View all Articles
Commentary By Charles Blahous

Social Security and Longevity Increases: Getting the Facts Right

Economics, Economics, Economics Tax & Budget, Healthcare, Employment

Click here for a printer-friendly version of this article.

This last week, the Huffington Post derided President Obama’s fiscal commission co-chair, former Senator Alan Simpson, suggesting that his proposals to increase Social Security’s eligibility age were founded upon gross ignorance of the details of longevity trends (disclosure: I worked for the Senator fifteen years ago, but have not consulted with him about either the interview or this piece). Paul Krugman and Ezra Klein thereafter continued this attack on their blogs with the New York Times and the Washington Post respectively.

The aim of the attacks was to spread the impression that the Simpson-Bowles commission’s Social Security eligibility age recommendations were premised on misinformation. An examination of the policy realities, however, demonstrates otherwise.

The particular chronology was as follows: a reporter contacted Senator Simpson and read him some figures suggesting that life expectancy had grown since 1940 by much less than the Senator had been claiming. Senator Simpson reportedly expressed some unfamiliarity with the particular figures cited, whereupon the reporter stated that they came from the Social Security Administration. An online article resulted, alleging that the Senator suffered from an “apparent unfamiliarity with the history and foundation of Social Security.” (For good measure, the piece mocked Senator Simpson for saying that Social Security was originally “not a retirement program”; more on this later).

The basic background: since Social Security’s inception, American life expectancy has increased – not in equal measure for everyone, but nevertheless for both sexes and throughout the income spectrum. Such longevity increases are one but not the only factor placing strains on Social Security finances (other important factors include changes in fertility rates as well as a host of legislated benefit expansions).

Importantly, under both current Social Security law and the Simpson-Bowles proposals:

  • Individuals would be able to claim retirement benefits earlier than when Social Security was first established;
  • Beneficiaries would generally expect to live significantly longer – whether at birth or at age 65 -- than when Social Security was first established;
  • Ergo, beneficiaries would generally expect to collect for a larger number of years;
  • Future annual benefits at any given claim age would be far higher than under original Social Security -- whether measured in real dollars or replacement rates.
  • Future (like current) retirees would further benefit from the provision of annual COLAs, not an original feature of Social Security.
  • Summarizing: under Simpson-Bowles, beneficiaries could claim retirement benefits earlier, expect to live longer, collect for more years, receive higher initial benefits, and thereafter receive annual COLAs that Social Security’s original beneficiaries did not.

These bottom-line facts are being presented here because it would be virtually impossible to glean them from recent accounts. Now on to the salient analytical points:

Point #1: The Simpson-Bowles proposals were developed based on assumptions incorporating the same longevity figures cited in the Huffington Post article. The HuffPo article implies that more modest adjustments than Simpson-Bowles’s “dramatic changes” would be sufficient if longevity is properly measured. This is incorrect. The cited longevity assumptions actually substantiate that the amount of changes offered by Simpson-Bowles is required, for their proposals were developed around those very assumptions.

Point #2: Life expectancy growth even at age 65 (the measure emphasized in the HuffPo piece) has far surpassed the rise in Social Security’s eligibility age, supporting the commission view rather than its critics’. Through today, Social Security’s Normal Retirement Age (66) has risen by only one year since 1940. According to the SSA figures cited, life expectancy at age 65 has simultaneously grown by over five years, both for men and for women. By the time the NRA reaches 67 under current law, life expectancy at 65 will have grown by over six years. By the distant time Simpson-Bowles would increase the NRA to 69 (four years total increase since 1940), life expectancy at age 65 will have grown by over nine years.

Point #3: The article covering the issue conflates two different measures of life expectancy, likely accounting for some of the mutual confusion.

Two different measures of life expectancy are provided in the Trustees’ report. “Period” life expectancy is observed life expectancy for a particular year. “Cohort” life expectancy by contrast takes into account future longevity increases affecting all people alive in a particular year. Thus, “period” life expectancy tells you that by 1940, males who had reached the age of 65 were already living an average of 11.9 more years. “Cohort” life expectancy instead tells you that 65-year-old males alive in 1940 could expect to live another 12.7 years, that is, a bit longer due to improvements during those following 12 years.

For life expectancy at age 65, the differences between these two measures aren’t that big. But for life expectancy at birth, the differences are substantial. It seems obvious that Senator Simpson was citing “period life expectancy” whenever he made his repeated observation that Social Security’s original eligibility age was set higher than total male life expectancy (period life expectancy at birth for males in 1940 was 61.4, whereas cohort life expectancy was 70.1). But the reporter not only shifted the reference frame from life expectancy at birth to life expectancy at 65, but also shifted from period to cohort life expectancy as well (as can be demonstrated by matching up the article’s figures with the SSA tables).

Some bloggers treated the Senator’s surprise over hearing different figures as prima facie evidence of ignorance. Those familiar with the Trustees’ history of life expectancy presentations would not have done so. For most of that history, only the “period” figures cited by Senator Simpson were published. The first year that the Trustees also included the “cohort” figures was in 2001, following a recommendation of a 1999 technical panel of the Social Security Advisory Board. Senator Simpson was most likely surprised by the cohort figures because he had long been citing the other ones -- the period expectancies that have, after all, been historically more prominent in the annual Trustees’ reports.

Point #4: Despite the misimpression given by these articles, the Simpson-Bowles proposal would not require future workers to delay retirement relative to those of 1940. In reality, under the proposal, workers would be able to claim benefits earlier, collect them for longer, and receive substantially higher annual benefits regardless of the age of claim, than could workers in 1940.

One of the blog entries reporting on the HuffPo article states that under Senator Simpson’s proposals, “low-income Americans” in tough jobs would be told to “work those jobs longer because people who aren’t in them have seen large increases in life expectancy.” Insofar as this comment relates to the controversy comparing today’s demographics to those of 1940, this assertion is unambiguously false.

Social Security benefit levels are a function of several factors, including:

  1. The Normal Retirement Age (NRA): this establishes the age at which “full benefits” can be claimed.
  2. The Early Eligibility Age (EEA): this establishes the earliest age at which benefits can be claimed.
  3. The PIA initial benefit formula: this establishes the benefit level paid at NRA.
  4. The Actuarial Reduction Factor (ARF): this determines the annual benefit one receives if claiming before NRA. It causes the payment of a lower annual benefit than at NRA, adjusting for the fact that benefits are expected to be paid for a greater number of years.

Now let’s examine how these factors have evolved together under current law.

At Social Security’s inception, the NRA was 65. It is now 66. Under current law, it will gradually rise further to 67. Under Simpson-Bowles, it would ultimately reach 69 by the mid-2070s.

Original Social Security did not offer early retirement. This was established at age 62 in 1956 for women and in 1961 for men. Workers thus now typically claim benefits when younger, despite greater longevity, than in 1940. This EEA would not change under current law. Under Simpson-Bowles, it would ultimately reach 64 by the mid-2070s. In other words, beneficiaries could still claim benefits earlier than they could in 1940, despite the acknowledged increases in life expectancy since then. To repeat: Under Simpson-Bowles, by the 2070s, when life expectancy at age 65 is expected to be nine years greater than in 1940, individuals could still claim retirement benefits one year earlier.

Benefits claimed at ages 64 and 65 would of course be less under any scenario than they would be if claimed at age 69, due to the ARF. But they would nevertheless still be far higher than those paid to retirees in 1940. This is because of numerous benefit increases enacted since then. As a result, future beneficiaries under Simpson-Bowles would not have to retire at a later age than those in 1940, even to collect a substantially higher benefit.

Simpson-Bowles contains various other benefit restraints to sustain program solvency, beyond raising the retirement age. Yet even accounting for these other provisions, the point still remains: under Simpson-Bowles, future workers could claim higher benefits, earlier, relative to the 1940 frame of reference.

The data:

Year Worker Turns 65                                                                                            Initial Social Security Benefit, Age 65, Scaled Medium-Wage Worker, in $2010 (inflation adjusted)
1940 (Current Law) $4,069
2010 (Current Law) $16,752
2080 (Simpson-Bowles) $28,079

An allegation was made that low-income workers in tough jobs specifically would be harmed by Senator Simpson’s agenda. In reality, the benefit growth afforded at 65 to the low-income, long-working manual laborer under Simpson-Bowles would be proportionally far greater because of benefit increases crafted specifically for this group.

Year Worker Turns 65                                                                                              Initial Social Security Benefit, Age 65, Scaled Low-Wage Worker, in $2010 (inflation adjusted)
1940 (Current Law) $2,250
2010 (Current Law) $10,164
2080 (Simpson-Bowles) $22,335

Note furthermore: these figures are for initial benefit levels. COLAs were not given to 1940 retirees (they were enacted in 1972). This further reinforces the point that under Simpson-Bowles, the retirees of the future would not be asked to delay retirement relative to those of 1940, and would actually live much more comfortably if they choose to claim benefits at the same age.

Readers may be surprised at the low benefit levels for 1940. These figures look quite small to modern eyes despite this table’s adjusting them for inflation through 2010. Per above, Senator Simpson was derided for observing that original Social Security was conceived as an income supplement rather than as a full retirement program. But it was only much later that Social Security benefits were expanded to the point where they provided anything close to the share of total retirement income that they do today. In fact, because initial benefits were not originally indexed, they for a time diminished from the levels above in inflation-adjusted terms (a 1948 medium-worker’s initial annual benefits, for example, were only $3,053 in $2010 dollars).

Some readers will note that over time, wages have grown faster than inflation, prompting the follow-up question of how these numbers would look if adjusted instead to keep pace with national wage growth. Might it be that Simpson-Bowles would offer lower future benefits relative to 1940, adjusted for wage growth rather than inflation? Again, no – not even close. Wage replacement rates for medium-wage workers retiring at age 65 in 2080 under Simpson-Bowles would be roughly 25% higher than in 1940. For low-wage workers, replacement rates would be over 75% higher for the same claim age(!).

Reasonable people can differ on whether the Simpson-Bowles commission’s recommendations represent optimal policy. But while there is clearly not universal understanding of these issues, the president’s commission is not the place where the knowledge gaps lie.

Charles Blahous is a research fellow with the Hoover Institution and the author of Social Security: The Unfinished Work.