Forget the dubious constructs of race and IQ for a moment.
Suppose there really was a genetically distinct race of white-skinned people inhabiting a large, hypothetical island in the Pacific Ocean; that IQ really could be reliably measured; and that we knew, for a fact, that while the measured IQs of Caucasians, blacks, Hispanics, Jews, Native Americans, and all other identity groups in the United States had converged to an identical average, members of this one hypothetical race had IQ scores that measured 5 points lower on average. Additionally, suppose that the average IQ of nations as a whole had been indisputably linked to educational attainment and GDP. Would it be legitimate to bar that lower IQ group from immigrating?
To me, doing so would be wrongheaded.
Even setting aside my strong preference for policies rooted in individualism and the dangerous, inherently problematic nature of singling out a specific racial group for disparate treatment, barring the hypothetical low IQ people would imply that intelligence determines worth, and that our project as a nation is intimately tied to constantly maximizing material wealth.
I wouldn't go so far as to say that recruiting human beings with impressive skills is illegitimate. In fact, I think it is prudent, and I'm glad that lots of talented scientists, athletes, artists, and programmers want to come here. More, please. I'm glad that lots of farm workers and janitors want to immigrate too. I recognize that the economic contributions of the two groups are different, but I don't conclude that the low skill immigrants are less worthy of citizenship or less valuable citizens. Are they kind? Honest? Wise? Fun?
The debate (or to be more accurate, the lack thereof) triggered a couple of tangential thoughts. The first is that existing immigration policy in many developed countries already has an IQ filter. Australia and Canada’s skilled immigration systems are often pointed to as being among the most successful; so successful in fact that they are the two OECD countries where second generation immigrants outperform students with native parents in the PISA tests - see here, here and here. A large part of improvements in Swiss PISA test scores was attributed to immigration changes in the 1990s.
The immigration reforms that triggered the Heritage Foundation’s report also contain a skills-based component, including a points systems like that used in Australia and Canada. The United States is effectively implementing some of Richwine’s recommendations. (Since I first drafted this post, I see that Ed Realist has pointed out how some of Richwine’s ideas were doing just fine until the storm around the Heritage report.)
Another thought is that IQ-barriers are pervasive within countries. Tests for entry into college or university (such as the SAT) are highly correlated with IQ scores. IQ test results predict success in universities and awarding of scholarships. Many jobs have IQ-testing as part of the application process, particularly in police and fire departments (which often makes them the subject of litigation about exclusion of minorities). Intelligence is also a filter for who we are friends with and who we marry. Being of low intelligence has significant costs.
We can have a high level of confidence that the difference in IQ scores within developed countries has a genetic component. Estimates of the heritability of IQ from twin and adoption studies are robust. This means that within many countries we already actively exercise discrimination based on genetic factors, on both an institutional and personal level.