Perry and Romney debate a program in need of reform
Rick Perry and Mitt Romney have begun a lively debate about Social Security, but it’s not the one we need. Their debate mostly concerns whether it was a good idea to establish the program. Perry has made it abundantly clear that he thinks the answer is no. The program’s creation was an affront to the Constitution. Romney objects to this line of argument. He implies that he would have voted for it in the 1930s.
Neither of them is saying much about the actual question the country faces today: Given that we have a Social Security program and are not going to abolish it or devolve it to the states (as Perry has daydreamt), what should we do about it? Romney says he is open to various reforms, though he has not made any specific proposals. Perry has been even vaguer.
Romney is almost certainly right that forthright opposition to Social Security in principle is a losing proposition in American politics. It remains one of the most popular federal programs, and polls consistently find that voters want to prevent benefits from being cut. When pollsters have divided respondents by party and ideology, they find that this consensus includes strong majorities of self-described Republicans, conservatives, and tea-party sympathizers. (Nobody, to my knowledge, has even done polling on whether the federal government should withdraw from the field altogether.)
Polls also consistently find that the public understands that Social Security is in trouble. But Americans want the program to be saved, not eliminated or replaced. If he becomes president, or even the Republican presidential nominee, Perry will have to have a politically realistic agenda for Social Security. But for any set of reforms, resistance is likely to be greater if the public thinks they are motivated by a desire to undermine the program.
Governor Perry is also right, however, to suggest that the program is deeply flawed — and its flaws, as he has hinted, go beyond its much-discussed funding shortfall. Rates of elderly destitution are much lower since the development of Social Security, but this gain came at a cost. Martin Feldstein, the distinguished Harvard economist, has long maintained that the program markedly reduces the nation’s capital stock — and thus economic growth and wages — by diminishing the incentive to save. It diminishes the incentive to work, too, by encouraging retirement in the early and mid-60s, even though lifespans have increased and hard physical labor has declined.
An increasing number of studies demonstrate that the program shrinks the average American family, too, by socializing the returns to child-rearing. Parents contribute to the program’s future twice, first by paying payroll taxes and second by raising children; but the program recognizes only the first set of contributions.
The program also has some baleful political effects. It makes the electorate much more dependent on the federal government, and thus on the political class, than it otherwise would be. Social Security did not just cause the federal government to outgrow the constraints of the Constitution; it also did a lot to create a citizenry unmindful of those constraints.
The program’s structure was designed to spread false beliefs about it that would make it politically unassailable, and that now make it hard to reform. The purpose of creating a payroll tax supposedly dedicated to the program was to convince Americans that they were paying into a special fund and would later withdraw what they contributed. In reality, the program bears only a passing resemblance to this picture. But many liberals, from the dawn of the program to this day, have not believed that Americans would support the amount of redistribution the program entails if this redistribution were transparent.
Finally if most pressingly there is the question of the program’s solvency. Social Security is currently paying out more in benefits than it takes in. While it may run a surplus again if we ever see a strong economic recovery, within a few years the increasing number of retirees — ten thousand Boomers will quit working every day for the next 20 years — will push the program into deeper and deeper deficits starting just a few years from now.
For a while, the program will pay the excess benefits out of its “trust fund”: the pile of IOUs it accumulated when the program ran a surplus and the rest of the federal government borrowed from it. To repay those IOUs the federal government will have to either raise taxes, cut other spending, or go even deeper into debt. When the IOUs run out, the law imposes a sharp cut in Social Security benefits. The current projections have it that the program will cut benefits by 23 percent starting in 2037. Under current law, then, the program will make federal deficits much worse for more than two decades and then sharply cut benefits. Neither part of this future is desirable, or politically sustainable.
There are five reasons the program’s spending has been growing faster than its revenues and will continue to do so. First, we have more retirees. Second, their retirements are lasting longer. Third, birth rates have fallen, so there are fewer work years to support each year of retirement. Fourth, there is a cap on the taxes the program imposes on any individual. But as income inequality has risen, a higher proportion of wages has been above the cap and thus gone untaxed.
Fifth, the program’s benefit levels rise with wage levels throughout the economy. In his 2010 book Social Security: The Unfinished Work, Charles Blahous, who advised Pres. George W. Bush on the issue and now serves as a trustee for Social Security and Medicare, noted that a medium-wage worker retiring in 2010 at the normal retirement age would receive an annual benefit of $17,700. A medium-wage worker retiring at the normal retirement age in 2035 is expected, on the other hand, to receive $24,000 annually (plus whatever inflation occurs between now and then).
Of course, that future worker will also be putting more into the system because he will be paying taxes on a higher wage level. So it might seem as though that benefit level is only fair: It replaces the same percentage of a worker’s wages when he retires tomorrow as it replaces today. But this is the wrong way to think about the problem. If the retiree population grows faster than the work force, the only way to maintain that replacement ratio is to increase taxes constantly — and while raising taxes on higher-income workers can generate some of the needed income, much of any plausible tax increase would have to fall on this medium-wage worker himself. Neither a free lunch nor a free retirement can be assumed.
The unfunded liability of Social Security — the excess of projected benefits over projected revenues — is estimated at $16 trillion. Yet there are people familiar with the projections who nonetheless regard Social Security as a second-tier problem for the federal budget. The liberal version of the argument goes like this: Health-care inflation rather than population aging is responsible for the bulk of our future spending explosion, and therefore our main goal should be to impose cost controls on health care rather than change benefit formulas to account for new demographic realities. The conservative version simply notes that Medicare has an even larger unfunded liability: $30 trillion. In the long run, it is primarily responsible for increased spending, and Social Security is an afterthought. Both of these arguments have helped reduce the pressure on presidents and presidential candidates to put Social Security reforms on the table.
The liberal argument is simply mistaken. By all means we should look for intelligent ways to moderate health-care inflation. But the Congressional Budget Office reports that 64 percent of the increased cost in Social Security, Medicare, and Medicaid between now and 2035 is a result of population aging. The conservative argument is stronger, but it may take too long a view. Between now and the early 2030s, Social Security is expected to remain larger than Medicare, and over the next 15 years its growth will put more strain on the budget than Medicare’s.
Fix Social Security now, and you still have time to solve the larger but longer-run problem of Medicare spending. Fix Medicare now, and the country still moves into very rocky fiscal territory in the next two decades.
Fixing either program is of course politically difficult, or else it would have happened already. But fixing Social Security is conceptually easy: Raise the eligibility age and moderate the growth of benefits. Let the medium-wage worker who retires in 2035 receive the same benefit level, adjusted for inflation, that the medium-wage worker who retired in 2010 receives. We could change the benefit formula so that benefits for the lowest earners continue to rise even when adjusted for inflation.
Jed Graham, a writer for Investor’s Business Daily, proposes an additional tweak in his book A Well-Tailored Safety Net: Make sure that benefits for the oldest of the old are higher than benefits for the newly retired. This goal too is easy to achieve in theory. Reform initial benefit levels as discussed, so that they rise with inflation. But then give every beneficiary an increase every year of retirement that more than keeps up with inflation.
Paring back the growth of benefits would make the program solvent while reducing its effects on the length of working life, the savings rate, the incentives of voters, and the size of families. (On this last point, however, there is no good alternative to instituting a large child tax credit that compensates parents for their contribution to the program.) Raising taxes would, on the other hand, make the program’s negative side effects, such as its discouragement of work, worse.
For the last 15 years or so, most conservatives have also sought to introduce personal savings accounts into Social Security: allowing workers to retain possession of some of their payroll taxes so that they could invest for their own retirement. This proposal had two great virtues. It attempted to make all workers into small-scale capital owners, so that the electorate would become less entitlement-oriented.
And it was a mechanism to pre-fund retirement. Politicians have long claimed to want to reserve excess Social Security revenues to pay for future benefits rather than use them to pay for other operations of the federal government. This was the point of Al Gore’s much-mocked “Social Security lockbox.” The accounts would have been several million lockboxes. If the federal government had used the Social Security surpluses of the last two decades to launch personal accounts, the federal government would not have been able to tap these funds and might have had to go on a diet. If that had happened, the accounts would have been a net increase in national savings.
But we aren’t running surpluses anymore. Starting the accounts means moving some federal expenses from the future into the present budget. A worker who participates in the accounts agrees to get a smaller payment from the traditional program in the future in order to put money into the account now. The feds, in return, pay less later but have to allow funds to flow into the accounts now. In a time of massive deficits and debt — with both set to rise — that trade-off is not going to look attractive. And the accounts are now less likely to yield an actual increase in national savings: We are probably not going to cut more spending than we have to cut in order to fund the accounts. Even if a Republican presidential candidate runs on a platform including personal accounts, then, he will probably have to quietly shelve them if he wins.
It is just barely possible to imagine the next president’s moderating the growth of Social Security’s benefits so as to make them affordable. It is a matter of some urgency that this scenario happen. Governor Romney and Governor Perry should do what they can to make it more likely — and refrain from doing anything to make it less likely. If they are not going to advance credible reform proposals, perhaps it would be best if they refrained from talking about the issue altogether.