Blog Name

Further Thinking on the Costs and Benefits of Deficits

Jason Furman (PIIE) and Lawrence H. Summers (Harvard University)



In recent months the debate over the future of fiscal policy has intensified. For example, Olivier Blanchard, in his presidential address to the American Economic Association this year, highlighted the implications of the empirical fact of sustained low real interest rates for fiscal policy, and others have debated the purported changes in the nature of economic theory itself—most notably modern monetary theory (MMT). In January we made our own contribution to the fiscal policy debate in Foreign Affairs. We offer here our current perspectives on a number of issues that have emerged as most salient in the months of discussion since then.

1. Is the political system so biased towards deficit increases that economists have a responsibility to overemphasize the costs of deficits?

Sometimes the political system is biased towards generating inefficiently large deficits, but other times it is biased in the opposite manner—the role of economists is to share the often nuanced economic reality.

A number of critics of our arguments in Foreign Affairs appear to believe that the political system naturally tends to increase deficits too much out of short-run expediency and in turn believe it is the responsibility of economists to perpetuate the moral principle that deficits are very, very bad. For example, the Washington Post’s Robert Samuelson, whose principled advocacy and analysis we greatly admire, argued “If Washington feared debt, Congress would long ago have tamed budget deficits…. What both Democrats and Republicans actually fear are the highly unpopular steps—spending cuts or tax increases—they might have to take to reduce or eliminate the deficits, which are huge.”

Samuelson is indeed right that the political system can be biased towards greater deficits. But the opposite is often the case as well. Among the largest economic policy errors of the last decade was that the fiscal stimulus was too small in the wake of the Great Recession—that is, deficits were too small, not too large. This is not just an isolated example. In the 1960s and 1970s one of the top concerns of economists at places like the Brookings Institution was “fiscal drag,” or the fact that the system would generate too much deficit reduction. The deficit reduction in the late 1990s was probably excessive, especially in light of the license it gave for unproductive tax cuts. Germany is one of many countries that has also shown that politicians can generate excessive fiscal prudence. Japan has done the same at many points in the 20 years after its financial bubble burst in1989—the most recent example being the self-induced recession caused by raising its consumption tax in 2014.

With interest rates so low that they could limit central banks’ ability to respond in a downturn everywhere in the industrial world, the dangers posed by anti-deficit dogma in the next recession could be enormous.

As a general rule, we think the role of economists is to analyze the economy and not to lean one way or the other to counteract some presumed bias of politicians. The fact that politicians are not always biased towards making deficits too large, and may become more or less biased over time, argues for providing an accurate assessment of the economic realities, not guessing how that assessment should be skewed to counteract perceived political biases.

2. Do the changing economics of deficits mean that anything goes and we do not need to pay any attention to fiscal constraints, as some have inferred from modern monetary theory (MMT)?

No. The costs and benefits of deficits have changed, but we still need a limiting principle in order to conduct fiscal policy.

The economics of deficits have indeed changed in profound ways. Anyone repeating unchanged talking points and policy nostrums over the last several decades has clearly missed out on the fact that real interest rates have fallen from nearly 4 percent in 2000 to around 1 percent today despite the fact that the debt has exploded. This is the result of what one of us has termed secular stagnation and appears to be due to a combination of demographic changes, rising inequality, and declines in the cost of and taste for capital goods.

The idea that we need deficit reduction to keep interest rates down and thus encourage borrowing and investment may have made sense years ago, but it would be an absurd diagnosis of today’s economic problems. In fact, if we had actually kept the budget in balance after 2001, the result, according to research coauthored by one of us, would have been even lower underlying real interest rates, making it likely that the Federal Reserve would have still kept nominal rates at zero—and even then we would not have adequate demand.

When interest rates are less than the growth rate, then the government can perpetually run primary deficits (that is, a deficit excluding interest) and still have debt falling as a share of the economy. It cannot, however, run unlimited primary deficits without the debt rising substantially as a share of the economy.

Fiscal constraints are real. Even if one accepted the MMT idea that inflation is the only constraint on government spending, there is still some constraint on what government can afford. As long as we eventually face a budget constraint, then the government needs to prioritize among the many worthy, and some not so worthy, claims on scarce resources. Without this limit, how can we say no to any sympathetic group or force tax increases on any except the most reviled?

In our Foreign Affairs article we proposed that the political system should adopt a “do no harm” approach, paying for new proposals but not necessarily making it an urgent priority to do any more than that. Adopting this principle would have the benefit of requiring policymakers to think harder about whether to adopt the next seemingly popular tax credit or spending program. Many ideas that seem appealing judged against an unspecified future cost are less appealing when you make their costs explicit today.

3. You advocate doing no harm, but is that enough to stabilize the debt at a reasonable level?

Yes, it likely is enough, if we also do what is widely agreed, which is to eventually close the shortfalls in both Social Security and Medicare hospital insurance and not use the savings to pay for other priorities.

Social Security and Medicare hospital insurance revenues fall short of spending by about 1 percent and nearly 0.5 percent of GDP, respectively, over the next 75 years. The budget forecasts assume that these shortfalls are not remedied and the trust funds in these programs are exhausted but full benefits continue to be paid. The general principle that Social Security solvency should be restored has been widely accepted across the political spectrum, from Senator Bernie Sanders to Senator Ted Cruz. Restoring solvency, and not using the savings to pay for other programs, would reduce the present value of the deficit by about 1.5 percent of GDP—roughly enough to stabilize the debt as a share of GDP at a roughly reasonable level.

As part of this solution, Social Security benefits should in fact be raised for the majority of recipients. Despite the fact that Social Security is an efficient, defined benefit program, its benefits are not generous by global standards and leave too many seniors in or close to poverty. Expanding benefits only works, however, if people are willing to pay more during their working years—meaning that revenue, and not just from high-income households, has to be an important part of the solution, as in the plan advanced by Congressman John Larson. The advantage of this approach is that it would expand aggregate demand, as the additional spending by the elderly would outweigh any reductions in spending associated with higher payroll taxes, in part because the plan could lead households to save less.

Some additional Medicare savings are possible, including through strategies such as more aggressively shifting to advanced payment models, introducing more competition in the Medicare Advantage program, reducing overpayments for prescription drugs, and reforming the structure of Medicare benefits. However, we are not optimistic about how large these savings would be, especially absent broader systemwide healthcare reform, so additional revenue will likely be needed here as well.

To the degree the imbalance between revenue and spending in Social Security and Medicare is eventually reduced—much of it through higher revenue—those savings should not be available to pay for new spending but instead count towards the existing spending we have already promised. This would make the reform effort net deficit reducing compared with the standard projections from the Congressional Budget Office and others that are widely cited. The result, based on current projections, would be to stabilize the debt as a share of the economy at a reasonable level.

We are not motivated by the conventional argument that entitlement spending needs to be contained in order to keep federal deficits under control. Entitlement programs need to be fortified for fundamental social reasons. But a byproduct of rational Social Security and Medicare policy would be to improve the debt outlook, as long as policymakers allow it to do so.

4. Isn’t action on the deficit urgent in order to reduce the risk of a fiscal crisis?

No. Currently the benefits of a reduced probability of a fiscal crisis do not outweigh the costs of deficit reduction. If the underlying conditions change, which is possible, deficit reduction can be done rapidly.

People have been predicting fiscal crises in the United States for decades, and instead the opposite has happened—interest rates have fallen, not risen. Countries that borrow in their own currencies and run independent monetary policies have substantial latitude on fiscal policy. Not unlimited latitude, as some political advocates of MMT argue—just look at when the United Kingdom had to turn to the International Monetary Fund in the 1970s—but much more than indicated by the overwrought predictions of many advocates of more urgent deficit reduction.

It is notable that policymakers routinely tolerate a much greater degree of risk in the financial system: Banking and other financial crises are much more common than fiscal crises. We could limit private debt or require banks to hold much higher levels of capital, but we choose not to because of a (possibly mistaken) view that the costs of these constraints would outweigh any of the benefit of the lowered risks, a point Blanchard makes.

How much would an extra 10 percent of GDP reduction in the debt reduce the probability of a crisis? Probably not much, but even that magnitude of debt reduction would be costly. A much larger amount of debt reduction would reduce risks more but would also be substantially more costly. Advocates of greatly reducing the risk of a fiscal crisis should do much more to quantify both the costs and the benefits of their approach and also explain why they do not apply the same lexicographic preference for risk reduction to the private financial system.

We are not sure about the future. The 90 percent confidence interval on the deficit five years from now is plus or minus 4 percent of GDP. In the good case it would be a mistake to make unnecessarily painful changes now, and in the bad case we can always make the changes later.

5. Do you think anything about fiscal policy is urgent?

Yes. We should be making contingency plans for the next recession and addressing our fundamental problems while paying for them in an efficient and equitable manner.

In an era of secular stagnation the response to the next recession will have to draw heavily on fiscal expansion because conventional monetary policy is limited to cutting rates. We should be actively making contingency plans now for the next recession by identifying infrastructure projects, finding equitable ways to reduce sales taxes, and considering ways of giving tax benefits that maximize the propensity for them to be spent. We should also consider improved automatic stabilizers that would kick in during all future recessions, for example providing aid to states and extending unemployment benefits when they are needed most. It is important to recall that a bad recession can also cause a bigger and more lasting increase in the debt than many of the policies we argue about on a daily basis.

The budget deficit is not our only or even most important national deficit. Those worried about placing burdens on the next generations should be even more concerned with an inadequate infrastructure, an education system that fails most young people, places where employment rates are too low, insufficient investment in scientific leadership, and a government that increasingly lacks the capacity to do fundamental tasks like collecting taxes, caring for veterans, and enforcing the law.

At the same time, while the deficit is not the most urgent problem, there is no reason to make it worse in order to deal with these other problems. Some additional spending reductions could play a role in preventing other policies from adding to the deficit, but additional revenue will have to play a major role.

This path forward represents a reasonable balance between competing priorities and is understandable and politically sustainable. It would also be more fiscally responsible than what we have seen in the last few years. Most importantly, it would better address our even more urgent priorities.

More From

More on This Topic