John Hood: Stop kicking Social Security can

RALEIGH — Although the politicians in Washington almost never talk about the issue anymore, our entitlement programs are on a collision course with fiscal reality.

Let’s zero in on Social Security. Its “trust fund,” which consists entirely of debt the federal government has issued to itself, will be exhausted in a decade or so. As a practical matter, this isn’t as big a deal as it sounds. Social Security is already running a cash deficit — more payments going out than payroll taxes coming in. To cover the difference, the federal government draws from its trust fund of federal bonds.

In other words, it draws from its general revenues to pay Social Security benefits. If the trust fund didn’t exist, Washington would . . . draw from its general revenues to pay Social Security benefits.

Nevertheless, as a legal matter, the coming exhaustion of the trust fund will trigger a crisis. Let’s say it happens in 2032. At that point, federal law will require that Social Security benefits be lowered across-the-board until they can be financed entirely by payroll taxes. That would be an immediate 21% cut.

And it’s not going to happen. That much is obvious. No president or member of Congress could withstand the blowback. They’ll either raise the payroll tax rate, eliminate the current cap on income to which the tax is applied, or make more-complex changes in benefits to shield lower-income recipient from anything like a 21% hit.

Or perhaps they’ll change the law to run larger deficits. But that will only postpone the fiscal reckoning a bit. Federal debt is not a perpetual-motion machine, much as some political activists might want it to be. Servicing it will require some combination of tax increases and reductions in spending — either enacted directly or applied indirectly through a renewed and rampant inflation.

Now, if you’ve been thinking about Social Security’s unfunded liabilities at all, you probably have a preferred solution. I happen to think the best solution at this point is to apply a means test to benefits, making the program more explicitly a redistributive safety net rather than pretending it can or should be the centerpiece of household savings for retirement.

But whatever we do — general tax hikes, subjecting upper-income households to more payroll tax, raising the retirement age, changing the inflation calculator to reduce benefit growth over time, etc. — we need to get on with it. The longer we wait, the costlier the intervention will be.

That’s the main conclusion of a new National Bureau of Economic Research paper. The authors — John Shoven and John Watson, both of Stanford University, and Sita Nataraj Slavov of the American Enterprise Institute — didn’t just look at estimates of the effects of various reforms on Social Security’s structural deficit. They also examined the costs of political dithering about reform.

“Knowing in advance which reform will be implemented allows for better planning and therefore has value,” they wrote. “The government could resolve that indecision by deciding today what steps it will take to close Social Security’s shortfall.”

If Americans are informed ahead of time that, say, their benefits are going to be reduced, or the age they can retire with full benefits is going to be raised, they have time to react. If they aren’t informed ahead of time, the study found, that could cost them many thousands of dollars.

I know why Democrats and Republicans alike don’t want to talk about this. Each of the practical options for bring Social Security costs in line with revenues will make a large bloc of voters angry. Twenty years ago, when there was some momentum in Washington to construct a bipartisan solution, politicians could have made virtually all of the benefit changes prospective, affecting only future retirees, and given people lots of time to accumulate savings in personal accounts. It is, however, too late now to engineer a soft landing.

It was a missed opportunity, a failure of political leadership and imagination. And, alas, a costly one.