Krugman (PoKro?) has gone back to Social Security after promising to deal with health care for a month or so. This is too bad, because Krugman was making good points on health care, even if some of his statistics were misleadingly framed. His latest on Social Security is deliberately deceptive, however. Here we go:

Sure enough, a close look at President Bush's proposal for "progressive price indexing" of Social Security puts the lie to claims that it's a plan to increase benefits for the poor and cut them for the wealthy. In fact, it's a plan to slash middle-class benefits; the wealthy would barely feel a thing.

Under current law, low-wage workers receive Social Security benefits equal to 49 percent of their wages before retirement. Under the Bush scheme, that wouldn't change. So benefits for the poor would be maintained, not increased.

No, no, no. This is a common deception used for political ends, whereby one party can brag that they increased spending while the other party claims they decreased it. Nobody's benefits will be going down under this Bush plan. None. Nobody's. No spending will be cut. None.

Here's what's actually been suggested by the president. Benefits will continue to be increased every year, but the rate at which they are increased will be less for wealthier earners than the rate at which they currently increase. Right now, SS benefits are indexed to wages, which usually rise faster than prices in an expanding economy. It hasn't always been this way. During the Carter era of high inflation, prices briefly grew much faster than wages. At that time, the switch was made from price indexing to wage indexing because the government apparently thought that trend would continue for a while. But it makes more sense to tag benefits to prices, so that what your pension buys now it can keep buying in the future.

So what Bush is essentially prosposing is to keep the extra-generous wage indexing for those who depend most on Social Security, while adopting the less-generous index for those who depend on it least, and everybody in between would sort of split the difference. If you're a member of the middle class, planning to retire after 2015, and you've ever actually estimated your Social Security benefits and used them to plan your retirement out to the dollar, you have a right to be angry. But I'm not sure more than five people in this entire country have done all those things.

What Krugman would prefer is to raise the cap on wages subject to Social Security taxes, currently capped at $90,000. This would be a transfer from the wealthy to the middle class, instead of from the middle class to the poor. I'm morally comfortable with either option, but Krugman's argument that reducing the benefits to the middle class will make Social Security vulnerable in the future seems pretty far out. He may be right, but giving the very wealthy an incentive to do away with the program by making them pay for virtually all of it sounds like a better recipe for electoral evisceration than giving the middle class slightly less free money.

Any time a NYT columnist calls himself a "control group," I hope a red flag goes up in your mind. Tierney compares his non-privatized Social Security pension with the one his friend Pablo from Chile is going to get. Chile privatized their pension plan in 1981. Tierney claims to control for income disparity between himself and Pablo, and concludes that the privatized system gives back roughly a third of American social security. The implication is that privatized plans make for three times the retirement.

But there's one problem with Tierney's "experiment." The test subject and control group are in two very different countries. Here is a chart of the S&P 500 since January 1981 and the Santiago IPSA index since 1989:

Chile has been growing at an astonishing pace in the past twenty years. It's easier for a developing nation to grow at a fast pace than a developed one -- I've heard it compared to cyclists drafting off the lead biker. As Chile's economy has boomed and its financial markets have matured, its stock market has exploded. Even giving the S&P an eight-year head start, the IPSA's growth on a percentage basis has been triple the S&P.

This comparison would be more accurate if I had data on the Chilean markets going back to '81, when the shift to private pensions happened. But I can't find it. Since financial markets have an upward bias, the gap between Chile and the USA growth is probably even bigger. But maybe not. If anybody knows where I can get better Chilean stock numbers, let me know.

Bush proposed to index benefits to prices for high-end earners, fixing their benefits at a constant inflation-adjusted income. Lower income workers would continue to have their benefits indexed to wages, as they are now. Wages traditionally increase more quickly than prices. People between the extremes would see their increases indexed to some sort of average between the two rates.

The press jumped on this proposal by calling it a "benefit cut," which is kind of true, but not in the way most people understand a "cut." Decades from now, it means middle- and upper-class earners will get less than the law currently anticipates giving them, but at no point will benefits be reduced from what they were the previous year. The increases will simply be smaller.

Once people got that straight, the idea that's been bouncing around the echo chamber has been the idea that this is all a plot to undermine the popularity of Social Security in the long run. Right now Social Security is a bad deal for the middle class, but if it were an even worse deal, it would somehow cease to be popular. (e.g. This Slate essay, titled "How to Make the Middle Class Hate Social Security") But even if this were Bush's motives, wouldn't it make more sense to infuriate both poor and middle-class people (by increasing the payroll tax rate or raising the retirement age) or the wealthy and powerful (by raising the cap on the taxes)? Or is the suggestion that we keep the program as it is and run a deficit? Considering deficits are the basis for financing every other government program, especially the military, I don't see why nobody has suggested doing this. It's not like we aren't looking the other way on Medicare.

When I read these criticisms of the Bush-Pozen plan, I don't understand why nobody is worrying about the poor. It's taken as given that wages rise faster than prices, so benefits for the poor will rise faster than benefits for the wealthy. But the reason we have wage-indexed benefits now is because in the late '70s prices were rising faster than wages -- and the government switched to a wage index to slow the increase in benefits. Ironically.

So in times of prosperity, or normal or even tepid growth, the poor will see their benefits rise faster than inflation. Great! But in times of stagflation, prices rise quite a bit faster than wages. The wealthy will see their benefits stay at the same inflation-adjusted level, so that part of their retirement income will keep pace with a surging cost of living. But the pension of the poor will lag behind the cost of living, possibly by quite a bit. And because the other half of stagflation is a stagnant economy, things will have to get pretty bad before they can get better. This is what happened in the late '70s and early '80s. Advances in economic theory have made stagflation less likely -- but not impossible.

So this bifurcated indexing would give the middle-class a smaller, but more stable, Social Security benefit. The poor would get a benefit that increases more rapidly in good times, but actually contracts when they need it most. The poor would still come out ahead on average. But those who depend on Social Security and expect it to keep up with inflation could experience real hardship if the economy were to really tank the way it did during the '70s oil shocks. Instead of worrying about what Joe Suburb is going to think about a price index, we might consider what could happen if, or when, the economy tanks again to the poorest seniors.

Tomorrow I'll tie this all together (and justify this title) with some remarks on the private retirement plans the government already gives you, and why privatized Social Security might be a better deal for retirees for precisely the opposite reasons being offered by the Republican party.

The president would like to put about four percent of our pre-tax income into private accounts. But you don't need him to put a gun to your head to do it. It's easy to put four percent of your pre-tax income into a private savings account. Depending on where you work, what your income is, and what you think is the best bet for you, you can open an IRA, a 401(k), or a Roth IRA.

If you're like most Americans, you don't know what these actually mean. I'm in the same boat -- I've been trying to find a good summary all day, and the link above is the best I can find. But all function on the same principle: your income grows tax-free within the account. You never have to pay taxes on dividends and capital gains once your money is in the account. You can reinvest that extra money and not pay taxes on the dividends or capital gains earned with the taxes you didn't pay. Over decades, that small difference in the effective rate of compounding means a lot of extra money. And there are other tax advantages as well. Up to a certain limit, what you put into a traditional IRA is tax deductible. (It's line 25 on the 1040). You don't pay income tax on it until you withdraw it as a retiree, presumably in a lower tax bracket than while working. The 401(k) seems to be a corporate variation on the theme, often with matching funds thrown in by one's employer for free. (E.g., I get an extra $.35 in my 401(k) for every dollar I put in. If I worked there long enough, I could get a full dollar.) The Roth IRA works on the opposite principle -- you deposit after-tax money, but it grows tax-free and you don't owe anything on it when you retire and start withdrawing. All of these plans have severe tax penalties if you withdraw your money before you reach a minimum age.

Figuring out which of these plans works best for you would require time, a thick mass-market paperback, a couple pencils, and patience. There are rules and regulations I'm sure I'm failing to address in this quick summary. I'm not a financial planner. But the basic principle of the IRA is comparable to the basic principle of the private Social Security account. I think. I am flying blind on this comparison, so if there are any accountants out there, let me know where the gap lies.

So here is the true situation: we have been sold a debate on private Social Security accounts that pretends to be between the New Deal, protosocialist pay as you go system where big government runs the show, versus an "ownership society" where autonomous individuals have some choice in the investment of their federal pension funds. But the debate is really more comparable to the one over socialized medicine. The pay as you go system can't be scrapped, so private accounts would be a new program, not a replacement. The Democratic side would preserve the current patchwork system of tax-deferred investment, a patchwork where some people get great benefits from work (401k), some people buy their own plans (IRA), and some people just go without -- similar to our system of health insurance. Under the Bush plan, everybody would be forced to participate in such accounts, and would pay for it with a mandatory payroll chunk -- just like socialized medicine.

The best reason to set up private Social Security accounts, then, is not because it will make us an "ownership society" where even the poor own a piece of the market. That potential is already out there. The best argument is actually a big government one -- we're too stupid to save if nobody makes us do it.

How stupid are we? A growing body of economic literature has been studying companies with 401(k) plans with matching funds. The main determinant of particpation in these plans is not age, income, the generosity of the matching program, or education. It's the default action. The percentage of employees who particpate tends to be about forty percent if employees are asked if they would like to enroll. It's about seventy percent if they're asked if they would like to not enroll. The difference is whether they have to check a box.

You might think that gap is mostly young people who don't think about retirement much. But once you hit a certain age, you can begin to withdraw from your 401(k) without penalty, and even these older employees still participate at only about an eighty percent level if they have to opt in when they're hired -- even though they could withdraw their savings and matching funds any day without penalty. They're giving away FREE MONEY. We don't need to be turned into a society of owners. We might need big government to save us from our own stupidity.