«Der Open-Access-Publikationsserver der ZBW – Leibniz-Informationszentrum Wirtschaft The Open Access Publication Server of the ZBW – Leibniz ...»
Der Open-Access-Publikationsserver der ZBW – Leibniz-Informationszentrum Wirtschaft
The Open Access Publication Server of the ZBW – Leibniz Information Centre for Economics
Gundlach, Erich; Navarro de Pablo, José; Weisert, Natascha
Education is good for the poor: a note on Dollar and
Provided in Cooperation with:
Kiel Institute for the World Economy (IfW)
Suggested Citation: Gundlach, Erich; Navarro de Pablo, José; Weisert, Natascha (2004) :
Education is good for the poor: a note on Dollar and Kraay, In: Shorrocks, Anthony F. Hoeven, Rolph van der (Ed.): Growth, inequality, and poverty, ISBN 0-19-926865-7, Oxford Univ. Press, Oxford [u.a.], pp. 92-106
This Version is available at:
Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Documents in EconStor may be saved and copied for your Zwecken und zum Privatgebrauch gespeichert und kopiert werden. personal and scholarly purposes.
Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle You are not to copy documents for public or commercial Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich purposes, to exhibit the documents publicly, to make them machen, vertreiben oder anderweitig nutzen. publicly available on the internet, or to distribute or otherwise use the documents in public.
Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, If the documents have been made available under an Open gelten abweichend von diesen Nutzungsbedingungen die in der dort Content Licence (especially Creative Commons Licences), you genannten Lizenz gewährten Nutzungsrechte. may exercise further usage rights as specified in the indicated licence.
zbw Leibniz-Informationszentrum Wirtschaft Leibniz Information Centre for Economics
Education Is Good for the Poor:
A Note on Dollar and Kraay (2001)* by Erich Gundlach, José Navarro de Pablo, and Natascha Weisert revised, March 2002 Erich Gundlach Kiel Institute of World Economics P.O. Box 4309 24100 Kiel, Germany firstname.lastname@example.org José Navarro de Pablo Group Policy and Research UBS P.O. Box Zurich 8098, Switzerland email@example.com Natascha Weisert Graduate Institute of International Studies Rue de Lausanne 132 1211 Geneva, Switzerland firstname.lastname@example.org * We thank Bert Hofman for basic ideas and two anonymous referees and participants of the WIDER Development Conference on Growth and Poverty for constructive comments on an earlier version.
Education Is Good for the Poor:
A Note on Dollar and Kraay (2001)
ABSTRACTA recent paper by Dollar and Kraay (2001) finds that higher primary educational attainment of the workforce does not increase the income of the poor except for its effect on average income. We test the robustness of their finding by using a broader measure of human capital that accounts for international differences in the quality of education. Our findings suggest that more quality-adjusted education does increase the income of the poor in addition to its positive effect on average income. Hence effective education policies should be an essential component of any poverty-reduction strategy.
JEL: O 15 Keywords: Poverty, income distribution, education, growth I. Introduction and Background A recent paper by Dollar and Kraay (2001) finds that growth is good for the poor, but that the income of the poor does not respond systematically to supposedly "pro-poor" policies such as public expenditure on education. Using a sample covering 137 countries over the period 1950they report that the income of the poor rises one-for-one with average income. However, the primary educational attainment of the workforce (and the level of primary enrollment, as in an earlier version of their paper) does not seem have a measurable effect on the income of the poor beyond its effect on average income. Hence, their work tends to suggest that a focus on education rather than on growth might be misplaced as an essential component of any poverty-reduction strategy.
We test the robustness of the findings by Dollar and Kraay (2001) by using a broader measure of human capital, which considers all levels of education and accounts for international differences in the quality of education. Contrary to Dollar and Kraay, we find that a higher stock of human capital increases the income of the poor, not only through its effect on average income, but also through its effect on the distribution of income. Our results appear to be robust to a number of alternative specifications. We interpret our findings as suggesting that effective education policies would be a first-best poverty reduction strategy.
Our interpretation of the empirical evidence seems to be more in line with a policy strategy favored by the Development Report 2000 (World Bank 2000a) than the paper we seek to criticize, which in fact emanated from the World Bank’s research department. With its focus on attacking poverty, the Development Report goes significantly beyond the message conveyed by Dollar and Kraay (2001) in that economic growth is merely considered to be a necessary condition for achieving development and
reducing poverty, but it is not deemed a sufficient force. Effective antipoverty strategies are meant to focus on three additional issues:
strengthening the participation of poor people in local decision-making and fighting discrimination; reducing vulnerability of the poor to economic and natural shocks, sickness and violence; and lastly, expanding economic opportunity and access to assets, such as education, capital and land. An additional study by the World Bank on growth and poverty (World Bank 2000b) further emphasized the centrality of education in the development process. This study argues that human capital appears to be the main asset of most poor people. Hence, investment in the human capital of the poor should be a powerful way to augment their assets, redress asset inequality and reduce poverty.
Recent analyses of international differences in output per worker and growth rates have also raised the awareness of the role of human capital in development, either as a direct or as an indirect factor.1 The endogenous growth literature emphasizes the centrality of human capital for innovation and technological progress. Most empirical cross-country studies of long-run growth now include some measure of human capital.
Regardless of the underlying model, it is a fairly robust empirical finding that a country's human capital is almost always identified as an essential ingredient for achieving growth.2 However, the quantitative impact of human capital on growth has not been precisely estimated up to now.
See, e.g., Mankiw et al. (1992), Benhabib and Spiegel (1994), Hall and Jones (1999).
See, e.g., Temple (1999) and Krueger and Lindahl (2000).
The centrality of education in poverty-reduction policies stems from the belief that education is a powerful equalizer. However, this belief cannot command strong theoretical support. Ram (1989) reviews several theoretical frameworks linking the level of schooling and its dispersion with income inequality, such as human capital or dual-economy-type models. He finds that these models do not generate any clear theoretical hypotheses about the effect of education on income inequality or absolute poverty.
For instance, traditional human-capital models of earnings provide two opposing insights with regard to the relationship between education and income-distribution. First, holding other things equal these models imply a partial positive relation between the mean level of schooling and earnings inequality, such that if the mean level of schooling rises, wages of educated workers go up relative to wages earned by non-educated workers. But these models also feature a partial positive relation between schooling inequality and earnings inequality in that a more equal distribution of schooling leads to a more equal distribution of earnings.
Knight and Sabot (1983) show these effects in a dual-economy version of the human capital model. Educational expansion has again two different effects on the distribution of earnings and thus on overall income inequality as it raises the supply of educated labor. On the one hand, the composition effect (or Kuznets effect) increases the relative size of the group with higher education (and higher earnings) and thus tends to increase inequality. On the other hand, the wage compression effect resulting from the relatively greater supply of educated labor reduces inequality. Which effect dominates is again unclear and will ultimately depend on the country's level of development, the relative size of the different educational groups, the degree of substitutability between workers with different levels of education, and the wider social, political and economic aspects that affect the structure of relative wages for different educational groups and the demand for labor.
To the extent that formal schooling is a significant component of human capital investment, the recent endogenous growth literature might provide a more conclusive theoretical framework regarding the relationship between educational expansion and income distribution. Tamura (1991) explains income convergence in the developed world by an endogenous growth model with human capital spillovers and heterogeneous agents. In his model, human capital convergence results in income convergence.
Human capital convergence can be induced by educational expansion and the promotion of research activity, and arises because for a given stock of existing knowledge, agents with below average human capital have a higher rate of return to human capital investment.
With a more explicit focus on the formal schooling component of human capital investment, Glomm and Ravikumar (1992) construct an overlapping generations model with heterogeneous agents that provides similar results. The human capital possessed by each individual agent is a function of the parents' stock of human capital, the level of schooling acquired, and the quality of education provided, which is modeled as an increasing function of tax revenue and determined endogenously by majority-voting. Furthermore, they assume that the learning technology exhibits at least constant returns to the quality of schools and the parents' stock of human capital. While they are mainly interested in comparing the effects of public and private investment in human capital on growth and the distribution of income, they also show that income inequality unambiguously declines over time in an economy with a public education sector where the quality of schooling is homogenous. Since the growth rate of any agent's income is inversely related to his initial level, income convergence results in their model.
By contrast, the endogenous growth model suggested by Lucas (1988) does not predict income convergence. In this model, the human capital is supposed to generate internal and external effects, where the latter means that the average level of education also contributes to the productivity of all other factors of production. Assuming that a given percentage increase in human capital requires the same effort independent of the level of human capital already attained, the model generates sustainable growth through the accumulation of human capital. Due to the presumed linearity in the production of human capital, the model is capable of predicting permanent income differences of any size. Incomes would not converge because the incentive to invest in human capital, as measured by the rate of return to education, would be the same across all levels of income and human capital.
Given the various theoretical possibilities, it is probably not surprising that it has proved to be difficult to identify a clear empirical link between education and income inequality up to now. Intertemporal studies are rare in number and, as Ram (1989) notes, also do not appear to point to general conclusions regarding the relationship between education and inequality.3 Fields (1980) and Psacharopoulos and Woodhall (1985) provide extensive surveys of the empirical literature. Some older crosssection studies tend to confirm the equalizing function of education. Ram (1984) challenges these findings by pointing out that the empirical evidence appears generally inconclusive. More recently, a study by De Gregorio and Lee (1999) based on international panel data finds that higher educational attainment (and a more equal distribution of education) plays a significant role in making the distribution of income more equal. Their finding appears to be in conflict with the results by Dollar and Kraay (2001).
Our paper is structured as follows. Section II presents the data and the basic specification for our empirical analysis. Section III presents our empirical results. Section IV summarizes our argument and points out directions for future research.
II. Data and Specification of Variables
2.1 Income distribution As the source for internationally comparable data on the distribution of income, we draw on the data set initially provided by Deininger and Squire (1996). This data set contains Gini coefficients and cumulative quintile shares for 111 countries over a period of 40 years. In line with Dollar and Kraay (2001), we define the average per capita income of the There may also exist several indirect mechanisms which influence the relation between educational expansion and reduced inequality. In particular, there appears to be some empirical evidence for the favorable impact of female education on reducing inequality. For instance, Ram (1989) notes that the expansion of female schooling may improve the income distribution through increasing female labor force participation as well as through reducing fertility.
poor as the average per capita income of the poorest 20 percent of thepopulation.