The Heritage Foundation has published a controversial paper arguing that “amnesty” for undocumented immigrants will cost the federal budget a net $5 trillion over the long-term, generating a big fight between conservative proponents and opponents of amnesty and higher immigration. There’s a lot I’m not an expert on regarding these issues, but I suspect that caveat applies to many in this debate. For what it’s worth, some thoughts:
- It’s not the number, it’s the sign. Heritage estimates that amnesty would generate a $5.3. trillion long-term increase in the budget deficit. I’m sure I could nitpick that number down a bit. But could I nitpick it to something very small, or to a reduction in the budget deficit? Heritage isn’t the only group that can run numbers; their critics should put their fingers on some calculators as well. The paper surely has faults — any big project does — but are they big enough to shift the overall conclusions?
- It’s not just about the short term. CBO projected that the 2006 immigration proposals would improve the budget deficit over 10 years. But over the first 10 years or so, newly-legalized immigrants will pay taxes but won’t be eligible for government benefits. After that, they can start collecting EITC, Food Stamps, and so forth. Recognize that pattern from anywhere? Conservatives saw the problems with this type of accounting when it applied to Obamacare; they should be just as careful here. It’s the long term that matters.
- Dynamic effects cut both ways. Many commentators argue that the Heritage paper didn’t account for the “dynamic effects” of immigration reform. Yes, immigration will raise GDP, since more workers equals higher output. Of course, due to the lower wages of immigrants, it will also lower GDP per capita. But the more interesting dynamic effects are reduced wages for native workers — particularly those with lower skills. For instance, CBO’s dynamic analysis concluded that “the average weekly wages of workers who had not completed high school (about 12 percent of all adults employed last year) would be about 4 percent lower.” But Republicans connected so well with working-class Americans in the last election that I don’t think they need to worry about these folks.
- Amnesty is almost certainly a loss for Social Security. About half of undocumented immigrants pay Social Security taxes, according to SSA, but obviously can’t collect benefits. Once legalized, they’ll pay taxes and collect benefits. Illegals who currently work under the table will start paying taxes, which helps in the short term, but will collect benefits later. And because these are principally low-wage workers, they’ll collect more in benefits than they paid in taxes. It’s hard to tell a story in which amnesty helps Social Security’s financing.
- No one is correctly modeling the effects of higher immigration on Social Security. Immigrants have lower earnings, higher fertility and lower mortality than native-born Americans; moreover, the children and even grandchildren of Hispanic immigrants also have below-average wages. All of these factors affect the ratios of taxes paid to benefits received and immigration’s overall effect on Social Security’s solvency. CBO models the earnings distribution of immigrants but, as far as I know, not the other factors. SSA, I believe, accounts for none of these differences. Maybe a lot of it comes out in the wash, but you really don’t know until you check.