Trillion Dollar Bills on the Sidewalk: The Borjas Critique
George Borjas‘ new Immigration Economics contains the first intellectually serious critique of the increasingly mainstream view that open borders is a big stack of “trillion dollar bills on the sidewalk.” Borjas begins by clearly explaining what’s at stake.
[W]hat types of gains would accrue to the world’s population if countries suddenly decided to remove all legal restraints to international migration and workers moved to those countries that afforded them the best economic opportunities? In contrast to the immigration surplus calculated for the receiving country’s native population in the previous sections, it turns out that the “global immigration surplus” is huge and seemingly could do away with much of world poverty in one fell swoop.
The critical variable is R, the First World/Third World wage ratio for equally-skilled labor:
As in the original Hamilton-Whalley (1984) study, the exercise reveals that the gains to world income are huge. If R=2, for example, world GDP would rise by $9.4 trillion, a 13.4 percent increase over the initial value of $70 trillion. If R=4, world GDP would increase by $40 trillion, almost a 60 percent increase. In fact, if R were to equal 6, which may be near the upper bound of the range of plausibility suggested by the available data, world GDP would rise by $62 trillion, a near-doubling. Note, moreover, that these gains would be accrued each year after the migration occurs, so that the present value of the gains would be astronomically high.
The Borjas critique:
Even putting aside the political difficulties in enacting such a policy [open borders], this argument in favor of unrestricted international migration glosses over two conceptual obstacles.
First, the calculation assumes that people can somehow start at a specific latitude-longitude coordinate and end up at a different coordinate at zero cost… The absence of legal restrictions prohibiting the movement of people from one country to another does not circumvent the fact that it would be very costly to move billions of workers.
As noted in chapter 1, large wage differences across regions can persist for a very long time simply because many people choose not to move. In a world of income-maximizing agents, the stayers are signaling that there are substantial psychic costs to mobility, perhaps on the order of hundreds of thousands [sic] dollars per person… Kennan and Walker (2011 p.232) for instance, estimate that it costs $312,000 to move the average person from one state to another within the United States…
Although these costs seem implausibly high, moving costs must be around this order of magnitude to account for the observed fact that people do not move as much as they should given the existing regional wage differences. If moving costs were indeed in that range, it is easy to show that the huge global gains from migration become substantially smaller and may even vanish after taking moving costs into account. [emphasis original]
Borjas then does some back-of-the-envelope calculations and concludes:
The “breakeven” cost of migration given in the last row of Table 7.3 is around $140,000. In short, the entire present value of the global gains is wiped out even if the costs of migration were only half of what is typically reported in existing studies.
Qualitatively, Borjas’ argument is entirely true. Human welfare rises by less than GDP because of material and psychic relocation costs. But quantitatively, Borjas’ argument is ludicrous. Yes, he’s seriously using the average valuations of Americans to estimate the marginal valuations of Third Worlders! Yet any decent econ undergrad can tell you that:
1. The marginal migrant minds moving less than average. Indeed, given the strictness of the current regime, many marginal migrants would probably take a wage cut to exit their homelands. Think of every Iraqi Shiite who can’t sleep tonight because the Sunnis are coming, and every Iraqi Sunni who can’t sleep tonight because the Shiites are coming.
2. Third Worlders are almost certainly willing to pay vastly less than Americans to stay in their homelands because location is a normal good. Indeed, location is probably a luxury. Does Borjas really think that most Haitians would forego $140,000 in income because they’re in love with Port-au-Prince?
An excellent econ undergrad might add that:
3. Due to diaspora dynamics, psychic relocation costs endogenously fall over time. The more migrants there are, the easier it is to say adios to your country of birth. Forget the Maine, but remember Puerto Rico.
Points #1 and #2 are so basic that I reviewed all of Borjas’ attendant footnotes to see if he addresses them. He grudgingly accedes to #1 in footnote 22:
Only a subset of persons in the data are actually observed to move, so that the subsample of movers may have moving costs that differ substantially from (and may be much lower than) the “average” estimates for hypothetical movers.
But to the best of my knowledge, Borjas never even hints at #2 – a bizarre omission given his childhood flight from Cuba. He does however further undercut his critique in footnote 20:
It is worth noting that a disproportionately large fraction of the global gains can be accrued even if only a fraction of the potential movers migrate to the North. For example, global GDP would increase by around 17 percent when 10 percent of the potential movers move (assuming R=4).
At this point, you may be asking, “Wait, didn’t Borjas promise us two ‘conceptual obstacles’?” He did. Sadly, his second is a throwaway objection with a single citation.
[T]he gains reported in Table 7.3 depend crucially on the assumption that the intercepts of the labor demand curves in the North and South are fixed. However, the North’s demand curve lies above the South’s demand curve, not simply because that is just the way things are, but because of very specific political, economic, institutional, and cultural factors that endogenously led to the development of different infrastructures in the two regions…
As the important work of Acemoglu and Robinson (2012) suggests, “nations fail” mainly because of differences in political and economic institutions. For immigration to generate substantial global gains, it must be the case that billions of immigrants can move to the industrialized economies without importing the “bad” institutions that led to poor economic conditions in the source countries in the first place. It seems inconceivable that the North’s infrastructure would remain unchanged after the admission of billions of new workers. Unfortunately, remarkably little is known about the political and cultural impact of immigration on the receiving countries, and about how institutions in these receiving countries would adjust to the influx.
That’s all Borjas has to offer. Is he really unaware that plenty of research on the political consequences of immigration is already out there? (See Gochenour and Nowrasteh’s literature review for starters) Is it really so difficult to picture major mitigating and countervailing factors? And why would one of the world’s foremost immigration scholars bemoan our “unfortunate” ignorance of “the political and cultural impact of immigration on the receiving countries” instead of rolling up his sleeves and investigating the issue? Yes, we all have a limited time budget, but isn’t the prospect of “doing away with much of world poverty in one fell swoop” slightly more pressing than, say, measuring the effect of migrant Soviet mathematicians on academic mathematics?
To be blunt, I suspect that Borjas prefers to remain agnostic about the political and cultural effects of immigration. That way, no matter how mighty the economic case for open borders, he’ll never have to say, “Good God, how could I have been so blind? There’s a whole stack of trillion dollar bills right there on the sidewalk!”
P.S. I’ve scheduled my first Open Borders Meet-Up at my house on August 3. All friends and fellow travelers of open borders are welcome to email me for an invitation. Pursuant to my writings, avowed immigration restrictionists may only attend with my explicit permission.