Matthew Klein of Barron’s exhibits less concern than other observers about Social Security’s long-term viability. Still, Klein points out factors that could mean bad news for the program’s future.

The legitimate concern is how U.S. society will address the expected aging of the population. A decade ago, about 12% of Americans were age 65 or older. That share is now 16% and is projected by the United Nations to increase to 21% by 2030. By the end of this century, about 28% of Americans are expected to be at least 65 years old.

This should be manageable. The Social Security trustees expect spending on benefits “rises from 4.9 percent of GDP for 2018 to about 6.1 percent by 2038, then declines to 5.9 percent by 2052, and then generally increases to 6.1 percent by 2092.” …

… Of course, long-range projections have a history of being wrong. Relatively small changes to assumptions about fertility, immigration, and mortality can have large effects when stretched out that far into the future. The current crop of estimates is likely too optimistic, especially when compared to projections made before the financial crisis. …

… For example, the trustees believe the large drop in birth rates in the past decade will reverse once the economy recovers and women who postponed having children in their 20s end up having kids in their 30s.

By contrast, the U.S. Census expects the fertility rate to stay where it is now. …

… The most questionable forecasts have to do with their economic projections. …

… Much more important is the question of productivity. The more each worker produces per hour, the easier it will be to transfer real resources from workers to retirees without the workers minding it. It would be equivalent to getting a smaller share of a much bigger pie. The problem is that the Social Security trustees project nonfarm business productivity to grow at a significantly faster rate in the future (2.06% per year) than it has in the past (1.84% per year since 1966).