After the election campaign, the nation likely will turn in one way, shape or form to dealing with the budget. Several analysts and bipartisan groups have had their say on what the ultimate plan should be. Among those statements is a paper by Andrew G. Biggs, Kevin A. Hassett and Matthew Jensen of the American Enterprise Institute, entitled “A Guide for Deficit Reduction in the United States Based on Historical Consolidations That Worked.” This paper, released in December 2010, has received an enviable amount of attention for a fairly technical enterprise.
To tell you what I am going to tell you: The authors argue that the United States should reduce its deficit much more (they pick 85 percent) by reducing spending, and thus much less by raising revenues, than the most widely recognized bipartisan plans (which are at about 50-50). I think they overplay their statistical hand. This post gets a bit nerdy, but in my view the reasoning comes down to a fairly fundamental issue. So all but the faint of heart, please read on.
As conveyed in the title of the paper, the concept is thoroughly sensible. We should learn all possible lessons from successful (and unsuccessful) experience around the world. However, I have some problems with the execution in this paper, which I believe could lead us down a path toward failure – either in achieving an agreement in the first instance, or the performance of the resulting agreement if we do. Let me first present their findings – in an admittedly abbreviated form – and then explain my concerns.
The co-authors (following the practice of the Econ tribe, I will refer to them collectively as BHJ for short) identify the “fiscal consolidations” that they want to study in two alternative ways. The first is based on reductions of the cyclically adjusted (that is, corrected to remove the effects of changes in the unemployment rate) non-interest budget deficit, observed ex post. The other definition is limited to actual purposeful changes in public policy. The latter criterion yields far fewer historical examples than the former, which can show up on the radar screen because of any developments that yield budget improvement. BHJ then identify “successful” deficit-reduction programs according to arithmetic criteria (the amount of the reduction of the ratio of debt to the GDP) three years after the episode began. The nations studied are the members of the Organisation for Economic Cooperation and Development (OCED), from 1970 through the latest data available at the time when they undertook their study.
BHJ find that successful (by their criterion) fiscal consolidations have relied primarily on spending cuts. Combining all of their results, spending reduction provided between 66 percent and 120 percent of the savings (the 120 percent meaning that revenues were reduced while spending was cut). So they take that finding, pick a target share of spending reduction – 85 percent – from the approximate middle of their range, and put together a recommended 35-year U.S. plan that meets that standard. A deficit-reduction program that employed 85 percent spending cuts would be quite different from the approximately 50-50 spending cut-revenue increase balance of the recommendations of the Bipartisan Policy Center Debt Reduction Task Force, for example.
So what to conclude? Is 85 percent the correct number for the spending-cut share of any U.S. deficit reduction plan? I have a lot of quibbles, many of which BHJ point out themselves. I will mention only my most super-quibble quibble. But that one leads to a potentially more significant problem which I believe is worth airing.
First of all, I am not sure that one can say that there has been an episode of “fiscal consolidation” simply because the budget improves – even after adjustment for changes in the unemployment rate, and even omitting fluctuations of interest expense. All sorts of spurious developments could cause even sizable periods of budget improvement that satisfy BHJ’s criteria, and reversal of those developments or subsequent totally unconnected events could push the budget in the other direction. Example: In one year, a budget windfall arises because of a strong national stock market, satisfying the BHJ definition that a “fiscal consolidation” is underway, caused by a “revenue increase.” In subsequent years, the country in question throws a party with the windfall, but then the underlying good news evaporates – resulting in an “unsuccessful consolidation.” Whether the party included a spending increase or a tax cut, this episode looks after the fact as though an attempted fiscal consolidation using a tax increase has failed to improve the budget.
Let me raise one real-world example to illustrate why BHJ’s arithmetic test to identify episodes of fiscal consolidation doesn’t work. Look at the United States. Ask a budget-process wonk when the Congress and the President have made major assaults on the deficit, and the two instances at the top of the list will be 1990 and 1993. But BHJ’s test identifies only one U.S. attempt at fiscal consolidation: 1991 (which may be the fiscal-year equivalent of the deal enacted during calendar year 1990). And using their formula, BHJ label it a failure – even though by 1999, the U.S. Treasury was actually buying back its outstanding debt in the open market. Not to complain about my job, but this stuff really is complicated. Formulas won’t answer your questions. You need people with grey hair (or no hair at all, as the result of repeated pulling) who know where the gimmicks are buried.
That problem, in my judgment, calls into significant question those among BHJ’s findings that rely on their first definition of a fiscal consolidation as any after-the-fact observed significant improvement in the budget. I believe that conclusions about episodes of fiscal consolidation require picking instances where nations actually changed policy purposefully to slow the growth of debt. And once you go down that road, the number of instances of fiscal consolidation becomes much less. BHJ identify only 36 such attempted fiscal consolidations among the OECD countries from 1970 through 2010, with the number of successful consolidations ranging from seven (involving only four countries) to 10 (among only six countries), depending on the precise criteria they employ. Thus, the statistical pickings are pretty slim (belated apologies to Slim Pickens).
And this highlights my major problem. BHJ believe that the results from this small group of nations should point the way for the United States. But these countries – Belgium, Denmark, Finland, and Sweden by BHJ’s narrow definition, adding Italy and the United Kingdom in the broader one – are quite different from the United States in one important respect: They all have much higher taxes than we do.
As of 2009 (the latest actual numbers generally available on a comparable basis), of the nations that attempted fiscal consolidations by these definitions, the lowest revenue share of the GDP would be that of the United Kingdom, at 34.3 percent of the GDP. The second-lowest of this group is Finland – much higher, at 44.1 percent of GDP. The highest would be Denmark, at 48.0 percent of its GDP. Revenues in the United States were only 25.5 percent of the GDP – not three-quarters those of the U.K., more than 40 percent lower than Finland, and barely half those of Denmark. In fact, the only OECD countries with lower tax burdens than the United States are Mexico (17.4 percent) and Chile (19.3 percent) – worthy nations, but quite different from the world’s wealthiest, which is responsible for much of the burden of maintaining international order.
This does not prove anything, but it raises a meaningful (in my judgment) question: Is it necessarily the case that budget solutions for high-tax, high-spending countries are the best options for one of the lowest-tax, lowest-spending countries in the world? A country that bears the burden of the maintenance of global security? A country that faces an aging demographic that will increase significantly the cost of health care and public pensions, even if those programs are successfully reformed?
I have further concerns about how BHJ construct their 85-percent-spending-cut deficit reduction plan, and the very notion of taking estimates of three-year budget results elsewhere and using them to guide a 35-year budget plan for this nation, when our budget process is almost always based on 10-year estimates. The plan you get might not satisfy any useful criteria. But those are questions for another day. My major concern is generalizing from European high-tax, high-benefit economies to a U.S. system that some would call “cowboy capitalism” – with much lower taxes and much smaller social programs. In fact, many who have cited BHJ’s results might be taken aback to think that they are accepting lessons from “Old Europe.”
To sum up, I have not proven that an 85-percent-spending-cut is too high for a U.S. deficit-reduction package; but I don’t believe than BHJ have proven that it is right. The United States is unique (some would say “exceptional”) among nations, and it must build its own plan. A package that is 85 percent spending cuts and is rejected by 50 percent of the chambers of our Congress will not achieve fiscal consolidation. But a 50-50 package that achieves meaningful healthcare and retirement reform, simplifies and streamlines our income tax system, and achieves (I am sure grudging) bipartisan support – and therefore will not be repealed immediately after the next election – might meet the key criterion of dispelling uncertainty for the population and for the financial markets. I do not know what our elected policymakers will do after this election and in the new year, but I worry much less about the percentage of spending cuts in their product than I do that they might fail to act at all.