Resumen de The effects of land inequality on economic growth: a regional and comparative study, 1950-1970
The scarcity of resources, or the unequal distribution of them, has always been one of the main fields of study in economic history. This thesis analyses the unequal distribution of one of the most basic economic resources: land.
Land is a basic asset in economy, and it is one of the main factors of production. In fact, Adam Smith’s most famous work, the Wealth of Nations (1776), already emphasized that any society developed to produce a good needed three productive factors: land, labour, and capital.
Several authors have investigated how the distribution of land has been able to impact the development and economic growth of countries. For example, Barrington Moore in his work, The Social Origins of Dictatorship and Democracy (1966), was one of the first researchers to point out that countries with strong landed oligarchies where land was unevenly distributed experienced more difficulties in becoming democracies, and to open up to modern economic growth. Years later, various authors such as Galor, Deininger, or Frankema, among others , defended the hypothesis that the way in which land has been owned and distributed among the population has left a mark on subsequent economic growth. In fact, some authors such as Deininger and Squire even hypothesize that unequal land distribution is the real reason behind economic inequality, and not differences in income distribution.
How can the unequal distribution of land affect growth? Theoretically, the hypothesis defended by these authors is that in countries where there was unequal distribution of land, economic and political power was concentrated in hands of landed elites, who tended to design structural frameworks that defended their economic and social interests to the detriment of the majority of the population, thus delaying subsequent economic growth.
To understand why land was unequally distributed, or how this inequality was linked to elites that hold back development, and what real effects this inequality had on growth; this thesis analyses the distribution of land and its evolution in Latin America in the mid-20th century. Additionally, this thesis provides a new and original data on land distribution.
The reason for choosing Latin America is simple. Latin American countries have been characterized throughout their history by having unequal distribution of land. Furthermore, they have distinguished themselves for being predominantly agrarian countries, and for having powerful landed elites. On the other hand, the period of study chosen (1950-1970) is of great interest. The reason is simple, in these years great agrarian reforms were carried out around the world, and specifically in Latin America. During these years, in countries such as Bolivia, Chile, Colombia, or Mexico, great agrarian reforms were designed and carried out with the aim of improving access to land. Additionally, in the middle of the 20th century, a large part of the Latin American countries applied policies to promote industrialization, with the aim of turning Latin American countries into more industrial than agrarian. These policies sought to reduce the importance of the agricultural sector in the economy, improve access to land, and above all facilitate industrialization in Latin America. Obviously, these circumstances affected the distribution of the land, which is why this period is so interesting to study. In these years we can see more clearly if there was unequal distribution of land in Latin America, if this inequality could be negatively affecting growth, and if the decrease in this inequality, thanks to the reforms, had any economic effect.
To meet the objective of this thesis, I analyse the distribution of land in Latin America in the mid-twentieth century providing new and original data on the distribution of land, I also empirically analyse the supposed effects that this inequality can have on growth through three channels: political stability, human capital formation, and access to credit.
Consequently, in this thesis, Chapter 1 deals with the scarcity of data on land distribution, and with the different measures and indices found in the literature to analyse land distribution. First of all, it should be noted that the measure chosen in this thesis to study the unequal distribution of land is the landgini index. This index enjoys great support in the empirical literature as a statistical tool to analyse the unequal distribution of land. Secondly, it is necessary to mention that data on land distribution is scarce, published and unpublished, and particularly for these years. For this period, only Taylor and Hudson in 1972, Deininger and Squire in 1999, and Frankema in 2008 offered 36, 28 and 29 landgini estimates respectively, at the global level. For this reason, the main contribution of this thesis is to fill in the gaps in the literature by providing the largest collection of data on land inequality. For this I have compiled the agricultural censuses of 18 Latin American countries in the mid-20th century, and I have calculated the landgini of all of them at the national and sub-national level for the period 1950-1970. However, due to the need to have the greatest possible understanding, and a global vision of the evolution of land inequality in Latin America, other countries are included in the study. New estimates of landgini for Canada, Western Europe, and certain African countries are also offered in this study in an original way for these years. Thanks to the agricultural censuses of each country, I can offer a total of 89 new estimates of landgini at the national level, and a total of 649 new estimates at the sub-national level.
Chapter 2 focuses on one of the channels through which land inequality can affect economic growth: political stability. In this chapter I explore the possible relationship between unequal land distribution and democracy by investigating the role of land inequality in holding free and fair elections. According to the main claims in the literature, countries with high land inequality have been linked to latifundia systems and strong landed elites. These elites have opposed the holding of free and fair elections out of the fear of losing their lands. These landed elites, who own most of the land in their countries, are wary of truly representative political systems where political parties not related to them may win elections. True democracy would cause the redistribution of its main asset: land, the loss of cheap labour in countryside, and higher taxes on its assets. While countries with low inequality have been linked to small family farms, and commercial elites. In these countries, both small and medium-sized farmers and business elites do not fear free and fair elections. The land is distributed in a more equal way, so the risk of expropriation faced by the majority of the agrarian population is minimal. Furthermore, the wealth of the commercial elites in these countries does not come from the exploitation of large farms (latifundia system), or from the use of cheap labour, and further still, their assets and commercial transactions are more difficult to impute than those of the landowners. To analyse this hypothesis, I use, on the one hand, the index of free and fair elections, which indicates if free and fair elections are held in a given country and to what degree. And on the other hand, my new and original database of landginis indexes. This new landgini database has been calculated from the agricultural censuses of each country. In this chapter, the sample of countries contains the 18 Latin American, 12 European, 4 African, the United States and Canada. The results show that, in general, Latin American countries registered higher levels of land inequality than European countries, the United States and Canada. On the other hand, using ordinary panel least square, fixed and random effects approach, I find that countries with higher land inequality, Latin American or not, held less free and fair elections. Furthermore, countries with high land inequality showed more voting restrictions than those countries that were organized on family farms with low land inequality. In most cases, land inequality was linked to politically influential landed elites that were detrimental to free and fair elections, particularly in regions such as Latin America.
Chapter 3 focuses on the effects that land inequality can have on human capital formation. In this case, the main claim of the literature is that countries with higher land inequality will be united with strong landed elites, which will oppose the expansion of mass literacy. This fact would cause greater illiteracy and worse human capital formation, and as a consequence lower economic growth. It is true that throughout history, in Latin America there has been high land inequality, and strong landed elites who have opposed mass literacy for economic reasons (higher wages for workers, better working conditions, and possible loss labour due to increase job opportunities for workers), and for political reasons (the literate population tends to demand more voting rights, and they are more politically aware). For its part, in the United States historically agriculture has been organized mainly on small and medium-sized family farms; farm families themselves have had more freedom and possibility to decide on their training and that of their children. Moreover, in the United States the elites have been primarily commercial and have seen no threat from the spread of literacy. To analyse this hypothesis raised by the literature, according to which countries with high land inequality, large landowners systems and elite landowners will have worse literacy levels than those countries with low land inequality, organized in family farms and with low land inequality, a new database on illiteracy is calculated in this chapter. Using the population censuses of 11 Latin American countries and the United States, I calculate the illiteracy rates for the 14+ and 5-10+ age groups, at the national and sub-national levels for all countries in the sample. These new illiteracy rates together with my also original landgini estimates, and together with the estimation of econometric models with ordinary least square, fixed and random effects, show us that the hypothesis raised by the literature must be accepted. Land inequality negatively affects literacy. Countries with higher land inequality showed, in general, higher levels of illiteracy. However, not all Latin American countries fit this general description. Some regions of Costa Rica showed lower levels of land inequality and illiteracy than some states in the United States. Within the United States, southern states such as Louisiana had higher levels of land inequality and illiteracy than certain Latin American regions. These results, although they may seem surprising, fit the hypothesis raised in this chapter, since some of these American states had a latifundia past, with strong landed elites, and high levels of land inequality similar to many Latin American regions. Furthermore, the results suggest that land inequality tends to accentuate illiteracy more among adults than among young people.
Finally, chapter 4 of this thesis analyses the possible impact that land inequality can have on access to credit. According to the literature, land inequality causes most of the population to lack collateral to access the banking system. To obtain a loan, it is generally necessary, or at least helpful, to be able to provide collateral. The unequal distribution of land means that a small part of the population concentrates most of the land in their hands, which means that a large part of the population does not own any piece of land to offer as collateral when they need to obtain a loan. Moreover, land inequality is associated with a politically and economically influential landed elite who can use their influence to skew credit for their benefit. Landlord elites may have an interest in skewing credit to maintain their economic and social position. Preventing the rest of the population from having access to credit prevents them from making investments, increasing productivity, and in general competing with the landed elite. Additionally, preventing the population from owning properties encourages the population to have to work for other people, which ensure a supply of cheap labour for landowners. In this case, to examine this hypothesis, this chapter focuses on the case of Mexico. This country has been an agrarian country for most of history, with high land inequality, strong landed elites, and a shortage of rural credit. Moreover, the case of Mexico is a unique case because an attempt was made to solve this situation, an agrarian reform. The Mexican reform is one of the most important agrarian reforms in history. This reform had the objectives of reducing land inequality, limiting the power of the landed oligarchy, and improving access to rural credit by granting land to landless peasants who could use it as collateral to obtain loans. To confirm if land inequality had a negative effect on access to rural credit, and if the agrarian reform decreased land inequality and improved access to credit; I elaborate an original and unique dataset of the rural credit received by the farms, and of land inequality, all done state-by-state that I have obtained from the original censuses of Mexico. With these new estimates, I estimate econometric models with ordinary least square, fixed and random effects. The results reflect that land inequality, although harmful, did not affect access to credit in any Mexican state. In fact, access to credit depended on political factors and not on high land inequality. The revolution and the Mexican agrarian reform caused a change in the ruling elites of the country, the landed elites were replaced by new political elites, who designed land reform in such a way that redistribution was a continuous process. Beneficiaries of land redistribution did not always obtain full property rights to the land received. These beneficiaries, if they received land, could not always sell, rent, or use this land as collateral. For this reason, the beneficiaries to obtain credit had to turn to the new government, which granted land and credit based on the political support it received. This fact caused that despite implementing an agrarian reform, neither land inequality nor access to rural credit improved in the case of Mexico.
Focusing from more to less, first with an international analysis, then with an analysis focused only on America, and finally, analysing only the case of Mexico. This thesis analyses land inequality and its effects on economic growth.
The results show, first, that land inequality in the mid-twentieth century was still high in some countries, especially in Latin America. Second, most of the time, in line with the literature, this inequality was linked to a strong landed oligarchy, which used its political and economic influence to protect its interests, slowing the economic growth of the countries. Third, the channels through which land distribution can affect growth are diverse, but this thesis has focused on political stability, human capital formation, and access to credit. The main reason for this choice has been to try to cover the main factors that can affect growth. Through political stability, a given country can allow for democratic systems or policies that promote growth. In this case, this thesis has shown that countries with high land inequality are linked to institutions that do not allow the proper political functioning of the country. Generally, countries with high land inequality hold more fraudulent elections, which benefit elites that do not apply adequate policies for growth. On the other hand, through the formation of human capital, this thesis has shown that countries that have historically had high levels of land inequality have had worse human capital formation. As mentioned above, land inequality has been linked to landlord oligarchies who have seen mass literacy as a threat to the status quo. As a result of this situation, these oligarchies have slowed down the formation of human capital, thus damaging growth. Finally, through access to credit, this thesis has shown that although land inequality can produce a lack of collateral to access credit flows, in the case of Mexico land inequality was not a determining factor, being political factors that made it difficult to access credit.
Keeping in mind the limitations that an international analysis may have, this thesis argues that land inequality is harmful to growth, especially through political stability and the formation of human capital.
