Does immigration account for relatively stagnant median income?

It's a common claim that increasing immigration accounts for the slowdown in U.S. median income growth.  No, I don't mean the claim that immigrants lower wages.  Even according to George Borjas, that effect isn't large enough to much budge the median and most of it is concentrated on workers without a high school degree.  Rather I am referring to the mere addition of immigrants to the rolls, which itself shifts the median of the distribution.

I don't see that the immigration effect explains much of the data.

1. Look at Dew-Becker and Gordon, p.62.  They consider panel data for hourly compensation, and they still find a stagnant median.  Also see p.78, where they summarize their results that the stagnation extends upward through the 90th percentile of the income distribution.  The U.S. didn't come close to taking in enough immigrants to produce that result.

2. Look at Lane Kenworthy's figures, p.41.  Median income growth for the category "White, non-Hispanic" is only barely better than for the general population.

3. This paper measures the effect of immigration on the Gini coefficient and finds that a given year's worth of immigration has a peak effect on the Gini at four years' time and the effect disappears altogether within six years' time.

4. Researchers who work in this field are well aware of the phenomenon of immigration and generally they do not use it to dismiss the issue of declining median income growth.

I've heard from one reader that the eBook is now available in France, from another that it is in Australia, but Canada will take a little more time.


Comments for this post are closed