What the Social Security Administration Knows, and Could Tell Us

Last December I wrote a quick post expressing concern that the U.S. might have reached peak transparency, now that the Democratic Party, as a result of the rising burden of public employee pensions, has turned against the dissemination of accurate, factual information about government and society. Joining the Republicans, who have been against providing access to such information for a couple of decades.


Since then I’ve seen the same concern expressed by many others, now that Donald Trump, hardly Mr. Transparency himself, is President, with reports of government bureaucrats spiriting away statistical information to a secure location before the change of regime, lest it be deleted. Even so, I’m always on the hunt for alternative sources of actual facts, and this January I happened to think of one – the Social Security Administration. And wrote a letter to the Deputy Chief of the Office of Long-Range Actuarial Estimates, the office “responsible for estimates for up to 75 years in the future, based on economic/demographic assumptions developed for the annual Trustees Report.”

I didn’t receive an answer. Given that people need to keep their jobs until they can collect their pensions, and having worked for the government for 20 years myself and knowing what it’s like, I didn’t expect one. It is fair to say that I wrote the letter that follows for the purpose of publishing it on this blog after a reasonable period of waiting for a response had passed.

January 24, 2017

Lawrence D. Littlefield



Email address


Karen Glenn, Deputy Chief Actuary

Office of the Chief Actuary

Social Security Administration


Dear Ms. Glenn,

I am writing to suggest that Social Security Administration records could be used to analyze the extent to which different generations of workers have earned more or less at different points in their lifecycle and a career, and to request that such an analysis be undertaken and published.

For much of my working life I have been reading studies and analyses showing that those at the back end of the Baby Boom, and the generations to follow, have earned less (even progressively less) on average at each point in their working lives than the generations that preceded them. Less in their 20s, less in their 30s, less in their 40s, etc., when adjusted for inflation.

I discussed a couple of recent studies in a couple of recent posts on my blog. The Federal Reserve Bank of St. Louis found that each later born cohort of workers by birth year, adjusted for inflation and other socio-economic factors, have had less income and wealth than those born between 1935 and 1945, the best off generation.


Other studies that do not adjust for educational attainment put the start of falling income by generation later, in late-1950s, because the Baby Boomers were so much better educated than the generations preceding.

The Census Bureau did an analysis of the characteristics of young adults at different times, using Census data. It showed that those in my children’s generation were worse off in young adulthood than those in my generation had been at the same age. A disadvantage that may continue as today’s young adults pass through later points in their careers. I referenced that study in this post.


Crucially from the point of view of the Social Security Administration, if younger generations are being paid less on average (or have lower self-employment income, as a rising share of the workforce is contract labor) they will also be entitled to lower Social Security payments on average in retirement.  This disadvantage would compound the lower payments at each retirement age resulting from the rising “normal” retirement age, which some want to increase further.

Perhaps offsetting these disadvantages is the higher share of workers, and thus larger number of family members, in the workforce among later-born generations, with more people set to collect twice under current law.

Past analyses of the financial status of different generations of Americans have been based on federal statistical samples: the former long form of the decennial census, the American Community Survey, the Survey of Income and Program Participation, and the Current Population Survey. The work earnings data in these surveys is self-reported, perhaps inaccurate, and in the case of the decennial census, only available for census years. When one generation was, for example, at ages 46 to 55, the census year might have been a recession year, but for another generation the census year when they were ages 46 to 55 may have been in an economic boom.   The data for the two generations would not really be comparable.

Looking at my Social Security Statement, however, it occurs to me that the Social Security Administration must know the wages, salaries, and self-employment work earnings of every current and retired worker in the United States when they were at every age, adjusted for inflation. In order to calculate the level of future benefits.   And that this data is more detailed, and more accurate, than the survey data previously relied upon.

Based on this data the SSA is therefore able to calculate, for example, the median earnings between age 26 and 35 of those in the so-called Silent Generation, those born between 1936 and 1945, compared with those in the first half of the Baby Boom, those born between 1946 and 1955, compared the second half of the Baby Boom, those born 1956 to 1965, and compared with those in Generation X, Generation Y, the Millennials, etc.

And the income of a worker in the 80th percentile, in the upper middle class, in each generation at that age. And in the 20th percentile, among the working poor, at that age.

The Social Security Administration could compare the generations at each point in their lives, based on average earnings, rather than just ages 26 to 35. And make projections for points in their lives more recently born generations haven’t reached yet, based on the assumption that they would continue to as far ahead or behind prior generations at older ages as they had been at younger ages.

As for the possibility that falling median work earnings per worker has been (or had been) offset by a rising share of adults in the workforce, the SSA could calculate this as well. By calculating the percentage of people in each generation that were working at each age, based on its own records and census bureau population data by age. Indeed I suspect the SSA must already be doing this to produce its long-term projections.

The SSA could thus provide a table of median work earnings for each generation at each point of their lifecycle similar to the table of suicide rates for each generation at each point in the lifecycle that I posted here.

Just imagine median inflation-adjusted income over a decade in each of those cells, rather than the suicide rate. And a projection of median total career work earnings by generation, made by projecting the relative level of advantage or disadvantage thus far forward. And thus future Social Security benefits.

Just today, the question of how much worse younger generations will have to be in retirement came up in congressional testimony, for the confirmation of the future head of the Office of Management and Budget under the Trump Administration. Why is that always the assumption, and why is it always assumed that older generations that wanted to pay less are also entitled to more? Why is this never questioned?

Have later born generations, those now approaching retirement, worse off than those who came before, the recently retired? Are our children worse off still? And how will that affect our and their health and well being in old age? I have my own opinions, but these are based on the facts as they have been presented to me. I’m interested in better, more accurate, more telling facts and ask that the Social Security Administration provide them to the American people.

I’d be happy to answer any questions about this.


Larry Littlefield


Lets review some of the assertions I made in the letter in turn. Here is a summary of yet another study showing later-born generations are poorer, data source unknown, data behind a paywall.


Using panel data on individual labor income histories from 1957 to 2013, we document two empirical facts about the distribution of lifetime income in the United States.

First, from the cohort that entered the labor market in 1967 to the cohort that entered in 1983, median lifetime income of men declined by 10%–19%. We find little-to-no rise in the lower three-quarters of the percentiles of the male lifetime income distribution during this period. Accounting for rising employer-provided health and pension benefits partly mitigates these findings but does not alter the substantive conclusions.

Recall that I tested the assumption that falling wage income has been offset by rising employer contributions for retirement and health care benefits, using Bureau of Economic Analysis data, and found this to be false starting around 1990.


For most private sector workers non-wage benefit income has been going down as pensions are down away with, employer contributions to “defined contribution” plans are reduced or eliminated, employee co-contributions to employer-sponsored health insurance are increased, and more and more workers were re-classified as “freelancers” or “contract workers.” These are ongoing trends going back decades, with later-hired workers generally worse off in union contracts, and the private sector in general, as well as in government. All to pay for deals of those who came before.

For women, median lifetime income increased by 22%–33% from the 1957 to the 1983 cohort, but these gains were relative to very low lifetime income for the earliest cohort. Much of the difference between newer and older cohorts is attributed to differences in income during the early years in the labor market. Partial life-cycle profiles of income observed for cohorts that are currently in the labor market indicate that the stagnation of lifetime incomes is unlikely to reverse.

Basically, average female earnings increased relative to decades when women were traditionally confined to limited lower-paying occupations, and had lower educational attainment. But following that catch-up, median female earnings at a given education level have started to fall too, as I showed using American Community Survey data too.


Far from seeking a better alternative for this type of information, the Republicans have long wanted to get rid of the one that already exists, the ACS.

Second, we find that inequality in lifetime incomes has increased significantly within each gender group. However, the closing lifetime gender gap has kept overall lifetime inequality virtually flat. The increase within gender groups is largely attributed to an increase in inequality at young ages, and partial life-cycle income data for younger cohorts indicate that the increase in inequality is likely to continue. Overall, our findings point to the substantial changes in labor market outcomes for younger workers as a critical driver of trends in both the level and inequality of lifetime income over the past 50 years.

While the press reports that workers are upset about wage stagnation in the wake of the Great Recession, the trend has been wage and then benefit decline, not stagnation, over the 44 years since 1973, not in the seven years since 2008.

Despite earning less, Americans spent more, with more family members working, and then a virtual elimination of saving for retirement, and then soaring public and private debts.

Now that the first disadvantaged generations are in the their 50s and are facing retirement into poverty, and the only thing anyone can think of as a solution is to have the privileged generations put even less in (more tax cuts), and take even more out, at the expense of their worse off children. While these generational inequities are generally kept out of public discussion, even the beneficiaries see the writing on the wall.



Bottom of Form

There’s a time in everyone’s life to save. There’s also a time when you’re supposed to spend. That time is commonly known as retirement.

Millions of Americans aren’t doing that, however, which has put the U.S. in a perverse situation. Younger generations aren’t saving enough as their income slips further behind previous generations. Older Americans meanwhile sit atop unprecedented piles of assets built through stock market and real estate booms.   Yet these retirees, or at least the affluent ones, aren’t spending it. It turns out they’re afraid of the unknown.

Studies “have found affluent older Americans hoarding money. Last year, a studyin the Journal of Financial Planning found that the wealthiest fifth of U.S. retirees were spending 53 percent less than they could have. Meanwhile, the poorest 40 percent generally spend more than they safely should; the median retiree spent about 8 percent less than the safe amount.

Researchers looked at all the logical reasons why affluent retirees might be so tight-fisted, including the desire to leave an inheritance or worries about future medical needs. The big motivator turned out to be some version of fear they would run out of cash too early.

The situation for wealthier older Americans couldn’t be more different than that facing younger generations. A study released by the National Bureau of Economic Research last month found the typical American man who entered the workforce in 1983 earned up to 19 percent less over his lifetime compared with one who started working in 1967. (Women’s incomes rose over that period, but that’s because earlier generations of women earned very little money.) Based on more recent data for younger people who are still in the workforce, the authors wrote, “the stagnation of median lifetime income seems likely to continue.”

As noted, if people are earning 19.0% less they will be getting less out of Social Security since lifetime earnings is one of the factors in determining lifetime benefts. In addition, the Social Security trust fund is scheduled to run out of money around the time I am 70 years old and looking to retire, leading to a further automatic cut in benefits of 20 percent or so. These two factors will multiply by each other. And later born generations are dying younger, leading more of them to die before getting anything out of Social Security and Medicare, after a lifetime of paying in. And some want to cause even more of them to die young and unassisted, by raising the age of Social Security and Medicare – but only for those born after 1960.


Proposals to increase Social Security’s retirement age are beginning to resurface. The notion is that if people are living longer, they can work longer.  But the retirement age has little to do with how long people work, and a lot to do with how much money they get. Increasing the retirement age is a benefit cut.

Social Security’s Full Retirement Age (FRA) under current law is in the process of moving from 65 to 67. To keep lifetime benefits for the average worker roughly constant, benefits claimed earlier than the FRA are actuarially reduced and benefits claimed later are actuarially increased. That is, an FRA of 66 means that people will get 100% of their benefits at 66, 75% if they retire at 62, and 132% if they claim at 70. The comparable numbers for age 67 are 100% at 67, 70% at 62, and 124% at 70. 

So as the FRA rises from 66 to 67, the worker retiring at 62 sees his monthly benefit cut from 75% to 70%of the full benefit. The worker who increases his retirement age from 66 to 67 sees no cut in the monthly benefit but receives benefits for one less year, reducing his lifetime benefit. So raising the FRA is a cut in benefits either way. 

Continuously increasing the FRA is particularly hard on the lower paid, because they tend to retire early. This impact can be seen in the Figure, which shows the average labor-force participation rate for those with a high-school diploma or less (roughly the bottom third of the education distribution), those with some college but no four-year degree (the middle third), and those with a four-year degree (the top third). Although education isn’t a perfect proxy for lifetime earnings, the two are highly correlated. If the average retirement age is defined as the age at which half the group is still in the labor force, then the average retirement age is about 62 for the low earners, 64 for the middle earners, and 66 for the top third.

OK, so who is going to pay for that extra 20 percent of Social Security once the trust fund runs out – after the federal debt, having already soared as a percent of GDP, soars further to “pay back” the trust fund?

There is an under the table argument about how, and in what form, and with what distribution of the pain, poorer later born generations are going to be made even worse off to pay for the richer benefits of Generation Greed. The argument will remain under the table until everyone in Generation Greed is retired, collecting, and no longer paying payroll taxes. Then they will vote to grandfather themselves from any and all sacrifices, even at the cost of having their children due younger in ill health after having received nothing from Social Security and Medicare at all.

That is what is going on in Social Security. And every decision for 35 years has been just like it.

Taken the subway lately? Then perhaps you might have noticed the impact of 25 years of paying for maintenance with debt, on those who are still around to pay the debt back. And so the politicians whose campaigns have been funded by the winners, and whose policies have favored the richest and most entitled generations, are all claiming they not responsible for the NYC subway. From Mayor DeBlasio and Governor Cuomo what we get is this.

In Illinois there is a similar argument about who destroyed the entire state, with blame assigned to various politicians and political parties, not to the generations, and interests within generations, that benefitted from the past sale of the future (now the present).





Services for the young and poor have been gutted. Many services are near collapse. Property taxes have soared. Newly hired teachers will not only not get Social Security, but will also contribute all or perhaps more than all of the cost of their own pensions to pay for the richer benefits of those who went before. And yet two things cannot be considered – making retirement income taxable for state income taxes on the same basis as equal wage income, and reducing pension benefits for existing and near to retirement beneficiaries.

The argument over who is to blame could be going on in New Jersey, in Connecticut, in Dallas and Houston, and eventually in New York.

And, in less than 15 years, a similar discussion is scheduled for Social Security. By then, the actual perpetrators will have made the getaway. So they don’t want that discussion to take place now, using data from the Social Security actuaries. They want more federal tax cuts, and the deceptive assertion that everything will be OK.

1 thought on “What the Social Security Administration Knows, and Could Tell Us

Comments are closed.