It found a pattern of growing mortality inequality by SES.

"[T]he least educated men and women saw improvements from 1979–2011 of 1.5 percent and 0.5 percent per year, respectively, compared to 2.5 percent and 1.2 percent per year for the most educated," the study said, adding that mortality had improved more during the period for men than it had for women.

The study sought to show "how this growing gap in life expectancy has impacted inequality in the length of time that individuals spend in retirement relative to working."

Read: Senior boomers will be costly to sustain

It divided the study subjects, by socioeconomic status, into four quartiles, from lowest income to highest. Then it assumed that all individuals began working at age 22 and retired at age 65, and compared the number of years that each quartile of SES spent in retirement.

It found that "those in the lowest quartiles cannot work as long."

Males in the lowest quartile would spend 43 years working (age 65 – age 22) and 12.5 years in retirement, which amounted to spending 0.29 years in retirement for each year they worked.

But for men in the highest socioeconomic group, the results were higher: they spent 0.32 years in retirement for each year they worked.

For women, those in the lowest quartile spent an average of 0.40 years in retirement for each year worked, while those in the highest quartile spent 0.43 years in retirement for each year worked.

Further, the study said, "women in the lowest three quartiles would see their ratio reduced if they worked to Social Security's future 'full retirement age' of 67."

While the study reported that, because of improvements in mortality, all workers would live longer, those in lower quartiles would be disproportionately affected by any increases in retirement age—something, it cautioned, that policymakers should be aware of, since "policies seeking to extend worklives that treat all workers the same will tend to cut into the retirement of low-SES workers more than high-SES workers."