Note: This blog post is a little blase about the stock versus flow distinction, namely, the distinction between immigrant numbers (the stock) and net immigration levels (the flow). I didn’t belabor the distinction because I don’t think the distinction was particularly relevant to any of the points I was making, but others might disagree.
Here’s a plausible theory that has emerged from recent blog posts and comments on this site, as well as other material I’ve read, regarding the effect of immigration crackdowns on illegal immigration at the federal and state levels:
- Ceteris paribus, if a single US state cracks down on illegal immigration in that state, illegal immigrant numbers and proportions would tend to fall in that state. Most of this, however, would be to the detriment (from a restrictionist perspective; open borders advocates might call it “benefit” instead) of other US states: immigrants would move to the other states, and potential immigrants would find other states more attractive to settle in.
- If, however, all states were to crack down on illegal immigration simultaneously, or if a federally coordinated crackdown were initiated, then the net downward effect on illegal immigrant numbers would be much lower than the effect for any single state, because the immigrants are less likely to leave the country than they are to move to neighboring states. In other words, estimates of the success at a national level of strong immigration enforcement based on their “success” in individual states are likely to lead to overly optimistic (from a restrictionist viewpoint) estimates of the decline in immigration levels and/or total immigrant numbers.
My main criticism of economic determinism (which says that immigration numbers are determined by economic trends rather than immigration policies), at least in the US context, has been that states that embraced immigration crackdowns saw their illegal immigrant numbers decline. At the same time, the numbers regarding overall cross-border migration flows seem to suggest that state-level immigration crackdowns were relatively insignificant in their effects on these flows. The above offers a possible reconciliation.
My co-blogger Nathan Smith, in his book Principles of a Free Society, says something similar in the context of a border fence to what I’ve just said about crackdowns (you can download the PDF or DOC of the immigration chapter from which I quote):
Experts like Douglas S. Massey insist that all the increases in immigration enforcement since the 1980s have failed not only to stop, but even to slow illegal immigration, and even that they have increased permanent illegal migration by making the alternative of seasonal migration more difficult. I find it hard to believe that enforcement has been that ineffective; yet there is an easy way to make sense of the claim. Border enforcement increases the cost, in money and hardship, of migrating, but that cost is still small compared to the value of migrating.
If U.S. GDP per capita is $46,400, while Mexico’s is $13,500, and if we assume the economic growth rate in both countries is about the same and the discount rate is 1% more than economic growth, the benefit of living in the U.S. rather than Mexico for a typical migrant might be on the order of $1.3 million. The rewards, then, may just be too large for the border enforcement measures applied so far to make much of a difference in the incentives for migration. […]
Economically, the United States is rather homogeneous: GDP per capita differs from state to state by a factor of two or less. For a prospective illegal immigrant, one way into the country may be about as good as another; but there is a strong incentive to get in somehow. If there is a fence at one point on the border, a rational migrant will go to another part of it. If there were a fence along the whole border, it would be time to go over or under or around. A person can get over a fence with ladders, or tunnel under it, go around it by boat through the Gulf of Mexico or up the Pacific Coast, or fly over it in a small plane, or perhaps even as a human cannonball with a parachute.
Although the thesis outlined above sounds prima facie plausible, I’m ambivalent about its truth. On the one hand, the substitution effect between states is closer than the substitution effect between countries, so that does point to a plausible reason why the “success” of enforcement measures in individual states would overstate their success if adopted simultaneously by all states, or if adopted through a federally coordinated initiative in which all states participated actively. On the other hand, if the measures were adopted in all states simultaneously, this might deter potential migrants to each individual state more because they know they have fewer fallback options. If Arizona alone had strong enforcement, a person might still migrate to Arizona, knowing that he/she could always escape to California if things got tough. If, however, Arizona and California both have strong enforcement, this makes Arizona even less attractive.
These conflicting effects mean that looking at state immigration crackdowns to predict the effects of an “all-state” or federal immigration crackdown is fraught with uncertainty both ways.
PS: Just to be clear, I think that measures being “successful” in the sense of reducing illegal immigrant numbers does not make them worth enacting (obviously, or I wouldn’t be posting this on an open borders blog). In fact, the very “success” of these measures might cause an adverse economic impact. Also, I think that the morality of many strong enforcement measures is questionable, though this is not the place to voice those concerns. The questions above are more about empirical predictions, not about prescribing a particular course of action.