Reihan Salam asks me to consider the cost of my revenue-positive immigration plan. In other words, the $750 billion dollars isn’t coming out of thin air. I wrongly took this for granted earlier, but on second thought it brings a fresh perspective to this argument. The explicit costs will clearly fall on the firms and local governments which choose to purchase permits to “import” (for wont of a better word) employees.

I also noted this might create a downward pressure on immigrant wages, which can be considered a “tax” as well:

It’s crucial to note that the burden of permit financing would fall on both employers and employees, depending on elasticities of demand and supply. The dearer the permits, other things equal, the less a potential employer is willing to pay for the same level of output, realized as a lower wage. This can be thought of as a migrant financing his own permit.

Let’s consider a firm operating in an imperfect labor market – as most inevitably are. If the market rate for a permit at a given time is $n, the only immigrants who will be hired are those for whom employers expect the discounted value of all future earnings to be greater than n. For these workers, the costs are explicit $10,000 each. Consumer surplus here is represented by (NPV – n).

The more important welfare loss derives from the implicit cost expropriated on firms that want to hire workers whose NPV is less than the market value of permits. That is governmental intervention prevents an otherwise profitable transaction. In this sense, welfare loss will be roughly proportional to the ratio of firm demand for immigrants to the number of permits supplied on the open market by the government.

This isn’t a mathematically rigorous statement, rather an intuitive heuristic. If demand for migrant workers falls, the permit price on the open market will drop, resulting in fewer excluded transactions. Same logic on the denominator, wherein government can increase the supply to cut permit costs.

By this point it’s clear that the real price of permit auctioning is the cost of closed borders. Standard economic theory tells us that open immigration is the Kaldor-Hicks efficient solution, and any regulation thereof will inflict deadweight losses. At this point, it’s worth comparing (if briefly) my proposal to our current solution.

The deadweight loss comes from the difference between employer demand for migrants and actual cleared licenses. But there are two, huge benefits of auctioning n permits rather than allocating the same number on a first-come-first-serve basis as we do today:

1. The deadweight loss is lower because auctions will almost definitely command a higher quality of immigrant. If firm A wants to bring a highly profitable doctor, and B a similarly “skilled” professional whose just not as competent, the former will be willing to pay more on the open market. This information cannot be captured in any other way. (Nobel Laureate Al Roth has written about how good auction design compels market participants to divulge useful information).
2. Even if the deadweight loss is equal, in my revenue-positive proposal, at least the government captures some of it. In today’s system, all is lost. I suppose more surplus is captured by firms who “make it first” or have large systems that can maneuver government bureaucracy efficiently (like, say, Google).

In fact, there’s a Schumpetrian superiority to permit auctions – over our current system or even the better Canadian “points” program. Byzantine systems requiring “proof of need” etc. give an unfair advantage to large and established players. Startups, the blood of American innovation, are discriminated in the present system. The Canadians and Australians will face a similar problem, if to a much lesser extent. Therefore, the burden of argument is on those who would rather surplus be captured by monopolistic corporations rather than the government.

Though this isn’t the point of my post, I want to conclude with some notes on Reihan Salam’s last remark (which I can’t seem to find, now), which echoed the idea that either of us could design a policy much better than status quo, but that political deliberation makes that impossible.

I don’t think he’s being fatalistic here – he has argued for the Canadian option – but that’s what makes this interesting. It’s hard to argue the American system doesn’t cater to vested interests with regard to immigration. However, what are the “loopholes” really, of the Canadian system? It’s a very transparent, skills-oriented, rubric which cannot be “gamed” in any meaningful way – at least not to my knowledge.

As far as American policy goes, why is my (or whatever his preferred choice is) solution “idealistic” whereas Canada’s is somehow more politically sound? I say this because I’m rather surprised at the paucity of “creative conservatives” arguing for an auction-oriented approach to immigration. I really believe if this idea finds more traction, it’s at least as realistic as a point system. Indeed, both Reihan and I worry more about native wages than overall welfare gain – a topic which puts me at odds with many liberals like Matt Yglesias and deserves a post of its own – and I think few ideas command the “most valuable” immigrants as the quantity-regulated open market.

Theo Clifford has a thoughtful reply to my recent post calling for immigration on the open market:

That said, it isn’t the only way you could do market-based immigration policy. One alternative would be to have an immigration tariff, as advocated by Gary Becker. Instead of auctioning a fixed number of permits, the government would sell as many permits as demanded at a fixed price. The benefits of this kind of system are broadly the same as the benefits of an auction Ashok describes in his post. However, there are a few subtle differences that for me suggest a tariff might be a better way to go.

Now, before we go on, let’s note that standard econ theory would tell us the optimal policy is an infinite number of monthly permits, or a tariff priced at nothing – at which point the two systems would converge into open borders. Before I go through Theo’s argument, it’s important to note that in a government-controlled industry, we can control either price (and let quantity float by demand) or quantity (and let price float by demand).

When it comes to immigration – people – I assert that the latter is wildly more preferable. On what basis do we price immigration? It’s tough to come up with a non-arbitrary algorithm to achieve this. Immigrants, after all, are not “externalities” to be priced and taxed. On the other hand, it’s eminently possible to create a non-arbirary framework through which the number of immigrants are controlled.

A country like the United States may decide that it wants to target an n% labor force growth rate annually. Based on native fertility rate, the monthly auction can be sized to hit this target. Similarly, European countries with below-replacement fertility can target constant population. Therefore the comparison to international trade isn’t fair:

Ultimately, I think the parallel with international trade holds – tariffs are better than quotas. An immigration tariff would be more sensitive to the needs of the market, less bureaucratic, in some ways more predictable, and maybe even more politically palatable than an auction system.

To the extent neither of us are endorsing freely open borders – which would be the Kaldor-Hicks efficient solution – we accept that there are cultural constraints that dictate policy. Frankly, I don’t care how many tons of steel are imported each year, so long as my buildings are efficiently priced. But immigration is a whole different story – these are people you interact with. They will assimilate into your culture, to a large extent, but you into theirs. To abstract the migration of human families into human trade misses the very reason why neither of us advocate free borders (or perhaps he does, in which case a tariff is suboptimal).

But even beside that important point, I think a permit auction holds strong.

The first reason to prefer a tariff is that it means the market is more responsive to shifts in supply and demand. […] Using the auction system, the same number of people come – all that changes is the price. Under a tariff regime, on the other hand, the number of migrants is free to fluctuate according to the needs of the economy […] Yes, the government could adjust the number of auctioned permits according to economic circumstances, but it is better that changes in the number of migrants be market-driven, rather than decided by bureaucratic assessments of ‘need.

I think I’ve addressed most of my concerns with this claim above, i.e. that targeting price level is arbitrary whereas number is not. But even that aside, an auction is sufficiently robust to meet changes in demand. I’m not asking for a once-in-a-lifetime bonanza. The government would host an electronic auction monthly. Furthermore, state and local governments have the explicit right to petition for more permits to be floated on the market in response to acute changes in immigrant demand.

I would argue, too, that permit markets are far more conducive to deeper markets, by encouraging secondary and tertiary exchanges. A big part of my proposal piggybacks on the old, American, Jeffersonian ideal that states are in competition with each other – thereby bringing out the best in each. Michigan, as I noted, has shown a remarkable interest in immigrant labor. Under my proposal, it would be easy for Michigan to buy a block of permits and float them on an internal market open to Michigan’s only businesses. Theo’s proposal is quite similar to a permit market endorsed by Matt Yglesias, but I believe my employer-focused plan places greater emphasis on ensuring only the most valuable immigrants make are allowed. (Note I said “valuable” not “skilled”).

Another argument for tariffs is certainty of barriers to entry:

Having a fixed tariff price, or at least a planned price schedule, also gives a level of certainty to the market. If I live in Mexico and my family is saving up to buy my way into the USA, I want to know how much I’m going to need to squirrel away. With an auction, the price will vary from year to year, perhaps dramatically, and there will be no guarantee that my savings will prove sufficient to get me a permit this year, or even next year. Tariffs solve this problem.

Ask yourself this question. You’re the American president. You have to decide between providing some level of certainty to your own people about what their country will look like in a generation. Or you can provide certainty to some potential Mexican peanut farmer about how much he has to save to maybe make it.

But I’m not even convinced that prices will be so predictable in a tariff system. One of Theo’s main arguments was that the labor market can flexibly respond to changes in demand by allowing more people, but this is a double-edged sword. What if the price is too high, which would strangle innovation and economic growth? Can the government credibly promise business leaders it will keep it high? And what if the price is too low vis-a-vis nativist preferences, which are always in flux? Any sensible government would change the price according to national and business sentiment: but this removes any “predictability” to the whole thing.

Generally, for international trade, I’d support tariffs (if anything at all). But neither price nor quantity is arbitrarily defined. In labor, it makes a lot more sense to target a number than a price. The former can definitely be credibly sustained, with minor exceptions as per immediate request.

Auctioning permits gives the government far more control over its long-term demographic profile, which is ultimately the heart of all immigration debate. There’s a Cato post here that makes pretty much the same points as Theo, but ignores that there’s not something magically more “market” about controlling price than quantity. They (correctly) argue that America needs more immigration but then seem to assume that a tariff would not be overpriced relative to demand but an auction would be. I suppose the old joke about the economist on a stranded island assuming a can opener is appropriate.

Oh and by the way, if we’re extrapolating this debate, I wonder how the folks at Cato would feel about a tariff-based Sovereign debt system rather than an auction. There are scenarios where targeting price is better (international trade) and those where targeting quantity trumps (immigration). It would be wrong to equivocate two very different markets.