A Preliminary Analysis of Sen. Graham's Social Security Proposal Revised December 14 to correct the estimate of the 10-year budgetary cost of the proposal ($1.4 trillion); originally published November 18, 2003.
The Social Security reform proposal introduced by Sen. Lindsey Graham of South Carolina this morning is serious and worthy of bipartisan consideration. It will not be easy to find the funds to pay its large transition costs (about 1 percent of GDP for 15 years), but the long-run budgetary savings would be well worth it.
Outline:
A Serious Proposal
Basics of Sen. Graham’s Bill
Budgetary Impact
Perspective
The deteriorating long-term budget outlook requires that legislators reconsider entitlement reform. Without a renewed effort to balance the budget and reform the entitlement programs, the four largest entitlements -- Social Security, Medicare, Medicaid, and Interest (on the public debt) -- would begin to swamp the federal budget after the baby boom generation retires (see Figure 1).
Figure 1.
A Serious Proposal: To his credit, President Bush touched the “third rail” of Social Security policy during the presidential campaign in 2000 and not only survived, but proved that many Americans are open-minded about a serious debate on reform.
The President’s hand-picked Social Security Commission also got off to a good start, with an interim report that clearly explained the issues facing Social Security.
However, the Commission ended up presenting not one, but three reform proposals. Moreover, the analysis of those proposals in the Commission’s final report was very complex and difficult to understand.
The multiple reform proposals and the exceeding complex report signaled that the Bush Administration was not really prepared to follow through on Social Security reform. The Commission’s final report almost seemed to say “It’s been fun talking about Social Security reform, but we’re not ready to truly suggest a reform proposal and clearly explain its pros and cons to the public and Congress.”
To make matters worse, the Bush Administration has never included the cost of Social Security reform in its budgets. That is a further signal that the Administration is not serious about doing the public and legislative groundwork that would be required to actually enact a Social Security reform bill.
In essence, the public debate on Social Security ended when the President’s Commission disbanded. Some reform bills have been introduced, but they have not sparked a thoughtful national discussion beyond the usual political rhetoric: Republicans applauding personal accounts and Democrats denouncing privatization.
Now, Senator Lindsey Graham of South Carolina has introduced a thoughtful Social Security reform bill that contains many appealing features. This bill should provoke a new debate on substantive pros and cons of reform -- it gives the analytic community and the congressional support agencies a good starting place for a full and detailed discussion.
This report does not pretend to be a full analysis. The Social Security actuaries’ cost analysis of the Graham proposal was released only yesterday, and the Congressional Budget Office and Social Security Administration have not yet published any distributional analysis (estimating the impact of the proposal on workers earning different amounts or working for different periods).
Instead, this report attempts to describe the main features of the Graham bill, and assesses its overall budgetary impact through 2030 in the context of baseline projections of the impact of the retiring baby boom generation. That is not a full analysis, and it may need to be revised or expanded as additional details become clear and official analyses and estimates are conducted.
Basics of Senator Graham’s Proposal: Like most reform proposals, the Graham bill would not change Social Security for current retirees and workers currently aged 55 or over.
The bill would give younger workers a choice: pay more to stay in the current Social Security defined benefit system, or join a new system that combined personal accounts (whose accumulated balances would gradually increase over time) with gradual reductions in defined benefits.
Pay to Stay. To stay in the current defined benefit system, workers would have to pay an additional payroll tax of 2 percent. Taxable payrolls account for roughly 50 percent of the overall economy, or GDP. Therefore, a payroll tax increase of 2 percent is equivalent to roughly 1 percent of GDP, roughly half the amount by which Social Security is estimated to be underfunded in the long-term (see Figure 2).
Figure 2.
The extra payroll tax might have to raised again in the distant future if the 2 percentage point increase proved insufficient to pay benefits for workers electing to stay in the old system.
Reform Option. The reformed option would include personal accounts, lowered defined benefits, and increased minimum benefits for low-income workers. The defined benefit would be adjusted lower by indexing initial benefits to price increases rather than (higher) wage increases. Once initial benefits were calculated there would be no further reductions over time -- i.e. no reductions in COLAs or cost of living adjustments.
Personal Accounts. Workers choosing the reformed system would receive the equivalent of 4 percentage points of payroll taxes in their personal accounts, up to a maximum federal contribution of $1,300 a year.
Workers could make personal contributions of up to $5,000 a year to the accounts. The contributions would be after-tax, but the income or “inside build up” of personal contributions would not be taxed each year or at withdrawal.
Low-income workers would be eligible for extra matching funds for personal contributions they make. The government would match the first $500 of personal contributions of workers with earnings of less than roughly $30,000 a year. Each year, the government match would be $100 for the first dollar, and 50 percent for each additional dollar up to a maximum additional government contribution of $500. Low-income workers could designate their Earned Income Tax Credit (EITC) refund for this purpose.
Computing Initial Defined Benefits: Price Indexing vs. Wage Indexing. Initial defined benefits promised under current Social Security law are computed based on annual increases in wages. This so-called “wage indexing” means that initial Social Security benefits increase faster than price increases or inflation. Therefore, future workers are virtually guaranteed higher retirement incomes from Social Security -- on an inflation adjusted basis -- than current workers.
However, once the initial benefit is computed, subsequent benefits are indexed higher based on the increase in prices (the Consumer Price Index, or CPI). After Social Security beneficiaries retire, they are guaranteed benefit increases sufficient to keep their inflation-adjusted income constant.
The Graham bill would continue to use price indexing to adjust post-retirement benefits. But it would switch from wage indexing to price indexing to compute initial benefits as well.
Switching from wage indexing to price indexing for initial benefits would essentially freeze Social Security benefits at current levels, adjusted for inflation. But that implies lower initial benefits than under current law, because inflation will grow much slower than wages.
Thus the main benefit reduction in the Graham bill would take place very slowly over time, as workers in the reformed system gradually earned benefits through a formula linked to a slower rate of growth. However, workers could be assured of receiving defined benefits that were the same as those received by current workers, accounting for inflation.
2.7 Percent “Offset Interest Rate.” Initial defined benefits under the Graham bill would be further reduced by an amount equivalent to the value of the basic (4 percent) government contributions to beneficiaries' personal accounts, had those accounts earned a “real” or inflation-adjusted interest rate of 2.7 percent.
The 2.7 percent figure was chosen because it represents an average inflation-adjusted rate of return on long-term government bonds. This is roughly the inflation-adjusted return of the safe "G Fund" within the Thrift Savings Plan, the 401k-style retirement plan used by federal employees. The Graham bill would use the Thrift Savings Plan as a model for the investment options available via Social Security. The G Fund is a special government bond fund that earns the interest rate of a long-term bond without any price fluctuation.
Therefore, most workers who invested their personal accounts in this type of safe fund would break even on this adjustment. Their personal accounts would essentially grow by a sufficient amount over time to compensate for the offset interest rate reduction.
Workers who invested in funds earning more than 2.7 percent or who made additional personal contributions (or received government matching funds) would have enough money in their accounts to compensate for the offset interest reduction, with funds left over to supplement their incomes.
Workers would be allowed to invest in stock funds and other types of bond funds whose prices could fluctuate considerably and lose money. For workers whose investments returned less than 2.7 percent, therefore, the offset interest reduction would mean an additional benefit reduction compared with current law.
Benefit Improvements. The Graham bill essentially guarantees that a worker with a 35-year career would receive a minimum defined benefit equal to 120 percent of poverty, regardless of any other benefit reductions and not including any amounts in a worker’s personal account. This would increase the benefits for many workers whose wages were low during their pre-retirement years. It would also create a floor under which benefits could fall. For example, if workers' personal accounts had a very poor yield (perhaps due extremely bad timing of investment allocations between stock and bond funds) their minimum retirement benefit would still keep them above poverty.
The calculation of widow’s benefits would be improved to 75 percent of the couple’s previous benefits (up from 50-67 percent under current law). This would also have an anti-poverty effect.
Budgetary Impact: Social Security costs are currently just under 4.5 percent of GDP. Over the next 25 years, the program's spending will accelerate to just over 6.5 percent of GDP, and will continue rising gradually from that point.
The Graham Social Security proposal would raise Social Security costs by about 1 percent of GDP for the first 15 years. The bill's reform option would begin in 2006, but it would be after 2020 before there was a noticeable reduction in the cost of the new system. However, spending would begin to decelerate in earnest after that, and Social Security costs would gradually decline to about 4 percent of GDP by the end of the 75-year estimating period (see Figure 3).
Figure 3.
(These cost figures are based on a preliminary analysis of the budget impact of the Graham proposal performed by the Social Security Administration's Office of the Actuary. However, the figures presented here have been converted to percentages of GDP by Centrists.Org, which is responsible for the resulting calculations.)
Another way to think about the Social Security's budget situation is in terms of the program's cash-flow deficit or surplus. Based on current trends in payroll taxes and other revenues dedicated to Social Security, the program's surpluses will switch to deficits in 2018. The deficits will grow to over 2 percent of GDP by 2075.
Under Sen. Graham's proposal, the Social Security would begin running a small deficit in 2006, when personal accounts were introduced. The deficit would grow at a slower rate than under current law, but it would still hit a peak of about 1.7 percent of GDP between 2025 and 2030.
After 2030, however, the Social Security deficit would begin to shrink, gradually becoming a surplus again before the end of the projection period (see Figure 4).
Figure 4.
These estimates assume that all eligible workers will enroll in the reform and personal account option, rather than choosing to pay extra to stay in the old benefit system.
Perspective: The Graham Social Security bill would be expensive during the transition period. The proposal would cost about $1.4 trillion between 2004 and 2013, the current budgeting period for the federal government. However, these are transition costs required to fund the personal accounts, which, in turn, would allow seniors to maintain a reasonable level of retirement income even while Social Security spending was significantly reduced in future decades.
The problem is, politicians are not planning for these costs. President Bush's budget and the Congressional Budget Resolution do not make allowances for the transition costs of Social Security reform. Putting a reserve fund or allocation in the President's budget would help persuade politicians on both sides of the aisle that the Social Security debate was becoming more serious.
Many Democrats are just not interested in Social Security reform. But some are, often because of the opportunity to help low-wage workers to build financial wealth. Others are motivated by a passionate concern for the well-being of future generations of taxpayers.
To gain a bipartisan consensus on reform, some additional funding mechanisms will probably have to be added to the Graham proposal. One idea is to raise the cap on payroll income subject to Social Security taxes from the current level of $87,000 to about $140,000. That would put the overall percentage of payroll income subject to the tax at 90 percent of total wages earned, up from the current percentage of about 85 percent. (The percentage was 90 percent in the early 1980s, but it has drifted down to the current level over time.)
Figure 5 shows the Social Security cash-flow situation under the Graham proposal, assuming a gradual increase in the tax cap to 90 percent of total wages earned, effective from 2006 until Social Security surpluses were re-achieved in mid-century. This would save roughly $325 billion within the 10-year budget window between 2004 and 2013 (assuming no extra benefits accrued from the extra taxes).
Figure 5.
There may be other suitable financing options to help pay for the transition costs of reforms in Sen. Graham's proposal. The limit on payroll taxes shifted to the accounts ($1,300) could be reduced -- that would lower the account balances of middle- and higher-income workers, but it would also save in transition costs.
Financing options need not be limited to Social Security-type measures, and perhaps there are other ways to rein in the costs or enhance the benefits in Sen. Graham's plan to help spur bipartisan interest, or at least curiosity.
In the end, a serious proposal like that put forward by Sen. Graham deserves a thorough and continuing analysis and is worthy of bipartisan consideration. The proposal should help re-focus the dormant and ritualistic debate on Social Security reform.
Links:
Social Security Administration, Office of the Actuary Memorandum Estimated OASDI Financial Effects of S.1878, The Social Security Solvency and Modernization Act of 2003 as introduced by
Senator Lindsey Graham (November 18, 2003)
CentristPolicyNetwork.Org Legislative Resource Page for Sen Graham's Social Security Proposal (S. 1878), including summary of the bill issued by Sen. Graham's office.
CentristPolicyNetwork.Org Put Social Security Reform in the President's Budget (November 17, 2003)
Centrists.Org Raising the Cap on Payroll Taxes Doesn't Solve the Social Security Problem (November 17, 2003)
Centrists.Org Issue Summary: Wealth Building (Basics)
Centrists.Org Issue Summary: Social Security
