Poorly Designed Lock-Box Proposals -- 7/24/00

Poorly Designed Lock-Box Proposals That The Senate And House
Have Passed Pose Risks to The Economy
by James Horney and Robert Greenstein

The House and Senate each recently adopted "lock-box" legislation that would establish new rules ostensibly intended to ensure that Social Security and Medicare Hospital Insurance trust fund surpluses are used to pay down the federal debt. The House version of the lock-box is contained in H.R. 3859, a bill sponsored by Rep. Wally Herger and approved by the House on June 20. The Senate adopted two somewhat different and competing versions of the lock-box on June 29, one sponsored by Senator John Ashcroft and the other sponsored by Senator Kent Conrad. Both of the Senate lock-box measures were adopted as amendments to the Labor, Health and Human Services, and Education appropriation bill for fiscal year 2001.

Using the Social Security and Medicare HI surpluses to pay down the debt is a wise policy amidst today's booming economy and burgeoning surpluses. Nevertheless, these lock-box measures raise extremely serious concerns. It is not sound economic or fiscal policy to enact rules that limit the use of the Social Security and Medicare HI surpluses to paying down debt under any and all circumstances — including during an economic slowdown — unless three-fifths of the Senate agrees to do otherwise. Yet that is precisely what two of these lock-box measures would do. Both the House bill and the Senate-approved measure that Senator Ashcroft sponsored would require Congress and the President to raise taxes or cut programs when the economy falters and recession looms, unless three-fifths of the Senate vote to waive these rules.

Such tax increases or program cuts would be required even if the nation already was pursuing an unusually austere fiscal policy for a recession and the total budget was in surplus. Moreover, the President would be required under the House and Ashcroft proposals to submit a budget calling for tax increases or program cuts in the middle of a recession; the President would not be permitted to waive this requirement.

When the economy weakens, tax revenues fall or grow more slowly, while expenditures for programs like unemployment insurance increase. A slowdown in the economy thus could easily cause the non-Social Security, non-Medicare HI budget to fall out of balance. To maintain balance, budget cuts or tax increases would be required. Cutting government programs or raising taxes when the economy is slowing markedly, however, is the opposite of what sound economic policy would prescribe. Such actions would risk tipping a faltering economy into recession or making an incipient recession deeper and more protracted, thereby harming businesses and workers alike.

Some might argue that it would not be hard to obtain a three-fifths vote in the Senate to waive these requirements if there is an economic slowdown or some other unforseen circumstance develops that causes projected surpluses to disappear. But securing a three-fifths vote could prove problematic. There would likely be disagreement about how serious a slowdown has to be to justify waiving these requirements. It might not be possible to secure three-fifths approval until the economy is already in a recession.

The economy could weaken considerably and head toward a possible recession over a significant period of time before a recession actually began. During such a period, with growth slowing significantly, a non-Social Security, non-Medicare HI budget that otherwise was in close balance would threaten to become unbalanced. Under the House and Ashcroft lock-box requirements, that would necessitate budget cuts or tax increases, unless a three-fifths Senate majority could be secured to suspend the lock-box strictures. In such circumstances, however, a three-fifths majority could be difficult to obtain, since the economy was continuing to grow, although at a slower rate. Yet imposing austerity measures would only steepen the economy's descent.

Aggravating these problems, if economic growth fell sharply or a recession began, Congress would not have data confirming the sharp decline in growth or the onset of a recession for close to half a year or possibly more after the slowdown or recession started. For example, economic data show the last recession started in July 1990. But it was not officially declared a recession until April 25, 1991, some 10 months later. Furthermore, the effects of a slowdown can vary greatly by region of the country, which could make it difficult to garner the support of three-fifths of the Senate for suspending the lock-box requirements until the downturn had spread across most of the nation.

Finally, any time there is a requirement for a supermajority vote, that makes it easier for a minority to withhold its votes as bargaining chips to obtain the support of other lawmakers for something the minority wants to achieve, including policy changes of dubious merit that otherwise would not pass and that have nothing to do with the economy or the budget.

The Congressional Budget Office (CBO) projects that non-Social Security surpluses will total nearly $2.2 trillion over the next 10 years.⁽¹⁾ It also projects that Medicare Hospital Insurance trust fund surpluses will total $360 billion over this period, leaving projected non-Social Security, non-Medicare HI surpluses of $1.8 trillion. If the costs of maintaining current policies in such areas as aid to farmers, extension of current tax credits, and preventing the Alternate Minimum Tax from hitting millions of middle-class families are taken into account, however, the projected non-Social Security, non-Medicare surpluses are reduced by about $600 billion. Of the remaining $1.2 trillion amount, roughly $500 billion or more is likely to be needed as part of reform packages to restore the long-term solvency of the Social Security and Medicare HI programs, since few policymakers in either party will contemplate the types of benefit and payroll tax changes that are required to restore long-term solvency without general-revenue transfers of at least this magnitude.⁽²⁾ The $700 billion remaining of the projected surpluses could finance policy initiatives that address some of the nation's most pressing needs if funds are not squandered to pay for expensive tax cuts that primarily benefit better-off Americans or other dubious policy changes.

Under current conditions, it is a sound idea to use all of the Social Security and Medicare HI surpluses to pay down the federal debt. Doing so should help strengthen the economy over the long term and put the federal government in a better position to meet the demands that will be placed on the budget in future decades as increasing numbers of baby-boomers become eligible for Social Security and Medicare. Economic and other circumstances in the coming years, however, may turn out quite different than current budget projections assume. Under some circumstances, such as an economic slowdown, it would be wise to use some or all of the Social Security and Medicare HI surpluses to cover a portion of the costs of the rest of the government, rather than to pay down more debt. The alternative — instituting tax increases or deep spending cuts in the middle of a recession, or being unable to enact temporary increases in programs such as unemployment insurance during a downturn, simply in order to maintain balance in the non-Social Security, non-Medicare HI part of the budget — would be precisely the wrong response to an economic slowdown. Such a course would weaken the economy further and make the economic slowdown more severe. It also should be emphasized that running a non-Social Security, non-Medicare HI deficit does not harm Social Security or Medicare or reduce the assets of the trust funds and that adhering to the proposed lock-box rules would do nothing to extend the solvency of these programs (see box on next page).

The House version of the lock-box would greatly restrict the ability of Congress and the President to deal with a weakening economy unless and until three-fifths of the Senate, along with the House Rules Committee and majority of House members, agreed to waive the lock-box rules. The House bill would do the following:

Do These Lock-box Proposals Protect Social Security and Medicare?

Polls indicate there is a widespread belief among the public that a primary reason Social Security faces long-term financing problems is that the reserves accumulating in the Social Security trust funds to help pay the benefits of future retirees have been depleted by the use of Social Security revenues for other purposes. Lock-box proposals such as the House, Ashcroft, and Conrad proposals are often portrayed as shoring up Social Security and Medicare by ending these raids on the trust funds.

In fact, the Social Security and Medicare trust funds have not been "raided," and such portrayals of lock-box legislation are incorrect. The Social Security trust funds now have reserves of nearly $1 trillion, which are scheduled to rise to $6 trillion by 2025. The Medicare HI trust fund currently has reserves of more than $150 billion. The long-term financing problems of Social Security and Medicare are the result of entirely different factors. Moreover, lock-box proposals such as these would not shore up Social Security or Medicare or address its. Doing so entails tackling Social Security's long-term financial problems directly.

When the Social Security and Medicare trust funds take in more in revenues in a year than the trust funds need to pay Social Security and Medicare benefits that year, the Treasury borrows the surplus funds and provides the trust funds with Treasury bonds in return. This occurs regardless of whether the non-Social Security, non-Medicare HI budget is in deficit or in surplus. The House, Ashcroft, and Conrad proposals do not change this.

If there is no deficit in the non-Social Security, non-Medicare HI budget, the Treasury uses the surplus Social Security and Medicare HI revenues to pay down the debt. If there is a deficit in the rest of the budget, the surplus Social Security and Medicare HI revenues are first used to cover that deficit, and any remaining surplus revenues go to pay down the debt.

The point is that the Social Security and Medicare trust funds receive the same amount of Treasury bonds regardless of whether the Treasury uses Social Security and Medicare surpluses to help fund other government programs or to pay down debt. Social Security's and Medicare's assets increase by the same amount either way.

Deficits in the non-Social Security, non-Medicare budget thus do not diminish the assets the Social Security and Medicare trust funds hold and do not result in raids on these trust funds. The existence of such budget deficits until the past few years is not why Social Security and Medicare face long-term financing problems. Those problems are primarily due to demographic changes that will result in more retirees and slower labor force growth in future decades, increases in average life expectancy, and rising health care costs.

Since the trust funds do not lose assets when the Treasury uses part of the Social Security and Medicare surpluses to cover deficits in the non-Social Security, non-Medicare HI budget, the lock-box proposals would not alter the level of assets in the trust funds. This is why such proposals do not extend the dates on which Social Security and Medicare are projected to become insolvent (i.e., the dates after which these programs cannot continue paying full, as opposed to partial, benefits).

To be sure, using the Social Security and Medicare HI surpluses to pay down debt is a sound idea; it is helpful both to the long-term prospects of the economy and to the nation's long-term fiscal health. Doing so increases national saving, which in turn should result in modest increases in the long-term economic growth rate. A larger economy can more readily afford to provide the resources to finance Social Security, Medicare, and other needs in the future. Paying down debt also reduces the interest payments the federal government must make on the debt, thereby creating more room in the budget to help finance programs such as Social Security and Medicare in the decades ahead. But it is not a sound idea to require using all of the Social Security and Medicare HI surpluses to pay down the debt under any and all circumstances. If the economy slows or other emergency needs arise, using at least part of the Social Security and Medicare HI surpluses to help cover some of the costs of the rest of the budget until the slowdown or emergency ends would be much sounder policy than cutting other programs or raising taxes.

The Senate lock-box that Senator Ashcroft sponsored would achieve essentially the same effects as the House bill, although it would do so in a somewhat different manner. It would:

Unlike the House bill, the Ashcroft lock-box does not provide an exception for Social Security or Medicare reform legislation.

The third verison of the lock-box, the Senate version sponsored by Senator Conrad, takes a different approach that averts the worst — but not all — of the pitfalls of the House and Ashcroft approaches. It would:

The major distinction between the Conrad measure and the other two approaches is that the Conrad proposal does not bar a budget resolution from showing a non-Social Security, non-Medicare HI deficit, although it does prohibit a budget resolution from assuming enactment of legislation that would create or enlarge such a deficit. Thus, if a weakening economy was causing the non-Social Security, non-Medicare HI budget to show a temporary deficit, Congress and the President would not be obliged under the Conrad measure to erase the deficit by raising taxes or cutting programs when the economy is sputtering.

The Conrad approach does, however, parallel the House and Ashcroft proposals in one unfortunate respect. Under all three approaches, counter-cyclical measures to moderate recessions — including supplemental unemployment insurance benefits during a recession for workers whose regular unemployment benefits are running out and who cannot find jobs — would be barred. It would take 60 votes in the Senate, along with the approval of the House Rules Committee and a majority in the House, to waive this prohibition and provide supplemental unemployment benefits.

The House bill and the Conrad proposal also include a few additional provisions. The House bill contains an unusual provision that would make it illegal for any official government document to contain both on-budget and off-budget numbers, even though the legislation also requires the budget resolution to contain both such numbers. The Conrad proposal would exempt Medicare HI spending from sequestration under the Deficit Control Act.

The Lock-Box Proposals Would Weaken Policymakers' Ability to Manage the Economy

All three versions of the lock-box would restrict the flexibility of Congress and the President to deal with an economic turndown, a domestic or international crisis, or other circumstances that might lead to a substantial drop in revenues or a large increase in expenditures.

CBO currently projects that the non-Social Security, non-Medicare HI surplus will be $86 billion in 2002 if current policies are unchanged. CBO has made clear, however, that such a projected surplus can easily disappear even if no new laws are enacted that cut taxes or expand programs. CBO calculated in January that if it used slightly different assumptions about the path of the economy and Medicare and Medicaid spending — alternative assumptions that CBO described as "apparently reasonable" — projected non-Social Security, non-Medicare surpluses in 2002 would drop by about $125 billion. A swing of $125 billion in 2002 would turn CBO's currently projected non-Social Security, non-Medicare HI surplus of $86 billion into a $40 billion deficit. If some or all of the projected surpluses have been used already to fund tax cuts or programs increases, as is likely to be the case, the non-Social Security, non-Medicare HI deficits that could materialize if the economy grows more slowly than anticipated would be larger.

Suppose the economy and the budget turn out to follow the modestly less-optimistic CBO alternative forecast. In that case, complying with the Ashcroft and House lock-box proposals would require that the budget resolution call for tax increases or program cuts totaling about $40 billion in 2002 (or more than that if tax cuts or spending increases enacted in the meantime have already used part of the currently projected surplus).

As noted, the Conrad lock-box differs in this area because it does not call for tax increases or cuts in programs if non-Social Security, non-Medicare HI deficits are projected under current policy assumptions. The Conrad proposal does not require that damaging fiscal policies be pursued if the economy falters unless three-fifths of the Senate vote to suspend the lock-box procedures. The Conrad proposal also does not prohibit the President from proposing a budget containing sensible fiscal policies when the economy turns down, including policies that entail running a non-Social Security, non-Medicare HI deficit during such a period.

Even so, the Conrad proposal still would create a significant problem since, like the other proposals, it would not allow the budget resolution or subsequent legislation to provide for tax cuts or program increases that could be needed to respond to a deteriorating economic situation or another crisis, unless 60 senators voted to allow that.

Imposing austerity measures when the economy slows is the reverse of how sound fiscal policy should work. The House and Ashcroft lock-boxes pose a risk of tipping a faltering economy into recession and making a recession deeper and more long-lasting. If the required tax increases or program reductions are large enough, the resulting contractionary effect on aggregate demand could have a significant negative effect on a slowing economy.

These proposals also would pose risks to the economy during the early stages of a recovery. The largest fiscal effects of an economic slowdown are a reduction in corporate and personal income taxes. Those taxes often are paid with a lag in time. As a result, revenue collections usually remain low for a period of time after a recession has technically ended and a recovery has begun. Unemployment insurance and food stamp expenditures also often lag behind the slowdown and continue rising during the early part of the recovery. This reflects the fact that unemployment rates typically continue to rise for a number of months after a recession has ended and a recovery has begun. For example, the last recession ended after the first quarter of 1991 (January through March of 1991). The unemployment rate, however, continued to rise from 6.7 percent in April 1991 to 7.8 percent in June 1992 and did not fall below 6.7 percent until November 1993.

It should be noted that the point at which a recession technically ends and a recovery begins is the point at which the economy hits bottom, stops falling, and begins to grow again. Thus, the point at which a recovery begins is the lowest point of the economic cycle, and the economy often is quite weak at that time. Non-Social Security, non-Medicare HI deficits are likely to persist for one or a few years after a recession is officially over.

Yet the fact that a recession has officially ended could make it difficult to overcome the points of order the House and Ashcroft lock-boxes would create against a budget resolution that simply recognizes the temporary non-Social Security, non-Medicare HI deficits that a weak economy has generated. If such a point of order could not be overcome, most likely because of an inability to secure a three-fifths vote in the Senate, Congress would find itself compelled to cut programs or raise taxes during the early stages of a recovery when the economy is still fragile and unemployment is still rising or remains at a high level.

Such policies could abort the nascent recovery. Congress also could find it difficult to pass regular appropriation bills in such circumstances. (See the box below.)

Rules Pose Risks for Appropriations Bills

The House and Ashcroft lock-boxes could place appropriations bills at risk. Suppose Congress approved a budget resolution that assumed enactment of program cuts or tax increases to eliminate a projected non-Social Security, non-Medicare HI deficit that had emerged due to slower-than-forecast economic growth. Suppose Congress then proved unable to pass the program cuts or tax increases (or passed them but not in the full amount the budget resolution assumed). In such a circumstance, even if the Appropriations Committees have met all of their targets, the last appropriations bill or two would probably be said to "cause or increase" a non-Social Security deficit.

Such appropriations bills consequently would trigger a point of order under both the House and the Ashcroft lock-box proposals and require 60 votes to pass in the Senate. The operation of the points of order in these lock-box proposals consequently appears to be skewed against the appropriations process.

One of the nation's most respected economists, Robert Reischauer, president of the Urban Institute, delivered testimony that bears on this issue in an appearance before the House Budget Committee in May 1992 in his capacity as director of the Congressional Budget Office. Reischauer explained that when the economy is weak, the government acts automatically to stabilize it, since revenue collections automatically slow and expenditures for unemployment insurance and means-tested benefit programs automatically rise. Reischauer observed that "This automatic stabilizing occurs quickly and is self-limiting — it goes away as the economy revives — but it temporarily increases the deficit. It is an important factor that dampens the amplitude of our economic cycles."⁽⁴⁾ Under a requirement that the budget resolution must show balance in the non-Social Security, non-Medicare HI part of the budget even if the economy is weak, the government's automatic stabilizing role would be cast into doubt.

Defenders of the House and Ashcroft proposals dismiss these concerns, arguing that the House Rules Committee, a majority of the House, and three-fifths of the Senate can set these restrictions aside when the economy weakens. Such claims, however, make it sound easier to overcome these problems than it is likely to be. It is unlikely, for example, that a three-fifths majority would materialize in the Senate until after the economy is already in or very near a recession and the overly restrictive fiscal policies the lock-boxes have engendered have caused some damage.

The Office of Management and Budget and the Congressional Budget Office have never forecast a recession before one started. We usually do not know we are in a recession until the downturn is at least several months old. As a result, these budgetary procedures could lead to stiffer austerity measures being taken in years when the economy was weakening but not yet in a full-blown recession. Such measures could trigger a recession that otherwise might not have occurred or have started so soon. Such austerity measures also could make a recession deeper.

Adding to this problem, a three-fifths majority could be particularly difficult to garner if a recession were regional rather than national in nature, as is usually the case at least at the start of an economic downturn. It might not be possible to obtain a three-fifths majority until a recession had spread to a substantial majority of states.

Past recessions generally have started in some regions and taken time to expand; they also have hit some regions much harder than others. In the last recession, New England and the mid-Atlantic states began experiencing declines in employment a full year before employment turned down in most of the rest of the country. If rising unemployment insurance costs and falling revenues in several regions threatened to push the non-Social Security, non-Medicare HI budget out of balance, would enough Senators from states where economic problems were not yet evident be willing to allow a budget resolution to show a deficit in the non-Social Security, non-Medicare HI budget? If not, fiscal retrenchment might be required that could make the downturn deeper, and regions needing federal recession relief might find it was not forthcoming when they needed it.

Economic downturns also tend to last longer in some regions than others. In the last recession, employment losses lasted two to four years in California and much of New England and the mid-Atlantic states, while a number of other states recouped their employment losses in a matter of months. This pattern was reversed in the recession of the early 1980s; that downturn was sharpest and lasted longest in the Midwest and South. The regional patterns that characterize economic downturns raise further questions about the wisdom of enacting legislation that prohibits a budget resolution from showing a deficit in the non-Social Security, non-Medicare HI budget and allows this stricture to be relaxed only when three-fifths of the Senate so approve.

As noted, the Conrad lock-box does not suffer from this flaw; it would allow the automatic stabilizing effect of lower revenues and higher expenditures for unemployment insurance and other means-tested benefit programs to operate and help dampen the effects of an economic slowdown. It is vastly superior to the House and Ashcroft lock-boxes for this reason. The Conrad proposal would, however, make it difficult for Congress to pass legislation that may be needed to provide additional countercyclical fiscal relief and make assistance available to Americans hurt by an economic downturn. The coverage and benefits of the unemployment insurance program often are highly inadequate in a recession, and Congress has found it necessary in recent recessions to enact temporary extensions of unemployment insurance benefits to ensure that benefits are available to workers who are out of jobs for long periods through no fault of their own. During the economic slowdown of the early 1990s, for example, a temporary Emergency Unemployment Compensation program was enacted that provided almost $19 billion in unemployment benefits over three years. Under the Conrad lock-box (as well as under the other two lock-box proposals), it would not be possible to secure Senate passage of similar legislation during a future economic downturn unless three-fifths of the Senate so agreed. Here, too, it could prove difficult to obtain three-fifths support despite a need for such benefits until the recession had spread nationwide and considerable hardship had occurred.

The potential harm from the Conrad lock-box would be somewhat reduced if a provision were added to the proposal that suspended the lock-box requirements in case of an economic slowdown. The Gramm-Rudman-Hollings law provided for suspension of its deficit targets by a simple majority vote in the House and Senate (and agreement by the President) if reported real economic growth fell below 1 percent for two quarters or if CBO or the Administration forecast a recession. The lock-box legislation proposed last year by Senators Spencer Abraham and Pete Domenici provided for automatic suspension of the lock box without any vote by Congress if economic growth was less than one percent for two consecutive quarters.

Since, as noted above, a recession is not likely to be forecast before it starts and the negative effects of a recession can linger long after the economy begins to recover and grow again — and some areas of the country could suffer severe economic hardships even if the economy as a whole is growing slowly — including a provision to suspend the Conrad lock-box procedures when growth is below one percent would not eliminate all of the potential problems. But it could help significantly in some circumstances, particularly if the suspension is automatic and does not require a vote.

It is important to note, however, that providing for suspension of the lock-box procedures in the case of low growth would not significantly improve the Ashcroft or House lock-boxes. As indicated above, revenues could drop far below currently projected levels before economic reports would trigger a low-growth suspension, and revenues could remain low and unemployment benefit payments high for a substantial period of time after a recession has ended and the economy has begun growing at a rate of somewhat more than one percent. A procedure for suspending the Ashcroft or House lock-boxes when a slowdown is certified would not keep those lock-boxes from requiring fiscal policies that could tip a weak economy into a recession, make a recession deeper, or slow a fledgling recovery.

Of course, an economic slowdown is not the only circumstance in which relief from the lock-box rules would be desirable. Other unexpected events — such as a war or other international crisis, a severe natural disaster, or a financial emergency like the savings and loan crisis of the late 1980s and early 1990s — could produce unexpected increases in spending or reductions in revenues. If Congress and the President have enacted tax cuts and program increases that have used up all of the projected non-Social Security, non-Medicare surpluses, then responding to an international conflict, a natural disaster, or some other emergency would push the budget into deficit unless the cost of responding to such a problem is offset by budget cuts or tax increases. Requiring offsetting budget cuts or tax increases in such circumstances, however, would make it harder to respond in a timely manner to emergency needs. As a result, it would be a good idea also to provide a process for suspending the lock-box rules in the case of non-economic emergencies. Such a process could include a definition of "emergency" and a mechanism for determining when an emergency has occurred, to prevent the misuse of this process.

Because of the serious economic problems that would result if the House and Ashcroft lock-boxes were adhered to when the economy weakened, these proposals would likely produce another result as well — the use of heroic assumptions and patently dishonest estimates in budget resolutions so the resolutions do not show deficits in the non-Social Security, non-Medicare HI budgets. We saw similar developments in the late 1980s, when highly unrealistic estimates were used to avert sequesters under the Gramm-Rudman-Hollings Act. With the advent of the Budget Enforcement Act of 1990 — which enforces budget discipline through ceilings (or "caps") on discretionary programs and pay-as-you-go rules that require the costs of tax cuts and entitlement increases to be offset, rather than through rigid deficit targets that are not responsive to economic cycles — the budgets again reflected relatively honest numbers. The House and Ashcroft lock-boxes would likely usher in an era in which budget resolutions resort to highly unreliable assumptions when problems emerge.

1. Congressional Budget Office, The Budget and Economic Outlook: An Update (July 2000).

2. James Horney and Robert Greenstein, How Much of the Enlarged Surplus Is Available for Tax and Program Initiatives?: Available Funds Should be Devoted to Real National Priorities, https://www.cbpp.org/6-29-00bud.htm, (June 29, 2000).

3. Under current law, transactions of the U.S. Postal Service, as well as those of Social Security, are off-budget. But while the Social Security surpluses are projected to average more than $200 billion a year over the next 10 years, the Postal Service runs surpluses or deficits of at most a few billion dollars a year. Over time, the Postal Service surpluses and deficits net to approximately zero.

4. Statement of Robert D. Reischauer, Director, Congressional Budget Office, before the House Budget Committee, May 6, 1992.