Arnold Kling  

Social Security Reform

PRINT
Immigration Policy... Meeting the Enemy...

The Social Security Administration has evaluated a proposal by Peter Ferrara for private social security accounts.


This plan would establish voluntary, progressive individual accounts for workers who are under age 55 on January 1, 2005 and would provide for a reduction in the Social Security retirement and aged survivor benefits for those who participate. All participating workers would be guaranteed that the total benefits available from the combination of the OASDI program and their personal account would be at least equal to OASDI benefits scheduled under current law if they choose the default investment option...
Individual account (IA) assets would be invested by individual workers through a central administrative authority with a default allocation 65 percent in broad indexed equity funds and 35 percent in broad indexed corporate bond funds.

To make up for the loss of tax revenue to pay Social Security beneficiaries, the Ferrara plan contemplates a combination of lower government spending elsewhere and increased government borrowing.

The analysis makes the assumption that the real return on stock market accounts will be 6.5 percent. If the overall economy grows at 3 percent and corporate profits also grow at 3 percent, then these stock market returns must come from a drop in the risk premium, which eventually has to fall below zero.

I have raised this issue before. I think that the assumption of 6.5 percent real returns for stocks distorts the analysis of Social Security privatization. Although I am sympathetic to privatization, I believe that the analysis ought to be conducted as carefully and reasonably as possible.

[note: I initially misread the SSA document and thought that they assumed 7 percent returns, not 6.5 percent, on the stock market]

For Discussion. Under Ferrara's proposal, those of us under 55 could direct a portion of our Social Security taxes to private accounts. We would be guaranteed a minimum return equal to what we would have received from Social Security, provided that we invest in the "default portfolio" of stock and bond index funds.

Assuming that the guaranteed minimum return induces everyone to choose the "default portfolio," how does Ferrara's proposal differ from one in which the government itself invests payroll tax receipts in the default portfolio?


Comments and Sharing


CATEGORIES: Social Security



COMMENTS (28 to date)
Eric Krieg writes:

I don't get it. Social Security's return is NEGATIVE for me (32 years old)!

Mcwop writes:

Agree with Eric on this one. What return? Varies greatly be person and income earned over time.

Rather than use a 7% real return, the should do some Monte Carlo analysis to look at probabilities for worst, best, and median scenarios. The worst scenario could be the guarenteed minimum, though I would rather see no guarentees.

For money directed into private accounts retirment date funds (where asser allocation shifts as you get older) should be the only option. That should prevent people from making any real big mistakes.

Boonton writes:

Why wouldn't society's long run return on equity be the same as long run economic growth? If the economy grows 3% per year for the next 50 years how can equity return 7%!?

dsquared writes:

Eric: Remember that there is an implied insurance premium in Social Security investments.

McWop: But then the argument just moves to what parametric assumptions one puts into the MC model.

Boonton: In principle, the investments could be put into the securities of other, faster-growing economies, although I must say I regard that possibility as highly unlikely from a practical standpoint.

john top writes:

Great. Another additional burden on us young folks. If we go to a privatized system then I will have to pay for two retirements: the current old and mine (since when I'm old the young won't be paying for my social security).

I've been thinking that one way around the problem is to take care of the current social security obligations (current retirees and what non-retirees have already paid in) by issuing perpetual bonds. Future generations won't be paying into social security so they can pay off the bonds from the benefits of *not* being in this lousey system.

Bob Dobalina writes:

Is anyone else terrified of the corporate welfare that will likely result of privatization?

If "retirement date" funds are required, what will a fund company have to do to get approval? Will the government encourage funds-of-funds that add an additional layer of management fees to already pricey mutual funds? Will there be an office in the SSA that mandates asset allocation for each set of retirement years? Will certain benchmarks be favored at the expense of others? Will we see the "about-to-be-added-to-the-S&P" effect writ large, resulting in speculation for firms ranked 501-550 in market cap? Will arb firms trade these companies, inflating their prices prior to the index inclusion to take advantage of a government regulation? Will the government let me underweight financials and overweight mining companies? If a balanced fund is required, can I avoid highly interest-rate sensitive equities? Will I have to own Philip Morris if I wish to own stocks in my retirement plan? Will the government make me own corporate bonds?

et cetera.

Just kill Social Security altogether.

Eric Krieg writes:

>>Eric: Remember that there is an implied insurance premium in Social Security investments.

Typical left-winger response.

The disability portion of SS is a small fraction of its cost. Being generous, say 10%.

The real cost of SS is the middle class entitlement. You know, setting up Grandma in Sarasota in the lifestyle to which she has become accustomed.

Lawrance George Lux writes:

Dealing with Social Security funding is like dancing with the Devil. The first constraint which must be understood, lies in the context of Irreducible Expenses of Households. American society will insist on Elderly Provision. Elders who cannot pay their Irreducible Expenses will apply for, and receive, their Welfare transfer packages. This is reality. Would utilization of Welfare programs be more expensive than Social Security benefits? This is also reality.

Any form of Privatization of Social Security ties eventual benefits to the Stock Markets. This is not necessarily bad is the Short-run, but long-run performance insists such participation will reduce Profits and Dividends; as Corporations continue additional Stock Issues to maintain price, and pay usurious rate of remuneration to Corporate Executives. The Issue of Irreducible Expenses remain, as does the fact that Privatization of Social Security will vastly expand other Welfare transfers at greater expense; while diluting Dividends for all Corporate Shareholders. lgl

Mcwop writes:

D2 - good point, though even I have enough confidence that responsible government agencies can get the inputs mostly right.

Bob - I would not have fund companies do it. Perhaps the Federal Reserve could manage them at cost, and use a combination of active and passive management.

I do agree that there become poliitical and personal preferences involved with allocation selection and secirty selection. My model would be similar to the federal thrift program, and pray that sensibility rules the day. Probably, won't be that much worse than the current system.

gerald garvey writes:

"If the economy grows 3% per year for the next 50 years how can equity return 7%!"

In the past, equity in the US has grown faster than 7% on average. But we also need a reason, since this observation is full of selection bias and also a lot o' noise.

One obvious answer is leverage; fixed costs and debt payments. Corporate profits are only a very thin slice at the bottom of a company's income statement. If you think of the average firm's sales as growing at 3% then it is not at all hard to have profits growing at 7%. OTOH, this also implies lotsa risk. E.g., if the economy's growth stalls, this also implies that profits should fall dramatically. Kinda what we see.

Bob Dobalina writes:

"I would not have fund companies do it. Perhaps the Federal Reserve could manage them at cost, and use a combination of active and passive management."

If I own Ford, and the guy at the treasury (who manages a trillion dollars in Social Security funds) decides he's going to buy GM, I'm seriously, seriously pissed off. No way the government ought to show favoritism in these affairs and I think it's unavoidable.

Boonton writes:

"In the past, equity in the US has grown faster than 7% on average. But we also need a reason, since this observation is full of selection bias and also a lot o' noise."

In the past winning lottery tickets have also generated returns on the order of hundreds of thousands of percent. Of course, if we went back in time and everyone brought the winning number the real return would be less than a dollar because the pot would be divided among everyone who brought tickets.

So to my mind the 7% average is like all the winning lottery tickets in the past. The more relevant question is how much realized returns did the stock market give people in the past? This would incorporate all the dumb decisions that people made like selling their Microsoft shares in 1992 because they wanted to buy a new stero. A chunck of those 7% returns depended on people doing dumb things like that just like the big winner at the casino only exists because there's thousands of little (and big) losers.

If you went back in time and put all of the SSI money into privatized accounts that encouraged a 'default portfolio' favoring stocks you WOULD NOT see the same 7% returns. Such a dramatic change would have to alter history meaning the returns would be different.

In the end Keynes's lesson is important IMO. In a sense it is impossible for a society to save unless they actually bury goods and services in a time capsule for later consumption. Whatever paper is brought today will not change the fact that tomorrows retired (as well as tomorrows disabled, babies, welfare cases etc.) will have to be fed on tomorrows output.

Boonton writes:

Bob:

The argument that gov't can't 'invest' SSI funds because of political bias is not convincing. Here is a simple way to do it, buy shares by using a random number generator. Another way would be a market index. Simply take all the available stocks, figure their market capitalizations (share price times # of shares) and then figure out what portion of the entire market capitalization each one represents. Divide your portfolio along those lines.

In both these cases the gov't (or fund manager) would be showing no bias towards any stock and relative prices would be set by the market (although I'm sure the influx of SSI money would push all prices upwards). It is my thesis, though, that this would accomplish nothing. The total returns of the entire system cannot exceed the overall return (growth) of the economy.

If stocks had an impressive return in the past you cannot deny that part of it had to do with the fact that SSI indirectly channels money into T-bonds. If SSI money went into stocks the price of all stocks would rise thereby lowering stock returns. Bond returns, though, would increase since bond prices would collapse.

Bob Dobalina writes:

Boonton:
"Here is a simple way to do it, buy shares by using a random number generator"

Unbiased, but still unfair.

"Another way would be a market index. Simply take all the available stocks, figure their market capitalizations (share price times # of shares) and then figure out what portion of the entire market capitalization each one represents."

You'd then be using the Wilshire 5000, I presume? Some of those stocks are very illiquid. Then the government would be filling the coffers of Knight, Lazard, Island, Instinet, and others.
To avoid the market makers screwing the taxpayer, then, we'd need an even bigger agency.

"If stocks had an impressive return in the past you cannot deny that part of it had to do with the fact that SSI indirectly channels money into T-bonds. If SSI money went into stocks the price of all stocks would rise thereby lowering stock returns. Bond returns, though, would increase since bond prices would collapse."

Decompose the return of stocks into earnings and their multiple. Decompose earnings into real earnings and inflation. I don't have the data handy, but I think you'll find that the reason for the extrordinary returns of equities is an expansion of the p/e multiple. Many would attribute this to low interest rates. And I disagree that the price spike that the influx of capital would cause will be permanent. Over time, people who aren't idiots would rather own a car wash or laundromat at 5 times earnings than Gilette at 30 times.

If you indeed feel that bond prices would collapse, and interest rates would head north, wouldn't you suspect that firms would change their capital structures, trading debt for equity? Someone earlier in the thread mentioned that leverage is what enables corporate earnings to outpace aggregate economic growth. In a high-interest rate scenario, much of what we've seen in the past 15 years will be turned on its head.

Boonton writes:

"Decompose the return of stocks into earnings and their multiple. Decompose earnings into real earnings and inflation. I don't have the data handy, but I think you'll find that the reason for the extrordinary returns of equities is an expansion of the p/e multiple. Many would attribute this to low interest rates. "

An expansion of the p/e multiple basically means people are willing to pay more for a $1 of earnings coming from the stock market than they were in previous times. Obviously if you could go back in time and get into the market, you will benefit with a healthy return as society started to favor getting their earnings at the stock market rather than the bond market.

"If you indeed feel that bond prices would collapse, and interest rates would head north, wouldn't you suspect that firms would change their capital structures, trading debt for equity? "

Of course and if more firms started paying off their bonds by selling stock what impact would that have on stock prices? It would dampen upward movement in stock prices and if you do not have upward movement in stock prices then you've blown out one possible way to get a return with stocks...increases in the equity. That leaves earnings as a possible route to achieve returns. But earnings are capped at the ability of the economy's overall growth. Any company whose earnings are growing faster than the economy will eventually have to slow down...unless that company becomes the entire economy!

BfloGuy writes:

Government must not be allowed to invest the money. Anyone who thinks that politics wouldn't influence these investments is naive.

The temptation to steer this immense amount of money into politically fashionable industries would be huge.

And, no, I wouldn't want my portion invested by some random computer pick. I'll do it myself, thank you, and benefit or suffer accordingly.

Jim Glass writes:

"Why wouldn't society's long run return on equity be the same as long run economic growth?"

It is and always has been, or close to it.

E.g., the Dow Jones Industrial Average has risen an average real 2.9% for the last 80 years, which is a little less than the average economic growth rate over the period.

(The DJIA was 95 in 1923, which inflation-adjusts to about 1025 today, which compounds at a little less than 3% for 80 years to give today's 9800.)

"If the economy grows 3% per year for the next 50 years how can equity return 7%!?"

This comes up again and again. The answer is as simple as the difference between simple and compound interest.

The famous 7%long-run return to stocks is simple interest -- 3% growth plus 4% in dividends and stock share repurchases on net.

After people have spent that 4% of their annual simple return from stocks on other things, the remaining 3% has compounded over time.

Regarding an investor's typical life cycle of saving for retirement one can imagine an early-years period of letting dividends compound to get the full 7%, then a later period of spending dividends and having principal grow at 3%, then a period of liquidating principal, all multiplied by millions of investors at all different stages of the cycle, averaging out to a 3% compound growth for stocks overall.

Frankly I don't understand why so many people keep saying this long-term compound return on stocks that basically matches the growth rate of the economy is "unsustainable" (especially since it's been sustained for 200 years so far). Nobody's ever explained that to me.

There seems to be an endemic error of people thinking the compound growth rate of stocks has been 7% -- the Dow 36,000 mistake of thinking dividends are both spent and reinvested simultaneously.

BTW, if one looks at the capitalization of all stocks in all US markets rather than just at the DJIA or S&P 500 the same basic picture remains. The compounding growth rate in recent decades is more like a little over 4% than under 3%, but the businesses represented are tapping new business opportunities all over the world, in Asia and elsewhere, and aren't restricted to inevsting in just the US economy, so personally I don't have any problem with that growth rate either.

Jim Glass writes:

"The analysis makes the assumption that the real return on stock market accounts will be 6.5 percent. If the overall economy grows at 3 percent and corporate profits also grow at 3 percent, then these stock market returns must come from a drop in the risk premium, which eventually has to fall below zero. I have raised this issue before. "

[From which...]
"We also disagree about the long-run return to capital, which he says can be 'rationally' expected to exceed the growth rate of the economy by a large amount... If the market value of stocks is going to outgrow the economy, then either the price-earnings ratio has to rise or the earnings-to-GDP ratio has to rise...."

But ... but ... but (as Brad DeLong would say) note how "long run return to capital" in the first sentence transforms into "If the market value of stocks is going to outgrow the economy" in the second. And those are two different things!

Facts: Over the last 80 years the first of these, the average annual real return to the DJIA (7%), *has* exceeded the growth rate of the economy; but the second, the growth rate of the real market value of the DJIA (2.9%), has *not* exceeded the growth rate of the economy.

I still don't see why this situation is supposed to be unsustainable. Since the market value of stocks has *not* been outgrowing the economy, what's the objection?

And why isn't this same objection made about corporate bonds? The average return on them exceeds the growth rate of the economy too (as a quick look at Moodys verifies) but nobody ever says that's unsustainable.

I can only imagine that's because everybody knows that bonds don't compound at their average annual rate of return, while for some reason thinking stocks do.

Bob Dobalina writes:

Jim: The Dow isn't representative of the market as a whole, especially considering the margin by which small companies have outperformed mega-caps over the past 75 years.

Boonton: "But earnings are capped at the ability of the economy's overall growth".

Again you fail to account for leverage. Why?

Jim Glass writes:

"Jim: The Dow isn't representative of the market as a whole..."

Which is why I specifically noted that the same thing is true for all stocks on all US markets, even giving the difference between them.

The Dow *is* representative of the market as a whole on this. If small stocks (which after all represent only a small portion of the market's capitalization) appreciated faster than the growth rate of the economy over the same period the small stock indexes today would be 10x higher than they are.

And nobody would ever invest in big stocks and give up that 4+ points of average compound return per year. ;-)

Bob Dobalina writes:

"If small stocks (which after all represent only a small portion of the market's capitalization) appreciated faster than the growth rate of the economy over the same period the small stock indexes today would be 10x higher than they are."


But the small stocks become midcaps and no longer participate and contribute to the smallcap indices.

And, Jim, you're in the business, right? Surely you've seen the Ibbotson SBBI chart which breaks smallcaps from largecaps.

Arnold Kling writes:

Jim wrote, "Facts: Over the last 80 years the first of these, the average annual real return to the DJIA (7%), *has* exceeded the growth rate of the economy; but the second, the growth rate of the real market value of the DJIA (2.9%), has *not* exceeded the growth rate of the economy."

Going forward, make the simplifying assumption that stocks pay no dividends. Thus, the return on capital is simply the capital gain on stocks.

Under that assumption, if the return is 6.5 percent, then stock prices have to go up 6.5 percent. That means that either earnings have to grow faster than the economy, or the P/E ratio has to grow indefinitely, or both.

Adding dividends gives us another term to keep track of in the model, but it does not alter the economics.

Jim Glass writes:

"Going forward, make the simplifying assumption that stocks pay no dividends. Thus, the return on capital is simply the capital gain on stocks."

Going forward, make the simplifying assumption that all interest on corporate bonds will be 100% reinvested in bonds so that bonds compound faster than the growth rate of the economy, and income from them is never spent. Therefore bond interest rates are unsustainable. Is that an argument too?

But, but, but ... why would fundamental human behavior change all of a sudden after 200 years so that what has always been sustainable suddenly becomes unsustainable?

People have always wanted income from their investments, including from stocks. Why would people suddenly not want income from investing? What would be the point of investing then, if income could never be spent -- because it all had to compound forever?

It doesn't look like a simplifying assumption to me, but one that changes fundamental reality.

(If we make that simplifying assumption as of 1923, which seems as good a date as any, then the Dow is over 200,000 today -- and no income from stocks has been spent in 80 years.)

BTW, a second thing I don't understand about the argument: Stocks do not = 100% of capital. There are plenty of investments that return far less than the growth rate of the economy, down to short-term paper.

I'll take it that the *average* return on capital should approximate the growth rate of the economy (disregarding investments abroad) but in the great national portfolio of all investments wouldn't we expect some types to return less than the average and other types more than the average on a sustained basis?

E.g., if short-term paper is going to return less than the average growth rate of the economy continuously, doesn't some other kind of investment have to return more than that average on a sustained basis to get total return up to that average?

Jim Glass writes:

Bob wrote:
"And, Jim, you're in the business, right?"

I'm in a business, not the investing business. ;-)

"Surely you've seen the Ibbotson SBBI chart which breaks smallcaps from largecaps."

Sure. Ibbotson says small cap stocks have produced an excess return over large cap stocks of 1.9 points since 1926 (having trailed for up to decades at a time.) That leaves them compounding at 5% not 7%. And, of course, by definition, they are only a small portion of total market capitalization.
~~~

Our gracious host wrote:

"Adding dividends gives us another term to keep track of in the model, but it does not alter the economics"

I'd say it alters the economics *of investing* quite fundamentally.

Say I own a mature business that has no growth opportunities but which for a while will be a cash cow yielding 10% simple on my invested capital. How can I reinvest that money in the business to earn 10% compound? I can't -- it is *impossible*. I must take that money out of the business one way or another.

Now say the economy as a whole does not offer businesses in the aggregate growth opportunities exceeding the growth rate of the economy, say 3%, although they can earn yields exceeding that rate, say 7%. How can businesses invest their yields to earn a 7% compound rate? They can't -- it is impossible.

And it *is* impossible -- because 7% compound is much more than 3% compound plus 4% simple. It is a higher rate ... and business owners want to earn the highest rate possible ... so if it was possible to get it, they'd already have their businesses getting it!

IOW moving from 3% compound + 4% simple to 7% compound is not a "simplifying" assumption -- it is assuming (1) a whole different *higher* rate of return; and also (2) that investors have voluntarily passed on it for all these years, for some unknown reason.

The logic that long run stock returns cannot compound at much above the growth rate of the economy is quite persuasive -- and entirely consistent with the data.

I think you underestimate its power -- it's already been governing things for all these years.

gerald garvey writes:

Boonton:

The 7% historical return is NOT the return to lottery ticket winners only. It assumes neither good nor bad markt timing. In terms of your example, if someone mis-times the market he will earn say 3% but this implies that the party on the other side earned 11%, for an average of 7%. The argument you are making is instead straightforward paternalism. Some fools will make bad trades and SS is good because it does not let them do it.

Jim Glass: thanks for pointing out leverage again. Not sure why the rest of the posters cannot deal with this. Equities are leveraged. They earn higher average rates of return than the underlying assets/economy. By this same token, they are also more volatile. The higher average return compensates for the higher risk. Maybe a homeowner analagy will help? Suppose I buy a home for 50,000 and put down only 10,000. If I sell my home for 60,000, that represents a 100% return on my investment. The growth of the house value is only 20%. The risk part comes because if the house price falls by 20%, I lose all my money.

Eric Krieg writes:

How does one leverage a 401k account?

Jim Glass writes:

My comment about Ibbotson's data on small stocks seems confused when read in daylight. I'll try again briefly.

Ibbotson says small caps returned 1.9 points more on average than large caps since since 1926. Rounding, that means a real 9% versus 7%. *Assuming* that extra 2 points was all compounding then that would have been 5% with a 4% dividend return. But there's no reason to assume that as not all small caps are growth stocks, there are many mature small cap businesses. I don't have the data on at hand at the moment to say.

Boonton writes:

"Boonton:
"Here is a simple way to do it, buy shares by using a random number generator"

Unbiased, but still unfair."

I don't see how. Unlike appointing a national 'pension fund manager' who can be lobbied by interest groups to direct the investments...a random purchase of stocks would not permit companies to do anything that would increase their chances of being brought as part of a SSI 'pension fund'. If done in a random way (and I admit that it would be more complicated than just a random number generator) then the return would basically be the stock markets return.

My point is that if the stock market really had superior returns, they could be captured without privitized SSI. There are a finite amount of returns in the system which have to be capped by the economy's overall growth rate, therefore if SSI is making 'horrible returns' while the stock market is making great returns...SSI funds cannot be moved to the stockmarket unless stock returns fall.

The whole point of returns is goods and services. I put $100 into the stock market in January and in December I sell my shares for $107. I can now demand $7 more goods and services when I buy my Christmas presents. The economy must supply those goods and services. If it cannot then somewhere else the return must have been less than 7%!

Comments for this entry have been closed
Return to top