................................................................................. 47 STATISTICAL ANNEX ........................................................................................................................... 53 BACKGROUND PAPERS (INVOLUME11) . . Ariel Fiszbein. Paula Inks Giovagnoli and Isidro Aduriz. 1ARGENTINA'SCRISISANDITSIMPACTONHOUSEHOLD WELFARE 2 RISK-ADJUSTED POVERTYINARGENTINA:MEASUREMENTDETERMLNANTS . AND . Guillermo Cruces and Quentin Wodon. 3 ECONOMIC CRISISAND SOCIAL CRISISINARGENTINA SOCIAL ORGANIZATIONS.COMMUNITIES . AND FAMILIES . Sandra Cesilini. Irene Novacovshy. Carolina Mera and Paula Giovagnoli.  4. POVERTY AND THE LABOR MARKETTHEARGENTINECRISIS1998-2002. IN Emesto Kritz 5. GOVERNMENT SERVICESFORTHE POOR. Cristo'bal Ridao-Cano. 6. THEARGENTINE HEALTH SECTORINTHE CONTEXT OFTHE CRISIS. Juan Pablo Uribe and Nicole Schwab. 7. THEIMPACT OFTHE CRISIS ONTHE ARGENTINE EDUCATIONAL PROCESS. Sergio Espaiia, Suhas D. Parandekar and Maria Paula Savanti. 8. POVERTYMEASUREMENTINARGENTINA Haeduck Lee 9. THESOCIAL IMPACT OF THE ARGENTINECRISIS ONTHE INFRASTRUCTURESECTORS Vivien Foster and CEER  Argentina -Crisisand Poverty 2003 A Poverty Assessment EXECUTIVE SUMMARY The EconomicCrisisandthe Poor Argentina has gone through a major crisis in the past year, resulting in severe social dislocations and a reduction of welfare by its poorest. The collapse of the Convertibility Plan, the freezing of bank deposits, and the default on foreign debts andthe resulting highinflation, falling output and exchange rate devaluation, carried with it severe consequences for the poor. The rate of poverty as traditionally measured rose from 37% in 2001 to a peak of 58% by the end o f 2002. The numbers o f people living below the indigence line doubled. Income distribution worsened, suggesting that the poorest segments of the population suffered more than the more well off. Unemployment rose, but not nearly as much as expected, perhaps because o f the mitigating effects of government employment programs. Nevertheless, the rate of unemployment continues to be highat 18%. The sharp rise inpoverty in great part has been due to the rise inthe prices o f foods, a major portion of the expenditures of the poor. Unlike other economies, foods are an important export inArgentina, andtheir prices rose with the devaluation. The break with the Convertibility Planalso meant that the adjustment inthe labor market occurred more through wages, rather than by an increase inunemployment. Inflation reduced real wages substantially. Unlike previous recessions, moreover, unemployment arose largely from the formal sector, with an increase inemployment in the informal sector. Inprevious recessions, most of the unemployment came from the informal sector. Thus, the informal sector acted as a type o f cushion for cutbacks inthe formal sector. However, the jobs created in the informal sector inthis case tend to be temporary, low paidjobs that effectively obscure the depth of the crisis. While employees inthe formal sector were entitled to severance payments under the labor law, there i s evidence to suggest that they received amounts substantially less than their entitlements as employers negotiated partialsettlements. The impact of the recession has been difficult for many groups. Many of the middle class, faced with both declining wages and freeze on bank assets, moved into poverty for the first time. These "new poor" have characteristics somewhat different than the traditional poor, including higher levels of education. Inaddition, they appear to have a more difficult time adjusting to new circumstances. Interviews with civil society organizations representing the poor point to higher levels of family violence, alcoholism, and crime. Families appear to cope through a variety of strategies, including increased production o f goods inthe home, the entry into the workforce o f those not previously employed, and reduced consumption of food and other products. Inaddition, they often substituted toward the purchase o f cheaper goods and lower cost foods. Expenditures on school materials were often reduced, and while enrollments have not declined there may be a decline in attendance. Inmany cases, the poor have purchased through delayed payment schemes and/or raised money by selling assets. i  The GovernmentResponse The Government has made a major effort to maintain or expand social spending, particularly social spending targeted to the poor (safety nets). Total social spending has increased as a share o f the budget, but declined 32% inreal terms. For spending targeted to the poor, spending has been increased by 21 percent, a considerable achievement considering the budgetary problems facing the Government duringthe crisis. Inaddition, it i s inmarked contrast to previous recessions when social spending for safety nets was actually reduced. However, the large increase in the numbers of poor people duringthe crisis means that real spending per poor person has actually declined by 16%.The Government has made a major effort to organize a new workfare program, the Plan Jefes, which provides a payment of A$150 per month for those engaged incommunity work or training. This program presently has about 2 million beneficiaries, and may be the explanation of why unemployment declined between May and October 2002. Budget cuts have had serious consequences inthe social sectors. Ineducation, spending designed for investments in improved services have been diverted to shorter-term safety net programs, such as school feeding. Shortages of funds have producedteacher strikes and the interruption in learning. Insome cases, teachers have been paid with provincial government issued quasi-money instruments. The severe budget problems have prevented the continuation of the education reform strategy launched by the Government in 2000. Inthe health sector, the health insurance system which depends on wage contributions from formal sector workers, has seen major reductions inrevenues. As a result, healthproviders (the obrus), have not been able to provide the mandatory healthbenefits. Revenues for the PAMI, the health system for retirees, have fallen by 40%, also resulting in cutbacks. The result has been an increase in the demand for services from public hospitals, which normally provide services to those without health insurance coverage (workers inthe informal economy). The public hospitals, which are generally runby the provinces, have already seen areductioninfundingfrom bothprovincial sources and federal transfers. Household surveys also reveal that many people are spending less on preventive care, such as taking children for immunizations and checkups. More broadly, the use of quasi-monies for the payment o f pensions and public wages has often resulted inreductions in real incomes, since many of these provincial quasi-monies have been trading at a discount. The tariffs for basic utility and transportation services were `pesified' (fixed inpeso terms) and have remainedfrozen since January 2002, incontrast to other basic goods whose costs have risen on the order of 30% on average. Nonetheless, as a result o f falling incomes, the poorest households face significant problems in affording these basic services. The first quintile i s currently spending 22% of its income on water, energy and telephone calls, while about 20% to 25% of these households report having been disconnected from electricity and telephone as a result o f non-payment. At the same time, the policy o f freezing tariffs threatens the financial sustainability of the infrastructure service providers, who have already defaulted on about half of their US$16billion stock of foreign currency denominated debt. Some tariff increase for these services therefore appears to be inevitable, and raises further concerns about affordability. .. 11  Reversing the Trends Policies for the Immediate Future - A recovery of the economy has begun, butthe future remains uncertain. While other countries have hadrobust recoveries from crises, such recoveries take a year or two to develop. At the sametime, the crisis itself constrains the actions of the government interms of how it mightrespond to the social problems createdby the crisis: resources are scarce and options are limited. Macro Policies. Clearly, the root cause o f the recent growth inpoverty and unemployment can be traced to the problems of weak economic management, and the consequent slow economic growth. A primary concem i s to find policies that will stimulate a rapid and sustained recovery, but in a way that also increases employment o f low-skilled poor people. These policies include strengthened public finances, improved governance, providing a clearer regulatory environment for the private sector, restoring a healthy banking system, and clearing public sector arrears and normalizing relations with external and domestic creditors. The move to a floating exchange rate, and the consequent reduction inreal wages, has provided the basis for a potential recovery. The reduction inreal public sector wages and pension obligations with inflation, while nominal revenues increased, has temporarily provided for an improved fiscal balance. The government, however, still faces serious issues infiscal and financial sector management. Past actions, including the freeze on deposits, and the break o f the dollar-peso link, have reduced faith inthe financial system and eliminated the sources of credit needed to finance a recovery. While the Government has lifted the freeze on checking and savings deposits (the corralito),the financial system remains weak. The task of generating a adequate revenues to cover the costs o f the financial sector while providing social services will be difficult without efforts to broadenthe tax base and better enforce existing tax obligations. O f particular importance inthe longer term would be a strategy of moving away from the regressive value-added tax to greater reliance on a progressive income tax. A stronger tax base would also permit a reduction intaxes on some key factors (exports, labor) while permitting an expansion of social programs while maintaining a balanced budget. Overall, good macro policies, leadingto sustained growth, remain the single best route to poverty reduction. Safety Nets/Severance Payments. Government programs still do not provide an adequate safety net for the unemployed on a permanent basis. Workers have had to rely on their own savings, and their severance payment benefits, as their major safety net. Duringthe past crisis we have seen that both were inadequate: savings accounts were frozen, and severance benefits were reduced through negotiationsbetween workers and employers. Emergency public programs have helped. However, while the expanded Jefes program has absorbed 2 million unemployed, and it accounts for 7.5% of the current expenditures of the national government, it has reduced unemployment by at most 4 percentage points (assuming that all entrants to the Jefes programwere unemployed before, which i s not true). Ingeneral, social expenditure programs still tend to fall with a crisis, just when the numbers of poor and unemployed increase. However, it is also true that social programs, once established, tend to be viewed as an entitlement, and are difficult to cut once a recovery takes place. Safety net programs are scattered across the government, and many programs are costly and duplicate other programs. Smaller, inefficient programs should be combined. ... 111  A major safety net for workers inthe formal sector is the severancepayments owed to formal-sector workers uponjob termination. For bankrupt firms, it i s likely there could be no payment at all since these obligations are not pre-funded. A better solution would be a major expansion in unemployment insurance, with either a fixed benefit, or a benefit based on some proportion of average wages, and a fixed limit on total payments. Any such reform should be extended to include more of the employees now inthe informal sector, who have neither severance payments nor unemployment benefits and are outside of the social security system. Such an extension, however, implies that these employees move from the informal to the formal system, and pay social security andother wage taxes, andare covered by the labor code. This would further reduce labor market flexibility, given the rigidities inherent inthe present labor code, Thus, unemployment insurance reform should be accompanied by the liberalization of the labor code, which itself will bringbenefits inmakingthe labor market more flexible, and reducing labor costs. Education. The provision of educational services needs to be maintained, particularly with regard to payment of teachers, as the welfare effects go beyond the immediate income levels of teachers. While demand for social assistance (cash and food transfers) has accelerated, diversion of education funds to provide these short-term needs may conflict with the needto reach educational objectives. While cash-transfer programs may be helping to maintain school attendance, their effectiveness needs to be evaluated against other strategies for improving education. Federal government education allowances to provinces should be adjusted to insure equity between provinces, with a bias toward those with the greatest poverty. Highpriority shouldbe given to the strengthening of information systems and the use of them to inform educational policy. Such measures should be complemented and strengthened through the encouragement of community participation. Health. Inthe midst of the on-going social crisis, the first priority inhealth is to guarantee access for all of the population, particularly the poor who lack insurance andhave less ability to pay, to a minimumpackage of treatments andmedicines aimed at the major health risks. Inthis context, the Government should protect financing o f priority programs such as those for maternal-child health, immunizations, andpublic health (epidemiological surveillance, control of transmittable diseases and others). More specifically, the gradual implementation o f an infant and maternal health insurance and the definition o f and agreement on provincial health goals (with monitorable health indicators) seem to be two key initial steps towards more effective protection for the poor uninsured. Other needed government actions must simultaneously address chronic inefficiencies in the health sector, freeing up scarce resources to support the above and improving the effectiveness of health care for the poor. Main actions here include eliminating inefficiencies and overlaps innational-level insurance programs such as PAM1and the Obras Sociales, and in the interface between the public hospitals and the social security system (enhancing cost recovery by the former); strengthening the regulatory framework o f the health sector, as well as the institutional capacity of the regulatory agencies; and improving the performance o f the public hospital network, emphasizing primary health care. i v  Longer-Term Poverty Reduction Measures Inthe longer-term, anti-poverty efforts will needto focus on two broad areas: 4 First,reforms andpolicies that will leadto a pattern of growth that will be more rapid overall, and feature a higher level of employment per unit o f output. 4 Second, improving the access of the poor to basic services, that will both raise their overall welfare and, by improving their humancapital, improve the productivity andtheir ability to compete in an increasingly globalized economy; Generating Labor Intensive Growth. Macro-economic policies that permit rapid and stable economic growth' without inflation are an essential first step to a significant decline in poverty. This i s more likely to happen if Argentina's labor markets operate efficiently. However, Argentina's labor market i s one o f the most rigidand regulated in the developing world, preventing wage adjustments from taking place easily. Some o f key short term reforms that would facilitate a more orderly operation of the labor market include: ... elimination o f centralized or sectoral collective bargaining agreements which are automatically extended to all workers in a sector, even if not signed and even when expired; reduce the highcost of labor by reducing labor taxes, including severance payments; allow temporary employment that i s not subject to payroll taxes, as under the former modalidades promovidas, and 4 extending programs which permit exceptions for small scale enterprises. Inthe longer term, the critical problemremains that a large part of the labor force inthe informal sector lacks any form of pension or unemployment insurance coverage. A major reform o f the labor laws that would reduce their present highlevel of protectiveness should be followed by an extension of at least minimumcoverage to small firms and others inthe now informal sector. It would also help reduce the gap inwages between the formal and informal sectors, and thus help make more equal the present unequal distribution of income. IncreasingAccess to Services. A major effort should be undertaken to raise the level andquality of education available to the poor, and increase their accessto secondary andhigher education. One o f the key problems i s that children of poor families are more likely to drop out of school for various reasons. A viable strategy in education would include: . 4 greater investments in secondary schools inpoor neighborhoods; expanded cash grants to poor families conditional on keeping children in school particularly at the secondary level, in order to offset the economic incentives from school leaving and the effects of unemployment; 4 the establishment of a system of partial cost recovery from students at public universities, . who generally tend to be from non-poor families, and the establishment of a nation-wide system of scholarships for students from poor families. expanding the capacity of the current public university system, both by improvements in operatingefficiency and through further investments. 1The topic of how to attain sustained economic growth, however, go beyond the scope of this report. V  While the situation inhealthcare i s less critical, nevertheless poor, informal workers mustrely on the public hospital system, while formal sector workers have health insurance. Greater efficiency inthe health sector could improve the quality of service available to the poor. Particularly, the Govemment should focus public health care expenditures on those without health insurance, by improving cost recovery from those with insurance and the ability to pay, and by improving the operating efficiency of the public hospital system. While granting more autonomy to public hospitals can improve their efficiency, care needs to be taken to avoid buildinginincentives that will reduce services to the poor. Eventually, health insurance coverage should be extended to those inthe informal sector not presently covered. Existing programs of maternal and child health (PROMIN) need to be expanded, andlinked with family planning and reproductive health services for the poor, inorder to reduce their currently highrate of fertility among the poor. Deficiencies ininfrastructureboth reduce the productivity of the poor, and limit human resource development. The urban poor live inareas usually devoid of adequate sanitation and safe water, and often without paved roads. Provision of such public services inpoor neighborhoods can improve health outcomes. But attention also needs to be focused on building upcommunities, especially inurban areas, that lack roads, lights, and other services, and do not have legal titles to their land. The urban poor are particularly vulnerable to problems of crime and violence, and attention needs to be paid to alcohol and drug abuse, and improvements in police protection and access to justice. There i s considerable consensus that, inorder to resolve the dilemma between the affordability of basic infrastructure services and their continued financial sustainability, some kindof social policy is requiredto protect the poorest from the impact of any future tariff increases. Lookingfurther ahead, as incomes recover and affordability improves, the focus of social policy for the infrastructure sectors should switch away from social tariffs andtowards social connection programs, particularly inthe water, sewerage and natural gas sectors, where considerable coverage deficits remain and those without access suffer substantial economic and social costs. vi  Policy Actions for Reducing Poverty Area Actions Needed Establishment of Macro policies that will provide a stable environment for sustained, pro-poor private sector-led growth, including a healthy financial economic growth system, and an efficient regulatory network for infrastructure. Improvingthe fiscal base to be able to fund social programs directed at the poor, and moving away from the regressive value-added tax to a more progressive income tax. Establishment of a strong Providing for a more secure form of unemployment safety net compensation, such as unemployment insurance or pre- funded severancepayments. Improvedevaluation of existing safety net programs, and consolidatiodclosing of weak and ineffective programs. Reforming the labor code to make it more flexible, and thereby allow an extension of coverage to those inthe informal sector. Improvingeducation, Health: health and infrastructure o Protect and restore key programs benefitingthe services poor o Develophealth insurance for people inthe informal sector, esp. maternavinfant o Eliminate inefficiencies and overlap between systems o Increasethe efficiency and autonomy of public hospitals Education: o Maintain payments for teachers and operating costs o Focus provincial allocations onthe poorest regions o Increase investments insecondary schools o Provideincreased cash grants to poor families to promote secondary school attendance o Expandthe public university system Infrastructure and other Services o Expandbasic infrastructureinpoor urban areas, including water, roads, electricity; o Adjust pricingof infrastructure servicesto benefit the poor; o Take steps to reduce crime and violence inpoor neighborhoods. vii   ARGENTINA :CRISISAND POVERTY 2003 CHAPTER1 THEEVOLVING CRISIS AND THE POOR Background Since the last Poverty Assessment was written (March 2000),2 Argentina has gone through a severe economic and financial crisis. After three years o f continuing recession, the economic and financial crisis deepened during 2001. Various attempts were made throughout 2001to spur growth as a preludeto improvedpublic finances andthe debt profile but to no avail. Market perceptions of risk of default increased, especially inmid-2001. Efforts to protect the banking system from wide-spread withdrawal of deposits (the corrulito) resulted instrangling of liquidity and economic activity and generated popular discontent. This finally culminated inthe resignationof then President de la Rua inDecember 2001, a quick succession o f appointed Presidents, a formal announcement of default on public sector borrowings, and inearly 2002, the abandonment of the Convertibility Plan. Now inits fourth year o f recession, the economy in 2002 i s estimated to have declined by 10.9 percent, bringingthe decline in GDP to over 20 percent since its 1998 peak. With the peso depreciating sharply since its float Argentina is experienced significant inflation for the first time since 1991. Between the decline ineconomic activity and the currency's depreciation, per capita GDPhas declined sharply, to an estimated US$2,695 in2002 (down fromUS$8,210 at its peak in 1998). The previous Poverty Assessment focused on the periodbetween 1990 and 1998. During that period, the government's stabilization plan, including the convertibility plan for the peso, generated economic recovery, stable prices, and declining poverty up to 1995. Between 1995 and 1998, poverty rates and unemployment rates both rose, and income distribution deteriorated. On the basis of major urbancities, the poverty rate went from more than 40 percent in 1990, to a low of 22 percent in 1994, to 29 percent in 1998. ~~ World Bank. 2000. "Argentina: PoorPeopleinA RichCountry." ReportNo. 19992-AR. Available on the web at httD://www.worldbank.or8/lac, search under Argentina. 1  The Relative Size of the Argentine Crisis in2002. The table below shows some indicators for various countries in crisis, including Mexico during the Tequila Crisis, Four of the East Asian Tigers during the crisis of 1997-98, Russia in 1998199,and Argentina itself in 1995. In comparison with these crises, Argentina i s second only to Indonesia interms of decline in GDP and the drop in real manufacturing wages, ranks first in terms of the drop in the exchange rate, and one of the highest in terms of the resulting rate of unemployment and increase inpoverty. Table 1.1Comparative Statistics on GrowthandPoverty Source: Fallon and Lucas (2002), and World Bank staff estimates. Clearly, this situation, which was not very good in 1998, has become even more difficult as of 2002. Preliminary evidence shows that poverty now (October 2002) has reached the level of about 58 percent for urban areas (see figure 1.1).3Unemployment has risen from 13 percent in 1998 to a peak o f 22 percent inMay 2002 (inurban areas). Compared to other countries experiencing crises, the parameters for Argentina are more severe than most interms of poverty, unemployment and falling wages (see box and Table 1.1). Falling revenues have further cut social programs, adding to the poverty of the poorest, and leading to increasing social unrest, crime, and violent political demonstrations. Growing uncertainty has ledto a decline inforeign direct investment and a cut-off in access to foreign capital, further reducing the prospects for long-term growth and recovery. This report will not attempt to cover all of the ground inthe original Poverty Assessment4.Rather, it will focus on broadtopics that are relevant to the current crisis: 0 First, it will examine the trends inpoverty, income distribution, andunemployment. 0 Second, it will look at how people have adjusted to the crisis, how it has affected the poor, and what types of coping strategies they are using. Third, it will look at how the crisis has impactedgovernment spending programs, particularly those considered to be pro-poor and that constitute the social safety net. 0 Finally, it will look forward and suggest what might be done better in the future to reduce poverty Data since 1998 uses a poverty line developed by INDEC. For a summary of the original poverty assessment, see box. 2  Summary of the Poverty Assessmentof 2000 This report found that since 1991, Argentina had gone through aperiod of adjustment that had led to a remarkably sharp drop in rate of inflation, the privatization of state-owned industries, and the opening of the economy to foreign commerce. Progress on the economic front produced real gains in terms of reducing poverty and improving welfare. Poverty rates fell from 40% in 1990 to a low of 22% in 1994. However, since 1995 poverty had grown slightly as a percentage of the population, and income distribution had deteriorated. By 1998 the poverty rate for urban areas was estimated to be 29%. The deterioration of income distribution reflects the fact that while overall growth had been positive, and average per capita income had risen, the gains had gone largely to the more skilled and educated inthe labor force, and not to the poor. The report noted that, in general, poor families have low levels of education, have a large number of dependents, and are younger than families that are not poor. Large family sizes are the result of much higher fertility rates among poorer women, a factor that tends to perpetuate poverty. They live in areas lacking often in water and sanitation services, roads, and other public amenities, live in areas affectedby flooding, and live in overcrowded conditions. The often lack titles to the land they inhabit, and therefore lack the incentive and the collateral to invest in their housing. The distribution of urban services is uneven between urban areas; some seem to do a betterjob than others in meetingthese basic needs. The report found that Government programs that were specifically targeted to the poor generally worked well, and were well targeted. The more general problem was one of coverage. Only about 25% of poor families received any form of direct public assistance, in the form of cash, food, etc. The report estimated that public and private transfers together probably reduced overall poverty by 4 percentage points, and were particularly important for the elderly. Shifting demand for labor had put a high premium on education. While rates of return to primary education were extremely low (about 3%), returns to tertiary education was estimated to be 29% (1998). Despite these high returns, the poor often did not complete secondary school and were underrepresented in higher education. Repetition rates were high, as were dropouts. Only 24% of those aged 18-24 among the poor we found to have a secondary education. The low quality of education, particularly inpoorer areas, and the needto work, all worked against school completion. The report noted that limited information available suggestedthat there was substantial poverty among the rural population, particularly inthe Northwest andNortheast. Most of these poor were not farmers, but farm and non-farmworkers who were often unemployed and lack skills and education. The indigenous people of the rural areas seemto be particularly poor, since they live in remote areas away from public services. Source: World Bank (2000)Poor People in a Rich Country: A Poverty Reportfor Argentina, World Bank report no.199962- AR, March25,2000 Poverty Trends Poverty i s measured usingthe government's poverty line and the Encuesta Pemanente de Hogares, the biannual householdsurvey covering about 30,000 urban households. Data for individual urban areas are combined to produce a national urban poverty estimate, but these urbanareas themselves cover only about 70percent of the urbanpopulation. Thus, the surveys leave out the population in smaller urban places and all of the rural population, much of which 3  probably has a higher than PovertyLines inArgentina average poverty rate5. Thisreport classifies the poor with the income thresholds that INDEC Despite improvements in has adopted as its official poverty lines since their first introduction in poverty measurement inthe 1988. Poverty lines consist of the basic food basket and nonfood consumption bundle whose combined values arejust sufficient to allow past, Argentina still lacks a a typical individual to achieve a minimumlevel of material welfare. The comprehensive living basic food basket was constructed by looking at the food consumption standards survey that patterns of a reference household group between the 2"dand 4" income adequately covers rural and deciles in the 1985-86 Survey of Income and Expenditure administered urban areas. Inaddition, all for Metropolitan Buenos Aires (GBA). The indigent food line is the value of the basic food basket that would allow a representative adult of the numbers on poverty male age 30-59 years with moderate activity level to achieve reported here have the consumption of a daily energy intake o f 2,700 calories. Updated for drawback of being based on 2002, the adult equivalent indigent line for residents in Metropolitan income. It i s widely Buenos Aires i s estimated to be 81.8 pesosper month in April 2002 (or recognizedthat consumption about US$28.80). Observation of nonfood consumption among households in the same group gave the empirical parameter (inverse i s a better basis for measuring Engel coefficient) to estimate the adult equivalent poverty line, which welfare, since it allows for was estimated to be 193.8 pesosper month (or about US$68.23; all inter-temporal adjustments, conversions are at the April 2002 exchange rate of 2.84 pesosper and i s more accurately dollar). measured than income. Between 1998 and 2001, there was a steady increase inboth poverty and unemployment. The poverty rate rose from 28 percent of the population in May 1998 to 37 percent by October 2001,as the economy stagnated and the economdfinancial crisis gradually worsened (see Figure 1.1and Statistical Annex Table A.1).Likewise, unemployment rose from 13 percent to 18 percent. The collapse of the convertibility plan, major exchange rate devaluation, and default on extemal debts in January 2002, however, produced a sharp and sudden increase inpoverty. By October 2002 the poverty rate reached 58 percent (see Fig. 1.1). Unemployment peaked at 22% inMay, butdeclinedto 18%inOctober. At the same time, extreme poverty, definedas those without sufficient incomes to buy a basic basket of food, has gone from 6 percent to 28 percent. Few countries inthe world have seen such a rapid rise inpoverty-an increase inthe poverty rate of more than 50 percent. O f particular interest i s that the implied poverty elasticity, which was estimated to be about-1.8 inthe 1998report, appearsto be about-3.0 duringthe crisis.6 'A limited survey of 2,800 household undertaken by the World Bank in 2002 found a poverty rate of 53 percent in urban areas and 73 percent inrural areas (see Fiszbein, Giavagnoli, and Anduriz, November 2002. "Argentina's crisis and its impact on household welfare", Background Paper No. 1, found in Part I1of this report.) Assuming -1 1percent growth in 2002 and population growth of 1.3 percent, per capita income will have fallen about 12 percent. Assuming an average poverty rate for the year of 55%, the rate then increased 49 percent, or an implied elasticity of about -4.1. The estimate of -1.4 is based on the longer, 1990-98 period. This illustrates the danger, however, of using a fixed poverty elasticity based on historical relationships for estimating short-term impacts of macro crises. 4  Figure1.1.Unemploymentand Povertyas Percentageof Population,andRealPer Capita HouseholdIncome,1990-2002 70.0 350.0 60.0 300.0 $ 2c 0 250.0 .E 0 s5 50.0 8 40.0 200.0 -c: 3 83 Q h 0 0 , 30.0 5 150.0 2$5 5Q .t= m 0 & 20.0 100.0 3& Q 0 10.0 50.0 Q ti5 0.0 0.0 The sharp rise inpoverty can be attributed to two broad effects: higher prices as a result of the devaluation, particularly in food andother wage goods, and unemployment. Unlike other countries, where basic foods are not traded commodities, inArgentina basic foods are traded, andthus the devaluation hada direct impact on livingcosts. Poor people, with highlevels of food in their consumption basket, were affected particularly hard.Thus, the poverty line for Buenos Aires increased between April 2001 and April 2002 by 26 percent, while the indigence line (purely food) went up by 29 percent. B y contrast, the total consumer price index went up by only 18 percent in the same period. Kritz7estimates that every point rise inthe price index for the basic food basket produces 50,000 more indigent poor inthe country. Furthermore, he estimates that the decline inreal income between May 2001 and May 2002,30 percent can be explained by the fall inemployment, 20 percent by a decline innominal wages (including fewer hours worked), and 50 percent by the increase inprices (see figure 1.2). 'Ernest0Kritz,December2002, "Poverty andthe Labor Marketinthe Argentine Crisis, 1998-2002," Background PaperNo. 4, Part11. 5  While all incomes have fallen, the incomes of the poorest groups have declined Figue1.2 FactocsExplainingtheM i n einRealLaborIncane more than those o f the richer groups. Thus, the "-" w tendency for income distribution to worsen, noted inthe 2000report, has accelerated duringthe crisis. As shown intable 1.2, the msddmMlin share of the top 20 percent income distribution Ektdthe "rt30% increaseinpicts FC6@h&Bbm has risen from 54.8 percent in 1998 to 57.2 4% percent in October, 2002. Conversely, the share of the poorest 20 percent has fallen from 3.8 percent to 2.8 percent.* The Gini 21% coefficient has risen from 0.504 to 0.532, placing Argentina near the middle of the list of S r u a r D u 4 p d ~ ~ l i o n ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ - Latin American countries rankedby income distribution (see table 1.3). Because the Argentine data exclude rural areas, which are M a y October Year 1990 1992 1994 1996 1998 1999 2000 2001 2002 2002 Per capita income by decile(in 1999 pesos) 1" decile (lowest) 38.3 45.6 43.2 28.1 31.7 30.4 26.3 17.1 9.0 16.1 5thdecile 153.4 181.1 188.1 158.1 167.3 164.2 155.1 136.3 97.9 95.8 lothdecile 825.4 1,004.7 1,060.7 992.3 1,114.2 1,028.2 1,041.9 993.8 769.9 705.3 Relative measures Top 20% share 50.7 51.0 51.6 53.7 54.8 53.8 55.1 56.8 58.2 57.2 Bottom 20% share 4.6 4.5 4.2 3.6 3.5 3.5 3.2 2.6 2.1 2.8 Gini coefficient 0.454 0.456 0.467 0.493 0.504 0.494 0.510 0.530 0.551 .532 Tophottom 20% ratio 11.0 11.3 12.3 14.9 15.7 15.4 17.2 21.8 27.7 20.4 The potentially explosive situation with regardto income and income distribution can be better measured by considering the actual real incomes of each income group. As shown in table 1.2, the real incomes of the lowest 10percent have fallen 49 percent since 1998. The richest 10 percent have experienced a decline inincome of 37 percent. Inboth cases, most of the decline has come just inthe last year. O f particular interest i s the very major change between October 2001 and May 2002, when income distribution worsened, and then the subsequent improvement between May 2002 and October 2002. Over this period, the income of the lowest 10% fell 47% inrealterms, andthen increased by 79% (see Table 1.2). It is impossible to identify exactly what caused these sudden movements, but the implementation of the Government's Plan Jefes program, which provides income supplements o f A$150 /month to about 2 million people might explain part of this the sudden improvement. * These ~ are based on the data for October from the Encuesta Permanente de Hogares (EPH) surveys. 6  It is important to note that the discussion of income distribution does Table 1.3. GiniCoefficientsfor Various Countries not consider the impact of any changes (data for 1998income,unlessstatedotherwise) inthe value of assets. With the freeze on bank deposits many people were Country Gini Latin-America unable to withdraw their funds, which Brazil .61 were subsequentlyeroded by inflation. Colombia* .58 Those holdingdollar deposits saw their Chile .57 deposits converted into pesos. Thus, MBxico .53 many of the non-poor who depended Argentina (urban, 2002) .53 Venezuela S O on savingsinbanks as a type of self- Dominican Republic .47 insurance against risks, found that they Uruguay (urban) .43 were unable to use their personal Other countries safety net. The middle class were also France (1995) .33 affected by defaults on bonds, and Russian Federation .49 Japan (1993) .25 declining real estate prices. Since the States (1997) .41 poor have few, if any, bank deposits/ *United consumptiondata bonds/ real estate they may have been Source: World Development Indicators,2002 (World Bank, isolated from some of the worse WashingtonD.C.), table 2.8, except for Argentinawhich is based onBank staffcalculationsfor 2002, and Uruguay,which aspects of these adjustments.But these comes from the most recent povertyreport of the World Bank factors could be important for the (ReportNo. 21262UR). middle-class who are now poor, or the "new poor". There i s some evidence to suggest, furthermore, that large depositors were able to foresee the coming devaluation andmake adjustments, while small depositors got caught by the corralito andthe devaluation. It is difficult to say, however, how many of these small depositors were Table 1.4 PovertyProfileinUrbanArgentina, 2002 Poverty Group Indigent Moderate poor Non-poor Overall Family size 4.9 4.0 2.7 3.5 Adult equivalent family size 3.8 3.2 2.2 2.8 % female-headed households 24.0 22.8 32.1 28.1 Average age 23.3 29.2 38.0 31.3 Years of schooling among adults aged 20 and over 7.9 9.3 11.7 10.2 Labor force participation (among those of age 15-64) 59.2 62.2 69.8 64.9 % informal sector 39.0 39.5 33.2 36.3 Unemployment rate 34.9 20.7 9.7 17.8 Hours worked 29.1 38.6 42.0 38.7 % working < 30 hrslweek 37.8 25.9 17.5 1 24.6 Unemployment is clearly a phenomenon that affects poor families severely and i s a major component inexplaining their poverty. If we divide the population into the indigent poor (those below the food poverty line), the moderate poor (those between the indigent line and the normal See Halac and Schmukler (2003) for a discussion of these possibilities in both Mexico and Argentina. 7  poverty line), and the non-poor we see certain clear characteristics. The unemployment rate among the indigent i s close to 35 percent, compared with only 21 percent for the moderate poor and 10percent for the non-poor (October, 2002, see Table 1.4). While unemployment has risen since 1998 from 12percent to 18 percent, the unemployment rate for each subgroup has remained about the same. This situation suggests that as people move down to a lower group they take on the unemployment rate of that group. Thus, total unemployment rises without a rise inthe unemployment of each group. It also suggeststhat many of the "new poor" areworking at lower incomes, and that unemployment i s only part of the cause of their poverty; the other part i s clearly lower incomes. There i s no evidence that rising unemployment has led to lower labor force participation (discouraged worker effect). If anything, there has been a slight increase in labor force participation among the indigent and moderately poor, but the increase i s only about 1percentage point compared with pre-crisisconditions (see Table A.2). Another important characteristics o f the poor are their large family size and high dependency rates. The average family size for the indigent poor i s 4.9, compared with 4.0 for the moderately poor and 2.7 for the non-poor (see Table 1.4). As a consequence, a large number of children are poor". Thus, while the overall poverty rate rose from 29 percent to 53 percent over the 1998-2002 interval, the poverty rate for children (age 0 to 14 years) rose from 45 percent to 70 percent. However, it i s also useful to note that female headed households tend be less poor than male headed households. But poverty seems more related to labor market characteristics; not only i s the unemployment rate of the indigent poor much higher than the non-poor (see box), average hours worked are also much lower; 29 per week for the indigent compared to 42 for the non-poor. About 38% of workers inindigent poor families work less than 30 hours per week, compared to 26% among the moderately poor, and only 18% among the non-poor. Thus the poor are poor not only because they unemployed, or employed inlow payingjobs, but becausethey cannot find enough work to fill a 40 hour week. An important characteristic of the poor is that inboth good times andbad, they face not only lower income, but also more unstable incomes. Usingpanel data that allow a view both forward and backwards, we can estimate deviation from mean income by income class." As shown infigure 1.3, the coefficient of variation (CV)12 of the poorest quintile have historically been higher than those of the upper quintiles. Furthermore, the gap i s increasing, indicating that the income o f the poor has become more unstable and uncertainnow than inthe past, while the income o f the richer groups has maintained its relatively stability. InMay 2002, the C V of the poorest group i s roughly twice that of the upper income quintiles. We also know that poorer groups have more limitedcapacities for consumption smoothingby using credit or savings. Hence, they face low and unstable incomes with little private means of coping. While risk and poverty are related, there are differences. Education seems to reduce risk, independent of its ability to raise income. People with higher risks include migrants, households with informal loHowever, this conclusion derives from a poverty estimate that does not allow for economies of scale in household consumption; that is, large families can realize economies of scale in the consumption of housing and utilities that improve their welfare over small families. Recalculations of poverty with the introduction of various scale economy estimates indicate that over the most feasible range for the scale economy coefficient large families are still poor (see H.Lee, Background Paper No. 8). Calculated from data in G. Cruces and Q.Wodon, "Risk Adjusted Poverty in Argentina: Measurement and Determinants." Background Paper No. 2, Part 11. l2The coefficient of variation (CV) is the standard deviation divided by the mean. 8  workers, and/or unemployed andinactive workers. The elderly face low incomes but also low risks, probably becausetheir incomes derive from more stable pensions. To the extent that risk reduceswelfare, one could convert data on variability into a risk- adjustedmeasure of poverty. With such a measure, the gap betweenrich andpoor will be greater given that the poor have more variability than do the poor. For a discussion of such adjustments, see Part 11,Background PaperNo. 2. Figure 1.3. Coefficientof Variation of HouseholdIncome by Quintile Greater BuenosAires, Waves 1995-2002 0,700 - 0,600 -- / 0,500 - 0,400 - -' / 0,300 0,200 0,100 - May 95 Oct, 95 May 96 Oct, 96 May 97 Oct, 97 May 98 Oct, 98 May 99 Oct, 99 May 00 Oct, 00 May 01 Oct, 01 May 02 -Q1 -Q2 Q3 -Q4-Q5 Labor Marketsand Adjustment The current crisis has createdmajor changes inthe labor market,13and many of these changes have impacted the poor adversely. During the convertibility period, adjustmentsto the real exchangerate could only occur through deflation of nominal wages and prices, becausethe nominal exchangerate could not be adjusted and inflationcould not be usedto reducereal wages. As the country moved into a state of increasing crisis, nominal wages did not fall, and labor market adjustmenttook the form of steadily risingunemployment. Despite the fact that wages were more flexible inthe informal sector, such wage flexibility was not sufficient to maintain full employment. Rapid growth inthe early years meant that output growth exceeded labor force growth, and productivityrose (see table 1.5). During the period 1991-96, productivity grew by 7.6 percent per annum, and the employment elasticity was very low at 0.14. During 1996-98, employment grew at a record 5.2 percent per year, implying an elasticity of 0.6. But since the Tequila Crisis, the demandfor labor has becomeeven more pro-cyclicalinboth directions. With a fall in output of 2.8 percent in 1995, employment fell 2.9 percent, and unemployment reached 18.5 percent. l3E. Kritz, BackgroundPaperNo. 4. 9  Table 1.5. Growthof GDP,Employment,and Unemployment I 1994-2002 Year Growth rate of real Growth of Growth of GDP employment unemployment 1994 5.8 -0.5 34.9 1995 -2.8 -2.9 34.4 1996 5.5 0.9 21.4 1997 8.1 6.3 -10.0 1998 3.9 4.1 -7.4 1999 -3.4 .9 .7 2000 -0.5 0.4 12.5 2001 -4.4 -0.7 16.2 1* Excluding 2002 -10.9 -1.1 18.8 2002* -10.9 -3.6 29.3 impact of Plan Jefes. Source: INDEC (EPH), various years, as calculated by Kritz. Inthe post-convertibility period of 2002, with a resumption of Microeconomic Labor Management: What Have Enterprises Done Since Devaluation? inflation, real wages have taken more of the adjustment, and there has been A special survey of 200 leading enterprises reveals some a relatively smaller increase in interesting facts about how enterprises have adjusted employment unemployment. Output i s estimated to and wages to the crisis, including the following key points: have fallen 11.O percent in2002, One out of three enterprises reduced its labor force, although this tendency has decelerated inrecent months. while employment has fallen 1.1 The average labor force reduction was 17percent. percent (see Table 1.5). The small Practically all leading enterprises took some action in relation drop inemployment may be to the to wages. On average, general adjustments produced a 21 impact of Plan Jefes, which alone percent increase over base salary. This includes the 100 employs now 2.0 millionpeople. pesos decreed by the government, which alone represented Without Plan Jefes employment between 8 percent and 10 percent of base. This means that in leading enterprises the adjustment doubled the amount would have fallen by 3.6%, still established by the decree. But even so, it represents half of relatively small when compared to the the rise in consumer prices accumulated in the first three 11% drop in GDP. However, there i s quarters. some evidence to suggest that as Seventy percent of firms made adjustments on non-salary many as 40% of participants inPlan costs. The most frequent measure was to reduce training programs, but they also took actions that affected the income Jefes were not inthe labor force and of employees: 50 percent reduced overtime hours and 31 unemployed before participating in percent suspended them completely; 35 percent reduced the program. Hence, adjusting benefits such as cars, cellular telephones, and others; 33 employment figures to exclude Plan percent reduced pension plans and 19 percent suspended Jefes may not be entirely appropriate. them; 32 percent lowered rewards and bonuses and 15 percent suspended them. Last, but not least, three out of four enterprises suspendednew personnel hires. As notedin the Poverty Seventy-nine percent of layoffs were mutually agreed upon. Report of 2000, formal sector Only in 21 percent of the cases was there a dismissal, or a employment tended to be relatively termination of the labor contract decided by the enterprise. stable during the 1990s.Most of the This policy is partly related to corporate image , but also unemployment was generated by the reduced the size of the severance payment that firms had to informal sector or byjobs of lower Pay. productivity. However, inthe present Source: Kritz, E. BackgroundPaper No.4, Part 11. 10  crisis job losses have been predominant inthe formal sector. BetweenMay 2001 and May 2002, 825,000 jobs were lost, of which 90 percent were inthe formal sector and of which two-thirds were salaried employees. Of 14.3 million people that are inthe labor market at present (May 2002), only about 5 million have stable14employment; of these 1.7 million belong to the public sector. Private, formal-sector, salaried workers constitute about 3.3 millionpeople, or about a fourth of the total labor force. Ifone adds private-sector, non-salaried (self-employed) workers, the private formal sector employs about one third of the labor force. At the beginning of the crisis (1998) the situation was different: unemployment was 9 percentagepoints lower, and the weight of the private sector inthe labor force was 8 percentagepoints greater. Conversely, the informal sector has grown duringthe crisis, adding 170,000 employees between October 1998 and May 2002. The growth ofjobs inthe informal sectorinthis periodwas concentrated inshort-term temporary jobs offering low incomes. Work consists of a few months or days, followed by periods of short-term unemployment. This type of intermittent work employs about 21 percent of the labor force. Thus, while total informalemployment hasrisen, informal, salariedemployment of a long-term nature has fallen at the same time. Hence, the current role of the informal sector has reverted back to a more traditional one of absorbing employment from the formal sector, and even from stable employment within the informal sector. The most stable sector of employment remainsthe public sector, of which about 15percent of all households dependinwhole or in part. This very stable source of employment i s generally enjoyed by the middleclass, however, not the poor. Only 6 percent of all the public-sector employees belongto the lowest 20 percent of the population. These changes are more important for the poor than for the non-poor. Of the 40 percent of the population with the lowest incomes, 60 percent of the economically active work in the informal sector; and of these, half are % Working between % Working more than intermittent workers. 10-30Hoursper week 30 Hours3er week These two lowest quintilesaccount for 55 Year Month Indigent Moderate Non- All Indigent Moderate Non- All poor poor Urban poor poor Urban percent of the Areas Areas unemployment. In 1998 May 18.7 14.1 12.6 13.2 32.1 54.1 73.4 68.2 addition, poorer workers October 17.6 15.5 12.8 13.6 30.2 55.0 74.1 68.4 tend to be employed less 1999 May 18.3 15.4 12.6 13.4 27.0 53.2 72.6 66.3 than 40 hours per week, October 18.4 15.2 13.0 13.7 29.4 54.0 72.2 66.6 andthe numberof 2000 May 17.6 16.0 12.5 13.5 26.5 50.0 71.3 64.2 underemployed has been October 15.9 16.4 12.9 13.8 27.7 51.3 72.0 65.1 growing steadily since 2001 May 18.3 15.7 12.6 13.7 24.3 52.2 71.3 63.5 the early 1990s.At October 18.3 18.1 13.5 15.0 24.3 46.2 68.9 59.4 present, only about 26 2002 May 20.9 17.7 14.4 16.6 23.7 49.4 68.0 54.1 percent of the indigent October 30.8 21.5 14.9 20.5 25.5 51.5 71.0 55.1 poor work more than 30 hours per week; the rest l4 Definedby Kritz as formal sector workers with permanentlabor contracts. 11  are either unemployedor partially employed (see Table 1.6). Even among the moderately poor, only about half have full-time work. Furthermore, the numbers engaged for 10-30 hours have continued to rise over the past five years (see Annex Table A.7). More flexible real wages, with inflation, during 2002 probably has helped to limit the amount of unemployment created by the crisis. This outcome can be seenby looking at labor costs and employment indifferent sectors. For the whole private sector, between December 2001 and September 2002, real labor costs (the real wage bill) fell 30 percent, or 6 percentage points more than the fall inreal wages. Although many formal sector firms increased nominal wages by even more than the 100-peso increase decreed by the government, these increases fell short of the increases inprices (see Table 1.7). However, changes inreal labor costs and wage bills vary greatly by sectors. Incommerce, for instance, sales fell 28 percent incomparison with a year earlier, real unit labor costs fell 26 percent, and employment fell by 11.5 percent. Thus, the total wage bill fell by 34.5 percent, and the labor cost per peso of sales is 9 percent lower than a year earlier. Accordingly, wages and other unit labor costs have become the main mechanism for adjustment inthe labor market, and account for the deceleration of unemployment growth inthe thirdquarter of 2002. As showninTable 1.7, real andnominal wages for formal sector employees were relatively unchanged after 1995 through 2001, but show a dramatic drop in 2002. For informal workers, nominal and real wages show more flexibility, but this flexibility was not sufficient to restore full employment. As inmany other countries, it i s difficult to reduce nominal wages. Table 1.7: Average Wages - Greater BuenosAires 1992-2002 (A$j e r month, cc ;tant and c rent prices) Salaried Salaried workers Formal Informal workers Formal Informal total workers workers CPI total workers workers Real (base = May 2002) Nominal 1992 772 840 634 82.27 1,109 1,207 911 1993 852 957 643 91.86 1,096 1,231 827 1994 893 1,016 644 95.84 1,101 1,253 794 1995 825 939 576 100.29 972 1,106 679 1996 808 935 543 100.08 954 1,104 641 1997 816 944 603 100.73 957 1,107 707 1998 825 970 573 101.91 957 1,125 664 1999 823 976 561 101.15 961 1,140 655 2000 813 979 529 100.07 960 1,156 625 2001 812 970 549 99.84 961 1,148 650 2002 616 769 363 118.17 616 769 363 The Law of Economic Emergency inJanuary 2002 suspended the right to terminate employees without justified cause and doubled the size of the applicable severance payment. Firmscouldrequest a reduction inthe severancepayment to 50percent of normal if they could prove to the Ministry of Labor that they were suffering from a loss of business resulting from economic conditions. However, Article 241 of the Labor Law allows for the termination of employment by mutual consent. Infact, a survey of 200 firms reveals that this i s the preferred 12  mechanismfor terminations. The negotiated settlement i s usually 50 percent to -75 percent of what is normally requiredby law. While this method reduces labor costs somewhat, and makes the system more flexible, it also tends to mean that the safety net afforded by severancepayments i s uncertain and negotiable. There i s currently a draft law inthe Congress to severely restrict the application o f Article 241 inthe future. 13  CHAPTER 2 HOW THE POOR COPEWITHA CRISIS Giventhe severity and depth of the present crisis, how have poor people coped with a sudden change intheir incomes? To better understand the dynamics of the social and economic situation, we carried out two surveys: one a survey of civil society organizations (CSOs)working with the poor, and the other a more quantitative survey of living conditions. Impactof the crisis inArgentineanfamilies A Surveyof CSOs. - The study of CSOs used consisted of focal groups, in-depth interviews and questionnaires with civil society organizations, with these or anizations representing both the structurally poor and organizations representing the new poor." It was carried out inthe most populatedurban areas of the country: metropolitan area of Buenos Aires, Cbrdoba, Corrientes, Resistencia, Rosario and Tucumiin (with regional representation). This survey ledto the following observations: The structurally poor are considered to be those who live in areas with deficiencies in public services, inbasic social services and dwell in sub-standard houses. They have a higherthan average number of children andhave children at earlier ages. The education level of the householdmembers i s low. Labor force participation of these groups i s concentrated inmanual tasks, with minimalqualifications and highlabor instability, usually inthe informal market. Impoverished middle-class groups (or the new poor) live inneighborhoods with more efficient public services, generally own their home and many of them have a car. They are characterized by having higheducation levels and qualifications. Their situation changed duringthe last decade due to the macroeconomic changes that took place, which left many o f them unemployed. An unprecedented characteristic of the current crisis inArgentina is that it has severely affected middle-class groups, and not only the structurally poor, who had already been sufferingfrom marginality and poverty for several decades. Ina society like Argentina, with higheducation levels, the presence of a new group o f poor people -- constitutedby those who traditionally had access to goods and to important economic and cultural services-- caused a major domestic shock. l 5Methodological development of the ERES team, based on the revision of methodologies used in several publications of the Bank, like "Voices of the Poor", "Empowerment and Poverty Reduction", and other literature produced by the Civil Society and Social Development team. Working with CSOs' representatives allows us to capture the social emergency from a point of view that, on one side, comprises the perception of the segment as a political actor, and, on the other side, has better capacity o f conceptualization, synthesis and analysis of community and family situations, and of relationships with other institutions. 14  Inthe face of this crisis, the new poor -- who were not usedto resortingto survival practices -- developed different strategies to manage it, while the structurally poor redefined their role before the state, butnot ina massive way (e.g., the Voices of the Poor -Argentina, 2002 appearance of the "ppiquetero" movement agitating for labor "For the young people it is very hard and I guess it has to programs). do with this, with the anger and toughness, the confrontation with those who are killing them, and they don't know what this is about. I think the youth feels Impoverishedmiddle-class groups despair and that their lives are going be miserable, or state that they have suffered a major their lives are going be just begging, stealing and while they go out on the streets, trying to find some way to get reduction intheir income due to the quick money, but if they don't get anything they'll be loss o f their jobs and decreased killed. You can hear this talk in all the youngpeople all the purchasing power, and have time and it's awful." implemented new forms of "Just go around the city of Resistencia.Before there were generating income like trading 2 or 3 comers withprostitutes but now we can see places everywhere and the saddest part of this is that these are products and services, organizing increasingly younger girls in prostitution who go into this informal markets to sell objects, as a way to survive. startinghome-made production and "Before, you didn't see that much violence. Now you see it micro-enterprises. They have also much more in young kids: they get high, kill and steal. All tried to put new members of the this because there is nothing holding them back. " "I'mdark, I'm young, I go downtown and Iwind up in trouble." household into the labor market. "Now there is even discrimination against you for your Other strategies include combining skin color in the agencies that are supposed to help us. efforts with friends and/or family Even inside and outside your own neighborhood: evenjust members in order to purchase goods the merefact of having a different haircut." by taking advantage o f sales and "Insecurity resultsfrom no education, no work, no health, wholesale prices, choosing more no housing. The government doesn't have any answers...... economic brands, replacing costly because the govemment's answer is an iron hand." non-essential products with cheaper ones, andincreasing the development o f home-made products. Structurally poor groups, who have been experiencing unemployment andprecarious jobs for a long time, have resorted to increased participation o f women and children in subsistence activities, like cardboard collection and other strategies. Some companies that have gone bankrupt have gone back into business as cooperatives (constituted by their own employees about to be fired) or as community micro-enterprises managed by grassroots organizations that incorporate staff financed by the Plan Jefes. The structurally poor are increasinghome-made production for self-consumption (preserved fruit, bread, pasta andtheir own-produced vegetables), and community purchases inwholesale stores. Nevertheless, even applying these strategies, they have had to adjust their consumption of basic products, e.g. reducing milk and sugar, andreplacing meat for soy, cereals, noodles or rice. Regarding access to services, these families make illegal use of electricity and substitute natural gas with bottled gas and/or firewood. They consider that services are too expensive, particularly when their only income is the subsidy they get from the Plan Jefes. 15  As the new poor are concerned, they have experienced an increase inoverdue payments and service cuts. Insome cases, they resort to the use of deferred payment and controlled consumption (e.g. intelephone, electricity) and their capacity to honor debts and credits has declined. This brought about the loss of goods like automobiles, real-estate property and even of family homes. Likewise, they have dropped recreational and cultural consumption, like going to the movies, theatres, restaurants, buying books and traveling. Changes can be seen intheir transport habits: cars are sold or left at home, while the use of public transport, bicycles, motorcycles and walking increases. As a result of these changes, there i s a psychological impact of perceiving themselves as being increasingly closer to the poor sector. The new poor find it difficult to accept and acknowledge their new situation, to ask for help and reducing expenses that were previously considered as indispensable, and to change their role by being forced to resort to charitable organizations inwhich they used to participate as donors I n the educationarea, the crisis affects both groups in different ways. Structurally poor groups state that children have reduced school attendance due to the lack of school elements, shoes, clothes and means of transport.16At the same time, the mainreason for attending school is the fact that they receive their mealthere. These groups have troubles causedby teachers' strikes, because duringthose days children do not receive the food that i s provided to them by the school cafeteria. They suffer from the discontinuity and lack of coverage of scholarship programs, the suspension of which discourages teenagers' attendance to school. It is getting increasingly difficult for the impoverished middle-class groups to afford the fees of private schools and to continue with University studies. Insome cases, University studies are dropped inorder to enter the labor market. New alternatives of education services arise among groups o f parents, friends and neighbors. More use i s made of popular libraries. I n the health area, the situation seems to be even more serious. Issues related to the deterioration of the quality oflife arise among the structurally poor, especially poor nutrition: malnutrition i s increasinginchildren and adults; there i s a new outbreak o f diseases that had been eradicated, like tuberculosis, hepatitis B and dengue. The situation of children, elderly people and terminal patients i s extremely precarious and vulnerable. As access i s concerned, the impossibility of gettinghospital attention has increased due to the lack of money to afford means of transport. Lack of resources prevents patients from following treatments and obtaining prescribed medicines. Hospitals also lack medicines, elements and sufficient equipment to meet the new demands. Free access to contraceptives has also declined (inthe few provinces where this was implemented), causing an increase in abortions and hindering preventive care. The impact o f the crisis on the new poor struck inthe changes of forms of medical coverage o f these families, like changing from pre-paid medical services to a social l6 However, this i s not borne out by the data, which show no decline in school enrollments (see Chapter 3). 16  security system, or full coverage pre-paid services to others with less coverage, which implies a deterioration inthe quality of the service; on the other hand, the numberof families with no medical services becausethey have lost their regularjob and cannot afford the cost of a pre-paid servicehas increased, with the consequent growing demands on public hospitals. Medicalvisits havebeen minimized, and cheaper medicines are bought. As with the poor groups, they claim that there i s a reduction infree access to contraceptives. I n general, CSOspoint to an increase in community violenceand social conflicts. The structurally poor mention anupsurge of violence related to robbery andinsecurity. They have witnessed an increaseinpolice abuse and impunity, especially youth persecution. Systematic violation of civil rights, as well as ethnical and social discrimination, go along with the growing process of poverty criminalization. New forms of protest and demand, such as pickets (road blockades), taking of businesses and buildings, preventing access to factories and public buildings, emerge and become more acute, as they aim at obtaining responsesto emergency issues. Among the new poor, the growth of crime rates fosters fear and insecurity, risingthe concern over crimes against property. Conflicts escalate, becauseof the new situations and living conditions. Among bothgroups, there is a stronger commitment (both female andmale) with civil society organizations (CSOs) arises, as male participation incommunity activities buildsup, due to the Plan Jefes. The new poor have started to perform community work, makeclaims as consumers and take part indemonstrations and protests that they had ignored until then. The crisis has a strong impact onfamily compositionand on the traditionalfamily model. Inall groups, these changes occur becauseof migrations to cities andrural areas insearch of new sources of income, as well as due to regroupings of family memberswho, now unemployed, go back to the family home to live, which worsens overcrowding and promiscuity conditions. The crisis of the traditional family modelbrings about a change inthe division of labor by gender. There is aredefinitionof roles anddutiesbothin householdsandinthe public field. At home, women strengthentheir role as heads of household, weakening the traditional role of menas family chief and support. Inthe public sphere, women take part inCSOs and inthe new forms of politicalparticipation such as pickets, assemblies, pot banging (cacerulazus) anddemonstrations. As a result, new conflicts emerge, breakingup family union and demeaning the image of parents before their children. Within families, there i s an increased sense of fear and impotence, specially over the situation of the youth concerning the loss of future projects, which deepens despair and pessimism. Child and teenage pregnancies increase, as well as infantile prostitution. Intraditionally poor groups, violence translates into fights, aggressions, separations, batteredchildren. Inmost cases, violence i s directed to women andchildren. Inter-generational conflicts and the impossibilityof control of youth increase. Impoverished middle-class groups witness growing rates of aggressions, 17  divorces and higher intolerance infamily relationships. Inboth segments, stress, depression, phobias, suicides and drug and alcohol addictions have grown. Results of a Special Household Survey To provide more quantitative evidence o f the impact, the Bank also commissioned a special household survey inJune/July 2002, covering 2,800 households indifferent regions (see The first part of the questionnaire asked for information for all householdmembers and includes questions on demographic characteristics, employment, income, migration status, education level, health coverage, and contributions to social security. Retrospective questions were asked for several of these questions, usingOctober 2001as the reference point. Individuals The Special Household Survey, 2002 Inaddition to the survey of civil society organizations, this chapter usesa special household survey specifically designed to identify socio-economic conditions in Argentina in 2002, and changesbetween 2001 and 2002. The survey is nationally representative and covered 2,800 households in different regions.' Unlike most other surveys in Argentina, this one included small localities (less than 2,000 inhabitants) and thus provides a unique perspective on rural areas. The survey, however, did not include dispersed areas and thus does not present a full picture of Argentina's heterogeneousrural sector. For urban areas, the sample design was based on a stratification by city size by region. For the rural areas, localities were selected randomly but, even though an effort was made to include localities from various regions, the sample is not large enough to be representative at the regional level. Overall, the survey design allows comparisons with the EPH for the urban areas. The field work was conducted during the months o f June and July of 2002. For additional information, see A. Fiszbein, et. al., "Argentina's Crisis and Its Impact on HouseholdWelfare", Background Paper No. 1, Part 11. were asked about their individual labor and non-labor incomes (cash and in-kind) as well as official and nonofficial income transfers. The second part of the questionnaire focused on the household as a unit and inquired about the use of savings, changes in consumption patterns, and participation in social programs and community activities. The head of household (or the best- informed individual inthe household) answered the questions.I8 The survey inquired about the most frequent strategies used by households inreaction to the worsening economic conditions experienced duringthe previous eight months. Changesin Consumption Table 2.1 summarizes the type of changes that families adopted intheir expenditure patterns, focusing primarily on reductions inconsumption levels and substitution for less- expensive items.Overall, all families seem to have reacted to the economic crisis through a combination of consuming less and substitutingfor less-expensive goods bothfor essential and nonessential items." That said, two patterns are observable. First, higher-income groups appear, l7The survey was conducted by IBOPE . For details, see Background Paper No. 1by Fiszbein, Giavagnoli and Anduriz. '*Ifthe head of household did not know the answer to a particular question (for example, changes in consumption patterns) the household was visited up to three times to find the member who could respond to the specific question. l9The consumer price index (CPI) increased by 30 percent duringthe fist six months of the current year. The increase in the prices of food items was higher (39 percent) than average. Using the expenditure shares for the 18  not surprisingly, to have avoided to a greater extent such changes. This i s particularlytrue inthe case of food, where the difference betweenincome quintiles i s striking. Second, the substitution for cheaper products (or second-handproducts when available) appears to be more pervasive thanreductions inconsumption levels. Inother words, the informationsuggeststhat families have triedto shieldthe consumption of food as muchas possible by reducing the consumption of other goods and switching to less-expensiveproducts. The lower-income groups, not surprisingly, appear to havebeen less successfulin doing so. Changes in consumption patterns I I1 I11 IV V Total Food Reducedconsumption 90.4 83.1 73.2 69.0 59.1 74.9 Substitutedfor cheaper products 97.6 95.4 92.5 91.5 84.8 92.3 90.5 87.7 81.5 76.8 68.3 81.0 Substitutedfor cheaper products 89.5 89.3 80.4 80.2 76.6 83.2 Purchased second handproducts 51.7 40.2 34.8 33.0 24.3 36.7 Dismissedheducedthe use domestic services* 43.0 27.8 40.5 37.9 A question that deserves particular attention is whether there is any evidence of increased school dropout rates as a result of the economic crisis. Table 2.2 looks at three age groups: 6-12, 13-15, and 16-18 years. The second column provides rough estimates of coverage rates for each group, without any referenceto the level of schooling. With all its limitations, these data confirm the rates traditionally observedthrough household surveys and do not show signs of reductions incoveragerates. We have further analyzed whether there are any indications of dropouts duringthe first months of the current school year among those children reported not to be at school (that is, those not included incolumn two of the table). Household heads were askedwhether someone in the household (without identifying who it was2') had either dropped out or postponed starting school. The third column reports the percentage of children inthe age group that are not in school andbelong to a family inwhich there i s at least one member that dropped out or postponed starting school. Of course, this might be overestimating the presence of dropouts to the extent that there is more than one household memberwho haddropped out or postponed starting school (including higher education), andinthat sense it is an estimate of the maximum potential size of the effect. The first thing to notice i s that, given the very small number of cases involved, the estimates are not sufficientlyprecise (that is, they have a highcoefficient of variation). lowest quintile derived from the 1997 consumption survey, we estimate that the cost of living for the poorest groups *'inThis avoided society increased45 percentduring this period. a 'finger-pointing' effect that might have led to underreporting of cases of dropouts at the expense of losing information on exactly which member of the household was involved. 19  Goes to school (% of all in group) Dropout estimate Age group Total areas Urban area Rural area Total areas 6 to 12years 97.0 97.3 92.1 0.2 13 to 15 years 92.5 92.5 90.9 0.6 16 to 18 years 80.2 81.1 60.1 2.3 Takingthis into account, it is apparent that there is no evidence of increased dropout rates for children inthe 6-12 or 13-15 age brackets. For the 16-18 age bracket, the estimated dropout rate i s about 2 percent of the age group, but the calculation i s not statistically significant. Looking at the distribution of this latter group, over half (54 percent) were found to come from the lowest quintile. The limited evidence of children dropping out of school as a result of the crisis could indicate that families try very hardto protect what i s perhaps the most important type of investment they can make for their children even under difficult circumstances. Inaddition, with reduced opportunities inthe labor market, the incentive to remove children from school so they can work i s reduced. In addition, as noted above, schools provide a source o f food for children through feeding programs. However, as evidence from other crises suggests, such negative effects might not be felt immediately but rather, mightbe felt several months after the beginning of the crisis (for example, mightbe reflected inenrollment rates in 2003). Another possible strategy i s to shift children from private to public schools. However, there i s little evidence from the survey data (see table 2.3), which suggests a resistance to make such changes even inthe context of a sharp economic deterioration (this contradicts the conclusions of the focus group discussions). A very large proportion (72 percent) o f the households reported reducing the purchase of schools materials, with as yet unclear effects on the quality of schooling. The effect was more pronouncedamong the poor, with almost 90 percent of the households inthe lower-income quintile reporting reductions inthe purchase o f school materials compared with only 43 percent in the highest-income group. The difficulties experienced by several provinces inkeeping up with spending levels suggest the presence of restrictions on the supply side that may be more serious than those on the demand side, at least at this early stage of the process. Table 2.3. AdjustmentsinEducationExpenditures (% of households) Substitutedprivate school for public one 2.0 Turned to cheaper private schools 3.1 Reduced purchase of school materials 71.9 Inthe case of health services, there is evidence of severenegativeimpacts as of early on inthe year, both interms of difficulties within the social healthinsurance andpublic health subsystems. The information collected through the survey confirms the trends identified previously. Approximately 12percent of the individuals experienced some change inhealth insurance coverage. More than 60 percent of those lost coverage altogether, mostly those in the 20  lowest-income groups. For the approximately 40 percent o f individualsexperiencing a change in health insurance coverage, the movements were toward lesser coverage. The loss of insurance coverage is leading an increasingnumber o f people to rely on the use public health facilities. O f those households reporting to have substituted public for private health facilities as a reaction to the economic crisis, 60 percent have experienced some form of loss or limitation inthe coverage offered by their health insurers. More specifically, 16percent lost their insurance altogether, 5 percent have outstanding bills that do not allow them to use their insurance coverage, 13 percent cannot afford the co-payments, 8 percent experienced a reduction in services coveredby their insurer, and 13 percent are retirees ex eriencing difficulties with the services offered by PAMI.(health program for retirees) 2p Overall, the survey found that families have, inone way or another, cut down on health services as a result of the economic crisis. Almost 23 percent o f the households reportedthat at least one member experienced lack of access to health services. Three-fourths of those reported that the reason was "lack o f money" to pay for medicines (44 percent), transportation costs (26 percent), andfees (5 percent). More specifically, 37 percent o f the households with children under the age o f 12reported having reduced the frequency with which they take their children for medical checkups. Almost 45 percent of those belong to the lowest quintile of the income distribution. But this effect was also felt inthe second (29 percent) and third (18 percent) deciles. Another coping strategy that households have adopted i s to cut back on the use of utility services, or to delay payment thereby effectively consuming on credit. Overall, residential consumption of these services has fallen b y 10% since the onset o f the crisis. Furthermore, between 30% and 50% o f the households reported delaying payments. The percentage of households reportingbeing cut off for non-payment (electricity, gas, water, telephone, cable, internet) range from 2% percent (inthe case of water supply) through 10% (for electricity and fixed line telephones) up to 14percent for cable. Disconnections are concentrated inthe lower quintiles of the income distribution, with between 20% and 25% o f first quintile households reporting that they have been disconnected from electricity and telephone. Nonetheless, the disconnection statistics do not tell the whole story. Almost half o f first quintile households who report having been disconnected from the electricity service following the crisis, say that they currently have access to the service, suggesting that they must subsequently have become reconnected, although it i s not clear whether this i s through legal or illegal means. Similarly, households have tended to substitute less-expensive means o f transportation, as shown intable 2.4. More than half the households previously using only public transportation now report having increased the use of a bicycle or simply walking; for the poorest households the figure i s significantly higher. These figures are consistent with those emerging from the official data on the use of public services" that indicate, for the first five months o f the year, a 22 percent reduction inthe number of passengers on urban railways, a 9 percent drop in subway use, and a 14percent decrease inuse of metropolitan bus services. A similar switch from cars and taxis to public transportation is observed as well. 21PAM1is the national health insurer for retirees. 22These statistics are collected monthly by INDEC and constitute the basis for its Indicador Sintetico de Servicios Publicos (ISSP, Indicator of Public Services). 21  Table 2.4. Changes inthe Use of Transport Services I I1 I11 IV V Total 1 Carhaxis by public transportation 57.4 56.7 47.8 45.5 32.4 46.1 Car/taxi/ public transportation by bicycle/walking 73.9 62.5 61.3 49.0 33.5 55.4 Income-Generation Strategies Poor households can use various strategies to attempt to raise incomes, or to offset income reductions, in a crisis. As shown intable 2.5, approximately 13 percent of the households sampled placed a new member of the household into the labor force. This move was particularly strong among the lowest-income group, where the figure rises to 28 percent. These are typically secondary workers: in 50 percent of the cases it i s the son or daughter, and in 25 percent of the cases it i s the spouse of the person reported to be the head of household. Quintiles I I1 I11 IV V Total Labor market strategies Adding new workers to labor market 28.0 16.8 12.2 6.2 1.4 12.9 Working more hours 11.4 15.6 16.3 11.5 13.4 13.7 Time Increased home manufacture 74.4 73.0 62.6 52.5 43.2 61.1 Dismissheduce domestic services 43.0 27.8 40.5 37.9 Migration 4.3 2.3 6.0 5.1 4.6 4.5 This coping strategy, however, has limitedimpact duringa time of risingtotal unemployment. Although some households are successful inhaving new entrants find work, the effects are more than offset by risingunemployment of those already inthe work force. As shown intable 2.6, comparing October 2001 and July 2002, most o f those who moved out of "inactive" status wound up unemployed; only 17 percent found work. Those who found work were often not those inthe poorest groups, and the majority of work was in temporary jobs. Where work was found, it was largely by those previously unemployed. At the same time, many of those who had worked became unemployed, and many of those unemployedbecame "inactive," suggesting that labor market conditions discouraged further job search. As shown in table 2.6, the rate of unemployment of the sample went up from 8.6 percent to 9.9 percent during the period, a trend consistent with the macroeconomic data.23 23The unemployment rate calculated here is as a share of the total population, not the labor force. Unemployment as a share of the labor force would be 21.6 percent in October 2001 and 24.6 percent inJuly 2002. 22  Table 2.6. ChangesinLabor MarketStatus (October 2001-July 2002) I 1 I I 1 Previous status I Unemployed Inactive Total Unemployed - 1000s 1,435 1,228 15 555 3,233 % 50.6 6.2 0.2 17.1 9.9 Working more hours i s another strategy followed by all income groups. However, it tends to be mostly salaried workers inpermanentjobs with a particular concentration incommerce who appear to succeedinthis effort to increasethe hours of work. Inparallelwith these changes inemployment strategies, it is possible to observeatendencyto substituteleisure for in-home work as a way of coping with the crisis. Approximately 60 percent of the households report that they have increasedthe time allocated to the preparation of home-produced goods (for example, food) to substitutefor purchasedgoods. As shown intable 2.5, the lowest-income groups heavily use this strategy ("home manufacture"). The survey also inquiredas to whether ahouseholdmember hadmigratedsince October 2001 or if someone inthe householdis considering that option. On average, a little over 4 percent of the householdsreported having at least one member who migrated, a percentage distributedmore or less evenly acrossthe income di~tribution.~~The mainreasons for migration included "lack of work" (58 percent) and seeking abetter quality of life (15 percent). More than 20 percent of the householdsreported that at least one family memberi s considering migrating. The key option beingconsidered (80 percent of the households) i s migrating to another country, particularly within the highest-income groups inthe city of Buenos Aires. Since the beginning of the crisis a larger proportionof the households are usingtheir savings, selling assets, or borrowingas a means of smoothing consumption (table 2.7). Perhaps the most interesting fact to notice i s how low-incomehouseholds rely on the use of informal credit at neighborhood stores (compru alfiudo) whereby payment i s delayed. This coping strategy i s usedby 15 percent of the poorest families, but only 1percent of the upper quintile. 24This excludes those migrating as a result of marriage, sickness, or other family reasons. 23  Quintiles I I1 I11 IV V Total Selling assets 5.9 3.7 3.3 2.7 1.1 3.3 Using savings 2.8 3.5 4.0 8.0 5.6 4.8 Borrowing from banks 0.9 3.6 1.8 0.6 2.0 1.8 Purchase with delayed payment 14.6 13.1 9.5 2.3 0.7 8.0 Social Network Strategies Table 2.8 summarizes some of the key indicators of use of social networks by households. Overall, we observe that more than one-third of the households report benefiting in at least one aspect from access to a network of social support. The use of such strategies i s somewhat higher for households that experienced negative nominal shocks intheir incomes and i s significantly higher for households at the lower end of the income distribution. % of household % of households using reporting a reduction in income using Assistance from other persons who don't live in the housez5 16.3 17.0 Borrowingfrom friends or family 10.7 13.7 Social plans 6.9 7.5 Participation in community activities26 20.9 26.4 Barter 11.1 15.2 Using at least one 37.0 42.5 Using more than one 14.6 19.9 The survey provides some evidence that social networks might be cutting across the income distribution. Interms of receiving or giving assistance to persons not living intheir households, those inthe lowest quintile tend to be net recipients and those inthe highest quintile net givers (see table 2.9). For the poorest households, 10percent gave and 26 percent received assistancefrom persons outside the household. For the richest quintile, the reverse i s true: 38 percent gave and 7 percent received. Table 2.9. Use of SocialNetworks by Income Level 1 1 I1 I11 IV V Total Assistance from other persons who don't live in the house 26.4 15.5 20.9 15.8 7.3 17.1 Assistance to other persons who don't live in the house 10.0 19.7 17.1 21.2 37.9 21.2 Borrowingfrom friends or family 21.2 15.7 10.6 5.8 3.0 11.3 Social plans 18.5 6.1 7.5 1.3 0.1 6.7 Participation in community activities 29.4 23.6 22.8 15.7 15.7 21.4 Barter 20.2 15.4 11.7 5.5 3.9 11.3 25A higher percentageof households (22 percent) report providing assistanceto persons not living inthe same households. 26Community activities include the following categories: school canteens, soup kitchens, communal purchases of goods, community workshops, fundraising, community childcare, among various others. 24  Although the survey registers an increase inthe use of various forms o f social networks, with respect to October 2001, the largest increases are inthe use o f barter and participation in community activities, as shown intable 2.10. Table 2.10. Changesinthe Use of Social Networks % of household Before Oct 2001) % of using household using Assistance from other persons who don't live in the house 16.3 13.2 Borrowing from friends or family 10.7 7.0 Assistance from government or nongovernmental organization 6.9 2.7 Participation in community activities 20.9 11.1 Barter 11.0 3.2 The intensity of use of barter among low-income groups i s documentedby the fact that the per capita income of households usingbarter i s less than that of the entire population (143 pesos versus 214 pesos, based on adult equivalency). Households using barter tend to have more unemployed members (43.3 percent versus 22.4 percent) and more temporary workers (42 percent versus 23 percentbfurther proof of the higher vulnerability of those households. A similar patternis observed for those households that report receiving some type of official help. Their per capita income i s below average (122.7 pesos), they have a higher incidence of unemployment (65 percent have an unemployed household member), and those who are employed tend to have mostly temporary jobs (80 percent). Comparisons with Other Countries A quick comparisonwith evidence from other countries experiencing similar large economic shocks shows some interesting similarities as well as some specific differences. The analysis by Fallon andLucas (2002) on the effects on labor markets27indicates a patterno f rising open unemployment rates and important changes inthe sectoral composition o f employment (sharpreductions inconstruction and manufacturing-the latter as the corporate sector i s hit by the higher costs o f imported materials, difficulties in obtaining credit, and the rising burden o f debt inforeign currency) similar to that observed inthe case o f post convertibility Argentina. The evidence on the distributional impact o f crises shows some differences between countries. Thomas and others (1999) estimated that in the case o f Indonesia there was a larger decline inper-capita expenditures inboth the top and bottom quartiles of the distribution than for households inthe middle of the distribution. Incontrast, there i s evidence o f at least a weak redistribution of income from the middle classes to the rich inThailand. Our estimates suggest that reductions infamily income inArgentina have affected mostly the middle class. Regarding coping strategies, as inthe case o f Argentina, Frankengerg, Thomas, and Beegle (1999) note that informal assistance from friends and family members was particularly important inIndonesia during that crisis, with about a quarter o f the households receiving such 27Countries studied and crisis years: Indonesia (1998), Korea (1998), Malaysia (1998), Thailand (1997), Argentina (1999, Mexico (1995) and Turkey (1994). 25  assistance. Its median value was considerably higher than that of assistance from formal services. Sudarno, Wetterberg and Pritchett (1999), also found case differences incoping strategies along the Indonesian income distribution. For instance, middle-class families responded by working more hours, reducing consumption, drawing down savings, andselling assets, while lower- income households resorted to more drastic measures, such as taking children out of school. Regardingthis last point, the evidence on the matter i s mixed. For example, inMexico duringthe 1982 crisis, highschool dropout rates increased slightly while primaryschool dropout rates fell. Both changes were part of long-term trends throughout the 1980s (Lustig ,1998). Similarly, Adam and Chamberlin (1999) did not find evidence o f significant reductions in school participation inThailand even though they also recognized the measurement difficulties and delayed responseswe mentioned above. The evidence for Thailand, though, did find a gap in dropout rates between poor and non-poor households, as i s found in for the case of Argentina. 26  CHAPTER 3 THEGOVERNMENT RESPONSE ToTHECRISIS:SOCIAL SPENDING~' Backgroundand GeneralTrends Conscious of the extreme stress in the economy and the attendant impacts on social conditions, the national government of Argentina has been taking the bulk o f the responsibility for formal public action.29 These actions have been constrained, however, by the government's own declining fiscal resources. The main government initiative has been directed at the formulation o f a Social Emergency Program. The mainthrust of this program i s to provide the material support needed to avoid the most deleterious impacts o f the crisis on the poor and vulnerable with actions inhealth, education, nutrition, temporary work, income support, and community action. These investments are critical in ensuring continued human capital development among the poor and vulnerable groups inArgentina and inprotecting the existing humancapital from serious deteriorations under the severe economic crisis. Table 3.1 shows that between 2001 and 2002 there has been a combination o f overall fiscal contraction and a pro-poor shift inthe composition o f public pend ding.^' This ability to reallocate funds has been inpart possible thanks to the government's temporary default on its external debt, so it i s not clear how sustainable this reallocation strategy will be inthe near future. While total spending in constant prices fell 38% in this period (public sector), social spending fell 32%, indicating some attempt to give priority to the social sectors. Social spending, however, continues to have a strong universalnature and is not necessarily beneficial to the poor. Targeted social spending, that part of total social spending which funds programs directed at the poor and vulnerable segments of the population, has experienced a significant increase in nominal terms and as a percentage of social pen ding.^' This increase reflects the implementation of the Social Emergency Program, particularly the introduction o f the workfare program, Plan Jefes y Jefus de Hogur. This nominal increase has been more than sufficient to ** This chapter i s based on three background papers: Ridao-Cano, C. 2002 "Government Response to the Crisis: Social Spending." Background Paper No. 5; Schwab, Nicole, and Juan Pablo Uribe. 2002, "Impact of the Crisis on the Health Sector in Argentina." Background Paper No. 6; and Espana, Sergio, MariaPaula Savanti, and Suhas Prandendakar. 2002, "Impact of the Crisis on the Education Process in Argentina 2002." Background Paper No. 7, all in Part I1of this report. 29 See chapter 2 on the coping strategies adopted by Argentine families. 30 Overall social spending is given in terms of both the national government and consolidated public sector, which includes the provincial and municipal governments. The discussion in this section focuses on social spending of the national government for because detailed information by individual program is only available at the national level, and this information is key for an accurate characterization of social spending, particularly targeted spending. In addition, the national government has taken the bulk of the responsibility for formal public action to counteract the social impact of the crisis, although many national programs are administered by provincial and municipal governments. 31For a description of all targeted social assistanceprograms, see Guia de Programas Sociales (Sistema de Informacih, Monitoreo y Evaluaci6n de Programas Sociales [SIEMPRO] 2002), Ministry of Social Development, Government of Argentina. 27  counteract the effect of inflation, resulting in a 21 percent increase of targeted social spending (TSS) inreal terms. When put inthe context of the expected 11percent decline inreal GDP, TSS shows, in contrast with previous crises, an anti-cyclical pattern. When put inthe context of a 42.8 percent increase inthe poverty rate, however, this increase inreal terms of TSS has not been sufficient to cope with increasing poverty, resultingin a 15.8 percent decline in spending per poor person inreal terms. This decreaseinTSS per poor person i s higher than the decline in GDP or GDP per capita (11.9 percent). As shown later, however, only about 61 percent of targeted social spending in 2002 takes the form of direct cash transfers to families. Direct cash transfers have increased by 81.7 percent inreal terms and by 25.9 percent inper poor terms. Given the fiscal constraints, inflation and the sharp rise inpoverty, the government's efforts in reallocating funds to targeted social spending, particularly inthe form of direct cash transfers, are notable, and an improvement over their performance inprevious crises. Table 3.1. PublicSector SocialSpending, 2001-02 2001 2002 % change Public Sectorj2 Total Spending Millions current pesos 96,153 91,111 -5.2 Millions 1993pesos 91,643 57,302 -37.5 Social Spending Millions of current pesos 59,112 60,962 3.1 Millions of 1993pesos 56,340 38,341 -31.9 National Government Total spending Millions of currentpesos 48,935 46,09 1 -5.8 Millions of 1993pesos 46,640 28,988 -37.8 Social spending Millions of current pesos 28,65 1 29,587 3.3 Millions of 1993pesos 27,308 18,646 -31.7 % Total spending 58.55 64.19 Targeted social spending Millions of currentpesos 3,308 6,056 83.1 Millions of 1993pesos 3,153 3,816 21.0 % Social spending 11.55 20.47 77.2 % GDP 1.23 1.70 38.2 Perpoor person 240 202 -15.8 Source: Ministryof Economyand SIEMPR0.33 Table 3.2 shows that the combination o f reallocation of national government spendingto targeted social spending (TSS) and relocation within TSS across functions has resulted in substantial increases ineducation, health, and, particularly, employment, which becomes the most important spending category. Within social assistance, there have been significant increases 32Includesnational and provincial government expenditures. 33The spendinginformation for 2002 refers to accrued expendituresat 12/31/2002.I norder to make the comparison betweenTSS in 2001 and 2002 possible, funds allocatedto national fund for housing (FONAVI) are excluded from TSS in 2001 (about 656 million pesos). FONAVI is no longer consideredpart of TSS in 2002 (it is still part of social spendingthough) as FONAVI funds became unconditional transfers to provinces in 2002. Real figures are basedon acombinedindex of wholesale and retail prices, using 1993 as the base year. The poverty number for 2002 is based on the EncuestaPermanente de Hogares(EPH) of May 2002. 28  infood andcashtransfers.These changesare consistentwith the priorities set forth by the Social Emergency Program. Table 3.2. Targeted Social Spendingof the National Government by Function (million Desos) Social assistance 1,595 1,520 48.22 1,802 1,136 29.76 Social security 1,014 966 30.65 1,115 703 18.41 Education 131 125 3.96 353 222 5.83 Employment 266 253 8.04 2,308 1,455 38.12 Housingand urban development 17 16 0.5 1 16 10 0.26 Water and sewerage 13 12 0.39 14 9 0.23 Total 3,308 3,153 100.00 6,056 3,816 100.00 SocialAssistance Programs Within social assistance therehasbeen a significant increaseinfunds allocated to food and conditional cashtransfers.34 The bulk of the food assistanceis channeled through the Programa de Emergencia Alimentaria (272 million pesos)35andthrough FOPAR (Programa Fondo Participativo de Inversion Social), whose funds have been reallocated to finance community kitchens and have increased substantially (from 8.5 million pesos to 16.5 million pesos). These programs provide daily food rations for poor families, particularlythe most vulnerable (that is, pregnant women, children, and the elderly), through community kitchens. It i s worth notingthat an important partof food assistancehas been channeled through already existing programs, with the aim of trying to facilitate a quick and effective delivery to the poor andvulnerable. Because of the problems associatedwith in-kindtransfers the Federal government no longer provides food directly to poor people. It now provides cash assistance to poor families, or financial resources to support various ongoing nutritional programs runby provincial and municipalgovernments. The problemis that many of these have become de facto block transfers with little or no impact on malnutrition. Most of the funds allocated to food assistancefor 2002 have been disbursed. This i s not the case of FOPAR, which has been subject to important implementation delays, resulting inmuchlower disbursements than expected for 2002. For 2003, however, FOPAR is expected to reach 54.2 million pesos. The IDH(Ingreso para el Desarrollo Humano) i s the most important conditional cash- transfer program (69.8 million pesos). The program, which startedbeingimplementedinthe secondhalf of 2002, is aimed at promotingthe development, health, and schooling of children through the provisionof cash transfers to poor families that are conditionalon school attendance and healthcontrols, and that dependon the number of children (up to a maximumof 150pesos 34 The most important part of social assistance funds, however, continues to be transferred to provinces for targeted social programs (1,079 million pesos), mainly covering basic social infrastructure. 35 The amounts reported in this section are in current pesos. 29  per family per month). The design and implementationof this programbuilton the successesof programs such as PROGRESA inMexico interms of school and health outcomes of children. The IDH, however, has been subject to significant implementation delays, resulting inmuch lower disbursements and coverage than expected for 2002. For 2003, however, IDHi s expected to reach 227.5 million pesos. As a whole, these transfers of cash and food are thus mainly targeted at those who are unable to work, such as pregnant women and the elderly, and those who should not work, such as children in school, which makes them an appropriate safety net, at least theoretically, given that an important part of the rise inpoverty i s explained by declining real wages. Within employment,and targeted social spending as a whole, the most important programby far i s Jefes y Jefas de Hogar (2,248 million pesos; see box). This i s an ambitious workfare program that i s estimated to have reached around 2 million beneficiaries by the end o f the year, providing them with 150pesos per month in exchange for work incommunity projects or training. The transfer amount was set at a level slightly below the going wage for full-time work for unskilledworkers, so as not to discourage people from seeking more permanent jobs as soon as they become available, and thereby avoid distortions to the labor market. This transfer amount should also promote effective targeting of the program to poor households. Although quantitative causal statements are difficult to make, the rapid expansion of the Jefes program in 2002 appears to have been the main reason behind the decline in unemployment from 21.5% inMay to 17.8% in October. However, not all beneficiaries were previously unemployedbefore participating inthe program. The design and implementation of this program i s built on the success of a predecessor workfare program, Trabajar, in mitigating the loss of income resulting from unemployment (as opposed to lower real wages). Ina sense, it makes up for the low coverage of the unemployment insurance program. The Trabajar provided up to six months per year of work to unemployed adults. The Jefes program was originally intendedfor unemployedhouseholdheads who have children at home. However, the large number of women participating inthe program suggests that the definition of who i s a household head i s flexible in practice. Compliance with key eligibility requirements (households containing children under the age of 18 and presentation of IDnumber) has improved gradually from its early beginnings. It i s now estimated that ineligible persons participating are less than 20%. For 2003, the Jefes program i s expected to reach 3,055 million pesos, thanks, inpart, to the new loan from the World Bank. Inadditionto Jefes, a similar employment program calledPEL (Programde Emergencia Laboral) was created to provide for those unemployedindividuals who do not meet the eligibility criteria of the Jefes program. It currently enrolls about 300,000 people, and is targeted mainly to adults that do not have children (single individuals) or whose children are older than 18 years. Inaddition the Government i s trying to encourage private businesses to offer employment of 6 months to Jefes participants. Under the scheme, the company would only need to top up the workers' salary by the difference between the amount o f the Jefes transfer and the salary level inthe collective agreement.36 36 IsmaelBermddez, "Lanzaron el planparasubsidiar el EmpleoEstable", Clarin, Jan.22, 2003. 30  Plan Jefes y Jefas de Hogar Desocupados (Program for Unemployed Heads of Households) Type of program: The "Jefes" program is an emergency workfare program. Unlike its predecessor, Trabajar, the Jefes program was originally intendedfor heads o f households only. However, the data shows that there were actually as many if not more heads of household employed underTrabajar as under Jefes. Part of the reason for that i s the type of work that is offered, such as work in community kitchens, which i s more appealing to women than to men. The program i s funded by the national government and runby the Ministry of Labor. Stated objectives: To provide financial assistanceto male/female household heads with children in order to guarantee the Family Right of Social Inclusion, ensuring (1) the school attendance and health of children; (2) access of beneficiaries to formal job training and education; and (3) their participation inproductive projects or community services. Inpractice, however, program administrators are not collecting the information on school attendance and heath. Likewise, most beneficiaries participate intraditional workfare activities (community services and small construction projects) and only very few take advantage of the education, training and private sector employment options. Implementation: The implementation of the program is decentralized through municipalities and communes, in collaboration with the corresponding municipal or communal council. These councils are responsible for the monitoring, transparency, and implementation of the program. The municipalities assign the activities to be undertaken by program beneficiaries inreturn for the financial assistancethey receive. These activities can also be proposed by any public or private nonprofit organization. Intended beneficiaries: Male/female household heads with children who are either 18 years of age or younger or disabled. Likewise, households in which the female head, spouse, concubine, or cohabitant partner of the male household head suffers from serious health conditions are also eligible. To receive benefits, the household member must provide standard set of information and supporting documentation (birthcertificate, national identification number) to the municipal employee handling the registration. The beneficiary is supposed to sign a statement swearing that the information provided is correct. Before making the monthly payments, cross-checks with other data basesare done usingthe IDnumber to ensure that the beneficiary is not participating in another program of the Ministry, receiving unemployment insurance, has a formal sector job, or is receiving a cashtransfer from another program of the national Government. Benefits: Monthly payments o f 150pesos. Ineducation, andcomplementing the IDH,the Social Emergency Programseeks to maintain the provision of education services and to facilitate access by vulnerable groups through two means: (1) providingbasic learning materials, which may otherwise be out of reach of poor families, through the new program Utiles Escolares (30.8 million pesos) and (2) offering incentives through the program Becas Estudiantiles (200.1 million pesos in 2002), and school lunchprograms (70 million pesos), through the programAcciones Compensatorias en Educacidn, to encourage school attendance. Most of the funds allocated to these programs in 2002 have been disbursed, although Utiles Escolares has experienced implementation problems. Inhealth, the Social Emergency Programaims to protect the most vulnerable andto maintain essentialpublic healthprograms with support for the functioning of the public hospitals network, provision of essentialdrugs for the most vulnerable, and protectionof priority public health measures.These measuresinclude immunizations, disease control, and ensuring services relatedto women's reproductive healthcare and prenatal controls, such as anemia prevention, child delivery, newborn reception, postpartum and child healthcare, through the provision of 31  pharmaceutical products, medical supplies, and iron-fortified milk.Essentialdrugs are provided free o f charge to poor people by a new health emergency program called REMEDIAR (7.3 million pesos) through primary health care centers. The program started being implemented in July 2002 and has been subject to significant implementation delays. These delays have resulted inmuch lower disbursements and coverage than expected for 2002. The same is true, for the same reasons, for programs that provide drugs to combat specific illnesses, such as the national program to combat acquired immunodeficiency syndrome (AIDS) and sexually transmitted diseases (STDs) (2.7 million pesos). For 2003, however, REMEDIARi s expected to reach 157.8 million pesos while the assistance for AIDS and STDs drugs i s expected to reach 123.8 million pesos. Vaccinations have beenprovided through the National Immunization Program (127.5 million pesos). Complementing the efforts o f food programs, actions related to maternal and child health and nutrition have been channeled through PROMIN(61.3 million pesos). Finally, an important part of targeted social spending inhealth continues to be focused on actions directed at those benefiting from non-contributory pensions (144 million pesos). Most of the funds allocated to the last three programs in2002 have been disbursed. Within social security, funds allocated to pay unemployment benefits seemto have responded somewhat to the rise of unemployment inthe formal sector (from 297 million pesos in 2001 to 472.3 million pesos in2002). However, funds allocated to non-contributory pensions have decreased slightly (from 672 million pesos in 2001 to 639.8 million pesos in 2002). Withinhousing andurbandevelopment, the most important programi s PROMEBA which supports community projects aimed at upgrading poor neighborhoods. The program has been subject to significant implementation delays resulting in an unexpected decline infunds disbursed by this program with respect to 2001 (from 12 million pesos in2001to 6 million pesos in2002). Inwater and sewerage, fundingfor the Programfor the Provisionof Potable Water and Basic Sanitation (PROPOSA) has increased slightly from 9 million pesos in 2001 to 9.6 million pesos in 2002. Efficiency and Effectiveness of Targeted Social Spending Inthedebateoverpoliciestofightpovertythat hastakenplaceoverthepastfew years, there has been a general recognitionthat important amounts of targeted social spending could be reallocated to finance more effective programs that fight poverty by subsidizing income, either in cash or inkind (for example, food, medicines, learning materials, health treatment). Within income subsidies, cash transfers are considered to be more effective tools for poverty alleviation than in-kindtransfers, particularly inthe short-run, and to be less subject to the inefficiencies and administrative difficulties related to in-kindtransfers. An important part o f targeted social spending comprises programs that fight poverty through interventions in specific markets such as housing, health, and education. To determine more precisely the type of intervention to alleviate poverty used b y each program, SIEMPR037has proposed the categorizationo f targeted spending that appears intable 3.3. 37SIEMPR0,2002 ,Estructura y Evolucidn del GastoSocial Focalizado: 1997-2002,Informe de Gasto Social No. 4, Consejo Nacional de Coordinacih de Politicas Sociales, Presidencia de la Nacih, Argentina. 32  2001 2002 Nominal Real %TSS Nominal Real % TSS Income subsidies and social assistance 1,709 1,629 43.11 4,224 2,662 69.75 Incash 1,343 1,280 33.86 3,691 2,326 60.95 In kind 366 349 9.24 533 336 8.80 Targetedsectoral interventions 527 502 13.29 750 474 12.38 Unclassified transfers to provinces for 1,073 1,022 27.06 1,079 680 17.82 targeted social spending Total 3,308 3,153 100.00 6,096 3,816 100.00 As table 3.3 shows, between 2001 and2002 there has been a significant reallocationof funds toward cash transfers, which accounts for 61 percent of targeted social spending in2002. This reallocation mainly reflects the introduction of Jefes y Jefas de Hogar. As mentioned earlier, the design and implementation of Jefes y Jefas de Hogar, as well as IDH,buildfrom the successful experiences of similar programs. These cross-cutting programs are part of a new effort to formulate more effective andefficient social policies inthe face o f the deep crisis inthe labor market andthe failure o f more traditional social policies. Incontrast with Jefes y Jefas de Hogar, however, IDHhas been subject to serious implementation difficulties, resulting in significant delays. This i s partly explained by the different targeting mechanisms used by each program: Jefes y Jefas de Hogar i s a self-targeted program while IDHi s a means tested program. SIEMPR03*points out that, given the magnitude and seriousness o f the social crisis, an immediate first step should be to reallocate funds to cash transfers. To this extent, it identifies substantial cost-saving opportunities inthe funds allocated to social promotion anddevelopment (which includes programs aimed at buildingthe capacity o f individuals) and in-kindtransfers. Part of these funds is used to finance an important number of small-scale programs with high operational costs and highdegree of duplication. Inthese areas, it could be possible to obtain gains ineffectiveness and efficiency by merging programs and overlapping institutions and through the implementationof management models that emphasize results. The fragmentation of programs within institutions and overlapping program objectives across institutions i s a reflection o f a broader and more serious problem, which i s the historical lack of coordination o f social policies, and thus the lack o f a clearly defined safety net strategy. This situation appears to have improved in2002 with the introduction of the ConsejoNacional de Coordinacidn de Politicas Sociales (CNCPS). An important tool inthis new coordination effort i s the description o f social programs provided in Guia de Programas Sociales3'. Historically, very little has been publicly known about the coverage and efficiency of social-assistance programs. Basic monitoring data on the number o f beneficiaries and budgets have been often lacking or unreliable. This basic information i s now provided by SIEMPRO for 38SIEMPRO, 2002, Estructuray Evolution, op.cit. 39SIEMPRO, 2002, Guiade ProgramasSocialesNacionales2002,Consejo Nacionalde Coordinacihde Politicas Sociales, Presidenciade la Nacih, Argentina 33  national and provincial programs, although the latter i s still limited.40 Inaddition, an Integrated Monitoring System (SIM) has been recently re-introduced by SIEMPRO, which tracks the performance of programs interms of indicators of implementation, product indicators (for example, unitarycosts), coverage, results, and context. Although these are steps inthe right direction, it i s still unclear how relevant and effective these new instruments are inpractice. Targeting i s an issue closely linked to the efficiency and effectiveness o f social-assistance programs. The Index of Unsatisfied Needs (NecessidadesBasicas Znsatisfechas --NBI) has been the most common indicator used for targeting social programs. This index mainly measures structural poverty, weighting more heavily those aspects that persist over time as well as housing conditions. Usingthe 1997Encuesta de Desarrollo Social (EDS), the Bank's report Poor People in a Rich Country (2000)found that while social-assistance programs appeared to be strongly pro-poor overall, there was considerable leakage to non-poor (35 percent o f the benefits). More troublesome, 70 percent of the poorest quintile were not getting any benefits. The same picture holds when looking at food and nutrition programs, while housingprograms were found to be badly targeted to the poor. On employment, the workfare program Trabajar was found to be well-targeted to the poor.41.Infact, Trabajar performedbetter than any other social-assistance programfor which incidence analysis has been carried out, which proves the power of self- targeting mechanisms. Finally, on communit development, the program FOPARhas been found to be the second best well-targeted program4 ,l which comes from a combination of careful geographic targeting and good self-targeting. Inan attempt to overcome the deficiencies of the NBI, SIEMPRO has recently developed the Index of Quality of Life (ICV), which approximates householdper-capita income through 22 variables that cover a broad range of structural as well as transitory aspects o f poverty. This index i s currently under trial and has been used to identify beneficiaries for the IDH.This index i s part of a broader System for the Identification and Selection of Current and Potential Beneficiary Families of Social Programs (SISFAM), which i s a database that includes socioeconomic and demographic information of the poor and vulnerable at both provincial and municipal levels. This tool will be used to identify beneficiaries for different types of programs usinga variety of indicators as well as to identity new areas of intervention. Itwill also facilitate the analysis of benefit incidence. However, it i s not possible to know how well targeted the 2002 social programs are with the tools available. Inparticular, the only socioeconomic survey that i s collected on a regular basis (that is, EPH) does not collect information on program participation, except for the October 2002 wave, which includes information on Jefes and Jefas de Hogar. This deficiency calls for the need to develop more appropriate data instruments. This problemis common to impact evaluations,a key component in assessingthe performance of programs inreducing poverty. Despite their strategic importance, impact evaluation studies are almost non-existent. Infact, the only true assessment o f causality has been done for Trabajar (Jalan and Ravallion, 1999).As regards programs included inthe 2002 40 Despite the efforts to consolidate public spending information from the national government, provinces and municipalities,the budgetinformationby individual programis only available at the nationallevel. 41Jalan, J. and M.Ravallion, 1999,Income Gains to the Poorfrom Workfare: Estimatesfor Argentina's Trabajar Program, World Bank, Washington, DC. 42SIEMPRO, 1997,ProgramaFondo Participativode InversidnSocial (FOPAR): Evaluacidnex post de proyectos, ConsejoNacionalde Coordinacidnde PoliticasSociales, Presidenciade la Nacih, Argentina. 34  budget, structured impact evaluation i s contemplated for Jefes and Jefus, using a special module included inthe 2002 October's EPH. Evaluating the impact of, at least, major social assistance programs is particularly important incrises as it provides the necessary feedback for possible corrections. Impactof the CrisisonEducation Inthe beginning of the 1990s, Argentina embarked on a major series of reforms of the education sector. The two most important elements of the reforms were (1) completion of the process of decentralizationo f the educational system to the provinces, and (2) an extension o f the years of compulsory schooling from 7 years to 10years. Under the decentralizationreform, the provinces would runtheir own schools, while the federal government was assigned the responsibility for maintaining the quality andthe equity of education provided by the provinces At the time that the major effects of the fiscal crisis began to take effect at the beginning of 2000, a snapshot of the different educational systems of the provinces reveals different stages of completion o f the reform process initiated 10years before. Even as inequality of educational opportunity inArgentina had been reduced significantly, the crisis came at a time when differences remained very high,particularly inregard to educational quality and access to secondary education. Provinces had shown different levels of political support for the reforms and had usedtheir own resources as well as transfers from the federal government with widely varyinglevels of efficiency. It is possible that the halt inthe progress of the educational reform that did take place after 2000 would have occurred regardless of the initial macroeconomic manifestations of the crisis. Inany case, it can be said that the crisis was coterminous with the reformbeing halted in its tracks. As shown inchapter 2, coping strategies o f households Figure3.1 School Attendanceand Levelsof Poverty have not included the withdrawal of children from school. We see that 120 I as poverty increased steadily from 1998 through 2002, the enrollment of children aged 6 to 15 years 80 remainedconstant at around 98 -0 / t458 percent (see figure 3.1). 5 60 0 .-s 40 v) Looking at year-to-year 8 -- 35 dropout rates using the fact that 75 20 percent o f households are retained 0 I I I 1 I I I - 30 inthe sample of the EPHfrom one M 9 8 0 9 8 M 9 9 0 9 9 M 0 0 0 0 0 M 0 1001 M02 round to the next, we find no I difference indropout behavior prior *Attendance, children 5-16 -a- Poverty to the crisis and duringthe crisis. For those aged 18 to -25 years, there appears infact to be a slight increase inenrollments. The household behavior can probably be explained by high tastes for schooling and the lowered opportunity cost o f time away from school. There is a vast literature that has examined the question of whether enrollments are pro- 35  cyclical or anti-cyclical. On the whole, this literature suggests that enrollments are anti-cyclical, though a lot depends on specific circumstances of the country. On the negative side, households didreduce their consumptionof school supplies to cope with sliding incomes. As the crisis induced the governments to scale back their own provision of educational materials, the reduction inpurchases of school supplies i s disturbing, as it would have a higher effect on students from poor families, giventheir greater dependence on provision from the school and initial low levels of private supplies and consequent highmarginal impacts. The total of public education expenditures has not changed much because of the crisis, whether one looks at federal expenditures or provincial expenditures. However, a look at the total expenditures does not reveal the complete picture. Approximately 75 percent of federal education expenditures go to the federal universities, which have a potent lobby inthe political process, and university expenses were virtually untouched. The federal government also used to provide transfers to provincial governments with regard to salary expenditures for non-university higher education, and a so-called incentive fund that provided universal increments to teachers from an earmarked tax. The picture i s murkyregarding aspects of the federal salary-related transfers, but these expenditures have been cut back. The main impact of the crisis has been the drastic reduction of federal aid to provinces for compensatory programs and federal transfers to provinces to complete the reform process of extending the years of compulsory schooling. The federal government has replaced these expenditures with programs that are more of the nature of social assistance, such as nutritional programs to provide school lunches and cash-transfer programs to children from poor families. As regards the provinces, the main fiscal impact has been regardingthe cash position of provinces. There has been a significant increase inarrears on teacher salaries, and an increase inthe use of provincial bonos to pay teachers, with consequent erosion of purchasingpower.43 Payment problems for teacher salaries aggravated the labor-market conditions of teachers that are fraught with political tension even innormal times. The resultant loss o f teacher days resulting from officially registered strikes i s a lower bound on the real number o f school days lost, as in some of the provinces the school was officially open, but the only function was to provide the school breakfast or the school lunch. Teachers who suspended teaching activities apparently do not want to face the opprobrium of the community, particularly inthe case of poor communities where the nutrition providedby the school i s seen as a major benefit for the students.Even as the official data suggest tens of days lost becauseof strikes insome provinces, there i s anecdotal evidence to suggest that inmany cases, schools opened even on holidays as feeding centers for children. Even as highpriority was paid to school feeding programs, the erosion of the school's primary function as a place to provide education may result indamaging long-term effects, especially as school quality was not good in some schools inthe first place. The federal government discontinued the financing of earlier programs to help provinces with school construction and teacher training associated with the educational reform, as well as compensatory programs that provided school equipment and materials to schools serving children from poor families. The provincial governments, strapped for finances, have not been able to replace the loss of federal transfers with their own revenues or savings generated from 43These bonos are essentially provincially issuedmonetary instruments, which are accepted as a medium of exchange only within the province that issues them, and usually trade at a discount with respect to the peso. 36  efficiency improvements. Beginning in2001, but particularly in2002, the federal government began to devote substantially higher resources for transfers to provinces for school lunches (before the crisis, an exclusively provincial item of expenditures) and increased the amount o f financing for means-tested scholarships for secondary education. N o data are available that enable monitoring of the effect of the crisis on test scores of students. However, regression analysis on older, cross-sectional data on test scores can be used to measure the impact of variables that would have suffered changes because o f the crisis. A simple regression analysis usingtest score data from 2000 shows appreciable effects of what we term as "transitory variables." For instance, the magnitude of the effect of 10 days o f schooling lost i s equivalent to the effect of reducing by one year the education o f the mother. The analysis of impact on school quality shows the possible harmful long-term educational effects of the crisis. As children from poor families tend to have worse academic results because of a combination of family influences as well as poor provision of quality education, children from poor families suffer disproportionately when there i s a systemic decline ineducational quality. Impact on the Health Sector Four mainfactors can explain the acute effects of the crisis on the health sector: (1) a progressive deterioration inthe financing of the different subsystems (fell by 15.5 percent during the first 8 months of 2002); (2) a rise inthe cost o f imported medications and biomedical inputs as a result of devaluation; (3) an increase inthe demand for public hospital services; and (4) higher epidemiological risks. The first two factors reduce the effectiveness o f service supply, while the last two increase the demand for services. Health Insurance The crisis harmed the finances and ability to provide health care o f the principal subsectors offering formal health insurance coverage in Argentina - the SNSS, PAMI, the provincial public insurers and the private sector: SNSS. (The National Health Insurance System --SistemaNacional de Seguro de Salu#). Between the first semester of 2001 and the first semester of 2002, total revenues o f the SNSS-which covers 10.8 million citizens (roughly 30 percent of the population, mostly belonging to the formal e~onomy)~~-fell 12percent. By January 2002,62 percent o f the obras sociales nacionales (union organizedhealth programs --OSNs) inthe city of Buenos Aires did not comply with the provision of the mandatory health benefits package (PlanMkdico Obligatorio,PMO)to their affiliates. Inother regions of the country, noncompliance surpassed 80 percent. Ifcurrent regulations (on service provision, financial, economic, and administrative requirements) for national insurer funds were applied, a significant percentage o f the 280 existing OSNs would require in-depth restructuring or exclusion from the system. Today, 38 OSNs (covering 10percent of the insured population) are legally bankrupt. %Argentina, payrolldeductionsfinancetheSNSS.ThesystemisrunbytheObrasSociales, whichareunion- organized health insurance entities responsible for providing services for their members, mostly by contracting with private providers. 45This figure doesn't include the population affiliated to the Social Security Institute for Retirees and Pensioners (INSSJyP or PAMI). 37  PAMI. This programprovideshealthservices for retirees andpensioners andis responsible for 3.2 million people, most of them retired elderly. It saw its revenues fall by nearly 40 percent duringthe last months of 2001 and the beginning of 2002. Medical assistance (and pharmacy services) are periodically cut off for PAMIbeneficiaries throughout the country. The suspension o f basic services, the growth inwaiting lines for elective procedures, and the increase inpatient referrals to public hospitals all signal deteriorating quality. Provincialhealthfunds.The situationof most provincial public healthinsurers (obras sociales provinciales, OSPs) i s expected to be similar to that of the OSNs andPAMI, with some sporadic exceptions. But up-to-date consolidated information on the situation of the 24 autonomous OSPs, which provide insurance coverage to approximately 5.5 millionpeople (15 percent of the population, mostly provincial public employees and their dependents) i s not available. Private Subsector The private health insurance subsector has contracted as a result of the crisis. Between October 2001 and June 2002, the proportion of the population coveredby private insurance schemes fell 8 percent to under 3 million. Likewise, the percentage of the population with double insurance coverage decreased. The government has attempted to confront this health insurance crisis mainly by increasing revenues of national-level insurers via higher mandatory employer contributions. New legislative measures have increased employer contributions to OSNs and PAMI. These actions have incremented net revenues with respect to 2001, but not to prior levels, and mid- and long- term effects still need to be discussed. Simultaneously, an attempt has been made to strengthen the redistributioncapacity o f the Solidarity Redistribution Fund(FondoSolidario de Redistribucidn, FSR), which i s responsible for a partial redistribution of funds from higher- income workers to lower-income ones exclusively within the SNSS. Such actions, however, do not address chronic inefficiencies inthese subsectors. Structural reforms are pending. Reform attempts (before the crisis) that sought to contain costs and improve service provision inPAMI, the OSNs, and several of the OSPs failed. Chronic debts have kept growing, while public hospitals andout-of-pocket expenditures remain the escape goat for poor service responsiveness. Among the OSNs, the accumulated debt to external providers was approximately A$1,040 million inAugust 2002. InPAMI, by the end of 2001, it reached A$1,867 million (more than 75 percent of its total annual budget). This total was reached in spite of the national treasury's "additional" contributions of nearly A$1,700 million between 1995 and 2001. Public Subsector Public hospitals inArgentina, most of which belong to the provinces, are responsible for providing services to uninsured citizens, most of whom are poor. As a result of the crisis, these hospitals are confronted with (1) delays inthe flow of funds (a consequence of the poor fiscal situation of provinces plus difficulties with federal transfers), and (2) increases inthe demand for 38  services. Between October 2001 and May 2002, the percentage o f the population that had health coverage only through public hospitals rose from 36 to 43 percent. As a result, there has been inrecent months a serious crisis of medical supplies throughout the public hospital network. The lack of inputs (especially medications and surgical materials) has affected the quality of care. The situation inthe primary healthcare network i s also worrisome. Anecdotal evidence reveals difficulties intimely provision of basic preventive care (for example, toxoplasma and syphilis detection tests during pregnancy). The pressure for maintaining high-complexity expenditures may further shrink the already low resources allocated to primaryhealthcare (usually close to 8 percent o f the total sector budget in most provinces). A similar situation could occur with public health programs, precisely when they need to be reinforcedbecause o f greater demand and epidemiological risk. But, as with the health insurance subsector, it is necessary to acknowledge the existence of long-dated deficiencies inpublic hospitals that aggravate the current situation. An important deficiency i s the presence of a chronic regressive subsidy from public hospital budgets (intended for the poor uninsured) to health insurers. This subsidy arises when services are renderedto insuredpatients inpublic facilities without subsequent cost recovery by their third-party payers. It is estimated that around 30 percent of individualsreceiving care inpublic hospitals in Argentina have some type o f health insurance.46Billingto andcost reimbursementfrom insurers, however, is minimal, accounting for far less than 5 percent o f the total operatingbudget of those few hospitals that get involved in this process. Since 2000, and contrary to what would be expected, total public hospital billing to social security has gone down, and insurers are rejecting more bills. Impact on the Health of the Population As noted earlier, one coping strategy those affectedby the crisis use is to reduce out-of- pocket expenditures for health services and medicines, including reducing doctor visits. In addition, many people have lost, or have been forced to change their health insurance coverage. The deterioration of the social andeconomic conditions of a large part of the population, added to weakened national and provincial basic public health programs, i s expected to result in an increase inmorbidity and mortality among the population. Initial data suggest that mother and child health are at risk.The information system o f the PROMINproject reveals that between July 2000 and September 2001 there was a fall inthe coverage and quality o f mother-infant services ina group of healthcenters locatedinpoor urbanareas. As mentioned above, resultsfromthe World Bank Crisis Survey support these findings. While, on average, 38 percent of households report that since October 2001, they take their children to preventive medical controls less frequently, this change has been more severe inthe poorer quintile, in which 57 percent o f the households report a decrease inthe frequency o f preventive controls for their children. Vaccination programs have also suffered from irregular supply, and the Ministryof Health estimated that in 2001, 30 percent of the infant population did not have access to one o f the vaccine provided by the PAI, (Public Immunization Program) and effective coverage dropped in the national aggregate. 461997 Social DevelopmentSurvey and 2002 World Bank Crisis Survey. 39  Inadditionto maternal and childhealth, the deterioration of the social and economic conditions of the population let us anticipate that, if timely actions are not taken, the country will experience an increase inendemic and seasonaldiseases. The epidemiological surveillance system reports an increase in some endemic disease cases such as AIDS, Chagas disease, and leishmaniasis inthe first semester of 2002, as compared with 2001. As of August 2002, there was no increase inthe number of notified cases of acute diarrheic diseases, upper respiratory infections, or meningitis. But, inthe latter case, the percentage of cases with "nonspecified" etiology doubled compared with the same period in2001. Finally, there i s evidence of an increase inviolence against individuals and of mental healthproblems. According to the World Bank Crisis Survey, 20 percent o f the population reported that they have been the victims of crime or violence instances between October 2001 and June 2002. Moreover changes inthe mentaland emotional status of individuals are being observed, with a higher percentage of surveyed households reporting that they feel discouraged or without hope for the future, compared with how they felt before October 2001. Mental illnesses have already recorded a doubling of emergency consultations, and suicide i s a particularly worrying subject as it seems to be increasing among adolescents. Impactonaffordabilityof basic infrastructureservices The Government's immediate response to the crisis inthe infrastructure sectors was to pesify (convert from dollars into pesos) all tariffs for infrastructure services at a rate of 1:1and freeze them pending a general process of renegotiation with the private concessionaires. Despite initial plans to conclude the renegotiationprocess within 120days, utility tariffs have remained frozen innominal terms throughout 2002, incontrast with other basic consumer goods whose prices have risen by around 30% on average over the same period. Although this has avoided the social consequencesof a tariff increase, it i s not financially sustainable given that a significant proportion of the operating and financing costs of the infrastructureproviders are denominated in foreign currencies. As a result, operators have already defaulted on about half of their US$16 billion of dollar denominated debt. Some tariff increases therefore appear unavoidable if the solvency of the sector i s to be maintained. This i s evidently a significant social concern given that-ven at current frozen tariff levels-expenditure on householdutilities (electricity, gas, water, sewerage and telephone) accounts for 22% of the income o f the first quintile, and 16% of the income of the second quintile (Foster and CEER, Background Paper No. 9). These values represent a doubling of the household budget share allocatedto these services, when compared with the situation that prevailed in 1997 (when the last national expenditure survey was conducted). They are also high relative to typical international parameters o f between 10% and 15%. Clearly, any future tariff increases would further raise these shares, indicating a genuine affordability problem that needs to be addressed. 40  Sourceof finance Natureof benefit Beneficiaries ADESPA Cross-subsidy,tax exemption Exemptionfrom emergency Those meetingpovertycriteria or subsidy tariff increase as per census or household surveys. ADIGAS Cross-subsidybetween Exemptionfrom emergency Jefes y Jefas and low income customers. tariff increase pensiones,with 4 0 % average consumption. Electricity sector Off-settingtax reductionin tax Exemptionfrom emergency Unclear burden on the sector tariff increase Gas sector Off-settingtax reductionin tax Exemptionfrom emergency Unclear burden on the sector tariff increase Congress Exemptionfrom VAT, plus Tariff discount of 40% Beneficiariesof other social 'concessionairefinance' programs Metrovias Direct government subsidy Tariff discountof 20% Beneficiariesof other social (from Fondo Fiduciario) programs This situation has ledto widespread demands for the development of a social tariff, that would insulate the most vulnerable households from the impact of any future tariff increases. A number of specific proposals have been circulated (Table 3.4). Most rely on providing discounted tariffs for subsistence levels of consumption, or for families that are already beneficiaries of existing social programs, such as the Jefes y Jefus program, and most propose financing via cross-subsidies or fiscal credits. Nationalschemes Company specific schemes Electricity Electricity Tariff CompensationFund, financed Electricity distributorsin 18Provinces apply socialtariff from $0.024/kWh surchargeon wholesale schemes, some funded directly from the ETCF, others electricity market, raises annualresourcesof through cross-subsidy. $1" to finance a wide variety of tariff discountsto ruralconsumers, industrialcustomers, and social cases inthe Provinces. ~ Gas Patagoniangas subsidy, funded from surchargeof None. $0.004/m3on gas sales, raises annual resourcesof $1" to finance tariff discountsof around 75% on basic fuel requirementsin the Patagonian region. Telecom UniversalService Fund,financedfrom 1%levy on Major operatorsoffer discountedpayment plans to small telephonebills, raises $115mannually.These are consumersandpensioners,as well as discounted public intendedto cover a rangeof social and public payphoneservices inlow incomeareas andisolated telephony programs. However,the fund is not yet communities. operational. Transport Transport InfrastructureTrust Fund, financedfrom Most operatorspracticeanumber of specific franchises 18.5%tax on diesel, allocates40% of its resources for pensioners, school students, andother privileged to subsidizeurbanrailandbus services in GBA, groups. amountingto $170mannually. Water None. Water utilities inGBA and 7 Provincesapply social tariff schemes, some fundeddirectly from the General Budget, othersthroughcross-subsidy. Source: Adapte 41  The introduction of a social tariff of this kindi s complicated by the absence of any coherent safety net policy for the infrastructure sectors upon which such a mechanismmightbe built. Initially, the social dimension of infrastructure services was largely overlooked duringthe Argentine privatization program of the 1990s.This omission rapidly became apparent in numerous situations, prompting the development of a wide variety of ad hoc social policy mechanisms to finance tariff discounts in different sectors indifferent jurisdictions. These mechanisms primarily take the form of special surcharges on electricity, gas and petroleum products, which are earmarked for specific social purposes. Major examples include the Transport Infrastructure Trust Fundproviding operational subsidies to the urbantransport sector inGBA, the Electricity Tariff Compensation Fundfinancing a diverse array of tariffdiscounts in the Provinces, and the Patagonian Gas Fundreducingthe cost of gas for residents of the extreme south (Table 3.5). There has been a growing tendency for these resources to be channeled through extra- budgetary trust funds, rather than coming directly from the public budget (Figure 3.2). Although these instruments absorb a substantial and growing volume of resources, estimated to amount to $680m in 2003, the available evidence suggests that they are not particularly effective in reachingthe poor. This i s particularly unfortunate given the present need to mitigate the social concerns raised b y the renegotiation process andresultingtariff adjustments for essential services. Figure 3.2: Evolution of user subsidiesfor energy and transport services 1 800 Special Funds -2700 T " t ~ 600 National e E 500 Budget g 400 Transport 3 5 300 Special 0 3 Funds 200 Energy 100 0 National I 0 Budget 1999 2000 2001 2002 2003 Energy About three quarters of the resources channeled to the existing subsidy schemes take the form of supply side subsidies, primarily to transport operators. These are equivalent to an implicit price reduction of $0.09 per bus passenger, $0.14 per metro-rail passenger, and $0.25 per suburban railpassenger. Given that these subsidies benefit all customers, with no particular attempt to target resources to the poor, it i s not surprising to find that the distribution o f these resources i s skewed towards the upper middle income groups (Figure 3.3a). A recent study estimates concentration coefficients for subsidies receivedby metropolitan commuters that range from +0.09 for suburban rail, +O. 16for bus, and +0.36 for metro-rail. 42  Most of the remaining subsidy resourcestake the form of demand-side subsidies for energy consumption targetedto certain geographical areas, basedon criteria that have little to do with poverty. Thus, 92% of the resources of the ElectricityTariff Compensation Fundgo to the smaller Provinces of the interior which have only 44% of the nation's poor. Moreover, Provinces have a large degree of discretion over how these funds are used and it i s not clear to what extent they are allocatedto social tariff schemes. Similarly, 100%of the resources of the Patagonian gas subsidy go to the far south which has only 3% of the nation's poor. Moreover, even within Patagoniathe gas subsidy was found to have a concentration coefficient of +0.08, indicating that its effect on income distribution i s mildly regressive(Figure 3.3b). Seen from a national perspective, the Patagoniangas subsidy excludes 95% of poor gas consumers inArgentina, while 63% of its beneficiaries are non-poor households. Figure 3.3: Distributional incidenceof subsidy schemesinenergy and transport 35% , ~ I$35 ' =Average 30% $30 monthly 30% weekl; subsidy $1.50 subsidy 25% $25 25% (multi-modal) 20% $20 -+& Percentage 20% of 15% $15 beneficiaries 15% beneficiaries 10% $10 10% $0.50 4Percentage 5% of subsidy 5% +Percentage ~ $5 0% $0 0% I 1 2 3 4 5 $O.O0 of subsidy 1 2 3 4 5 Income quintile home quintile ~ (a) Patagonian gas subsidy (b) Metropolitan urbantransport subsidy Itis noteworthy that the water sector has no national social policy to comparedwith those found for electricity and gas. Notwithstanding, social tariff schemes have beendeveloped in GBA, as well as seven Provinces. These use awide variety of targeting mechanismsincluding neighborhood, retirement status, income level, and housing characteristics. The limitedevidence available suggests that these are relatively effective at targeting resourcestowards the poor. However, they are small inscale, benefiting no more than 100,000 households nationwide with an aggregatebudget of $10m. Simulations of alternative uses of these various funds show that targeting performance could be substantially improved if resourceswere allocated on the basis of a weighted average of poverty proxies based on housing characteristics, along the lines of what i s currently practiced in a number of Provincesthat provide social tariffs for water (Foster and CEER, Background Paper no. 9). Despite extensivediscussion of the needfor social tariffs, little has beendone to estimate the cost of such a policy. Inorder to provide an illustrationof the possible costs, it i s assumed that the policy objective would beto keep the cost of subsistence water, sewerage, electricity and gas bills for poor householdsbelow 15% of income. At current prices, the total cost of this policy would be around $300 million per year, which i s well within the current overall subsidy envelope for these sectors, but i s substantially more than the $210 million that are currently estimated to be spent on subsidizing water and energy tariffs. This assumes, however, that prices remainfixed 43  at current levels, which are financially non-viable for the providers. If all prices were to increase by IO%, the cost of the policy would escalateto $390m, while a 30% tariff increase would generate a cost of $560m. Looking further ahead, as incomes recover and the affordability concerns raisedby the current crisis begin to ease, the social policy framework should also consider measures to promote access to water, sewerage, and natural gas services among the lowest income groups, where substantial coverage deficits still remain. Inthe case of water and sewerage, there i s an evident public healthcase for seeking to achieve universal access. Inthe case of natural gas, those who lack access suffer a significant economic disadvantage, given that the closest substitute, LPGcylinders, cost five times as much per effective unit o f energy delivered. Inboth instances, the substantial connection charges and intra-household investments associated with connecting to the network, represent a major barrier to access for low income households. Furthermore, connection subsidies are by far the most effective way o f reaching genuinely poor households. Simulations performed in a recent study, find that the concentration coefficient associated with connection subsidies are strongly progressive, with values rangingfrom -0.30 to -0.40, compared with only -0.10 to -0.20 for even well-targeted consumption subsidies (see Background Paper No. 8). However, any subsidies given need to be carefully targeted and be affordable, and not become a fiscal drain on the government's general revenues. Such general subsidies, once enacted, become both costly anddifficult to remove. 44  CHAPTER4 POLICY IMPLICATIONSAND THE FUTURE This brief survey of current economic and social conditions describes the impact of the current crisis on the poor and how the current crisis has generated more poverty inthe country. There are some hopeful signs that the economy may be recovering in first half of 2003. These include an increase ineconomic activity, a stabilization o f the exchange-rate, and a reduction in the rate of inflation. Nevertheless, the economic situation remains precarious, and the prospects for a rapid and sustainable recovery o f the economy are uncertain. At the same time, the crisis itself constrains the actions of the government interms o f how it might respond to the social problems created b y the crisis: resources are scarce and options are few. The previous analysis, however, does point to some positive steps that a new government might consider. Short-Term Issues for a Pro-Poor Recovery Growthand Macro Policies4' Clearly, one o f the best ways to reduce poverty is to have high and sustained growth. Growth inturn depends on a stable macro-environment and sound economic management, both of which have been absent inrecent years. The Convertibility Plan, as successful as it was in controlling inflation after 1990,also brought with it all o f the negative effects of a fixed exchange-rate regime. Without the ability to adjust the real exchange rate, the government was left with two choices: borrow money or reduce nominal wages andprices. Given rigidities, the latter proved impossible, andwhen credit access was cut off after 1998, the long-term position became untenable. The restorationof a viable exchange rate, and the use of inflation to adjust real wages, has provided the basis for an improved fiscal balance and a potential recovery. The government, however, still faces serious issues in fiscal and financial sector management. Past actions, including the freeze on deposits, and the break o f the dollar-peso link, have reduced faith inthe financial system and eliminated the sources o f credit needed to finance a recovery. While the Government has liftedthe freeze on deposits ,the financial system remains weak. There i s a need to generate fiscal surpluses to reduce the large public sector deficit, which tripled inthe past year with respect to GDP as a result both o f devaluation, and the need to compensate banks for the conversion of their assets and liabilities into pesos. The task of generating a fiscal surplus will be difficult without efforts to broadenthe tax base and better enforce existing tax obligations. Fiscal relations between the federal government and the provinces need to be recast into a sounder andmore sustainable framework. Issues o f governance need to be addressed. Untilthese problems can be overcome, a strong recovery, and a substantial reduction inpoverty are unlikely. While good macro policies can provide a stable environment that promotes investment and growth, positive actions need to be taken by the Government to encourage strong, private sector led growth. One o f the most important i s to restore property rights and the health o f the 47For a more complete review of these issues see the forthcoming World Bank report on "Sources of Growth" 45  banking system. Other areas of importance include improving the regulatory framework for infrastructure, eliminating tax distortions, and reducing the costs of startinga business. Another area i s the need to provide more incentives to exporters, by eliminating export taxes and/or accelerating reimbursements of taxes paid by exporters. Recommendations to reduce taxes on exports, and on labor (see below), while increasing spending on education, health and social protection, while at the same time maintaining a fiscal balance may appear contradictory. Such steps are viable if taken inconcert with a general reform of the tax system, particularly one that would move to greater reliance on a progressive income tax and less reliance on the more regressive value-added tax. Such an approach would helpreduce the wide disparities inincome distribution which have been characteristic of the Argentine economy. Safety NetdSeverance Payments Government programs still do not provide an adequate safety net for the unemployed duringa recession. Itis clear from the history of the past ten years that Argentina continues to be faced b y recurring crises, with the potential for reduced wages and unemployment. Furthermore, the poor seem to be more prone to risks, interms of changes incomes, than the poor. Following the framework laidout inSecuring Our Future4*,we know that risks can be mitigatedeither through self-insurance, or through risk-pooling arrangements. There are few opportunities for riskpooling, since the present system of unemployment insurance is smallandinadequate. Workers who relied on self-insurance via accounts inbanks lost much of their safety net when these accounts were frozen, andconverted from pesos to dollars. Emergency public programs have helped. However, while the expanded Jefes programhas absorbed 2 million unemployed, and it accounts for 7.5% of the current expenditures o f the national government, it has reduced unemployment by at most 4 percentage points. Ingeneral, overall social expenditures still tend to fall with a crisis, although programs targeted to the poor have increased in2002. However, with the large increase inpoverty, targeted social spending per poor person has actually declined. Safety net programs are scattered across the government, and many programs are costly and duplicate other programs. Smaller, inefficient programs should be combined. Education The provision of educational services needs to be maintained, particularly with regard to payment o f teachers, as the welfare effects go beyond the immediate income levels of teachers. Whereas there may be heightened demand for social assistance (cash and food transfers), it needs to be considered carefully whether the objective of school-based transfer programs i s to provide or strengthen a safety net, or to further educational objectives. The danger of ambiguous objectives i s that there may be a failure to meet either one o f them. There may be more effective and efficient mechanisms of providing social assistance than through the educational system, and the possible illusion that a social-assistance policy i s educational policy may lead to a neglect of educational objectives. While cash-transfer programs designed to encourage school attendance may be helping to maintain school attendance, their effectiveness needs to be evaluated against other strategies for improving education. Federal government education allowances to provinces 48 DeFerranti et. al. 46  should be adjusted to insure equity betweenprovinces, with a bias toward those with the greatest poverty. Highpriority should be given to the strengthening of information systems andthe use of them to inform educational policy. Such measures should be complemented and strengthened through the encouragement of community participation. The all important question of the wage billandteacher deployment to schools is amenable to solutions through better use of existing information, - absent a concerted effort on these grounds, the employment o f teachers and school principals would continue to focus on generating public sector employment rather than providing quality educational services. Health Inthe midstof the on-going social crisis, the first priority inhealthisto guarantee access for all of the population, particularly the poor who lack insurance and have less ability to pay, to a minimumpackage o f treatments and medicines aimed at the major health risks. Inthis context, the Government should protect financing of priority programs such as those for maternal-child health, immunizations, and public health (epidemiological surveillance, control of transmittable diseases and others). More specifically, the gradual implementation o f an infant and maternal health insurance and the definition of and agreement on provincial health goals (with monitorable health indicators) seem to be two key initialsteps towards more effective protection for the poor uninsured. Other needed government actions must simultaneously address chronic inefficiencies in the health sector, freeing up scarce resources to support the above and improvingthe effectiveness of health care for the poor. Main actions here include eliminating inefficiencies and overlaps innational-level insurance programs such as PAM1andthe Obrus Sociules, and in the interfacebetween the public hospitals and the social security system (enhancing cost recovery by the former); strengthening the regulatory framework of the health sector, as well as the institutional capacity of the regulatory agencies; and improving the performance o f the public hospital network, emphasizing primary health care. Longer-TermPoverty Reduction Measures Inthelonger-term, anti-poverty effortsneedtofocus ontwobroadareas: First,reforms andpolicies that will leadto apatternof growth that will be more rapid overall, and feature a higher level of employment per unit of output. Second, provide a better and more secure safety net; and Third, improve the access of the poor to basic services, that will bothraise their overall welfare and, by improving their human capital, improve the productivity and their ability to compete in an increasingly globalized economy; Generating Labor Intensive Growth. Macro-economic policies that permit rapid and stable economic growth without inflation are an essential first step to a significant decline in poverty. Growth alone i s not sufficient, if it does not increase demand for labor, particularly the low skilled labor of the poor. This i s more likely to happen if Argentina's labor markets operate 47  efficiently. However, Argentina's labor market i s one of the most rigid andregulated inthe developing world, preventing wage adjustments from taking place easily. Some of key short term reforms that would facilitate a more orderly operation of the labor market include49: elimination o f centralized or sectoral collective bargaining agreements which are automatically extended to all workers in a sector, even if not signed and even when expired; reduce the highcost of labor by reducing labor taxes, including severance payments; allow temporary employment that i s not subject to payroll taxes, as under the former modalidadespromovidas, and extending programs which permit exceptions for small scale enterprises. Inthe longer term, the criticalproblemremains that alargepart of the labor force inthe informal sector lacks any form of pension or unemployment insurance coverage. A major reform of the labor laws that would reduce their present highlevel of protectiveness should be followed by an extension of coverage to small firms and others inthe now informal sector. At present, the highcost of complying with the labor laws serves as a major disincentive to firms moving to the formal sector; labor reform needs to reduce taxes on labor and improve real benefits so that there i s an incentive to being formal. The gap between formal and informal workers also helps aggravate the problem of the unequal distribution of income. Providing a More Secure Safety Net. The main safety net for workers inthe formal sector i s the severance payment owed to formal-sector workers uponjob termination. Inthe current crisis, however, the practice has been that this type of unemployment compensation leads to a negotiated settlement rather than a guaranteed benefit. For bankrupt firms, it i s likely there could be no payment at all since these obligations are not pre-funded. One possibility would be to obligate firms to pre-fund severance obligations in a segregated account. A better arrangement would be a major expansion unemployment insurance, with either a fixed benefit, or a benefit based on some proportion of average wages, and a fixed limit on total payments. These benefits might be financed from individual accounts supplemented by some guaranteed minimumpayment from the public sector. Such a system has to be carefully designed to ensure that unemployment insurance does not act as a major disincentive to searching for work, and that firms comply with requirements to collect and pay unemployment taxes. Benefits that decline over time, and stop altogether after a fixed period, help provide incentives for workers to continue searching for work. Alternatively, severance payment obligations of corporations could be fully funded and sequesteredfrom financial accounts of employers. However, any such reform should be extended to include more of the employees now inthe informal sector, who have neither severance payments nor unemployment benefits and are outside of the social security system. Such an extension, however, implies that these employees move from the informal to the formal system, and pay social security and other wage taxes, and are coveredby the labor code. This would further reduce labor market flexibility, given the rigidities inherent inthe present labor code. Thus, 49Admittedly, these reforms were more urgent during the Convertibility era, when flexibility indomestic wages and prices were the only adjustment mechanismsavailable. Inaddition, some reforms were undertakenduring the 1990s, and unemploymentincreased, raising questions about the strategy. However, the overall impact of reforms during this period had only a minor impact of labor market rigidity, and unemployment situation was aproduct of other factors, including slow growth and public sector privatizations. 48  unemployment insurance reformshould be accompaniedby a liberalization of the labor code, which would also have longer-term benefits interms of encouraging private sector growth (see above). While such an extension would have to be accompaniedby institutionalimprovements incollectionandenforcement, the muchhigherlevels of formal employment inneighboring countries, such as Chile and Uruguay, suggest that such an extension is quite feasible. Increasing Access to Services. A major effort should be undertakento raise the level and quality of education available to the poor, and increasetheir access to secondary and higher education. One of the key problems is that children of poor families are more likely to drop out of school for various reasons. A viable strategy ineducation would include: greater investments in secondary schools inpoor neighborhoods, such as by extending the presentPlan Social Educativo; expandedcash grants to poor families conditional on keeping children inschool particularly at the secondarylevel, in order to offset the economic incentives from school leaving and the effects of unemployment; the establishment of a system of partial cost recovery from students at public universities, who generally tendto be from non-poor families, and the establishment of a nation-wide system of scholarships for studentsfrom poor families. expanding the capacity of the current public university system, bothby improvements in operating efficiency andthrough further investments. While the situation inhealthis less critical, greater efficiency inthe health sector could improve the quality of service available to the poor. Particularly, the Government should focus public health care expenditures on those without health insurance, by improvingcost recovery from those with insurance and the ability to pay, and by improving the operating efficiency of the public hospital system. While granting more autonomy to public hospitals can improve their efficiency, care needs to be taken to avoidbuildinginincentives that will reduce services to the poor. Eventually, healthinsurancecoverage should be extended to those in the informal sector not presently covered. Existingprograms of maternal and child health (PROMIN) need to be expanded, and linkedwith family planning and reproductive health services for the poor, in order to reduce their currently highrate of fertility among the poor. Deficiencies ininfrastructure bothreduce the productivity of the poor, and limit human resourcedevelopment. The urban poor live in areas usually devoid of adequate sanitation and safe water, and often without pavedroads. Provision of such public services inpoor neighborhoods can improve health outcomes. But attention also needs to be focused on building upcommunities, especially inurbanareas, that lack roads, lights, and other services, anddo not have legaltitles to their land. The urban poor are particularly vulnerable to problems of crime and violence, and attention needs to be paid to alcohol and drug abuse, and improvements in police protection and access to justice. With the privatization of infrastructure services, regulated pricing of services needs to bereviewed to that an implicit cross-subsidiesbetween users benefit the poor. Cross-subsidiesthat reduce the costs of connection charges for water and gas would be one usefulmechanism provided it was properly targeted. 49  Year relative to crisis Country andcrisis year -3 -2 -1 0 1 2 3 GDP growth (percent) Argentina, 2002 -3.4 -0.8 -4.4 -10.9 Indonesia, 1998 8.4 7.6 4.7 -13.1 0.8 4.9 3.1 Korea, 1998 8.9 6.7 5.O -6.7 10.9 9.3 3.O Malaysia, 1998 9.8 10.0 7.3 -7.4 6.1 8.3 0.4 Russia, 1998 -4.1 -3.4 0.9 -4.9 5.4 8.3 Thailand, 1997 8.9 9.3 5.9 -1.4 -10.8 4.6 4.7 Argentina, 1995 11.9 5.9 5.8 -2.8 5.5 8.1 3.8 MCxico, 1995 3.6 1.9 4.4 -6.2 5.1 6.8 5.O Chile, 1982 8.7 8.1 4.7 -10.3 -3.8 8.0 7.1 Change in exchange rate (percent) Argentina, 2002 0.0 0.0 0.0 210.0 Indonesia, 1998 3.9 4.0 19.5 70.9 -27.5 6.7 12.3 Korea, 1998 -4.2 4.1 15.4 32.1 -17.9 -5.1 11.9 Malaysia, 1998 -4.8 0.5 10.6 28.3 -3.3 0 0 Russia, 1998 11.0 11.5 40.4 60.6 12.5 3.6 11.0 Thailand, 1997 -0.7 -0.9 1.7 19.2 24.2 -9.4 5.7 Argentina, 1995 3.7 0.0 0.0 0.0 0.0 0.0 0.0 MCxico, 1995 2.5 0.7 7.7 47.4 15.5 4.0 13.3 Chile. 1982 17.7 4.7 0.0 30.5 54.8 25.00 63.3 Argentina, 2002 14.2 15.1 17.4 19.7 Indonesia, 1998 4.1 4.7 5.4 6.4 6.1 Korea, 1998 2.0 2.0 2.6 6.8 6.3 4.1 Malaysia, 1998 3.1 2.5 2.5 3.2 3.4 3.1 3.9 Russia, 1998 9.5 9.7 11.8 13.3 13.4 11.4 9.5 Thailand, 1997 1.1 1.1 0.9 3.7 5.2 3.0 2.4 Argentina, 1995 7.0 9.3 12.2 16.6 17.3 13.7 12.4 MCxico, 1995 3.1 3.2 4.2 5.7 4.3 3.4 2.9 Chile, 1982 13.6 10.4 11.3 19.6 14.6 13.9 12.1 Scenariosfor Recovery and Long-Term Poverty Reduction It is difficult to estimate the nature of apossible recovery path for the economy, given uncertainties inmany areas. Table 4.1 gives some comparative data for eight countries emerging from crises inrecent years. It shows that for most countries, the year of the severest recession was typically followed by a year o f substantial growth. On average, the year after recession in the sample inTable 4.1 shows a positive growth rate, except inthose countries (Chile, 1982; Malaysia, 1997) where the crisis lastedtwo years. B y the second year, all countries inthe sample show positive growth, averaging here about 7% in the second year and about 4 9%inthe third (however, remember that this i s a small sample of countries). Interms of exchange rates, no clear pattern emerges. While some countries show exchange rate appreciation, others show 50Average of May and October EPHsurveys. 50  continued exchange rate depreciation. Unemployment, however, seems to be relatively stable despitethe recovery inmost countries, showing no tendency to decline markedly despite rapid GDPgrowth. For the current year, most forecasts estimate arecovery of GDP of about 4% for the economy. Infollowing years, if the historical record i s any guide, growth couldbe strong. However, forecastshere are difficult to make, and a great deal depends on the economic program of the new administration that took over inMay 2003. Poverty Measurement For many years, Argentina has suffered from a lack of informationon income and social conditions inrural areas. As poverty i s likely higher inthese areas, and these areas are probably underservedby government services, their exclusion from the survey tends to understatethe magnitude of the poverty problemand distorts the calculation of income distribution. Infact, INDEC has plansto extend the EPHto ruralareas, starting with surveys to be undertaken in 2004. Work i s also beingdone to develop regional prices indices. However, poverty measurementshould move from income measuresto a consumption aggregate measurethat would give a more accurate measureof welfare, and be less prone to temporary fluctuations. A new expenditure survey i s plannedfor 2004 that will include rural areas. Expenditures surveys need to be undertaken on a more regular basis, perhaps annually. For a fuller discussion of these issues, see Background Paper No. 8 by Haeduck Lee. 51  References Adam, Sarah. and C. Chamberlain. 1999. "Thailand Social Monitor. Coping with the Crisis inEducation and Health." World Bank Report 24467 TH. Washington, DC. Available at: http://www.worldbank.or.th/social/pdf/social-part-i.pdf. Datt, Gaurav and MartinRavallion. 1992. "Growth and Redistribution Components of Changes in Poverty: A Decompositionwith Application to Brazil and India", Journal of Development Economics, Vo1.38, April, pp. 275-295 De Ferranti, David, Guillermo Perry, Indermit Gill, Luis Serven, Francisco H.G. Ferreira, NadeemIlahi, William Maloney, and MartinRama. 2000. Securing Our Future in a Global Economy. World Bank Latin American and Caribbean Studies, Washington, D.C.: World Bank. Fallon, P.R., and R.E.B. Lucas. 2002. "The Impact of Financial Crises on Labor Markets, Household Income and Poverty: A Review of Evidence." The World Bank Research Observer 17(1). Frandengerg, E.,D.Thomas, and K.Beagle. 1999. "The Real Costs of Indonesia's Crisis: Preliminary Findingsfromthe IndonesiaFamily Life Surveys." Labor andPopulationProgramWorking Paper Series 99-04. SantaMonica, Calif.: Rand Corp. Halac, Marina and Sergio L.Schmukler. 2002. "Distributional Effects of Crises: The Role of Financial Transfers." Preliminary draft, January 21. Jalan, J. and M.Ravallion. 1999. "Income Gains to the Poor from Workfare: Estimates for Argentina's Trabajar Program." World Bank, Washington, D.C. Lustig,N. 1998. "Mexico: The Remakingof an Economy." WashingtonD.C.: Brookings Institution. Sumatra S., A. Wetter berg, and Lant Pritchett. 1999. "The Social Impact of the Crisis in Indonesia: Results from a Nationwide Decimating Survey." East Asia Environment and Social Development UnitReport 21249. Thomas, D., E.Frankengerg, K.Beegle, and G. Teruel. 1999. "Household Budgets, Household Composition and the Crisis inIndonesia: Evidence from Longitudinal Household Survey Data." Paperpresentedat the 1999 Population Association Meeting of America, New York. 52  STATISTICAL ANNEX Table A.1 PovertyRates, Poverty Gap, and Seventy of Poverty inUrbanArgentina, 1990-2002 A.2 LaborForceParticipationandUnemploymentRateby Poverty Group, 1990-2002 A.3 LaborForceParticipationandUnemploymentRateby Gender, 1990-2002 A.4 Changes inIncomeDistributioninUrbanArgentina, 1990-2002 AS Changes inExtremePovertyby Region, 1990-200 A.6 ChangesinPovertyRateby Region, 1990-2002 A.7 EmploymentandUnder-Employmentby PovertyGroup, 1990-2002 A.8 Poverty ProfileinUrbanArgentina, 2002 A.9 Poverty Rateby Age Group andGender inUrbanArgentina, 2002 53  Sources: INDEC,Encuesta Permanente de Hogares (EPH), 1990-2002. 54  TableA.2: LaborForce ParticipationandUnemploymentRateby PovertyGroup, 1990-2002 I I I I Labor ForceParticipation (populationof i :e 15-64) Year bonth I Indigent I I UnemploymentRate Moderate ooor Non- All Urban Moderate Non- All Urban poor Areas Indigent poor poor Areas Sources: INDEC, Encuesta Permanente de Hogares (EPH), 1990-2002 55  TableA.3: LaborForce Participationand UnemploymentRateby Gender, 1990-2002 Labor Force Participation (populationof age 15-64) UnemploymentRate Year Month Male Female All Gender Male Female All Gender 1990 May 81.4 42.0 60.7 9.3 9.5 9.3 October 81.3 47.7 63.9 15.9 20.8 17.3 1997 May 81.3 48.8 64.5 14.2 19.8 16.1 October 80.8 48.9 64.3 12.0 17.0 13.7 1998 May 81.6 49.0 64.6 12.5 15.0 13.2 October 81.O 48.5 64.0 11.5 14.5 12.4 1999 May 80.3 50.2 64.6 13.9 15.9 14.5 October 80.2 49.9 64.4 13.1 15.9 13.8 2000 May 79.4 49.8 63.9 14.9 17.8 15.4 October 79.8 50.8 64.6 13.7 17.1 14.7 2001 May 80.0 50.2 64.4 16.4 17.2 16.4 October 79.3 49.4 63.7 19.2 18.6 18.3 2002 May 78.6 48.8 63.0 23.2 21.1 21.5 October 78.8 52.4 64.9 18.5 5.2 17.8 56  TableA.4: ChangesinIncomeDistributioninUrbanArgentina, 1990-2002 .... ..........,.. ....,.. ......,,. .. .. Y E A R ~ .... .~ xcapitahousehold 1990 1992 1994 1996 1998 1999 2000 2001 2002 2002 come ...., Oct. .... ..... ......... ........ ... ... Oct. Oct. ............ ..... .....,. .... Oct. Oct. Oct. Oct. Oct. May Oct.. verage income (in 1999 uesos) Overall 240.3 292.7 302.8 270.4 294.8 280.6 275.7 253.3 189.7 176.8 1 38.3 45.6 43.2 28.1 31.7 30.4 26.3 17.1 9.0 16.1 2 73.3 87.3 85.2 67.9 70.1 68.8 62.2 49.7 31.6 34.2 3 96.8 117.7 116.2 96.9 100.3 98.3 89.9 77.0 51.6 52.3 4 122.2 150.3 149.7 125.9 132.3 128.4 119.8 104.0 73.7 72.5 5 153.5 181.1 188.1 158.1 167.3 164.2 155.1 136.3 97.9 95.8 6 187.2 223.8 229.9 198.7 209.5 206.1 197.8 176.4 129.7 121.8 7 227.1 276.3 285.0 249.8 267.8 260.5 251.2 228.0 170.1 155.6 8 287.1 351.8 367.4 326.3 353.5 340.1 336.2 306.3 228.6 207.8 9 391.5 487.8 502.0 460.0 500.9 480.6 476.8 444.4 334.7 306.3 10 825.5 .................... 1,004.7............. .............. 1,060.4....... 992.3 " 1,114.2 1,028.2 1,041.9 993.8....... ?!??%" 705.3 .. ..... come share (%) 1 1.6 1.6 1.4 1.o 1.1 1.1 1.o 0.7 0.5 0.9 2 3.1 3.0 2.8 2.5 2.4 2.5 2.3 2.0 1.7 1.9 3 4.0 4.0 3.8 3.6 3.4 3.5 3.3 3.0 2.7 3.0 4 5.1 5.1 4.9 4.7 4.5 4.6 4.3 4.1 3.9 4.1 5 6.4 6.2 6.2 5.8 5.7 5.9 5.6 5.4 5.2 5.4 6 7.8 7.6 7.6 7.3 7.1 7.3 7.2 7.0 6.8 6.9 7 9.5 9.4 9.4 9.2 9.1 9.3 9.1 9.0 9.0 8.8 8 11.9 12.0 12.1 12.1 12.0 12.1 12.2 12.1 12.1 11.8 9 16.3 16.7 16.6 17.0 17.0 17.1 17.3 17.5 17.6 17.3 ....................... 10 ., ............ 34.4 34.3 .........35.0 36.7 37.8 36.6 37.8........ ..__ ~....... ............. .... 39.2 40.6 39.9 ..........,... )p 20% share 50.7 51.0 51.6 53.7 54.8 53.8 55.1 56.8 58.2 57.2 ittom20% share 4.6 4.5 4.2 3.6 3.5 3.5 3.2 2.6 2.1 2.8 )p/Bottom20% !lo........... ................................ 11.0 11.3 12.3 14.9 15.7 15.4 17.2 21.8 ....... " ....... 27.7 20.4.,. . . inicoefficient 0.454 0.456 0.467 0.493 0.504 0.494 0.510 0.530 0.551 0.532 )pulation presentedby the mple (1,000) 14,715 16,079 17,920 17,772 20,674 20,581 20,920 20,882 20,967 21,360 urces:INDEC, EncuestaPermanente de Hogares (EPH),October of 1990, 1992, 1994, 1996, 1998, 1999,2000,2001, and May and October of 2002. 57  TableA.4: Changes inIncomeDistributioninUrbanArgentina, 1990-2002(Cont'd) Y E A R "..... ........ ~ " 'er capita 1990 1992 1994 1996 1998 1999 2000 2001 2002 2002 . ousehold income Oct. ,..,.,,.. Oct................, ..Oct. ..,........ ,.. oct.,,... Oct............,......................,....,.. Oct. " Oct. Oct: .. ..MaY Oct. iverageincome (in 1999pesos) Overall 240.3 292.7 302.8 270.4 294.8 280.6 275.7 253.3 189.7 176.8 1 38.3 45.6 43.2 28.1 31.7 30.4 26.3 17.1 9.0 16.1 2 73.3 87.3 85.2 67.9 70.1 68.8 62.2 49.7 31.6 34.2 3 96.8 117.7 116.2 96.9 100.3 98.3 89.9 77.0 51.6 52.3 4 122.2 150.3 149.7 125.9 132.3 128.4 119.8 104.0 73.7 72.5 5 153.5 181.1 188.1 158.1 167.3 164.2 155.1 136.3 97.9 95.8 6 187.2 223.8 229.9 198.7 209.5 206.1 197.8 176.4 129.7 121.8 7 227.1 276.3 285.0 249.8 267.8 260.5 251.2 228.0 170.1 155.6 8 287.1 351.8 367.4 326.3 353.5 340.1 336 2 306.3 228.6 207.8 9 391.5 487.8 502.0 460.0 500.9 480.6 476.8 444.4 334.7 306.3 10 825.5 1,004.7 1,060.4 ........................................ 992.3 1,I 1,028.2 14.2 1,041.9 .... 993.8 ... ....................... 769.9 .705.3 ... ncomeindex (base=1999) Overall 1 85.6 104.3 107.9 96.4 105.1 100.0 98.3 90.3 67.6 63.0 2 3 126.0 150.0 142.1 92.4 104.3 100.0 86.5 56.3 29.6 53.0 4 106.5 126.9 123.8 98.7 101.9 100.0 90.4 72.2 45.9 49.7 5 98.5 119.7 118.2 98.6 102.0 100.0 91.5 78.3 52.5 53.2 6 95.2 117.1 116.6 98.1 103.0 100.0 93.3 81.o 57.4 56.5 7 93.5 110.3 114.6 96.3 101.9 100.0 94.5 83.0 59.6 58.3 8 90.8 108.6 111.5 96.4 101.6 100.0 96.0 85.6 62.9 59.1 9 87.2 106.1 109.4 95.9 102.8 100.0 96.4 87.5 65.3 59.7 ....... ............,....., .............,..................... 10 ,.. 84.4 ..................................................... 103.4 .............108.0 95.9 ............................. ...................... 103.9 100.0 98.9 ... 90.1 ........................ ............... 67.2 61:1 Population representedby the sample (1,000) 14,715 16,079 17,920 17,772 20,674 20,581 20,920 20,882 20,967 21,360 Sources: INDEC,EncuestaPennanente de Hogares (EPH), October of 1990,1992,1994,1996,1998,1999,2000,2001, and May and October of 2002. 58  i All Year Month GBA Northwest Northeast Cuyo Pampeana Patagonia Regions 1990 May 12.5 19.1 19.1 14.6 9.3 11.9 12.6 October 6.6 10.2 17.1 8.8 5.9 4.1 9.9 1991 May 5.1 8.7 12.0 4.2 4.6 6.3 5.7 October 3.0 8.5 12.5 4.9 4.0 4.6 4.5 1992 May 3.3 9.7 10.9 6.1 4.5 4.3 4.8 October 3.2 9.2 11.4 4.2 4.1 5.5 4.5 1993 May 3.6 8.6 10.6 4.1 4.5 5.4 4.7 October 4.4 8.5 9.4 3.5 4.6 3.7 5.0 1994 May 3.3 8.4 11.1 3.4 4.2 5.2 4.3 October 3.5 7.6 10.4 3.9 5.1 5.7 4.6 1995 May 5.7 8.4 11.0 5.6 8.5 5.5 6.8 October 6.3 11.2 12.6 8.6 7.5 7.2 7.6 1996 May 6.9 11.6 14.8 7.8 8.5 7.0 8.2 October 7.5 13.1 17.9 8.8 10.1 7.6 9.1 1997 May 5.7 10.9 14.8 9.2 7.5 6.0 7.3 October 6.4 11.8 15.0 8.7 7.8 6.2 7.8 1998 May 5.3 11.7 15.0 6.8 7.6 6.5 7.1 October 6.9 11.7 17.7 7.6 8.5 8.0 8.5 1999 May 7.6 12.9 17.7 8.0 8.0 6.9 8.9 October 6.7 11.0 17.6 7.3 8.4 7.5 8.3 2000 May 7.5 12.2 18.6 8.3 8.8 7.6 9.0 October 7.7 13.3 19.8 10.7 9.0 7.2 9.6 2001 May 10.3 14.4 22.9 11.0 11.3 6.9 11.6 October 12.2 16.2 26.9 12.2 13.7 7.3 13.6 2002 May 22.7 29.5 38.8 24.7 25.1 15.4 24.8 October 24.7 35.1 41.9 29.7 27.2 21.o 27.5 1990-1994 -47.0% -25.5% -39.2% -55.7% -13.6% 39.0% -53.5% 1994-1998 97.1yo 53.9% 70.2% 94.9% 66.7% 40.4% 84.8% Sources: INDEC, EncuestaPermanentede Hogares(EPH), 1990-2002. Note: % changes are measuredfor four-yearintervalsfrom October of the base year to October of the endyear. 59  Table A.6: ChangesinPoverty Rate by Region, 1990-2002 All Year Month GBA Northwest Northeast Cuyo Pampeana Patagonia Regions 1990 May 42.5 52.5 54.5 46.9 37.5 35.5 42.7 October 33.7 43.5 50.5 37.2 27.2 16.3 38.1 1991 May 28.9 39.0 47.7 30.1 25.5 22.4 30.1 October 21.5 38.8 44.7 28.2 21.5 17.9 24.9 1992 May 19.3 37.1 43.8 28.9 23.6 19.0 23.9 October 17.8 36.9 42.6 24.1 20.4 18.2 22.0 1993 May 17.7 35.6 40.8 24.4 20.6 18.4 21.5 October 16.8 35.4 38.6 22.1 19.8 14.9 20.6 1994 May 16.1 35.5 39.6 21.4 20.1 16.4 20.1 October 19.0 37.2 40.8 22.8 21.o 18.3 22.4 1995 May 22.2 37.2 43.8 29.2 27.1 18.9 26.1 October 24.8 41.5 45.3 36.0 27.2 22.2 28.7 1996 May 26.7 44.2 46.2 1 31.3 29.2 20.8 30.1 October 27.9 44.3 50.2 34.9 30.9 22.9 31.5 1997 May 26.3 40.6 48.8 36.4 29.2 22.0 30.0 October 26.0 41.9 48.3 32.0 28.3 21.7 29.5 1998 May 24.3 40.1 48.5 32.8 28.6 21.o 28.8 October 25.9 42.2 50.3 32.7 29.2 22.6 30.2 1999 May 27.1 11 45.5 50.2 33.8 29.3 22.3 31.3 October 26.7 41.2 50.6 33.3 29.3 24.5 30.6 2000 May 29.7 44.5 53.0 36.5 30.8 23.1 33.4 October 28.9 43.6 53.1 39.2 30.4 24.3 32.8 2001 May 32.7 47.5 56.6 38.6 33.8 23.9 35.9 October 35.4 48.3 57.2 39.6 37.1 23.2 38.3 2002 May 49.7 63.5 69.8 54.9 52.7 39.1 53.0 /October I 54.3 1 69.4 I 71.5 1 61.3 1 56.7 1 45.6 1 57.5 % Change (October to October): 1990-2002 I 59.5% 1 41.6% I 64.8% 1 108.5% 1 179.8% I 50.9% -%changesaremeasuredforfour-yearintervalsfromOctoberofthebaseyeartoOctoberoftheendyear. Sources: INDEC,EncuestaPermanente de Hogares (EPH), 1990-2002. Note: 60  % Working less than % Working between % Working more than ---t-t- I I 101I Hoursber week I 10-30Hours Der week 11 30 Hours-De week I I I I I I Year IMonth 1Indigen Moderatc Non- All Indigent Moderate Non- All Indigent Moderate \Jon- All poor poor Urban poor poor Urban poor poor Urbar Areas Areas Areas 0.8 0.8 0.9 16.5 12.5 10.8 11.7 46.4 69.4 61.4 75.3 1.1 1.0 1.2 10.9 11.7 11.6 11.5 57.0 74.4 60.9 78.0 1.9 1.1 1.3 15.4 11.3 12.0 12.0 43.8 71.3 78.9 76.3 October 1.3 1.0 1.1 11.9 10.9 11.5 11.5 48.1 72.9 60.6 78.7 2.6 1.3 1.5 14.2 12.8 12.1 12.2 43.7 68.2 78.9 76.5 October 4.7 2.0 1.2 1.4 14.9 11.5 11.7 11.8 38.0 66.4 79.8 77.1 1993 May 4.0 3.5 1.8 2.0 10.9 12.0 11.9 11.9 31.3 60.9 76.2 73.1 October 4.5 2.5 1.4 1.6 13.9 12.9 11.6 11.8 31.2 61.2 77.1 74.0 1994 May 5.3 2.9 1.9 2.1 14.6 12.4 12.1 12.2 27.8 59.6 75.8 72.7 October 5.2 4.2 1.7 2.1 10.4 12.8 11.6 11.7 27.6 56.7 75.6 71.8 1995 May 4.8 2.9 1.7 2.0 10.3 12.2 11.7 11.7 23.8 50.0 70.3 65.0 October 6.8 4.3 2.0 2.7 14.9 14.5 12.2 12.7 24.5 51.3 71.0 65.3 1996 May 6.0 3.9 2.0 2.6 13.7 12.6 12.2 12.3 28.9 52.8 71.8 65.8 October 4.7 4.4 71.4 64.6 1997 May 7.7 4.9 71.7 66.1 October 8.8 4.4 73.6 68.0 1998 May 6.2 5.1 73.4 68.2 October 7.0 5.4 74.1 68.4 1999 May 8.7 4.3 72.6 66.3 October 7.0 5.4 72.2 66.6 2000 May 9.1 5.6 71.3 64.2 6.5 72.0 65.1 5.3 71.3 63.5 5.8 68.9 59.4 5.5 2.5 4.6 20.9 17.7 14.4 16.6 23.7 49.4 68.0 54.1 October 7.0 4.4 2.6 4.1 30.8 21.5 14.9 20.5 25.5 51.5 71.0 55.1 61  Poverty Group Indigent Moderatepoor Non-poor Overall Family size 4.9 4.0 2.7 3.5 Adult equivalent family size 3.8 3.2 2.2 2.8 % female-headedhouseholds 24.0 22.8 32.1 28.1 Average age 23.3 29.2 38.0 31.3 Years of schooling among adults aged 20 and over 7.9 9.3 11.7 10.2 Labor force participation(am0ng those of age 15-64) 59.2 62.2 69.8 64.9 % informal sector 39.0 39.5 33.2 36.3 Unemploymentrate 34.9 20.7 9.7 17.8 Hoursworked 29.1 38.6 42.0 30.7 % working < 30 hrslweek 37.8 25.9 17.5 24.6 62  Total population Female population Age Group % indigent I % poor % indigent 1 % poor Total 27.4 57.5 26.7 56.1 0-4 40.0 71.4 42.3 72.9 5-14 42.0 74.9 40.9 74.1 15-19 37.6 70.2 35.5 69.0 Sub-total, 0-19 40.4 72.9 39.8 72.5 20-34 23.1 53.7 23.4 53.6 35-49 25.5 55.5 25.1 55.2 50-64 18.0 44.6 16.7 42.5 Sub-total, 20-64 22.7 52.2 22.4 51.6 65 & older 7.1 30.4 6.5 27.2 63