Journal of Commerce, April 19, 1989
A major Census report in early 1989, raising the possibility of a decline in the U.S. population in the second half of the next century, has triggered renewed warnings that the United States needs to act now to increase fertility and raise immigration.
But it would be reckless for the government to commit itself to population expansion on the basis of tenuous long-range predictions or unclear needs. Pro-natalist programs and new immigration incentives can be expensive and disruptive and have long-term cumulative social and demographic side effects. Once started they are hard to turn off.
Flawed population forecasts are not unusual. Herbert Hoover's experts saw the U.S. population topping out at about 150 million in the 20th Century. And Franklin Roosevelt's technocrats markedly underestimated population growth between 1935 and 1980.
Before rushing toward hasty population policies we should be clear about what the recent Census report is saying. First, it does not predict population trends, it projects them. The difference is important. Projections are the mathematical outcome of assumptions about future trends in fertility, mortality and net migration.
While the projections are by definition correct, the assumptions underlying them often require revision. President Roosevelt's demographers could not have foreseen that the birth rate would rise continuously between his administration and John F. Kennedy's, or that it would fall abruptly with Lyndon Johnson.
Then, the Census report projects 30 different population outcomes based on combinations of high, medium and low assumptions. But the media and armchair demographers have concentrated on only one, the middle scenario, as our most likely population future. The "most likely" scenario assumes continued fertility of 1.8 per woman, life expectancy of 81.2 years, and net annual immigration of 500,000. Under those circumstances, the U.S. population would peak at 301 million in 2040 and gradually fall to 292 million in 2080.
How likely is this "most likely" scenario? Historically, immigration rates, like fertility, have been volatile. Persistent illegal immigration and refugee flows and rising demand for family reunification immigration suggest that the Census Bureau's high assumption of 800,000 annual immigration comes closer to current reality. Similarly, steady improvements in health care portend a life expectancy in the next century closer to Census's high assumption of 88 years.
Small changes in assumptions yield vast changes in population outcomes. A modest rise in fertility to replacement level of 2.2 children per woman more plausible with the sustained high fertility of increasing numbers of immigrant women combined with immigration of 800,000 and life expectancy of 88 years, would push the U.S. population to half a billion by 2080. Even if fertility and immigration remain at the assumed "most likely" levels of 1.8 and 500,000, an increase in life expectancy to 88 years by itself would bring the population to 318 million in 2050 with no prospect of decline.
If Americans' demographic behavior were to follow the Census Bureau's middle scenario, would the nation suffer? Some warn of declining national power, chronic labor shortages, and loss of economic dynamism. But the numbers hardly support such a cataclysmic view. In 2080 the U.S. would have 292 million people, 50 million more than now and an immense pool of consumers, workers and potential military recruits by any standard.
Those fearing labor shortages assume that the phenomenal job growth rates of the past two decades are a permanent condition. At current labor force participation rates the U.S. in 2080 would have 148 million available workers for an economy that will have become less labor-intensive. To be sure, the society of 2080 will be "older", but with an elderly population of greater vigor and longer working lives than ever.
The Census Bureau has given us useful alternative views of our population future. The "most likely" future they project need not frighten us. But it is the billions of unpredictable demographic choices of hundreds of millions of individual Americans over decades that will define that future - not Census models.
David E. Simcox, former head of the State Department's Office of Mexican Affairs, is the former Chairman of the Center's Board of Directors.