Putting the Debt on a Downward Path

Yesterday, President Obama suggested we need about $1.5 trillion in deficit reduction on top of what has been enacted so far, a claim which matches a recent analysis from the Center on Budget and Policy Priorities showing $1.4 trillion as sufficient to stabilize the debt. As we gear up for another round of budget negotiations, we'd encourage lawmakers to raise the bar and put the debt on a clear, downward path relative to the economy.

Although we agree with CBPP's analysis that $1.4 trillion would be sufficient to keep the debt stable through 2022 -- at least relative to their baseline which is quite similar to the CRFB Realistic Baseline -- we believe this amount of deficit reduction would fall short of what is needed in the long term. To be sure, $1.4 trillion of deficit reduction would be a very welcomed package of savings, but would almost certainly need to be followed with additional deficit reduction in order to put the debt on a truly sustainable path.

Among our specific concerns are:

1. A plan that is projected to just barely stabilize the debt leaves no room for error. Slower economic growth, higher interest rates, or new deficit-increasing legislation could all push debt back on an upward path.
2. Simply stabilizing the debt over ten years would likely not be sufficient to keep it stable beyond the ten year window. Given budgetary pressures from health care, aging, and rising interest payments, $1.4 trillion is highly unlikely to stop debt from rising in the early to mid-2020s, let alone over the longer-term.
3. 73 percent of GDP may be too high of a steady-state level for the debt. That level is higher than the international standard of 60 percent, and there would be little fiscal flexibility to accommodate economic or national security emergencies.

No Margin For Error

Aiming for the debt-to-GDP ratio to be exactly stable carries with it tremendous risks, because both economic and technical projections may be off and because future policies may worsen the situation.

On the former point, the Congressional Budget Office (CBO) bases all of its projections on specific economic forecasts and assumptions about GDP growth, inflation, interest rates, immigration, health care cost growth, unemployment, and other factors. For example, if the economy recovers more slowly (CBO projects a full recovery by 2018), steady-state growth is lower (CBO projects about 2.3 percent real growth), or interest rates are higher (CBO projects a 5 percent interest rate on a 10-year note by the end of the decade), debt projections could worsen.

As an illustrative example, we estimated the effects of average annual GDP growth being about a quarter of a point lower than CBO's projections each year, which would simultaneously increase the numerator and decrease the denominator in the debt-to-GDP ratio. Rather than stabilizing the debt at 73 percent, under this scenario a $1.4 trillion deficit reduction plan would put the debt on an upward path nearing 77 percent by 2022.

^{Source: CRFB calculations based on CBO data.}

To be clear, we do not believe GDP will grow a quarter point slower every year than CBO's projections, and there is a similar probability that the economy could grow a quarter point faster. However, the consequences of slower-than-projected economic growth, or higher-than-projected interest rates, are real enough to be taken seriously and reason enough to leave a margin of error to ensure stable debt levels.

This is especially true when considering the second type of projection error: a policy projection error. Both CBPP and CRFB rely on "current policy" baselines which create a base case projection of how Congress will act. In reality, however, it is likely Congress will enact more deficit-increasing policies than what are even in those baselines. CBPP points out one case -- the "tax extenders" -- where Congress has tended to resort to deficit financing; and they rightfully point out that if the extenders are not dealt with as part of a debt deal (for example, through tax reform) then as much as $400 billion in additional savings would be needed. Other similar examples abound as well.

Extending bonus depreciation (described here), a temporary measure now in its sixth year, would have substantial costs as would, to a lesser extent, continuing extended unemployment benefits beyond this year. On top of that, funding for Hurricane Sandy disaster relief, an extension of the lower Stafford loan interest rate beyond June of this year, the extension of higher Medicaid payments for primary care physician payments beyond 2014, the small bucket of "health extenders" outside of the doc fix, and any of a number of policies we cannot yet predict could all add to the deficit. A strict adherence to pay-as-you-go rules -- which both CRFB and CBPP support -- would reduce the effect of these policies on the debt. But so long as one year of costs is paid for over ten years of savings, deficit effects would not be eliminated entirely.

No Long-Term Stability

Even if $1.4 trillion proves sufficient to stabilize the debt through 2022, it is unlikely to stabilize debt in the second decade or beyond. With a debt stock of 73 percent of GDP ($18 trillion in 2022), interest alone will cost the government 2.9 percent of GDP ($700 billion) per year by 2022. By our estimates, even if the primary deficit (programmatic spending minus revenue) were held constant as a percent of GDP after 2022, debt would continue to grow relative to the economy due in large part to these interest payments. More importantly, the combination of population aging and health care cost growth is projected to put substantial upward pressure on the costs of Social Security, Medicare, and Medicaid while slowing the upward pressure on revenue.

Addressing these realities will require something of a "running start" in order to keep interest payments at bay and make the necessary changes to "bend the health care cost curve" and slow the growth of entitlement costs in a way that is sufficiently gradual to allow workers to plan and protect those already in retirement from dramatic benefit changes.

To demonstrate why a $1.4 trillion plan is unlikely to be sufficient, we put an illustrative plan into our long-term budget model. Specifically, we took CBPP's $1.191 trillion of primary savings ($185 billion in 2022) and $177 billion of interest savings ($51 billion in 2022) and assumed the primary savings is split equally between revenue and spending (CBPP's "Scenario B"). For illustrative purposes, we assume all of the health and other mandatory savings from the President's budget, with additional spending reductions coming from discretionary spending.¹

As the graph below demonstrates, this deficit reduction would be sufficient to keep the debt stable only through 2023 after which it would begin to rise again -- reaching 94 percent of GDP in 2035 based on CRFB's Realistic Baseline. Although the savings grow over time, particularly the interest savings, they do not grow fast enough to combat the effects of health care cost growth and population aging. In other words, we should enact more savings in order to get out in front of these effects.

^{Source: CRFB calculations}

To be sure, not all $1.4 trillion is created equally. A plan which included substantial long-term controls to Social Security, Medicare, and Medicaid could indeed keep the debt stable over the long-run. However, it is difficult (though not impossible) to imagine a package with only $1.4 trillion in savings this decade creating enough running room to allow for these changes, to the extent such changes would require the type of political pain policymakers would only endure in a "Go Big" scenario. As we've said before, putting everything on the table for consideration in a "Go Big" approach can improve the chances of actually controlling debt by providing the political tradeoffs necessary to generate savings and reforms sufficient to control the debt. 

CBPP rightfully points out that we don't currently know all the ways to control health care cost growth, and learning more about what works and doesn't work over the next several years can give us a better idea about controlling costs over the long term. However, this is no excuse for waiting to enact those policies that we know, or at least have significant evidence to believe, could help control costs now without causing unnecessary damage. In December, CRFB put forward nearly 100 options to reduce health spending, including policies recommended by MedPAC, the White House, House Republicans, the Simpson-Bowles Commission, the Dominici-Rivlin plan, the American Enterprise Institute, the Center for American Progress, and the National Coalition on Health Care. Many of these policies could be enacted today, with further reforms following in future years as we learn more about the health care system.

Policymakers must take action soon to control costs over the long term because many health care and Social Security reforms take time to phase in. But those savings compound over time. Waiting to make those changes further delays that process, but also may reduce total savings if more and more current or future beneficiaries are exempt from certain changes.

No Fiscal Flexibility

Even if a $1.4 trillion plan were designed in a way that stabilized the debt over the long run, it is not clear that stopping at 73 percent of GDP would be low enough. While this is below what Reinhardt and Rogoff found to be the "danger zone" of 90 percent of GDP, it is twice the nation's historical average and well above the international standard of 60 percent.² Higher debt levels leave the country vulnerable, both budgetarily and economically.

Would debt at 73 percent of GDP be a disaster? Probably not. In fact, it is about that right now. But as the economy recovers that level would lead to more "crowding out" of investment than would a lower debt level. More importantly, it would leave little flexibility for further debt accumulation in the case of an economic or national security crisis. Since our debt was 36 percent of GDP prior to the previous financial crisis, it gave the government substantial room to rescue the financial industry, inject stimulus into the economy, provide relief for those out of work, and avoid immediate austerity. It is unclear whether that same flexibility would have been available had debt levels been twice as high as they were.

Policymakers may not bring the debt down below 60 percent in the next couple of years or even in the next decade. But so long as our debt is at elevated levels, they should be working to put it on a clear downward path relative to the economy.

By reducing rather than merely stabilizing the debt, policymakers can help to ensure fiscal sustainability for future generations.

^{1 Longer-term calculations are based on the long-term CRFB Realistic baseline, adjusted for changes in the CBO August baseline and the American Taxpayer Relief Act. From 2014 through 2022, we assume year by year primary and interest savings as specified by CBPP. We assume that in each year in the ten-year window, half of the primary savings will come from revenue and generated in a way that does not fundamentally slow or accelerate the growth trend of revenue. We assume that in each year in the ten-year window, the other half of the primary savings comes from spending reductions. For illustrative purposes, we use the President’s budget as a starting point for where CBPP might assume those savings are generated. Specifically, we assume $325 billion of health savings reflecting CBO’s updated year-by-year estimates of the President’s health policies, updated for timing and legislative changes. We also assume roughly $175 billion of other mandatory savings based on the President’s budget. Finally, we assume any remaining spending reductions to hit CBPP’s target come from the discretionary side of the budget. Beyond 2022, we generally rely on our long-term model to extrapolate savings. We assume additional revenue grows at the same rate as baseline revenue in light of real bracket creep and other structural forces. We assume discretionary and mandatory savings, like discretionary and mandatory spending, are held constant as a share of GDP. In general, we assume that health care savings grow over time in line with the growth in health care spending. In three cases, we assume health care savings grow faster than health care spending and roughly extrapolate those savings based on our own calculations. Specifically, we assume that the President’s proposal to slow the growth rate of post-acute care payments will result in compounding savings over time that will exceed health care cost growth; we assume that proposals to change cost-sharing rules for new beneficiaries only will increase with the growth of both health spending and new beneficiaries; and we assume the proposal to increase income-relating for Part B and D premiums and freeze the income-related thresholds until they reach 25 percent of beneficiaries will grow rapidly due to a combination of rising per-capita costs and an increasing number of beneficiaries subject to the premiums. Finally, we assume that interest savings will grow as a result of the lower debt stock, relative to our baseline, in each year through 2035.}

^{2 In Reinhardt and Rogoff's study, the data looks at gross debt across countries for comparability. However, international comparisons of gross debt more closely resemble public debt for the United States.}