Immigration and Globalization

Frequent commenter ishmaelabroad asks for a reaction to this exchange between Andrew Leonard at Salon and George Borjas on his blog. Keeping in mind that this is a reaction to a comment on an analogy, here goes.

Borjas makes a reasonable point that journalists, more than many other professional occupations, should understand the impact of an influx of competitors with low reservation wages on the prevailing wage level. Leonard accepts this point, but then suggests his own analogy between globalization and the futility of erecting barriers to illegal immigration:

To think that one can turn back the tide of competition unleashed by the Net is a lot like thinking that in a globalized world one can ameliorate the wage impact of illegal immigration by building a border fence or by passing laws imposing strict sanctions against employers who hire illegal immigrants. The work forces of China and India and eastern Europe and of course Mexico have joined the world economy just like bloggers have joined the media universe. In both cases, technology has played a huge enabling role, and, unless the world experiences a truly massive and unprecedented energy crisis, that technologically-midwifed change is not going back in the bottle. In a globalized world, massive disparities between the living standards of individual nations will create more pressure than ever before for some kind of equalization,
whether that means workers finding their way from the developing to the developed world, or capital headed in the other direction.

I think he's right insofar as he notes that there will be "some kind of equalization." I think that he's concluding prematurely that immigration will play, needs to play, or should play a large role in that equalization.

The current Wikipedia entry for globalization points to economic integration in four types of markets: goods and services, capital, technology, and labor. I'm an economist with libertarian views, so I am all for greater opportunities to exchange the first three across national borders, provided that property rights are protected. I'll further assert that if those opportunities were enhanced, there could be less pressure for labor to move across national borders. (See this earlier post.)

The difference between labor and the other three markets is that a laborer is a person, and people have rights unrelated to their economic lives. As I've said before, I don't believe in guest-worker programs that authorize a second-class citizenry. So if people have immigrated, they have the right to vote, and their votes may move policies away from those that would otherwise prevail. As another example, they have the right to make claims against a social welfare system, and their claims may outweigh their contributions through the tax system. I'm sure others could come up with more examples.

So immigration, as distinct from other forms of globalization, imposes a distribution of costs and benefits on the rest of society that is intermediated through the political system. A citizen of the U.S., even one who embraces the other three types of globalization, could reasonably conclude that the costs of immigration imposed through the political system outweigh the economic benefits. This is particularly true if the benefits to the immigrants themselves are excluded. (See the distinction between being generous and being fair to would-be immigrants in this recent post.)

There are obviously many citizens of this country who have come to a different conclusion about these political costs net of economic benefits, and so they are naturally advocating for a more permissive immigration policy. However, if a citizen had come to that conclusion, then the next question is whether the costs of deterring illegal immigration are sufficiently smaller than these political costs net of economic benefits. Leonard (if he is asking this question) seems to believe that the ability to deter (e.g., through a border fence or employer sanctions) is so low or that the costs of deterrence are so high that the answer to this question would almost always be "no." I'm not so sure.