Immigration— The Congressional Budget Office Report (CBO on S.744)
I will post my little paper on the effects of immigration later today. In preparation for that, I downloaded the CBO report on the Senate bill. I find it unimpressive, except for the clarity of its writing style, which I do commend. Some notes:
1. The analysis assumes that immigration will raise total factor productivity. It alludes to increased specialization because of increased size of the economy, and increased innovation because of having extra high-skilled immigrants. Both are highly speculative. Yes, more Hyderabad programmers will increase innovation; that’s why the billionaires support the bill (they profit from those innovations— which is fine). But it’s dangerous to allow a fudge factor like “increased productivity from innovation” into an estimate— it can drive too much. It would be good to see these estimate with zero change in TFP.
2. I didn’t see anywhere how many of these immigrants are supposed to be high-skilled. I suspect that’s because it’s embarassingly small. If the report came out and said that we’d be adding 9.5 million low-skilled workers and .5 million high-skilled workers and productivity was going to rise as a result, people would laugh.
The report doesn’t have details of whether it separates out skills of labor in its computation of total factor productivity, but my guess is that it doesn’t. If that is true, saying total factor productivity will rise is really a joke.
3. It’s definitely this increase in TFP that is driving the wage increase estimate for 2033, as the report, to its credit,admits. I see that two very good academic economists, Lawrence Katz and George Borjas, provided comments. Maybe they agreed to not to make fun of the report if the CBO put in the right caveats like that. Or maybe the CBO gets the credit. Anyway, notice what this is saying: there is going to be a big increase in the amount of labor in the economy, so big that the return to capital is going to rise and people will invest in increasing the capital stock because of this extra profitability. But at the end of the day, with more labor, and, especially, more *unskilled* labor, the average wage is going to rise. Put a bit differently, if you increase the quantity supplied of a good and reduce the quality, the claim is that the price will rise. Nope.
4. It would be deep in the background analysis, but I wonder whether the report is assuming that the capital/labor ratio will return to its present state eventually because of the increased investment. That’s something to worry about, because I’d expect that the long-run equilibrium interest rate would rise because of the extra labor. Think of it in the extreme to understand why. If we multiplied the labor force by 10, then the return to capital would go way up as wages fell. Suppose the return on the stock market rose from 6% to 20%. Savings would rise because of the high return, so capital would increase, driving down the return from 20% to 18% to 16%,e tc. WOuld it go all the way back to 6%? No. Savers were only willing to have so much savings because the return was 20%. If it went all the way down to 6%, they would unsave, spending all that extra capital. So the interset rate should settle down to, say, 10%. THis is important, because if the K/L ratio ends up smaller than it is today, wages end up lower too.
5. On p. 3 there’s a funny switch from using increases in GDP to using changes in GNP per capita. The footnote explains that GNP is a better measure of national welfare— but why not use it throughout, then?
The footnote is correct because GDP measures what’s produced in the US, whereas GNP measures what’s produced by labor and assets owned by Americans. The footnote says, correctly, that GNP includes earnings by US citizens overseas but excludes earnings by foreigners on their investments in the US. That’s misleading, though. It leaves out what is the key thing with immigration: US GNP excludes remittances by immigrants to their country of origin. I wonder if they include that important reduction of US GNP in their estimates? If immigrants send their paychecks home to Mexico or save them and retire to spend it in Honduras, what they earn is part of US GDP, but it has to be subtracted to get GNP.
6. Point 5 is related to a fundamental problem: The report does not look at how the bill affects the welfare of Americans. Instead, it looks at how it affects output in the US, average wages, return to capital, and so forth. A lot of the effect is on foreigners— Mexican immigrants and European investors.
7. If they’d played it straight, here is what we’d expect from basic economics. The bill will allow a lot of low-skilled foreign labor to enter the United States (or remain here, if we think of the alternative as being enforcement of existing laws). This will drive down low-skilled wages and drive up the return to capital. The effect on middle- and high-skilled wages is unclear, but my guess is that it would increase them. In addition, the immigrants will compete with poor natives in the market for consumer goods and will drive up prices of items with limited supply— for example, low-quality housing. The dollar benefit to capital-owners will exceed the dollar loss to labor. Since some of capital is owned by foreigners, however (Chinese govt., Europeans, etc.), it might be that the benefit to Americans will be less than the loss. Foreign labor will benefit to the extent that US wages remain above foreign wages.
8. All of that pertains to the effects on US markets. What happens to government taxes and expenditure is a separate topic, equally important. Adding lots of poor people to the economy is not going to help national or local budget deficits, whatever handwaving may be going on.
9. This is all on the level of a blog post, or of what I’d have said to the authors of the CBO report if they asked me to comment. The report doesn’t include the background mathematical modelling, and maybe that can address some of these problems. Or maybe I’ve missed something in my reading of it. But I will stand by the basics of supply and demand: the burden of persuasion is on those who argue for the surprising claim that adding more poor people to America will increase American prosperity.