Incentives to be accurate about what forms of enforcement work

Does “self-deportation” (also known as attrition through enforcement) work, if the goal is to cut down on (unauthorized) immigration? For those who want to cut such immigration down, this is a key question to answer. Restrictionist groups such as NumbersUSA, CIS and VDARE have generally taken the stance that attrition through enforcement is a workable strategy (see here, here, and here). The Immigration Policy Center, on the other hand, has argued otherwise.

I laid out hints of my own position on the matter here and here. Briefly put, yes, attrition through enforcement does work, but it also imposes costs and collateral damage on citizens/residents. After some reflection, though, it occurred to me that even without knowing much about the details, one would be led to suspect this.

I begin with the assumption that restrictionists actually want to achieve what they say they want to achieve — that there should be less future immigration (legal and illegal) and that most current illegal immigrants should leave the country. It is, of course, possible that people like Mark Krikorian don’t really want their stated agenda to be implemented, because it would make their jobs as restrictionist advocates superfluous. I find this unlikely, because it seems to me that restrictionists are generally extremely talented people who can find employment in a number of areas that involve the skillful and convincing presentation of a weak case — as lawyers, political aides, or lobbyists — and that their sticking to the relatively less lucrative area of immigration restrictionism is at least partly explained by genuine conviction.

If restrictionists sincerely want what they say they want, this means that their incentives are very much aligned towards determining how to achieve it. Thus, when they evaluate policies like attrition through enforcement and border security, they’re probably the best judges of the effectiveness of these policies.

On the other hand, open borders advocates and immigrant rights groups don’t want immigration enforcement (for the most part) to succeed — or at any rate, not without corresponding liberalizations of immigration policy. Their incentives, therefore, are not well-aligned towards an unbiased evaluation of the success or failure of these policies. To an extent, they face conflicting incentives. If they see a particular enforcement measure that is cheaper and more “effective” at achieving restrictionist ends, do they acknowledge this — and end up providing free service to the restrictionist cause — or disingenuously deny it, coming up with reasons against? I suspect that this is a very real dilemma, and many immigrant rights groups resolve it by fooling themselves into thinking that the effective methods of enforcement are ineffective. It’s similar to how drug use legalization people might feel at being asked to evaluate the effectiveness of drug raids at imprisoning drug users.

For these reasons, you’d see why one might have a strong prior that restrictionists would be more likely to have figured out the best methods of achieving their goals than open borders advocates and immigrant rights groups.

Nonetheless, there is another side to the picture: cost and collateral damage. Here, the incentives for restrictionists are bad. Without knowing the details, one would be led to suspect that restrictionists would systematically underestimate the cost and collateral damage of their proposed methods to citizens and natives. The reason is that hardcore restrictionists have a much higher preference for getting rid of illegal immigrants than the general public, so in order to sell their message to a relatively less (but still highly) restrictionist public, they’d need to underplay the collateral costs of their policies. On the other hand, immigrant rights and open borders advocacy groups would be quick to point out the collateral damage to citizens, and run sympathetic stories of citizens, authorized immigrants, and tourists on valid visas getting detained and harmed by over-enthusiastic enforcement.

My own take on this matter is that the best response for a hardcore open borders advocate to discussions of what forms of enforcement work is to first acknowledge that his/her own complete disagreement with the end goal makes him/her an extremely bad person to consult regarding the means to be used towards that end. One can think of this almost in the sense of a “conscientious objector” who refuses to participate in a system that his/her conscience goes against, or a person who refuses to testify against himself/herself in court. If it is necessary to offer an opinion, there are two alternatives. The first is to be genuinely honest about what forms of enforcement work, but argue that something “working” does not mean that you endorse it. The other alternative is to deliberately dissemble about what forms of enforcement work. While the former is more intellectually honest, I suspect that one can make a case for the latter from a consequentialist point of view, at least in the rare cases that one’s opinion could actually influence or shape immigration enforcement. At any rate, however, the trade-off between intellectual integrity and a conscience that forbids lending support to a policy one perceives as evil should be undertaken consciously rather than out of a reflexive desire to disagree with restrictionists.

UPDATE (March 31, 2014): When I wrote this post, I was broadly of the view that interior enforcement is more effective than border enforcement. However, this blog post by Alex Nowrasteh cites a literature sumary by the Council of Foreign Relations that suggests the opposite. Assuming this summary is correct, restrictionists’ focus on interior enforcement seems puzzling, and the broad claim of this post might well be misguided or false.

10 thoughts on “Incentives to be accurate about what forms of enforcement work”

    1. I only endorse this for situations where one expects one’s voice to be influential. Unfortunately, Open Borders is far from influential in that sense, so I don’t expect that any of the people here would have reasons to lie. Also, if and when we do become widely influential, figuring out how to carry out mass deportations effectively would not be on the agenda. So we have little reason to lie or dissemble on that front.

      What do you think are some areas where open borders advocates have a strong incentive to lie or dissemble (assuming that they are genuine in their belief regarding the case for open borders)? Not to say that they always do lie or dissemble, simply that their incentives are skewed that way.

      1. Well, one of the most egregious examples is “Republicans will benefit from amnesty’s effects on the electorate” (offered usually by people who are explicitly rooting for Republican defeat, and essentially never by Republican political analysts who aren’t eager for amnesty on other grounds).

        But your question is a bit strange: open borders advocates have incentives to lie about any issue that affects the desirability of support or opposition to more open borders. Listing them out individually would just recap everything relevant to immigration.

        And explicit lying isn’t very necessary, just selective fact-checking and communication.

        Much of the time the immediate cause is the citation or elaboration upon a sensational factoid from a professional advocate. The claim is taken at face value because it serves the cause, without checking, and is repeated, while sensational claims with anti-immigration implications are not repeated except in arguing against them.

        In other cases people make an analysis, but they re-examine and adjust their analysis to ‘fix’ anti-immigration errors or issues, but don’t bother to fix pro-immigration errors.

        People who disbelieve a false claim may refrain from undermining an ally and just fail to mention their inconvenient belief, and someone who changes his mind may not retract false claims. This works very well with a division of labor:

        if there are n false claims supporting cause X, as long as at least 1/n advocates wind up believing each lie through error or bias, then you can assign the task of writing about each claim to someone who believes the lie, and use every available lie in the public square without anyone thinking of themselves as dishonest.

        The Open Borders blog is much better on average than the typical political site, but even here, my experience seeing and independently checking many testable claims on the blog is that the empirical errors strongly tend to favor the pro-open borders case in the dispute at issue.

        1. My original question was too open-ended, but I can be clearer. Here are two questions:

          1. What are the areas where open borders advocates (and/or restrictionists) have incentives to lie that are largely independent of the case for or against open borders?
          2. What are the areas where the incentives to tell the truth are very different for open borders advocates and restrictionists, and this judgment of differing incentives can be easily made while staying neutral on the question of open borders as well as on the area at hand?

          The example of the comparative evaluation of different enforcement policies would be is an example that meets both criteria: it is largely (not not completely) independent of the case for open borders, and the incentives to tell the truth are asymmetric between open borders advocates and restrictionists, and one can note this asymmetry without taking a stand either on open borders or on enforcement policies.

          Most examples won’t fit either criteria. For instance, on the crime question, the crime question is entangled with the case for open borders, and both open borders advocates and restrictionists have symmetric incentives to exaggerate on their respective sides.

          The Republican electoral strategy example fits criterion (1), but not criterion (2) — there is no a priori reason why one side should have incentives to lie while the other side should have incentives to tell the truth — in order to figure out which side has more incentives to lie, you’d first have to figure out who is right. The folks at VDARE trying to make inroads into the Republican Party have an incentive to exaggerate the “natural Democrat” nature of immigrants, while pro-immigration Republicans would like to call immigrants (potential) “natural Republicans” — so without actually getting into the controversial debate of who’s more right based on empirics and counterfactuals, we’d have no reason to believe that one side is right. (you might remember we discussed this here).

          Most of the examples that would fit (1) and (2) would be examples similar to my enforcement strategy example in that they are about comparative evaluations of strategies for the means to achieve ends that one side desires and the other side doesn’t. Or at least, that’s how it appears to me.

        2. Quick addendum to my previous comment: you’re right that if a Democratic Party strategist offered Republicans advice on the matter, one should suspect incentives to lie. However, my focus is on open borders advocates, who, qua open borders advocates, don’t have that sort of incentive.

            1. Symmetrically, restrictionists have an incentive into tricking Republicans into more opposing open borders (incidentally, a website called claims exactly such a strategy originating from a John Tanton memo — — a claim I haven’t yet had time to scrutinize). There doesn’t seem to be anything asymmetric about the open borders advocates’ incentives in the matter relative to restrictionists’ incentives.

              On the other hand, there is obviously plenty of asymmetry between the Democratic political strategist’s incentives and those of the Republican politicial strategist, in terms of the fact that the former wants the Republican party to fail, while the latter wants the Republican party to succeed.

              1. Yes, re all advocates having incentives to exaggerate their case.

                One asymmetry is that lots of pro-immigration Republicans say that amnesty would be electorally devastating for the GOP, but not the other way around. Thus all the GOP proposals of guest worker programs without citizenship.

                Also, the asymmetry of evidence mustered by the two sides is striking: those who think the GOP would suffer have the direct empirical evidence of voting records, surveys of political beliefs, demographic records, and the electoral responses to the Reagan amnesty and the proposed Bush and McCain amnesties.

                The “immigration will help Republicans” folks mostly offer data-free qualitative speculations to meet the burden of proof in predicting a huge unprecedented change in Hispanic-American electoral behavior. Karl Rove’s “ownership society” idea neglected confounders, and never offered the necessary magnitude of swing to get a net gain for Republicans: it would be an instance of losing money on every sale but making up for it in volume.

  1. BK raises an interesting point about open borders advocates having an incentive to overestimate the political benefits to Republicans of embracing freer immigration.

    But I would make a distinction between LYING and BIAS. I might have many reasons to be biased in favor of thinking Republicans would benefit from embracing freer immigration. For one thing, I like to think well of my countrymen. It’s unpleasant to think that they’re obsessed with expelling poor immigrants. There’s also a sort of availability bias. I probably have a lot more open-borders-supporting friends than most people do, because of my GMU and development backgrounds, and also because I’ve persuaded some people. Even people whom I haven’t persuaded completely, I’ve sometimes influenced in my direction. Probably some of those who disagree with me most strongly have just learned to keep quiet. I’m more likely to make friends with people I meet who share my leanings on this issue, and more likely to distance myself from strong restrictionists. Probably this leads to availability bias.

    I try to compensate for this, and I think it’s true at this particular moment in time that Republicans would benefit from supporting freer migration. Still, in general, while I think I’m pretty scrupulous about not lying, my strategic advice to politicians would probably be vitiated by a considerable amount of bias.

  2. This is a bit late to the party, but I think the case for honesty about results is stronger than is stated in the post. Consider this is analogous to a prisoner’s dilemma. If you lie, you get a result you prefer while preventing the opposition from getting such a result. In a one-time-off event the incentives are always to lie or cheat in the prisoner’s dilemma. But we aren’t in a one-time event. We are in a long conversation about the benefits and detriments to open borders and that means reputation counts.

    The proper approach in this case might be something along the lines of “OK yes method A would be the best for decreasing immigration if that is your goal, but the costs will be x, y, z nonetheless.” What this does is establish a reputation for honesty which can then be parlayed into higher trust of other statements made by open borders advocates. The more we “nobly” lie or allow biases to influence our predictions (the latter of which isn’t as bad for reputation, but is still less than ideal), the less credible our cause seems overall. Thus the cause for honesty even when we have an immediate incentive to dissemble I think is strong.

Leave a Reply