You wouldn't like retirees when they're angry.

Photographer: Scott Olson/Getty Images

Social Security or Savings? Um. Try a Little of Both.

Megan McArdle is a Bloomberg View columnist. She wrote for the Daily Beast, Newsweek, the Atlantic and the Economist and founded the blog Asymmetrical Information. She is the author of "“The Up Side of Down: Why Failing Well Is the Key to Success.”
Read More.
a | A

The other day, I tackled the vexing question of why so few Americans seem able to save enough. Today I’ll tackle the vexing corollary to this question: the assertion that if individuals can’t manage to save, the government must step in. Corporate pensions are dying, the argument goes, and private savings is laughably inadequate. Obviously, what we need is a much more generous Social Security system.

Dylan Matthews is the latest to make this argument: “Instead of trying to repair our broken privatized retirement system, why not give the public sector a shot? Why not expand Social Security so that it functions not just as a backstop but as the primary source of retirement income for ordinary Americans?”

There are so many reasons "why not." I'll get to those. To start with, in fairness, Matthews is highlighting real problems with the current approach, as he argues for junking 401(k)s and corporate pensions.

After all, corporate pensions never covered everyone and turned out to be plagued by risks from market downturns and corporate failures (which tended to happen at the same time, leaving retirees without adequate savings to fund their retirement), and anyway, they’re dying. Matthews is far from the first to suggest that 401(k)s are a laughably inadequate substitute. They primarily benefit the more affluent Americans, and people who have them often (maybe usually) don’t save as much as they should, and 401(k) holders who retire when the market is down end up in deep trouble. The Social Security system, on the other hand, does not depend on the vicissitudes of the market; it relies on the government’s taxing power.

So why not eliminate pensions and 401(k)s and use Social Security to primarily fund everyone’s retirement? Look at how much it would cost, and where the money would come from.

We currently fund Social Security with a  roughly 12 percent tax on all payrolls below a low-six-figure income. This tax has some nice features, from the government’s point of view. It ties the payout from the system to the amount that is paid in -- yes, yes, you in the back, I know that Generation Xers who retire will, mostly, have paid more into the system than they’ll ever get out. I’m not saying that there is a one-to-one relationship between how much you “contribute” and how much you get back. But there is a relationship. This is thought in many quarters to enhance the program’s political support, by making it seem more like an insurance program than like welfare. And judging by the number of people who respond to any attempt to alter the benefit formulas with an indignant claim that they earned those benefits, that strategy is working.

This sort of payroll tax is also nice because it’s very steady. Other government revenue may be whipsawed by the vicissitudes of the economy, but payroll taxes on middle-class workers deliver about the same amount of revenue, year in and year out. Since that’s also how the benefits have to be paid, this is a nice bit of political economy.

That said, the outflow began to exceed the inflow several years back. The difference is modest, right now, but in 15 years, or thereabouts, when Social Security’s stock of special-purpose government bonds is exhausted, the deficit will be about 20 percent. Making up that deficit is already going to be expensive.  Taxing all wage income, instead of just the first six figures, would get us about 75 percent of the way there, but would represent a 12 percent surtax on an awful lot of income. So much so that high earners in high-tax coastal states could be subjected to tax rates that have distorting economic effects on their output, as they decide it’s just not worth it to become a high-earning professional, or to put in those extra hours at the business.

It would also represent just about all of our remaining power to tax high incomes, devoted to a single program which is by no means our biggest fiscal problem. (That honor goes to the health-care entitlements: Medicare, Medicaid and Obamacare.) And of course, it would break the link between contributions and payouts, making Social Security simply another welfare program, with all the political problems that might stem from such a designation.

Alternatively, we could add another 5 percentage points to the existing payroll tax, without raising the cap on income subject to it. That would close the gap even 75 years out, but make a lot of workers very upset.

Okay, but what about the general fund? Our federal tax take remains well below the average in OECD countries; there’s surely room to raise the money from a variety of sources: corporate taxes, capital gains taxes and so forth. This is absolutely true.

However, let’s return to our argument in favor of abandoning corporate pensions and individual savings, which is that … they are too exposed to the market. Guess what’s also highly exposed to the market? That’s right, corporate income taxes and especially capital gains income, which goes off a cliff when the market falls. That’s why tax revenue tends to fall in recessions by much more than the decline in GDP; the payroll taxes are still chugging along, but corporate taxes, capital gains and the performance-driven incomes of high earners plummet.

In other words, if we peg our retirement income to something other than the payroll tax, we aren’t getting rid of the market risk; we’re just sticking it on the government, rather than the corporation or the individual.

Ah, you will say, but unlike the individuals and the corporations, the government can easily borrow during a recession. Just so -- for now. But transferring the risk does not end it.

As I have explained before, the absolute hardest thing for an economic system to do is eliminate risk. Usually instead it is transformed into some form that is harder to see, thereby buying the illusion of safety. That’s what banks do: They bundle lots of little deposits together and loan out the money. Then one day too many borrowers have missed their payments, and the bank fails, and depositors lose their money, and it turns out that the risk never went away.

Of course, at FDIC-insured banks, much of the risk is transferred to the government. But all that does is transform idiosyncratic individual risks that frequently produce bad individual outcomes into massive systemic risks that, much more rarely, result in total catastrophe for everyone in the country. When a bank fails, it takes some people's savings -- but those people may have saved in more than one bank, or they may have relatives or friends who saved in different banks and can help them out. When a pension fund fails, we can say much the same thing. When the government entitlement goes belly up, every retiree gets hit at the same time, and the more they depend on that pension, the harder they get hit.

The more entitlements we load on government, the more likely such failures become, simply because we’ve reduced the fiscal flexibility that a government has to deal with a crisis. Too much of the economy is funneled into mandatory payments, and the constant pressures of funding those promises tempt the government to take on high levels of debt to avoid resorting to unpopular taxes.

And just as corporate pensions and individual savings have their own particular additional risks -- the risk of undersaving, the risk of poor market timing --government retirement programs also have their own special problems. Specifically, the moral hazard that tends to undercut the solvency of the promises they’ve made.

In order to have the money later to pay for the benefits we’ve promised decades hence, we need an economy where the workforce produces sufficient output to support both workers and non-workers -- children, retirees, stay-at-home spouses, the disabled. How do you get that? It’s arithmetic: How big is your workforce, and how high is the productivity of each worker? Unfortunately, generous social security systems probably undercut both.

A shift away from the risks of 401(k)s and corporate pensions would be a shift toward the risks of an overburdened and overgenerous national retirement system.

That’s why virtually all pension experts talk about the “three-legged stool”: The government provides some, to keep people from becoming destitute; employers help a bit; and individual savings make up the rest. The flaws of each type of savings don’t exactly cancel each other out, but having all three legs somewhat limits the extent of the disaster if one fails.

For all the problems with our system, we are the envy of pension experts elsewhere for the way we have structured our system to distribute the burden and avoid the distortions and disasters that other countries have already begun to see.

The question, then isn’t "why not shift most savings to Social Security," but rather "why shift most savings to Social Security"? It's an attractive option to a certain class of coastal professional. Those people, more than most other Americans, are caught in a zero-sum bidding war over educational amenities for their children. As I described this week, they are trying to capture a slice of something that is fixed in supply -- land in the best school districts, slots at pricey elite colleges. And the prize often goes to the parents who most recklessly shortchange their own financial wellbeing to pour the money into the bottomless maw of these school systems.

For those people, a government alternative looks really attractive, because their financial difficulties are a collective-action problem. If everyone were forced to divert 15 percent of their income to retirement, everyone would remain in the same place on the competitive ladder and would get to retire in decent comfort.

But this describes a pretty small portion of the electorate, and one whose problems are not really going to be solved by any foreseeable government program, because the U.S. government is simply not going to ensure that every six-figure earner in the country can retire in the style to which they have become accustomed.

Whatever system we get will focus on the problems of most people. And most people have the more prosaic problems of deciding between tangible consumption now, or comfortable retirement later. Whatever method we choose is going to force them to consume less stuff, whether through higher taxes or more savings. Given that, we should probably choose the method that beefs up the relatively underweight portion of the three-legged stool -- individual savings -- rather than the portion that’s already overrepresented in the "savings" (taxes) and incomes of the elderly.

It's not a sure bet. Fundamentally, developed nations today are trying to do something that has never been done before: give most of the population a long retirement, rivaling the length of their working life. I’m not sure that’s feasible with any system. But if we’re going to try, we should go into the project with as many funding sources as we can, not try to put all our nest eggs in Social Security's basket.

  1. If that's Matthews's goal, hey, great news: Social Security already is the primary source of income for most retired Americans.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

To contact the author of this story:
Megan McArdle at mmcardle3@bloomberg.net

To contact the editor responsible for this story:
Philip Gray at philipgray@bloomberg.net