One of the more frustrating aspects of the CBO analysis of the Senate immigration bill is that while it provides an estimate of the bill’s impact on the size of the U.S. population, it doesn’t provide much detail regarding the baseline on which it builds:
CBO estimates that, by 2023, enacting S. 744 would lead to a net increase of 10.4 million in the number of people residing in the United States, compared with the number of people projected under current law. That net increase comprises an increase of about 10.4 million permanent residents; an increase of about 1.6 million temporary workers and their dependents; and a decrease of about 1.6 million unauthorized residents. (CBO estimates that about 8 million unauthorized residents would initially gain legal status under the bill, but that change in status would not affect the size of the U.S. population.)
I assume that the net increase of 10.4 million can be attributed entirely to immigration, unless the CBO is also taking into account larger immigrant family sizes. For the sake of argument, let’s say the CBO envisions that the Senate bill will lead to a net increase in the number of immigrants. This is in addition to the immigration level the CBO already anticipates in the absence of the legislation. I don’t have access to the baseline number of immigrants the CBO assumes will settle in the U.S. between now and 2023, but Yuval Levin kindly provided me with a March 2011 CBO report detailing the assumptions behind the CBO’s labor force projections, including its assumptions regarding immigration levels (and I’m more generally grateful to Yuval for pointing me to the volume of immigration the CBO assumes will follow from the Senate immigration bill):
The Congressional Budget Office (CBO) relies on its own projections of net immigration, in which total net flows rise from about 440,000 in 2010 to more than 2.3 million in 2015, then fall to about 1.3 million by 2020. CBO’s estimate for 2010 is much lower, and its projection for 2015 much higher, than are predictions from the Social Security Administration (SSA), the Census Bureau, or the Social Security Advisory Board’s Technical Panel on Assumptions and Methods; the differences arise largely because CBO’s projections explicitly take into account the impact of the recession and subsequent recovery, whereas the other forecasters do not. By 2020, CBO shows somewhat larger net inflows than SSA but somewhat smaller net inflows than the Census Bureau and the Technical Panel.
More details follow, but my rough guess would that the CBO is assuming that net immigration over the coming decade will be in the neighborhood of 10 million under current law. And so a net increase of 10.4 milion as a result of the immigration bill as compared to current law would mean that the U.S. immigration from 2013 to 2023 will be in excess of 20 million. I should stress that I’m not sure if my interpretation of the CBO analysis is correct. It is possible that my numbers are way off, and fortunately a number of crack reporters, including Brad Plumer of Wonkblog, are also trying to get to the bottom of this, so consider my take on this subject provisional.
The Census projects that the U.S. population in 2023 will be roughly 341 million. Let’s say the CBO is assuming that the population will instead be 351 million, of which 20 million reflects post-2013 immigration. This means that 5.7 percent of the U.S. population will consist of post-2013 immigrants as of 2023. By way of comparison, Matt Yglesias of Slate recently wrote on the historical data concerning how the foreign-born share of the U.S. population has changed over time. The foreign-born share peaked in 1890 at 14.8 percent and it was at its nadir in 1970, when it hit 4.7 percent. It has been increasing dramatically in the decades since, and it reached 12.9 percent as of 2010. The CBO anticipates that immigrants who will have arrived since 2013 will represent a larger share of the U.S. population in 2023 than all immigrants in 1970 or 1960 (5.4 percent), and only a slightly smaller share than the 1980 number (6.2 percent).
It would be foolish to simply add this 5.7 percent to the 12.9 percent 2010 share. But it seems reasonable to guess that we are looking at a foreign-born share of the U.S. population in the neighborhood of 16 to 18 percent, a level higher than the previous 1890 peak. The foreign-born share of the U.S. population remained quite high during the first decades of the twentieth century before drifting down by midcentury (14.7 percent in 1910, 13.2 percent in 1920, 11.6 percent in 1930, 8.8 percent in 1940, 6.9 percent in 1950). What is striking is that, as Matt acknowledges, the most egalitarian era of modern U.S. history coincided with a very low foreign-born share of the population. Matt argues, not unreasonably, that the failure to take into account the poverty-reducing effects of international migration is myopic. I’ve addressed how I balance the competing concerns of global poverty alleviation and facilitating the assimilation of current immigrants and combating entrenched poverty among the native-born elsewhere, and of course Matt and I disagree about the virtues of less-skilled labor in an economy like ours more broadly.
But for now I’ll just say that if you believe that culture has important political and cultural implications, the prospect of such a rapid increase in the foreign-born share of the population ought to be cause for concern. It is worth noting that the foreign-born share varies across U.S. states. California (27 percent), New York (22 percent), and New Jersey (21) already have much higher foreign-born shares than the U.S. will have under the scenario outlined above. What is interesting is how the texture of civic life might change in other communities, where the foreign-born share is currently relatively low. Leaving geography aside, one wonders how U.S. politics might evolve as the foreign-born share of the working-age population grows relative to the retired population. Rising public expenditures associated with Social Security, Medicare, and other programs devoted to the old might engender resentment among the young, particularly when cultural cleavages line up with generational cleavages.
And then there is the matter of the skill distribution. The following is drawn from the CBO report on the economic impact of the Senate immigration bill:
S. 744 would allow significantly more workers with low skills and with high skills to enter the United States—through, for example, new programs for temporary workers and an increase in the number of workers eligible for H-1B visas—and would allow somewhat greater numbers of workers with skills in the middle of the distribution to enter as well. JCT expect that a greater number of immigrants with lower skills than with higher skills would be added to the workforce, slightly pushing down the average wage for the labor force as a whole, other things being equal. Taking into account all of those flows of new immigrants, CBO and JCT expect that a greater number of immigrants with lower skills than with higher skills would be added to the workforce, slightly pushing down the average wage for the labor force as a whole, other things being equal. [Emphasis added]
It is widely believed that as the the size of the retired population increases, and as the cost of medical care increases, that the resources devoted to old-age social insurance programs will have to increase, and that this will necessitate increases in the lifetime net tax rates of younger workers. If a disproportionately large share of the workforce consists of younger workers with modest levels of educational attainment, a substantial tax burden might prove very problematic. A short while ago, a friend pointed me to Robert Dahl’s brief discussion of Argentina during the first half of the twentieth century. I wouldn’t rely too heavily on Argentina’s experience to guide us as we go forward, but it is a reminder that dramatic changes in the composition of a population can have dramatic cultural and social consequences.