At the heart of President Bush's plan to sell Social Security private accounts is a simple notion: You're always better off investing your retirement money than letting the government do it.
By doing it yourself, you can stow some money in the stock market, and over the long run will get a better return on that investment than today's Social Security system offers.
The idea is broadly accepted. That's why the administration's plan to partially privatize the system sounds appealing to many. But that better return won't always happen.
Just ask Stanley Logue of San Diego.
For 45 years, the defense-industry analyst paid into the system until his retirement in 1994. But with all the recent hoopla over reform, Mr. Logue, a Massachusetts Institute of Technology graduate, decided to go back and check his own records. Would he have done better investing his money than the bureaucrats at the Social Security Administration?
He recorded all the payroll taxes he paid into the system (including the matching amount from his employer), tracked down the return the Social Security Trust Fund earned for each of the 45 years, and then compared the result with what he would have gotten had he been able to invest the same amount of payroll tax money over the same period in the Dow Jones Industrial Average (including dividends).
To his surprise, the Social Security investment won out: $261,372 versus $255,499, a difference of $5,873.
It's an astonishing finding. The DJIA represents blue-chip stocks. Social Security invests in US Treasury bonds. Over long periods of time, stocks have consistently outperformed bonds. So, you would think that Logue's theoretical stock investments from 1950 to 1994 would have surely outpaced the return on government bonds.
The fact that they didn't illustrates one of the hard truths about stock investing: Timing matters.
Although Logue started pouring money into Social Security in the 1950s and early 1960s, some of the best years for stocks, he hadn't accumulated a lot of money.
So the gains of his theoretical stock portfolio would have been limited.
By the time he had substantial sums, the market swooned for long periods. From 1965 to 1982, for instance, the DJIA made no progress. Logue retired before the real run-up in stocks in the latter half of the late 1990s.
So the real lesson from his analysis is that any pension plan based on stock investments carries extra risks.
Advocates of privatization point out - correctly - that Logue's analysis compares theoretical stock returns with what the Social Security Trust Fund earned - not what he himself would get from the system.
From that perspective, the investment approach looks better, they argue. Over the long run, a typical worker can expect to earn 4.6 percent a year (after administrative costs) on a diversified portfolio of stocks and bonds and only about 2 percent or less from Social Security, according to federal estimates reported by Michael Tanner of the Cato Institute, long a proponent of privatization. Hypothetically, someone earning $30,000 annually would at the end of a 40-year career receive nearly twice as much under the investment approach ($344,000) than with Social Security ($185,000).
Who's right: Logue or Mr. Tanner?
The debate hinges considerably on what people want their retirement system to be. Social Security has always been an insurance program. It was never intended as an investment scheme. So everyone - retirees, the disabled, widows, and orphans - receive guaranteed monthly income. The "return" on their Social Security contributions depends largely on how long they live. Those in their 90s have enjoyed superb returns. Those who don't live as long benefit less.
Private accounts, by contrast, involve far more variability, both sides agree. Individuals who enter and exit the market at the right times would undoubtedly do better under privatization.
But under Britain's privatized pension system, so many retirees are doing so poorly at this moment that a commission warned this fall that widespread poverty among the elderly may be returning, which could require massive new government spending.
Presumably, President Bush's plan would offer the choice to meld insurance and private investment: much less guaranteed income in return for the opportunity - and risk - of earning more in the markets.
"Because financial asset returns are volatile, benefits under a personal account system would fluctuate," notes Bill Dudley, an economist at Goldman, Sachs & Co., a New York investment bank. "On a risk-adjusted basis, the privatized account ... becomes much less compelling."
There are other problems with private accounts. Administration expenses of the present Social Security system are minuscule compared with the size of the benefits provided. The Bush administration so far has provided no details on its private accounts plan. But if these are handled by Wall Street, the fees could be sizable, dissipating some of the return from investing in stocks. Logue takes no account of such expenses in his analysis.
Further, administrative costs and difficulties for private business could be large as companies, big and small, try to deduct the right amount from a payroll and put it into a private account in a timely fashion.
A study by the Congressional Research Service (CRS) notes some complexities: 650,000 employers go out of business or start new businesses each year. More than 4 million employers have 10 or fewer employees, often having record-keeping problems and errors. About 12 million to 15 million individuals are self-employed and presumably would have to send money directly to a private account.
So the complexities of change are substantial. If the extra return from privatization is not very advantageous, "why even consider changes that all agree would be very disruptive?" asks Logue.