Paul Crider’s recent blog post critiquing keyhole regimes, and his subsequent comment reply to Nathan regarding the appropriate keyhole solutions regarding the decline in the standard of living of some natives (due to wage decline) that might be engendered by migration got me thinking. Roughly speaking, we can summarize the effects of migration as follows:

• Migrants benefit greatly from migrating. Economically, this is due to the huge place premium that many migrants benefit from.
• The overall size of the economy that receives the migrants increases. On a per capita basis, the effect on the incomes and wages of prior residents of the economy is expected to be somewhat positive, but much less than the gains experienced by migrants.
• There are some subsegments of the native population, particularly those at the low-skilled end, who might in expectation experience small economic losses due to migration.

So, some natives are expected to lose out. A number of people believe that political changes must be, to the extent possible, Pareto improvements: nobody’s worse off than before. This is not to be taken literally: it’s impossible to be sure that nobody would lose out from a change as large as significant liberalization of migration. So I’ll use Pareto improvement in the sense of Pareto improvement in expectation at the clearly identifiable subgroup level: no large subgroup of the population that can be identified clearly in advance should, in expectation, be worse off than before. On this view, then, one native being killed by somebody who migrated due to liberalization of migration does not violate migration being a Pareto improvement. But if high school dropouts can, in expectation, expect to see their wages go down by 5% due to migration liberalization, that is a no-no and we need to get back to the drawing board.

To the extent that “pure” open borders is not a Pareto improvement, i.e., we can clearly identify subgroups of the population that would, in expectation, lose out from open borders, what is the best that can be done to fix the problem? The simplest approach is to compensate them through lower tax rates, greater welfare state benefits, or direct one-time cash compensation. The compensation method should change how they fare in expectation to the point where they are no longer worse off than the status quo. Suppose you agree with all this. The question arises: how should the compensation be funded? Paul’s post and comment hint at the two competing schools of thought regarding this.

Funding through the migrants

Migrants are the ones who experience the largest per capita gains, and who make the most voluntary decision (by actually moving). Thus, it seems to make the most sense to put the direct obligation of funding native compensation on migrants’ shoulders. This doesn’t mean, of course, that the money is borne completely by the migrant, as any student of tax incidence can tell you. But it still is a tax on migration and all the activities made possible through migration, even if not completely borne by the migrants.

Co-blogger Nathan Smith has blogged about the three types of ways that money may be raised from migrants: auctions, tariffs, and taxes (his own preferred scheme, which falls in the “taxes” subcategory, is what he calls DRITI). Nathan’s goal in the post was a combination of discussing means of revenue generation and addressing concerns about swamping. And therein lies the problem: any attempt to raise money from migrants or migration-related activities curtails migration flows. Some ways of curtailing migration flows are more morally defensible than others. Selecting migrants through a random or opaque process is among the least morally defensible. Having a publicly available tariff or surtax is both more morally defensible and more efficient. A surtax on earnings after one migrates is more defensible than a tariff because it does not exclude cash-constrained migrants and allows people to migrate for free for non-work purposes. On the flip side, the surtax discourages migrants from working, although this effect would be most relevant for migrants who have moved for non-work purposes and are considering work on the side. (It would still be better than the status quo, which generally forbids most temporary immigrants from working unless they have a visa that explicitly allows them to do so. For instance, while people on student visas in the US can work for their universities, their spouses who come on spouse visas cannot work at all). [Incidentally, co-blogger Hansjoerg Walther has a somewhat different proposal, namely immigration-backed bonds. I still haven’t gotten around to thinking through the proposal in depth, so I won’t say anything specific, but read it if you like to consider this sort of stuff.]

Funding through the generally larger economy

Paul Crider’s comment points out a moral problem with funding through the migrants, and suggests the alternative — fund it through the economy at large (emphasis mine):

As I hinted in the post, I’m not overly concerned with the plight of low-skilled rich country natives. I believe it’s necessary to let people down easy when government-promised explicit entitlements are going to be reformed away, as could one day happen with social security or medicare in the US. But wage diminution by immigration liberalization is something else. It’s more similar to deregulation leading to rapid adoption of new technologies, to the detriment of the previously protected firms. It is just allowing creative destruction to do its work, where before it was prohibited. I don’t think morality demands buying off the losers of every new innovation. And to the extent that that is what you want, redistributing from the new blood to the old blood is not the right tool. If you want to soften the vicissitudes of the market, general social safety nets are better tools for the job.

To understand the rationale for this, it might be worth reiterating a few points.

• From the point of view of the natives who lose out, losing out to immigrants isn’t qualitatively worse than losing out due to free trade or technological innovation, and should not be treated differently. If immigration tariffs or surtaxes are a good idea, so are trade tariffs and innovation taxes.
• Immigrants were not causally responsible for the fact that some natives may have enjoyed higher wages than the wage rate under open borders, and they are not morally responsible or culpable for the wage decline that the natives experience. To the extent that migrants are “winners” it is only relative to their earlier (coercively) enforced loser status. The transfer cannot be justified on the grounds of moral culpability or blame.
• In absolute terms, many of the migrants who will pay the huge tariff or surtax will be poorer than the natives who are being compensated. This form of compensation can therefore not be justified on egalitarian grounds.
• Many writers with free market sympathies have argued that companies that hire low-skilled workers and pay them low wages are not morally responsible for giving the workers a wage well above where the market clears, and in so far as there are general moral responsibilities, these responsibilities need to be fulfilled by “society” as a whole, not by the particular employer (see here for instance). A similar argument could be applied in our context: migrants are not morally responsible for making up the wage decline that some natives experience.

Fortunately, the prediction is that liberalizing migration will expand the total economy considerably, and is also likely to increase the per capita income of prior residents on the whole (though probably not by a huge amount). Assuming that certain forms of government spending, such as defense, will not go up in proportion to the population increase, this leaves more room for an expanded social safety net, with the expansion largely targeted at subgroups of the native population that are most likely to lose out in expectation from migration liberalization. Further, since some subsegments of the native population, particularly the richer ones, are likely to benefit more from free migration, taxes on them could be raised by a small amount. Some concrete proposals in this direction:

• Make the tax system more progressive while keeping it revenue-neutral. This effectively means some redistribution from high-earning people (who are likely to benefit the most from low-skilled migrants) to low-earning people (who are likely to benefit the least, and may even lose out). Note that the migrants themselves are low-skilled and end up paying low taxes, unless a separate surtax is imposed on them. A further modification would be to charge a surtax for migrants that brings their taxes up to the level that natives had to pay prior to the change in the progressive direction.
• Provide one-time cash compensation or increased welfare benefits to natives based on income or skill level, funded through greater tax revenues on the economy as a whole (without necessarily changing tax rates). This compensation will not be available to migrants, but otherwise they face the same tax and legal regime as natives.

Where I stand

In this blog post, I considered, in the context of a keyhole solution A, the following three options for which one can have a rank-order preference:

1. Open borders without keyhole solution A.
2. Open borders with keyhole solution A.
3. Closed borders and/or status quo.

Most people sufficiently committed to open borders, or to significant liberalization of migration in general, have a rank-order preference of either (1) > (2) > (3) or (2) > (1) > (3): open borders, with or without keyhole solutions, are preferable to the status quo. This is in contrast with some migrant rights activists who view some keyhole solutions as so bad that they prefer closed borders to the acceptance of such solutions — rank-order preference (1) > (3) > (2) or (3) > (1) > (2), and also in contrast with some keyholists who consider the keyhole solutions absolutely essential before they can even consider liberalizing migration — rank-order preference (2) > (3) > (1) or (3) > (2) > (1); hardcore restrictionists are either (3) > (1) > (2) or (3) > (2) > (1).

For the specific (potential) problem discussed in this blog post, namely, the existence of identifiable subsegments of the native population that are expected to lose out from migration, there are two alternate types of keyhole solutions being offered, namely:

• (2G): Fund compensation of natives who lose out through the general economy.
• (2M): Fund compensation of natives who lose out through migrants.

We can then consider rank-order preferences between (1), (2G), (2M), and (3). My personal rank-order preference is (1) > (2G) > (2M) > (3), which I believe would be the same as Paul’s, judging from his remarks. However, it would more accurately be described as (1) > (2G) > (2M) >> (3) — I think that the differences between (1), (2G), and (2M) are insignificant relative to their difference with the status quo. My relative rank-ordering between (1), (2G), and (2M) is also not too strong. While I do broadly agree with the arguments that Paul laid out, I don’t see this as an issue of comparable moral significance to closed borders. (Of course, when I say (2G) and (2M), I am assuming that the fees or taxes are “reasonable” in the sense of not being so large as to effectively work out to being equivalent to the status quo. It would be hard to quantify this, though, so I’ll keep things vague). Note that my rank-order preference deliberately focuses on permissibility and desirability and does not incorporate considerations of feasibility and stability, which would of course become very important when making proposals for actual political consideration. It is possible that accounting for questions of feasibility turns the rank-order preference to (2M) > (2G) > (1) >> (3), which is (I believe) the rank-order preference espoused by Nathan.

Vipul Naik

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