By James Kwak
According to its CBO score, the Budget Control Act of 2011 (a.k.a. the debt ceiling agreement) initially reduced aggregate budget deficits over the next ten years (2012–2021) by $917 billion, with a provision that ensures that deficits will be reduced by another $1.2 trillion (either through an agreement in the joint committee that is ratified by Congress, or through automatic spending cuts). The chatter in Washington is that even with the $1.2 trillion, this is still too small, and there is still this massive deficit hanging over our heads. This is true to an extent, but not the way you are being led to believe.
The first question is this: How big is the deficit anyway? The answer is pretty complicated—complicated enough for S&P to mess up (although in my opinion they made a rookie mistake, as I’ll explain later). Warning: lots of numbers ahead, though the only math is addition and subtraction.
What’s complicated is that the CBO score doesn’t say what the projected deficit is: it just says what the impact of the Budget Control Act was on some baseline. And the baseline was—wait for it—the March 2011 baseline (included in the annual Analysis of the President’s Budgetary Proposals), modified to include the extrapolation of the final FY 2011 appropriations (remember the first spending cut showdown?), and excluding overseas contingency operations such as Iraq and Afghanistan (which are in the January Budget and Economic Outlook, Table 1-7). So let’s go through this step by step.
In March, this is what the CBO baseline looked like:
For example, the federal government was projected to run a primary surplus of $78 billion and pay $807 billion in interest in 2021, leaving a deficit of $729 billion, or 3.1 percent of GDP.
This is the effect on outlays of extrapolating the changes in 2011 appropriations:
(The CBO didn’t estimate the effect on debt service, but it’s sure to be trivial.) The CBO has to do this because, by law, its baseline forecast must assume that all discretionary spending grows at the rate of inflation. So projections for discretionary spending in future years (which, by definition, has to be appropriated by Congress each year) are just the actuals for the most recent year plus future inflation.* In this case, those are positive numbers, so they reduce future deficits.
Then we have phase one of the Budget Control Act:
The numbers in the top row add up to $917 billion (actually, $918 billion since my spreadsheet doesn’t do the rounding correctly). Those are the new caps on discretionary spending, which are, in aggregate, $917 billion lower than the numbers in the old CBO baseline.
Various extreme right-wingers, including GOP presidential frontrunner Michele Bachmann, have been claiming that those spending cuts are fake because they only affect future spending. This is silly. First of all, there’s no other way to reduce future discretionary spending, because you can’t actually reduce it until appropriations season rolls around. Second, these caps are below the CBO baseline, which grows at the inflation rate. The CBO baseline has generally been considered too optimistic (from a deficit-minimizing point of view), because historically discretionary spending growth has been closer to the rate of economic growth than to the inflation rate. So setting caps below the inflation rate is a major reduction from the general path of policy.
Add those numbers together, and this is what you get:
Now, in 2021, the federal government has a primary surplus of $210 billion, interest payments of $767 billion, and a deficit of $557 billion, or 2.3 percent of GDP. That is roughly where the official baseline should stand after the Budget Control Act. (The CBO score also includes $1.2 trillion of unspecified future deficit reductions because of part two of the BCA.)
But, you may be saying, the CBO baseline is unrealistic to begin with. This is true. So I then adjusted the numbers above by applying three alternative assumptions from the January Budget and Economic Outlook: the planned drawdown of forces from Iraq and Afghanistan (since we know that spending on the wars is not going to grow at the inflation rate);** extension of the Medicare doc fix; and additional patches to the AMT that have the effect of indexing the AMT to inflation. These are the three policies from the CBO’s list that are the most likely to actually happen. Here they are:***
(Note that in this case the CBO estimates changes to outlays and then effect on debt service separately; the first number does not include the second, so you have to add them together to get total deficit impact.) You can see at a glance that the drawdown of forces slightly outweighs the two other adjustments, so the overall effect is to reduce future deficits. So when you add these to the previous numbers, you get:
That is, from 2018 through 2021 the government runs primary surpluses around $200 billion (0.9–1.1 percent of GDP), pays about $700 billion in interest (3.0–3.2 percent of GDP), and runs a deficit around $500 billion (1.9–2.3 percent of GDP). In total, we’ve sliced about $1.3 trillion off the 2021 national debt, bringing it down from 76 percent of GDP (in the March baseline) to 70 percent.
If you’ve made it to this point, there should be three things you should be thinking. (Actually, four, including “Doesn’t James have anything better to do?”)
First, since when is a stable deficit of 2 percent of GDP a national emergency? I agree the situation is not ideal. We won’t have a lot of fiscal space to deal with future shocks (e.g., financial crises), and we’ll be in a race pitting economic growth against the real interest rate on the debt. But at that point we’ll basically be paying interest on stupid policy decisions from the past (take your pick).
The real problem is that over the following two decades, things will slowly and then rapidly get worse as more Baby Boomers retire and health care costs continue to increase. But that takes place outside the ten-year window—and it’s mainly a health care problem. So why do we have a joint committee of Congress mandated to slash $1.2 trillion from the budget within that window? I know the political reason, but that doesn’t mean it makes any economic sense.
Second, the not-so-bad scenario painted above has one major assumption: that we let all expiring tax cuts, well, expire. If, instead, we make them all permanent, we’ll never have primary surpluses, we’ll have deficits over 5 percent of GDP in 2019–2021, and the national debt will be around 90 percent of GDP in 2021—with the demographic wave only beginning. But it seems to me this could be a political opportunity. President Obama can say: “I have a bulletproof, real plan to bring the deficit down to a sustainable level within this decade. I will veto any tax cut extension unless it is fully offset. The entire existence of a deficit crisis over the next decade is predicated on extending the tax cuts.” (I know, the fact that I used the word “predicated” rules me out as a political speechwriter.) Then he can negotiate from a position of strength. But, of course, we’ve been through this before.
Third, and less importantly: What was S&P thinking? According to their downgrade explanation, the problem is that total government debt (including all levels) is growing from 74 percent of GDP in 2011 to 85 percent in 2021—and they’re counting the $1.2 trillion in deficit reduction from the second phase of the BCA. I don’t know the state/local figures, but by my spreadsheet, if you count that $1.2 trillion (which I didn’t count above), federal debt falls from 69 percent at the end of 2011 to 65 percent at the end of 2021. The only way they get numbers that high is by working off of an alternative scenario in which all the tax cuts are extended and then assuming poor economic growth. (Poor economic growth seems like a good assumption, but it also argues for less deficit reduction rather than more in the short term.) In other words, S&P downgraded the U.S. because they assumed that the Bush tax cuts will be extended again. Again, President Obama could call them out on this if he were willing to let the Bush tax cuts expire—but we’ve been there before.
So if there’s one thing you should take away, it’s this: The ten-year deficit problem is a tax cut problem. No tax cuts, no problem. This won’t solve the all the long-term problems, but it’s a good start.
* S&P’s mistake was thinking that the “baseline” included discretionary spending growing at the rate of GDP. This is a big mistake because, by law, there is only one CBO baseline, and it has discretionary spending growing at the inflation rate. The CBO does publish projections under “alternative policy assumptions,” but they don’t call them baselines. Furthermore, the CBO’s score for the Budget Control Act—published in August 1, four days before the downgrade—was painfully explicit about what baseline it was using, and even repeated, “In CBO’s baseline projections, appropriations for discretionary programs are assumed to grow each year with inflation from the amounts provided for the most recent year.”
** This adjustment should be applied. The Budget Control Act excluded overseas contingency operations. So think of all discretionary spending in the March baseline as (a) overseas contingency operations and (b) everything else. We captured the reductions to (b) above. Since (a) was not affected by the BCA, the alternative policy assumption from January is still applicable.
*** These are from January and were not updated in March, so they could be off by now by a few billion dollars. In this context, that is not real money.