William Gale wrote: http://www.cato.org/publications/cato-online-forum/get-fiscal-house-order
And now he responds:
Recently, I wrote an article on the role of fiscal policy on economic growth. I argued that, if we want to raise living standards of future generations, a major priority should be reducing the long-term ratio of public debt to GDP. (I also suggested that, since the benefits of higher economic growth disproportionately accrue to high-income households, those households should bear the brunt of the costs of fiscal consolidation.)
In response, Berkeley Economics Professor Brad Delong asked, “Why would anyone seek today to relatively downweight virtually any other economic policy priority in order to focus on the deficit?” At the risk of oversimplifying, Delong offers two classes of reasons for asking his question:
In his view, the fiscal situation is not that bad. Interest rates are low; The 25-year fiscal gap is only 1.2 percent of GDP. Favorable developments are likely to occur. Congress may implement a carbon tax in the next two decades, which would boost revenues. The Affordable Care Act might slow health care cost growth, which would reduce federal spending on Medicare and Medicaid (and ACA subsidies) compared with projections.
Our political system “can’t handle the truth,” as Jack Nicholson would say. The system has had a hard time distinguishing between short-run needs for stimulus and long-run needs for deficit reduction and, because of that, attention to long-term deficits has generated bad policy in the last several years.
I think Delong asks a good question (and a question that other people I respect like Paul Krugman and Henry Aaron also ask. Both of them would add to the list of Delong’s points that fiscal projections are highly uncertain, so why bother paying attention to them). So, I would like to give a fairly complete answer, with apologies in advance for a response that is longer than the original post. My bottom line is:
Projected fiscal shortfalls are a real long-term problem – either society will need to make some difficult choices or growth will suffer. There are three ways to see this:
First, the stated fiscal gap significantly understates the fiscal problem because it aims to maintain the current debt/GDP ratio, which is already the highest in history other than around World War II and is not a reasonable long-term goal.
Second, in the absence of action, debt is projected to rise further.
Third, the current and projected debt levels will significantly reduce long-term growth relative to moving back to historical levels, according to all models I have seen.
Nor are policy developments necessarily likely to be as positive as Delong lays out. I would be extremely surprised if we get a carbon tax as part of a revenue-raising package (but if we do, it will be precisely because people did talk about the need to raise revenue – i.e., the long-term fiscal problem). I am cautiously optimistic about ACA, but I see other policy developments (like games played with revenue projections) and other assumptions (no war, no recessions, etc.) that are more troubling from a fiscal perspective.
The fact that the political system may misuse economic analysis is not a reason to suppress the analysis. It is hard to see how suppression of information will lead to more enlightened policies. Providing economic analysis of long-term fiscal issues does not denigrate other policy needs or priorities; nor does analyzing other policy needs denigrate the need to get the fiscal house in order.
I. The Political System and Long-Term Debt Issues
What is the role of economic analysis in a dysfunctional political system?
I agree that the political system is dysfunctional. It is an unfortunate fact, as well, that Republicans have wanted to use the long-term debt situation as a reason to impose near-term cuts in spending. I would not call that dysfunctional though; I would call that Republicans using whatever arguments they can (just as Democrats sometimes do) to advance their policy agenda.
The distinction between the appropriate responses to short- and long-term deficits gets confused in the public arena. But, in my view, that is not a reason for an economist to ignore the long-term fiscal problem. My job (or at least, the way I see my job) as an economist is to analyze the economics of the situation, including the distinction between the two situations. If the long-term debt path is a problem for long-term growth, I am comfortable saying that. If the political system mangles that message, I will try to clarify it. But not reporting the problem because the political system mangles the message defeats the purpose of having the capability to do the economic analysis in the first place.
For example, I have advocated strongly for the notion that the political system should have been doing more the last few years in terms of stimulus, but – in my view – that does not take away from the fact that there is a projected long-term debt build-up and imbalance that needs to be addressed at some point.
Does the long-term fiscal situation mean we need to downplay current priorities?
Absolutely not. There is nothing about saying that there is a long-term fiscal problem that should take away priority from shorter-term or more immediate policy goals. As I note in the paper, the fiscal situation is not a crisis and the country has the resources to pay our bills “for the foreseeable future.” Nothing in my article says or implies that we should cut current spending, raise current taxes, or ignore or downplay other key issues. Nor do I think we should. My piece addresses long-term growth prospects, not short-term (say, the next 5 years) or even medium-term (5-10 years). The piece is explicitly motivated (in the first sentence) by aiming to “improve living standards of future generations,” so let’s call it a 25-year perspective.
The article does not state or imply that fiscal solutions are the only thing the government should think about in terms of long-term growth, just that they are one major issue. It is not hard to think of other issues that need to be addressed as part of a long-term growth package, including education, infrastructure, immigration, and so on. None of that precludes the notion that fiscal status matters as well.
Does saying that the long-term fiscal situation matters for long-term economic growth mean that we need to slash government spending?
Delong does not explicitly address this issue, but I think it underlies a lot of discussions of this topic, so I will add an answer here. The answer is “absolutely not.” It often seems to me like a lot of the liberal dislike of even talking about the long-term fiscal problem stems from the fear, often unstated, that the only way to deal with the problem is to slash key programs like Social Security and Medicare. I think it is reasonable to think that some spending cuts and reforms will need to be part of the overall fiscal solution, but the notion that we can’t talk about the need for fiscal reform and the long-term fiscal problem because it means undoing key programs is simply wrong.
President Clinton, for example, embraced the idea of saving Social Security first when surpluses started emerging in the late 1990s. That is, one reason to explicitly address, focus on, and emphasize long-term fiscal problems is precisely because one wants to save or enhance the major programs, not because one wants to destroy them. Indeed, under current law, the surest way to induce major cutbacks in Social Security and Medicare is to ignore the problem and hence let the trust funds run out of money. This would require, by law, significant cutbacks, as my colleague Henry Aaron has emphasized. The bottom line: people who want a strong, activist government need to talk about raising the revenue to finance that government.
II. How Bad is the Fiscal Problem
Understanding the fiscal gap
As Delong reports, CBO estimates the fiscal gap is only 1.2 percent over the next 25 years. That means that, just to get the 2039 debt/GDP ratio down to its current level, we would need immediate and permanent tax increases or spending cuts of 1.2 percent of GDP. If “1.2 percent” seems small, think of it as about $200 billion per year, or 7 percent of all tax revenues, or 15 percent of all income tax revenues, or 11 percent of all federal spending other than Social Security, Medicare, and net interest. Cuts of that magnitude are well beyond what the political system will bear. That seems worth mentioning and discussing!
An appropriate debt target and policies needed to reach it
But, in fact, the size of the fiscal problem is probably bigger than the fiscal gap. The fiscal gap shows what it would take to get back to the current debt/GDP ratio by 2039. But the gap calculation is silent on whether the current debt/GDP ratio is healthy as a long-term phenomenon.
By many standards, the current debt/GDP ratio – 74 percent – is too high as a long-term standard. Before getting into this, it is worth noting that one can disagree about what a reasonable long-term debt/GDP ratio is. Economics, embarrassingly, is remarkably weak on determining the absolute magnitude of the optimal debt/GDP ratio. We can tell you it goes up with higher expected population growth or economic growth, it rises as interest rates fall (one reason, I believe, that Delong is less concerned about debt right now, for example) and so on. But convincing models that give convincing estimates of an appropriate quantitative level are few and far between.
Nevertheless, I would argue that least two perspectives tell you that 74 percent is too high. First, over the 50 years before the Great Recession, the ratio averaged just 36 percent. It was 35 percent in 2007. Think about that. After six years of George W. Bush’s tax cuts and increased spending on domestic and military options, the debt/GDP ratio was less than half as large as it is today. We all understand why the ratio ballooned – the recession and efforts to stimulate. Nevertheless, if you (like Delong and I) thought that fiscal policy was on a reckless course under President Bush, it is worth noting that the debt/GDP ratio has more than doubled in the brief time since then. If Bush’s policy was reckless, surely the current situation merits notice and attention.
Another perspective that suggests that a long-term 74 percent debt/GDP ratio is greater than optimal is the behavior of our own country and other countries over time. Countries historically have avoided ratios (of net debt to GDP) this high whenever they could. Wars, depressions, or financial crises are generally the only reasons that countries find themselves in this position. This is also worth thinking about. If it were costless to have a debt/GDP ratio above 70 percent, countries would have been clamoring to raise their debt—after all, they could increase spending or cut taxes and not have to pay for it with compensatory policies.
Yet, advanced countries don’t generally go to those levels voluntarily. For example, in 2007, before the financial crisis hit, only one out of thirty OECD countries (Italy) had net debt as high as the U.S. does now. (To compare consistently across countries, we need to use net debt obligations at all levels of government. By that measure, the U.S. is at 85 percent of GDP.)
Not only is high debt rare, but typically it is short-lasting in healthy economies. In the U. S., for example, we cut the debt/GDP ratio in half in 10-15 years after World War I and World War II. In the current projections, though, debt/GDP rises continually after the first few years. Under the baseline CBO projection, the debt/GDP is projected to rise to 108 percent of GDP by 2040. Alan Auerbach and I calculate the projected debt/GDP ratio at 125 percent of GDP under current policy by 2040.
So, let’s do a calculation equivalent to fiscal gap, but aiming to get down to 36 percent debt/GDP ratio by 2040. This gives us 25 years – rather than the 10-15 it has taken historically – to cut the current debt/GDP ratio roughly in half and restore it to the average level of 50 years before the Great Recession. Auerbach and I show that to meet that goal would require an immediate and permanent policy adjustment of 3.1 percent of GDP – $533 billion per year, a 38 percent increase in income taxes, or a 29 percent cut in non-Social Security, Medicare, or net interest government spending. And if we wait five years to avoid cuts that start while the economy is still recovering, the required policy adjustment would be 3.9 percent, far, far beyond what the political system could bear.
The bottom line here is that getting the debt/GDP ratio back to the mid-30s is going to require difficult changes. But, as the next section shows, leaving it at its current level or letting it rise as projected will hurt growth prospects significantly.
Economic Effects of Fiscal Policy
As noted above, no economic model of which I am aware says that long-term debt does not hurt long-term growth. Illustrative calculations by Greg Mankiw and Douglas Elmendorf suggest that added national debt of 50 percent of GDP reduces net output by more than 3 percent. A study by IMF researchers suggests that a higher initial debt-GDP ratio of 10 percentage points reduces growth in subsequent years by 0.15 percentage points. The Congressional Budget Office has estimated that under their extended baseline projections – where debt rises to 108 percent of GDP by 2040 – GDP will be permanently 3 percent lower by 2040, relative to not having an increase in debt.
All of these estimates suggest substantial negative impacts on long-term economic growth of having the debt/GDP ratio rise from 35 percent in 2007 to 74 percent in 2014, and to between 108 percent and 125 percent over the next 25 years. Just maintaining today’s debt/GDP rather than being at the pre-crisis level means net output is permanently lower by more than 2 percent (Elmendorf-Mankiw), or that the annual growth rate of the economy is reduced by almost 0.6 percentage points (IMF). Allowing the debt/GDP to rise another 40 percentage points, around the midpoint of the range above, would double the impact on growth. That is, the effect of the difference between debt-as-projected and debt-GDP-ratios-in–the mid-30s is a more than 4 percent drop in the level of GDP on a permanent basis relative to what it otherwise would be (Elmendorf-Mankiw) or about a 1.2 percentage point drop in the annual growth rate relative to what it otherwise would be (IMF). These are enormous impacts; surely, they are first-order issues with regard to long-term growth and surely they merit discussion in any analysis of prospects for long-term growth.
The budget and economic projections are marked by significant uncertainty, So, yes, the problem may go away or be reduced, but it may get worse as well. A number of assumptions in the project seem optimistic from a fiscal perspective – no recessions, no wars, no new programs (in fact, sustained cuts in discretionary spending as a share of GDP). If those do not play out as assumed, the situation could end up being significantly worse than advertised. Indeed, increased uncertainty, as Alan Auerbach explains in a recent paper, should heighten the precautionary saving response, precisely because the likelihood of a bad outcome rises.
III. Policy Developments
A. Will we get a carbon tax? Will it raise net revenue?
I would characterize Delong as optimistic that we will get a carbon tax, and although it is not stated explicitly in his piece, he is arguing that we will get a carbon tax as part of a revenue-raising package (otherwise, it would not affect the fiscal situation). I certainly hope that happens. But I have doubts along two dimensions. First, whether we will get a carbon tax at all. I see a Congress that has a lot of Republicans that are either climate-change deniers or have embraced the newly fashionable, disingenuous line of “I am not a scientist…” as an excuse for not dealing with climate change. Second, whether we will get a carbon tax as part of a revenue-raising package.
Currently, more than 90 percent of Republicans in Congress have signed the “No New Taxes” pledge. And, one of the biggest selling points of a carbon tax, indeed perhaps the only selling point in some conservative quarters, is that carbon tax revenue could be used to reduce revenues from other taxes, in particular the corporate tax, rather than raising overall revenues. Finally, a necessary, but surely not sufficient condition, for the U.S. to enact a carbon tax as part of a revenue-raising package is that people talk about and justify the revenue-raising part; that is, talk about the long-term fiscal problem.
B. ACA improvements and health care projections
The recent slowdown in health care cost growth has greatly helped the long-term fiscal situation. Part of that seems due to ACA, with part to other factors. (There has been a global slowdown in health care cost increases, and some of the slowdown occurred before ACA.) Like Delong, I am hopeful that further improvements in health care efficiency can and will happen. I offer two caveats though. First, health care spending growth has slowed in the past, only to pick up again. Second, it is worth noting that the CBO projections for Medicare already have negative excess cost growth assumed for the next 10 years, which may strike people as optimistic. In work earlier this year, Auerbach, Ben Harris and I showed that even if there is no excess cost growth in health care the next 75 years, it would still take immediate and permanent spending cuts or tax increases totaling 1.3 percent of GDP to restore the current debt/GDP ratio in 2040 and it would take cuts of 2.6 percent to get the debt/ratio down to 36 percent by then.
C. Other policy developments
Delong mentions two reasons for what might be called “policy optimism.” While we can be hopeful that ACA will have the desired effects on health care efficiency and a carbon tax will be implemented as part of a revenue-raising package, it is also important to note pressures going the other way – that is, potential policy developments that raise the deficit. On the revenue side, the clearest gimmick would be a tax reform plan that is revenue-neutral in the 10-year budget window, but loses revenue beyond the window. There are numerous ways to bring about such an outcome, and some prominent packages like Rep. Camp’s sweeping tax reform proposals have this feature built in. The issue could be exacerbated if Congress moves to dynamic scoring of major tax bills. A second issue concerns the tax extenders. For some reason, Congress does not feel the need to finance the extenders every year, and that decision would surely dig a further hole into the deficit. Just this year, Congressional action added about $100 billion to future deficits over the next 10 years from a variety of changes, including the CROmnibus.