Why I believe David Card’s results on immigration and wages

As MR readers will know, a famous David Card paper shows that the presence of many immigrants in a city does not much lower wages, if at all.

The obvious rejoinder is that cities with growing wages might attract more immigrants.  The new immigrants will cause local wages to fall back down, resulting in a lack of regional correlation between wages and immigration.  Without the immigrants maybe those locales would have had higher wages.  So what does the Card paper really show?

Keep three points in mind:

1. The skeptical story does not consider all possible adjustments.  Had there been no new immigrants in the growing cities, American workers would have moved in to take advantage of the higher wages, thereby pushing wages back down again.  The U.S. has the most mobile labor supply of any developed country.  So the net wage depression effect of new immigrants — taking all supply adjustments into account — still would be zero or very small for many places.

Note that this point, while it responds to Card skeptics, also creates some problems for the Card paper.  It suggests that the labor market is defined at the level of the nation as a whole rather than the city or region.  It is then no surprise — no matter what your view of immigration — if local labor conditions do not much correlate with local wage levels. 

But invoking this point about the national scope of the labor market also blunts fears about how immigrants depress wages.  Suddenly the question is not how many Mexicans are pouring into Texas, California, and Arizona.  The question is rather how many Mexicans are pouring into the United States.  But the larger the relevant market size, the better a job we will do absorbing immigrants.  And the smaller the effect on domestic wages we should expect.  To run a contrasting thought experiment, just try putting them all in Geneva, or better yet Soglio.

2. Have I mentioned capital mobility?  Foreign capital flows into the U.S. all the time.  Even when the Chinese buy T-bills, this frees up domestic private capital to put more people to work.  Having more immigrants encourages more capital to flow in.  If both labor and capital enter the country, there is no reason to expect immigration to lower domestic wages.

3. Are rising wage levels the key factor in drawing Mexicans to a region?  El Paso appears to be one counterexample.  In many cases proximity to the border and clustering effects seem to be more important.  In that case the Card test has less of a problem with endogeneity.  Still, I do not know the formal evidence on this point, so in the comments please pass along your expertise…