December 2011 Document of the World Bank 66146 ON THE EDGE OF UNCERTAINTY Poverty Reduction in Latin America and the Caribbean during the Great Recession and Beyond DECEMBER 2011 CONTENTS Introduction 5 I. Poverty reduction in Latin America during the growth rebound of 2010 7 II. Understanding the changes in poverty 13 III. Going forward 17 Annex A. Why past poverty estimates may change? 20 B. Annex Tables 21 Bibliography 26 End Notes 27 This brief was produced by the Poverty, Gender and Equity Unit from the Poverty Reduction and Economic Management Team (LCSPP) in the Latin America and Caribbean Region. The work was led by Joao Pedro Azevedo, Louise Cord and Carolina Díaz- Bonilla. The team included Bernardo Atuesta Montes, Andrés Castañeda Aguilar, Gabriel Facchini, Amer Hasan and Viviane Sanfelice. The team benefited from comments from Paloma Anos Casero, Francisco H. Ferreira, Margaret Ellen Grosh, Gabriela Inchauste, Julian Messina, Luis Felipe López Calva, Marcela Sanchez-Bender, and from key inputs from Samuel Freije-Rodriguez, Leonardo Gasparini and Leopoldo Tornarolli. The views and interpretations expressed here are the sole responsibility of the authors and should not be attributed to the World Bank, the Board of Executive Directors or the countries which they represent. The numbers presented in this brief are based on a regional microdata harmonization effort (called SEDLAC) to increase cross-country comparability of selected variables from the official household surveys. Given this objective, often the indicators constructed through this process are different from the official statistics, reported by Governments and National Offices of Statistics. Such differences should not be interpreted in any way as an indicator of methodological superiority since both numbers depart from different, yet equally important and valid, objectives: regional comparability and best possible national representation. The welfare aggregate used in this study are income based. List of Figures Figure 1: Growth and Poverty Reduction Rebounded Strongly in 2010 7 Figure 2: The Strong Link Between Poverty Reduction and Growth Resumed in 2010 8 Figure 3: Growth Resumes in all Regions of Latin America 9 Figure 4: Poverty Levels in Central America and Mexico Are Less Responsive to the 2010 Economic Rebound 9 Figure 5: Rural households were more resilient to the 2008 and 2009 crises - yet were more closely tied to the 2010 economic recovery 9 Figure 6: In 2007 and 2008, higher food prices kept poverty from falling further than it did 10 Figure 7: Extreme poverty trends using food and overall CPI 11 Figure 8: Growth has been the predominant factor contributing to poverty reduction 14 Figure 9: Labor income is most important for poverty reduction 14 Figure 10: Men in the labor market were more vulnerable to the crisis than women. 15 Figure 11: Poverty Fell More Amongst Households With Small Children (0-14 year olds) 15 Figure 12: The extreme poor are more reliant on non-labor income than the overall population 16 Figure 13: Poverty reduction in the future may be slower 18 Figure 14: Selected labor markets in Latin America support poverty reduction in 2011 except for Mexico 18 Figure 15: The region’s fiscal deficit (% of GDP) has increased 19 List of Tables Table B.1: Datasets used by country 21 Table B.2: Regional poverty numbers 22 Table B.3: Extreme and Moderate Poverty by country and year 22 Table B.4: GDP per capita, PPP (constant 2005 international $) 23 Table B.5: Gini by Country and Year 24 Table B.6: Poverty Projections 24 Figure B.1: Moderate and extreme poverty trends using the food CPI 25 4 LAC POVERTY AND LABOR BRIEF Introduction D espite a 2.8 percent decline in Gross Domestic Product (GDP) per capita in purchasing power parity (PPP) termsi, poverty levels in Latin America (LAC) remained basically static during the Great Reces- sion, as the poor were shielded from the economic crisis in some countries and continued to benefit from growth in others.ii In 2010, poverty reduction resumed sharply in Latin America, as household incomes were once again closely linked to economic growth at rates similar to pre-crisis years. Moderate poverty de- clined by almost 2.5 percentage points to reach 28 percent in 2010, while extreme poverty fell by more than 2 percentage points to reach 14 percent. As 2011 comes to a close, once again the global economy and Latin America are facing risks of yet another economic slowdown. Using household survey data from 2010 and selected labor market indicators through the third quarter of 2011, this note identifies some basic facts on the impact of the crisis and the recovery on the poor and explores their implications for poverty reduction in the region going forward. LAC POVERTY AND LABOR BRIEF Growth in mean incomes and a more equitable distribution of income each accounted for about half of poverty reduction during the last two years (2009 and 2010). Disaggregating changes in household income reveals that rising labor incomes were the most important factor behind poverty reduction. Nevertheless, during the crisis, the contribution to poverty reduction of non-labor income (public and private transfers) grew. Not surprisingly, households experiencing increases in both labor and non-labor incomes had an extra boost in moving above the poverty line. Labor market incomes for men were highly vulnerable during the crisis and experienced an overall decline in 2009 that, holding all else equal, contributed to higher poverty levels. In contrast, rising female labor incomes played a larger role in reducing poverty in 2009, accounting for half of the poverty reduction, while households with increases in both male and female labor income accounted for the other half of poverty reduction. 5 In 2010, growth in male labor incomes once again became the driving force of poverty reduction in Latin America. Of all the demographic groups, households with young children (0-14) were the most vulnerable to the crisis, but also responded fastest to the recovery in 2010. This may reveal these households’ dependence on male labor force earnings, which suffered significantly in 2009, but quickly bounced back in 2010. Projections for 2011 show continued poverty reduction for Latin America, but there is significant uncertainty as to whether this trend will continue into 2012 given concerns of a global slowdown. The experience of 2009 and 2010 suggests that the impact of any future economic crisis will depend upon the extent to which household incomes move in tandem with GDP per capita growth, as well as the ability of poor male and female workers to experience gains in labor income, and to a lesser extent growth in non-labor income. A new slowdown could undermine poverty reduction more than the 2009 crisis should labor markets soften, especially for the poor, and given that governments face heightened fiscal constraints compared to 2009, mak- ing a compensatory expansion of transfers more difficult. Section I examines trends in poverty reduction in LAC during the growth rebound of 2010, section II identifies factors that contributed to changes in poverty in recent years (2005-2010), while section III draws lessons for policymakers going forward. LAC POVERTY AND LABOR BRIEF 6 I. Poverty reduction in Latin America during the growth rebound of 2010 S trong poverty reduction in Latin America resumed with the growth rebound in 2010, as both mod- erate and extreme poor households benefitted from the recovery, accelerating poverty reduction to rates similar to those witnessed between 2003-2006 (Figure 1). The estimated 5 percent increase in GDP per capita that LAC experienced in 2010 led to declines in moderate and extreme poverty of close to 2.4 and 2.1 percentage points, respectively. Moderate poverty dropped to 28 percent, resulting in 12.6 million less people living in poverty, while extreme poverty reached a record low of 14 percent, with 14.22 million people rising above the food-based subsistence line iii. These 2010 poverty estimates are based on actual data covering nine countries (51 percent of the region’s population), with estimates for the remaining countries.iv Figure 1: Growth and Poverty Reduction Rebounded Strongly in 2010 GDP per capita PPP (constant 2005 international $) 50 11.000 LAC POVERTY AND LABOR BRIEF 45 45,9 10.500 44,8 43,6 43,9 40 42,2 42,9 42,7 Poverty Headcount (%) 41,5 41,3 40,9 10.000 35 38,0 34,0 9.500 30 31,5 30,5 30,3 9.000 25 27,5 27,9 27,7 26,8 26,6 26,7 27,9 25,0 24,5 25,6 8.500 20 24,0 21,8 15 18,5 8.000 17,1 16,3 16,1 10 13,9 7.500 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 Extreme Poverty Moderate Poverty GDP per capita PPP (constant 2005 international $) Source: Author’s calculations using SEDLAC data, 2011 (CEDLAS and the World Bank) and World Bank Open Data. Note: Estimates of poverty at the regional level are population-weighted averages of country estimates. For years where comparable household surveys for a particular country are unavailable, poverty is estimated using the regional elasticity between growth and poverty. For details, please refer to Annex A. The extreme poverty line used is equivalent to $ 2.5/day in 2005 PPP, and the moderate poverty line used is equivalent to $ 4.0/day in 2005 PPP. 7 The strong connection between growth and poverty reduction that had helped to reduce poverty sharp- ly in the mid-2000s resumed between 2009 and 2010 (Figure 2). Fortunately, this growth-poverty link had weakened during the 2008-2009 crisis years, protecting poor households from the contraction and causing pov- erty levels to remain the same despite an almost 3 percentage point drop in GDP per capita. As growth reverted in 2010 to pre-crisis levels, its ability to pull households out of poverty (-1.6 elasticity) became comparable to 2003-2007 (-1.9 average elasticity) - the best years of poverty reduction in the region. The 2010 recovery was also accompanied by a reduction in inequality, which accelerated poverty reduction. The Gini dropped in 6 out of 9 countries for which we have 2009-2010 data (Table B.5). The strength of the economic rebound and its impacts on poverty differed across the region. While eco- nomic growth was widely experienced across LAC in 2010, poverty stagnated in Mexico and Central America revealing a weakening between poverty and growth. Growth resumed in all sub-regions, with the strongest performance in the Southern Cone (a 6.7 percent increase in GDP per capita), followed by Mexico and Central America, and then the Andean Region (Figure 3).v Growth had a strong impact on poverty levels in the Southern Cone and the Andean regions (poverty declined by 3.7 and 1.9 percentage points, respectively, in 2010). In contrast, in Mexico and Central America poverty decreased by 1.3 percentage points, despite a more than 4 percent increase in GDP per capita – suggesting that poor households did not benefit much from the expansion and/or that vulnerable households continued to drop below the poverty line. Rural extreme poverty fell significantly more than urban extreme poverty during the 2009 crisis and 2010 rebound (Figure 5). Food price increases buoyed rural households in 2008 and 2010, suggesting many were net sellers of food products or that wages and/or employment increased as an indirect effect of higher food prices. In addition, safety net programs may have helped buffer the impact of both crises on the rural poor relatively more than on the urban poor as these programs are more targeted to rural areas. Meanwhile, the urban extreme poor were disproportionately affected by the food crisis and the deteriorated labor market opportunities result- ing from the global financial crisis of 2009, as poverty levels rose slightly during the crisis year. It is surprising that the 2010 recovery had a greater impact on the rural poor than the urban poor, suggesting that the rural economy recovered faster and that the labor market in urban areas remains less dynamic for poorer workers. Figure 2: The Strong Link Between Poverty Reduction and Growth Resumed in 2010 6,0 LAC POVERTY AND LABOR BRIEF 4,0 1,7 2,0 0,2 0,0 -1,7 -1,4 -1,0 -1,3 -1,6 -1,3 -1,6 -1,6 % Change -2,0 -2,4 -2,4 -2,1 -2,5 -3,6 -4,0 -6,0 -8,0 -10,0 -12,0 96-95 97-96 98-97 99-98 00-99 00-01 01-02 02-03 03-04 04-05 05-06 06-07 07-08 08-09 09-10 % Change in GDP per capita % Change in Moderate Poverty Ratio Source: Author’s calculations using SEDLAC data, 2011 (CEDLAS and the World Bank). 8 Figure 3: Growth Resumes in all Regions of Latin America 8 6,1 6,6 6 5,5 5,3 4,1 4,4 4,3 % Change in GDP per capita 4 3,6 3,7 2,8 2 2,0 0,2 0 -2 -1,0 -1,6 -4 -6 -6,1 -8 Andean Region Mexico & Central America Cono Sur Extended 05-06 06-07 07-08 08-09 09-10 Source: Author’s calculations using SEDLAC data, 2011 (CEDLAS and the World Bank). Figure 4: Poverty Levels in Central America and Mexico Remain Somewhat Unaffected by the 2010 Economic Rebound 2 Percentage points change in Moderate Poverty 1 0 -1 1,5 0,9 -0,7 -0,8 -0,8 -2 -1,2 -1,5 -3 -1,9 -2,6 -2,5 -4 -3,1 -3,2 -3,4 -3,6 -5 -6 -5,8 -7 Andean Region Mexico & Central America Cono Sur Extended 05-06 06-07 07-08 08-09 09-10 Source: Author’s calculations using SEDLAC data (CEDLAS and the World Bank). Figure 5: Rural households were more resilient to the 2008 and 2009 crises - yet were more closely tied to the 2010 economic recovery 0,5 Poverty Headcount 1,0 Poverty Gap LAC POVERTY AND LABOR BRIEF Percentage points change in Extreme Poverty 0 0,5 -0,5 0,1 0 -1,0 -0,6 -0,5 0,6 0,2 -1,5 -1,2 -1,0 -0,2 -2,0 -1,3 -1,0 -0,5 -0,6 -2,5 -1,8 -1,0 -1,5 -3,0 -1,4 -2,0 -3,5 -3,0 -4,0 -2,5 -4,5 -4,2 -3,0 -2,6 06-07 07-08 08-09 09-10 06-07 07-08 08-09 09-10 06-07 07-08 08-09 09-10 06-07 07-08 08-09 09-10 Rural Urban Rural Urban Source: Author’s calculations using SEDLAC data, 2011 (CEDLAS and the World Bank). Note: Change in poverty rate for a subset of LAC countries with annual data available between 2006 and 2010: Dominican Republic, Ecuador, Paraguay, Peru, and Uruguay. Data for Argentina are available annually but do not included rural households – hence it is excluded. Triangulations of these results for Brazil, Honduras and Costa Rica until 2009 are qualitatively similar. 9 Despite the improvements, and although poverty remains largely an urban phenomenon in LAC in terms of numbers of people, both the severity and intensity of poverty continues to be greater in rural areas; in 2010, the poverty rate and the poverty gap are almost three times as high in rural areas as in urban areas. While high food prices may have been a relatively positive force in rural areas, they imposed economic hardships on the average poor household and kept the regional extreme poverty rate from experiencing a stronger decrease. At the more aggregate national level, the rise in food prices in 2008, and to a lesser extent in 2010, contributed to higher poverty, while some increases in household income and a better distribution of income (linked to the higher food prices) worked in the opposite direction to reduce poverty (Box 1). BOX I: Food prices and changes in poverty Price differentials faced by the poor and non-poor during the 2006-2008 period were partially off-set by a strong redistribution of income. Extending the standard Datt-Ravallion decomposition techniques to include food price changes (through the use of the food CPI) as proposed by Shorrocks and Kolenikov (2001) can shed light on the role played by economic growth, changes in the welfare distribution, or changes in prices differentials on the overall change in poverty observed between 2005-10. In 2005-2006, before the rise in food prices, growth played a substantial role in poverty reduction; yet during the 2006-2007 and 2007-2008 spike in food prices, this role shrank considerably (Figure 6). Had there been no increase in food prices in 2007-2008, all else equal, poverty reduction would have been 1.4 percentage points higher. Figure 6: In 2007 and 2008, higher food prices kept poverty from falling further than it did 2 Percentage points change in Extreme Poverty 1 0 -1 -0,6 -0,7 -1,0 -1,0 -2 -3 -4 -3,4 LAC POVERTY AND LABOR BRIEF 2005-06 2006-07 2007-08 2008-09 2009-10 Distribution Growth Food prices Total Source: Author’s calculations using SEDLAC data, 2011 (CEDLAS and the World Bank). Note: The nine countries are Argentina, Brazil, Costa Rica, the Dominican Republic, Ecuador, Honduras, Paraguay, Peru, and Uruguay. The 2010 data for Brazil, Costa Rica and Honduras are from 2009 but inflated using national accounts to represent 2010 levels. Shorrocks-Kolenikov (2001) decomposition. 10 Figure 7: Extreme poverty trends using food and overall CPI 25 22,8 23,1 Extreme poverty (%) 20 19,8 18,7 17,7 19,3 17,0 18,4 16,0 15 16,2 15,7 13,9 10 2005 2006 2007 2008 2009 2010 Poverty line constructed using Food CPI Poverty line constructed using overall CPI Source: Author’s calculations using SEDLAC data, 2011 (CEDLAS and the World Bank). Note: The nine countries are Argentina, Brazil, Costa Rica, the Dominican Republic, Ecuador, Honduras, Paraguay, Peru, and Uruguay. The 2010 data for Brazil, Costa Rica and Honduras are from 2009 but inflated using national accounts to represent 2010 levels. Given that internationally comparable poverty lines do not fully capture the effect of rising food prices on the poor, the regional and sub-regional trends presented earlier may overstate the welfare gains ex- perienced by poor households since 2007.vi Traditional international measures of poverty are not sensitive to food price volatility as they rely on the overall CPI rather than the food CPI. A poverty line that is sensitive to food price changes reveals the decoupling of poverty trends when using the overall CPI and the food CPI. For a subset of countries for which we have yearly data, we can assess directly how much food prices mattered for the extreme poor. As these changes in food prices may produce winners and losers there are clear implications for distribution. Not surprisingly, higher food prices, such as those experienced in 2007 and 2008, tend to be a drag on poverty reduction. However, using the overall Consumer Price Index (CPI) to adjust both the poverty line and household income over time does not take into account that the poor spend a larger proportion of their income on food and are therefore more sensitive to fluctuations in food prices. The use of a food-specific CPI to adjust the value of the extreme poverty line for 9 countries in the region suggests that extreme poverty may not have decreased as much as anticipated, with an average reduction of 1.4 percentage points per year between 2005- 2010 (Figures 7 and B.1) as opposed to 1.8 percentage points. LAC POVERTY AND LABOR BRIEF 11 12 LAC POVERTY AND LABOR BRIEF II. Understanding the changes in poverty T o better understand what happened to household incomes during the crisis and recovery, this section unpacks the contribution of labor and non-labor income to poverty reduction. An impor- tant finding of this section suggests that labor income has been a critical force for poverty reduction in the region, and that men’s labor income was the most important in pushing households out of poverty in 2010, though during the crisis it was female labor income that continued to move households out of poverty. Going forward, countries have to continue to closely monitor the skills required by labor markets, adopt policies that can help less favored households succeed in the labor market, and explore options to expand female labor mar- ket participation to help diversify household income. Growth and inequality were equally important forces behind poverty reduction between 2009 and 2010 (using data from nine countries for which more recent annual data is available), in contrast to the pre-crisis years when the growth effect dominated over a more equal distribution of income (Figure 8). LAC POVERTY AND LABOR BRIEF Labor market income is the most important driver for poverty reduction for both the poor and extreme poor, accounting for around 55 percent on average of the reduction of poverty in recent years, while non-labor income, which includes public and private transfers, accounted for 20 percent viii (Figure 9). The remaining 25 percent of poverty reduction was accounted for by a joint increase in both labor and non-labor income. During 2009, the contribution of rising labor incomes in reducing poverty shrank, all the while remain- ing the most important driver of poverty reduction. In contrast, the importance of increases in non-labor income grew, underscoring the importance of accessing expanded public and private transfers for poverty reduction during the crisis compared to other years. However, the most traction for reducing poverty during the crisis oc- curred in households benefitting from both rising labor and non-labor incomes. 13 Figure 8: Growth has been the predominant factor contributing to poverty reduction in the recent years 0 Percentage Points in Moderate Poverty -0,5 -1,0 -1,5 -1,4 -2,0 -2,5 -2,4 -3,0 -2,6 -2,6 -3,5 -4,0 -3,7 2005-06 2006-07 2007-08 2008-09 2009-10 Distribution Growth Total Source: Author’s calculations using SEDLAC data, 2011 (CEDLAS and the World Bank). Note: The nine countries are Argentina, Brazil, Costa Rica, the Dominican Republic, Ecuador, Honduras, Paraguay, Peru, and Uruguay. The 2010 data for Brazil, Costa Rica and Honduras are from 2009 but inflated using national accounts to represent 2010 levels. Datt-Ravallion (1992) decomposition.vii Figure 9: Labor income is most important for poverty reduction, but moving out of poverty during the 2009 crisis meant relying jointly on both labor and non-labor income Extreme poverty 0% Share of effects on the total change in -10% -2% -20% -21% -17% -30% -23% -24% -21% Extreme Poverty -26% -26% -26% -40% -34% -50% -40% -60% -51% -50% -58% -70% -80% -90% -82% 2005-06 2006-07 2007-08 2008-09 2009-10 2005-06 2006-07 2007-08 2008-09 2009-10 2005-06 2006-07 2007-08 2008-09 2009-10 Labor income Non Labor income Rank correlation Source: Author’s calculations using SEDLAC data, 2011 (CEDLAS and the World Bank). Note:The nine countries are Argentina, Brazil, Costa Rica, the Dominican Republic, Ecuador, Honduras, Paraguay, Peru, and Uruguay. The 2010 data for Brazil, Costa Rica and Honduras are from 2009 but inflated using national accounts to represent 2010 levels. Fournier (2001) decomposition vii Men’s labor income was the most important in pushing households out of poverty in 2010, similar to the LAC POVERTY AND LABOR BRIEF pre-crisis years (Figure 10). However, it was also the most vulnerable to the crisis, as falling income from men caused poverty to jump (as well as its depth to increase). Meanwhile, female labor income compensated for the fall in male labor income during the 2009 crisis. Most striking, however, was the dramatic importance of having dual gender earnings increase during the 2009 crisis. Households experiencing increases in earnings for both men and women had the greatest chance of existing poverty. Households with young children were vulnerable during the crisis, but recovered quickly during the 2010 rebound. While poor households with children experienced the greatest improvements in well-being during periods of growth, they also suffered the greatest negative impacts during economic contractions (compared to households with older children and adults; Figure 11). This may reflect the fact that these households were less likely to have both males and females in the labor force, given the presence of young children.x This suggests 14 Figure 10: Men in the labor market were more vulnerable to the crisis than women 0,30 Share of effects on the total change in 0,20 0,10 0 Extreme Poverty -0,10 23% -0,20 -9% -16% -18% -18% -0,30 -29% -27% -0,40 -32% -33% -0,50 -0,60 -51% -53% -59% -55% -0,70 -62% -61% 2005-06 2006-07 2007-08 2008-09 2009-10 2005-06 2006-07 2007-08 2008-09 2009-10 2005-06 2006-07 2007-08 2008-09 2009-10 Men Women Rank correlation Source: Author’s calculations using SEDLAC data, 2011 (CEDLAS and the World Bank). Note: The nine countries are Argentina, Brazil, Costa Rica, Ecuador, Honduras, Paraguay, Peru, the Dominican Republic and Uruguay. 2010 data for Brazil, Costa Rica and Honduras are from 2009 but inflated using national accounts to represent 2010 levels. Poverty calculated using only labor income in this exercise. Fournier (2001) decomposition. ix Figure 11: Poverty Fell More Amongst Households With Small Children (0-14 year olds) 0,5 0 -0,5 0,1 -0,1 Percentage points change in -1,0 -0,4 -0,4 -0,7 -1,0 Extreme Poverty -1,5 -1,2 -1,1 -1,2 -1,5 -1,4 -1,3 -2,0 -1,5 -1,6 -1,6 -2,5 -2,3 -2,4 -3,0 -2,9 -3,5 -4,0 -3,4 -4,5 -4,2 0 to 14 15 to 24 25 to 64 65 or more 05-06 06-07 07-08 08-09 09-10 Source: Author’s calculations using SEDLAC data, 2011 (CEDLAS and the World Bank). Note: Data are for subset of countries with data available annually: Argentina, Ecuador, Paraguay, Peru, the Dominican Republic and Uruguay. Descriptive statistics. Households may be in more than one category if members have corresponding ages. that households with children remain highly vulnerable to crises and that the depth and coverage of safety nets LAC POVERTY AND LABOR BRIEF remains inadequate to protect them. These results could reflect the fact that the amount of the safety nets for these families is too small, that some of these poor households fall through the cracks and are not reached by cash transfers, or that they represent the “new poor� who also remain outside most of the formal safety net pro- grams. In addition, the fact that many households with young children may rely on a single income earner who lost earnings during the crisis could contribute to its negative impact. During the 2010 rebound, households in extreme poverty saw their labor income fall relative to average households (Figure 12). Between 2005 and 2010, the level of labor income for the extreme poor increased by 3 percent, while the level of non-labor income increased even more (9 percent). These increases were modest compared to average households in LAC, which experienced the largest increases in labor income during this period (24 percent) and smaller average increases in non-labor income (17 percent). The relative decline in la- 15 Figure 12: The extreme poor are more reliant on non-labor income than the overall population 1,05 1,00 1,00 0,95 0,99 0,99 0,98 0,97 Ratio 0,90 0,94 0,92 0,91 0,91 0,91 0,90 0,85 0,84 0,80 2005 2006 2007 2008 2009 2010 Share of Labor Income for Extreme Poor/ Share of Labor Income for Moderate Poor/ Share of Labor for Overall Population Share of Labor for Overall Population Source: Author’s calculations using SEDLAC data, 2011 (CEDLAS and the World Bank). Note: Data are for a subset of six countries that have anual data available between 2005 and 2010: Argentina, the Dominican Republic, Ecuador, Paraguay, Peru, and Uruguay. bor income of the extreme poor was particularly sharp in 2010. These trends suggest that the ability of labor markets to lift households out of poverty may be weakening since 2005 and particularly in 2010 and/or that extremely poor households are increasingly dependent on public and private transfers for their income. The increased reliance on private and public transfers for the extreme poor may be difficult to sustain should the global and regional economy dip downward in the coming year. Going forward, countries should monitor labor markets and wages closely, especially for the extreme poor, poor and vulnerable, taking advantage of the rela- tively high-frequency labor market data to track the impact of economic trends on their wellbeing. LAC POVERTY AND LABOR BRIEF 16 III. Going forward P overty reduction in LAC will likely continue its downward trend in 2011, but at a slower pace, even in the most optimistic scenarioxi. Depending on the responsiveness of household welfare to growth, moderate and extreme poverty in Latin America will experience anywhere between a 0.2 and 1.4 percentage point decline (Figure 13). Both these results are much less than the 2.1 and 2.4 percentage point poverty decline the region experienced during the 2010 growth rebound. Quarterly labor market trends and the latest remittance data confirm that poverty reduction is likely to continue through 2011. The Labor Income Poverty Index (LIPI) measures poverty in terms of individuals who cannot obtain the basic food basket with their labor income, and acts as a leading indicator of povertyxii. Labor markets in several representative LAC economies saw poverty fall, as the LIPI index continued to decline in Brazil, Peru and Ecuador through September 2011, despite concerns over the summer of a double dip recession (Figure 14). Mexico has been the exception as its labor markets continued to face growing difficulties in provid- ing income opportunities for the poor. The share of individuals earning below the poverty line has been rising LAC POVERTY AND LABOR BRIEF gradually since the 2009 crisis, despite the economy’s relatively strong growth. The resumption of growth in remittances in 2011 is another positive trend contributing to poverty reduction in the region. According to the World Bank’s Migration and Development Brief, remittance flows to Latin America and the Caribbean region have resumed in 2011 after a sharp decline during the global financial crisis and after remaining almost flat in 2010. Available data until the third quarter of 2011 for Mexico, Colombia, El Salvador, Guatemala, Honduras, Jamaica, and Nicaragua - which together account for three-quarters of remittance flows to the Latin America and Caribbean region - show that remittance inflows grew by nearly 7 percent in the first three-quarters. Remittances to Mexico surged by 11 percent in the third quarter, in part because of the deprecia- tion of the Mexican Peso relative to the US dollar.xiii 17 Figure 13: Poverty reduction in the future may be slower 19.0 Extreme poverty scenarios 36.0 Moderate poverty scenarios 18.0 18.3 34.0 17.2 34.2 31.7 17.0 30.8 Poverty Headcount (%) 16.4 32.0 30.6 16.1 16.0 27.8 30.0 15.0 27.6 14.0 28.0 26.4 14.0 13.8 13.2 26.0 13.0 12.0 24.0 2006 2007 2008 2009 2010 2011 2006 2007 2008 2009 2010 2011 Observed High elasticity Low elasticity Source: Author’s calculations using SEDLAC data (CEDLAS and World Bank) and the regional projections for 2011 are computed using a GDP growth projection for the LAC region of 4.2 percent (World Bank, DECPG, December 2011). Figure 14: Labor market data suggest the extreme poor may face a more substantial economic slow-down 140 130 120 119 LIPI (Extreme Poverty) 110 100 90 85 80 78 70 60 58 50 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 2005 2006 2007 2008 2009 2010 2011 Argentina - Urban Peru - Urban Lima Uruguay - National Brazil - Urban Colombia - National Mexico - National Ecuador - National Source: author’s calculation using LABLAC Data, 2011 (CEDLAS and the World Bank). Note: 3 months moving average. Base: Q4, 2007 = 100. As 2011 comes to a close, there are some obvious economic uncertainties and challenges ahead for con- tinued poverty reduction in Latin America. At present, the projected economic slowdown does not loom as LAC POVERTY AND LABOR BRIEF severe as the 2009 Great Recession, nonetheless it could have important implications for poverty reduction in Latin America in 2012 and beyond. In terms of poverty levels, the key uncertainties are how labor markets will respond and transmit any crisis to poor households, whether or not governments can increase social spending to compensate for any declines in labor income, what will happen to remittances and other private transfers and finally trends in food prices. Large fiscal deficits in LAC suggest that it could be difficult for governments to expand social spending should there be another crisis, in contrast to the situation of 2009. On average, government deficits in LAC as a percent of GDP have increased since 2007 (Figure 15).xiv While countries may have had the space to expand social spending early in the previous crisis – the deficit was less than one percent of GDP in 2008 – the estimated average fiscal deficit for 2011 has more than doubled compared to 2008. 18 Figure 15: The region’s fiscal deficit (% of GDP) has increased 5 4,0 Government de cit (%GDP) 4 3,0 3 2,2 2 1,1 0,9 1 0 Years 2007 2008 2009 2010 2011 (Projected) Source: WEO, September 2011. Food prices remain another factor of uncertainty, although food price volatility has been lower in LAC than in other regions of the world. However, there is general consensus that high and volatile global prices will continue in the medium term due to structural factors, such as a rapidly growing global population and increas- ing linkages between food and energy prices, particularly with the production of biofuels.xv Finally, the experience of the last crisis and the recovery also suggests that aggregate trends mask some vulnerable countries and groups of households. First, countries hard hit by the last crisis have not fully re- covered and may already be affected by the current slowdown, as suggested by the LIPI data for Mexico. In these countries, households were more exposed to the crisis and have faced challenges in benefitting from the recovery and could be particularly vulnerable to further income losses. Several Central American and Caribbean countries would also fall into this category. Second, urban households were particularly hard hit by the last crisis compared to rural households, who tended to have greater access to safety nets. Poverty increased in urban areas across the region in 2009 and the recovery had a smaller impact on the welfare of the poor in urban areas. Looking forward, governments may wish to expand programs to the urban working poor who were vulnerable to losing their livelihoods in the last crisis. Third, in 2009, households with young children was the only demo- graphic group to experience an increase in poverty, which could reflect their dependence on male labor income which suffered important losses with the crisis. While these households also experienced a strong recovery in 2010, it is important to consider how they could be better protected in a future crisis. LAC POVERTY AND LABOR BRIEF 19 Annex A. Why past poverty estimates may change? Poverty rates are computed using a measure of welfare, a price index and a poverty line. In LAC, as well as in many other regions, internationally comparable poverty series are computed by comparing household income in purchasing power parity (PPP) terms with the international poverty lines of 2.5 and 4 dollars a day. Changes in any of the inputs for the poverty estimates of a given year will have implications for the regional estimate. Revisions to population, due to periodic census updates, as well income aggregate estimates, or revi- sions and updates of CPI series may cause the resulting national, and subsequent regional poverty estimates to change. Moreover, comparable surveys are seldom available for all countries every year. In order to produce a regional poverty number over time, an estimate of poverty for the missing countries must first be computed. The re- gional estimates reported in this brief are computed using the observed elasticity between GDP per capita and poverty rates. For this exercise, two periods are recognized in the series: 1995 to 2002 (relatively stagnant) and 2003 to 2010 (important reduction in poverty). As a result, two elasticities are computed -- one for each segment (Table A.1). These elasticities are estimated at the Regional level. The periodic changes and revisions in the na- tional account system can also have, and will also have an implication on the estimated elasticities and imputed poverty rates. The GDP per capita data used to compute the regional estimates come from the World Bank’s Open Data, which in turn relies on population estimates from the UN and GDP estimates from the World Bank’s World Develop- ment Indicators. Consumer price indices come from the ILO and the National Statistical Offices of each country that provides household survey data. The International Comparison Program seeks to provide up-to-date price comparisons through PPP factors. Table A.1: Elasticity Poverty-Growth LAC POVERTY AND LABOR BRIEF Elasticity Poverty-Growth 1995-2002 2003-2010 Region Countries $2.5 USD a day $4 USD a day $2.5 USD a day $4 USD a day LAC All 16 countries -3.04 -2.19 -1.98 -1.63 Andean Region Bolivia, Colombia, Ecuador, Peru, -1.19 -0.82 -0.84 -0.71 Venezuela Mexico & Central Costa Rica, Domincan Republic, El 1.14 0.79 -0.91 -0.66 America Salvador, Honduras, Mexico, Panama Cono Sur Extended Argentina, Brazil, Chile, Paraguay, -3.33 -2.37 -3.12 -2.49 Uruguay Source: Author’s calculations using SEDLAC data, 2011 (CEDLAS and the World Bank). 20 B. Tables Table B.1: Datasets used by country Sources and relevant information for Socio-Economic Database for Latin America and the Caribbean (SEDLAC) Country Name of survey Acronym Last Year Coverage Argentina Encuesta Permanente de Hogares-Continua EPH-C 2010 Urban-31 cities Bolivia Encuesta Continua de Hogares- MECOVI ECH 2008 National Brazil Pesquisa Nacional por Amostra de Domicilios PNAD 2009 National Chile Encuesta de Caracterización Socioeconómica Nacional CASEN 2009 National Colombia Gran Encuesta Integrada de Hogares GEIH 2010 National Costa Rica Encuesta de Hogares de Propósitos Múltiples EHPM 2009 National Dominican Republic Encuesta Nacional de Fuerza de Trabajo ENFT 2010 National Ecuador Encuesta de Empleo, Desempleo y Subempleo ENEMDU 2010 National El Salvador Encuesta de Hogares de Propósitos Múltiples EHPM 2009 National Honduras Encuesta Permanente de Hogares de Propósitos EPHPM 2009 National Múltiples Mexico Encuesta Nacional de Ingresos y Gastos de los Hogares ENIGH 2010 National Panama Encuesta de Hogares EH 2009 National Paraguay Encuesta Permanente de Hogares EPH 2010 National Peru Encuesta Nacional de Hogares ENAHO 2010 National Uruguay Encuesta Continua de Hogares ECH 2010 National Venezuela, RB Encuesta de Hogares Por Muestreo EHM 2006 National LAC POVERTY AND LABOR BRIEF Sources and relevant information for Labor database for LAC (LABLAC) Country Name of survey Acronym Last Year Coverage Argentina Encuesta Permanente de Hogares-Continua EPH-C December 2010 Urban-31 cities Brazil Pesquisa Mensual de Emprego PME June 2011 Urban-6 cities Colombia Gran Encuesta Integrada de Hogares GEIH December 2010 National Ecuador Encuesta de Empleo, Desempleo y Subempleo ENEMDU June 2011 National Mexico Encuesta Nacional de Ocupacion y Empleo ENOE June 2011 National Peru Encuesta Permanente de Empleo EPE June 2011 Lima-Urban Uruguay Encuesta Continua de Hogares ECH December 2010 National 21 Table B.2: Regional poverty numbers GDP growth (%) Extreme poverty rate Moderate poverty Change Extreme Change Moderate rate poverty rate poverty rate Region 2008- 2009- 2008 2009 2010 2008 2009 2010 2008- 2009- 2008- 2009- 2009 2010 2009 2010 2009 2010 LAC -2,8 5,1 16,4 16,1 14,0 30,6 30,4 28,0 -0,3 -2,1 -0,2 -2,4 Andean Region -1,6 2,0 23,1 22,3 20,5 40,3 39,5 37,6 -0,8 -1,7 -0,8 -1,9 Mexico & Central -6,1 4,4 15,3 15,8 14,0 30,4 31,3 30,0 0,5 -1,7 0,9 -1,3 America Cono Sur Extended -1,1 6,7 13,8 13,3 10,7 26,3 24,8 21,1 -0,5 -2,6 -1,5 -3,7 Source: Author’s calculations using SEDLAC data, 2011 (CEDLAS and the World Bank). Table B.3: Extreme and Moderate Poverty by country and year Country Extreme poverty headcount Moderate poverty headcount ($2.5 USD a day - Actual data) ($4 USD a day - Actual data) 2005 2006 2007 2008 2009 2010 2005 2006 2007 2008 2009 2010 Argentina 13.4 10.3 9.0 8.3 8.1 6.4 26.0 20.6 19.7 17.3 16.4 14.6 Bolivia 35.0 33.5 33.5 31.1 53.9 50.7 50.7 48.1 Brazil 22.9 19.7 18.2 15.7 15.1 38.4 34.9 31.9 29.2 27.6 Chile 5.3 4.3 15.8 11.8 Colombia 31.7 27.2 24.8 22.0 52.8 44.9 42.6 39.5 Costa Rica 12.2 11.6 8.2 7.6 8.1 25.4 25.2 20.1 18.9 19.6 Dominican 21.1 18.7 17.9 18.4 16.4 16.1 40.5 37.5 36.4 37.9 34.7 35.1 Republic Ecuador 25.6 20.0 20.2 19.6 19.4 15.9 43.6 38.3 38.6 37.1 37.6 33.4 El Salvador 27.1 19.9 18.8 21.1 23.1 45.3 39.8 37.9 41.8 42.7 LAC POVERTY AND LABOR BRIEF Honduras 47.6 42.2 37.0 37.6 36.2 64.3 58.8 56.0 52.8 52.1 Mexico 15.1 11.9 13.4 12.1 29.7 26.7 28.2 27.9 Panama 22.5 22.2 16.1 37.5 37.1 29.9 Paraguay 21.9 26.2 21.1 19.2 20.6 18.4 39.9 43.6 39.0 36.3 34.9 32.8 Peru 28.8 25.1 24.4 20.4 20.0 18.3 47.7 43.4 40.7 36.9 35.7 33.6 Uruguay 8.9 7.2 6.1 4.1 3.4 2.8 21.6 20.7 18.7 13.9 12.0 11.0 Venezuela, 30.3 19.8 50.3 37.9 RB Source: Author’s calculations using SEDLAC data, 2011 (CEDLAS and the World Bank). 22 Table B.4: GDP per capita, PPP (constant 2005 international $) Country 2005 2006 2007 2008 2009 2010 Growth Growth 2008-09 2009-10 Argentina 10,833 11,647 12,545 13,276 13,272 14,363 0.0% 8.2% Bolivia 3,772 3,885 3,995 4,172 4,244 4,353 1.7% 2.6% Brazil 8,509 8,753 9,196 9,583 9,438 10,056 -1.5% 6.6% Chile 12,168 12,598 13,047 13,394 13,045 13,596 -2.6% 4.2% Colombia 7,305 7,677 8,085 8,250 8,251 8,488 0.0% 2.9% Costa Rica 9,042 9,673 10,261 10,367 10,059 10,259 -3.0% 2.0% Dominican 6,380 6,961 7,445 7,728 7,887 8,387 2.1% 6.3% Republic Ecuador 6,553 6,829 6,862 7,251 7,172 7,325 -1.1% 2.1% El Salvador 5,694 5,909 6,155 6,275 6,022 6,048 -4.0% 0.4% Honduras 3,277 3,423 3,567 3,636 3,496 3,516 -3.8% 0.6% Mexico 12,191 12,658 12,905 12,932 11,994 12,498 -7.3% 4.2% Panama 9,167 9,778 10,779 11,675 11,856 12,541 1.5% 5.8% Paraguay 3,901 3,994 4,187 4,352 4,111 4,656 -5.5% 13.3% Peru 6,387 6,805 7,333 7,967 7,952 8,558 -0.2% 7.6% Uruguay 9,683 10,075 10,783 11,675 11,937 12,903 2.2% 8.1% Venezuela, RB 9,924 10,721 11,404 11,756 11,190 10,806 -4.8% -3.4% Source: World Bank, World Development Indicator, 2011 (serie ny.gdp.pcap.pp.kd) LAC POVERTY AND LABOR BRIEF 23 Table B.5: Gini by Country and Year Gini (from per capita household income) Country 2005 2006 2007 2008 2009 2010 Argentina 0.488 0.475 0.469 0.459 0.449 0.442 Bolivia 0.576 0.561 0.572 0.556 Brazil 0.564 0.559 0.548 0.542 0.537 Chile 0.518 0.519 Colombia 0.552 0.564 0.560 0.554 Costa Rica 0.472 0.487 0.489 0.484 0.502 Dominican 0.499 0.519 0.487 0.490 0.489 0.472 Republic Ecuador 0.536 0.529 0.539 0.502 0.489 0.489 El Salvador 0.497 0.461 0.468 0.466 0.481 Honduras 0.594 0.574 0.560 0.593 0.553 Mexico 0.509 0.496 0.502 0.473 Panama 0.538 0.549 0.521 0.519 Paraguay 0.528 0.545 0.542 0.521 0.507 0.522 Peru 0.517 0.510 0.517 0.489 0.491 0.481 Uruguay 0.459 0.472 0.476 0.463 0.463 0.453 Venezuela, RB 0.476 0.435 Source: Author’s calculations using SEDLAC data, 2011 (CEDLAS and the World Bank). Note: Observations with zero income were not considered. Table B.6: Poverty Projections LAC POVERTY AND LABOR BRIEF 2010 2011 Change (2010-2011) Low Elasticity High Elasticity Low Elasticity High Elasticity $2.5 USD a day Headcount (%) 14,0 13,8 13,2 -0,2 -0,8 Number of poor (millons of people) 74,8 74,8 71,6 -0,1 -3,3 $4 USD a day Headcount (%) 27,8 27,6 26,4 -0,2 -1,4 Number of poor (millons of people) 149,0 149,7 143,3 0,7 -5,7 Source: Author’s calculations with data from SEDLAC (CEDLAS and the World Bank) using population-weighted averages. 24 Figure B.1: Moderate and extreme poverty trends using the food CPI Moderate Poverty Extreme Poverty 40 39,0 Headcount - Poverty Line $4USD day 25 Headcount - Poverty Line $2.5USD day 38,7 23,1 35,3 35 33,9 33,7 20 22,8 19,8 Poverty Rate (%) 32,6 18,7 32,7 30,7 30 30,1 29,6 19,3 17,7 17,0 28,7 18,4 27,6 15 16,0 16,2 15,7 25 13,9 20 10 2005 2006 2007 2008 2009 2010 2005 2006 2007 2008 2009 2010 Poverty line constructed using Food CPI Poverty line constructed using Overall CPI Poverty Gap Poverty Gap 20 Poverty Gap - Poverty Line $4USD a day 12 Poverty Gap - Poverty Line $2.5USD a day 17,9 9,9 18 10 17,7 9,8 8,5 8,1 8 7,5 7,4 Poverty Rate (%) 16 8,2 8,0 6,9 14,9 6 6,9 6,8 15,7 14,2 6,0 14 14,7 13,7 4 13,0 12,9 12 12,7 2 11,4 10 0 2005 2006 2007 2008 2009 2010 2005 2006 2007 2008 2009 2010 Poverty line constructed using Food CPI Poverty line constructed using Overall CPI Source: Author’s calculations using SEDLAC data, 2011 (CEDLAS and the World Bank). Note: The nine countries are Argentina, Brazil, Costa Rica, the Dominican Republic, Ecuador, Honduras, Paraguay, Peru, and Uruguay. The 2010 data for Brazil, Costa Rica and Honduras are from 2009 but inflated using national accounts to represent 2010 levels. LAC POVERTY AND LABOR BRIEF 25 Bibliography Datt, Gaurav and Martin Ravallion, 1992. “Growth and redistribution components of changes in poverty measures: A decomposition with applications to Brazil and India in the 1980s,� Journal of Development Economics, Elsevier, vol. 38(2), pages 275-295, April. Fournier, Martin, 2001, Inequality decomposition by factor component : a “rank-correlation� approach illus- trated on the Taiwanese case, Recherches économiques de Louvain, 2001/4 Vol. 67, p. 381-403. DOI : 10.3917/ rel.674.0381 Ravallion, Martin and Gaurav Datt, 1996. How Important to India’s Poor is the Sectoral Composition of Economic Growth? The World Bank Economic Review, Vol. 10, No. 1, 1-25. Shorrocks, Anthony and Stanislav Kolenikov (2001), Poverty trends in Russia, Unpublished manuscript. World Bank, 2010a. From Global Collapse to Recovery: Economic Adjustment and Growth Prospects in Latin America and the Caribbean. Presentation, April 21, 2010. Washington, DC: Office of the Chief Economist, LAC. World Bank, 2010b. From Global Collapse to Recovery: Economic Adjustment and Growth Prospects in Latin America and the Caribbean. Washington, DC: Office of the Chief Economist, LAC. World Bank, 2010c. Did Latin America Learn to Shield its Poor from Economic Shocks? Washington, DC: LCSPP Poverty and Labor Brief, No. 1. World Bank, 2011a. A Break with History: Fifteen Years of Inequality Reduction in Latin America. Washington, DC: LCSPP Poverty and Labor Brief, No. 2 World Bank, 2011b. LAC’s decade: Ending or Beginning. Presentation, September 13, 2011. Washington, DC: Office of the Chief Economist, LAC. World Bank, 2011c. LAC Success Put to the Test. Presentation, September 1, 2011. Washington, DC: Office of the Chief Economist, LAC. World Bank, 2011d. Work and Family: Latin America and Caribbean Women in Search of a New Balance. Washing- ton, DC: Office of the Chief Economist (LAC) and LCSPP. World Bank, 2011e. World Bank Food Price Watch. November. LAC POVERTY AND LABOR BRIEF 26 End Notes i. The PPP conversion factors used in this brief are produced by the 2005 round of the International Comparison Program (ICP). The ICP is a worldwide statistical partnership to collect comparative price data and compile detailed expenditure values of countries’ GDP, and to estimate purchasing power parities (PPPs) of the world’s economies. ii. See: “Did Latin America Learn to Shield its Poor From Economic Shocks,� World Bank, October 2010. iii. Population estimation for the LAC region of 589,000,000 for 2010 (World Bank, WDI, electronic version, series sp.pop.totl). vi. Actual 2010 data are available for: Argentina, Colombia, the Dominican Republic, Ecuador, Mexico, Panama, Paraguay, Peru and Uruguay. Brazil does not release a household survey in 2010 since this is the year of the population census. Estimates for the rest of LAC are based on projections (see Annex A for more details). Har- monized comparable data from the Caribbean is lacking, aside from the Dominican Republic. The estimates of extreme and moderate poverty are based, respectively, on 2.5US$-a-day and 4US$-a-day poverty lines in PPP dollars. The extreme poverty line represents the income required to meet basic food needs, while the mod- erate poverty line covers other expenses for basic subsistence (clothing, shelter, etc.). All poverty estimates presented and discussed in this brief follow the tradition of LAC countries to measure poverty by income and not consumption. As a consequence, any consumption smoothing mechanisms which could partially explain the poverty dynamics in the period, such as access to credit markets and insurance mechanisms, do not apply to the analysis of this indicator. v. The poverty trends here presented for Mexico differ from the official trends due to the use of the international $-a-day poverty methodology so as to assure international comparability. While the official numbers suggest an increase in poverty in the 2008-2010 period, the international poverty numbers suggest a very small de- cline in poverty for Mexico. This discrepancy has also been noted in other countries, more recently Chile 2006- 2009, and is due to the limitation of the international poverty measurement methodologies, which do not include spatial price differentials (within countries) nor adjust the food poverty line exclusively by the food CPI as done in a few countries in the region. LAC is divided into the three regions: the Andean Region, the Cono Sur Region, and the Central America, Mexico, and Dominican Republic Region. The Dominican Republic (DR) is the only country from the Caribbean for which we have data. Thus we do not present results for the Caribbean sub-region and include the DR along with Central America and Mexico. vi. Traditional international measures of poverty are not sensitive to food price volatility as they rely on the overall CPI rather than the food CPI. A poverty line that is sensitive to food price changes reveals the decoupling of poverty trends when using the overall CPI and the food CPI. For a subset of countries for which we have yearly LAC POVERTY AND LABOR BRIEF data, we can assess directly how much food prices mattered for the extreme poor. As these changes in food prices may produce winners and losers there are clear implications for distribution. vii. Datt-Ravallion (1992) decomposition of changes in poverty that are due to changes in income growth or due to changes in the distribution of income. viii. As proposed by Fournier, 2001. Non-labor income includes pensions, transfers and other sources of non-labor income. Labor income, include wages from both salaried and self-employed workers from all sectors. ix. Fournier (2001) decomposition of poverty by factor components: changes in poverty due to changes in la- bor income, changes in non-labor income, and changes in the rank correlation between the income sources. Since there is no way to consider a variation of the marginal distribution of a specific income source, keeping constant both the marginal distribution of other sources and the correlation between sources, Fournier (2001) uses an alternative approach that looks at rank correlation. In this approach, one can do simulations that are based on totally non-parametric computations since they only use the rank structure of various income 27 sources. In addition, this method presents the major advantage of allowing for a distinction between changes in the correlation between income sources and changes in the marginal distribution of income sources, with- out using an ad hoc sharing rule for the correlation effect between sources. x. World Bank Regional Gender Study, 2011: “Work and Family: Latin America and Caribbean Women in Search of a New Balance.� xi. The regional projections for 2011 are computed using a GDP growth projection for the LAC region of 4.2 percent (World Bank, DECPG, December 2011) and the response of poverty to that growth using two alterna- tive regional elasticities. In the Best Scenario, poverty is projected using an elasticity of -2 that reflects a high response of the poverty estimate to growth (and corresponds to the average elasticity estimated for the years of best poverty reduction, 2003-2007). On the other hand, in the Worst Scenario poverty is projected using a low elasticity of -0.5 (corresponding to the average elasticity estimated during the 2007-2009 crisis years). The same elasticities are used for both Moderate and Extreme poverty. xii. For a more detailed explanation of the LIPI methodology please see World Bank, 2010b. Did Latin America Learn to Shield its Poor from Economic Shocks? Washington, DC: LCSPP Poverty and Labor Brief, No. 1. xiii. See World Bank. “Migration and Development Brief, No. 17, December 1, 2011. http://siteresources.world- bank.org/INTPROSPECTS/Resources/334934-1110315015165/MigrationandDevelopmentBrief17.pdf xiv. For a deeper analysis of fiscal issues see: World Bank, 2011c. LAC Success Put to the Test. Presentation, Sep- tember 1, 2011. Washington, DC: Office of the Chief Economist, LAC xv. For a deeper analysis of please see World Bank Food Price Watch. November, 2011. LAC POVERTY AND LABOR BRIEF 28 The World Bank 1818 H Street, NW, Washington, DC 20433, USA. www.worldbank.org