The federal budget outlook has improved over the last year. The budget always moves with the economy, and there is a tendency for the numbers to move more than had been projected when a business cycle bends around a turning point. There are reasonably predictable relationships among output, employment, wages and profits, but financial market valuations and the incomes that flow from them are a bit of a wild card. For whatever reason, the swings in those incomes always tend to be sharper than the budget mavens expect when the economy changes direction — more sharply upward in good times, more sharply downward in bad times.
So right now we are in an upswing, and that is good news. However, every now and then the hosannas seem to be just a bit louder than appropriate. It may be useful to try to put the recent changes into context.
Let’s use the May update from the Congressional Budget Office (see here). CBO saw no change in the economy, and no change in legislation. However, they did revise their numbers on the basis of “technical” factors — what a late wise colleague of mine used to call the EOI, that is, the “extent or our ignorance.” CBO observed that incomes grew more rapidly than anticipated in late 2012 — but with no observed change in the economy, they attributed that development to taxpayers realizing income and postponing deductions before the turn of the year in anticipation of the expiration of the upper-income tax rate cuts. CBO also observed a slowing in the cost of long-term care under Medicaid — good news, but not a development that could continue and accelerate forever. CBO also observed declines in Social Security Disability Insurance applications and award rates, and modest but essentially program-wide reductions in Medicare spending. All told, CBO expects deficits smaller than their January estimates by $56 billion in 2014, growing in a zigzag pattern to a reduction of $82 billion in 2023.
Again, it is much better that we have improvements than worsenings in the budget outlook. But these are relatively small slivers off of large projected deficits. And still there is some talk suggesting that the long-term budget picture has improved in a meaningful way. Those suggestions do require some scrutiny.
CBO produced a revised long-term budget outlook in September of this year (see here). And if we turn the clock back to the preceding long-term estimates, issued in June of 2012 (see here), the net change between the two is decidedly adverse. By far the biggest change in the budget numbers came from the January 2013 fiscal cliff deal. The budget “baseline,” of course, assumes that the current law is continued without change. Before the fiscal cliff deal, that meant that all of the temporary tax cut provisions were assumed to expire in January 2013. Then, in actuality, fiscal cliff deal did allow the upper-bracket tax rate cuts to expire, but it continued the more-costly lower- and middle-bracket rate cuts, and several other wide-reaching provisions. The result was a significant ongoing reduction of revenue. Other changes in CBO’s assumptions were mixed puts and takes — Social Security costs were assumed slightly higher and healthcare costs slightly lower — but those changes essentially cancel out, and are much smaller individually. On net, the new long-term deficit numbers are unambiguously worse now than they were one year ago.
So why do some now suggest casually that the long-term budget outlook is improved? The answer is two accidents of measurement that steer the discussion in curious ways.
For a brief moment of background: It is of course easiest to express a complicated numerical picture if there is some single-number index. When it comes to the decades-long budget outlook, the handy single-number index is the “fiscal gap.” The fiscal gap is the amount of annual budget savings (expressed as a percentage of the GDP) that would be needed to leave the debt held by the public (again expressed as a percentage of the GDP) at the same level at the end of the period (commonly 25 years hence) as it was at the beginning.
(One side-curiosity here: The debt as of the end of 2013 is at a perilously high 73 percent of GDP. This common definition of the fiscal gap would require only that the federal government hold its debt at that dangerous level. Thus, it is as though one were to claim credit for neutralizing the risk of a hiker falling by holding him or her at the edge of a cliff for an extended period of time. A second point is that a remedy for the fiscal gap would have the public debt at the same percentage of GDP in 2038 as in 2013, but the debt might be significantly higher or lower than that level either before or after 2038.)
The September 2013 CBO report states that the fiscal gap is a comparatively paltry-sounding 0.8 percent of GDP. That small number might suggest that it would be easy to solve the budget problem, or indeed that it might not be too risky to wait and see if the problem solves itself. However, the CBO document goes on to explain that the comparably measured fiscal gap one year prior had actually been a substantial negative number. The reason was that CBO’s baseline assumed that the current law would continue unchanged, which meant that all of the temporary tax cuts would expire — substantially reducing the deficit. Now, as was discussed a moment ago, all of those revenues have fallen off of the table. Thus, although the long-term fiscal gap might seem small, it is in fact much worsened.
The CBO fiscal gap might appear small to some for another curious reason, which in this instance has nothing to do with the budget. The Bureau of Economic Analysis (BEA) of the Department of Commerce, the generator of the national income accounts, does a complete revision of the accounts (and their GDP component) every five years, and this July was one such revision. These revisions can arise from new source data (even for many years ago), but also on occasion for conceptual changes. And this time around, the BEA chose to revise its handling of (most notably) spending on research and development (R&D). Instead of treating R&D spending as a simple expense, the BEA decided to treat it as an investment to create a knowledge asset, which would then be depreciated over time. The result was to increase the measured amount of GDP. CBO took that revised historical GDP series and extrapolated it out into the future to create its long-term budget forecast. CBO’s projection, relative to its numbers as of June 2012, had the GDP higher in every year, with a range of increases from about 2 percent to about 5 percent, and an average of about 3 percent. Of course, this revision in no way says that the economy is any stronger (or even different) than we previously thought. Past production of goods and services was what it was. The revision merely provides information about the estimated production and depreciation of a knowledge asset over that time.
A similar but perhaps more obvious situation arose back in the 1990s, when the BEA switched from a fixed- to a chain-weighted price index, which reduced the measured rate of past inflation. Because this revision did not change nominal GDP, it said that measured real GDP growth had been faster over all of the history of the accounts. When I spoke for the Administration at a conference of economists at the time, one attendee asserted to me that the faster real economic growth had to mean that the federal budget would improve more rapidly in the future. I had to explain that the BEA’s methodological change, though arguably a technical improvement showing a better economy historically, did not change the amount of cash that the federal government had in drawer, either at that moment or at any time in the past. If real economic growth — properly measured — had been a bit faster than we thought but the budget improvement was what it was, then in the future we might expect slightly faster measured real economic growth but slightly less budget improvement per unit of that growth. Having the experts at the BEA working their magic slightly differently in a windowless room on 15th Street Northwest in D.C. should not have changed our expectations of how well the economy or the federal budget outside would work by one iota.
The same is true today with respect to this new methodological improvement with respect to spending on R&D. But there is one big difference from the 1990s: This revision increases the measured amount of nominal GDP. Therefore, even though the budget is not changed by a single dime, every budget aggregate — including the fiscal gap, and the public debt — that is measured as a percentage of the GDP is going to look smaller. Thus, it is highly tempting to take this occasion to point out how small the fiscal gap looks, and how the near-term projections of the debt as a percentage of the GDP are smaller than they were six months ago. But these “improvements” are totally spurious, being based solely on a change in the measurement concepts for the GDP — not on improvements in the budget itself.
So the bottom line from the recent CBO reports is that the federal cash drawer is a little less empty than we had expected it to be, and that is a good thing. But by meaningful measurement standards, the outlook still is for a debt that grows faster than our economy, and that will explode at some future moment if we do not change course. Perhaps that moment will come just a bit later than we had thought. Still, we should work no less urgently, because the stakes remain enormous.
But for those who insist on taking the higher measure of GDP as a sign of improved prospects for the budget, I do have a suggestion. Economists have wondered for decades why the GDP includes the value of the services of paid domestic workers, but does not at all include the unpaid services of family members who clean, cook, and clear clogged drains. An estimate by economists at the BEA indicates that counting home production would have increased the GDP by 26 percent in 2010. Now, if they would just take the big step and include all of that uncompensated production in the GDP, that might solve the budget problem all by itself!
But seriously, folks…