Fixable Error, New Insight, and Social Security

by Dale Coberly

Fixable Error
New Insight
and Social Security
(amended version)
When errors are honest, and people keep talking in good faith, they are usually fixable. One of the errors I will talk about here is mine. The new insight is entirely the result of the good faith and honesty of Eric Laursen who has written a book about Social Security called The People’s Pension.

In the transcript ( ) of Mark Miller’s interview with Laursen about his book appears this statement regarding possible fixes for Social Security:

“A fallback would be to raise payroll taxes for everyone, but so slowly that it would have little impact – say, by 0.01 percent a year”

I knew this was wrong, and I thought I knew why: Lots of people get confused between 0.01 which is one percent of one, and 0.01 percent, which is one ten thousandth of whatever you are talking about. A tax increase of 0.01% on a 40,000 dollar per year income would be 4 dollars a year. This would not help Social Security. But it is marvelous that a tax increase of 0.05%, which is one half of one tenth of one percent, per year would solve the Social Security problem entirely. This is five times 0.01%, or about 20 dollars per year, per year, or about 40 cents per week, per year.

Laursen told me he had gotten the number from Monique Morrissey ( ).

” These projected longevity gains, if they materialize, could also be offset by a 0.01 percentage-point increase in the payroll tax per year from 2025 to 2084 (from 6.2% to 6.8% overall). Over this 60-year period, average wages are projected to nearly double in inflation-adjusted terms. “

So I read Morrissey’s paper to see where she had gotten the number. Morrissey is primarily arguing against increasing the retirement age, and she wishes to show that increases in life expectancy are not the drivers of the projected shortfall in SS:

” However, gains in life expectancy represent only a small part of the fiscal challenge facing Social Security. The increase in the normal retirement age from 65 to 67, currently underway, already offsets gains in life expectancy for workers born before 1960, and longevity gains for younger generations account for only a fifth of the projected Social Security shortfall. “

So the difficulty is not in the arithmetic, but in the fact that she is talking about only ONE cause of the projected shortfall, which by her reckoning is only one fifth of the problem.
That’s fine, but it can lead to confusion. As in fact it did for Laursen and me.

Morrissey’s essay is mostly written in opposition to raising the retirement age, and it is WELL WORTH READING.

She provided me with a new insight.

That said, I disagree with her conclusion.

Her conclusion is that raising the cap on income subject to the payroll tax, scrapping it entirely, is the answer to the Social Security projected shortfall. I think that such a solution has no chance whatsoever in this political climate. And I do not think it would be a reasonable solution for a program that is successful exactly because it is NOT WELFARE. Raising the cap would turn SS into welfare by “taxing the rich” for benefits they will not get.

Moreover the cost of fixing SS by raising the payroll tax under the current cap is so tiny… forty cents per week… that no one would even feel it. And it would preserve the fundamental nature of Social Security: the “tax” is paid for by the workers who will get the benefits.

Morrissey appears to think of Social Security as “social welfare”, in which the whole society… and that means the rich… pay for the needs of the needy.

But American workers don’t see themselves as needy; they don’t want welfare; and they very much like the idea that they “paid for it themselves.”

And the American “rich” are not going to pay for the workers retirement. They see themselves as having a hard enough time paying for welfare for the truly needy without being asked to foot the bill for every single worker’s expected twenty years in retirement.

But Morrissey provided an insight which I had missed.

I have been saying that the need for the payroll tax increase is because future workers are going to be living longer in retirement than past workers.

Morrissey shows that in some important ways this is not exactly true. Future workers will not, in general, live significantly longer than current retirees. The increases in longevity will be mostly among the rich.

She shows the previous increases in longevity have largely been paid for by other factors, including increases in population and increases in incomes. I’ll leave you to read her essay to learn about this. Essentially I will argue that those other factors have masked the “true” cost of Social Security, and it would be serious error to to try to make up for changes in them by changing the fundamental nature of Social Security from “the workers pay for it themselves,” to “social insurance,” or what we call “welfare.”

It is not that easy to correct demographic trends, or economic trends, or inequities in the earnings of the poor vs the rich. These are problems that should be addressed, but they should be addressed outside of Social Security. Social Security is insurance, designed to help people through bad times. It is a mistake to shoulder Social Security with the burden of correcting all the ills of society that may fall upon the poor.

But that brings me to my error. And why I am not going to “fix” it.

I came into the Social Security debate from a background that has nothing to do with politics or “social welfare.” This gives me a different “frame.” [note this is not Lakoff’s political framing, but more like framing a problem in physics so that it can be solved. Peter Peterson has framed the problem so it cannot be solved at all. Peterson has dedicated a billion dollars to making his framing the conventional wisdom, and I think Morrissey has accepted it without thinking too much about it.
Consider a universe in which, for the moment, there is no population growth, no increases in longevity, no economic growth, no inflation, no interest, no changes in longevity.

But in this universe people still work when they are adults and strong, and want to be able to retire at some point before they die. They will need to “save” for that retirement some of what they produce while they are working.

This leads to a Basic Retirement Equation:

years worked, times percent of earnings “saved,” equals years retired, times retirement income (as a percent of prior earnings).

So if you are going to work for forty years and expect to live ten years in retirement at 40% of your prior income, you would need to “save” ten percent of your income:

40 X 10% = 10 X 40%

It does not matter, in this universe, if you save this as gold in a box, or as a non-cash implicit agreement by which working adults pay for the needs of the elderly, or by a formal pay as you go “tax”.

In all cases you are deferring 10% of your present income so that you will have enough to live on when you get old. This can be done by hoarding “gold”, which is dangerous, or by implicit “honor your mother and your father” agreement between generations, or by an explicit government-managed (but not “government paid”) tax and benefit system.

Now imagine another universe in which nothing has changed except that you expect to live 20 years in retirement. The equation now becomes:

“40 years working, times 20% savings, equals 20 years in retirement, times 40% “replacement benefit.”

This may look like a lot of money to save, or be taxed, but it would be a perfectly reasonable solution to the problem of living for twenty years after you become too old to work, or want to work. The key here is you have to realize the 20% is “your money”. You saved it. And it will be you who spends it.
As long as the other members of the society agree that this is a reasonable way to apportion earnings between working years and retirement years, it is no problem. Especially if the workers agree that money not used by those who die earlier than the 20 year expectancy will be used to pay for the needs of those who live longer. This is not welfare. It’s insurance.

Now, in the real world, the fact has been that other factors including income growth and population growth… at least the population of workers in the Social Security program… have worked to reduce the needed “savings” or tax rate. [Also, for the past few hundred years or so, “interest” or “capital growth, including appreciation of stocks, has reduced the needed “savings” rate. But this has been accompanied by the risks of inflation and market losses, which is what Social Security was invented to provide insurance against.]

But these other factors are OUTSIDE of Social Security. And Social Security does not depend on them to work.

It may be that rising income inequality, reductions in birth rate, or other economic and demographic factors will affect the tax rate needed to provide a reasonable retirement “rate.” But the BASIC fact is that it is the ratio between working years and retirement years that determines “the cost of retirement.”

Social Security is designed explicitly to take care of the ups and downs in the economy, including demographics. If rates have to go up, or if benefits have to come down in response to hard times, so be it. That is the history of the human race: There are fat years and lean years.

Some people argue, in bad faith I think, that ups and downs in Social Security “return on investment” are a kind of generational crime. They prefer you don’t look at the ups and downs in return on investment in the private markets. Even that box of gold under the bed, or the ability of the working adults in a family to care for their own elderly relatives, or the ability of investors to realize a rate of return they need to fund their own retirement… will be subject to factors like demographics, wars, the state of the economy, inflation, market losses, and will result in greater or lower “returns on investment” from one generation to the next. But the enemies of Social Security want you to think that if SS does not make exactly the same return to each generation, some generation is being “robbed.”

As it happens we are looking at… according to the Trustees… a period of time where some of the factors that have held down the cost of Social Security in the past will not be as much of a help in the future as they have been in the past. These “anti-factors” include rising inequality of income, a falling birth rate, lower productivity gains.. a combination of factors that will require either a benefit cut of about 20% or a tax raise of about 4%, which can be phased in over about eighty years, at one half of one tenth of one percent at a time.

Cutting a benefit that is already at survival level by 20% is insane. Raising a tax 4% over about 80 years wouldn’t even be felt. [The 4% tax increase would occur while incomes are increasing more than 100%, so that, in round numbers, after eighty years your after tax income will have increased 96% instead of 100%… you are not going to notice the difference… unless of course you threw away your Social Security in order to “save” that 4%.]

But these factors are external to the basic mathematics of retirement, especially from the perspective of the individual. The basic relation is: {the number of years you work compared to the number of years you expect to be retired} determines {the ratio between tax rate and benefit rate}.

While it is true that relying on the “whole economy pays for everybody” could in principle produce the same result, on average, through general taxation and a “government pension.” This perspective leads to an understanding which is basically wrong, and which the politics of this country, and the psychology of both the workers and the “rich” is not ready to accept. We prefer an explicit relation between the money we “save” as individuals and the benefits we get as individuals. We prudently modify this principle enough to allow for an insurance factor whereby those who cannot save enough over a lifetime of work to pay for a decent retirement are helped out by the “excess” of those who were able to save “more than enough” in the protected savings plan we know as Social Security. But the principle is still essentially “you pay for what you get.”

Morrissey may be correct that only one fifth of the projected shortfall is “due to” rising life expectancies, but that is because those other factors that have previously reduced the cost due to rising life expectancies may no longer be able to do so.  

We cannot easily change those factors, but going from there to propose a “solution” that destroys the fundamental nature of Social Security by changing it from “worker pays” to “the rich pay,” will in the first place not be accepted by those rich who have the political power in this country, and in the second place would deprive the workers of both the satisfaction and the political power of being able to say “I paid for it myself.”

If there were no alternative… if the extra revenue needed for a decent retirement were beyond the reach of the workers, then a welfare solution might be necessary… though this would be a bad way to arrange an economy over the long run. But since the cost of “fixing” Social Security, whatever the reason for the shortfall, amounts to only forty cents per week per year, it is far, far better that the workers pay for it themselves than to have them go to Mr. Richfellow and say, “Please, Sir, can I have some more?”

Meanwhile I will keep saying the need for the “tax” increase is because we are going to be living longer. I believe this is the more fundamental truth. And it’s certainly a lot easier to explain to people than all the complications Morrissey addresses in her essay. And it leads to the correct solution: the workers need to pay a little more for their own retirement because they will be living longer.

When you go to the store you do not go up to a well dressed stranger and say, “Sir, it looks like you have more money than I have. Please pay for my groceries.”

Social Security is you paying now while you have the money for the groceries you will need later when you may not have the money. That is a lot better for the country, and a lot safer for the workers, than teaching them to expect “the rich” to pay for the predictable costs of a large, and normal, part of their lives.