We hear it all the time. During the recent primary debate, Joe Biden said illegal immigrants “increased the lifespan of Social Security.” America must have large numbers of immigrants to “rebuild the demographic pyramid,” in the words of Jeb Bush.
To be sure, Americans have had small families for quite some time, and life expediency has increased. Our population is aging. But there is a significant amount of research on how much immigration can offset population aging in low-fertility countries such as the United States, and the answer is clear — not much.
In a 1992 article in Demography, economist Carl Schmertmann explained that, mathematically, “Constant inflows of immigrants, even at relatively young ages, do not necessarily rejuvenate low-fertility populations. In fact, immigration may even contribute to population aging.” After reviewing population projections, the former chair of Princeton’s sociology department and the director of its graduate population-studies program, Thomas Espenshade, observed:
It becomes apparent that the effect of alternative immigration levels on population age structure is small, unless we are willing to entertain a volume of U.S. immigration of historic proportion. . . . Immigration is a clumsy and unrealistic policy alternative to offset a shortage of domestic labor or to correct a perceived imbalance in the pensioner/worker ratio in the United States.
After looking at all the population projections prepared by the United Nations, Oxford demographer David Coleman has concluded, “There are no feasible migration solutions to the age-structure change and its effects on social security.” Coleman and others have pointed out that immigration can prevent population decline — that is, it can add a lot of people to the country — but it does not significantly change the age structure in the way that many immigration advocates seem to imagine. If we wanted to use immigration to offset population aging, the level necessary would have to be truly enormous.
A recent paper I coauthored based on the most recent Census Bureau population projections examined the impact of immigration on the nation’s age structure. Assuming current levels of immigration continue, the latest projections indicate that the total U.S. population will reach 404 million in 2060 — 79 million larger than in 2017. Future immigrants and their descendants account for nearly all (75 million) of the increase. Under this scenario, 59 percent of the population will be working-age (16 to 64). By contrast, in a zero-immigration scenario, 57 percent of the population would be working-age in 2060. More realistically, if immigration were limited to half of the expected level, 58 percent would be working age.
Part of the reason immigration has such a modest impact on the working-age population is that, while it certainly adds a lot of new workers, it also adds children and retirees over time. Most people will recognize why a larger senior population increases government expenditures, but a larger child population does as well. The United States spent $668 billion just on public schools in the 2016–2017 school year.
Of course, the children of immigrants eventually grow up and become workers, but by the time this happens, a significant share of their immigrant parents will have reached retirement age. These two trends tend to offset each other over time. As a result, immigration does not have much of an impact on the share of the population that is working-age. Even if we exclude children from the analysis and focus only on the ratio of working-age people to retirees (age 65-plus), the impact of immigration is still modest. There will be 2.5 working-age people per retiree in 2060 if immigration unfolds as the Census Bureau projects; there will be 2.3 workers per retiree if immigration is cut in half. The impact, while not trivial, is modest.
In fact, it is likely that the current Census Bureau projections, on which our analysis is based, overstate the impact of immigration on aging. In a new paper authored by myself and Karen Ziegler, we find that immigrants are coming to America at much older ages than in the recent past. The Census Bureau’s projections do not fully reflect this change. The average age of newly arrived legal and illegal immigrants has increased from 26 in 2000 to 31 in 2017. The share of newly arrived immigrants who are 55 or older doubled from 5 percent to 12 percent. In other words, about one in eight new immigrants are now arriving in America old enough to move directly into a retirement community.
Several factors likely explain the rising age of new arrivals. Fertility is declining and life expectancy is increasing all over the world. Older populations in sending regions translates into older immigrants coming here, at least to some extent. Equally important, under current law any citizen can sponsor a parent, and there are no numerical limits on this immigration category. In the last three decades 1 million people per year were given green cards — permanent residence — which allows for citizenship after five years. As a result, the population of naturalized citizens has roughly doubled in the last two decades, and so has the number of people entering in the parent category. The significant increase in the age at which immigrants arrive makes it likely that the Census Bureau projections overstate the modest impact of immigration on population aging.
What would it take to use immigration to prevent population aging? Our analysis shows it would take a fivefold increase in immigration levels over what the Census Bureau foresees to maintain the working-age share and ratio of workers to retirees. This would create a total population of 706 million in 2060 — more than double the current population. Most U.S. residents would be post-2017 immigrants or their offspring. Of course, immigration at this level would profoundly affect the culture, politics, and social fabric of the United States in dramatic and perhaps unforeseeable ways.
So what is the solution to an aging society? Commenting on our population study at a panel discussion in March of this year, American Enterprise Institute demographer Nicholas Eberstadt put it succinctly: “Immigration cannot possibly over the long run maintain a youthful population structure.” He added that “raising the age of retirement has a bigger bang.” David Coleman makes much the same point in his analysis. Increasing the retirement age by two years has about the same impact on the working-age share or ratio of workers to retirees in 2060 as does the immigration-induced addition of 75 million people projected by the Census Bureau. If the retirement age were increased to 70 it would nearly maintain today’s working-age share of the population through 2060. This change certainly would take some getting used to, but it seems far less disruptive to society than adding hundreds of millions of people.
Another way to deal with the decline in the working-age population is to increase the employment rate. At present, the employment rate for those 16 to 64 is about 70 percent, which is low by historic standards. Bumping the working-age employment rate up to 75 percent would do more to change the share of the population working than 75 million immigrants and their progeny would. It is not unrealistic to raise the employment rate, as it was 74 percent as recently as 2000.
One could favor liberal immigration policies for many reasons, but it is no fix for an aging society. We will have to look elsewhere to deal with the challenges associated with population aging. Raising the retirement age and getting more working-age people back to work are much more efficient in dealing with the issue.