This post builds on my preceding post not-quite-open borders: keyhole solutions, complementary policies, and blanket restrictions. Although it’s not necessary to read that post first, doing so would provide better context for understanding this post.
The keyhole solutions page on our site dates back to the site’s inception, and discussion and comparison of different keyhole solutions has been an important part of our site and blog since then. The basic idea behind keyhole solutions: for any identifiable problem, try to find a solution that addresses the problem as specifically and narrowly as possible, while not forbidding or restricting other actions. For instance, if the concern is that immigrants’ use of welfare benefits will lead to fiscal bankruptcy, placing and enforcing stronger restrictions on immigrant welfare access would be the keyhole solution that allows migration, preserves the existing welfare state for existing users, and allegedly solves the alleged problem. More at the keyhole solutions page.
My co-blogger Nathan Smith had devised his favorite keyhole solution, Don’t Restrict Immigration, Tax It (DRITI), long before Open Borders: The Case existed.
Does the site advocate keyhole solutions? Not officially, but it’s clear that we give them serious consideration as one of the alternatives to pure open borders as a way of liberalizing migration. In one of the first systematic explorations of the subject, I noted that there are six possible rank-order preferences one might have between “pure” open borders, open borders with keyhole solutions, and the closed borders-ish status quo. For any keyhole solution A, the three options being ranked are:
- Open borders without keyhole solution A (that one might loosely call “pure” open borders).
- Open borders with keyhole solution A.
- Closed borders and/or status quo.
This loose trichotomy (pure open borders, open borders with keyhole solutions, and closed borders) has been loosely endorsed by other bloggers on the site, such as Samuel Wilson in his discussion of moral intuitions and the euvoluntary principle in connection with open borders, myself in my discussion of the permissibility, desirability, feasibility, and stability of keyhole solutions, and Paul Crider in his critique of keyhole solutions. Paul offers a great summary of keyhole solutions before taking them down (which, in the above jargon, means endorsing (1) > (2) > (3) over (2) > (1) > (3)).
Implicit in this discussion is the view that some approaches to a more liberal migration policy are identifiable as (closer to) pure open borders whereas others are identifiable as keyhole regimes. Broadly, this is true, which is why my post, and Paul’s, make eminent sense. And I think that keyhole solutions are an important idea in the lexicon of people thinking about realistic regimes with substantially more liberal migration policies than exist today. However, I think that the distinction between pure open borders regimes and keyhole regimes is quite fuzzy. But first, a little detour.
Keyhole solutions, complementary policies, and blanket restriction
As I discuss at more length in my post not-quite-open borders: keyhole solutions, complementary policies, and blanket restrictions, there are actually three slightly different types of policy adjustments and compromises that often get put in the broad bucket of keyhole solutions:
- 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.
- 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.
- 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.
Ceteris is not paribus: some mathematical background
(This section doesn’t strictly require, but can benefit from, broad familiarity with the concept of derivatives and multivariable calculus).
In economics and the social sciences, it is customary to consider questions of the form “ceteris paribus, how does a change in variable x affect variable u?” Here, ceteris paribus is understood to mean “other variables being left unchanged.” The concept of partial derivative is a particular encapsulation of this idea of trying to isolate the effect of one variable on another while holding the remaining variables constant.
Back when I taught multivariable calculus to economics majors, I used to emphasize the important fact that the concept of ceteris paribus is ill-defined, because the choice of “other variables” that you keep constant heavily depends on how you coordinatize the system (you can read more here, and also watch the embedded videos). The “real-world” example discussed on the page is quoted below (and isn’t directly related to migration, but will help illustrate the general line of argument):
Suppose a country’s military spending is determined by just two factors: its per capita GDP and its population. We want to study the relationship between per capita GDP and military spending.
There are two sensible ways (among many) of trying to do this:
- Study the relationship between per capita GDP and military spending holding population constant. In other words, take the partial derivative with respect to per capita GDP holding population constant. In this case, we are thinking of military spending as a function of the two variables: per capita GDP, and population.
- Study the relationship between per capita GDP and military spending holding total GDP constant. Prima facie, this is similar to the previous one, because total GDP is just a product of per capita GDP and population. However, what we have done effectively is consider a partial derivative of the function in a new coordinate system, where the two variables of interest are: per capita GDP, and (per capita GDP) times (population).
The point is that the partial derivatives will have different expressions depending on what we hold constant.
Ceteris is not paribus: the problem of distinguishing between “pure” open borders and a compromised solution
In the simplest telling, pure open borders refers to a situation where we “just open the borders” and keep all other policies essentially unchanged. Keyhole solutions (viewed broadly) involve opening the borders but making some changes to policies. These changes could be true keyhole solutions (operating at the intersection of migration and the relevant domestic policy), complementary policy changes (such as changes to general rules for welfare eligibility or changes to the minimum wage), or blanket restrictions on some types of migration.
However, this raises the question: what are those other things that we hold constant? If nothing else, open borders will lead to changes in the size of the population in many countries, and sometimes quite significant changes. What would it mean, for instance, to say that we open the borders without changing any policies or rules related to the welfare state? There are many possible interpretations:
- The rules for welfare eligibility, and the size of welfare benefits per capita for recipient, remains the same.
- The total size of the welfare state (i.e., the amount of money allocated to pay for welfare benefits) stays the same, or stays the same relative to the population, or stays the same relative to the size of the economy.
- The total number of welfare recipients remains the same, or remains the same relative to the population size.
These are all different senses of “staying the same.” In most contexts, these differences don’t matter much (or at least, don’t appear to matter much) because the overall global changes to the relevant quantities aren’t very large. However, if open borders is going to be a big deal, then it likely won’t be possible to hold all of these constant or even close to constant. Once we concede that not all aspects of existing policy regimes can be held constant, the conceptual distinction between pure open borders and keyhole solutions becomes more tenuous. Both “pure open borders” and “keyhole solution”-type policies offer plausible descriptions of a world with more liberal migration that nonetheless preserve or inherit some features from the status quo. Calling one set of policies “pure open borders” is mostly about exercising a value judgment regarding which features are central to and represent the essence of the system.
Similarly, consider minimum wage policies (the minimum wage is a topic that has been discussed extensively in the Open Borders Action Group, and we intend to do blog posts about it). Again, there are many different ways in which we can consider minimum wage levels:
- We can think of them as associated with specific currency units, not adjusted for inflation. So a minimum wage of $8/hour remains a minimum wage of $8/hour even if the dollar depreciates significantly in value.
- We can think of a minimum wage in terms of the associated consumption basket, i.e., indexed for cost of living.
- We can set the minimum wage in relation to median wage levels.
- We can set a minimum wage threshold based on the maximum amount of unemployment we are willing to tolerate.
There are arguably many different ways of thinking about the appropriate level of a minimum wage (and whether there should be a minimum wage at all). Our choice of preferred rationale for justifying the status quo will determine what sort of minimum wage regime we’d like to see in an open borders world.
For those who simply think of the minimum wage in terms of a currency unit, little head-scratching needs to be done. For those who associate minimum wages with consumption baskets, median wages, or tolerable levels of unemployment, however, the nature of changes under open borders could be quite different. Migration can change the cost of living structure by making some services cheaper and others more expensive. Thus, cost-of-living calculations could change quite a bit. Median wage levels could change both because the wages of existing residents change and because of compositional effects arising from inclusion of the new migrants. Finally, given differences in the skill level of migrants compared to natives, the minimum wage would probably need to be reduced to maintain a similar unemployment rate to the status quo.
With all that said …
I don’t want to exaggerate how dramatic open borders would be. I think a fairly liberal migration regime is fairly consistent with the nation-state roughly as we know it, and while it might push the world to a “no borders” condition over the long run, that won’t happen in short order. But a lot of the specifics will change, in ways that are somewhat but not entirely predictable, and in some cases not pretty. Thus, the concept of “pure” open borders (“just open the borders and don’t touch anything else”) is not as clear-cut as we might naively presuppose. At the end of the day, a policy implementation has to consider many other existing policies in conjunction with changes to migration levels.
With that said, it is still possible to advocate purely for open borders, without advocating for the concept of pure open borders. When I say “advocate purely for open borders” I mean advocate for change in the direction of open borders, without being picky about the specifics, and being flexible about the selection of specifics based on more detailed context-specific empirical analysis. This might mean advocating for changes that could go in directions you consider more “pure” versions of open borders, or advocating in directions you consider more like keyhole solutions or open borders with complementary policies.
In addition to inline links within the blog post, check out:
- Open Borders Action Group post by John Lee, linking to No bed of roses for Haitians in Brazil despite open borders by the Associated Press, March 7, 2015. See in particular the comments by Carl Shulman.