Census Expects U.S. Population to Start Shrinking by Century's End
New Census population projections, the first since 2017, estimate that the U.S. population will peak in 2080 at nearly 370M and thereafter decline. This marks the first time the Census has projected a population decline as part of its long-term outlook.
According to the latest Census Bureau projections, the U.S. population will stop growing before the end of the century. Under the scenario the Census regards as most likely, the population will peak at just under 370M people in 2080 and then fall slightly to 366M in 2100.
If and when this happens, it will be (almost) without precedent. The only historical year in which the United States has reported a population decline (of -0.06%) before was in 1918, during World War I and the Spanish Flu pandemic. In 2021, at the height of the Covid pandemic, the U.S. population nearly hit stall speed—dropping to only +0.17%. Excluding pandemics, the slowest year of population growth during the 20th century occurred in the depression year of 1933 (+0.59%).
The big difference in the Census’s projections is that we’re not just talking about a single atypical year of negative growth. Rather, we’re talking about a continuous deceleration of population growth that ultimately turns permanently negative. Indeed, that deceleration has already begun. Until 2011, aside from pandemics and the Great Depression, the U.S. population has never grown by less than +0.8%. Since then, it has never grown by more than +0.8%. What’s more, according to the Census’s 2023 projections, the annual growth rate will never again rise to +0.5%. It will fall to +0.3% by 2035; to +0.2% by 2041; to +0.1% by 2051; and then drop below zero in 2081 and after.
These numbers are the Census’s first updates to its population projections since 2017. The projections are based on assumptions about future birth and mortality rates, which don’t change much year-to-year and usually (but not always) trend in the same direction decade over decade. The projections are also based on net international migration, which is much less predictable either year-to-year or in long-term direction.
Over the years, the projections have been adjusted lower and lower amid declining yearly total fertility rates and higher-than-expected mortality rates. In 2012, for instance, the Census projected that the U.S. population would be 420M in 2060. Now the latest projection for that year is 364M, a -13% decrease of 56M people. That’s nearly the populations of California and New York combined.
The projections also highlight just how much the nation’s population growth rate has slowed and will slow further in the coming decades. Annual growth rates have declined from +1.2% in the 1990s to +0.5% today, and will fall to just +0.2% by 2040. By 2038, the number of deaths is expected to exceed births.
This decline in natural increase (births minus deaths) underscores the growing role that net immigration plays in determining positive or negative population growth. For the U.S. working-age population (age 20-65), yearly growth ever since the 2020 pandemic would already be negative without working-age immigration. After 2038, the same will be true of the total U.S. population.
Unfortunately, net immigration is hard to project with any certainty. It depends critically on the relative attractiveness of the U.S. economy and society to everyone living abroad. It also depends on the making and enforcement of U.S. immigration policy, which in turn depend on the shifting mood of American voters. Who can say how these drivers will change over just the next decade, much less over the next several decades? Unlike fertility and mortality, moreover, we possess no baseline social theory telling us whether we should expect net migration to rise or fall over time.