J. Bradford DeLong
U.C. Berkeley and NBER
Democratic Policy Committee
May 13, 2005, 10 AM
Any discussion of Social Security reform needs to begin with one too-rarely-asked question: Why is the American political system focusing its attention on Social Security? Is this really the aspect of American fiscal policy that should be absorbing our attention right now?
The answer is that we shouldn't. We shouldn't be focusing on Social Security right now. America has three problems with the fiscal policy pursued by the Bush administration:
- The current 5% of GDP on-budget deficit, the likelihood of major legislative changes (like extensions of expiring tax cuts) that will blow further holes in the budget, and the risks of economic crisis and recession and slowed long-run growth created by this Bush-league fiscal policy.
- The generational explosion of federal health-care costs we expect to see. From one perspective, this is not so much a problem but an opportunity: we expect our doctors, nurses, and druggists to do even more wonderful things for us in a generation. We would like for all Americans--not just those with thick wallets--to benefit from the advances in health care that we confidently anticipate. But this will be expensive: we need to figure out how much publicly-funded health care for the poor, the disabled, and the old we as a society wish to buy, and what taxes are going to fund these public health-care programs.
- The likelihood--not the certainty--that the Social Security system as currently structured will be in deficit by mid-century.
The first of these--the current Bush deficits--is the most urgent. The second of these--the health care funding "opportunity"--is the largest. The third of these--Social Security--is both the least urgent and the smallest. So why are we spending our time on it? There's no good reason. As Berkshire-Hathaway Chairman Warren Buffett, no bleeding-heart liberal he, said last week:
...a [Social Security] deficit of $100 billion a year, something, 20 years out, seems to terrify the administration. But the $400 plus billion dollars deficit currently does nothing but draw yawns....
The answer, I think, is that the Bush administration doesn't want to deal with our current deficit because it made the mess and hopes that if it pretends it doesn't exist nobody will really notice. It's very much the three-year-old approach to fiscal policy: "What broken jelly jar on the kitchen floor?" The answer, I think, is that the Bush administration doesn't want to deal with health-care financing issues because they have no idea what to do: the only person even close to the inner circle with serious views is Mark McClellan who could get nobody else to pay any attention, so he's parked himself at CMS in the hope that by writing rules he can improve the efficiency of the public health-care systems.
Focusing on Social Security thus performs the important political function for the administration of allowing it to pose as seriously concerned with responsible governance while at the same time ignoring the mess on the kitchen floor. And in this case, it's not a three-year-old who has dropped a jelly jar. It is the moral equivalent of a three-year-old who has climbed up on the counter and thrown everything breakable in the room down onto the floor.
Moreover, how serious is Social Security's long-run financing problem? The 2005 Social Security Trustees' Report reduced its estimate of Social Security's 75-year deficit from 0.7% to 0.6% of GDP. Using the same accounting methods, the 75-year deficit of the U.S. government as a whole under current law plus extension of the Bush tax cuts and an AMT fix is 4.8% of GDP. The Social Security Trustees assume, for their intermediate run, that productivity growth in the economy will average 1.6% per year. They justify this assumption by writing that:
[since] productivity growth can vary substantially within economic cycles.... [We] consider historical average growth rates for complete economic cycles. The annual increase in total productivity averaged 1.6 percent over the last four complete economic cycles (measured from peak to peak), covering the 34-year period from 1966 to 2000. The annual increase in total productivity averaged 2.2, 1.2, 1.3, and 1.6 percent over the business cycles 1966-73, 1973-78, 1978-89, 1989-2000, respectively...
And the average of 2.2, 1.2, 1.3, and 1.6 is 1.6, hence their 1.6% per year assumed growth rate--the same growth rate they assumed in 2004, and 2003, and 2002. In 2003 the Trustees learned that the previous year's rate of productivity growth had been 3.2%, but this good news did not lead them to increase the assumed productivity growth rate. Ditto for 2004, when they learned the previous year's rate had been 3.5%. Ditto for 2005, when they learned the previous year's rate had been 3.3%. Past groups of Social Security Trustees have not been hesitant to lower estimates of productivity growth when bad news comes in; this group of Trustees is extremely eager to throw away all of the good news about the underlying productivity trend of the American economy we have had since 2000. I cannot think of a defensible economic rationale for such a forecast construction procedure. Nobody else I have talked to--not Democrats, not Republicans, not Independents--has been able to think of a defensible rationale either.
Taking account of the three places where I think the Trustees' forecasts are clearly awry--productivity growth, life expectancy, and also immigration--if I were running the Social Security Trustees, I would be forecasting a 75-year deficit of 0.2%-0.3% of GDP. Even if we weren't facing other much more urgent, dire, and serious fiscal problems, that deficit would be too small for me to think that it would be a good idea to focus for a year the limited discussion and decision-making power of the U.S. government on this issue. The limited attention span of the government needs to be focused on dire and urgent problems, not limited and far off ones.
With that note--that it is really a bad idea for us to be here talking about Social Security rather than talking about bringing the on-budget deficit down to zero from its current value of nearly $600 billion a year--let me, finally, begin:
In my view, a Social Security reform plan needs to clear five hurdles before it is worth considering:
- The private accounts it offers people must be a good deal for beneficiaries.
- The plan must raise national savings.
- The plan must preserve the valuable defined-benefit nature of the current program.
- The plan should restore long-run solvency, and put in place mechanisms for automatic adjustment should the system fall further out of balance.
- We must have confidence that the plan will be competently implemented.
You have already heard from Robert Shiller on how private accounts as proposed by the Bush administration are not a good deal for beneficiaries. They do promise some higher returns, yes, but those higher returns are not worth the risk they make beneficiaries bear--the extra purchasing power gained in those states of the world when stocks do well does not match the losses beneficiaries see in states of the world when stocks do not do so well, because the lower your income is the more you miss the extra dollars that just aren't there. I don't have more than quibbles with Shiller's analysis, and none of my quibbles have any effect on his conclusions.
It is depressing to have to say this, because I am actually, in at least one of my hearts-of-hearts, the heart-of-hears of an Eisenhower Republican, a believer in private accounts. I agree with Marty Feldstein that the higher value of the equity premium in the United States over the past half-century tells us that the stock market has really not done very well at mobilizing the risk-bearing capacity of the American economy, and that any steps we can take to broaden and deepen stock ownership promise not just boosts the income and wealth of new stock owners but also significant improvements in the ability of America's business to raise capital for large risky and entrepreneurial ventures. I agree with former Bush II Treasury Deputy Assistant Secretary Kent Smetters that it is a scandal and an outrage that the poorest half of Americans have no easy, automatic, straightforward, and low-fee way of investing in stocks.
But the Bush plan's private accounts are not private accounts that anybody should endorse. The 3%-plus-inflation clawback rate is just too high given likely future asset reurns--as Robert Shiller has just demonstrated. When I conduct informal conversations with those who have been called senior administration officials and ask "Why? Why not a clawback rate that is the Treasury's actual borrowing rate, or even subsidize the private accounts a bit?" I tend to get one or more of four answers:
- We know. We're trying to fix it. Our voice isn't very loud in here...
- There are powerful people who think that if we set the clawback rate at the Treasury's borrowing rate, it will look like our plan is nothing more than letting people borrow from Social Security to speculate in stocks.
- We can't find any money for a subsidy to make private accounts a good deal: we don't have the resources to reach actuarial balance as it is.
- There's a "fuzzy math" problem in here. People think that "ownership" is a good thing by itself, and don't focus on the numbers.
Let me skip over the "increase national savings issue" even though it is very important. Let me just say that boosting national savings in the medium run is a dealbreaking essential for Social Security reform, that the most even those under tight message discipline like Greg Mankiw dare say is that the Bush plan will neither raise nor lower national savings, and that Alan Greenspan, among others, is very worried that Mankiw and company are wrong and fear that it might lower national savings and raise interest rates:
We don't know the extent to which the financial markets... trading in long-term bonds, are discounting the... contingent liability.... If indeed the financial markets do not distinguish... one would say, 'Well, if you wanted to go to a private [accounts] system, you could go... without any response in interest rates because... you're just merely switching assets....' But we don't know that. And if we were to go forward in a large way and we were wrong, it would be creating more difficulties than I could imagine.
The boosting-national-savings concern is the big reason that it is important that properly-designed and properly-implemented private accounts will be an add-on to Social Security as it currently exists, not a carve-out from it.
Now let me turn to the defined-benefit nature of Social Security: that it takes a share of your wages, and returns a certain, inflation-adjusted, stable monthly benefit check that replaces a reasonable (althought sliding-scale) proportion of your pre-retirement income. Such a defined-benefit retirement program is a very valuable thing: people like the idea a lot. Such a defined-benefit program is very hard to get elsewhere than Social Security these days: Businesses today are very hesitant and unwilling to assume the risk and liabilities involved in setting one up. And businesses are not large and stable enough to be able to bear the long-run risk: ask the steelworkers, or the employees of United Airlines, where there defined-benefit corporate pensions are. Or ask the workers of GM.
The government, however, is big enough to bear the risk of a defined-benefit pension program. And I think that where there is something that people regard as extremely valuable that only the government can provide, the government should provide it. In another of my hearts-of-hearts, you see, I am a social democrat. An it is in this context that I find myself very skeptical of progressive price indexing, for over the long-run it eliminates Social Security as a defined-benefit pension program, replacing it with a Social Security system that gives everybody the same value monthly check equal to about 22% of the average wage. It (a) cuts average benefits relative to current law by about 40%, and (b) changes Social Security from a defined-benefit program to which you contribute a share of your income to a simple flat benefit check, along the lines of what George McGovern proposed in the 1972 presidential campaign.
Now this is a progressive change in America's tax and transfer system. But Peter Orszag talks fast enough to be able to list 20 better ways to make America's tax and transfer system more progressive in less than thirty seconds. And this way does come at a major cost: eliminating people's defined-benefit pension component, a component that cannot be easily replaced in the economy we are moving into.
Now there are problems in analyzing what the administration's progressive price indexation proposal is. It came without numbers, which is why we are all using the numbers that Jason Furman has cranked out based on Pozen's original plan and on Jason's Talmudic reading of what actual modifications of Pozen's proposal would be needed if the words from speeches and press conferences were to actually make sense. And we have disturbing claims on Meet the Press that Pozen's plan is "'...really not necessarily the president's plan,' [Andrew] Card noted. 'It's directionally consistent with the president's plan'."
Robert Pozen proposes to fill the post-2050 projected funding gap in Social Security by reducing benefits and raising taxes. He would fill 10 percent of the gap by cutting Social Security disability payments. He would fill 60 percent by shifting to "progressive price indexing." He would fill 30 percent of the hole by tapping into income tax revenue. Is Bush's adoption of Pozen's ideas a backhand way of saying that he too wishes to fill the anticipated funding hole with income taxes? Does anybody know?
It is important to register the magnitude of the benefit cuts relative to current law that Pozen is proposing. For those retiring in 2075, Pozen would keep the replacement rate at 49 percent for the working poor—those making half the average income. But the replacement rate for those making more would be cut: At the average income, the replacement rate would go from 36 percent to 26 percent; at one-and-a-half times average, from 30 percent to 17 percent; at the Social Security maximum, from 24 percent to 12 percent. Medicare premiums are already deducted from your Social Security check. Deduct the claw-back for the private-accounts diversion as well, and by late in this century the odds are that—-at least for the upper middle class—-the standard Social Security check would be zero. Social Security would no longer be a universal program: It would be a program in which the half of America that is richer and more powerful and more likely to vote sees large chunks of its money going in and nothing coming out.
It is possible that this is the point, at least for some factions within the White House: perhaps creating a large and powerful class of Americans who get much, much less out of Social Security than they put in and for whom Social Security as a whole is demonstrably a very bad deal is the objective. It was early Social Security guru Wilbur Cohen who said that "in the United States, a program that deals only with the poor will end up being a poor program. ... " Loading a large chunk of the burden of fixing Social Security onto America's upper middle class may be the first step in the creation of a mid-21st-century political majority for the phasing-out of the program as a whole. But it is also possible that the White House doesn't have such a long-run Machiavellian plan in mind--that it is hoping to attract some more support for its proposals by making them more progressive, and is just trying to trick its right-wing supporters by pretending that this is a deep, long-run, Machiavellian plan to eventually eliminate Social Security entirely. The only thing that is clear is that it is very unlikely that the White House is proposing progressive price indexing because it believes the government needs to put its thumb on the scales to make the U.S. distribution of income and wealth more equal: that's not their game. (It is, however, Robert Pozen's game: that's why he advocates this proposal.)
Let me skip over point four, solvency, by pointing out that the Bush plan--excuse me, there is no Bush plan, there are only "ideas" that are "not necessarily" the Bush plan but are only "directionally consistent" with it--doesn't reach actuarial solvency unless Bush's embrace of Pozen includes his plans to rest part of Social Security on income taxes. If that is in fact what Bush's embrace of Pozen means, that's big news.
And let me end with point five: competence in implementation. A Social Security reform plan could clear hurdles one through four, and still fall flat on its face if incompetently implemented: the devil is in the details. And looking at the farm bill, the steel tariff, the return of deficits bigger than ever, the use of intelligence by the NSC, the absence of a real plan for post-invasion Iraq, the Medicare drug bill that ex-HHS Secretary Thompson now really wishes had been structured differently, et cetera, et cetera, it is hard to believe that any reform to be implemented by this administration will be implemented competently. It has the anti-Midas touch: whatever it grasps turns to mud.
Let me give a small example of this, from George W. Bush's statement on April 28:
In a reformed Social System, voluntary personal retirement accounts would offer workers a number of investment options.... I know some Americans have reservations about investing in the stock market, so I propose that one investment option consist entirely of treasury bonds, which are backed by the full faith and credit of the United States government.... Options like this will make voluntary personal retirement accounts a safer investment that will allow an American to build a nest egg that he or she can pass on to whomever he or she chooses.
The "build a nest egg" part... The "invest in Treasury bonds" part... On April 28 the Federal Reserve was reporting that the twenty-year inflation-protected Treasury bond was yielding 1.87% plus inflation per year. The money you divert into your private account under the Bush plan is clawed back--charged against your normal Social Security benefit--at a rate equivalent to 3% per year plus the rate of inflation. Were interest rates to stay where they are now over the next fourty years, if you were 25 now, made an average of $80,000 a year over your career, and diverted all 4% of your wages possible into your private account and invested them in twenty-year inflation-protected Treasury bonds... then your Social Security benefits--normal plus the annuitized check from your private account balance--would be $514 a month less than if you had said "No thanks" to private accounts and kept your money in the standard program.
"Building a nest egg"?! Given current interest rates, following George W. Bush's advice and investing Social Security private-accounts money in inflation-protected Treasury securities is a really bad idea. Yet they let him go out there and say that.
Did nobody inside the White House bother to run the numbers? Did nobody care? This breaks my heart--for in yet another of my hearts-of-hearts, I am a technocrat who believes in administrative competence, and think that the President of the United States should not be sent out to make speeches that only an underbriefed fool would write because of the nonsensical things that they say.
Until the center of policy making and implementation in this administration is moved outside the White House to someplace else where people seriously concerned with the substantive design and implementation of policy, nothing the White House proposes--nothing, no matter how good it sounds at first--can be expected to turn out to be anything other than a large pile of mud.