Why We Should Not Reduce Budget Deficits Now
Remarks delivered at a conference at Princeton University on February 22, 2020
Douglas W. Elmendorf

View the figure accompanying these remarks.

I am delighted to be back at Princeton, from which I graduated many years ago, and delighted to be on this panel with so many friends.

The title of the panel asks us whether we should stop worrying and love the deficit. I do not think we should love deficits, but I do think we should worry much less now than most of us worried five and ten and twenty years ago. I want to explain why we should worry much less now, and then explain the implications that I draw for budget policy. The principal implication, as you will see, is that we should not reduce budget deficits now, even though we will need to reduce them eventually.

My comments draw on joint research with Louise Sheiner four years ago and ongoing research with Karen Dynan.

Interest Rates Have Fallen Sharply

Let me begin. When I became director of the Congressional Budget Office 11 years ago, interest rates had fallen markedly over the preceding decade. The figure below shows the yield on 10-year Treasury notes after adjusting for inflation, presented as a 5-year trailing moving average to smooth through the year-to-year fluctuations. As you can see, the inflation-adjusted 10-year Treasury rate had dropped from more than 3½ percent during the 1990s to about 1½ percent on the eve of the financial crisis. Nonetheless, CBO was projecting that the rate would quickly rebound and stay at or above 3 percent for the coming decades.

In fact, interest rates fell further, even after the economy recovered from the financial crisis and severe recession. By the time I left CBO 5 years ago, the agency was projecting a more modest rebound in rates, to a little under 2½ percent. However, interest rates did not rebound and instead fell further. As we speak today, this measure of inflation-adjusted borrowing costs is roughly zero. So, CBO has lowered its projection yet again. CBO now expects that the inflation-adjusted 10-year government interest rate will stay below ½ percent for the next 5 years and below 1½ percent for the next 15 years.

This outlook is not idiosyncratic to CBO: It is quite consistent with readings in financial markets, where the nominal 30-year Treasury yield is only about 2 percent. And this situation is not unique to the United States: Interest rates on government debt have fallen in many countries over the past few decades, notwithstanding a surge in government debt in many countries, which would be expected to boost rates. Last summer, one quarter of all government debt worldwide traded at negative nominal interest rates; I have not seen more recent figures, but I doubt that fraction has declined.

The reduction in the cost of government borrowing is a dramatic change in the economic backdrop for decisions about budget policy. Inflation-adjusted interest rates that were expected to be 3 percent are instead running near zero.

Lower Interest Rates Imply that Government Borrowing is Less Costly for the Economy

It is well understood that low interest rates change the dynamics of federal debt: Lower rates imply lower interest payments, so for any given amounts of revenue and non-interest spending, deficits will be lower and debt will rise more slowly.

It is less well understood that low interest rates change the desirable levels of deficits and debt—but analysis suggests that they do. Analyzing the implications of low interest rates for the desirable amount of debt is complicated in part because the implications differ depending on the factors that have reduced interest rates, and there are many different factors that may be playing a role. Louise Sheiner and I presented our analysis in a published paper in the Journal of Economic Perspectives a few years ago and in a more technical, unpublished version available on the Brookings website. Our bottom line was: “Many—though not all—of the factors that may be contributing to the historically low level of interest rates imply that … federal debt … should be substantially larger than [it] would be otherwise.” Of course, our analysis is hardly the final word, but other economists are reaching similar conclusions, and I have not seen any analyses finding that low interest rates have no implications or have different implications for the optimal amount of debt.

Moreover, the intuition for why lower rates imply that the desirable amount of debt is larger is straightforward: The most common economic cost attributed to high debt is that the government’s demand for loanable funds crowds out private borrowers who would have used those funds for capital investments that would boost future growth. But low interest rates suggest that loanable funds are not very scarce and that private investments with even modest returns can be readily funded—so the government’s use of borrowed funds is not very costly to society. The next-most common economic concern about high debt is that investors might suddenly worry that the debt will not be honored, and then try to sell their holdings in large quantities and push up interest rates sharply. But low interest rates suggest that investors are not very worried about that possibility.

To be clear, I am not arguing that debt can rise indefinitely relative to GDP, which is what would happen under current laws. We will ultimately need to change laws to raise taxes and reduce spending. I am also not arguing that more debt has no economic cost—although research by economists may ultimately reach that conclusion. Instead, I am making a more modest claim, that more debt has less economic cost when interest rates are low.

With interest rates especially low today and expected to stay low for years to come, the optimal amount of federal debt is larger now than in previous decades. Therefore, there is less urgency to putting federal debt on a sustainable path than I and many other economists have argued in the past.

Still, one might ask, since we will need to put debt on a sustainable path eventually, should we not start now?

We Should Delay Reducing Budget Deficits

My answer to that question is no, we should not reduce budget deficits now. I will offer three reasons—one purely economic, and two that mix economics and politics.

First, reducing deficits now could cause a recession and would certainly make responding to other recessionary forces more difficult. Increases in taxes or reductions in government spending would slow the economy in the short run. Under the economic conditions we have experienced during most of our professional lives, that contractionary impulse would be offset by reductions in the funds rate by the Federal Reserve. But with the funds rate already under 2 percent, the Fed has little room to cut further. And if negative shocks hit the economy, the Fed will need all of the room to cut the funds rate that it currently has, and then some. Therefore, reducing budget deficits today would present a significant risk of reducing output and employment, with all of the attendant economic and social costs.

Second, the policymaking process has limited bandwidth. Policymakers cannot address all worthy issues at once; in recent years, the Congress has struggled just to pass appropriations bills and confirm presidential nominees. So, good policymaking is about prioritization, not just offering a wish list. If our biggest public problem was overall economic growth being constrained by lack of funds for private investment, then tackling budget deficits would be appropriate. But, in fact, we have many public problems more pressing than a lack of funds for private investment, such as the economic and social crisis confronting Americans without college degrees, uncontrolled costs for healthcare, and the emerging effects of climate change. We should focus our policymaking attention on those issues and others. 

Third, the specific actions that probably would be taken to reduce deficits now would cause more harm than good, even apart from the risk of recession. When national attention last focused on budget deficits, in 2011, the outcome was limits on discretionary spending, not the changes in taxes and benefit programs that people often recommend in gatherings like this. Even though those spending limits were raised in subsequent pieces of legislation, federal investments in infrastructure and in research and development are now lower relative to GDP than at almost any point in my lifetime. That is hardly helpful for future generations. Moreover, many of the budget plans offered by the current presidential administration or offered by House Republicans over the past decade would reduce government benefits for lower-income Americans, even though wage gains for Americans without college degrees have been very limited for decades, and a growing body of evidence suggests that government benefits enhance the life prospects of children who receive those benefits. Worsening the economic circumstances of Americans who already are being left behind is not fair and is not helpful for future generations.

In sum, I think that reducing budget deficits now would be counterproductive in both economic and social terms.

We Should Also Avoid Increasing Budget Deficits, Except to Respond to Recessions

One might ask, then, what sort of fiscal policy do I recommend?

First, although I would not reduce budget deficits now, I would also avoid increasing budget deficits, except in response to recessions. We are in a deep long-term hole with the federal budget, and I think we should mostly stop digging. In my view, any policies that would reduce taxes or increase spending should be accompanied by other policies with offsetting effects—what is known in the jargon of Washington budget policy as a “paygo” approach. This approach also has the advantage of reminding people that policy choices have tradeoffs.

But making an exception for economic downturns is crucial. With monetary policy constrained by low interest rates, countercyclical fiscal policy will be especially important when the next recession hits, and we should undertake such policy even if government debt increases substantially as a result. In particular: We should strengthen the automatic fiscal stabilizers now, and I have been developing some proposals in this line with Karen Dynan. We should also implement discretionary increases in spending and cuts in taxes when the next recession arrives. And we should keep vigorous countercyclical policy going until the economy has really recovered and not just started to improve again.

Second, I would lengthen the maturity of outstanding government debt as a hedge against increases in interest rates. Because interest rates might rise more than currently expected, we should lock in lower rates to a greater extent by issuing more debt with long maturities. Larry Ball, Greg Mankiw, and I published a paper 25 years ago in which we analyzed the implications for budget policy of the risk that interest rates could rise unexpectedly and economic growth fall unexpectedly, making a debt that had appeared to be sustainable suddenly seem unsustainable. The paper was titled “The Deficit Gamble.” We noted that Treasury interest rates have been low historically because they provide insurance against damaging, low-probability events, and we concluded that sensible policymakers should take out insurance against the risk of a debt becoming unmanageable. Lengthening the maturity of federal debt is one way to take out such insurance.

Let me conclude here. I do not think we should love deficits, but I do think we should worry much less now than in the past, because we have experienced a dramatic, long-standing, global decline in interest rates. That decline implies that the economic costs of government debt are much less than economists have long believed, and therefore that putting debt on a sustainable path is much less urgent. Moreover, we have significant competing demands on our attention and the public fisc. Therefore, we should not reduce budget deficits now, but we should hedge our bets against future increases in interest rates by lengthening the maturity of government debt and avoiding policies that would increase debt except in recessions. 

I look forward to the discussion. Thank you very much.