This was, as usual, a stimulating set of comments. I cannot respond to all of them, but I will try to respond to some and in so doing clarify my original proposal. It is apparent that a number of the commenters misunderstand my proposal. I accept responsibility for not having explained it adequately; in addition, I modify it slightly in this response, in light of the comments.
A number of comments suggest that Becker's and my proposals are anti-immigrant or anti-poor. That is incorrect. As Becker explains in his response, his proposal would facilitate immigration by unskilled workers--as would mine, had it not been for my reference to IQ, which I retract below! Both of us contemplate that our proposals would lead to greater immigration. That in turn would tend to redistribute wealth from rich to poor, because the vast majority of immigrants (other than some asylum seekers) raise their standard of living by coming to the United States. This, by the way, was surely true of the people who were able to immigrate to America in the eighteenth century solely by virtue of the institution of indentured servitude. Indentured servitude (which must not be confused with slavery) is a method of commitment that, like a mortgage, enables a person to obtain an economic advantage that he could not obtain otherwise. Unfortunately this is a point that is very difficult for people not steeped in economic thinking to grasp. But try!
One comment misdescribes my proposal as one "to sell immigration rights." The only sale component concerns immigration slots auctioned to rich people who, because of age or health, would be unlikely to be productive citizens; the auction price would compensate the rest of us for supporting them in their sickness and old age.
I realize that I created the impression that I wanted immigration officials to assess whether each individual prospective immigrant would be likely to make a net contribution to the American economy, or to American society more broadly, as a condition of permitting him to immigrate. That was error; it would be an excessively costly, perhaps indeed a completely infeasible, undertaking. What I should have said is that the government should adopt a few simple criteria, perhaps limited to age, health, and criminal history, which could in most cases be readily determined, to screen would-be immigrants, and couple that with a residency requirement for welfare benefits (see below). I should not have mentioned IQ, since as Becker points out we need additional unskilled as well as skilled workers and since it is difficult to design IQ tests that will yield comparable results for persons of different linguistic, cultural, and socio-economic background. Congestion and pollution externalities are potentially strong objections to high levels of immigration, but they should affect policy at the level of deciding whether to place some overall limit on the annual immigration rate; they cannot be used to screen individual applicants.
Employability is important, but age and health are proxies for it; and disentitling new immigrants to social services for a limited period of time (probably no more than a year or two) is a way of discouraging immigration by workers who may be young and healthy and have a clean criminal record yet who for one reason or another are not attractive to U.S. employers. The purpose of this temporary residency requirement for entitlement to social benefits is not, as one comment asserts, to curb immigration because of the welfare state; it is to discourage free-rider immigrants. To repeat, the overall effect of our proposals would be to increase the amount of immigration. Moreover, even though the present patchwork of immigration laws is inefficient, I believe that the net effects of immigration, today as in the past, legal and illegal, on American society are positive, consistent with the study by Smith and Edmonston cited in one of the comments.
Speaking of the immigration law patchwork: an excellent comment, diffidently offered by an undergraduate, asks what features of the present system of immigration rights is my proposal intended to replace? All of them? No; I said that I thought we should continue to grant asylum to victims of persecution. But I did not comment on the other grounds on which people are allowed to immigrate under existing law, such as national quotas, family reunification, lottery, and special skills. Although family reunification has obvious appeal, I cannot think of a good reason to specify immigration quotas by nation. A lottery would make sense only if the number of people who passed the screening test that I have proposed exceeded some overall ceiling on immigration derived from concern with congestion or pollution externalities; in that event, a lottery would be a cheap way of equilibrating applications to openings. A special-skills exception would be superfluous (and is costly to administer) if the screening approach were adopted, because almost everyone who had special skills would pass the test.
One comment contains an intriguing hint that might be elaborated as follows: illegal immigration, being costly, tends to filter out would-be immigrants who are either faint of heart or don't have a really strong desire to live in the United States, while letting in would-be immigrants who are daring, ingenious, and optimistic about their chances for success in the U.S., albeit who also may have a below-average commitment to legality. On balance, illegal immigrants may constitute a desirable class of immigrants. If this is correct, it supports the Bush Administration's amnesty proposal.