Simon Johnson's Blog, page 52

Ponzi Schemes for Beginners

By James Kwak

On the theory that the best defense is a good offense, Rick Perry has been insisting to anyone who will listen that Social Security is a Ponzi scheme. Probably hundreds of people have already explained why it isn't, but I think it's important to be clear about why Rick Perry thinks it is—or, rather, why his political advisers think he can get away with it.

A Ponzi scheme, classically, is one where you promise high returns to investors but you have no way of actually generating those returns; instead, you plan to pay off old investors by getting new money from new investors. Social Security is obviously not a Ponzi scheme for at least two basic reasons. First, there's no fraud involved: all of Social Security's finances are right out in the open for anyone who cares to look, in the annual report of the trustees of the Social Security trust funds. Second, a Ponzi scheme by construction cannot go on forever; no matter how long you can keep it going, at some point you will run out of potential new investors and the whole thing will collapse. I'm sure there are other obvious differences, but that's enough for now.

So why do people ever think it's a Ponzi scheme? It's the combination of two factors, each of which is relatively innocuous on its own.

First, Social Security is, from a cash flow perspective, a pay-as-you-go system. That is, payroll taxes being paid by current workers are immediately going out the door to pay benefits for current retirees. This is different from a pre-funded pension system where the pension plan already has enough money, today, to pay for all promised future benefits. Employer-sponsored pension funds in principle (though not always in practice) work this way: at any moment, the fund is supposed to have enough money to pay off future benefits. A commenter on an earlier post of mine said that if Social Security were a corporate pension fund, I would be attacking it as a fraud for this reason.

Not so fast. There's nothing wrong in principle with a pay-as-you-go system, as long as the future revenue stream is secure. With any pension plan, the fundamental question is whether the plan will be able to pay future promised benefits. We want private pensions to be pre-funded because we wouldn't trust a company that said, "We know we don't have enough money to pay the benefits we promised, but don't worry, we'll be really profitable starting twenty years from now, so we'll be fine." It's not a question of whether corporate executives are honest; it's that the future is uncertain. As we know, companies can go bankrupt almost overnight (remember Enron? WorldCom?), so the only way to ensure that they will pay future benefits is to require pre-funding. Even then, pre-funding is not a magic bullet because you have to make some assumption about future investment returns. Under today's rules, companies assume some rate of return on their pension fund investments (actually, I think they assume some discount rate for their future liabilities, which works out to roughly the same thing)—which means that even with pre-funding, there's no assurance the money will be there to pay benefits.

Social Security is different because it knows that, unlike future corporate profits, the trust funds' revenues will be there in future years. That's the great thing about being the federal government: you know people will pay their taxes. Yes, there is uncertainty about economic growth, but Social Security's future income stream is a good deal more predictable than a corporate pension fund's future investment returns.*

But, you may be saying, and this is the second factor, the trust funds will go bankrupt in about twenty-five years!** That is probably true, but it is not because Social Security is a pay-as-you-go system. A pay-as-you-go system can easily go on forever, as long as you have a constant rate of population growth (or even a constant rate of population decline): Generation A's benefits are paid by Generation B, which is 10 percent larger than Generation A; Generation B's benefits are paid by Generation C, which is 10 percent larger than Generation B; and so on. With a few more simple assumptions, this can go on forever.

The Social Security trust funds will run out of money because the current structure of taxes and benefits requires a certain minimum dependency ratio (the ratio of workers to retirees) and the actual dependency ratio will fall below that required minimum as the Baby Boom generation retires. It's the combination of those two facts—that Social Security is a pay-as-you-go system and that the trust funds will run out of money—that makes Rick Perry thinks it's a Ponzi scheme. Ponzi schemes are also pay-as-you-go systems that will run out of money, but that doesn't make Social Security a Ponzi scheme, just like all giraffes are mammals, but not all mammals are giraffes.

As all informed observers realize, you could close the seventy-five-year Social Security budget gap simply by raising the payroll tax rate by two percentage points (or by other means that have a similar financial impact, such as eliminating the cap on taxable income). This in itself should make clear that it isn't a Ponzi scheme.

Now, it is true that the benefit-tax structure was already changed in 1983, so it's a reasonable question whether the system has to be "fixed" every thirty years, which would raise the question of sustainability. But there's no good reason to think the dependency ratio will keep getting worse and worse, because the Baby Boom was a one-time historical event. Right now the dependency ratio and the funding gap are expected to level off around 2035 (as the last Baby Boomers retire) and remain roughly flat for half a century. (See the 2011 annual report, pp. 10–11.) So any policy change that brings Social Security in balance in 2035 will keep it more or less in balance as far as we can see. Yes, it's possible that the birth rate could drop again, but it could also go up; we just don't know. (And if the birth rate drops, what that means is that we need more immigration by working-age people, but that's another topic.)

Wait, Rick Perry might say: Doesn't the fact that we all know Social Security's future revenues are not enough to pay its scheduled benefits make it a fraud all by itself? Not in a legal sense, since all the information is out there for everyone to see. But it does raise a legitimate issue: By 1983, the Baby Boom had happened, so everything that is going to happen with the dependency ratio was easily predictable—yet Congress and the Reagan administration consciously underfunded the system.  (As of 1983, Social Security was in 75-year actuarial balance, but that was because a large surplus in the first thirty-seven years balanced a large deficit in the last thirty-eight years. See the 1983 Annual Report Summary, Chart F.) Isn't that a problem?

But saying that Congress underfunded Social Security is not a valid criticism of Social Security as a program: it's a valid criticism of Congress. As a logical matter, you can criticize our political leaders, past and present, for not fixing Social Security's long-term funding problem, but you can't use that fact to criticize the basic structure of the system, where people pay taxes while working and receive benefits while in retirement. There's nothing wrong with that, try as Rick Perry might to confuse the issue.

* Private accounts would solve this problem, but not in a useful sense. Because you would only be entitled to the money in your private account, Social Security would have no obligations, and hence could never be underfunded. But you would be subject to inflation risk and investment risk; today, Social Security absorbs both inflation risk and investment risk, so all you are subject to is political risk (the risk that benefits when it comes time for you to collect them).

** "Bankrupt" isn't quite right: what will happen is that the trust funds' accumulated balance will run down, so the only money they will have to pay benefits will be the current year's payroll tax revenues—which will not be enough to pay scheduled benefits. It's not clear that these "scheduled" benefits should even count as "promised," since both the law and the financial status of Social Security are public information, but that's another topic for another time.

The Importance of Time Off

By James Kwak

I used to be a real productivity nerd. I wrote an earlier post detailing some examples of my compulsion to be as efficient as possible. I worked at a company where efficiency was one of our highest implicit values.

I still care a lot about productivity—my own, that is. For example, after reading Anthony Bourdain's first book, I became much more attentive to the sequence of movements I make around the kitchen: when you open the refrigerator, take out everything you need; when you walk to the far end of the kitchen to get utensils, get everything you need; and so on. I used to make fancy dinners that could take an hour and a half or two hours to cook. Now that I have a child, I make simpler, multi-course meals in 30–40 minutes.

But I have a more nuanced understanding of productivity now, which is why I liked Derek Thompson's article on the importance of vacation—and taking breaks—so much.* Part of it is that since I left the business world, most of my work now is creative—not creative in the sense of creating original works of art, but in the more modest sense that it involves thinking about stuff and writing about that stuff. When you're a manager at a fast-growth, under-staffed company, it's not hard to spend huge blocks of time just responding to email, reviewing documents, and providing input on various issues. That takes thought, but it's somewhat mechanical. When I'm writing anything I care about, though, I can't force myself to crank out another paragraph at will. And if that paragraph isn't coming, I go kill some zombies, or "clean up outer space" (as my daughter puts it), or weed the lawn.

Thompson cites a number of studies showing that taking short breaks can improve either quality of work or output or both. Even modest amounts of lolcats or cute animals can increase productivity. And anyone who knows anything about software development knows that if you demand more output from people in the same amount of time, you're going to get lower quality—which means more work in the long run, when you factor in bug fixes and the increased effort required on customer sites. The idea that more time spent "at work" translates linearly into greater value for the employer is just silly, for reasons I go into more in my earlier post.

There's a bigger issue here, too. If working forty hours per week is better than working forty-eight, why is working forty better than working thirty-two? One of the more obvious solutions to the unemployment problem is job-sharing or, more radically, a four-day work week. Various European companies have implemented shorter work weeks (and paid people less), with no productivity losses (I believe—I'm basing this on what people I trust have told me). (There's the problem of fixed benefit costs, but there must be solutions to that.) I realize that this does nothing for economic growth and GDP. But it would modestly reduce the problem of unemployment-induced poverty, reduce welfare and disability claims on state and federal governments, and allow people to maintain their job skills, which is important for the economy in the long run. And, who knows, maybe it would actually make people more productive. Of course, in today's America this sounds like a radical, even "socialist" idea. But that's more a comment on America than on anything else.

* Derek Thompson is my editor at The Atlantic, so I have a small incentive to say nice things about him.

You Get What You Pay For

By Simon Johnson

Standard & Poor's downgrade of United States government debt last month has been much debated, but not enough attention has been devoted to the fact, reported last week by Bloomberg News, that it continues to rate securities based on subprime mortgages as AAA.

In short, S.&P. is suggesting that these mortgages are more creditworthy than the United States government – a striking proposition. Leave aside for a moment that S.&P. made a big mistake in its analysis of the federal budget (as explained by James Kwak recently in this blog). Just focus on all the things that can go wrong with subprime mortgages – housing prices can fall, people can lose jobs, the economy may fall into recession and so on.

Now weigh those risks against the possibility that the United States government will default. As we learned this summer, that is not a zero-probability event – but it would take either an act of Congress, in the sense of passing legislation, or a determination by members of Congress that they could not act. S.&P. finds this more likely to happen than some subprime mortgages going bad.

Now S.&P. might be right, of course. Or its assessment might be influenced by the fact that it is paid by the issuer of those mortgage-backed securities – which presumably wants a higher rating. The rating agency's employees may want to do an accurate assessment; management can reasonably expect to make higher profits if its ratings please the paying customers.

Perhaps we should just disregard what S.&P. and its competitors say. But this is not so easy, because many investors are guided by rules – either self-imposed or created by regulators – that tie investment decisions, and thus these investors' holdings, to ratings. Ratings changes undeniably can move markets.

How can we take seriously a rating agency that is compensated by the issuers of securities? This system has long outlived its usefulness and should be discontinued.

In a similar vein, let me ask why we should take seriously economic analysis offered up by a financial-sector lobbying group on behalf of its members — if, for example, it says that regulation of its members will slow economic growth? Surely, we should check the numbers in the analysis carefully and be skeptical of the policy recommendations.

A timely example comes from the Institute of International Finance, which calls itself "the Global Association of Financial Institutions" and whose board members are all from big banks. (Indeed, the institute is more than a mere lobbying group; in the recent Greek debt negotiations, it was in charge of coordinating the terms proposed by private-sector banks for their involvement in the debt restructuring.)

So what do we make of its policy recommendations? In a report released this week, "The Cumulative Impact on the Global Economy of Changes in the Financial Regulatory Framework," for example, the institute asserts that additional capital requirements for its members could result in "3.2 percent lower output by 2015 in these economies than would otherwise be the case" (see paragraph 5 of its news release).

In recent conversations with policy makers from the Group of Seven nations, I was told that the institute's previous, interim report on this same topic was largely without value (some said completely without value).

I hope the official policy community reacts the same way in this instance, because the institute refuses to acknowledge the vast cost imposed on society by the combination of big banks, high leverage and low capital that it endorsed through 2008 and that it defends, without only minor modifications, today. (James Kwak and I wrote directly about these issues in 13 Bankers – and we're now hard at work on the sequel.)

The institute's report is nothing more than lobbying masquerading as economic analysis. And just as S.&P. is paid for its ratings by the issuers, the institute is paid to represent the views of big banks. We would be wise to suspect that in both cases, the paying customer would prefer a particular outcome – irrespective of what the evidence says.

An edited version of this post appeared this morning on the NYT.com's Economix blog.  It is used here with permission.  If you would like to reproduce the entire post, please contact the New York Times.