Budgetary Effects of the Diamond-Orszag Social Security Proposal Jeff Lemieux
Revised December 31, 2003 to add a new Table 1 and re-number the previous Table 1 as Table 2. (Originally published December 27, 2003.)
Economists Peter Diamond and Peter Orszag have compiled a thoughtful, complete Social Security reform proposal. The Diamond-Orszag proposal would resolve Social Security's funding shortfall directly, mostly by raising taxes. There are no gimmicks or wishful assumptions. The new revenues include both payroll tax rate increases and expansions of the Social Security payroll tax base. The proposal targets some benefit reductions toward higher-wage workers, but most of those savings would be offset by enhancements to benefits for low-income workers.
The proposal does not attempt to "pre-fund" Social Security, either through personal accounts or a trust fund or "lock-box" approach. Instead, it pays for Social Security's rising costs as they are accrued. Therefore, the proposal has none of the transition costs associated with pre-funding -- it would reduce the federal deficit in every year. By adding state and local workers to Social Security, the proposal could also reduce the employee pension costs of state and local governments in the long run.
In essence, the Diamond-Orszag reform proposal shows the extent to which taxes would have to be raised to eliminate Social Security's long-run deficit. The upside compared with proposals that would pre-fund a portion of future benefits is that taxes would have to rise only relatively slowly. The downside is that, unlike pre-funded proposals, the tax increases would be permanent, not temporary. But the authors are to be commended for their honesty -- this is the first Social Security reform proposal that is fully "paid for."
Outline:
Defining Social Security's Budget Problem
Pre-Funding vs. Pay-As-You-Go
A Fully "Paid-For" Proposal
Costs of the Diamond-Orszag Proposal
Long-Run Costs and Savings of Diamond-Orszag and Other Proposals
Defining Social Security's Budget Problem: The most helpful way to think about Social Security's future budget problem is to compare the program's benefit obligations with its dedicated tax revenues (mostly payroll taxes). Using forecasts for the next 75 years from the Social Security Administration, Figure 1 shows that Social Security costs will rise from about 4.5 percent of the economy toward 7 percent of gross domestic product (GDP) as the baby boom generation retires. Meanwhile, dedicated tax revenues are projected to remain flat at about 5 percent of GDP.
Figure 1.
The difference between these costs and dedicated revenues is Social Security's effective cash-flow surplus or deficit. Because benefit costs are currently lower than revenues, Social Security is running a surplus. However, Social Security's current surplus is expected to disappear and become a deficit after 2010 as the large baby boom generation starts to retire in large numbers. After 2030, the Social Security deficit is projected to be about 2 percent of GDP (see Figure 2).
Figure 2.
By defining Social Security's surplus or deficit as the gap between direct revenues and costs, the measures used in this analysis get to the core budgetary reality -- in order to pay benefits (or create personal accounts) we must have tax revenues.
This definition of Social Security's budget situation does not consider trust fund balances or intra-governmental interest payments or transfers. That is because Social Security is essentially a pay-as-you-go social insurance program, not an advance-funded retirement plan. Workers pay payroll taxes roughly sufficient to cover Social Security's outlays at the time. Social Security's future budget problem is simple: payroll taxes will not be high enough to cover future benefits.
To fully solve Social Security's budget problem, a reform proposal must close the 2 percent of GDP gap between revenues and costs in the long run, without (1) creating overly high or unmanageable transition costs over the next 20 years, or (2) relying on financing gimmicks or "one-sided bets," such as overly optimistic assumptions about investment returns in personal accounts or absolute guarantees that Social Security benefits would never be reduced.
Pre-Funding vs. Pay-As-You-Go. Most Social Security reform proposals seek to partially pre-fund the program, so that Social Security would be less of a pay-as-you-go system and more like an advance-funded system. Pre-funding effectively requires higher government spending now so that future spending obligations can be reduced.
Any proposal that would pre-fund Social Security costs would incur transition costs over the next several decades. Following the transition period, however, pre-funded proposals would achieve "transition benefits," when the up-front investment paid off in lowered spending and tax burdens in later years.
There are two ways to pre-fund Social Security:
1. Trust fund or "lock-box" approaches, where the government would attempt to set aside funds from the rest of the budget that could be used to pay benefits in the future.
2. Personal accounts owned by workers, whose accumulated balances would provide investment income to compensate for future reductions in traditional Social Security benefits.
Lock-box approaches were always dubious, but they have been completely discredited over the last several years. Although there was a flurry of interest in the lock-box idea in the late 1990s, it is now again obvious that Congress does not have the political will to "save" the Social Security surplus in any tangible sense within the federal budget. The government makes spending decisions based on the deficit in its "unified" budget, including Social Security. Financial markets also focus on the unified budget deficit -- the amount by which the government has to borrow from the public -- as the measure of "deficit spending."
Therefore, the Social Security trust fund has proven to be an inadequate storage location for any sort of surplus funds or pre-funding scheme.
By contrast, personal accounts probably would be a more reliable storage location for attempts to pre-fund Social Security costs, because the balances in the accounts would be owned and controlled by workers. Outlays to add funds to the accounts would be gone forever -- they would not be part of legislators' subsequent budget calculations, either directly or indirectly.
The main problem with personal accounts as a storage location would be the temptation for future governments to allow workers to make pre-retirement withdrawals from the accounts -- perhaps for hardship reasons. That would dilute the accounts' ability to supplement workers' retirement income as intended. Account systems used in other countries generally do not allow pre-retirement withdrawals.
The Diamond-Orszag proposal does not attempt to pre-fund Social Security's future costs, either through a lock-box or a personal account approach. Instead, it is a pay-as-you-go proposal, which adds revenues as needed to cover benefit costs as the baby boomers retire.
A Fully "Paid For" Proposal: Social Security reform proposals can be grouped into three categories:
1. "Paid for" proposals, where the transition costs of reform are offset by tangible tax increases or spending cuts. These "pay-fors" can be within the Social Security system, such as payroll tax increases or benefit cuts, or they can be outside the system.
2. Partially funded proposals whose transition costs would be large, but not economically unmanageable. This type of proposal usually contains offsetting benefit reductions or other tangible or "scorable" provisions to at least partially pay for the transition costs of reform.
3. Unfunded "free lunch" proposals whose transition costs would be huge. These proposals make no tangible or score-able efforts to reduce or pay for the transition costs, and generally guarantee that no Social Security beneficiary could possibly lose money on reform (compared with the current benefits promised in the law).
The Diamond-Orszag proposal is fully paid for within the Social Security system. Because the proposal does not pre-fund Social Security's future spending, it has no transition costs. It would increase the Social Security surplus (or reduce the deficit) in every year. Likewise, the overall federal deficit would be reduced in every year.
The main tax increases in the proposal include a very gradual increase in the tax cap (currently $87,000) and a new tax of 3 percent on wage and salary income above the cap. Benefit cuts consist mostly of reductions in benefit "replacement rates" for high-wage workers. However, benefit enhancements of roughly an equal magnitude would apply to workers with low incomes, widows or widowers, and workers qualifying for disability benefits.
After 2023, the proposal requires increases in the regular payroll tax rate (currently 12.4 percent), increases in the new (3 percent) tax rate on wages above the cap, and further reductions in benefit replacement rates to keep Social Security on a sustainable path, with its deficit shrinking closer to zero toward the end of the 75-year estimating period.
Costs of the Diamond-Orszag Proposal: Between 2005 and 2050, the Diamond-Orszag proposal would not reduce Social Security costs, because reductions in benefits for some workers would be offset by the cost of benefit enhancements for others. After 2050, however, the proposal would reduce Social Security costs by about one-half of 1 percent of GDP, from about 7 percent of GDP to 6.5 percent (see Figure 3).
Figure 3.
The Diamond-Orszag Proposal would increase Social Security revenues gradually over time, from about 5 percent of GDP currently to about 6 percent of GDP by 2050, and about 6.3 percent of GDP by 2075 (see Figure 4.)
Figure 4.
Combining the changes in benefits and revenues, the Diamond-Orszag proposal would improve the Social Security cash-flow surplus through 2020, and reduce its deficit thereafter. The deficit would gradually shrink back toward zero after 2030, when most of the baby boomers were retired (see Figure 5.)
Figure 5.
By these measures, it is clear that the Diamond-Orszag proposal meets the criteria for solving Social Security's budget problem. It eliminates the long-run deficit without implausible transition costs and without financing or benefit gimmicks.
The downside is that the benefit reductions and tax increases are permanent and ongoing. Therefore, the Diamond-Orszag proposal illustrates the eventual cost of not pre-funding Social Security's future benefit obligations.
Long-Run Costs and Savings of Diamond-Orszag and Other Proposals: There are many ways to compute Social Security's unfunded obligations. Figure 6 below, uses a very simple calculation: the net present value of Social Security's cash-flow surplus or deficit (as defined above) between 2003 and 2075 using a nominal discount rate 6 percent. This measure does not credit Social Security for the current balance in its trust fund -- the calculation is completely forward looking.
Under current law, the present value of Social Security's projected surplus of deficit in those years amounts to a net or unfunded deficit of about $5.1 trillion. By comparison, the proposal by anti-tax activist Peter Ferrara actually increases Social Security's deficit (in present value terms) to about $7.6 trillion. The proposal by Senator Lindsey Graham would reduce Social Security's unfunded obligations to about $3.5 trillion.
Of the three proposals we have studied so far, the Diamond-Orszag plan performs best by this measure, reducing Social Security's unfunded obligation to only $1 trillion between 2003 and 2075.
Figure 6.
The reason the Graham and Ferrara proposals have higher long-term liabilities is that their transition costs are not paid for. The Ferrara proposal has huge transition costs, which would actually increase the federal deficit by over 2 percent of GDP for decades (see Table 1). Interestingly, the Ferrara plan would raise Social Security costs to almost 8 percent of GDP, higher than the projected maximum cost of 7 percent under current law, before costs started to fall in later years.
The Graham proposal has much more modest transition costs, but they would still amount to about 1 percent of GDP for almost 2 decades. In budgetary terms, the transition costs of the Graham bill would be about $1.6 trillion between 2005 and 2014 (the 10-year budget period that begins Jan 1, 2004) and another $2.2 trillion in the following decade (see Table 2.)
Table 1.
Cost (+) or Savings (-) From Social Security Reform Proposals (As A Percent of GDP)
| Selected Years |
2010 |
2020 |
2030 |
2040 |
2050 |
2060 |
| |
|
|
|
|
|
|
|
| Sen. Lindsey Graham |
1.1 |
0.9 |
0.3 |
-0.5 |
-1.3 |
-2.0 |
| Peter Ferrara |
2.4 |
2.2 |
1.3 |
0.2 |
-1.2 |
-2.0 |
| Diamond-Orszag |
-0.3 |
-0.3 |
-0.5 |
-0.8 |
-1.1 |
-1.5 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Source: Centrists.Org, based on information from Social Security Administration.
Table 2.
| Cost Comparisons (trillions of dollars, by calendar year) |
|
|
| |
|
|
|
|
|
|
| |
|
Current Law |
|
Ferrara |
Graham |
Diamond-Orszag |
| |
|
|
|
|
|
|
| Present Value of Surplus or Deficit (2003-2075) |
5.1 |
|
7.6 |
3.5 |
1.0 |
| |
|
|
|
|
|
|
| Transition Cost or Savings (-) in Current Dollars: |
|
|
|
|
|
| 2005-2014 |
|
N/A |
|
3.7 |
1.6 |
-0.4 |
| |
|
|
|
|
|
|
| 2015-2024 |
|
N/A |
|
5.5 |
2.2 |
-0.9 |
| |
|
|
|
|
|
|
| Source: Centrists.Org |
|
|
|
|
|
|
|
|
|
Links:
Social Security Administration, Office of the Actuary Memorandum
Estimated Financial Effects of the Diamond-Orszag Proposal (
dated October 8, 2003;
released December 10, 2003)
CentristPolicyNetwork.Org
2004 Social Security Reform and Personal Retirement Accounts Resource PageCentrists.Org Unfunded Transition Costs of the Ferrara Social Security Proposal (December 2, 2003)
Centrists.Org
The Fourth Entitlement: Interest (December 1, 2003)
Centrists.Org
A Preliminary Analysis of Sen. Graham's Social Security Proposal (November 18, 2003)
Centrists.Org
Raising the Cap on Payroll Taxes Doesn't Solve the Social Security Problem(November 17, 2003)
CentristPolicyNetwork.Org
Put Social Security Reform in the President's Budget (November 17, 2003)
Centrists.Org
Suggestions for Income Testing in Social Insurance Programs (October 27, 2003)
CentristPolicyNetwork.Org
A Challenge To Both Left and Right on Social Security Reform (September 16, 2003)
Centrists.Org
No-BS Long-Term Budget Baseline HomepageCentrists.Org
Issue Summary: Wealth Building (Basics)Centrists.Org
Issue Summary: Social Security
