April 18, 2017


Why Mass Migration Is Good for Long-Term Economic Growth (Vincenzo Bove and Leandro Elia, APRIL 18, 2017, Harvard Business Review)

 In a recent study we asked the following question: Is the diversity created by mass migration a good thing for economic growth? To find out, we mobilized a large-scale data set on international migration from 1960 to 2010, using information on the nationality of the immigrants to construct indexes of birthplace diversity.

For each country at every census round, we measured its fractionalization level, the likelihood that two individuals randomly selected from the population were born in different countries. Higher degrees of fractionalization indicate more diversity. We also computed a "polarization index," or the extent to which a country's population was made up of two groups of equal size. To give some context, among the most fractionalized countries in 2010 were Kuwait, Saudi Arabia, and Singapore, whereas the least fractionalized were China, Indonesia, the Philippines, and Somalia. In the same year, the most polarized economies were Luxembourg, Singapore, and most of the nations in the Arabian Peninsula, such as Bahrain, Oman, and Saudi Arabia. The least polarized were China, Indonesia, Lesotho, and Somalia.

Because countries with higher economic growth attract higher numbers of immigrants, as well as immigrants from many different cultures, we faced a challenge in figuring out whether immigrants and diversity were causing economic growth, or were a consequence of it. Our model did not account for important issues that are difficult to observe or quantify, such as specific immigration policies; open-door policies toward immigrants are likely to correlate with both good economic performances and high levels of diversity. Excluding factors like these could lead to the wrong inference.

To circumvent some of these issues, we constructed predicted indexes of diversity using variables such as the geographic distance, colonial history, or existence of a common language between origin and destination countries. This method allowed us to create indexes of diversity based on exogenous characteristics that are uncorrelated with economic growth, as well as with other unobservable country-specific characteristics, such as the existence of particular immigration policies. In doing this, we isolated the portion of the correlation between diversity and economic growth that was due to the causal effect of diversity and removed the portion of the variability of diversity correlated with other relevant variables omitted from the model.

Our empirical findings suggest that cultural heterogeneity, measured by either fractionalization or polarization, has a discernible positive impact on the growth rate of GDP over long time periods. For, example, from 1960 to 2010, when the growth rate of fractionalization increased by 10 percentage points, the growth rate of per capita GDP increased by about 2.1 percentage points. (This is the average effect across all countries in the world.)

But we suspected that diversity might play a different role at different stages of development. Richer countries are closer to the technological frontier than poorer countries, so the adoption of new technologies should be faster in developing economies, and the labor force's skills and knowledge should increase at a faster rate. In other words, the more developed the destination country is, the less economic impact we are likely to see from migration.

To test this expectation, we split countries into subgroups of developing and developed economies, and then replicated our previous models. We found that developing economies are indeed more likely to experience a sharper increase in GDP growth rate after their populations become more diverse. Our estimates suggest that, from 1960 to 2010, a 10-percentage-point increase in the growth rate of fractionalization (or polarization) boosts per capita output by about 2.8 percentage points in developing countries. (That is 0.7 percentage points higher than the global average described above.) The same models suggest that the effect of diversity in the developed economies is much weaker. This all implies that developing economies benefit the most from diversity.

Posted by at April 18, 2017 8:47 AM