Bryan Caplan  

Social Security: The Inside Story

Is Collecting GDP Data An Exer... What kind of Great Stagnation?...
The Social Security Administration's latest "Actuarial Note" is as fascinating as actuarial notes get.  Its main lessons:

1. Social Security used to be a great deal.  The poorest one-earner couples born in 1920 got a real return of 9.18%; the richest one-earner couples got 5.87%.

2. Social Security remains a very good deal for one-earner couples: about 6.5% for the very poor, 5.5% for the poor, 4.5% for middle-income, 3.9% for the rich, and 2.9% for the richest (people who max out their Social Security taxes).  These rates have held almost steady for cohorts born since 1940.

3. For singles and two-earner couples, Social Security's return has heavily fallen.  Even the middle-income (career average inflation-adjusted earnings of $41,655 per year) now get a crummy return of under 3%.  For the rich, the return is around 2%, and for the richest, about 1%. 

4. Over time, singles and two-earner couples have become the norm.  So the decline in Social Security's return is more pervasive than it looks.

All these return estimates are based on the current formula.  The SSA also considers what will happen if taxes are raised or benefits are cut to keep the program out of deficit.  The two scenarios:
The Increased Payroll Tax scenario raises payroll-tax rates, beginning with the year of Trust Fund reserve depletion, to finance scheduled benefits fully in every year. The payroll-tax rate increases from the present law amount of 12.4 percent beginning in 2033. The payroll-tax rate increases to 16.54 percent for 2034 and continues to increase year-by-year, reaching 16.89 percent for 2086. Under this scenario, the payroll tax rate increases further after 2086 due to continuing increases in life expectancy.

Under the third scenario, Payable Benefits, payroll-tax rates hold constant while benefits decrease for each year after Trust Fund reserve depletion so that, for the Trust Funds as a whole, benefits paid equal taxes received. The reductions from scheduled levels apply equally proportionally to all types of benefits paid during the year.
5. In both alternate scenarios, returns for recent cohorts (born 1970 and later) fall by about 1 percentage-point.  Single-earner couples aside, a majority of recent cohorts can expect a real return of 2% or less.  The richest singles and two-earner couples get a return barely over 0%.

6. These are the SSA's own estimates.  What would neutral outsiders' estimates say?

Comments and Sharing

COMMENTS (24 to date)
Dennis writes:

Why is 2-3% real return a bad investment? They look incredibly attractive these days. Also, a fair comparison would be what investors actually earn once you take into account various behavioral biases/tendencies and inefficiencies in the investment sector (e.g.:

Brian writes:


So the absolute worst case single rich person return is a guaranteed mostly tax free 1% real return? And typical is 3-4% and more real, tax free?

That's fantastic. Go find me a secure safe bond returning 1% over inflation, mostly tax exempt and indexed to future inflation increases. If you can find a deal like that we'll soon be billionaires. Find a 4% real return for middle class couples and we could be trillionaires.

So you're saying SS is a trillion dollar beneficial idea for Americans. Big government Democrats will agree.

Note: Today's 5 year TIPS yield is 0.13%. Earlier this month it was -.05% -- that's _minus_ 0.05.

Kevin Erdmann writes:

Commenters saying what a great deal this is, I would assume, then, that you have your private retirement savings in funds with expected returns lower than this?

Charlie writes:

Expected returns for Social Security should be compared to bonds, not equities. The alternate scenarios presented above hardly seem politically infeasible, so presuming that SS is anywhere near as risky as equities would be ridiculous. The income stream is certainly more bond-like.

I'm just going to ballpark some numbers. Consider someone employed and paying SS taxes for 40 years. The higher earning years are later on, so let's say that this averages out to 15 years duration.

15 year TIPS are around 0.70% real return. "A majority of recent cohorts can expect a real return of 2% or less" in the alternate scenarios. That isn't too bad, all things considered.

Thomas writes:

Social Security payouts represent "returns" only in the sense that the proceeds of bank robberies do. Even when SS receipts outstripped payouts, the surplus was spent on other government programs. It wasn't invested* and therefore didn't yield "returns."

Some will argue that it was "invested," to the extent that it went into things like education and highways, which boost productivity. But that doesn't account for the opportunity cost -- the things that the private sector would have done with the same money. And given that government spending, on the whole, is counterproductive (e.g., $1.6 trillion annual deadweight loss due to regulation, significant reduction in growth rate per the Rahn curve), I stand by the view that the "returns" on SS amount to nothing more than robbing Peter to pay Paul.

robbbbbb writes:

Social Security is indistinguishable from a Ponzi scheme. Like all Ponzi schemes, returns look real good for the initial investors. Less so for the follow-on investors.

I'll be 41 next week. I don't expect SS to ever pay me a dime, and my retirement is planned accordingly. That's a big, negative rate of return.

john hare writes:

I would opt out if I legally could. The 8-10 years savings until I reach SS nominal retirement age would be less than I should expect to get from SS. That would be an acceptable tradeoff/loss if it would count toward eliminating a horrible system.

Perhaps SS could be reserved for those that are satisfied with 1% ROI.

Michael T writes:

SS may provide greater gains when compared to TIPS or bonds (I'd bet it beats lighting your money on fire also), but comparing to "risk free" assets is not a good benchmark. Putting all your money into an asset that has a low, risk free return is considered a bad investment strategy for a retirement account, at least in the grand scheme of a 40 year work life. The common advice is you should start with 100% equities early in your career, then wind down to a 40/60 mix to reduce risk. This is a far better comparison of what people can do with their retirement money(and what rate of return to expect) if it isn't deducted from their wages.

For actual data comparison(i.e. what gains do people actually achieve with their retirement accounts), I'd be curious to see what the rate of return on a typical 401k has been since their inception. It's probably much better than the 0-1% I'm projected to receive.

Social Security payments aren't tax free for high income people. Part of it is taxed back now, and more will be in the future. Where do you think the money is going to come from to redeem those special bonds the SS trust fund holds?

mucgoo writes:

Social security isn't as low risk as TIPS. A government "default" (reduction in payouts) is widely predicted. A US debt default obviously isn't. A ballpark 1% risk premium over TIPS seems reasonable. That still makes it a good deal for all but the richest.

I'd view social security as part of your bond allocation (and with a better rate of return than you'd find available) and then weigh the equities more heavily.

BC writes:

Dennis and Brian, in exchange for your Social Security contributions, I will gladly pay you the "guaranteed" rate of returns, guaranteed in the SS sense: you have no legally binding claim to your expected future benefits, and I can change your benefits at any time at my discretion.

If Social Security was so great, participation would not be mandatory. Instead, we would need maximum limits on participation, similar to the limits on 401(k) and IRA contributions. Notably, Social Security's strongest supporters are the ones that most strongly oppose any sort of even partial opt-out or private accounts. That's how much confidence they have that people would choose Social Security voluntarily.

BC writes:

I knew that SS transferred wealth from rich to poor, but I didn't realize that singles and two-earner couples were penalized so severely. These two wealth transfer criteria would seem to negatively impact highly educated women disproportionately. Highly educated women tend to have high incomes and to either be single or part of two-earner couples.

Many efforts to achieve "equal pay" for women involve heavy regulation of labor markets, which can have unintended consequences. In contrast, reducing the size of SS to minimize its disparate impact would probably actually reduce market distortions. I would think that progressives and feminist groups would want to abolish or reduce this discriminatory program designed by dead, old, white men.

Mark Bahner writes:
I'll be 41 next week. I don't expect SS to ever pay me a dime, and my retirement is planned accordingly.

I don't expect you'll get out what you put in, but I don't think you'll get nothing.

MikeP writes:

I knew that SS transferred wealth from rich to poor...

Are you sure of that?

The poor start paying Social Security when they are 18 or younger. The rich start paying when they are 22, or 26, or 28. The poor retire later. The rich retire earlier. Finally, the poor die earlier, and the rich die later.

It's not at all clear that the poor get a better return than the rich from Social Security. And when you consider that the payroll tax is wildly regressive, taxing even the first dollar earned by 15%, the loss of wealth for the poor far outweighs the loss of wealth for the rich.

Social Security is a transfer of money from the poorest age cohort directly to the richest. And all so the actually poor elderly can pretend they are not on welfare. It is an abomination.

ThomasH writes:

The post implies the opinion that Social Security is not an ideal policy. Very few policies are. The use of the ROR of the SS taxes paid by wage earners (which includes, of course, the "portion" "paid by" employers) implies a comparison to a policy of simply not taxing retirement savings up to certain percentages of income until withdrawn after some arbitrary age.

How likely is such a policy to have improved welfare had it been in effect for the last 70 years?

My opinion is not at all. I would prefer to pay SS and Medicare benefits (indeed all federal expenditures) from a progressive consumption tax, but better the status quo than no SS at all.

Zeke writes:

Do these returns include employer matching payments when calculating returns? If not, they overstate the return because the incidence of payroll taxes fall at least somewhat on labor.

david condon writes:

Is this the rate of return until 65 or until the person dies? If it's the former then the rate of return is worse than expected relative to a private account because a private account would continue earning interest after 65. In the latter case, it's better than most are thinking because a retiree with a private account would likely be withdrawing funds faster than the interest gain and have a much lower rate as they chose safer investments.

ThomasH writes:

I suggest that robbbbbb auction off his SS benefits. Someone would give him something and by his estimate, he will lose nothing. I'd recommend that my daughter who is about robbbbbb's age invest a few thousand to get his SS benefits if he were game.

ColoComment writes:

ThomasH -- I, for one, would take you up on that suggestion, were it legal, for "something" equal to my contributions during my working career (first job = lifeguard in 1963).

At least at that point, I'd OWN the asset to which I've contributed all my working life, be able to invest it in my sole discretion for better or worse, & be able to pass to my heirs whatever portion I failed to spend during my lifetime.

Brian Donohue writes:

@Kevin Erdmann,

I'm not sure if this is an apt comparison. Social Security provides a minimum floor of income. Its very existence allows people to take more risk with respect to additional retirement savings.

It's the old "don't bet the rent" advice. Social Security covers the rent.

Also, disability. Also, higher paid people live longer. Also, it's social insurance- if I get my money back, I'm ok with this.

Brian Donohue writes:

If only all of out government budget problems were as easy to solve as Social Security...

Keith K. writes:

Social Security cannot be considered an investment because the money was not put into things which will actually generate wealth. It was used to pay the beneficiaries at the time and the surplus used to finance more government spending. The trust fund is a Ponzi Scheme; IOU's piled upon other IOU's. There is no actual money in it.

There is also the opportunity cost of sending that money to the state vs. actually investing it in the private (read: wealth generating) sector of the economy. Imagine for example if in the 1930's, the program had been setup like Chile's current system of private forced savings accounts. Even if the return were exactly equal, the economy would be much stronger now, because the money that went into those accounts would actually have financed wealth generating enterprises. Compared to where it went then and now: the idiots at the cavalcade of Alphabet soup agencies.

Ricardo writes:

I have a dilemma. I'd like to save $5000 for my retirement, but I'd also like to spend that money. Wait, I have an idea! I'll "save" $5000 by writing my future self an IOU. Then I'll spend the $5000 on fun stuff today. Win-win!

Michael T writes:

@Brian Donohue

"Don't bet the rent" is good advice generally, especially if you need the rent soon. But what if your rent isn't due for 30 years and your bet is expected to return 5-10% yearly over the long run? Bet the rent! The risk premium from equities is worth their volatility at early stages of a long working career, and then less so as you gradually approach retirement. In my view, this is one of the main problems with SS. It robs people of the opportunity to invest an extra 12% of their income in wealth producing equities while they're far away from retirement and should be optimizing ROI. Instead, the money is simply removed from paychecks of workers, then sent to retirees. No investment here as far as I can tell, unless you're using the politician's definition of investment = public spending.

Bryan references keyhole solutions to address concerns about open borders, and one could make a similar argument here. Any problem you have with private retirement accounts can be resolved with a keyhole solution and would be better than the status quo.

-People won't save enough money? Mandate a minimum % savings for retirement. The US already deducts 12% now, maybe more in the future.
-People will act on bad advice with their retirement? Regulate asset mix based on age. The US already has the equivalent of a 100% bond investment for life.
-Don't trust financial institutions? The "privatized" Singapore CPF system has a default fund that is government managed.

Comments for this entry have been closed
Return to top