Skip to main content

Greenstein: Balanced Budget Amendment Unsound Policy

Testimony of Robert Greenstein, President, Center on Budget and Policy Priorities Before the Senate Committee on the Judiciary

Thank you for the invitation to testify today.  I am Robert Greenstein, president of the Center on Budget and Policy Priorities, a policy institute that focuses both on fiscal policy and on policies affecting low- and moderate-income Americans.  As part of our work, we have been analyzing proposed changes in budget procedures for more than 20 years.  We have conducted extensive analyses of proposals to write a balanced budget requirement into the Constitution, among other possible changes in budget rules.

The purpose of changing our fiscal policy course is to strengthen our economy over the long term and to prevent the economic damage that could occur if the debt rises in future decades to unsustainable levels as a share of the economy.  But we need to choose our fiscal policy instruments carefully.  We want to avoid “destroying the village in order to save it.”

The goal of a constitutional balanced budget amendment is to address our long-term fiscal imbalance.  Unfortunately, a constitutional balanced budget amendment would be highly ill-advised as a way to try to do that and likely would cause serious economic damage.  It would require a balanced budget every year regardless of the state of the economy, unless a supermajority of both houses overrode that requirement.  This is an unwise stricture that large numbers of mainstream economists have long counseled against, because it would require the largest budget cuts or tax increases precisely when the economy is weakest.  It holds substantial risk of tipping faltering economies into recessions and making recessions longer and deeper.  The additional job losses would likely be very large.

When the economy weakens, revenue growth drops and revenues may even contract.  And as unemployment rises, expenditures for programs like unemployment insurance (UI) — and to a lesser degree, SNAP (food stamps) and Medicaid — increase.  These revenue declines and expenditure increases are temporary; they largely disappear as the economy recovers.  But they are critical for helping to keep struggling economies from falling into a recession and for moderating the depth and length of recessions that do occur. 

When the economy weakens, consumers and businesses spend less, which in turn causes further job loss.  The drop in tax collections and increases in unemployment and other benefits that now occur automatically when the economy weakens cushion the blow, by keeping purchases of goods and services from falling more.  That is why economists use the term “automatic stabilizers” to describe the automatic declines in revenues and automatic increases in UI and other benefits that occur when the economy turns down; these actions help stabilize the economy. 

A constitutional balanced budget amendment, however, effectively suspends the automatic stabilizers.  It requires that when the economy weakens, federal expenditures be cut or taxes increased to offset the effects of the automatic stabilizers and prevent a deficit from occurring — the opposite course from what sound economic policy calls for.

Over the years, leading economists have warned of the adverse effects of a constitutional balanced budget amendment.  In congressional testimony in 1992, Robert Reischauer — then director of the Congressional Budget Office (CBO) and one of the nation’s most respected experts on fiscal policy — explained:  “[I]f it worked [a constitutional balanced budget amendment] would undermine the stabilizing role of the federal government.”  Reischauer noted that the automatic stabilizing that occurs when the economy is weak “temporarily lowers revenues and increases spending on unemployment insurance and welfare programs.  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.”  Under the constitutional amendment, he explained, these stabilizers would no longer operate automatically.[1]

Similarly, when a constitutional balanced budget amendment was under consideration in 1997, more than 1,000 economists including 11 Nobel laureates issued a joint statement that said:

We condemn the proposed “balanced-budget” amendment to the federal Constitution.  It is unsound and unnecessary. . . .  The proposed amendment mandates perverse actions in the face of recessions.  In economic downturns, tax revenues fall and some outlays, such as unemployment benefits, rise.  These so-called “built-in stabilizers” limit declines of after-tax income and purchasing power.  To keep the budget balanced every year would aggravate recessions.[2]

In 2011, then-CBO director Douglas Elmendorf sounded a similar warning when asked about a constitutional balanced budget amendment at a Senate Budget Committee hearing.  Elmendorf observed:

Amending the Constitution to require this sort of balance raises risks. . . .  [T]he fact that taxes fall when the economy weakens and spending and benefit programs increase when the economy weakens, in an automatic way, under existing law, is an important stabilizing force for the aggregate economy.  The fact that state governments need to work . . . against these effects in their own budgets — need to take action to raise taxes or cut spending in recessions — undoes the automatic stabilizers, essentially, at the state level.  Taking those away at the federal level risks making the economy less stable, risks exacerbating the swings in business cycles.[3]

And in October 2011, Macroeconomic Advisers (MA) analyzed the economic impacts of a constitutional balanced budget amendment.[4]  One of the nation’s preeminent private economic forecasting firms, Macroeconomic Advisers provides analysis to major corporations and government entities, such as the President’s Council of Economic Advisers under Presidents of both parties, including Presidents Reagan and George W. Bush.

MA concluded that if a constitutional balanced budget amendment had been ratified and were being enforced for fiscal year 2012, “the effect on the economy would be catastrophic.”  If the 2012 budget were balanced through spending cuts, MA found, those cuts would total about $1.5 trillion in 2012 alone — and would throw about 15 million more people out of work, double the unemployment rate from 9 percent to approximately 18 percent, and cause the economy to shrink by about 17 percent instead of growing by an expected 2 percent. 

Even if a balanced budget amendment were implemented when the budget was already in balance, MA concluded, it would still put “new and powerful uncertainties in play.  The economy’s ‘automatic stabilizers’ would be eviscerated [and] discretionary counter-cyclical fiscal policy would be unconstitutional. . . .  Recessions would be deeper and longer.” 

MA also warned that “The pall of uncertainty cast over the economy if it appeared a [balanced budget amendment] could be ratified and enforced in the middle of recession or when the deficit was still large would have a chilling effect on near-term economic growth.”  MA concluded that a balanced budget amendment would have detrimental effects on economic growth in both good times and bad.

Proponents of a constitutional amendment often respond to these admonitions by noting that the proposed constitutional amendment would allow the balanced budget requirement to be waived by a vote of two-thirds or three-fifths of the House and the Senate, so the BBA could be set to the side in recessions.[5]  But this response is too facile, and the waiver provision does not solve the problem.  It is difficult to secure even three-fifths votes for anything; consider the paralysis that marks much of the work of the Senate.  Moreover, it may take months after a downturn begins before sufficient data are available to convince a supermajority of the members of both houses of Congress that a recession is underway.  Furthermore, it is all too likely that even after the evidence for a downturn is clear, a minority in the House or Senate would hold a wavier vote hostage to demands for concessions on other matters.  By the time that a recession were recognized to be underway and supermajority votes were secured in both chambers, if such support could be obtained at all, extensive economic damage could have been done and hundreds of thousands or millions of American jobs unnecessarily lost.

The bottom line is that the automatic stabilizers need to continue to be able to work automatically to protect American businesses and workers.  The balanced budget amendment precludes that.

Nor is a recession the only concern.  Consider the savings and loan crisis of the 1980s, or the financial meltdown of the fall of 2008.  A constitutional balanced budget amendment would have hindered swift federal action to rescue the savings and loan industry or to put the Troubled Assets Relief Program in place expeditiously.  In both cases, history indicates that federal action helped save the economy from what otherwise likely would have been far more dire problems. 

More generally, a balanced budget is not necessarily the right fiscal policy goal even when there is no immediate crisis. “There’s nothing magic about exact balance,” Alice Rivlin, a former director of both CBO and the Office of Management and Budget, has noted.  “The really important thing is to keep the debt from growing faster than the economy.”[6]  Keeping the debt ratio — the debt held by the public as a percent of gross domestic product (GDP) — from continually rising is the soundest approach.  The debt ratio should grow only during hard economic times or major emergencies, and it should be stable or decline during good times.  It isn’t necessary to balance the budget, however, in order to stabilize or reduce the debt ratio.  Between 1946 and 1979, the debt ratio plummeted from 106 percent of GDP to only 25 percent.  Yet in only eight of those 33 years was the budget in balance or in surplus.

For all these reasons, fiscal policy should not be written into the Constitution.

A parallel problem is that most proposed constitutional balanced budget amendments would make it even harder than it already is to raise the debt limit, by requiring at least a three-fifths vote of both the House and Senate to raise the limit.  This is playing with fire.  It would heighten the risk of a federal government default.  A default would raise our interest costs and could damage the U.S. economy for years to come.

Effects on Social Security and Military and Civil Retirement

A balanced budget constitutional amendment could also seriously disrupt Social Security and other retirement programs.  By design, the Social Security trust fund is now building up reserves — in the form of Treasury securities backed by the full faith and credit of the United States — which will be drawn down to help pay benefits when the number of retired “baby boomers” climbs higher in the 2020s and early 2030s.  Currently, Social Security holds $2.8 trillion in Treasury securities.  But under the balanced budget amendment, it would in effect be unconstitutional for Social Security to draw down these savings to pay promised benefits.  More precisely, Social Security would be allowed to use its accumulated Treasury securities to help pay benefits only if the rest of the federal budget ran an offsetting surplus (or if the House and Senate each mustered two-thirds or three-fifths votes to permit an unbalanced budget).  Otherwise, benefits would have to be cut, because all spending — including Social Security benefits — would have to be covered by tax revenues collected during that same year.

The military retirement and civil service retirement systems, which have their own trust funds, would be affected in the same way.  Because all expenditures would have to be covered by taxes collected in the same year — and the use of accumulated savings would thus effectively be unconstitutional — these trust funds would not be able to draw down their accumulated balances, currently amounting to $1.5 trillion, unless the rest of the budget ran offsetting surpluses.

Effects on the Banking System and Other Insurance and Loan Guarantees

The potential effects on the banking system also are cause for concern.  The Federal Deposit Insurance Corporation (FDIC) currently holds more than $60 billion of reserves, in the form of Treasury securities, to insure depositors’ savings.  These reserves are called upon when banks fail.  Similarly, the Pension Benefit Guarantee Corporation (PBGC) currently has more than $18 billion of reserves to draw upon if a corporation’s defined-benefit pension plan goes bankrupt.

Here, too, the balanced budget amendment would effectively make it unconstitutional for the FDIC and the PBGC to use their assets to pay deposit or pension insurance, since doing so would generally constitute “deficit spending.”  Such payments could be made only if the rest of the budget ran an offsetting surplus that year (or if Congress achieved the necessary three-fifths or two-thirds supermajorities to override the balanced budget requirement).

In general, a constitutional requirement that all spending during a given year be covered by tax revenues collected in the same year would undercut all U.S. government insurance and loan guarantees.  Those range from the “full faith” backing by the U.S. government to pay interest on Treasury securities to deposit insurance, pension insurance, FHA loans, small business loans, flood insurance, and the nuclear power industry’s liability insurance under the Price-Anderson Act.  Henceforth, the U.S. government would only be able to fulfill its legal commitments if their cost did not cause a deficit or if both houses of Congress voted by a two-thirds or three-fifths supermajority to waive the balanced budget requirement.

The entire purpose of deposit insurance and other U.S. financial commitments is to guarantee financing in case of calamity.  How reliable is the “guarantee” if a balanced budget requirement places it at risk or forces it to be withdrawn just when it is needed most?

Mistaken Analogies to States and Families

Proponents of a constitutional amendment sometimes argue that states and families must balance their budgets every year and the federal government should do so, too.  But statements that the constitutional amendment would align federal budgeting practices with those of states and families are mistaken.

While states must balance their operating budgets, they can borrow to finance their capital budgets — to finance roads, schools, and other projects.  Most states do so.  States also can build reserves (or “rainy day funds”) during good times and draw on them in bad times without counting the drawdown from reserves as new spending that unbalances a budget.

Families follow similar practices.  They borrow — they take out mortgages to buy a home or student loans to send a child to college.  They also draw down savings when times are tight, with the result that their expenditures in those periods exceed their current incomes.

But the proposed constitutional amendment would bar such practices at the federal level.  The total federal budget — including capital investments — would have to be balanced every year, with no borrowing allowed for infrastructure or other investments that can boost future economic growth.  And if the federal government ran a surplus one year, it could not draw it down the next year to help balance the budget.

I would also note that the fact that states must balance their operating budgets even in recessions makes it all the more important from the standpoint of economic policy that the federal government not be subject to the same stricture.  American Enterprise Institute analyst Norman Ornstein addressed this matter in an article a few years ago, where he wrote:  “Few ideas are more seductive on the surface and more destructive in reality than a balanced budget amendment.  Here is why:  Nearly all our states have balanced budget requirements.  That means when the economy slows, states are forced to raise taxes or slash spending at just the wrong time, providing a fiscal drag when what is needed is countercyclical policy to stimulate the economy.”[7]

S.J. Res. 6 and S.J. Res. 2 Raise Additional Issues

The foregoing concerns apply to all versions of the balanced budget amendment that have been introduced.  Some versions of the balanced budget amendment, such as S.J. Res. 6 and S.J. Res. 2, raise additional concerns, because they would write into the Constitution new barriers to raising revenues to help balance the budget and would also prohibit federal expenditures in any year from exceeding a figure such as 18 percent of the GDP in the previous calendar year.  These constitutional requirements could be overridden only by two-thirds votes in both the House and the Senate.

A requirement for a two-thirds vote to raise any taxes, which S.J. Res. 6 would impose, would be extremely unwise.  It would protect what President Reagan’s former chief economic advisor, Harvard economist Martin Feldstein, has called the biggest area of wasteful spending in the federal budget — what economists call “tax expenditures” and Alan Greenspan has called “tax entitlements.”

In 2015, tax expenditures amounted to $1.2 trillion, more than the cost of Medicare and Medicaid combined (which was $890 billion), Social Security ($882 billion), or defense ($583 billion).  Many of these tax expenditures are fully the equivalent of government spending.  Let me use child care as an example.

If you are low- or moderate-income, you may get a federal subsidy to help cover your child care costs, and the subsidy is provided through a spending program.  If you are higher on the income scale, you still get a government subsidy that reduces your child care costs, but it is delivered through the tax code, as a tax credit.  (Moreover, if you are a low- or moderate-income parent with child care costs, you likely will miss out because the spending programs that provide child care subsidies are not open ended and can only serve as many people as their capped funding allows.  By contrast, if you are a higher-income household — and there is no limit on how high your income can be — your child care subsidy is guaranteed, because the tax subsidy that you get operates as an open-ended entitlement.)  It is difficult to justify making the tax-code subsidy sacrosanct and the program subsidy a deficit-reduction target merely because one is delivered through a “spending” program and the other is delivered through the code.

As the child care example illustrates, sharply distinguishing between subsidies delivered through the tax code and those delivered through programs on the spending side of the budget also has a “reverse Robin Hood” aspect.  Low- and moderate-income households receive most of their government assistance through spending programs; affluent households receive most of their federal subsidies through tax expenditures.  Effectively barring reductions in tax expenditures from contributing to deficit reduction is a prescription for placing the greatest burden of deficit reduction on those who can least afford to bear it.

The problems do not stop there.  If a constitutional amendment requires a two-thirds vote to raise any revenue, another likely outcome is a proliferation of tax loopholes.  New loopholes — including loopholes that Congress did not intend but that high-priced tax lawyers and accountants have found ways to create — could become untouchable once they appeared, because it would require a supermajority of the House and Senate to raise any revenue.  It would become more difficult to close tax loopholes that opened up, since, under S.J. Res. 6, special-interest lobbyists could block such action simply by securing the votes of one-third plus one member in one chamber.

Finally, S.J. Res. 6 and S.J. Res. 2 would bar federal spending from exceeding 18 percent of GDP in the prior calendar year, which translates into a limit of about 16.8 percent of the current fiscal year’s GDP.  This limit applies to the totality of all federal spending, including defense, Social Security, and interest on the debt.  To hit that level would require cuts of a truly draconian nature. 

For example, consider federal expenditures under Ronald Reagan.  Under President Reagan, who secured deep budget cuts at the start of his first term, federal expenditures nonetheless averaged 21.6 percent of GDP.  And that was at a time before any members of the baby boom generation had retired and when health care expenditures throughout the U.S. health care system (including the private sector) were just over half of what they are today as a percent of GDP.  It also was before the September 11 terrorist attacks led policymakers to create a new category of homeland security spending, and before the wars in Iraq and Afghanistan led to increases in veterans’ health-care costs that will endure for a number of decades.

Estimating the Effects of Spending Cap in S.J. Res. 6. and S.J. Res. 2

To provide a more precise and detailed analysis of the impact that the spending cap in S.J. Res. 6 and S.J. Res. 2 would have, we examined what the impact would be if a 16.8-percent-of-GDP spending cap took effect in fiscal year 2023, as would occur if Congress approved the constitutional amendment now and the requisite number of states ratified it by September 30, 2018.  Here are the results.[8]

  • Congress would have to cut all programs (except interest on the debt) by an average of 24 percent in 2023 and 25 percent by 2026.  It would likely have to cut programs by $9.0 trillion through 2026.
  • If all programs were cut by the same percentage, Social Security would be cut $333 billion in 2023 alone and $2.5 trillion through 2026; Medicare would be cut $214 billion in 2023 and $1.6 trillion through 2026; and Medicaid and the Children’s Health Insurance Program (CHIP) would be cut $132 billion in 2023 and $1.0 trillion through 2026.
  • Veterans’ disability payments, compensation, and other such benefits would be cut $54 billion in 2023 and $413 billion through 2026.
  • Defense spending would be cut $154 billion in 2023 and $1.2 trillion through 2026, on top of the reductions made to comply with the discretionary spending caps that the Budget Control Act establishes and the further reductions made through sequestration.

Congress would not, of course, have to cut all programs by the same percentage and likely would not do so.  But if Congress chose to spare certain programs, others would have to be cut even more deeply.  For example, if Social Security were spared, the average cut to all other programs — including Medicare and defense — would rise by more than one-third, from 24 percent in 2023 to 34 percent.  Similarly, if the defense budget (including war costs) were increased by about one-quarter, placing it at 4 percent of GDP and maintaining it at that level, then all other programs — including Social Security — would have to be cut an average of 37 percent in 2023 under S.J. Res. 6 and S.J. Res. 2.  More than one of every three dollars of Social Security benefits, veterans’ benefits, and cancer research would disappear. 

Even if the so-called “plain vanilla” version of the BBA were pursued, rather than S.J. Res. 6 or S.J. Res. 2 with their severe spending limitation, the required level of budget cuts would be massive, assuming taxes are not raised to help balance the budget.  Congress would have to cut everything an average of 19 percent in 2023, an average of 26 percent if Social Security were protected, and an average of 30 percent if the defense budget were set at 4 percent of GDP and Social Security were not protected.


Policymakers need to begin to change our long-term fiscal trajectory.  We should stabilize the debt as a percent of GDP and aim to ultimately bring it down somewhat when economic conditions are conducive.  But establishing a balanced budget amendment in the Constitution would be exceedingly unwise.  It would likely exact a heavy toll on the economy and on American businesses and workers in the years and decades ahead.  It is not the course that the nation should follow.

End Notes

[1] Statement of Robert D. Reischauer before the House Budget Committee, May 6, 1992.

[2] This statement was issued on January 30, 1997.

[3] Federal Service, Transcript of Senate Budget Committee hearing, January 27, 2011.

[4] “Man Up: AJ(obs)A vs. J(obs)TGA,” October 21, 2011,

[5] Some balanced budget amendments require a three-fifths vote of sitting members for a waiver.  Others, including S.J. Res. 6 and S.J. Res. 2, require a two-thirds majority of sitting members for a waiver.

[6] Annie Lowrey, “Balanced Budget Dispute is Fiscal and Philosophical,” New York Times, March 12, 2013,

[7] Norman Ornstein, “Four Really Dumb Ideas That Should Be Avoided,” Roll Call, January 26, 2011.

[8] These estimates are measured relative to the CBPP baseline, which is similar in nearly all respects to the Congressional Budget Office baseline.  For an explanation of the differences, see the technical note in Robert Greenstein, Joel Friedman, and Isaac Shapiro, “Program Spending Historically Low Outside Social Security and Medicare, Projected to Fall Further,” Center on Budget and Policy Priorities, updated February 24, 2016,  Our calculations assume that the program cuts start this year and grow by regular, fixed amounts each year: the 2017 cuts would be twice the 2016 cuts, the 2018 cuts would be three times the 2016 cuts, and so on, until the target is achieved in 2023, the year the constitutional amendment is assumed to take effect.  In subsequent years, the cuts equal the amount needed for the budget to remain on target.  These calculations assume that the target for 2023 and subsequent years is that total spending cannot exceed 16.8 percent of GDP.  Note that by beginning the program cuts in 2016, we allow interest savings to compound over a longer period and so make the size of the program cuts needed in 2023 and subsequent years to meet the constitutional spending limit or balanced budget requirement smaller than they would be if the onset of the cuts were delayed a few years.