What I found most interesting in the comments on my latest post and in Bryan’s response to my post was the wide range of views about how many immigrants would come to the United States in just a few years if the U.S. government dropped immigration restrictions except for those with dangerous communicable diseases or a criminal record.

I gave what I thought to be an upper estimate of 300 million in a couple of years. Most people who commented on this, including Bryan, thought it to be too high. It well could be too high, but I found the arguments for why it’s too high largely unpersuasive. Let’s consider them in turn.

. Bryan’s argument: “I think rising rents and declining wages for low-skilled workers would lead to more modest rates. See New York City: Every American is free to move there, but low real wages and high rents convince most of us to live elsewhere.”

Here’s the problem: We’ve been free to move there for a long time. So there’s an equilibrium. People are not free to move to the United States. That’s why we’re having this discussion. So the evidence from NYC is no evidence at all.

. eccdogg’s argument: “The highest immigration rate ever was 1.4% of the population per year from 1847-1854. A time period with pretty much the policy you advocate.
US immigration would have to be over 5 times that rate to have 300MM immigrants come to the country over a 10 year period.
Granted that it is cheaper to immigrate now, but there would be substantial feedbacks lowering wages and thus the incentive to move that would keep it well under that number.”

eccdogg identifies one problem with his own argument but leaves out an important problem. He points out that it’s cheaper to immigrate now: that’s the cost to the immigrant. The thing that has changed in a major way is the benefit. The potential immigrant will weigh this lower cost against the benefit. What is the benefit? Roughly, speaking, the difference (in purchasing power terms) between the unskilled wage here and the unskilled wage at home. This is a multiple of what it was in the 19th century.

Ryan Vann argues: “Seems you have an overactive imagination to me. Average immigration per year is about 2 million. This means that immigration levels would have to increase 75 times that number to get 300 million in two years. No way that happens.”

Touche. But remember, Ryan, that you can’t generalize from immigration flows when immigration is highly restricted to immigration flows when it’s not. Here’s how I can (perhaps overactively) imagine 150 million a year for two years: 70 million from Africa, 50 million from Asia, 20 million from Latin America, and 10 million from Europe, Canada, and Austrailia combined. Then do it again the next year.