Executive Summary
Efforts by President George W. Bush to promote major reforms in the Social Security retirement program have not led to policy change, but rather to increased polarization between the two parties. And the longer we wait to address Social Security’s long-term funding problem, the bigger and more painful the changes will need to be.
But despite the current political stalemate, Social Security is a solvable problem. To be financially and politically stable, a future Social Security system will likely need several components, including a version of the current Social Security-defined benefit that becomes somewhat less generous over time; and a new universal, mandatory retirement savings component in low-cost individual accounts that is financed by Social Security payroll tax increases; and a reasonable minimum benefit for those with low lifetime earnings. Diversification of trust fund investments and new legislative mechanisms to facilitate early action as trust fund problems emerge would also be useful components of a Social Security reform package. Lessons from other countries point to common-sense ways to address these issues.
A new approach to Social Security reform also requires a different process of policy formulation and adoption. As a first step, the president and congressional leaders should agree on an overall mandate for a commission named through a bipartisan nominating process designed to generate a group that is likely to focus on practical, consensus-building solutions. Special procedures in each house of Congress would provide expedited consideration of the commission’s reform package and alternatives, while providing incentives for constructive congressional engagement in the reform process.
Policy Brief #166
Over the past three decades, population aging has led to profound changes in the public pension systems of most advanced industrial countries. Many countries have experienced multiple rounds of reform. The United States is an outlier: our main public pension program, Social Security, has remained virtually unchanged for a quarter century.
The last seven years have been particularly unproductive for Social Security reform. In 2001, President George W. Bush established a Social Security reform commission with a mandate to propose partial individual account opt-outs for Social Security. The commission’s final report, issued in the run-up to the 2002 midterm election, identified several alternatives but made no clear recommendation. It was roundly ignored by Congress. After his 2004 re-election, President Bush tried to put Social Security reform back on the agenda and failed to generate any enthusiasm among either politicians or the citizenry.
When Social Security has made it to the U.S. agenda, the outcome has been a hardening of positions and rhetoric rather a willingness to compromise. That is a pity for several reasons. First, with a long term shortfall in contributions relative to expected benefits of about 2 percent of payroll, Social Security is a solvable problem. Second, the longer we wait to address Social Security’s long-term funding problem, the bigger and more painful the changes will need to be. As the Pensions Commission in the United Kingdom recently pointed out, an aging population means that society’s only long-term options are later retirement, lower retiree benefits, devoting more tax resources to retirement income security, or increased savings. Third, delaying a Social Security fix diverts attention from health care financing, which is a much more difficult issue politically and economically.
The public pension reform experiences of other wealthy countries provide the United States with a panoply of potential reform options. The major problem in reforming Social Security is a political one. Finding a long-term solution to Social Security’s funding woes—indeed, having any outcome other than stalemate—requires liberals and conservatives to both put aside positions that are most unacceptable to those on the other side of the ideological divide and to accept some proposals from the other side with which they may strongly disagree.
What sorts of positions need to be put aside? Conservatives must drop efforts to “carve out” some portion of current Social Security payroll taxes to finance individual accounts, which liberals believe will undermine the current Social Security system as the core of U.S. retirement income policy. Liberals must recognize that conservatives will accept neither putting any more payroll tax resources into the current Social Security program nor injecting significant general revenues into the current Social Security program, except perhaps for very narrow, carefully delimited purposes.
As difficult as this is, it is the easy part: liberals and conservatives must also agree on a common overall vision for the future of the retirement income system in the United States including some provisions, backed by their opponents, that they strongly oppose. What might that vision look like? To be financially and politically stable over the long term, a future Social Security system will likely need to have several components, including a version of the current Social Security-defined benefit that becomes somewhat less generous over time; a new universal, mandatory retirement savings component in low-cost individual accounts; and an improved minimum benefit for those with low lifetime earnings. Future arguments should concern the relative size of those components rather than whether they should exist at all.
Two major substantive trade-offs are almost certain to be required in a comprehensive Social Security reform package. Conservatives will have to accept increased payroll taxes. They must further accept that the current Social Security-defined benefit will remain the biggest component of the retirement income system. In exchange, liberals will have to accept that all of those additional payroll taxes will go into a new system of low-cost individual accounts. Liberals will also have to accept cuts over time in Social Security’s guaranteed benefit. No agreement is likely to be reached unless both sides set as an objective that combined Social Security-defined benefit and mandatory savings will in the future replace something close to the same percentage of a retiree’s former earnings as the current Social Security system alone does now, and that it not lead to an increase in senior poverty among the most vulnerable groups of seniors.
In addition to a set of reform proposals that can address both fiscal and income security concerns, adapting Social Security for the new century requires: (1) a reform process that encourages compromise and protects politicians from their own predilection to cast blame rather than solve problems, and (2) a willingness on the part of politicians to compromise. The experience of other wealthy countries with aging populations suggests that a workable reform process can be created if politicians are willing to sacrifice some policy control over policy formulation, but it is unrealistic (and undesirable) to expect that they give up complete control. Once again, pragmatic compromise is needed.
Elements of a Reform Package
If the two sides of the ideological and partisan divide on Social Security are willing to work toward an agreement, rather than just bash each other for fun and political profit, a balanced reform package could be built from some or all of the following components:
Payroll Tax Increases and Individual Accounts: A key component of a reform package is likely to be an increase in the Social Security payroll tax of between 2 and 2.5 percent of earnings (split between employers and employees). All of the new contributions would go into individual accounts that would be mandatory for all workers, but no existing payroll taxes would be diverted to individual accounts. Workers would choose from a modest range of index fund options managed by private sector fund managers. Those managers would be selected in a competitive bidding process, as with the Thrift Savings Plan for federal workers. The Social Security Administration would manage the collection and flow of money from individual accounts into those funds and switches among fund managers in order to further lower administrative costs. At retirement, individuals could choose between annuitizing the funds in their individual accounts to guarantee a steady income stream or drawing the funds down by a set schedule. Lump-sum withdrawals would not be permitted.
Gradual Reductions in Defined Benefits: The initial Social Security-defined benefit would be reduced for future cohorts of retirees over time as the new individual accounts are phased in. Replacement rates will be set so that the combined “old” Social Security benefits and new mandatory savings accounts will roughly equal current benefits, given moderate estimates on rate of return for the mandatory savings component. Given the progressive nature of the current Social Security benefit formula, some changes in the Social Security benefit formula or injection of general revenues to finance benefits of low earners would be required. Workers would also need to receive better information about how working longer can lead to higher benefits, and about the increased risk posed by having their Social Security benefits depend partially upon the performance of individual investment accounts.
Alarm Bells and Fail Safes: Another potentially useful element of a Social Security reform package would be an “alarm bell” mechanism to focus attention and debate on future Social Security trust fund imbalances if they arise. One option would be a special procedure triggered by a finding of the Social Security actuaries in the trustee’s annual report that the Social Security trust fund was likely to be exhausted within thirty years.
When this alarm bell is triggered, the president would be required to make a report to Congress within a specified amount of time proposing specific legislative steps to address that shortfall. Each house of Congress would then be given a window of time to consider the president’s recommendations under special rules limiting debate and prohibiting amendments. This procedural vote would require a special majority of 60 percent of members voting in each chamber. Final approval of the president’s plan would require a normal majority vote in each chamber. The procedural vote hurdle would hopefully encourage the president to submit a plan that could win broad support.
Stronger “fail-safe” mechanisms—for example, automatically putting in place a combination of automatic tax increases and benefit cuts in specified proportions if the alarm bell is triggered and the president’s reform package fails to win approval—could also be employed. Sweden, Canada and Germany all have enacted various forms of fail-safe mechanisms that vary in their balance between benefit reductions and payroll tax increases. Sweden’s automatic balancing mechanism, which operates entirely on the benefit side, fully adjusts for adverse trends in demography and economic growth. But congressional and interest group resistance to adopting such a stringent fail safe would undoubtedly be very high, making it unlikely to win adoption.
Improvement of Minimum Benefits: Despite major progress over the past 40 years, the U.S. has one of the highest rates of relative senior poverty among the advanced industrial countries. Poverty is especially serious among elderly widows. Many privatization proposals could make this very vulnerable group even worse off.
To address this problem, Congress should include in a Social Security reform package a more generous minimum benefit for retirees who have worked (or whose spouse has worked) long careers at low wages in the United States. This benefit should be paid for out of general revenues. The current safety net income program for the elderly, Supplemental Security Income, serves very few people because its benefit levels are low and its assets tests are extraordinarily stringent.
International experience suggests that it is almost impossible to eliminate poverty among the aged without increased reliance on some form of means-tested program. In Canada, close to 40 percent of seniors receive the Guaranteed Income Supplement, which has moderate income tests, no assets tests, and streamlined re-application procedures designed to reduce stigma and increase take-up among those who are eligible. Using this approach, Canada has virtually eliminated senior poverty at a modest cost. Both Sweden and Germany included improved safety net pensions in their recent pension reform packages. To reduce costs, this new minimum benefit should not confer automatic eligibility for Medicaid, which would raise program costs substantially, but SSI-qualified seniors receiving the new benefit could still get Medicaid.
Help for Families with Young Children: Another potential component of a reform package, pioneered in many European countries, is a requirement that government make contributions to Social Security and mandatory savings accounts on behalf of a custodial parent of very young children who is out of the labor force or has only minimal labor force participation during his or her children’s first years of life. These contributions would be paid at (or topped up to) a flat rate, perhaps 60 percent of average earnings, and paid for out of general revenues. This benefit would disproportionately aid women, who are still much more likely than men to spend prolonged periods either outside the paid labor force or in part-time paid work because they bear a disproportionate share of the burden as caregivers to children and elderly parents. Making modest contributions on behalf of caregivers—at least for very young children—would recognize the social value of their work and help even out the balance of incentives in current government policies, which social conservatives have argued currently favor mothers going into the labor force rather than staying home with their children.
Diversification of Trust Fund Investments: Currently, Social Security trust fund surpluses are invested only in U.S. Treasury securities. Canada, New Zealand, Norway and Sweden all invest part of the public pension funds in equity, corporate bonds and other assets through independent entities in order to gain higher returns. These funds are clearly charged with maximizing fund assets for retirees rather than social investment criteria. The U.S. should consider doing the same thing. As in other countries, these funds should have a strict legislative mandate to maximize return for retirees. The U.S. could use multiple funds of limited size with heavy reliance on private fund managers, to prevent any disruption of capital markets. This approach not only would increase returns on Social Security contributions, it would ease the cash flow transition expected to occur in 2017 as Social Security shifts from serving as a source of financing for the federal government to being a drain on government revenues.
A Process and Timetable for Reform
Social Security reform can only succeed if it is built on a process that involves both Democrats and Republicans, and both the president and Congressional leaders. Equally important, both sides must be constrained from pinning the blame on the other for painful elements of a reform package. Given the political sensitivity of benefit cuts, retirement age increases, payroll tax increases, partial privatization and other reform proposals, an ever-present fear of blame-generating attacks from the other party gives politicians an incentive to criticize proposals from the other party without offering their own alternatives (as Democrats did in 2005), as well as incentives to play to your political bases by being intransigent in policy positions. These political risks make the prospects for stalemate very high.
Several countries with pension funding problems much more severe than the U.S. have found ways to insulate their pension reform process from such pressures leading to stalemate, although the steps taken have understandably reflected individual national political environments. In Sweden, a multi-party negotiating group appointed by a conservative coalition government in the early 1990s succeeded in large part because it was made up of politicians who were committed to reform and to working together and who resisted taking intransigent policy stands. Representatives of two small parties who rejected reform dropped off the working group. The parties remaining on the working group were committed to finding a solution that compromised on key issues, such as the inclusion of individual accounts, and defended their proposal against opposition from interest groups. In the U.K., a recent independent commission built public awareness of the need for pension reform in preliminary reports before issuing its final reform recommendations.
These national experiences suggest several lessons for design of a reform process in the U.S.. First, a partisan process, or a process that appears to give an advantage to one party, must be avoided. Second, a direct interest group role should be avoided. Finally, the government should take time to inform the public rather than try to push a reform package through quickly.
In the round of reform that culminated in the 1983 Social Security reform package, the president and congressional leaders shared responsibility for appointing members of a bipartisan commission. Enactment of that package was aided by two factors. First, doing nothing was not an option: unless a package was passed within a few months, there would not be enough money to pay promised benefits. Second, one of the most difficult changes, an increase in the standard retirement age added by Congress to the commission’s original package, did not need to go into effect immediately; it was delayed almost two decades into the future. Neither of the conditions that facilitated passage of the 1983 reform package exist today: there is no immediate funding crisis to force immediate action, and making a major dent in the long-term Social Security deficit will require phasing in changes sooner rather than later.
Given the absence of an immediate action-forcing mechanism and the need to phase in some painful changes in Social Security fairly quickly, what sort of reform process might work in the United States? One possibility is the procedure used to close military bases and ratify presidentially-negotiated trade agreements: Congress is presented with a package on which it must vote either up or down (with no option to add amendments) by a specified deadline. These procedures prevent bold reform packages from suffering “the death of a thousand cuts.” However, it is unlikely that Congress would limit its authority so severely on an issue that is as politically contentious and has such high political stakes as Social Security.
In the absence of up-or-down fast track authority, what sort of process has the best prospects for facilitating agreement leading to bold and relatively rapid policy change? One possibility is to have a four-step process of policy formulation and adoption that is partially insulated from political pressures, but gives politicians incentives to engage constructively in Social Security reform.
As a first stage, the president and congressional leaders of both parties should agree on a very general overall mandate for a reform commission. Unlike the 2001 president’s commission, the mandate for the new commission should not require members to sign on to endorse a particular substantive direction for reform. On the contrary, they should commit themselves to consider all reasonable alternatives. Second, members of the commission should not be named directly by politicians, but rather by a seven-person nominating body with two appointments by the president, one each by the House speaker, the Senate majority leader and the House and Senate minority leaders, and a chair jointly agreed to by the president and House Speaker and minority leader. The nominating body should be instructed to choose a seven-member commission that includes members with a combination of substantive Social Security expertise and general political and economic expertise. The slate of seven commission members, including the chair, would require the approval of six of seven members of the nominating body. This process should generate a body that is likely to focus on practical, consensus-building solutions.
Third, the commission would be charged with writing two reports. The first, to be published within eight months, would lay out the seriousness of the Social Security financing problem for the public and discuss the major options for change with their social and financial implications. After this report, the commissioners would hold a series of public hearings and publish a second report containing a specific legislative proposal within six months. After conversion to specific legislative language, that proposal would proceed to the fourth stage: debate followed by votes under special procedures in both the House Ways and Means and Senate Finance committees and on the floor of the House and Senate. In these committees, an initial vote would pit the commission’s proposal against an alternative posed by the committee’s ranking minority member, followed by the winner of that vote against an alternative proposed by the committee chair. The bill approved by the committee (if it approved one of the alternatives) would then go to the chamber floor, where after debate the ranking minority member and the committee chair would once again have sequential opportunities to pose alternatives, followed by an up-or-down vote on the winner of that competition. If bills passed by the two chambers differed, the normal conference committee process would be used to try to reconcile them. If a reform package was passed by Congress and signed by the president, it would become law. If no package won enactment, the commission could try again with another proposal within six months that would follow the same congressional ratification process, after which point it would go out of existence. Even if this process failed, it would force congressional leaders to come to grips with Social Security, generating both concrete reform proposals and a momentum for reform.
Willingness to Compromise
None of the specific reform proposals outlined above, individually or collectively, is a panacea for Social Security’s problems. Nor is the proposed reform process. But if politicians truly want to solve Social Security’s funding problems rather than score political points and pander to their political base, it should be possible to assemble and enact a balanced reform package that has elements appealing to both sides of the ideological divide, using reform processes that encourage consensus and minimize blame-generating.
Seeking a middle-of-the-road solution for Social Security is not easy. Texas politician Jim Hightower once wrote a book titled There’s Nothing in the Middle of the Road but Yellow Stripes and Dead Armadillos. That has certainly has been true of the recent debates on Social Security. If it remains true, it will be to the detriment of future generations of workers and retirees, and to the shame of today’s politicians. Social Security is a solvable problem, but only if policymakers exhibit both political courage and willingness to compromise.
R. Kent Weaver is a Senior Fellow in Governance Studies at the Brookings Institution and a Professor of Public Policy and Government at Georgetown University. He is the author of the forthcoming book Reforming Social Security: Lessons from Abroad.
