The issue of September 2015 of the prestigious World Development journal is devoted to the topic “Growth, Poverty and Inequality in Sub-National Development: Learning from Latin America’s Territories”. The volume contains the findings from a program conducted by a wide network of researchers from over 50 research centers in 11 Latin America countries, to answer three questions: (1) Are there rural territories that have experienced simultaneous economic growth, poverty reduction, and improved distribution of income?; (2) What factors determine these territorial dynamics?, and; (3) What can be done to stimulate and promote this kind of territorial dynamics?

As aptly put by the editors of the volume in the introductory article (Berdegué, Bebbington and Escobal, 2015), “rural development dynamics are usually described and analyzed either through macro narratives (…) or through case studies that give a rich picture of a particular place or situation. Very rarely is there an attempt to describe and explain large-scale questions in a way that combines a discussion of both the forest and of the trees”. This research is the most comprehensive attempt so far to do this.

The central finding is that between the mid- nineties and the mid- 2000s only 12% of about 10 thousand subnational territories in 11 countries experienced inclusive growth, defined as economic growth with poverty reduction and improved distribution of income. In the vast majority of territories, there was little or no improvement in one or more of these dimensions of development. According to the editors, the more significant explanations to the lack of inclusive growth are found in deeply rooted social structures and institutional arrangements (not in the traditional explanations based on resource endowments or on the geographic dimensions of density, distance and division emphasized by World Bank, 2009). The resistance to change due to institutional factors explains why inclusive dynamics are so rare, regardless of the aggregate economic or social performance of the country. Although the main explanation bears many similarities to Acemoglou and Robinson’s (2012) thesis of why nations fail, less emphasis is put on the historical roots of institutional arrangements, and much more on the actions, especially coordinated ones, that foster institutional change allowing territories to escape from poverty and inequality traps.

Given the richness and length of the volume, I focus here on the main empirical findings that are summarized in the article by Mondrego and Berdegué. It is fitting to start with one important stylized fact: inter-territorial inequality explains a relatively small fraction of total national income (or expenditure) inequality (between 11% and 23%, see Table 1). This implies that the huge inequality levels that characterize Latin American societies are also observed at the sub-national level, in a fractal manner.

In the nine countries considered, the general trend points toward spatial convergence of mean household incomes, with the remarkable exceptions of the two fastest growing economies, Chile and Peru (see Figure 1). However, only in Mexico is the speed of convergence sufficiently fast as to be significant from a territorial cohesion point of view. In terms of conditioning factors, econometric results discussed by the authors indicate that human capital is the most cross-cutting factor correlated with mean income or consumption changes. At odds with mainstream economic theories of agglomeration, larger local economies (such as capital cities) do not tend to grow faster. 

Growth also does little to reduce income concentration patterns within territories. On the contrary, in most countries faster income growth (at the local level) is associated with more income concentration (especially in highly unequal Brazil and Chile). No country shows a clear average pattern of territorial growth with decreasing inequality.

However, surprisingly –maybe due to data quality problems–, territorial income inequality is apparently not very persistent: the unconditional persistence coefficients are 37% for Chile, 40% for El Salvador, 22% for Mexico and 57% for Nicaragua. Importantly, the changes observed are highly idiosyncratic, probably due to the importance of territorial institutional frameworks that cannot be controlled for in the econometric exercises.

Growth is much more effective in reducing poverty. The conditional semi-elasticities of poverty rates to growth are consistently high and of similar magnitude across countries (between -34% and -56%). Poverty is also very sensitive to income inequality changes: after controlling for many other factors, each percent point of increase in the Gini coefficient of a location is associated with an increase of 0.4 percent points in the poverty rate. Interestingly, there is no consistent effect of agglomeration on poverty reduction across countries, implying that migrating to a larger city is not necessarily the safest escape from poverty.

For the authors, the most important policy implication of these results is that “in contrast to policy approaches that advocates for a spatially concentrated growth aided by spatially blind institutions (e.g., World Bank, 2009) … [the] findings build a strong case for place-based development policies in the region”. I could not agree more. However, what policies are more effective to that purpose is still an open issue.

References:

Acemoglu, D., J. Robinson (2012). Why nations fail. The origins of power, prosperity and poverty. Crown Publishers, New York.

Berdegué, J., A. Bebbington and J. Escobal (2015). Conceptualizing Spatial Diversity in Latin American Rural Development: Structures, Institutions and Coalitions.  World Development, 73, 1-10.

World Bank (2009). World development report 2009: Reshaping economic geography.

World Bank Publisher, Washington DC, USA.