Climatic differences can create path dependencies even within countries, with local institutions perpetuating inequalities and hurting economic development in the process, writes Evan Wigton-Jones.

Recent years have witnessed a renewed interest in issues of economic inequality, and the history of Brazil offers a unique window both on how it first arises and also how it impacts a country’s long-run development.

The climate is an important determinant of historical inequality in the Brazilian context. Data from a national census conducted in 1920 shows that warmer regions with high rainfall were characterised by plantation economies, with a wealthy agricultural elite and a large underclass of poor labourers. In contrast, cooler and drier areas were more conducive to smaller family farms, resulting in a more equitable society.

Farmers work together in Paraíba to get cattle feed ready for storage (Flávio CostaCC BY-NC 2.0)

This historical inequality had ramifications for local governance. In the early 20th century, municipalities with higher inequality were susceptible to “elite-dominated” institutions. That is, a small agrarian elite expended local public resources on their own interests instead of those of the general public. Importantly, these local institutions have persisted throughout the 20th century, since those localities which were elite dominated in the early 20th century continue to spend less on public goods and welfare in contemporary times.

This local inequality has persisted as well. Data from a demographic census conducted in the year 2000 reveals that areas which were unequal in 1920 are more unequal today. This long-term inequality is also detrimental for present-day municipal economic development, as localities with higher inequality are generally poorer and display lower levels of education and public health. This chain of events proceeds as depicted below:

Figure 1: Schematic diagram showing the indirect influence of climate on contemporary Brazil (author’s elaboration)

Agriculture has long been intertwined in the social and economic fabric of Brazil. It is therefore not surprising that different types of agricultural production would affect the local distribution of income and wealth. Crops such as sugarcane and coffee are known for their “increasing returns to scale” — or the large amounts of land and labour that are required for the efficient production of these commodities. Importantly, these crops also share an affinity for warm temperatures and high rainfall. In contrast, crops such as wheat and beans exhibited “decreasing returns to scale”, where fewer workers and less land are required for profitable production. Unlike coffee and sugarcane, these crops thrive in cooler and drier climates. When long-run municipal climate patterns are more amenable for plantation crops, greater inequality results. If, on the other hand, the climate is more favourable for the production of smallholder crops, a more equitable distribution of resources emerges.

More specifically, the 1920-era distribution of land can be used a proxy for the broader distribution of wealth and income. As Brazil’s economy was predominantly agrarian in 1920, the concentration of landholdings is a good proxy for that of economic welfare more generally. This distribution of land can be quantified with a Gini index – a standard measure of inequality that ranges from 0 (perfect equality) to 100 (one individual holds all land). Municipalities with a plantation climate typically have far higher land Gini indices than do municipalities with cooler and drier weather.

Municipal-government expenditure for the year 1923 (the earliest year for which data on municipal public finances are available) can be used to analyse the effect of inequality on historical local governance. Interestingly, localities with higher land inequality witnessed less public spending on education, health, public goods, and public electricity. For example, a one unit increase in the Gini index is associated with a .83 percentage point decline in such expenditure calculated for Brazilian municipalities for the year 2000.  There is a corresponding increase in expenditure on government administration and the salaries of local legislators. In the Brazilian context, such patterns are consistent with the domination of municipal governance by a local elite.

Historical inequality also has implications for contemporary economic development within Brazil. Here local development can be measured using the Human Development Index (HDI). This metric, which accounts for education, public health and income, has been calculated for Brazilian municipalities for the year 2000. Like the Gini index, the HDI is scaled from 0 to 100, with a higher score indicating a greater degree of development. Historically unequal areas score much lower on the HDI: a one unit increase in 1920 land inequality is associated with a reduction of .38 points in this index.

Children race between crops on their parents’ farm outside Brasilia (Rafael Zart, Ministério do Desenvolvimento Social, CC BY-SA 2.0)

Furthermore, the elite domination of some municipal governments has persisted. Contemporary local government expenditure on resources for the broader public is considerably lower in areas with higher historical inequality. Here I define these resources as municipal public spending on Assistance and Welfare, Health and Sanitation, Housing, and Regional Development. Specifically, a one unit increase in the 1920 Gini index results in a .49 percentage point decrease in such spending over the 1995-2005 time period.

Importantly, local inequality has persisted over the 20th century. Areas with greater land inequality in 1920 contain higher levels of income and land inequality today. For example, a one unit increase in the 1920 Gini index is associated with a 7% increase in municipal income inequality for the year 2000 (and as one might expect, more public welfare spending in 1920 is associated with greater levels of such spending over the 1995 to 2005 timeframe).

These results would suggest the following. First, inequality can exert a strong effect on institutional and economic development; furthermore, these effects are detrimental to both. These findings thus contribute to an increasingly important discussion within economics over the broader consequences of disparities in economic welfare. The wealth of subnational data in Brazil also allows for a more comprehensive analysis of inequality, enabling an examination of its effects on both human and local institutional development.

These results also contribute to a growing literature on geographic endowments and long-run development. A unique finding here is that natural endowments such as the climate can create a path dependency even within countries. When local governments have a degree of autonomy over the allocation of public resources, they may be just as susceptible to elite capture as those at a national level. In the case of Brazil, this would suggest that local institutions may perpetuate the historical distribution of resources, and inhibit local economic development in the process.

• The views expressed here are of the authors and do not reflect the position of the Centre or of the LSE
• Please read our Comments Policy before commenting

Evan Wigton-Jones – University of California, Riverside
Evan Wigton-Jones is a PhD candidate in the Department of Economics at the University of California, Riverside. His research interests include development and agricultural economics, as well as economic history. He can be reached at:

Print Friendly