Monday, January 17, 2005

Paul Krugman commits a cardinal sin of logic regarding Social Security, and repeats it, and repeats it...

That logical sin: looking at an option and condemning it without considering the alternative.

This is now the standard operating procedure of the defenders of the Social Security status quo who denigrate the higher returns that can be expected from real investments in private accounts as being "risky", "expensive" or whatever. As they do this, they never compare such returns to the expected returns from Social Security as they want to keep it. In a moment we'll see why.

Now Krugman is yet again banging out this tune on his one note drum. A little over three weeks ago he wrote in his column that the British system of private accounts is risky and has fees that are much too costly. In his most recent column he wrote that the British system of private accounts is risky and has fees that are much too costly. The difference? This time he quotes from an article in that impartial and esteemed* source of economic analysis, the American Prospect...

"Britain's experiment with substituting private savings accounts for a portion of state benefits has been a failure ...A shorthand explanation for what has gone wrong is that the costs and risks of running private investment accounts outweigh the value of the returns they are likely to earn.... Reductions in yield resulting from providers' charges can absorb 20-30 percent of an individual's pension savings."
Yes indeed. If one's investments earn a 6% average return and one pays an annual 1.5% fee -- entirely plausible -- then returns will be knocked down by 25% and the growth in one's savings will be reduced accordingly. Yes. So if you invest $1,000 when young in a private account that earns an average 6%, then 40 years later when you retire instead of having it grow to $10,286, due to that fee it will be a mere $5,816.

That's 43% less! Ouch!! We don't want to do that! But wait ... maybe we do?

To find out, let's consider the alternative -- the returns provided by the Social Security status quo -- before making up our minds.

The first thing we note in doing so is that the Social Security actuaries say that all annual cohorts of retirees after 2000 will get back from Social Security less than they contributed to it, using the federal bond rate as the discount rate. That is, all will lose money compared to if they had invested in the lowest-yielding, safest investment available, "risk free" government bonds. And this loss grows as the years pass.

Somehow, defenders of the status quo don't consider this loss a "risk" ... perhaps because it is a certainty? It is, as it results from the legislated tax contribution-to-benefit formula.

Now, within each annual cohort of retirees annual return varies according to personal situation (single, married, one-earner or two-earner couple, etc.) We can look at some specific examples.

The Social Security Administration's actuaries projected returns on contributions for the 1994 Social Security Advisory Commission. Here are some numbers for persons entering the work force in 1994 (they'd be about age 30 today).

These are actuarial present values (in 1994 dollars) for lifetime contributions to, and lifetime benefits to be received from, Social Security. Benefits include all benefits, not just retirement benefits...

Single males

low wage: taxes $39,024; benefits $38,004; - 2.6%

Does Paul Krugman, the concerned progressive, tell us that from now on even many of the poor will be made poorer by Social Security?

average wage: taxes $86,720; benefits $62,889; - 27.5%

maximum wage (taxable wage limit): taxes $208,124; benefits $99,789; -52%. Ouch!

Single females (women live longer than men, so they do a little less badly)

low wage: tax $40,778; benefits $46,025; + 12.9%

average wage: taxes $90,620 ; benefits $76,185; - 15.9%

maximum wage: taxes $217,430; benefits $120,801; - 44.4%

Note that among married couples if both spouses work and have near-comparable incomes each will receive benefits on their separate work records -- so these "single" amounts substantively apply to many married persons too.

The relative big winner under Social Security is the working husband with a non-working spouse -- the standard family structure when Social Security was created -- due to the spousal benefit that provides benefits for two lifetimes on one earnings record...

Married male with family

low wage: taxes $39,024; benefits $77,963; +99.8%

average wage: taxes $86,720; benefits $129,863; +49.7%

maximum wage: taxes $208,124; benefits $205,916; - 1.1%

.... but note that even here, less than doubling one's money over 40 years is a pitifully low return, less than 2%. One can get that difference over government bonds just by buying high-grade corporate bonds, with no risk on stocks at all.

And even the most favored person in Social Security, the male bread-winner with the stay-at-home wife plus kids, loses money to Social Security if his income is as high as $90,000 -- which isn't exactly "rich" for a person supporting a non-working wife and children.

Of course there are other personal scenarios in Social Security: if you die before retirement age you take a 100% loss on your retirement benefit. If you are married and work to bring in the family's second income, but your lifetime earnings are low enough so that you claim benefits on your spouse's earning record, you take a 100% loss on the taxes you paid. If you end a marriage in less than 10 years you may take a big loss of benefits ... (These are not risks?)

But when you wrap them all up the actuaries say everybody combined takes a loss compared to investing in government bonds, from this day forward forever more, and the loss grows larger every year.

And NOTE THIS: Even the returns given above for Social Security are all underfunded by 30%. And if this funding gap is closed on a "paygo" method -- through the mix of 'modest tax increases and benefit cuts' as recommended by Krugman, the editors of the NY Times, and other such defenders of the status quo -- then by the Iron Laws of Arithmetic those returns must drop another 30%. (For either a tax increase or a benefit cut further reduces the ratio of benefits to taxes.)

Now the low-income male instead of suffering a 3% loss suffers a more than 30% loss ... the average-wage woman scheduled to take a 15% loss takes a 40% loss ... and the high-wage workers scheduled to take a 50% loss take a 65% loss.

(Maybe now that investment earning 6% that merely quintupled one's money after being knocked down by that heavy 1.5% fee looks at little better?)

There is no way around this if Social Security remains as it is. This is its future. Indeed, there is a real risk that benefits will be reduced by more and sooner.

So perhaps we begin to see why those who condemn returns in private accounts as being too low and risky never actually compare them to the returns guaranteed to today's young workers from Social Security.

Hey, let's be considerate and do that job for them:

Historically, stocks have provided an average real return 5 points above government bonds. But let's say one's private investments earn much less than that, only 3.5 points. Perhaps one is unlucky, or diversifies across lower-return investments such as corporate bonds (even government bonds) for safety.

Now let's take a realistic annual expense rate from real-world American experience -- noting how Krugman searches the globe to cherry pick investment programs that he can most easily present (or misrepresent, various folks would say) as examples of failure, while carefully ignoring successfully managed retirement investment programs that are actually running right now here at home in the U.S. He wouldn't want anyone to form an opinion by looking at those!

Let's take as our model the Federal Thrift Savings Plan which right now manages private retirement investments for government workers with a 0.1% expense rate. If it's good enough for government employees one would think it's good enough for the rest of us.

OK, from our conservative 3.5% annual return we subtract 0.1% to get 3.4%. That's hardly the huge return from stocks that some people talk about. But still, using the bond rate as the discount rate, over 40 years: $1 invested in a government bond gives us $1 ... $1 invested at 3.4% over the bond rate gives us about $3.80 ... and $1 contributed to Social Security gives us less than $1.

Wouldn't it be nice if our low-wage young male worker could get back more than only 98 cents -- or maybe as little as 70 cents (after the funding gap is closed) -- for every $1 he pays in taxes to Social Security? Like a private account that delivers a mere $3.80 would provide?

This is the comparison that Krugman and the other defenders of the Social Security status quo never make -- and the question they never ask, as "progressive" and concerned for such low-wage workers as they pose at being.

Paul Krugman believes private investments are too risky and expensive to place in retirement accounts. It follows he must be saddened that he can't invest more of his own retirement savings in Social Security, rather than be forced to manage them himself at such great risk and expense as he does now. But I will make an offer to cheer him up!

Here's the offer: He sends me his retirement money. I will guarantee him -- pledging my home and all other assets as collateral -- that I will pay him on those funds the exact same less-than-the-federal-bond-rate return that Social Security assures today's young workers. And I will charge him no fee for this service!

Instead, I will merely invest the funds he sends me in a broadly diversified portfolio of market investments, and hope to eke out some return for myself from the difference between market returns and the sub-federal-bond-rate I must pay him. I will assume all the risk! On his behalf.

If he's serious in his beliefs, he'll e-mail me.

In fact, if Krugman is right, my offer will be universally attractive and I'll make it to anyone. I'll market it to everyone! I've found a market failure to exploit! I'll set up a mutual fund to safely duplicate Social Security-level returns, and wait for the cash to come pouring in from all those who only wish they had a safe, risk-free, efficient way to lose money on the way to retirement...

Nah, I don't think so, that's being disingenuous, I'm just kidding.

The other aspects of Krugman's non-kidding disingenuousness on this subject in his column of three weeks ago, in all its glory, were covered previously.

* fn. I'm teasing, of course. Going to the American Prospect's masthead right at the top we see the names "Robert Kuttner" and "Robert Reich" -- two fellows Krugman famously used to savage on a regular basis for what he deemed their analytical incompetence and factual ignorance regarding things economic. You'll have no trouble googling up some enjoyable exchanges.

Now Krugman recycles completely credulously -- breathlessly, even -- material they provide as the proof of his own beliefs. It seems the overwhelming liberal political imperative to assure that even low-wage workers will be made poorer by Social Security creates strange bedfellows.