Lifting the Crushing Burden of Debt

Thank you, Chairman Ryan and the other members of the Committee, for the opportunity to comment on the US economy and the risks for the federal budget and debt. I am currently Dennis Weatherstone Senior Fellow at the Peterson Institute for International Economics.I suspect that I was invited to this hearing titled "Lifting the Crushing Burden of Debt" because, for more than a decade, my research has focused on various types of financial crises, including their fiscal implications and other economic consequences. Specifically, some of this work has focused on the historical and international evidence on the links between public debt and economic growth.

The march from financial crisis to high public indebtedness to sovereign default or restructuring is usually marked by episodes of drama, punctuated by periods of high volatility in financial markets, rising credit spreads, and rating downgrades. This historic pattern is unfolding in several European countries at present. That situation is far from resolved and remains a source of uncertainty for the United States and the rest of the world. However, the economic effects of high public indebtedness are not limited to turmoil in financial markets. Quite often, a build-up of public debt often does not trigger expectations of imminent sovereign default and the associated climb in funding costs. But in the background, a serious public debt overhang may cast a shadow on economic growth over the longer term, even when the sovereign's solvency is not called into question.

In a paper written over a year ago with my coauthor Ken Rogoff from Harvard University, we examined the contemporaneous connection between debt and growth. I summarize here some of the main findings of that paper and as well as our recent related work and relevant studies from the International Monetary Fund (IMF) and European Central Bank (ECB).

Our analysis was based on newly-compiled data on 44 countries spanning about 200 years. This amounts to 3,700 annual observations and covers a wide range of political systems, institutions, exchange rate arrangements, and historic circumstances. The annual observations were grouped into four categories, according to the ratio of gross central government debt to GDP during that particular year: years when debt-to-GDP levels were below 30 percent; 30 to 60 percent; 60 to 90 percent; and above 90 percent. Recent observations in that top bracket come from Belgium, Greece, Italy, and Japan.

The main finding of that study is that the relationship between government debt and real GDP growth is weak for debt-to-GDP ratios below 90 percent of GDP. Above the threshold of 90 percent, however, median growth rates fall by 1 percent, and average growth falls considerably more. The threshold for public debt is similar in advanced and emerging economies and applies to both the post World War II period and as far back as the data permit (often well into the 1800s).

Debt Thresholds: The 90 Percent Benchmark

Mapping a vague concept, such as "high debt" or "over-valued" exchange rates, to a workable definition for interpreting the existing facts and informing the discussion requires making arbitrary judgments about where to draw lines. In the case of debt, it turns out that drawing the line at 90 percent was a critical one in detecting a difference in growth performance.

A hint about how important that cutoff is comes from the fact that countries rarely allow themselves to enter that high-debt range. Pooling the debt-to-GDP data for the advanced economies over the post-World War II period reveals that the median public debt-to-GDP ratio was 36.4. Fully three-quarters of the observations were below the 60 percent criteria in the Maastricht treaty governing the European Union. About 92 percent of the observations fall below the 90 percent threshold. If debt levels above 90 percent are indeed as benign as some suggest, one has to explain why they are avoided so often over the long sweep of history. (Generations of politicians must have been overlooking proverbial money on the street).

We do not pretend to argue that growth will be normal at 89 percent and subpar at 91 percent debt-to-GDP, any more than a car crash is unlikely at 54 mph and near certain at 56 mph. However, mapping the theoretical notion of "vulnerability regions" to bad outcomes by necessity involves defining thresholds, just as traffic signs in the United States usually specify 55 mph. Subsequent work suggests that we were generous in putting the threshold so high. An analysis at the European Central Bank, for instance, presents evidence that the negative impact of debt on growth may start at a lower 70-80 percent threshold for European countries.

Debt and Growth Causality

Our analysis looked at contemporaneous relationships between average and median growth and inflation rates and debt. Temporal causality tests are not part of the analysis. But where do we place the evidence on causality? For low-to-moderate levels of debt there may or may not be one. For high levels of debt, the evidence suggests causality runs in both directions.

Our analysis of the aftermath of financial crises presents compelling evidence for both advanced and emerging markets on the fiscal impacts of the recessions associated with banking crises. There is little room to doubt that severe economic downturns, irrespective of whether they originated with a financial crisis or not, will, in most instances, lead to higher debt-to-GDP levels contemporaneously with or without a lag. There is, of course, a vast literature on cyclically-adjusted fiscal deficits making exactly this point.

A unilateral causal pattern from growth to debt, however, does not accord with the evidence. Public debt surges are associated with a higher incidence of debt crises. In the current context, even a cursory reading of the recent turmoil in Greece and other European countries can be importantly traced to the adverse impacts of high levels of government debt (or potentially guaranteed debt) on county risk and economic outcomes.

There is scant evidence to suggest that high debt has little impact on growth. Kumar and Woo (2010) highlight in cross-country analysis that debt levels have negative consequences for subsequent growth, even after controlling for other standard determinants in growth equations. For a dozen European countries a study from the European Central Bank (Chechrita and Rother, 2010) provides further evidence of negative causality from debt to growth.

I will conclude on the same note of my testimony of about a year ago before your Senate counterparts. The sooner our political leadership reconciles itself to accepting adjustment, the lower the risks of truly paralyzing debt problems down the road. Although most governments still enjoy strong access to financial markets at very low interest rates, market discipline can come without warning. Countries that have not laid the groundwork for adjustment will regret it.

This time is not different.