Across the developing world, providing land through agrarian reform has been a popular strategy for expanding economic opportunities among the rural poor. Its use is commonly debated in several developing countries, including South Africa, China, India and many others in Latin America. A widely held view against providing land is that it traps recipient families in the countryside, forcing them to remain in the subsistence sector (
Banerjee, 2000). This limits their chances to move up the social ladder. However, in my
job market paper I find that the children of recipients may tell a different story.
I examine how applicants to the 1968 Colombian agrarian reform and their children fared in life. I use ID numbers to track them in various administrative data four decades later. Moreover, I use a regression discontinuity design to estimate causal impacts. During the reform, a selection mechanism combined data from several socioeconomic characteristics to assign applicants a score and allocate land to those above particular thresholds. Thus, the analysis compares recipients and non-recipients who were just above and below a score threshold but very similar along many socioeconomic dimensions.
The children of recipients were actually more successful in climbing up the social ladder. They were more likely to experience better living standards, work in formal and high-skilled sectors, move to large cities and accumulate more years of schooling - a bundle of measures that form the nexus of modern economic life.
Main Finding #1: Land is not necessarily a trap. Actually, it can become an opportunity.
Recipients experienced improved living standards. Having been eligible to receive land during the reform increased their household wealth index in 2006 by 0.2 standard deviations relative to non-recipients. They were also more likely to earn above minimum wages and to have left agriculture. However, they were not necessarily more successful in entering the formal sector, which limited their use of financial markets. This suggests positive but modest impacts of land.
Next, I look at the children to understand whether these impacts persisted across generations. The children of recipients exhibited, on average, better living standards relative to those from non-recipients and their parents. In 2006, the household wealth index of these children was 0.3 standard deviations higher. In 2010, they were also 22 pp more likely to earn above minimum wages (mean of 17%) and 25 pp more likely to work in formal and high-skilled sectors (mean of 39%). These results are indicative of considerable upward economic mobility.
These children also exhibited higher intergenerational mobility. I use a wealth index and years of schooling in 2006 as outcome variables. Similar to Chetty et al. (2014), I rank applicants based on their outcome levels relative to other applicants with children in the 1970s-1980s birth cohorts. I then rank the children of applicants based on their outcome levels relative to other children in the sample. I characterize intergenerational mobility for recipient and non-recipient families based on the slope of rank-rank relationships.
A 10-percentile point increase in recipients and non-recipients wealth rank was associated with 4.8 and 6.0 percentile increases in their children’s relative wealth ranks, respectively (see Figure 1). Furthermore, I find that upward mobility for the children of non-recipients in the bottom quarter of the wealth and education distributions was lower than for the children of recipients, which rules out that effects are caused by worse outcomes for better-off rural families.
In summary, these findings suggest that land can expand economic opportunities for the rural poor. Providing a father with a productive asset can significantly improve a family’s well-being and help their children climb up the social ladder. They do not merely reveal persistence, but rather, show amplifying effects across generations.
Main finding #2: If you are a recipient, your children will have more chances to end up living in a city.
But, how did land help the children of recipients to enter the modern economy? In contrast to a common argument that recipient families may have been tied to the countryside, I find they were more prone to move in search for better opportunities. Recipients were 11 percentage points more likely to migrate to cities, relative to a mean of 19%. Meanwhile, their children were 22 percentage points more likely to move to urban centers, compared to a base of 39%.
I also find that the children born after the reform accumulated more years of schooling, suggesting higher investment in their education. Consistent with a setting where an asset appears to have relieved credit constraints on urban migration costs and education, notarial records show that numerous recipients formally sold their land within a few years of the reform. Presumably, even more did so through informal land markets. This suggests that land provided an opportunity, but not necessarily the expected one.
Implications for Policy
Previous findings can have broad implications for development policy. Providing land through agrarian reform often consumes significant resources. Therefore, it is important to compare the potential benefits and costs of this policy. I do so by predicting the children´s lifetime earnings and contrast them with the fiscal investment made by the government per recipient family. A simple cost-benefit analysis of the 1968 agrarian reform reveals it had of rate of return of -80%, suggesting it was most likely not cost-effective.
In a world where economic mobility is of increasing salience, this policy may be an expensive way to improve it. If the reason that recipients benefit from accessing land is to sell it to relieve credit constraints, then policy-makers can think of alternative policies that would subsidize these costs. Recent and future research could shed light on whether, for example, other asset transfers or credit incentives can be a more effective tool for raising the well-being of rural families (e.g. Bandiera et al, 2016; Banerjee et al, 2015).
Juan Sebastián Galán is a PhD candidate in Political Economy and Government at Harvard University
I examine how applicants to the 1968 Colombian agrarian reform and their children fared in life. I use ID numbers to track them in various administrative data four decades later. Moreover, I use a regression discontinuity design to estimate causal impacts. During the reform, a selection mechanism combined data from several socioeconomic characteristics to assign applicants a score and allocate land to those above particular thresholds. Thus, the analysis compares recipients and non-recipients who were just above and below a score threshold but very similar along many socioeconomic dimensions.
The children of recipients were actually more successful in climbing up the social ladder. They were more likely to experience better living standards, work in formal and high-skilled sectors, move to large cities and accumulate more years of schooling - a bundle of measures that form the nexus of modern economic life.
Main Finding #1: Land is not necessarily a trap. Actually, it can become an opportunity.
Recipients experienced improved living standards. Having been eligible to receive land during the reform increased their household wealth index in 2006 by 0.2 standard deviations relative to non-recipients. They were also more likely to earn above minimum wages and to have left agriculture. However, they were not necessarily more successful in entering the formal sector, which limited their use of financial markets. This suggests positive but modest impacts of land.
Next, I look at the children to understand whether these impacts persisted across generations. The children of recipients exhibited, on average, better living standards relative to those from non-recipients and their parents. In 2006, the household wealth index of these children was 0.3 standard deviations higher. In 2010, they were also 22 pp more likely to earn above minimum wages (mean of 17%) and 25 pp more likely to work in formal and high-skilled sectors (mean of 39%). These results are indicative of considerable upward economic mobility.
These children also exhibited higher intergenerational mobility. I use a wealth index and years of schooling in 2006 as outcome variables. Similar to Chetty et al. (2014), I rank applicants based on their outcome levels relative to other applicants with children in the 1970s-1980s birth cohorts. I then rank the children of applicants based on their outcome levels relative to other children in the sample. I characterize intergenerational mobility for recipient and non-recipient families based on the slope of rank-rank relationships.
A 10-percentile point increase in recipients and non-recipients wealth rank was associated with 4.8 and 6.0 percentile increases in their children’s relative wealth ranks, respectively (see Figure 1). Furthermore, I find that upward mobility for the children of non-recipients in the bottom quarter of the wealth and education distributions was lower than for the children of recipients, which rules out that effects are caused by worse outcomes for better-off rural families.
In summary, these findings suggest that land can expand economic opportunities for the rural poor. Providing a father with a productive asset can significantly improve a family’s well-being and help their children climb up the social ladder. They do not merely reveal persistence, but rather, show amplifying effects across generations.
Main finding #2: If you are a recipient, your children will have more chances to end up living in a city.
But, how did land help the children of recipients to enter the modern economy? In contrast to a common argument that recipient families may have been tied to the countryside, I find they were more prone to move in search for better opportunities. Recipients were 11 percentage points more likely to migrate to cities, relative to a mean of 19%. Meanwhile, their children were 22 percentage points more likely to move to urban centers, compared to a base of 39%.
I also find that the children born after the reform accumulated more years of schooling, suggesting higher investment in their education. Consistent with a setting where an asset appears to have relieved credit constraints on urban migration costs and education, notarial records show that numerous recipients formally sold their land within a few years of the reform. Presumably, even more did so through informal land markets. This suggests that land provided an opportunity, but not necessarily the expected one.
Implications for Policy
Previous findings can have broad implications for development policy. Providing land through agrarian reform often consumes significant resources. Therefore, it is important to compare the potential benefits and costs of this policy. I do so by predicting the children´s lifetime earnings and contrast them with the fiscal investment made by the government per recipient family. A simple cost-benefit analysis of the 1968 agrarian reform reveals it had of rate of return of -80%, suggesting it was most likely not cost-effective.
In a world where economic mobility is of increasing salience, this policy may be an expensive way to improve it. If the reason that recipients benefit from accessing land is to sell it to relieve credit constraints, then policy-makers can think of alternative policies that would subsidize these costs. Recent and future research could shed light on whether, for example, other asset transfers or credit incentives can be a more effective tool for raising the well-being of rural families (e.g. Bandiera et al, 2016; Banerjee et al, 2015).
Juan Sebastián Galán is a PhD candidate in Political Economy and Government at Harvard University
Join the Conversation