The migration response to increasing temperatures

Cristina Cattaneo and Giovanni Peri

Journal of Development Economics, Volume 122 (2016), Pages 127-146, 

https://doi.org/10.1016/j.jdeveco.2016.05.004 

Review

This article examines the long-run effect of temperature changes on emigration and rural-to-urban migration in poor and middle-income countries. 

The empirical analysis is based on a reduced-form equation where emigration from both poor and middle-income countries is a function of temperature, and weather variables are interacted with a dummy variable taking a value of 1 if a country is poor. The authors use data on average temperature and on international migration and urbanization for 115 countries (30 low-income countries and 85 middle-income countries) between 1960 and 2000. The data shows that: 

  • The average ten-year net emigration rate is higher in middle-income countries (4.2 percent) than in low-income countries (1.8 percent). While emigration rates are stable in middle-income countries, there is more variation and a larger proportion of emigration rates that increase over time, relative to those decreasing. 
  • The average urbanization rate is much higher in middle-income countries (42 percent) than in low-income countries (19 percent), but both are well below the average urbanization rate in high-income countries (around 75 percent).  
  • Agriculture accounts for a much higher share of value-added production in low-income countries (35 percent) compared to middle-income countries (16 percent). 
  • Between 1960 and 2000, temperatures increased in most middle-income and low-income countries. 

Main results: 

  • In middle-income countries, higher temperatures increase urbanization rates and international migration. In a middle-income country with an average yearly temperature of 22 degrees Celsius, a one degree increase in temperature (about a 5 percent increase) would increase international migration rates by 20 percent and increase urbanization rates by 4 percentage points. 
  • In low-income countries, higher temperatures reduce urbanization rates and international migration, consistent with severe liquidity constraints. In a poor country with an average yearly temperature of 22 degrees Celsius, a one degree increase in temperature (about a 5 percent increase) would decrease international migration rates by 80 percent and decrease urbanization rates by 4 percentage points. This effect is strongest in low-income countries that are highly dependent on agriculture: a one degree increase in temperature (about a 5 percent increase) would decrease international migration rates by an additional 60 percent. 
  • Increasing temperatures may decrease overall emigration to OECD countries. Rising temperatures increase emigration rates from middle-income countries only to close and non-OECD destinations but have no significant effect on emigration to OECD countries. Low-income countries experience a decrease in emigration rates to any country.  
  • In middle-income countries, long-run temperature increase promotes a structural transition away from agriculture towards more urban and productive sectors, leading to growth in average incomes. Increases in temperature significantly decrease the agricultural share of GDP for middle-income countries. The mobility of workers into cities (with higher productivity potential) and the out-migration of poor rural workers has a positive effect on GDP per capita. 
  • In low-income countries, increases in temperature slow structural transformation and reduce income. Long run temperature increase slows structural transformation in low-income countries since poor rural workers become less likely to move to cities or abroad. Moreover, higher temperatures have a negative effect on GDP per capita. 

Higher temperatures affect migration by lowering agricultural productivity both in poor and middle-income countries. In middle-income countries, lower agricultural productivity increases incentives to migrate either to urban areas or abroad, which in turn leads to a further reduction in agricultural value-added. In low-income countries, lower agricultural productivity amplifies liquidity constraints and prevents people from migrating, and consequently lower agricultural productivity does not trigger a structural change from rural to urban economies. The authors conclude that in middle-income countries, migration provides an important adjustment to global warming, potentially contributing to structural change and even increasing income per worker, but this adjustment mechanism does not seem to work in poor economies.