Justin Wolfers had a nice piece in the Upshot about
new work on how growing up in a bad neighborhood has long-term negative consequences for kids. The key point of the new work is that the benefits of moving from bad neighborhoods may be particularly high for kids whose parents won’t voluntarily move, but only move because their public housing is demolished.
When many people think of bad neighborhoods, they think of housing projects in urban cities, or slums. But as someone who grew up on the back of a farm in a very small town (and then was lucky enough to move), I tend to think of the wop wops. So Justin’s post reminded me of a paper I had heard about but never tracked down before. “Habit formation and the misallocation of labor: evidence from forced migrations” makes the point that too few farmers leave their farms, and that “farmers could increase their long-term income by 70–80% by leaving agriculture.”
Forced migration of Finns
The setting here is post-World War II Finland. At the end of the war, Finland ceded its eastern parts to the Soviet Union and had to resettle the entire population living in those areas to other parts of Finland. In total 11 percent of the population were forced to migrate. Displaced farmers were given land and assistance to establish new farms in areas that had similar soil and climate as the origin regions, and former neighbors were resettled close to one another to preserve social networks. This resettlement was completed in 1948, and then they were free to sell their new land, and move anywhere else in the country like the rest of the Finns.
The authors use data from the 1950 census, which asked retrospective questions on place of residence and occupational status two months before WWII. They then link this data to the 1970 census and to 1971 income tax records. They take men who were 14 to 32 years old at the time of the war and working in agriculture, and then can look at outcomes when they are 46 to 64 years old.
They have three different approaches to estimation, and discuss the possible bias in each, with the idea being that they act in different directions, and can therefore provide bounds on the likely treatment effect. The first is to do simple OLS regression comparing those who were displaced to those who were not, with and without controls for pre-war characteristics. The second approach is a spatial discontinuity approach, comparing displaced farmers on one side of the new border to non-displaced farmers on the other. Thirdly, they compare the displaced farmers to farmers in the new locations they were relocated to.
Figure 1 below provides a striking graphical illustration of the spatial discontinuity approach. You see that men who were just on the displaced side of the border have much higher incomes in 1971 than those who were just on the non-displaced side (top panel), and that they are more likely to have stopped farming and moved into non-agricultural employment (bottom panel). The gain in annual income is estimated to be between 11 and 28 percent using the different approaches, and, under the assumption that the main gains come from leaving farming, are up to 70-80% for those who leave agriculture.
Why is there this effect and how relevant is it for developing countries?
The authors examine several reasons for the impact, and come down in favor of what they call a “habit formation” reason, in which farmers “consume” their farms and have a strong attachment to this particular place. When they are forced to move to an otherwise similar farm, they lose this consumption capital, and so then require a smaller compensating income differential for leaving agriculture – as one said “ since I cannot go back to my old land, it does not matter where I live”. Of course, the issue if this is the reason is how to value this loss in utility to the farmers from moving, and whether the gain in income is enough to offset it. This is not something the authors discuss, but is of course an issue if we want to move from “farmers earn more and are more productive when they leave their farms” to “farmers would be better off in the long-run if pushed off their farms and compensated for the cost of the farms but not for the utility of the farm habit”. This brings us back to the same policy debate as the new US findings on moving from housing projects shows – if the biggest gains are for those who won’t voluntarily move, what should policy do?
You might also ask whether Finland’s experience has any relevance for developing countries today? The authors note that at the time of the re-locations, Finland had one-half the population working in agriculture, had per-capital GDP comparable to that in current North Africa, and had only had independence for two decades after having gone through a short but brutal civil war in 1918.
When many people think of bad neighborhoods, they think of housing projects in urban cities, or slums. But as someone who grew up on the back of a farm in a very small town (and then was lucky enough to move), I tend to think of the wop wops. So Justin’s post reminded me of a paper I had heard about but never tracked down before. “Habit formation and the misallocation of labor: evidence from forced migrations” makes the point that too few farmers leave their farms, and that “farmers could increase their long-term income by 70–80% by leaving agriculture.”
Forced migration of Finns
The setting here is post-World War II Finland. At the end of the war, Finland ceded its eastern parts to the Soviet Union and had to resettle the entire population living in those areas to other parts of Finland. In total 11 percent of the population were forced to migrate. Displaced farmers were given land and assistance to establish new farms in areas that had similar soil and climate as the origin regions, and former neighbors were resettled close to one another to preserve social networks. This resettlement was completed in 1948, and then they were free to sell their new land, and move anywhere else in the country like the rest of the Finns.
The authors use data from the 1950 census, which asked retrospective questions on place of residence and occupational status two months before WWII. They then link this data to the 1970 census and to 1971 income tax records. They take men who were 14 to 32 years old at the time of the war and working in agriculture, and then can look at outcomes when they are 46 to 64 years old.
They have three different approaches to estimation, and discuss the possible bias in each, with the idea being that they act in different directions, and can therefore provide bounds on the likely treatment effect. The first is to do simple OLS regression comparing those who were displaced to those who were not, with and without controls for pre-war characteristics. The second approach is a spatial discontinuity approach, comparing displaced farmers on one side of the new border to non-displaced farmers on the other. Thirdly, they compare the displaced farmers to farmers in the new locations they were relocated to.
Figure 1 below provides a striking graphical illustration of the spatial discontinuity approach. You see that men who were just on the displaced side of the border have much higher incomes in 1971 than those who were just on the non-displaced side (top panel), and that they are more likely to have stopped farming and moved into non-agricultural employment (bottom panel). The gain in annual income is estimated to be between 11 and 28 percent using the different approaches, and, under the assumption that the main gains come from leaving farming, are up to 70-80% for those who leave agriculture.
Why is there this effect and how relevant is it for developing countries?
The authors examine several reasons for the impact, and come down in favor of what they call a “habit formation” reason, in which farmers “consume” their farms and have a strong attachment to this particular place. When they are forced to move to an otherwise similar farm, they lose this consumption capital, and so then require a smaller compensating income differential for leaving agriculture – as one said “ since I cannot go back to my old land, it does not matter where I live”. Of course, the issue if this is the reason is how to value this loss in utility to the farmers from moving, and whether the gain in income is enough to offset it. This is not something the authors discuss, but is of course an issue if we want to move from “farmers earn more and are more productive when they leave their farms” to “farmers would be better off in the long-run if pushed off their farms and compensated for the cost of the farms but not for the utility of the farm habit”. This brings us back to the same policy debate as the new US findings on moving from housing projects shows – if the biggest gains are for those who won’t voluntarily move, what should policy do?
You might also ask whether Finland’s experience has any relevance for developing countries today? The authors note that at the time of the re-locations, Finland had one-half the population working in agriculture, had per-capital GDP comparable to that in current North Africa, and had only had independence for two decades after having gone through a short but brutal civil war in 1918.
Join the Conversation