No, the 90 Percent Debt Threshold Hasn’t Been Proven

By Mike Konczal |

The deficit hawks at the Washington Post editorial board are worried. They are worried that the deficit is falling and the debt-to-GDP ratio is leveling off as a result of the numerous cuts and tax increases implemented over the past two years. Liberals know this and are starting to push back, either claiming that the deficit is coming down too quickly or arguing that the main medium-term deficit issues are taken care of and we should focus more on unemployment and other non-budget issues while implementing Obamacare reforms well. The CBPP has been leading the charge on this, noting various levels at which debt as a percent of GDP would level off in the following graphic:

The editorial focuses on the debt-to-GDP ratio leveling out too close to a 90 percent threshold. The writers also claim that there is a well-defined and well-established 90 percent threshold over which our economy will suffer. They write, “The CBPP analysis assumes steady economic growth and no war. If that’s even slightly off, debt-to-GDP could keep rising — and stick dangerously near the 90 percent mark that economists regard as a threat to sustainable economic growth.” This 90 percent threshold was proposed by Carmen Reinhart and Kenneth Rogoff in their 2010 article “Growth in a Time of Debt” (GITD). They found that economies with public debt over 90 percent of debt-to-GDP grew more slowly than other countries.

It’s always tough to figure out where consensus among economists lies. But economists don’t “regard” the 90 percent mark as definitive; in fact, this study and its claim have never even been peer reviewed by an economics journal. [1]

I don’t bring this up because something that’s peer reviewed should automatically be accepted as definitive, or that credentials are everything, or that only Very Serious Economics matter. (That’s a bad rule in general, and as an economics blogger that would be a doubly insane claim.) I bring it up only because the public should understand that the 90 percent threshold couldn’t survive peer review for a very important reason: It’s impossible to seperate the cause and effect here given the evidence collected. Policymakers and deficit hawks should reconsider if they’re running under the assumption that this is a well-established rule.

Remember that growth that is suprisingly slow will increase the debt-to-GDP ratio relative to expectations by definition. And periods of slower growth will lead to higher debt levels. That doesn’t mean that those debt loads caused the slower growth — in these cases it would be just the opposite. Reinhart and Rogoff present no techniques, tools or theory to break this problem down and determine what is the cause and what is the effect in this debt versus GDP relationship.

As John Irons and Josh Bivens of EPI noted in their review of the GITD paper (my bold):

First, the theory that governs the relation between debt and growth suggests strongly that causality runs more firmly from slower growth to higher debt loads. Slow economic growth, and especially growth that is slower than policy makers’ expectations, will lead to higher levels of debt as revenues fall and as automatic-stabilizer spending increases… Importantly, the timing matters. Persistent slow growth will yield high debt levels, and will thus mechanically yield to contemporaneous combinations of high debt and slow growth…

In short, the statistical evidence strongly suggests that the causality runs from growth to debt, and not the reverse. Given that theory and preliminary investigation agree in this case, it seems clear that the GITD analysis—which looks only at contemporaneous levels of debt and growth—is much more likely to capture causal relationships running from slow growth to high debt. This means there is very little reason for policy makers to think that there is a high-debt threshold that acts to slow growth.

As one economist wrote me in an email, “it is likely unpublishable in a top journal due to the fact that they have not developed any techniques to tease out causality in what are suggestive but non-conclusive correlations. For this work to be the *one* thing that politicians decide to take from economics is horrible.”

You can think that lower debt is better than higher debt ratios. You can be worried about interest payments, even though those are at a 30-year low and projected to go back to historical averages. But there isn’t a great reason to believe that that leveling out at 80 versus 90 percent of GDP matters that much when we have mass unemployment, low interest rates, and inflation in check. Growth matters just as much as GDP for this calculation, and it’s a terrible deal if we sacrifice either immediate growth or long-term investments in an attempt to bring down this debt-to-GDP ratio. There isn’t good evidence that the levels matter that much if the plan works, and it is likely the plan won’t work. Weakening growth is likely to balloon that deficit as well.

It’s important to get a sense of where the deficit hawks will focus next because, if it is true that the deficit wars are coming to an end, all those giant deficit hawk groups are still funded through the apocalypse. Their mission will be that of Peter Venkman in Ghostbusters: “Type something, will you? We’re paying for this stuff.” How will they keep busy and justify their taxpayer-subsidized funding? We may have just gotten an important glimpse.

[1] According to their C.V.s, it’s been published in the May 2010 issue of the American Economic Review, which is a special non-reviewed “papers and proceedings” issue.

Follow or contact the Rortybomb blog:


Mike Konczal is a Fellow with the Roosevelt Institute, where he works on financial reform, unemployment, inequality, and a progressive vision of the economy. His blog, Rortybomb, was named one of the 25 Best Financial Blogs by Time magazine. Follow him on Twitter @rortybomb.