Analytically as well as politically, migration is a more complex phenomenon than international trade. The latter involves two-way flows of goods and services that allow the real aggregate income to grow in each country involved. Therefore, the arguments for protection against imports typically rest on the detrimental effects on the income distribution or the existence of prior distortions in other markets.
Migration involves one-way flows, which calls for a distinction to be made between the "source" country and the "destination" country. In addition, one must separately consider the welfare of the migrant and also decide whether his or her welfare constitutes a part of the welfare of the source country or destination country. In turn, since the migrant physically resides in the country of destination and cannot be excluded from consuming public goods, society is confronted with the issue of fiscal burden imposed by the migrant on the native population.
The literature on the brain drain saw explosive growth in the 1970s and has been elegantly surveyed by Bhagwati and Rodriguez. This literature continues to provide a useful framework for analyzing many of the current issues in the migration debate.1 Therefore, my discussion below draws heavily on that literature. I begin with the introduction of the basic issues with the help of the conventional one-good, two-factor model.
The Simple Welfare Economics of Migration in a One-Good Model
The one-sector, two-factor model offers us the simplest framework within which we can introduce the key issues that migration raises. The model has the [End Page 229] obvious limitation that it does not admit international trade. As such, it is to be viewed as only the starting point for the introduction of the basic issues.
With this qualification, I call the two factors capital (K) and labor (L) and divide the world into two countries: the capital-abundant North and labor-abundant South. In figure 1, the horizontal axis OSONrepresents the total supply of labor worldwide. I measure labor employed in the South to the right from OS and that employed in the North to the left from ON. Using the only good in the economy as the numeraire, MPLS and MPLN represent the marginal product curves of labor in the South and the North, respectively.
Suppose now that point R gives the initial allocation of labor between the North and the South. It is then immediately clear that the initial northern wage, WN, exceeds the initial southern wage, WS. The area under each country's marginal product curve up to the labor-allocation point gives its total income. The rectangle formed by the height of the wage and the length of labor allocation determines the wage bill with the rest of the income going to capital.
Next, allow just one worker to migrate. Given the higher wage in the North, this migration will be from the South to the North. Assume for simplicity that the migrant does not own any capital either before or after migration. Under such circumstances, since the volume of migration is infinitesimally small by assumption, neither the welfare of those left behind in the source country nor that of the native population in the destination country is altered. Only the migrant benefits, who now receives the higher northern wage. Before emigration from the South, he added WS to the GDP and received WS; after immigration into the North, he contributes WN to the income and receives WN.
If migration is finite, however, the welfare of those left behind in the source country and of the native population in the destination country is affected. This is shown by the movement of RR' workers from the South to the North in figure 1. Continuing to assume that migrants do not own any capital either before or after migration, this change leads to a loss of the triangular area FGH by those...