The Senate Budget Committee recently reported legislation co-sponsored by Chairman Mike Enzi and Committee member Sheldon Whitehouse that would alter the congressional budget process in highly problematic ways. Although the bill includes positive elements (see box), it also includes a significant land mine: a requirement that Congress craft and consider deficit-reduction legislation under expedited “reconciliation” procedures if the Congressional Budget Office (CBO) finds that the debt projection for the final year that the congressional budget resolution covers is higher than that reflected in the budget resolution. Such a CBO finding would very likely occur, history shows, and it could drive deep cuts in mandatory programs such as Medicare, Medicaid, health insurance subsidies, SNAP, and Supplemental Security Income (SSI) — and at a time when the economy is weak, thus making the economy still weaker.
Enzi-Whitehouse (S. 2765) would shift the congressional budget resolution from an annual document to one that covers two years. Under the bill, Congress would approve a budget resolution in the first year of a two-year session. Then, on February 15 of the second year, CBO would calculate whether the debt ratio — the debt held by the public as a share of the economy (gross domestic product, or GDP) — for the final year of the budget resolution would exceed the debt ratio projected for that year in the budget resolution that Congress had adopted a year earlier. Such a CBO calculation would trigger a new, automatic deficit-reduction process. That new, automatic process, which would likely be even more expedited than the current reconciliation process on which it’s modeled, differs from that process in important ways and has major flaws.
Tying reconciliation to debt estimates that can vary widely from year to year is ill-advised for several reasons. It likely would drive the public debate toward hitting an ever-moving debt-ratio figure and away from using the budget to address key national issues. Most worrisome, the new, automatic deficit-reduction process would likely increase the chances that policymakers would make budget cuts (thus, imposing austerity) at times when the economy is weak or in recession, which would further weaken the economy. In addition, the new process would confer some new powers on the Senate Budget Committee in developing important fiscal policies at the expense of other committees, particularly the authorizing committees.[1]
Enzi-Whitehouse also would retain the existing reconciliation process, prompting the question of why a second one is needed.
Some elements of Enzi-Whitehouse could prove helpful, most notably a provision that could reduce dangerous brinksmanship around raising the statutory debt limit. The bill provides that, if Congress agrees to a budget resolution, legislation raising the debt limit to the level in that resolution would be deemed to have passed Congress and would go directly to the President, avoiding separate House and Senate votes solely on the debt limit.
Known as the “Gephardt procedure” (after the former Democratic majority leader who spearheaded a similar rule in the House), this provision would substantially reduce the risks of a harmful federal default and of artificial crises in which a minority threatened to force a default to obtain concessions on other issues. An even better approach would be to repeal the debt limita or, as a fallback, to suspend the debt limit while an approved budget resolution is in effect.
The Enzi-Whitehouse bill, as amended in committee, also affirms a Government Accountability Office ruling that the Impoundment Control Act does not authorize the President to withhold funding that Congress has appropriated until the time period covered by the appropriation ends and the appropriation lapses. The bill imposes penalties on officials who allow that to happen. In addition, it requires the Office of Management and Budget (OMB) to promptly report each apportionment of an appropriation that it makes or that an agency requests it to make.b
The bill also would move budget resolutions to a biennial schedule, which could be a more workable approach to designing budget plans, while retaining the role of annual appropriations bills to fund the government every year. While two-year budgets could provide more funding stability, their success would depend on whether they provide adequate funding in both years, not just the first.
The process that would trigger a new, automatic reconciliation process could force budget cuts that are disconnected from a solid policy rationale. The projected final-year debt ratio might exceed the level projected in the budget resolution:
- Because, even though the debt ratio is falling, it’s not falling as fast as the budget resolution assumed; or
- Because the budget resolution made overly optimistic assumptions about future economic growth; or
- Because Congress rejected a deficit-reduction plan of spending cuts and/or tax increases the previous year or the President vetoed it; or
- Because the budget resolution assumed “magic asterisk” budget cuts Congress never intended to implement, simply to show some level of deficits or debt that sounded appealing; or
- Because the Federal Reserve set interest rates somewhat higher than the budget resolution had assumed, increasing interest costs and thereby leading to higher deficits and debt; or
- Because of purely technical reasons related to underlying demographic and other assumptions about current-law spending and revenue levels, unrelated to any changes in policies that affect projected deficits. Stated another way, even if Congress made every change in spending and tax policy assumed in its budget resolution, CBO might still subsequently project a final-year debt-ratio level higher than the level that the budget resolution projected for that year.[2]
What programs are at risk? Congress can use reconciliation to change revenue and mandatory programs, except Social Security. The largest mandatory programs, more or less in order, are Medicare, Medicaid and the Children’s Health Insurance Program (CHIP), refundable tax credits (the Earned Income Tax Credit, the Child Tax Credit, and the Affordable Care Act’s premium tax credits), civil service retirement, veterans’ compensation and pensions, SNAP, military retirement, SSI, family support and foster care, farm supports and crop insurance, and child nutrition.
(Note: While policymakers cannot change Social Security in a reconciliation bill, the level of debt held by the public is driven by the entire budget, including Social Security, even though Social Security is designated by law an “off-budget” federal agency. Therefore, debt ratio projections also will vary from earlier expectations if Social Security cost or revenue projections change. As a result, under auto-reconciliation, other mandatory programs may have to be cut to make up for, say, a decrease in projected Social Security revenues.)
Of major concern, the media and the public may interpret a CBO finding that the projected debt ratio for the final year will be exceeded as an official, nonpartisan determination that Congress didn’t meet its budget commitments, and that policymakers now should reduce the deficit by the excess amount that CBO calculates. But that’s not what the CBO report would mean. More likely, it would mean that the budget resolution was overly optimistic, assumed “magic asterisk” cuts Congress never intended to enact, or both. To be sure, the bill gives the Senate Budget Committee and Congress the authority to set a different amount of deficit reduction (smaller or larger) than CBO projects, but the media and public will presume that the CBO estimate would be the most appropriate amount. Furthermore, they will presume that policymakers need to reduce the deficit even if current economic circumstances suggest the opposite, as discussed below.
Were this process in place for the last four budget resolutions that Congress adopted, for fiscal years 2010, 2016, 2017, and 2018,[3]the last three would have generated a CBO calculation projecting a tenth-year debt ratio above those in the budget resolutions, thereby triggering the auto-reconciliation process in each case.[4]
Under Enzi-Whitehouse, CBO is also required to estimate the size of the policy savings — mandatory program cuts or revenue increases — needed over the multi-year period of a budget resolution to hit its final-year debt ratios.[5] Those CBO reports would have shown the following amounts of budget cuts for a ten-year auto-reconciliation bill:
- $3.0 trillion for the 2018 budget resolution;
- $1.0 trillion for the 2017 budget resolution; and
- $5.5 trillion for the 2016 budget resolution.
The 2018 budget resolution is particularly instructive. It paved the way for the 2017 tax cut legislation, which increased the deficit by an estimated $1.9 trillion over ten years.[6] Under Enzi-Whitehouse, just two months after policymakers enacted the tax cuts, CBO would have estimated that the debt ratio in the resolution was being exceeded and that $3.0 trillion of deficit reduction would be required to meet that ratio.[7] That would have triggered the automatic process to reduce the deficit. In designing a deficit-reduction package, the Republican majority surely would have protected the new tax cuts, instead designing a package solely of cuts to mandatory programs. This approach would have exacerbated the inequities in the new tax law, which provided the overwhelming share of its tax cuts to the nation’s wealthiest households and most profitable corporations.[8]
Moreover, even though CBO’s projections are highly credible, they are — like all budget projections — subject to substantial variation that, under Enzi-Whitehouse, could trigger the automatic deficit-reduction process. CBO recently assessed its past projections of deficits and debt,[9] measuring the accuracy of each of its five-year debt projections since 1984, excluding the effects of legislation. While that’s not the same calculation that Enzi-Whitehouse would require, it highlights the impact of shifting economic and technical estimating assumptions. Although CBO found little or no bias in its projections, some projections were far too high and some were far too low. Its fifth-year debt projections missed the mark by an average of 7.1 percent of GDP. Five years from now, that would equal $1.8 trillion. And ten-year projections would have been off by substantially larger amounts. Nevertheless, CBO found that, since 1984, its multi-year economic forecasts have been as good as the forecasts of OMB or the Blue Chip consensus, showing that all such forecasts are inherently quite uncertain.
Most worrisome, the final-year debt ratio in a budget resolution might be exceeded mainly, if not exclusively, because the expected pace of future economic growth slowed in the year after Congress adopted the budget resolution. Periods of slower economic growth are exactly the wrong time to enact new budget cuts because they would make a weak economy still weaker and make recessions that are underway deeper.
Potentially, the Senate Budget Committee could address this challenge in three ways, but none seems realistic:
- The Committee could set the final-year debt ratio in the budget resolution much higher than anticipated given its policy, economic, and technical assumptions,[10] thereby reducing the risk that CBO would calculate in the following year that the debt ratio would be breached. But setting higher levels of debt in the budget resolution seems quite unlikely; indeed, the far likelier scenario, given the history of recent budget resolutions, is that Congress will be overly optimistic in its “out-year” projections in order to make the debt problem seem less serious than it is — thereby increasing the chances that a large, economically ill-advised deficit-reduction package would be triggered as the economy is slowing.
- If the final-year debt ratio is on track to be breached, the Committee or Congress could ignore the CBO estimate of the amount of deficit reduction needed to hit the assumed debt-ratio level and instead set a very low auto-reconciliation deficit-reduction target, such as $1. But deficit “hawks” would almost certainly deride such an approach as irresponsible gimmickry, even when it would be the most prudent economic policy. The public debate would likely fixate on reports about gimmickry, when instead it should focus on whether (or how much) economic stimulus (tax cuts and/or spending increases) policymakers should enact to strengthen a weakening economy, or at least the degree to which austerity would harm a weakening economy.
- If the Committee did decide to require deficit reduction in the full amount that CBO estimated at a time when the economy was weak, it might include year-by-year deficit reduction targets rather than just a single total over the entire multi-year period. And if so, it should call for all the deficit reduction to occur at the tail end of the period, when the economy might have fully recovered.
Agreeing to any of these approaches would diminish the economic harm caused by the proposal, but because these approaches are not likely to be politically realistic, this aspect of Enzi-Whitehouse represents unsound economic policy. Indeed, policymakers ought to amend the budget process to increase Congress’s tools to make appropriate fiscal policy when the economy weakens, rather than establishing automatic processes that can push Congress to make unwise policy — and place Congress in the unfortunate position of needing to circumvent those processes to make sound fiscal policy when the economy falters.
The bill’s automatic deficit-reduction process would also give the Senate Budget Committee some new powers.[11] When triggered, the process would initially charge the Committee with deciding how much deficit reduction to mandate and which Senate committees must make those program cuts (or, in the case of the Finance Committee, revenue increases). The initial Budget Committee allocation of program cuts would then go to the Senate floor, where it could be amended or defeated. In this respect, the process is similar to the existing budget process. But there are two key differences.
First, under the existing budget process, reconciliation is an optional part of an overall budget plan, while under Enzi-Whitehouse, initiating reconciliation is the entire purpose. Second, after a new reconciliation directive had been agreed to, if any Senate committee then fell short of its target, the Budget Committee would have the unprecedented authority to increase that committee’s cuts or draft new ones to meet the target, and those proposals would be automatically included in the auto-reconciliation bill that went to the full Senate. As under current reconciliation procedures, the Senate could amend or defeat the bill, but it could not reduce the amount of savings.
The question of what’s sound budget policy should not be driven by a presumption that austerity is always the right answer, nor by a mechanical calculation that could trigger austerity at the wrong time in an economic cycle. It should not be driven by uncertain projections of a moving target that may not indicate whether Congress is actually adhering to the budget policies it has set for itself.
In this analysis we discuss reasons why CBO might subsequently project that the debt ratio for the final year covered by a budget resolution that Congress approved the previous year might be higher than the resolution had assumed, triggering the automatic reconciliation process. There is an additional reason: in measuring the debt ratio, the Enzi-Whitehouse proposal relies on the federal government’s standard measure of its debt, which doesn’t fully account for the government’s financial position and could actually trigger the automatic reconciliation process even when no deterioration in the government’s position has occurred, and in some cases, even if deficits now are expected to be lower than the budget resolution assumed.
Here’s why:
The proposed automatic reconciliation process focuses on the estimates of “debt held by the public” — the debt from government borrowing in the credit markets — that are shown in the final year of a budget resolution (e.g., the tenth year of a ten-year resolution). Debt rises whenever the federal government runs a deficit, so a budget resolution with a decade of deficits will increase the debt by the sum of those ten deficits. But the federal government also has financial assets, such as cash, gold, and direct loan assets, that somewhat offset the debt.
Suppose, for example, that the Treasury Department borrows more money solely to build up its cash reserves that have fallen dangerously low, threatening its ability to pay federal bills on time. In that case:
- the federal government’s financial position would be unchanged (the debt would be higher, but the increase in debt would be offset dollar for dollar by an increase in cash); and
- the additional borrowing would not show up on the federal books either as an increase in spending or an increase in deficits.
So sometimes an increase in debt doesn’t reflect any deterioration in the government’s financial position. That is, the growth of debt over some period of time is often not simply the sum of deficits over that period. Debt can grow significantly more or less than the sum of deficits. In 2015, in fact, the deficit was $442 billion but the debt rose by only $337 billion. In 2016, by contrast, the deficit was $585 billion but the debt rose by $1.1 trillion.
As CBO has written, “Other factors — collectively labeled ‘other means of financing’ and not directly included in the budget totals — also affect the government’s debt. Those factors include the cash flows associated with federal credit programs, such as student loans, as well as changes in the government’s cash balances.”
Now, let’s look more closely at student loans, which highlight the potentially large policy implications of the automatic reconciliation process. Over the ten years from 2008 through 2017, federal deficits totaled $8.4 trillion. Over the same period, however, debt held by the public grew by $9.6 trillion. The overwhelming reason for the $1.2 trillion difference was the government’s 2010 conversion of the federal student loan program from loan guarantees to direct loans. Under the loan guarantee program, the financing had taken place off the government’s books, with banks lending money to students, the students paying back the banks over time, and the government responsible for defaults. Under the direct loan program, the financing takes place on the government’s books: the Treasury borrows from the credit markets and lends the money to students. The government’s costs over time are essentially the same either way, arising from defaults — but with direct loans, the government’s books show higher debt because the government borrowed from the credit markets to finance direct loans, and the books also show corresponding larger financial assets, which are the value of those direct loans to the government.
The Enzi-Whitehouse automatic reconciliation trigger accounts for only half of such transactions (the government’s additional borrowing), while missing the other half (the government’s acquisition of direct loan assets, when students repay the government). Such accounting would have missed more than $1 trillion in student loan assets that the government has acquired.[12] Indeed, when CBO estimated the budgetary effects of the 2010 conversion from guaranteed to direct loans, it concluded that the legislation would reduce spending and deficits in every year even though it would increase debt held by the public.
This aspect of Enzi-Whitehouse could prompt Congress to favor loan guarantees over direct loans, even when the former produce modestly higher deficits over time by requiring extra government payments to banks to induce banks to participate in the program.