- Learn more about the RWJF Health & Society Scholars Program.
- Learn more about RWJF Scholars, Fellows and Leadership programs.
In order to allocate Social Security funds and other benefits, the United States government has for decades calculated the life expectancy of Americans. Population mortality forecasts generated by the U.S. Census Bureau and the Social Security Administration predict the ratio of the working population to the elderly Americans who are supported by their taxes, and by extension the solvency of the government programs tens of millions of Americans rely on.
But the predictions of these models may be miscalculated by as many as two years, which could mean a significant and unexpected burden on the system in the future, says Robert Wood Johnson Foundation (RWJF) Health & Society Scholar (2008-2010) Samir Soneji, PhD. An assistant professor at the Dartmouth College Institute for Health Policy and Clinical Practice and the Norris Cotton Cancer Center, Soneji is coauthor of a study published in the July 2011 issue of Demographics Research that recalculates Americans' life expectancy.
"We wanted to look at this subject because forecasts are useful to so many federal governing agencies and the general public interest." Soneji explains. "It drives resource allocation… future Medicare and Social Security financing might be structured differently because part of the financing of these programs depends on who's alive to pay for it and who's alive to receive the benefits."
Traditional academic and government forecasts rely on demographic patterns to determine the mortality rate and do not account for information about the effects of risk factors with known health effects. But Soneji and his coauthor, Gary King, PhD, employed a new model that accounts for two common risk factors they deemed important to modern-day mortality predictions: smoking and obesity.
While obesity has rapidly increased in the United States and is reducing life expectancy, they found, it is offset by reductions in tobacco use, which has steadily declined.
"These are two of the biggest biological risk factors," Soneji says. "Today's reduction in smoking won't affect today's lung cancer mortality, but it will affect mortality 20 to 30 years from now… We were very transparent and explicit [in the study] about how we make assumptions and what input we use, using today's health to forecast future mortality."
Using their updated methodology, Soneji and King estimate gains in life expectancy for both men and women in the next 25 to 30 years. They calculated an increase in male life expectancy at birth from 76.2 years in 2010 to 79.9 years in 2030, which is 1.8 years greater than the projection of the U.S. Social Security Administration and 1.5 years greater than the projection of the U.S. Census Bureau.
Their study predicts a more modest increase in life expectancy for females because their decline in smoking hasn’t been as great as the decline in smoking for men, Soneji explains. The researchers calculate an increase in life expectancy for females from 80.5 years to 81.9 years—nearly identical to the Social Security Administration projections and two years less than the Census projection.
Although the researchers forecast gains in life expectancy for the entire population, the gains could have been greater if not for the rapid rise in obesity, Soneji says. "Obesity might be holding us back from achieving even higher gains than we are forecasting."
Implications for Social Security
The two year difference in the study's life expectancy for men and the Social Security Administration’s projection "is a considerable difference in structure and aging of the future population," Soneji says.
"We're forecasting a larger and older population than Social Security is preparing to handle," he says, which means more Americans drawing benefits than the program is prepared to finance. "Social Security may face greater financial concerns sooner than expected, on the order of several years."
Soneji and King have continued to examine the implications of their findings on the Social Security program. In a forthcoming paper, they will examine the solvency of the program and the challenges it will face, and calculate the "turnover point" when the program will not be collecting enough in taxes to distribute to beneficiaries.
"Right now, Social Security collects more in taxes than it gives out, but that may reverse sooner than Social Security expects," Soneji says. "The program may be in more dire straits than they're thinking."