This post is more of a utility post that grew as a spinoff of another post I’m drafting, which is getting a bit long. I hope to refer to this post in that other (forthcoming) blog post and in other forthcoming blog posts. I’ll be using the term “keyhole solution” a lot in this post, so if you haven’t already, I suggest you read the keyhole solutions page to get an overview of the term, and also check out all blog posts with the “keyhole solutions” tag on this website.
Consider keyhole solution A that has been developed to meet a specific objection to open borders. Suppose you are asked to specify a rank-order preference between three options:
- Open borders without keyhole solution A.
- Open borders with keyhole solution A.
- Closed borders and/or status quo.
Now, there are six possible rank-order preferences for this set of three options (we get the number 6 by taking the factorial of 3). I want to discuss each of these, and the kind of people who might want to take that position.
Open borders advocates who prefer open borders without the keyhole solution: rank-order (1) > (2) > (3)
These open borders advocates would prefer open borders without the keyhole solution. But, since the world is not an ideal world, they are willing to compromise and accept open borders with the keyhole solution. They are not genuinely convinced that the problem the keyhole solution addresses is a genuine problem that needs to be addressed through the keyhole solution. But they’re willing to go along with the keyhole solution as a form of dealism, or to placate existing interests and convert potential Pareto improvements into Pareto improvements.
John Lee seems to take this position with respect to the immigration tariffs keyhole solution. He writes in a comment on his own post:
Thanks Nathan. I think we agree on what’s probably the best achievable policy reform for now (immigration tariffs), but I am inclined to disagree with your moral preference for Pareto-improving policies here, as well as your characterisation of them as merely “a little unfair”. I think immigration tariffs would be a massive improvement, but remain a distant second-best policy (from a moral standpoint) to true open borders. I used to consider immigration tariffs the best policy here, but have been convinced by the economic evidence that I was placing far too much weight on natives’ welfare, and far too little on migrants’ welfare. (This change in my beliefs is something I plan to write more on, so I won’t elaborate too much on it here.)
The theme of compromise and the willingness to accept keyhole solutions to problems he’s not convinced are serious runs through John’s writings. For instance, in a comment, he writes:
Surely you’ve heard of the term “political compromise”? Keyhole solutions split the difference between restrictionists and hardcore open borders advocates. If you’re concerned about giving the other side a victory, that’s how it’s supposed to work.
Open borders advocates who prefer open borders with the keyhole solution: rank-order (2) > (1) > (3)
Here, the open borders advocate would support open borders both with and without the keyhole solution — the case for open borders is too strong for a keyhole solution to make or break it. But the open borders advocate still prefers open borders with the keyhole solution. Many open borders advocates take concerns about political externalities seriously, and are willing to consider keyhole solutions that would limit voting rights for recent immigrants as superior to open borders with voting rights. But even if they’re not able to get that preferred policy, they would still support open borders. The same is true for making immigrants ineligible for welfare: a lot of libertarian supporters of open borders would prefer that immigrants be ineligible for welfare benefits, but would support open borders even if this were not possible. Bryan Caplan makes this kind of argument in a response blog post to David Henderson regarding political externalities, something both consider a possible problem:
Now, under my scheme, the U.S. government would couple open borders with a 20-year residency requirement for U.S. citizenship and a requirement that one be a citizen in order to get any kind of welfare, including government schooling… But Bryan and I both know that you can’t always get what you want. The government is not some entity that he and I control. So what if the government did not couple open borders with this 20-year residency requirement or even a 10-year residency requirement? Would Bryan still advocate open borders?
Definitely. While I view libertarian rights as prima facie rather than absolute, I do insist on high certainty of very bad consequences before making exceptions. And the negative political consequences of letting lots of immigrants vote are at worst uncertain, for reasons I’ve previously explained.
David concludes:
And would he worry at all that the new residents would vote away the goose that lays the golden eggs? And if he wouldn’t worry about that, why wouldn’t he worry? Inquiring minds would like to know.
Of course I would worry. But mere worry is a poor excuse for depriving millions of their basic human rights to work for a willing employer and rent from a willing landlord. And I’m confident that immigrant voters would have one extremely pro-liberty effect: Protecting the right of free migration. As I think David will agree, American voters have sadly shown themselves to be totally unreliable on this vital issue for over a century.
Absence of keyhole solution a deal-breaker for open borders: rank-order (2) > (3) > (1)
This position is taken by (I think) people like BK, who, in this comment and elsewhere, express willingness to sign on to open borders with keyhole solutions — such as guest worker programs or immigration tariffs, but would reject open borders otherwise, preferring instead the status quo or closed borders.
Milton Friedman seems to have taken this position regarding immigration and the welfare state in the US. As Donald Boudreaux notes here, Friedman was willing to sign onto open borders conditional to the elimination of the welfare state.
The size and influence of this group is important because the open borders advocates of the (1) > (2) > (3) type are really peddling the keyhole solution largely in order to appeal to this group. BK, and possibly some other EconLog commenters, fall in this group. If the size of this group is negligible, then the (1) > (2) > (3) type open borders advocates (who aren’t genuinely convinced about the desirability of the keyhole solution) would stop supporting the keyhole solution since they don’t seem to be getting much added support from non-open borders advocates in exchange for the compromise. Of course, even in the absence of (2) > (3) > (1) types, the (1) > (2) > (3) may still end up compromising with the (2) > (1) > (3) types and support the keyhole solution, but at any rate they wouldn’t be trying to actively advertise and garner support for keyhole solutions.
Presence of keyhole solution a deal-breaker for open borders: rank-order (1) > (3) > (2)
These are people who would support open borders without the keyhole solution, but they are so opposed to the keyhole solution that they’d rather stick with the status quo than endorse the keyhole solution. For instance, many people who are influenced by local inequality aversion and territorialism raise the second-class residents objection to guest worker programs.
An example is the Immigration Policy Center’s opposition to the Red Card scheme may be thought of as something of this sort, though it’s unclear whether the Immigration Policy Center actually supports open borders or radical changes to the status quo. I blogged about the attitude of pro-immigration groups to open borders here.
Status quo preferable, but if you have to have open borders, do it with the keyhole solution: rank-order (3) > (2) > (1)
The majority of restrictionists seem to take this position. They want to push for the keyhole solutions anyway, even as they push for closed borders, because there is going to be some amount of immigration (legal and illegal) despite their best efforts. For instance, restrictionists in the US want to end birthright citizenship and various forms of welfare for immigrants, but not as a “keyhole solution” that could be used to justify expanded immigration, but rather, to complement their efforts to close the border.
As mentioned earlier, a lot of open borders advocates develop keyhole solutions largely to appeal to the deal-breaker folks ((2) > (3) > (1)). But given that the deal-breaker folks are likely much fewer in number than restrictionists who prefer the keyhole solution anyway as a safety measure, it’s understandable that open borders advocates are concerned that too much effort on their part to develop keyhole solutions, without managing to get explicit promises of more support for open borders from the deal-breaker folks, would just play into the hands of restrictionists. This, I think, partly explains the reluctance of some on the open borders side to work harder in promoting keyhole solutions.
Status quo preferable, but if you have to have open borders, do it without the keyhole solution: rank-order (3) > (1) > (2)
Some restrictionists are averse to specific keyhole solutions (for instance, due to the second-class residents objection), and if open borders were forced down their throats, they’d prefer not to have the keyhole solution. Bryan Caplan notes that this seems to be the case with emergency medical care for immigrants: much as restrictionists lament the overuse of emergency medical care by immigrants, many of them would be uncomfortable with the keyhole solution of denying emergency medical care. Similarly, Tino Sanandaji, not a fan of open borders, states very clearly that he is opposed to the idea of second-class residents. He writes here:
For me the choice is simple, I prefer Democracy to Open Borders. If a country decides to take immigrants, they have to be included 100% with full rights, and not permanent second-class citizens.
Overall summary
The interaction between these six rank-order preferences for every possible keyhole solution is important for determining the types of coalitions that may be formed around support and opposition for a keyhole solution. The analysis in this blog post is hopefully a starting point that will enable clearer thinking on this issue in the future.
Vipul Naik founded the Open Borders website in March 2012. See also:
Vipul Naik’s personal statement
Vipul Naik’s description of open borders on his personal site
Page about Vipul Naik on Open Borders
Thanks Vipul, I think this is a fantastic catalogue of potential views of open borders interacting with keyhole solutions. A couple thoughts in response.
First, I think the *kind* of keyhole solution we’re discussing may also affect where one falls in this matrix. I actually virtually completely agree with Bryan Caplan’s position on keyhole solutions with respect to citizenship and similar political rights. I’m much more open to keyhole solutions which restrict political rights (as long as they do not *extinguish* such rights altogether) than keyhole solutions which restrict more fundamental human rights, such as freedom of movement or contract.
Second, this isn’t much to do directly with keyhole solutions, but Tino Sanandaji’s comment reminded me that it is important to be clear on what kind of migrants and migration is being discussed. People frequently conflate “person who intends to immigrate temporarily” with “person who intends to immigrate permanently” so as to cherrypick the arguments they make about immigration. It’s not obvious to me why the first category of person should automatically be entitled to political rights, and it’s not obvious to me why the second category of person should never be entitled to political rights. And it’s not obvious to me why immigration policy should be designed to force people at the instant they cross the border into one of these buckets, which is exactly what most countries’ immigration laws do. (Under US immigration law, all foreign students must never intend to stay in the US after graduation; intent of doing so is grounds for refusal of a student visa, or worse. This policy is laughably absurd in the face of how humans actually behave; it’s like telling a Mancunian who attends LSE that he can only do so if he has no intention of living or working in London after graduation.)
The benefit of many keyhole solutions, I think, is that they approximate the effect of true open borders. Under true open borders, I can go to Paris and stay for a week, a month, a year, a lifetime. Under current immigration law, that isn’t true (at least, not without an immense amount of trouble). But the right set of keyhole solutions could maintain immigration restrictions while still enabling me and most people, more or less, to experience life as it would be under an open borders regime. Most of the problems people worry about when it comes to open borders are simply “open citizenship”, which few, if any, of us are advocating — keyhole solutions address the “open citizenship” problem head on while keeping the borders mostly open.
John, yes, you’re right that your rank-order preference is a function of the keyhole solution being discussed. That’s what I intended to say, but I just “fixed a specific keyhole solution” beforehand — a typical style of mathematical argumentation.
In technical jargon, if you have a bunch of different problems and a bunch of different keyhole solutions, you have a vector of rank-order preferences, one for each keyhole solution. You could complicate this analysis further by looking at rank-order preferences for keyhole solution “packages.” May be somebody’s rank-order preference for keyhole solution A + B cannot be predicted by looking at that person’s rank-order preference for keyhole solution A and rank-order preference for keyhole solution B separately.
Anyway, I didn’t want to get too technical in a blog post, which was meant more to lay out a framework I (and possibly others) can refer to in subsequent posts.
“Open borders advocates who prefer open borders with the keyhole solution: rank-order (2) > (1) > (3)”
One thing that this analysis may not make clear is that if people in this category think that (2)>>(1), but implementing (1) precludes (2) (e.g. it’s much harder to disenfranchise or expel people than not to make them citizens in the first place), then they have reason to resist (1) as long as there is still a significant chance of (2).
Toy example: you’re trying to boost long-run world GDP with migration policy. So if (2) would really boost world GDP by 50%, but (1) would only increase GDP by 10% after taking into account political externalities over time (leaving out non-GDP impacts like R&D externalities and global crises), then one should vote against open borders (if it were put to a vote, and hopefully also voting for things like increased foreign aid) so long as there is still over, say, a 1 in 4 chance of (2) coming along in the next few decades.
This effect can’t be captured in the rank ordering above, because it makes no mention of probabilities, or the possibility of later changes in state..
Excellent point, BK. I’ll definitely incorporate these ideas in future posts on the interaction dynamics, which, unfortunately, will have to wait a while.
Good taxonomy. To estimate the size of the groups I would look at divergence in polling data on attitudes towards immigration from different countries, and attitudes towards different policies like the H1-B and visas for scientists vs the bracero program, Reagan amnesty, etc.
“The size and influence of this group is important because the open borders advocates of the (1) > (2) > (3) type are really peddling the keyhole solution largely in order to appeal to this group.”
In the U.S., I think the most politically relevant constituency for keyhole solutions would be the amalgam of Republican officials and elites. They are eager for more labor to increase wealth, returns to capital, and markets on behalf of businesspeople, but they also fear demographic shifts towards groups with higher rates of government dependence, lower incomes, and anti-market biases will make the (slightly) small government and pro-business ideology too hard a sell.
Unlimited scale guest-worker programs, separating votes from jobs, have been put forward by Bush II and Gingrich as proposed legislation and as campaign platform. That is a sector that is persuadable, very powerful, and has self-interested reason to fear non-keyhole open borders.
Yes, the real world effect of open border advocacy is usually marginal increases in legal guest workers and a lack of enforcement against illegal immigration. This is mostly meant to drive down labor costs a bit without “breaking anything”.
I very much doubt that any of the keyhole solutions will last more than 50 years.
Restriction of voting rights for a group that could potentially grow to 20% or more of the country? There is no way they wouldn’t be able to muster enough political power to overturn that. Not in a democracy. In a monarchy perhaps (see Dubai) but not in the US.
No Welfare? Poor immigrants aren’t going to accept that if hard times hit. There is no way you can look a starving immigrant mother in the face and say “sorry, you’re not a citizen”.
Immigration tariffs. Well okay, I guess a one time fee is feasible.