Monday, January 17, 2005


One of the common arguments against Social Security is the rate of return. Basically it boils down to “why should I get X% on the money I pay in Social Security Taxes when I can get X+% just putting it in the bank!”

I would like to first explore what the actual rate of return is on Social Security and what it means to have been able to earn something in the past. Social Security for many people will have a positive rate of return and for those who need it most the return will be quite good. Kevin Lansing had an interesting article published by the Federal Reserve Bank of San Francisco back in 1999. shows the real rates of return he found depending on birthdates. In general you are probably look at a real return around 1-2%. Keep in mind that a real rate of return is the rate you actually see minus inflation. This is what the author grudgingly concludes:

The authors find that, under current OASI rules, today's lowest paid workers (those in the bottom 20% of the income distribution based on lifetime earnings) can expect internal rates of return between 4% and 5% after adjusting for inflation (Panel A). Today's middle income workers can expect real rates of return between 1% and 2%. Today's highest paid workers can expect real rates of return below 1% and may even face negative rates of return if born after 1975.

If you explore deeper you’ll note that blacks and whites have nearly the same rate of return, however women outpace men quite a bit. This is caused by Social Security’s nature. Your rate of return is going to be higher if you live longer (which women do) or if you can collect from someone else’s contributions (which many women do as widows).

Private savings, which is both 401K’s and other types of savings in bank accounts, home ownership, even comic book collections has a different return structure. If you are higher income then you should be able to get higher than average returns because a larger account should have smaller transaction fees as a portion of investment. In theory a well educated low income person could find savings options at financial institutions that accept small deposits and do not charge high fees and also find mutual funds and index funds with low fees. However, this requires a great deal of study and I believe on average the typical low income saver is going to end up seeing a larger share of his return eaten up in fees than a high income saver.

Also, except in the unusual case where a person buys annuities, returns are lump sums. In other words, if you and your friend have the same 401K balance & return at 65 but you live 20 years longer than he does you get nothing extra except having to stretch the balance for that much longer.

A pattern that emerges is that Social Security provides a different type of return than private savings accounts. It provides a counter balance to what one expects from private savings. For the low income it offers much lower fees while private savings demands higher fees. For the long lived it promises larger returns while private accounts simply require one to stretch things out. While we are on that subject, let’s note that Social Securities benefit to the long lived is automatic. With a lump sum private account a person has to estimate how long they will live and spend accordingly. Since they cannot be certain about the future they will feel the need to keep some in reserve. If they still under-estimate they may end up with a serious problem, running out of money at the end of their retirement rather than the beginning. That would be at the point where they are less likely to be fit enough to take corrective measures like returning to the workforce.

Returns are Supposed to be Equal (with risk considered)

Let’s get to the meat of what returns are. Returns are paid to you by someone else. There is no magic ‘interest fairy’ who rewards you for putting $100 aside. Someone pays you for doing that and the reason they pay you is that they are able to take your $100 and do something with it to not only pay you back your $100 and your interest but also have some profit for themselves.

Imagine the economy that has no money and produces only cookies. Say a cookie baker is hungry. If you lend him ten cookies he will eat them and have the energy to bake 20 cookies at the end of the day. For this he agrees to pay you five cookies in interest. So the economy that starts out with ten cookies in the morning ends the day with 20 cookies…you have 15 (ten cookies plus five interest) and the baker has five of his own. Everyone wins, you are richer and so is the baker. Return wise, economic growth was 100% (ten cookies to twenty cookies) and your return was 50% (ten to fifteen). The balance of the growth was given to the baker.

At some point, total returns are capped at economic growth. A casino generates an average return based on how they set their odds (but they always set the odds so that in the long run they and not you win). Like a casino, however, everyone doesn’t split the loss or gain evenly. Some people win big and others lose big. At the end of the day, if the casino took in net $1M from 1000 people then the average person must have lost $1000. But if one person won $3000 then someone else must have lost more than a simple $1000.

In an economy the returns realized by people is capped by economic growth. If one person is able to get returns larger than the growth of the economy then someone else must be paying for it. This leads to confusion because historical returns for both stocks and bonds make it appear that you can get a better return than economic growth.

The element that is missing is risk. Everything except maybe death and taxes is a risk. All assets carry with them a host of different risks. Cookies can become stale and worthless. Your favorite stock may turn out to have lied about their books like Enron. The gov’t may change the tax laws so you can’t cash in on some really profitable investment without taking a huge tax bite.

Another missing element is realized returns. A stock starts the year at $50 a share and ends it at $100. That’s a 100% return even though the economy only grew at 5%. But what was the realized return? Price quotes for stocks are based on actual sales so we can be sure at the start of the year at least one block of 100 shares was sold for $50 each and at the end at least one block was sold for $100. Someone realized a gain of $5,000. But could the entire economy have pulled off the same trick? No. We know supply and demand will tell us if a huge number of people tried this the price at the beginning of the year would soar. Likewise the price at the end of the year when everyone tries to cash in would be less. The more money that tried to capture this lucky stock the less its paper return would have been.

This doesn’t stop some commentators from saying silly things like “If only people could have taken their social security taxes and brought that stock…they would have had more money in retirement than they would have gotten from social security if they lived to be 190!” What goes for a particular stock also goes for a mutual fund or anything else. If all of Social Security had been ‘invested’ in some mutual fund that yielded 20% returns in the 70’s and 80’s all that would have changed would be that funds real returns.

Realized returns also are the returns that really matter because those returns are what puts food in your mouth. In the example of the stock that went from $50 to $100 in a year, all that tells us that at least one trade happened in the beginning of the year and one at the end. All the people who own the stock during the year will see their ‘balance’ double but that doesn’t become real for them until they actually sell their shares and take the profits.

So in order for private investments to generate realized returns in excess of economic growth to finance retirements other returns have to be less than economic growth…even negative.

This leads me to returns all being equal given risk. There’s a host of different returns at any given time but there’s a trade off with higher returns. They have higher risk. If they didn’t people would sell off their assets and purchase the asset that generated a high return with lower risk. This would increase the price of the asset (thereby decreasing its return) until the return was in line with a low risk-low return asset. If markets work, then returns will fall along a line of increased return to increased risk similar to the CAPM model.

The author concludes that Social Security’s return is inferior:

The "average" U.S. worker faces a rate of return on contributions that is quite low--less than 2% after adjusting for inflation. By comparison, the real yield on a 10-year inflation-indexed Treasury Bond is currently around 3.5%. In addition to being low, rates of return from Social Security must be viewed as risky because they are subject to change from future political actions that will be needed to ensure long-term solvency of the program.

What’s wrong with this? Well Social Security has only political risk and economic risk. Political risk in that the gov’t may change the rules of the game. Economic risk in that the entire economy may not be able to fund benefits in the future.

The 10 year inflation indexed bond has a stated yield of 3.5%. Bonds, however, are no different than stocks. They can increase or decrease in value. The 3.5% yield is only what you earn if you hold it to maturity and the interest doesn’t compound. Every year will have coupon payments that you will hope to invest for at least 3.5% Plus remember transaction fees. While you can purchase bonds directly many times they are sold through brokers and mutual funds. As soon as this happens you incur fees that lower your return. If you do not have enough money to buy a whole bond you have to buy them in a fund which automatically exposes you to fees. For an asset as safe as a 10 year inflation protected T-Bond, we can identify all of Social Securities risk elements:

1. Political risk – That the gov’t may not honor the bond given the horrible state of the overall Federal Budget. That gov’t may ‘change the rules’….it may apply taxes against the interest income of the bond, it may change the rules for capital gains taxes on the bond, and so on.

2. Economic risk – the Economy may collapse leaving the gov’t unable to pay off the bond or its interest.

But the bond also carries two additional risks that social security dodges:

1. Interest rate risk – Interest rates may fall leaving you unable to re-invest your coupon payments at 3.5%.

2. Market risk – The price of this bond goes up and down every day. If you sell the bond before it matures you cannot predict ahead of time how much you’ll receive until you actually do it.
So looking at it objectively it doesn’t appear that the case for Social Security being a raw deal has really been made.

Sunday, January 16, 2005


Social Security: Not a Problem

Hi Everyone, sorry I've been away for so long. I want to get back to ecnomics and I'm especially interest in the Social Security debate. The next few posts will concentrate on Social Security:

Roger Lowenstein has an excellent piece on SS in this sunday's NYT Magazine

But is it? After Bush's re-election, I carefully read the 225-page annual report of the Social Security trustees. I also talked to actuaries and economists, inside and outside the agency, who are expert in the peculiar science of long-term Social Security forecasting. The actuarial view is that the system is probably in need of a small adjustment of the sort that Congress has approved in the past. But there is a strong argument, which the agency acknowledges as a possibility, that the system is solvent as is.

The projections:
Politicians and other commentators tend to speak about these long-range trends, or at least about Social Security's finances, with an air of precision. This is almost amusing, since few economists can predict the swings in the federal budget even a year in advance. Joshua Bolten, head of Bush's Office of Management and Budget, said of Social Security last month, ''The one thing I can say for sure is that if left unattended, the system will be unable to make good on its promises.'' But the Social Security Administration itself pretends to no such certainty. Its actuaries (about 40 are on staff) frankly admit that the level of, say, immigration in 2020, or of wages in 2040, is impossible to forecast. ''The only thing we are sure of is that it won't happen precisely as we project,'' says Stephen Goss, the chief actuary at the agency. And the trustees' annual report, which is based on the actuaries' analysis, takes pains to say that it is not making a prediction. It makes a projection -- three different ones, actually -- that amount to informed but very rough guesses. The agency's best guess, labeled its ''intermediate'' case, is that the system will exhaust its reserves in 2042. At that point, as payroll taxes continue to roll in, it would be able to pay just over 70 percent of scheduled benefits. That would leave a substantial deficit, but one that Congress could easily avert if it were to act now when the projected problem is more than a generation away.

So the 'huge liability' amounts to only 30% of benefits. Therefore if nothing is changed 70% of benefits will be paid for by not touching taxes. If current taxes will take care of 70% how exactly is the remaining 30% going to swamp the entire GDP????

Past performance is no indication of future success but:

No one can definitively predict that outcome, either, of course, but David Langer, an independent actuary who made a study of Social Security's previous projections compared with the actual results in 2003, thinks the ''optimistic'' case is its most accurate. Over a recent 10-year span, the trustees' intermediate guesses turned out to be quite pessimistic. Its optimistic guesses were dead on, and its pessimistic case -- sort of a doomsday situation -- was wildly inaccurate.

And, contrary to widespread belief, recent demographic trends have been modestly better (from an actuary's gloomy standpoint) than anticipated. For instance, longevity hasn't increased as much as expected. Partly as a result, since 1997 the agency has pushed back, by 13 years, the date at which it projects its reserves will be exhausted. In other words, as the cries of impending doom started to crescendo, the guardians of the system have grown more optimistic.

Interesting how critics of global warming are so eager to jump on predictions that failed topan out.

BTW, rising wages do push SS more into the black. Contrary to what some critics have said, rising incomes do allow us to 'grow' out of any problems:

Rising wages are also a boon to Social Security's finances. Forecasting wages is difficult, as the trustees' report frankly admits, but it seems undeniable that as society ages, businesses will be harder pressed to find workers, and that should push wages higher. The trustees, however, project that real wages will grow at only 1.1 percent a year -- roughly equal to the level of the last 40 years

This point has been explored by other like Brad DeLong. Basically only the beginning benefit is indexed to wages but after that it is indexed to price. If wages are rising and I retire tomorrow, that will boost my benefit. But if wages continue to rise then the tax revenue coming into the system will increase while my benefits remain constant...adjusted only for inflation.

Putting the cost into perspective:

One way or another, societies with more old people have to devote more resources to them. Right now, benefits amount to 4.3 percent of G.D.P. The trustees' most likely projection assumes that over the next 75 years that figure will rise to 6.6 percent. In the more optimistic case, benefits will rise to 5.2 percent. Given the substantial increase in the elderly population, neither of these figures seems rash or out of proportion. The increased cost would be on a par with that of making Bush's first-term tax cuts permanent, which is projected to be about 2 percent of G.D.P.

This page is powered by Blogger. Isn't yours?