THE WORLD BANK Inequality in Focus The Challenge of High Inequality in China Terry Sicular I Figure 1 Income Inequality in China, 1981–2012 n March 2013, before leaving office, China’s then-premier Wen Jiabao delivered his final assessment of the state of Chinese society to the National People’s Congress. In it, he gave his administration mixed marks on their pledge to reduce income inequality. Wen warned that Chinese development was “unbalanced” and that China “still face(s) many difficulties and problems in (its) economic and social develop- ment” (New York Times 2013). China recently completed a decade under the leadership of former president Hu Jintao and Wen, the former premier. During the period from 2003 to 2013, China pursued a “harmo- nious society” policy agenda that emphasized equitable growth. China implemented a wide Sources: Gini coefficients for the years 1986–2001 are from Ravallion and Chen (2007), range of policy measures designed to reduce 2002 from Gustafsson et al. (2008), 2003–2012 from the National Bureau of Statistics. disparities and to protect the economically vul- nerable. These measures included agricultural support policies, then it has declined slightly, but remains well above 0.45.1 social welfare transfers, targeted tax reductions, minimum wage With a Gini approaching 0.5, China’s level of income inequal- increases, and increased spending on poverty alleviation. ity is in the same ballpark as that of relatively high-inequality Despite these policies, income inequality in China has in recent Latin American countries such as Mexico (0.51), Nicaragua years remained stubbornly high. As shown in Figure 1, China’s (0.52), and Peru (0.48), although still lower than Brazil and Gini coefficient (a measure of inequality ranging from zero, which Honduras (0.56–0.57).2 China is now among the least equal 25 represents perfect equality, to one, perfect inequality) rose from percent of countries worldwide.3 Very few Asian countries belong about 0.3 in the early 1980s to more than 0.45 in the early 2000s. to this group. After 2000, the Gini rose further to a high of 0.49 in 2008. Since What, then, explains inequality in China? Recent research by participants in the China Household Income Project (CHIP) The author is professor of economics, University of Western provides some answers. The CHIP is an international collaborative Ontario. This article draws on findings from the China House- survey research project that began in the late 1980s with the aim hold Income Project (CHIP), an international collaborative of tracking changes in incomes, poverty, and inequality in China. survey research project. Fuller analysis of income inequality and The CHIP organized a series of nationwide household surveys that poverty based on the CHIP survey data can be found in Rising collect detailed information on incomes and related variables. Al- Inequality in China: Challenges to a Harmonious Society, edited though the most recent survey was for 2007, it provides the most by Shi Li, Hiroshi Sato, and Terry Sicular. up-to-date, nationwide micro-level data. In a new book by Li, Sato, Poverty Reduction and Equity Department : : www.worldbank.org/poverty : : Volume 2, Number 2 : : August 2013 and Sicular (2013), these data have been Figure 2 Per capita Household Income by Decile, 2002 and 2007 used for an in-depth analysis of the pat- tern and structure of income inequality. In this article, which draws on findings from this book, I focus on three key sources of inequality in China: 1. Inequality in China is not the result of stagnant or declining incomes among poorer groups, but of more rapid growth in incomes of richer groups. 2. Inequality in China is strongly linked to urban-rural differences. 3. Income from private property is a newly emerging and potentially long- term source of inequality. Incomes of the Rich Are Rising Faster than Incomes of the Poor China’s persistently high inequality does not reflect a deterioration of living standards for poorer groups. Between 2002 and 2007, years for which we have CHIP data, growth of per capita household income of poorer deciles in Source: Calculations based on the CHIP data. the income distribution was substantial (figure 2).4 During this five-year period, income increased by per year (figure 3). In comparison, between 2000 and 2010, nearly 50 percent for the poorest decile, and by nearly 60 percent China’s Gini coefficient increased by nearly 10 percent. This for the second-poorest decile. Rising incomes for these low- increase was associated with robust income growth for the rich income groups contributed to a marked drop in poverty. Using a combined with slower, although by no means slow, growth in PPP (purchasing power parity) $1.25-per-day poverty line, Li and incomes for the poor. From 2002 to 2007, the per capita income colleagues (2013) estimate that China’s poverty rate fell from 19 of households in the top two deciles grew by a remarkable 14 percent in 2002 to 8 percent in 2007. percent per year. The bottom two deciles grew by a respectable 8 Despite the strong income growth for poorer groups, income to 10 percent per year (figure 3). Thus, both China’s poorest and inequality still increased because the incomes of richer groups grew richest households experienced faster income growth than their even faster. Between 2002 and 2007, the richest two deciles of the counterparts in Brazil. One could argue, then, that even though income distribution saw their income nearly double (figure 2). As a the pattern of income growth in China increased inequality, it was consequence, the income gap between the richest and poorest de- preferable to that in Brazil. ciles widened from 19:1 to 25:1, and national inequality increased. High and rising inequality in China has occurred at a time The Income Gap between Urban when inequality in some other parts of the world, notably Latin and Rural Households Is Large America, has been declining. The Brazilian experience provides China’s urban-rural income gap has widened since the early an interesting contrast with China. According to Lustig et al. 1980s (figure 4). By 2002, per capita incomes for urban house- (2012), between 1998 and 2009, Brazil’s Gini coefficient declined holds were, on average, more than three times higher than those by 5 percent. This decline reflected robust growth in the incomes for rural households. Since that time, the urban-rural income of the poor combined with slow growth in incomes for the rich. ratio (measured as the average income per capita of urban house- From 2001 to 2009, per capita income of Brazilian households in holds divided by the average income per capita of rural house- the bottom two deciles increased by 5 to 6 percent per year, while holds) has remained well above 3.0. This urban-rural income that of households in the top two deciles increased by 1 percent differential is very high by international standards (Knight and 2 : : August 2013 : : Inequality in Focus Figure 3 Growth Incidence Curves for Household levels of human capital (Gustafsson et al. per capita Income, China and Brazil 2008; Knight et al. 2013; Sicular et al. 2007). Differences in educational op- portunities and outcomes between urban and rural areas persist. Labor earnings, however, are only part of the story. The urban-rural income gap also reflects dif- ferences in non-wage income. Urban non- employment incomes—including pen- sions, government transfers, and returns on private assets such as interest income and imputed rents on owner-occupied housing—have increased rapidly in recent years. By 2007, on average, non-employ- ment income accounted for about 40 percent of urban income as compared to only 15 percent of rural income. Many of Note: From 2002 to 2007 for China and from 2001 to 2009 for Brazil; in constant prices. these unearned components of income are Source: Figure 2; Lustig et al. (2012). associated with government policies that Song 1999, p. 138; see also World Bank 2009b). have disproportionately benefited the registered urban popula- China’s urban-rural income gap has been a central factor tion, such as the privatization of urban housing, the pension pro- underlying national income inequality in recent years. The gap’s gram, and a range of other subsidies and social welfare programs contribution to overall inequality was 45 percent in 2002, and 51 (World Bank 2009a). percent in 2007 (Li et al. 2013). China’s urban-rural income gap has a distinct regional dimen- Urban-rural differences arise in part from China’s household sion, though we are unsure as to why. The urban-rural income registration or hukou system, established during the Maoist ratio is largest in western and eastern China (in 2007, 3.85 and era. The hukou system is an internal passport system that was 3.44, respectively). Between 2002 and 2007, excluding large initially adopted in the late 1950s to control domestic popula- municipalities such as Beijing and Shanghai, this ratio rose by tion movements, especially from rural to urban areas. For many a remarkable 43 percent in the east, as compared to 27 percent years, individuals who wished to move their place of residence in the central part of the country, and only 3 percent in the west were required to apply to the relevant bureaucracies for permis- (Li et al. 2013). The reason for these regional differences merits sion, and approvals were tightly controlled. Since the mid-1990s, further investigation. the hukou system has undergone a series of reforms that have led to a reduction in constraints on geographic mobility and the The Creation of Private Property rapid increase of rural-urban migration. In fact, earnings from Is a New Source of Inequality migrant work have become an important source of income in During the Maoist era, private property was prohibited. In the rural areas, contributing to rural income growth and moderating 1980s, tentative steps were taken to allow private ownership of the urban-rural income gap. Nevertheless, barriers to permanent some forms of property such as livestock, vehicles, and equip- relocation—such as employment discrimination, high housing ment used for family businesses. In the 1990s, new policies costs, and low access to public services like education and health opened the way for private ownership of a wider range of assets care—continue to affect rural migrants. Consequently, most rural- and on a larger scale. This included the private ownership of urban migration has been temporary or short-term. urban housing, which began in the 1990s and was basically In recent years, the income gap has not been the result of completed by 2005. Other complementary policy measures stagnant rural incomes, but rather of faster—indeed very rapid— promoted the development of urban real estate markets for both growth in urban incomes. Between 2002 and 2007, rural incomes residential and commercial property. It was during this time that grew at an average annual rate of more than 7 percent, but urban the government began to foster the development of domestic incomes grew even more rapidly at 11 percent (in constant prices). financial markets and stock markets, and allowed the emergence Part of the gap is explained by differences in rural-urban wages of private businesses. These changes created new mechanisms and labor earnings. Important in this regard are differences in for household saving and wealth accumulation, and Chinese Inequality in Focus : : August 2013 : : 3 Figure 4 China’s Urban-Rural Income Ratio, 1985–2012 involve more difficult political choices. China currently has a fairly extensive set of social and poverty-alleviation programs in place, many of which were implemented as part of the “harmonious society” agenda. Yet, analysis by the World Bank and others has identified room for improvement in the implementation, targeting, and design of these programs (World Bank 2009a, Li et al. 2013). Such improvements could provide ways for China to maintain and improve distributional outcomes even as macroeconomic growth slows. Redistributive policies, however, can only do so much in reducing inequality if there are still forces that exacerbate or create new sources of inequal- ity. China’s urban-rural income differential and the inequality associated with asset income are two such forces. China’s urban-rural income differential reflects the long-term consequences of past and Source: National Bureau of Statistics, 2012. current institutions and policies that have created unequal access between urban and rural residents households were rapidly transformed into property owners. In to education, employment, assets, and social welfare benefits. Chi- the mid-1990s, few Chinese households owned their homes; only na recently announced a new policy of accelerated urbanization, 10 years later, more than 90 percent of Chinese households were with the aim of increasing the share of the population that lives homeowners (Sato et al. 2013). in cities. This new urbanization policy, however, does not appear Private property generates income for households through to address the factors that underlie urban-rural income differen- interest earnings, dividends, rents, and capital gains. Because of tials. It is possible that the new urbanization program will replace data limitations and the underreporting of these forms of income the rural-urban gap with an urban-urban gap, wherein the urban in the survey data, some uncertainty exists about their size and population is divided between the privileged, already established distribution. Available estimates, however, show substantial urban population, and a new urban underclass composed mainly increases in the importance of asset income. For example, Sato of migrants and former rural residents. et al. (2013) report that the share of asset income in household The evolution of asset ownership by Chinese households is still income nearly doubled in the five years from 2002 to 2007, from in its early stages. Yet, we can safely predict that, over time, house- about 8 percent to 15 percent. hold wealth in China will increase—and it will do so unequally. A As in most countries, asset income in China is distributed more lack of equity in the process of wealth accumulation makes this fac- unequally than wage earnings and other components of income. et of inequality politically complex and controversial. Timely policy Therefore, the growth in asset income for those who own prop- reforms to regularize financial and real estate markets, strengthen erty has contributed to increased levels of inequality. Based on property rights especially for rural and poorer households, institute calculations using CHIP data, we found that in 2002 asset income taxation of property and wealth, and limit opportunities for irregu- contributed to 8–10 percent of national income inequality; by lar and illegal gains, could help offset the impacts that the expan- 2007, it contributed to 13–19 percent. sion of private wealth will have on inequality. China’s new leaders, President Xi Jinping and Premier Li Concluding Thoughts Keqiang, thus inherit from their predecessors continually evolv- China’s shift from a low-inequality to a high-inequality economy ing challenges in reducing inequality. In the context of a changing occurred during a period of rapid macroeconomic growth. This economy and emerging sources of inequality, meeting those chal- growth contributed to improved incomes for all socioeconomic lenges will require new and creative policy approaches. groups and provided resources for redistributive and social programs. In the wake of the world financial crisis, however, China has Notes entered a period of slower growth, which creates new challenges 1. The Gini coefficients in Figure 1 are calculated using income for promoting equity. China can no longer rely solely on macro- data from China’s National Bureau of Statistics. economic growth to raise the incomes of poorer socioeconomic 2. The Gini coefficients for the other countries reported here are groups. The use of public funds for redistributive policies may now for 2005, are measured over household income per capita, and are 4 : : August 2013 : : Inequality in Focus from the UNU-WIDER WIID2c database, http://www.wider.unu. lenges to a Harmonious Society. New York: Cambridge Univer- edu/research/Database/en_GB/wiid/ (accessed August 12, 2011). sity Press. Note that the Gini coefficients for Brazil and Honduras are the Li, Shi, Hiroshi Sato and Terry Sicular, eds. 2013. Rising Inequal- highest among all the countries listed in 2005-2006. ity in China: Challenges to a Harmonious Society. New York: 3. Based on national Gini estimates reported by the World Bank, Cambridge University Press. http://data.worldbank.org/indicator/SI.POV.GINI/ (accessed June 20, 2013). Lustig, Nora, Luis F. Lopez-Calva and Eduardo Ortiz-Juarez. 4. Estimates of income and inequality based on the CHIP survey 2012. “Declining Inequality in Latin America in the 2000s: data use a fuller definition of income than those based on NBS The Cases of Argentina, Brazil, and Mexico,” World Develop- income data. The CHIP income figures include imputed rents on ment, 44, 129–141. owner-occupied housing and implicit subsidies on subsidized National Bureau of Statistics. Various years. China Statistical urban rental housing. Yearbook. Beijing: China Statistical Press. References New York Times. 2013. “China’s Wen Warns of Inequality and Gustafsson, Björn, Li Shi, and Terry Sicular, eds. 2008. Income Vows to Continue Military Buildup.” Andrew Jacobs and Inequality and Public Policy in China. New York: Cambridge Chris Buckley. March 4. University Press Ravallion, Martin, and Shaohua Chen. 2007. “China’s (Uneven) Knight, John, and Lina Song. 1999. The Rural-Urban Divide: Progress against Poverty.” Journal of Development Economics, Economic Disparities and Interactions in China. Oxford: Oxford 82(1), 1–42. University Press. Sato, Hiroshi, Terry Sicular and Ximing Yue. 2013. “Housing Knight, John, Terry Sicular, and Ximing Yue. 2013. “Educational Ownership, Incomes, and Inequality in China, 2002–2007,” Inequality in China: The Intergenerational Dimension.” Chapter 3 in Li Shi, Hiroshi Sato and Terry Sicular, eds. Ris- Chapter 4 in Li Shi, Hiroshi Sato, and Terry Sicular, eds. Ris- ing Inequality in China: Challenge to a Harmonious Society. New ing Inequality in China: Challenge to a Harmonious Society. New York: Cambridge University Press. York: Cambridge University Press. Sicular, Terry, Ximing Yue, Björn Gustafsson, and Shi Li. 2007. Li, Shi, Chuliang Luo, and Terry Sicular. 2013. “Overview: “The Urban-Rural Income Gap and Inequality In China,” Income Inequality and Poverty in China, 2002-2007.” Chap- Review of Income and Wealth, 53(1), 93–126. ter 2 in Li Shi, Hiroshi Sato, and Terry Sicular, eds. Rising Inequality in China: Challenge to a Harmonious Society. New The World Bank. 2009a. From Poor Areas to Poor People: China’s York: Cambridge University Press. Evolving Poverty Reduction Agenda, an Assessment of Poverty Li, Shi, Hiroshi Sato, and Terry Sicular. 2013. “Introduction: and Inequality in China. Washington, DC: World Bank. Rising Inequality in China.” Chapter 1 in Shi Li, Hiroshi The World Bank. 2009b. The World Development Report 2009: Sato, and Terry Sicular, eds. Rising Inequality in China: Chal- Reshaping Economic Geography. Washington, DC: World Bank. Inequality and Unrest in Brazil: A Conversation with Ricardo Paes de Barros R ecently, the Inequality in Focus editorial staff interviewed lower-cost public transportation, but eventually evolved into a wide variety Ricardo Paes de Barros, the secretary of strategic actions in of demands on the state. The interview discusses some of the reasons behind Brazil’s Ministry of Strategic Affairs. He is an expert on pov- these protests, as well as Brazil’s current state of development, equality, and erty and inequality issues and has been intimately involved social policy. This is a summary of the conversation. in some of Brazil’s signature social policy initiatives, includ- ing Bolsa Familia, the national conditional cash transfer program. Foreign Inequality in Focus: In your opinion, what has been the Policy magazine recently identified Paes de Barros as one of “the 500 most main cause of the recent protests in Brazil? powerful individuals on the planet.” The two-hour interview touched on Ricardo Paes de Barros: It is important to acknowledge that there many issues, but within the context of the widespread popular protests that is no clear and simple answer to that question, because nobody really began in Brazil in June 2013. The initial demands of the protests were for knows. The causes are puzzling to many with neither the government Inequality in Focus : : August 2013 : : 5 nor the opposition ever expecting this. There were never any social or there is a general demand to allocate resources differently. economic indicators that could have predicted this level of discontent. The other contributing factor is that the World Cup has drawn Actually, it is kind of amazing how a country with relatively low levels attention to the allocation of public spending. And what people are of unemployment, falling poverty rates, and reduced inequality is questioning is why we are spending so much on preparing for the experiencing such unrest. World Cup when resources are so scarce for education, health, public That does not mean that social and economic conditions in Brazil safety, and transportation. Why are we investing more in building are perfect. Levels of poverty and inequality are still high. But, they are roads to the stadium than building roads to work? Although much of falling. For example, infant mortality rates are very high—still two this World Cup investment is actually private, it’s possible that the re- deaths per thousand—but the rate at which this is declining in Brazil cent Confederations Cup football tournament lit the flame in people’s is among the top five fastest in the world. heads; not so much about whether the quality of services is good or Still, recent opinion polls show increasing levels of public dissatis- bad, but whether the allocation of public funds is right or wrong. faction with the quality of government services. Interestingly, it seems But this is all speculation, and all we really know is that we are that the public’s opinion of government services is changing much seeing a movement. Maybe it is a movement that is seeking a slightly more radically than the actual quality of services. In other words, the modified development path, or maybe it is seeking major change. satisfaction with the quality of public services has declined rapidly, al- Maybe all of this has something to do with greater representation and though the actual measurements of quality have shown improvement. changes in public spending. But it is not yet clear what these messages I think this is the most basic and immediate explanation for the are. And that is the difficulty in reading too much into the messages. protests. It is purely a change of perceptions—not that quality has deteriorated and people want quality levels back to what they were. IF: So should the government respond by shifting its focus It’s amazing to what level people’s expectations have increased; from more to the provision of “pure” public goods (e.g., justice, law an environment in which expectations were holding steady—people enforcement, logistics, etc.) and away from the provision of were happy—to all of a sudden having a situation where hopes have private goods (e.g., health, education, transfers)? climbed and where what previously seemed good now seems bad. PB: Focusing on the provision of “pure” public goods and services is fine for a country with no inequality. In theory, if everyone had equal IF: What caused this change in perceptions? opportunities, the state could stand in the corner, not worry about PB: Most importantly, because of increasing incomes and economic redistribution, and focus on the provision of pure public goods and growth you have 40 million people who have moved out of poverty services. This would benefit the entire public equally. However, in a to the middle class in Brazil. When you take such a large number of country that has high levels of inequality, like Brazil, the priority must poor people and place them squarely into the middle class, then sud- denly what was important to them before—surviving—has changed. first be to reduce inequality. In a very short period of time you had 40 million more people enter What does it mean to reduce inequality? It means that you must the formal labor market, pay taxes—and who now reflect on how improve the situation of the poorest at a faster rate than you do for the their tax money is used. Before, these people had no time to contem- richest. To do this, you must target social policy toward the poorest, plate these things, they had no idea how much they were paying in and not toward everyone equally. In the end, this means that social taxes, they were more worried about what would happen tomorrow policy is not designed to deliver pure public goods, but rather to pro- and whether they would have money for food or bus fare. vide publicly funded private goods—like education, health, transfers, At the same time that you have these 40 million people moving etc. —for the most impoverished people. This is what Brazil has done into the middle class, you have another 30 million people who make and we have done it quite successfully. We have invested relatively up the richest 15 percent of the population and their incomes are more in publicly funded private goods for the poor than in pure not improving much. What is the flip side to reducing inequality? public goods, which would stand to benefit the non-poor the most. To reduce inequality, you must have an unequal growth rate. If every Therefore, it is not a huge surprise that the population of a country income group grows equally, you do not reduce inequality. To reduce with an expanding middle class would begin to demand improved inequality, poor incomes must grow faster than rich incomes. In Brazil transportation services and greater security; that is, more typical over the last 10 years, the richest 15 percent had a positive growth “pure” public goods. rate, but well below the national average. This means that for one of I think Brazil had made a decision, which was “we will take seri- the first times in history, the richest 15 percent of Brazilians have seen ously the issues of inequality and poverty, and therefore, we will themselves falling closer and closer to the middle class. As a conse- focus our attention on the poor.” Much of Brazil’s success in reducing quence, they are also less satisfied and increasingly question how their inequality was based on the decision to prioritize the needs of the tax money is spent. poor above all else. Today’s slogan in Brazil is, “A wealthy country is Although the spending decisions by the government have not one without poverty.” changed much over the past 10 or 20 years, 70 million people have When you really look at it, it is counterintuitive that such a policy started to reach the conclusion that the allocation of public spending would be politically feasible in a country with only a 10 percent il- is not exactly what they want. If you look carefully, you can see that literacy rate and only 20 percent of its population squarely living in 6 : : August 2013 : : Inequality in Focus poverty. However, despite their own self-interests, large portions of were happy, and businesses were happy because people didn’t have to Brazilians from all socioeconomic classes believe in this slogan. worry about their children and could show up on time for work. But now I’m doing this other thing.” This guy is already on another level. IF: But as you mentioned, there is an emerging middle class There is nothing like having an active and intelligent mayor interact- in Brazil and their values and expectations of the govern- ing with the community everyday to see what problems need to be ment are changing. What do you believe these new values solved and what changes to policy need to be made. to be? At the same time, there are things the mayor will not see simply by PB: If you look at the evidence, the values of the middle class tend walking around a city. There may be some environmental problems to be a mix between the values of the richer classes and the poorer that are not seen because nobody wants to see them. Maybe a factory classes. It is a middle group that thinks in the middle. Who hates to is generating pollution, but the factory creates jobs and is a boon for pay taxes the most? The rich hate them the most, the poor the least, the local economy so nobody wants to acknowledge the negatives. and the middle class somewhere in the middle. Who thinks the qual- For this reason, using objective data and empirical evidence is also ity of public services is bad? The poor think it’s bad, the rich think it’s important. In the case of Petrolina, the mayor is tapping into this lousy, and the middle class is in between. So far we have not seen any empirical information as well. For example, he is using census data to evidence that the middle class is actually revolutionizing, moderniz- identify the scope of the challenges. of Brazilian society. ing, or rethinking the values ​​ What I am saying is this: Empirical evidence is more than welcome But of course, with the rise of the middle class, the needs and and in some cases irreplaceable. But you will never design public aspirations of the average Brazilian are changing, and this represents policy 100 percent based on evidence, which is often late or incom- a tremendous challenge for public policy. How can you adapt policies plete. The idea is to collect as much information as possible while also to these changes? I think an improved policy is one that is no longer keeping a keen ear to the ground. solely focused on the poor but also takes into account the needs of the emerging middle class. The government is very aware of this. So, IF: How do you then go about improving public policy now Brazil’s slogan is changing to, “The end of poverty was just the evaluation in Brazil? beginning.” PB: Brazil, as a nation, is quite good at evaluating public policy. We What this means for public policy in Brazil generally is, “How do have the institutions, the money, and the demand to evaluate the you handle success?” A successful public policy also means the death performances of both the program managers—like ministers, state of that policy. For example, if you find an excellent remedy for a secretaries, governors and mayors—and the programs themselves. certain disease and, as a result, cure the disease, you no longer have What Brazil is lacking is an institution, not to evaluate—it is better a need for that remedy. It is the same thing here, Brazil has been suc- that the private sector, universities, and nongovernmental organi- cessful in reducing poverty, but because of this success the old social zations do this—but to coordinate all of the evaluation efforts; to policy is less relevant. The question then becomes, “how do you determine which evaluations are sound and which are not and to ensure that public policy continues to remedy the problems.” Well, make this information visible to everyone; to create a list of certified you must continue to adapt public policy as success is attained. If evaluators and maybe a list of institutions in search of evaluators; and you have a new middle class, you have to adjust the policy to address many other things. So, if you need an evaluation of a certain social their needs. A continuously new perspective needs to be incorporated program, this agency could not only direct you to it, but also provide into policy making in all facets—from the justice system, to public a review of the evaluation. safety, to public transportation. Practically, such an effort could be managed by Brazil’s Federal Ac- Of course, in practice the solutions are not that simple. Redesigning countability Office (Tribunal de Contas da União), by the Institute of social policy requires other factors as well. How do you adjust the tar- Applied Economic Research (Instituto de Pesquisa Econômica Apli- geting of these policies while maximizing the ability to recover costs? cada), or by a stand-alone government agency. The point is to have a How can you alter public-private partnerships in order to pay for or government agency that regulates, guides, and funds the evaluation deliver these new services? Ideally, you want the new social policy to process, but doesn’t actually do the evaluating. If it did both there be highly efficient. This means that the communities and the private would be potential conflicts of interest. sector manage these services and the government stays in a position IF: This brings up an interesting question on the role of the to regulate and fund them. public and private sectors in the provision of social services. IF: What would need to happen to be able to more effec- What do you believe the role of each should be? tively adjust Brazilian social policy? PB: What we want is a world of high quality services. To do so you PB: A constant evaluation of the social and economic situation in need a fully integrated service. What I mean is a service that is equally Brazil is required, as well as the ability to enact change. For example, accessible to the poor, the rich, and the middle class and of high I recently visited the mayor of Petrolina, Pernambuco, in northeast- quality for all. Ideally this would be a service provided by the private ern Brazil. He asked me, “Remember how we increased access by sector, paid for by the public sector, and certified and regulated by the families to child care last year? It had such a positive impact: people public sector. Such a system would be free to the poor—or at least Inequality in Focus : : August 2013 : : 7 they would be reimbursed for their expenses—while the rich would a segregated world, rich communities would have higher quality have to pay the state for the service they are receiving. schools than poor communities. In other words, if you wanted a Such a system already exists in Brazil for child-care services in world with high quality education that is accessible to all, it would be Petrolina. In this case, the state has a contract with a private company more important to change the community than it would be to change to construct the child-care center, which is paid for by a state-admin- the school. istered fund specifically used for basic educational services. These Therefore, the main role of the state in social policy is not the man- private providers are free to earn profits and compete for contracts. agement of the service, but the design of it. It may be better that the Such a system is beneficial because competition will ensure a high government leaves the management of the teachers and physicians to quality service. Furthermore, such a system can be easily expanded the private sector and concentrates instead on what people want and because the state provides educational funds based on the number of making sure they hire the right agencies to deliver it. In other words, students, not as block sums. So as more students are enrolled in this social policy design is not something that can move to the private system, more resources will be diverted to it. sector. The market should be used to determine what courses can be But such a system would be highly difficult on a larger scale. Not taken and private providers can tell the government what they need, because of its design, but because of inequality in Brazil. In order to but the government would need to regulate the functioning of the have a system of equal quality and access for poor and rich alike, you market and to guide supply and demand to meet its social goals. would have to have residential desegregation—not rich communi- ties and poor communities, but rather communities of rich and poor. This interview took place on July 19, 2013, between Ricardo Paes de Barros Ideally, these integrated communities would be a microcosm of Brazil and Pedro Olinto, Fernanda Luchine, and Maximillian Ashwill of the In- as a whole. In this way, a high quality service can be provided to all. equality in Focus editorial team.The interview, in Portuguese, was recorded What I mean is that in order to have a public service of equal quality and transcribed by Fernanda Luchine. The above dialogue is an interpreta- and availability to all you must also have equality. For example, in tion of the conversation written by Maximillian Ashwill and Pedro Olinto. The Inequality in Focus series aims at informing the public debate authors and do not necessarily represent the views of the World on equity, inequality of opportunity, and socioeconomic mobility. Bank, its Executive Directors, or the countries they represent. It features articles written by World Bank staff, as well as research- ers and policy makers from the broad development community. The Inequality in Focus series is not copyrighted and may be The views and interpretations in the articles are those of the reproduced with appropriate source attribution. Editorial Committee: Pedro Olinto (managing editor), Maximillian Ashwill (senior editor), Jaime Saavedra, Francisco Ferreira, Luis-Felipe Lopez-Calva, John Newman, Gabriel Demombynes, and Anna Reva Editor: Mary Anne Mulligan THE WORLD BANK Poverty Reduction and Equity Department Poverty Reduction and Economic Management Network (PREM)