According to the Committee for a Responsible Federal Budget, “life expectancy at age 20 has increased by 9 years for men and 10 for women.”  By comparing this change to the increase in retirement age, the CRFB is implying that people who reached age 20 in 2006 can expect retirements 7-8 years longer than people who reached age 20 in 1940.  This is a gross exaggeration.

There are two reasons why this is dangerously wrong.  First, there is a technical point.  There are two main measures of life expectancy, and the CRFB chooses unwisely for their comparison.  Second, life expectancy only measures the length of retirement for those who are of retirement age.  By choosing both to use life expectancy and then choosing the less relevant measure, the CRFB grossly exaggerates the actual and projected experiences of workers.

Measures of Life Expectancy

When the CRFB says that life expectancy was 61 for men born in 1940 and 75 for men born in 2006, these are projections that are based on the health care and death rates at the time.  In fact, the improvements in health care in a numbers of areas have substantially reduced death rates for men born in 1940.  Social Security currently projects the average lifespan of men born in 1940 to be 70.1, not 61.4.  For men born in 2006, the figure is 81.7—not 75.1.

The CRFB’s “period” life expectancies are lower, but have grown much faster than have actual “cohort” life expectancies.  With many fewer years for the benefits to compound, the CRFB’s period life expectancy and cohort life expectancy are much closer when measured at age 65, but period life expectancy is entirely irrelevant to the discussion.  What really matters is how long actual and expected retirements have changed over time, and life expectancy is not a good measure.


Person A and Person B are both age 20 in 1940.  Person A will live to age 57 and Person B will live to age 77.  On average, they will live to 67—two years more than age 65.  This seems to suggest that they average two years of retirement, but that is not correct.  Person A, has no retirement while Person B has ten years of retirement.  Correctly computed, their average length of retirement is five years, not two.

(We write here of expected years of life past normal retirement age.  While most workers do retire early, their benefits are reduced correspondingly.)

Thus, when the CRFB says life expectancy for men was 67 for those aged 20 in 1940 in comparison to a normal retirement age of 65, it leads readers away from the fact that those 20-year-olds could expect to enjoy a retirement of nearly 11 years.  Even back in 1919, 20-year-olds would look forward to an average of 8.3 years in retirement even though (based on period life expectancy) the average 20-year-old would not live to 65.

Life expectancy — especially period life expectancy — simply does not describe the retirement experiences of young workers.

Increased life expectancy and work

The CRFB also steers readers away from important data.  They write, “In light of continued increasing longevity, it only makes sense that individuals would work longer. Raising the normal retirement age would serve as an important signal to encourage them to do so, and raising the early retirement age would go even further to promote longer working lives.”

Unfortunately, the CRFB conveniently leaves out that (on average) increased life expectancy has meant that individuals have worked longer over time.  Men turning 20 in 1969 will live on average 42.0 years between ages 20 and 65.  By comparison, men turning 20 in 1919 lived only 39.0 of those years.  On average, men’s working-age lives have grown by three years—even before considering the increase the retirement age!  Over the century, the average working life of a 20-year-old man will have grown by six years.

For women it is even more stark.  A woman turned 20 in 1980 could expect a retirement of 17.1 years—only one month longer than a woman turned 20 in 1957.  The increase in the retirement age from 65 to 67, however, has meant additional working years with virtually zero increase in expected retirement.

Falling Age of Retirement

Finally, CRFB asks why it all matters, given that the average age of retirement (never mind that statutory age) has fallen. But this is the average of the people who retire.  The CRFB wants its readers to believe that people who would have delayed retirement past age 65 back in the 1950s are people who retire early today.  More likely, the people retiring early today are people who would not have made it to 60 in decades past.

“Why all the pushback” indeed!

Year Age 20

Expected Years of Life 20-Retirement


Expected Years of Life During Retirement