Monday, June 18, 2007

Economics and Immigration

(Cross-posted from Vox Nova)

In the Vox Nova comboxes on a recent post on immigration, I argued that the negative wage effects on low-skilled workers from an influx of immigrants were limited. Jonathan took issue with this claim, referring to the work of George Borjas. His point warrants an answer, and empirical analysis trumps mere assertion (in the reality-based community at least!).

Let's address the big picture first. There is little doubt that immigration increases overall GDP per capita, or the overall pie, through its direct effect on the growth in working age population. This allows for higher growth, lower inflation, favorable fiscal outturns, and less problems with longer-term sustainability arising from pressure on aging-related expenditure (especially since immigrants have higher fertility rates). It is by no accident that the countries most open to immigration (US, UK, Ireland, Sweden etc.) have grown most strongly over the past decade while immigration-phobic countries have lagged. As an example, the UK Treasury has boosted its long-term growth rate by 1/4 percent as a direct result of higher immigration inflows. This is by no means insignificant!

So far so good. But the argument usually revolves around the losers from immigration, and there are losers. If there is an influx of workers of a certain skill type, then the increased supply of labor can be expected to depress the wages of native-born workers of a similar skill type. Nobody disputes this basic economic logic. What people dispute is the relative magnitude of the effect.

One of the leading labor economists researching this topic is David Card of UC Berkeley. Card shows quite convincingly that local labor market outcomes of low-skilled native-born workers are barely affected by changing labor supply. Card also shows that, despite large-scale immigration in intervening period, the wage gap between dropouts and high school graduates is virtually unchanged from 1980. In the New York Times, Eduardo Porter provides an illustrative example. Many immigration opponents look at the experience of California, where the wages of high-school dropouts fell by 17 percent from 1980-2004. The effects of immigration? Well, no. The wages of high-school dropouts in Ohio fell by 31 percent over this same period, which did not have an influx of "illegal" immigration. Clearly, something else is going on here, a long-term pattern of increasing inequality that is worthy of further debate, but is not really relevant in the immigration context.

The sanguine findings of Card were disputed by others, notably Borjas and Katz. These researchers estimated the "illegal" Mexican immigrants between 1980 and 2000 lowered the wages of high school dropouts in the US by 8.2 percent, a not insignificant number. For a start, we need to realize that this result is an outlier, and indeed, Borjas and Katz refer to Card's results as the conventional wisdom. But there are a number of methodological problems with this work. First, as noted by Brad deLong, these coefficients are imprecisely estimated (large standard errors), in the sense that there is a one-sixth chance that the effect is actually positive. More pertinently, Borjas and Katz failed to adequately account for the fact that businesses react to increased labor by increasing investment. Indeed, certain businesses would probably not even exist without cheap immigrant labor. Taking these effects into account knocks the Borjas-Katz number down to 3.6 percent.

Overall, then, the consensus is that the effect of low-skilled immigration on native wages at the low end of the market is negligible, and declining as Americans become better educated and number of high-school dropouts continues to fall. As David Card says there is a "surprisingly weak relationship between immigration and less-skilled wages." A consensus figure lies closer to 1 percent, with Borjas-Katz providing the upper bound of the range of estimates.

How should we, as Catholics, analyze this debate? We need to look at the common good, and that means the global common good, not just the welfare of those persons who happen to possess US passports. Clearly, the Mexican immigrant benefits by increasing his and his family's economic opportunity. The US as a whole also benefits. There are some losers at the lower end of the wage distribution, but the effects are small. This does not mean they should be ignored, as that would violate the preferential option for the poor. But stamping down on immigration will do little to help them, while causing a great deal of suffering elsewhere. Subsidized training and other active labor market policies (like in Denmark) might be an apt response. Finally, something we do not hear much about in the debate, is the economic effect on Mexico. Surely, if there is a movement of labor between Mexico and the US, any downward wage pressure in the US will be accompanied by upward wage pressure in Mexico? Low-skilled Mexican workers will be made better off. As Catholics, their welfare matters too.

1 comment:

Stuart Buck said...

There are a couple more recent studies that you ought to take into account. See number 5 here: