Keyhole solutions

There are many plausible and valid objections to open borders, particularly harms to immigrant-receiving countries.

However, the fact that there are harms arising from free immigration does not mean that the best way to mitigate these harms would be to close the borders completely. Rather, there may be strategies to mitigate the harms while preserving the tremendous plus side of immigration. Such strategies are part of what economist Tim Harford calls keyhole economics — when trying to address a problem, find solutions that are specifically tailored to that problem. In other words, don’t draw a sword to kill a mosquito.

Examples of keyhole solutions

Below are some of the important classes of keyhole solutions:

  • Immigration tariffs: Charge a fee for immigration, possibly influenced by market forces, rather than putting hard quotas.
  • Guest worker programs: Expand the scope of existing guest worker programs, and create new ones, that do not offer citizenship/voting rights or welfare benefits/schooling subsidies.
  • Linguistic and cultural fluency requirements: Require a certain minimum of linguistic and cultural fluency as a prerequisite for immigration.

An interesting scheme that combines ideas from both immigration tariffs and guest worker programs is the Don’t Restrict Immigration, Tax It scheme proposed by Nathanael Smith, who also blogs for this site. The full text of the chapter where Smith proposes this scheme can be downloaded as a Word Document or as a PDF.

The concept of keyhole solutions

In his essay Why Should We Restrict Immigration? (part of Cato Journal Winter 2012), Bryan Caplan writes (Page 17, concluding paragraph of essay):

Even if all these empirical claims are wrong, though, immigration restrictions would remain morally impermissible. Why? Because there are cheaper and more humane solutions for each and every complaint.If immigrants hurt American workers, we can charge immigrants higher taxes or admission fees, and use the revenue to compensate the losers. If immigrants burden American taxpayers, we can make immigrants ineligible for benefits. If immigrants hurt American culture, we can impose tests of English fluency and cultural literacy. If immigrants hurt American liberty, we can refuse to give them the right to vote. Whatever your complaint happens to be, immigration restrictions are a needlessly draconian remedy.

In an earlier blog post for The Economist titled Immigration Restrictions: A Solution in Search of a Problem (which he later expanded into a video), Caplan writes:

Every blogger I know tells me the same story: The fastest way to provoke angry comments is to post a kind word about immigration. In the blogosphere, as in real life, complaints about immigrants abound. The funny thing, though, is that the complaints are diverse, but the cure is almost always the same: Cut immigration quotas, reinforce the border, and deport the illegals.

What would happen, though, if we actually wrote down specific complaints about immigrants and tried to figure out specific solutions? While we’re at it, why not focus on specific solutions that are cheap and relatively humane?

Suppose, for example, that the complaint about immigrants is that “They take advantage of the welfare state.” If that’s the problem, the simplest solution is not the get rid of immigrants, but to make them ineligible for benefits. Make them pay the usual taxes, but make it clear that welfare, unemployment benefits, Medicare, Social Security, and the like are only for native-born citizens.

(More at the link).

Here is a video where Caplan and Alex Nowrasteh discuss some keyhole solutions:

Converting potential Pareto improvements to actual Pareto improvements

Keyhole solutions help address one objection to open borders: although open borders are a potential Pareto improvement over the status quo (i.e., winners gain more than losers lose), they are not an actual Pareto improvement, because at least some people lose. Keyhole solutions can be designed so as to convert these potential Pareto improvements into actual Pareto improvements. Michael Clemens, a proponent of freer migration, has said something to the effect that keyhole solutions can convert “win-win-lose” scenarios to “win-win-win” scenarios.

Nathan Smith, in a paper titled Open Borders with Migration Taxes are the Optimal Policy, and a blog post of the same name, explains, both verbally and mathematically, why open borders with migration taxes (a type of keyhole solution) would be optimal.

Development economist and open borders advocate Lant Pritchett has made the argument, although he uses the term “instruments-to-targets approach” to describe keyhole solutions. Here is a quote from Page 94-95 (part of Chapter 3) of his book Let Their People Come: Breaking the Gridlock on Global Labor Mobility (free PDF chapters, Amazon paperback):

This is about the political traction of the idea that because it lowers the unskilled wage and increases wage inequality, labor mobility should be opposed. Though the general popularity of the notion might not have much to do with economists, it is worth a word or two about the economists’ response to this idea. The economists’ typical response to objections to potential Pareto-improving policies that worsen the income distribution is to say “Instruments to targets.” That is, nearly any macroeconomic or microeconomic policy reform—trade liberalization, airline deregulation, macroeconomic stabilization—will produce some change in the economywide income distribution. But the usual approach is to recommend a policy if it is potentially Pareto improving and to also recommend that broad distributional concerns be addressed through the best available instruments for redistribution. That is, there is a broad tradition (to which many object but is nevertheless widespread) that one cannot burden every single policy with a complete general equilibrium analysis of not only the aggregate but also the distributional impact of each policy and use every policy as a distributional instrument. Rather, the “instruments to targets” literature suggests that the most effective policy is to have the best instrument for each target. In this view, while any policy (say trucking deregulation) might have distributional consequences, the policy recommendation should maximize output with one set of instruments and redistribute with another.

Keyhole solutions are a breakthrough idea

In a blog post titled Milton Friedman opposed a Pareto improvement, Bryan Caplan notes that despite Milton Friedman’s concern about immigration and the welfare state, he did not give much thought to keyhole solutions.

Q: Instead of a green card [resident alien status], can the USA issue a blue card which does not give welfare?

A [Friedman]: If you could do that, that would be fine. But I don’t believe you can do that. It’s not only that it is not politically feasible, I don’t think that it is desirable to have two classes of citizens in a society.[…]


My last, best way to rescue my intellectual idol is his admission that he’d never thought about the proposal before. But isn’t that bizarre? Friedman, who could run mental circles around lesser geniuses, never asked himself, “What is the least un-libertarian way to deal with the conflict between free immigration and the welfare state?”? The lesson, perhaps, is that what Tim Harford calls “keyhole surgery” (looking for the least intrusive solution to market and/or political failure) is a major intellectual advance – obvious in retrospect, but deeply counter-intuitive nonetheless.

Keyhole solutions as dealism

Why should people who are unconvinced that immigration creates a problem agree to a keyhole solution to that problem? Another way of looking at keyhole solutions is in terms of what economist Robin Hanson has called dealism. Hanson outlines the idea of dealism in his blog post Efficient Economist’s Pledge.

Objections to keyhole solutions

For more on this topic, see the blog post Keyhole solutions: permissibility, desirability, feasibility, and stability.

There are many broad classes of objections to keyhole solutions. The broad classes of objections are listed here. Specific formulations of these objections are discussed and linked to on pages on the specific keyhole solutions.

  • The implementation of keyhole solutions is inegalitarian, discriminatory or un-libertarian. For instance, requiring immigrants to display English language proficiency is discriminatory: The counter-argument to this is that even though such practices may violate some ethical principles, they are still better than a blanket denial of people’s right to migrate on the account that they might be bad at English. The point is not whether a given keyhole solution is better or worse than placing no restriction. The point is that the keyhole solution is better than a blanket restriction.
  • Keyhole solutions lack political feasibility: This is probably true as of now, but some keyhole solutions already have historical precedents and/or are adopted in partial form in the present day. Further, the relevant question is not the feasibility of keyhole solutions in absolute terms, but their feasibility relative to unconditional open borders.
  • Keyhole solutions lack political credibility (for their long term survival): For instance, a keyhole solution that denies immigrants the right to vote lacks credibility because all it takes is one bunch of politicians eager to grab their votes a few years down the line. The counter-argument is that this may be true, but restrictionists could redirect their energy from preventing immigration to trying the strengthen the credibility and permanence of keyhole solutions among the people, politicians, and movers of public opinion.

Rank-order preferences for open borders with and without keyhole solutions

The following blog posts discuss various rank orderings related to open borders and keyhole solutions: