A naive thinker about open borders might think of it as simply a continuum of possibilities: we can open the borders a bit, or a lot. This distinction is discussed on our moderate versus radical open borders page.

The idea of keyhole solutions has added more dimensions (in a very literal mathematical sense, in addition to the more metaphorical one) to the discussion. Rather than thinking simply of “how much more open should borders be?” the question shifts to “what sort of policy combinations can allow for us to get the most benefit out of migration, in the ways we care about?” The term “keyhole solutions” has come to represent the general idea of exploring a larger space of possibilities with respect to how migration can be expanded.

The purist in me isn’t too happy about this, because “keyhole solutions” as I believe the term originated had a more narrow meaning: to refer to narrow, targeted solutions that address the particular (real, perceived, or predicted) problem at the intersection of migration and whatever other domain is being considered, while trying to interfere as little as possible with the rest of the universe. But meaning is imbued by usage, and I’m okay with the meaning expanding and getting more fuzzy. In this post, however, I discuss some important distinctions between different approaches to “compromise” on open borders policy. There are a few additional subtleties that I’ll deliberately refrain from here, thereby meaning that my post is not reflective on my full thinking on the topic. I’m making that trade-off to keep the post simple.

A simple illustration of the distinction between true keyhole solutions, complementary policies, and (selective) blanket restriction

Consider the (abstract) problem that high levels of migration, along with current de facto rules for the rules for eligibility for welfare benefits, could lead to fiscal bankruptcy. Consider three potential “solutions” (note that these don’t even come close to exhausting the space of possible “solutions” — but they help to illustrate the distinction I’m trying to draw here):

1. Improve the effectiveness with which immigrants (perhaps limited to the additional immigrants under migration liberalization) are walled off from the welfare state (this could involve changing the rules, or enforcing existing rules more effectively, or a combination).
2. Reduce welfare benefits across the board for the whole population (see also our contraction of welfare state page).
3. Forbid the migration of people for whom the probability of welfare benefit use, or the extent of such use (in expectation) exceeds a threshold.

In a loose sense, these are all “keyhole solutions” insofar as they attempt to address the (perceived or predicted) problem of migrant welfare benefit use.

However, they are all different in important ways:

1. The first addresses the perceived problem at the intersection of migration status and welfare eligibility. Prima facie, this targets the problem most narrowly and is most deserving of the “keyhole solution” label. I’ll call this type of solution a true keyhole solution.
2. The second addresses the problem but focuses on the broader issue of welfare use and welfare eligibility. Rather than focusing on migrants per se, it addresses a potential problem that is made more severe due to migration flow, but it addresses it in a way that does not per se discriminate on the basis of migration status. I’ll call this type of solution a complementary policy change.
3. The third seeks to preserve the status quo as far as possible with respect to domestic policy, and addresses the potentially dangerous interaction with migration by forbidding the forms of migration perceived as risky. I’ll call this type of solution a blanket restriction. To emphasize that the blanket restrictions don’t apply to everybody, we might call it selective blanket restriction.

However, from another perspective, (2) and (3) are examples of keyhole solutions, insofar as they directly address problems created by migration. Whatever names you choose, I want to claim that there is an important conceptual distinction.

(1): True keyhole solutions: unequal treatment within a territory

True keyhole solutions are most likely to openly run afoul of moral scruples, particularly those related to equal treatment, from the territorialist perspective (i.e., from a perspective where the welfare of people within a geographical territory matters regardless of their country of origin, whereas the welfare of people outside the territory matters much less). In particular, true keyhole solutions strongly rub against people’s local inequality aversion. Because they specifically focus on policy at the intersection of migration status and the relevant domestic policy issue, they treat migrants differently from natives with respect to that policy issue. To be concrete, the true keyhole solution for welfare access denies migrants access to welfare benefits that natives can access. They discriminate between people in the same territory based on their migration status. Or, as Paul articulated in his critique of keyhole solutions:

Charging immigrants special taxes just because they are immigrants is discriminatory, violating the principle of equality before the law. It also ignores the common sense principle that, where possible, you should tax bads instead of goods. Immigration is not a bad.

Apart from their open violation of the principle of equal treatment within a territory, true keyhole solutions are problematic in another sense: they can be hard to enforce in practice. This is particularly true of the true keyhole solutions that involve migrant behavior after they’ve already moved (as opposed to those that are applied at the point of entry). The migrants are part of the general population now, and can mingle freely with it. Maintaining a separate set of laws for them would require identifying migrant status for them, as well as for natives who might be confused for migrants. Just requiring migrants to carry identifying documents isn’t enough — non-migrants also need to do so. The local law enforcement or welfare offices need to cooperate with the enforcement of the distinction based on migration status. These aren’t impossible problems if the true keyhole solution involved otherwise seems reasonable to those tasked with enforcing it, but this may not always be the case. For instance, denying migrants welfare benefits requires the cooperation of state and local offices where people submit documentation for welfare eligibility. What if the officers there are reluctant to put in extra work to ascertain a person’s migration status, or are happy to look the other way if migrants proffer false documentation, or simply aren’t able to judge the authenticity of documents?

Running afoul of equal treatment can also be problematic in cases such as the minimum wage, for different reasons — if immigrants have a different minimum wage than natives, this might actually hurt natives because their competition can credibly promise to legally work at a lower wage than they can. David Henderson made the point here:

First, [Bryan Caplan] says that if we didn’t get rid of the minimum wage, legal immigrants would just get black-market jobs. That’s true, but incomplete. Precisely because they would now be legal, they could comfortably go to a wage board in their state and complain after the fact and get back pay. An employer, looking forward to that outcome, would not hire them. I pointed out here that ironically, the way the minimum wage law is enforced is what gives illegal immigrants (who would no longer be illegal under Bryan’s preferred policy) an advantage in the competition for jobs. Illegal immigrants can credibly commit not to turning in a minimum wage violating employer. Legal immigrants can’t credibly commit.

(See also Caplan’s response here).

The minimum wage is a separate topic that deserves several blog posts of its own (but check out previous Open Borders Action Group posts about the minimum wage, in particular this post). Let’s get back to our current topic now.

(2): Complementary policy change: complex, and gets us into general debates about politics and policy

Complementary policy changes are protected from the charges of inegalitarianism that can be used to attack true keyhole solutions.

On the other hand, complementary policy changes affect domestic policy not just for migrants, but also for existing natives. They thus have a direct and obvious effect on natives. Discussing the wisdom of complementary policy changes therefore involves discussing the general wisdom of particular political stances. Ultimately, once we advocate open borders along with complementary policy changes, we’ve moved from solely debating or advocating migration policy (if we were ever doing that) to advocating an overhaul of sorts of the overall political and social system. I don’t mean to exaggerate: in the short run, open borders is perfectly compatible with the nation-state as we roughly know it, but to the extent that complementary policies are necessary or desirable to fully realize the gains from open borders, it does involve a lot of reimagining of the specifics of the nature and role of government. This is a topic worthy of a lot more blog posts, so I’ll leave it here for now.

(3): (Selective) blanket restriction: inflexible and generally inefficient but arguably necessary in some cases

Blanket restrictions are the restrictionist solution. Some types of migrants would (in expectation) cause or exacerbate a particular problem? Don’t allow them to come!

Blanket restrictions basically revert us to the original one-dimensional view of migration, where the question is just of how many and who the people allowed to migrate are. It is beset with the same problems as any external attempt to micromanage and control a voluntary market interaction. Some people could move and become better off, and make the world better off, but now aren’t allowed to do it because they happen to belong to a category that is forbidden from migrating. Much better to let them move and use a true keyhole solution to prevent any harms they may cause.

Nonetheless, there are arguably some contexts where blanket restrictions might be the only or best approach. Open borders advocates generally restrict these contexts to terrorists, seasoned criminals and people carrying communicable diseases (see this post of mine for more exploration of the crime question).

Immigration tariffs and bonds, and schemes such as DRITI are almost true keyhole solutions, but for a subset of the affected migrant population, they effectively function as blanket restrictions. Nathan has previously clarified that DRITI differs from blanket restrictions in the respect that deportation is off the table even for people who violate the terms of DRITI — they are treated as tax evaders rather than deportable offenders. Nonetheless, some people are likely to still be priced out of the system.

Fuzziness of the distinction

How clear is the distinction in practice? That is a hard question. Even open borders advocates are sometimes inconsistent and confused about the relevant distinction. For instance, many moderately pro-open borders people would balk at (3) as a “keyhole solution” to welfare use, but are okay with considering or even endorsing the corresponding “keyhole solution” to crime (for more on blanket restrictions for crime, see this earlier post of mine).

The distinction between (1) and (2) can get tricky for these two related reasons:

• On the one hand, policies that are ostensibly targeted only at migrants can in fact have wide-reaching effects on natives who are among their friends, family, employers, employees, co-workers, or customers. Nathan Smith made this point really well in the context of blanket restrictions when discussing the right to invite and the communitarian case against migration restrictions. But the same point applies to keyhole solutions such as taxes and tariffs. As any person who has studied the economics of tax incidence can tell you, the person who has the direct responsibility of paying a tax isn’t necessarily the only person who effectively pays for it.
• On the other hand, it is possible to devise policies that are ostensibly neutral and appplicable to all people in the geographical territory, but are effectively targeted largely at migrants and their children. For instance, educational policies relating to education in languages other than the country’s native language(s) are largely targeted at migrants and their children, although they apply equally to all. Similarly, racial policies may be ostensibly neutral to migration status, but may disproportionately affect migrants because of differences in the racial composition of natives and migrants.

Immigration tariffs and bonds, discussed at the end of (3), show that the distinction between (1) and (3) can also be fuzzy.

Related reading

• Slippery slopes to open borders by Vipul Naik, February 11, 2014.
• Keyhole solutions: permissibility, desirability, feasibility, and stability by Vipul Naik, August 16, 2013.
• Six possible positions on a specific keyhole solution by Vipul Naik, November 1, 2012.
• Auctions, tariffs, and taxes by Nathan Smith.