July 13, 1999 | Testimony on Social Security
Testimony before the Task Force on Social Security / Committee on the Budget - United States House of Representatives
We appreciate the opportunity to appear before you today to discuss the costs of transitioning to a solvent Social Security system. At the outset, let us state that the views expressed in this testimony are our own, and should not be construed as representing any official position of The Heritage Foundation.
High transition costs will be a fact of life for Social Security regardless of whether the program is radically reformed or just left as it is. While some consider transition costs to apply only to proposals that would reform Social Security, this is not the case. Since the existing program will begin to run cash flow deficits in 2014, the transition costs of various reform proposals should be measured against the costs associated with doing nothing at all. In fact, it would probably be more accurate to talk about "preservation costs" instead of transition costs.
"Transition Costs" Defined
We define the transition costs for the Social Security retirement program as the total amount of money that must come from sources other than Social Security payroll taxes at the current level and the small portion of the income tax on benefits paid to certain higher-income retirees. Thus, increasing the payroll taxes would count as part of the transition costs as would any general revenue that is transferred to the program.
The easiest way to measure this cost is to look at the annual cash flow deficits of the Old-Age, Survivors and Disability Insurance (OASDI) trust funds under both the existing program and the various proposals that have been made. For this analysis, we use estimates made by the Social Security Administration (SSA) Office of the Chief Actuary with only one change. While the SSA measures these amounts in percentages of taxable payroll, we translate them into constant 1999 dollars in order to make them more understandable.
When considering the various reform plans, we compare the operating deficits (if any) that would result after subtracting trust fund income (mainly payroll tax revenues) from trust fund costs (the aggregate benefits that would be paid under the reformed system). However, let us emphasize once again that our definition considers only payroll taxes at the current level. Increased payroll taxes, whether by raising the wage cap or increasing the tax rate or other revenues that are used to fill the operating deficit count as part of the transition cost. This is true regardless of whether the general fund revenues come from a budget surplus. In either case, they represent additional resources that are used to pay Social Security benefits.
This analysis is limited to measuring the effect of various policy options on Social Security revenue and outlays, and by extension on the government's budget. It does not measure changes in the retirement benefits received by individuals. For instance, both the existing system and the Archer-Shaw plan assume that benefits would be paid at levels called for under current law, while the Gramm plan assumes that the combination of traditional benefits and individual accounts would equal at least 120 percent of that amount. On the other hand, the Kasich plan assumes that retirement benefits would remain at the current level instead of growing in real terms as is called for under existing law. Kolbe-Stenholm, meanwhile, assumes that Social Security retirement benefits from traditional sources would gradually decline as the amount available from individual accounts grows. These effects on the individual can be measured only indirectly in a discussion of transition costs.
Applying This Definition to the Existing System and Various Reform Plans
Three quick examples show how this definition applies to the existing system and various reform plans. Looking at the SSA's intermediate prediction for the existing program in 2020, the OASI trust fund is estimated to take in $634 billion in taxes and pay out $737 billion in benefits. Even if the $104 billion operating deficit is covered by liquidating some of the assets in the OASI trust fund, that money comes from sources other than the payroll tax, and should be considered part of the transition cost.
This is not to imply that the special-issue Treasury bonds held in the trust fund are worthless or will not be repaid on schedule. However, the Analytical Perspectives volume of President Bill Clinton's fiscal year 2000 budget accurately characterize the assets in this trust fund when it says:
These balances are available to finance future benefit payments...only in a bookkeeping sense. They do not consist of real economic assets that can be drawn down in the future to fund benefits. Instead, they are claims on the Treasury that, when redeemed, will have to be financed by raising taxes, borrowing from the public, or reducing benefits, or other expenditures. [p. 337 -- italics added for emphasis]
On the other hand, the Kolbe-Stenholm plan would divert an amount of the Social Security tax equal to 2 percent of income to individual accounts and adjust the traditional benefits to reflect the gradual ability of individual accounts to pay some portion of Social Security benefits. In 2020, after making these adjustments, Kolbe-Stenholm assumes that $57 billion would be transferred from general revenues to cover the operating deficit. A further $27 billion would be brought in from the effect on income tax collections of re-estimating the consumer price index, for a total of $84 billion.
The Archer-Shaw plan would make no changes to the existing payroll tax rate or to the existing benefit structure. However, it would fund individual accounts with an amount of general revenues equal to 2 percent of income and require general revenue funds to pay Social Security benefits. Thus, in 2020 the transition cost for Archer-Shaw would include both $98 billion for the personal accounts and $72 billion to pay benefits for a total of $170 billion.
Where Does the Money Come From?
The source of the money for general revenue transfers to pay Social Security benefits is extremely important. In short, no matter where the money comes from, Congress must always face opportunity costs. They range from foregoing expansion of a non-Social Security program in order to pay the interest on borrowed money to being forced to raise taxes to support Social Security outlays rather than spending those funds on urgent needs in education or defense. Depending on a variety of factors, there also could be an effect on the growth rate of the entire domestic economy.
The phrase "there is no free lunch" has never been more true than in this situation. To the extent that the money is borrowed, future generations will bear a significant interest cost that will be in addition to the base transition cost. If the federal government borrowed a significant amount for Social Security, under some circumstances this could cause an increase in interest rates for the overall economy. This in turn could lower economic growth, thus reducing payroll tax collections below anticipated levels.
Congress also needs to consider the effects on the rest of the budget that stems from increased general revenue spending to meet Social Security's challenges. It is easy to assume that this can be paid out of surpluses, but there is a catch to the recent good news on that front. Over the past six months, the White House's estimate of the cumulative 15-year budget surplus has gone up by $1.1 trillion. However, the beginning of an economic downturn, which is inevitable at some point, might cause a downward revision of an equal amount. It is a fallacy to assume that these surpluses are inevitable.
In that case, future Congresses may face the choice between tax increases and significant reductions in other programs. If there is no surplus, where will the $104 billion come from that will be required to redeem assets of the trust fund in 2020? Will the 115th Congress sitting in 2020 be forced to reduce spending for education, highways, defense, or other programs to pay Social Security benefits? When considering transition costs, it will be important to consider aggregate deficits, how long they will last, and how large the individual annual deficits are. To a very real degree, the actions of this Congress and the next one will limit the ability of future Congresses to meet national needs.
You Can Pay Me Now or You Can Pay Me Later
As it will become clear from looking at both the forecast for the current system and various reform options, there is no easy solution to Social Security. No matter what, future taxpayers will bear a significant additional burden to pay the benefits of that time's retirees. The only questions surround when the annual deficits will begin, how big they will be, and how long they will last.
The benefits of individual accounts would take some time to develop. Even if a taxpayer were allowed to begin them tomorrow, the accounts would not grow large enough to offset any significant amount of the traditional benefits for a good 20 to 30 years.
If a portion of the existing Social Security tax were diverted to individual accounts, there would be a direct relationship between the amount that could go into an individual account, the size of the initial deficits, and how soon the deficits could end. Diverting part of the tax would reduce Social Security's income and cause almost immediate deficits. On the other hand, the more that went into the individual accounts, the faster they could grow to a significant size and replace much of the traditional benefit. However, because the early deficits could be so large, most reform plans initially would limit individual accounts to an amount equal to 2 percent of income.
If Congress did nothing, annual cash flow deficit would begin in 2014. They would amount to about $21 billion in 2015, $252 billion in 2030, and $516 billion in 2070. It appears that the annual deficits would continue and grow in size so long as they could be measured. On the other hand, most reform plans would start to run overall deficits sooner, but they would tend to be smaller over time, and in a few cases they eventually would end.
As expected, plans such as Kolbe-Stenholm, Kasich, and the Senate bipartisan plan that finance individual accounts with part of the existing Social Security taxes would begin to run deficits almost immediately. While these plans adjust the traditional benefit that is financed solely from payroll taxes, these reductions would begin to reduce costs only after the individual accounts had the chance to grow for 20 to 30 years. This necessary delay in cost reduction would cause these plans to run significant deficits that grew for about 30 years and then began to decline steadily.
For instance, Kolbe-Stenholm would reach a maximum annual deficit of $133 billion in 2030, but by 2070 the deficit would decline to only $38 billion. In several years in between, there actually would be a surplus. The Kasich plan also would reach a maximum deficit of $187 billion in 2030, but deficits essentially would end after 2065. This pattern also would be true for the Senate bipartisan plan, where deficits would reach $130 billion in 2030, but would end in 2065.
Because our definition includes all outside revenues other than the existing level of payroll taxes, this pattern also would hold true for the Archer-Shaw plan. Even though under Archer-Shaw, the Social Security trust fund would not begin annual cash flow deficits until 2014, the level of general revenues that would be necessary to fund the add-on individual accounts would cause an almost immediate aggregate deficit. Thus, while Social Security would run a surplus in 2000 of $69 billion, the $74 billion for general revenues that went into the Archer-Shaw accounts would cause a $5 billion aggregate deficit. This deficit would climb to $255 billion in 2030, before declining to $185 billion in 2070.
Thus, what this Congress does will have a major impact on the future. In 2000, this country could face a Social Security surplus if Congress did nothing. Passing Kolbe-Stenholm, Archer-Shaw, the Senate bipartisan plan, or Kasich would result in a deficit of about $5 billion.
By the time a child born in 2000 reaches the age of 30 in 2030, the annual deficit will climb to $252 billion under the existing system. Under Kolbe-Stenholm, it would be $133 billion, while the Archer-Shaw deficit would be $255 billion. That same year, the Senate bipartisan plan would run a deficit of $130 billion, and the Kasich plan deficit would be $187 billion.
In 2070, the existing system will run a $516 billion deficit. For Kolbe-Stenholm, it would be $38 billion, for Archer-Shaw $185 billion. However, for both the Senate bipartisan plan and the Kasich plan, there would be no Social Security deficit.
Of course, there is more to be considered than just aggregate costs. The simple fact is that for millions of low- and moderate-income families, Social Security is the only retirement plan they have. Unfortunately, for most of them today's Social Security is not a good investment. At a time when the Standard & Poor's 500 has gone up 20.5 percent in the past 12 months, Social Security "earns" the equivalent of only 1.3 percent.
The objective of Social Security reform must be more than just restoring the financial health of the system. It is time to allow every American family -- no matter what their income level -- to have the opportunity to fully participate in our economy. Social Security reform must also improve the retirement income for low- and moderate-income workers.
The real question is how responsible this Congress and the one following want to be to future generations. It can do nothing, and place a significant burden on our children and grandchildren -- or it can act responsibly and reduce that burden. Your decision will have an even greater impact on our children's grandchildren. What kind of a legacy do we as a people want to leave to the future?