Garett Jones on Open Borders: A Belated Reply
By Bryan Caplan
Last November, Garett Jones wrote two responses to my Open Borders. The first was “Measuring the Sacrifice of Open Borders,” a short paper on the distributional effects of free migration. I replied here.
Soon afterward, however, Garett also wrote me this open letter. Since I didn’t want to hastily respond to serious criticism, I waited until I had time to carefully respond. Now I’m ready. Here’s my point-by-point response. This format works especially well because Garett’s letter is largely a list of questions. Garett’s in blockquotes; I’m not.
As you know, my key disagreement is a theoretical and empirical one: the policy of Open Borders flows fairly naturally from the view that a nation’s level of productivity—total factor productivity or TFP to be pedantic—is largely exogenous to the experiences, backgrounds, and skills of a nation’s citizens. As you know (but for the benefit of the audience), economists call it “total factor productivity” because in some places and times, just about everything is more productive than at other places and times: people produce more, machines are more efficient. All the factors of the economy work better with higher TFP. I don’t think this exogenous TFP view is warranted, but whether or not I think it’s warranted, it’s worth letting other readers know that the assumption of largely exogenous productivity is crucial to the case for Open Borders.
This isn’t quite right. Suppose, like Garett, you estimate nations’ per-capita GDPs as a function of national IQ. Then you simulate what would happen to a nation’s per-capita GDP if immigration lowered its IQ. Does this simulation assume that TFP is “largely exogenous to the experiences, backgrounds, and skills of a nation’s citizens”? Absolutely not. It assumes that TFP controlling for IQ is “largely exogenous to the experiences, backgrounds, and skills of a nation’s citizens,” which is quite a different thing.
The same holds if you estimate nations’ per-capita GDPs as a function of national ancestry (“Deep Roots”). Simulating what would happen to per-capita GDP if immigration reduced your national ancestry score does not assume that TFP is “largely exogenous to the experiences, backgrounds, and skills of a nation’s citizens.” It assumes, rather, that TFP controlling for national ancestry is “largely exogenous to the experiences, backgrounds, and skills of a nation’s citizens,” which is quite a different thing.
Big Picture: The main point of research on the macroeconomic effects of IQ and ancestry is to understand the root causes of TFP. So when I use these very models to estimate the effects of open borders, you cannot fairly accuse me of assuming that TFP is exogenous. Instead, I am showing that – according to your favorite endogenous TFP models – open borders would have amazingly good effects.
Specific models aside, Open Borders does not casually “assume” that migration will leave TFP unharmed. Instead, I investigate the main channels through which migration plausibly harms TFP, and conclude that there is, at best, little evidence of harm.
Thus, since many critics worry that immigrants will reduce TFP by voting for bad policies, I examine the extent to which immigrants favor different policies than natives. The same goes for cultural and fiscal objections to immigration. True, I don’t use the phrase “Total Factor Productivity” in the main text of Open Borders. Nevertheless, my book contains three full chapters that hunt for ways that immigrants might harm TFP. As you know, I don’t find much – and it’s not for lack of looking.
Admittedly, there is one TFP-relevant mechanism that I barely discussed in the book. Namely: How open borders would markedly enhance TFP growth by (a) integrating the world’s greatest minds into the global economy, and (b) raising global IQ by swiftly eliminating severe childhood deprivation. In hindsight, this was a grave omission on my part.
Related assumptions float around in the other arguments for Open Borders that I’m aware of: there’s a routine assumption that some nations are exogenously more productive than others. There might be some attempts to explain those differences, but these explanations discuss little more than proximate causes like “institutions” without explaining why institutions differ across countries.
I disagree. The “routine assumption,” rather, is that when critics of immigration claim that immigrants will harm productivity or “damage institutions,” we should examine the extent to which their claims check out. Most of Open Borders does precisely this. Punchline: Many of the claims don’t check out at all; the rest are mild, especially compared to critics’ hyperbolic rhetoric.
It’s true that I don’t examine amorphous claims that immigration will harm productivity in some totally unspecified way. Given the many decades human beings have debated and studied immigration, it’s unreasonable to put much weight on such fears. I assume you’ve heard of Pascal’s Mugging?
Sometimes geography is mentioned as well.
Indeed. Is this mechanism somehow suspect? In long-term growth regressions, the benefits of good geography turn out to be markedly more robust than the ancestry channels that you’ve emphasized.
In his Lucas Critique, Nobel laureate Robert Lucas argued that it’s hard to offer good economic policy advice unless you know the true structure of the economy.
To be frank, this is a silly claim. Do we have to “know the true structure of the economy” to advocate the legalization of Uber? Pollution taxes? Housing deregulation? Should governments have prevented the Industrial Revolution or the Information Revolution until they “knew the true structure of the economy”? Yes, “The bad indirect effects of a seemingly good policy might overwhelm its direct effects” is always logically possible. Yet when critics offer no specific reason to think that a seemingly good policy has strongly negative indirect effects, I say the “good economic policy advice” is “Ignore them.”
Do you think you have a good model of the true structure of the global economy, of why some countries are more productive than others? If so, what is it?
Tell me if this counts as a “model.” All of the following are independent fundamentals:
1. The accumulated body of technological and managerial knowledge.
2. Elite support for free-market policies and rejection of very bad policies.
3. Popular support for free-market policies and rejection of very bad policies.
4. National IQ.
I don’t know if my model is “good,” but it’s good enough for practical purposes.
Do you generally agree with my claim that Open Borders has much in common with the Real Business Cycle literature, where exogenous, mostly unexplained productivity shocks that last a year or so explain booms and recessions?
No. I say the “shocks” that matter for long-run national prosperity are fairly well-explained, long-lasting, and self-perpetuating.
What would serious explanations of cross-country productivity differences look like—and do those explanations have little to do with the backgrounds, experiences, and skills of each nation’s citizens?
See above. The last piece of the puzzle is that, with few exceptions, first-generation immigrants assimilate enough to avoid messing up functional systems, and second-generation immigrants assimilate almost fully. I know you dismiss this position as ill-informed or wishful thinking, but I obviously disagree.
By now quite a few people-based theories of national total factor productivity exist: my own Hive Mind theory, but also Dohmen et al.’s “Patience and Comparative Development” along with a variety of Deep Roots stories where the migration-adjusted past is prologue. How plausible are these sorts of explanations of why total factor productivity differs across countries?
I think your national IQ story is part of any good explanation, with the caveat that low IQ in poor countries is, in the long run, at least 40% endogenous. Probably more like 60%.
I spent over a month reviewing research on ancestry (“Deep Roots”). In the end, I think there’s not much there. Hive Mind‘s story is much better.
By contrast, on a scale of, say, 0 to 10, with 10 being metaphysical certitude with no caveats, how plausible is what I understand to be the Open Borders view: that aside from some transition period, we can mostly treat a nation’s level of total factor productivity as exogenous?
You misunderstand the Open Borders view – or at least my Open Borders view. I strongly disagree with the position you attribute to me; I’d give it less than 1/10 (where 0 is “metaphysical impossibility”).
In Jonesian terminology, what my book really says is:
1. We can make TFP endogenous in either of your favorite ways by treating it as a function of national IQ or national ancestry, and the standard extremely optimistic estimates still hold.
2. Cultural assimilation is an endogenous process, so when people migrate from less-developed countries to more-developed countries, their kids end up closely resembling their native peers.
3. The causal effect of poverty on IQ is yet another endogenous process. When people from low-IQ countries have kids in high-IQ countries, their kids end up much smarter than they would have been if they had grown up in their parents’ country of origin.
Now if you asked me, “on a scale of, say, 0 to 10, with 10 being metaphysical certitude with no caveats, how plausible is the Open Borders view: that the migration of billions of people from the Third World to the First World over the next fifty years would durably raise Gross World Product by 50% or more,” I’d give you at least a 9/10.
Let me close with one question for Garett. To the best of my knowledge, you have offered no reason why the U.S. should not have open borders for anyone above its current average IQ of 98.
On your own terms, this liberalization should be just as awesome as I claim; maybe better. Yet as far as I can tell, you have failed to endorse this reform.
Will you do so? If not, why not?