Sumner's Switzerland

When I claim that "migration is not euvoluntary", I mean to acknowledge the BATNA disparity that tends to exist across national borders. Many countries have poor political and economic institutions, low productivity, appalling civil rights (both de jure and de facto), and high rates of poverty. Very often, migration restrictions mean that willing migrants find themselves obliged to accept the worse of the but-for-immigration-quotas options.

Scott Sumner recently addressed some of Bryan Caplan's claims about the benefits of open migration by example. ATSRTWT. A slice:

I'd rather explain this problem using Switzerland, as it will be easier to make my point. Switzerland has 8 million people living in a small mountainous country. The built up environment is very neat and tidy, and indeed often quite beautiful. It's safe. I once accidentally left a wallet with lots of cash clearly visible on a ledge in a Geneva train station. When I came back later it was still there, with the cash. That doesn't happen in lots of other countries. With open borders Switzerland might attract tens of millions of immigrants. It would no longer be a "land of the Swiss." It would be much more crowded, much poorer, less well-educated, more crime-ridden, etc. Culturally it might become more like Kyrgyzstan, another small mountainous country. The Swiss greatly value the neat and orderly characteristics of their country. I find it quite plausible that they would feel worse off.

Thankfully, Sumner does not commit the uneconomic error of claiming fixed-sum labor transfer (dey took our jerbs). But his other claims bear closer inspection. Since I don't know what it means, I'll ignore the "etc." in there after "crime-ridden." Everything else should be fair game.

On congestion. Complaints about crowding are difficult to square with the prosaic observation that net migrations tend to be toward cities rather than away from them, even within nations (see R. Lucas's famous 2004 JPE piece). Congestion is a bit like pollution: it's an unwanted by-product of a desirable process. I have no doubt that visitors to Zurich would prefer a more sparsely-populated Switzerland—charming views of the mountains and the brisk blue sky unrent by the serration of skyscrapers are part of the appeal. But I urge anyone thinking about this closely to consider Adam Smith's habit of the mind: exercise analytical sympathy. Resist the urge to apply tourists' aesthetics to residents' best interests. Would you have Lower Manhattan vacated because you preferred emptier streets on your stroll to Central Park or do you recognize that the people of New York are made better off because other people live there? For tourists, crowded is a bug. For residents, it's a feature.

On poverty. n = 3 people have w1 = $100 each. Their average wealth is Σ(w1)/n = $100. Three more people show up with w2 = $20 each. Average wealth is Σ(w1, w2)/n = $60. Average wealth has fallen by $40. But the three original people are no poorer than they were before. Simplistic? Yes, but that's basically how national income is counted. It is my sincere hope that an economist as thoughtful as Scott Sumner isn't making this sort of elementary error, conflating national income accounting statistics with actual flesh-and-blood humans. To make the claim that Switzerland would be poorer, it's wise to distinguish between incumbent citizens and immigrants, and then to show what impact immigrants would have. For the incumbent citizens to be actually poorer, you'd have to disprove the First Welfare Theorem (surely a Nobel-worthy accomplishment) or you'd have to demonstrate how the median immigrant is a net [discounted] lifetime tax consumer rather than a net [discounted] lifetime tax payer. I admit a lacuna in my own knowledge here, as I'm insufficiently familiar with existing Swiss statutes on transfer payment eligibility for immigrants. I do know from work by friends of EE Zac G. and Alex N. that immigrants to the US are emphatically in the latter category: even in spendthrifty California, immigrants are still net lifetime taxpayers.

On education. First-generation immigrants tend to be less well-educated than the native-born population. Second-generation immigrants tend to be better educated than their more established peers. At least in the US. Please contact me at spivonomist@gmail.com if you'd like to see my data and regressions.*

More crime-ridden. No. First-generation immigrants are marginally less prone to commit crime than incumbent citizens (h/t M. Clemens). I think the reason this myth persists is because the average crime rate in net emigrant countries is higher than net immigrant countries, but this confuses averages with margins. The marginal emigrant is not the average resident. The marginal immigrant tends to have lower discount rates (evidenced by household savings rates and remittance payments), higher conscientiousness (again, see high educational attainment for second-generation immigrants) and when controlling for other factors such as language acquisition and education, higher IQ (again, contact me for GSS data & regressions). The myth of the criminal immigrant is just that: a myth. The laws that "illegal" immigrants break are the ones written specifically to target them. In a nation with institutions that refrain from indulging in dominion and discrimination, the median immigrant would be more law-abiding than the median native-born citizen.

On culture. When I was a boy visiting the DC Metro area, if I wanted Chinese food, I'd have my choice of "Hunan" or "Hunan". It was wretched. Not so anymore, and that's just food, a frippery. American culture is enriched by the people that voluntarily, joyously, felicitously choose to call it home. However, this is already wordy enough, so I'll stow a more detailed critique of the cultural arguments for another day.

*[edit] Second-generation in this context is defined as people born in the US, whose parents were also born in the US, but whose four grandparents were all board abroad. n=2213