The authors analyze periods of major public indebtedness and identify 26 episodes over the past 200 years that share certain characteristics. First, once a debt overhang has lasted five years, it is likely that it will continue for an extended period of time. Second, a public debt overhang substantially slows down economic growth. Third, a public debt overhang does not necessarily coincide with rising real interest rates.

The authors focus on the 26 episodes in various advanced economies when gross public debt exceeded 90% of GDP on a sustained basis and analyze the interactions between public indebtedness, GDP growth, and real interest rates.

First, the authors find that a public debt overhang that has lasted for a minimum of 5 years is likely to continue for at least 10 years. The average duration of the debt overhang episodes they research is 23 years.

Second, the authors demonstrate that a public debt overhang correlates with a significant slowing of real economic activity. Although the average growth rate during periods without debt overhang is 3.5%, it is only 2.3% during episodes of a debt overhang. The authors review a number of recent studies to shed light on the difficult question of causality: Does the debt overhang cause lower growth, or is it the result of slow growth? They conclude that, especially at high levels, public debt does have a negative effect on GDP growth. Given the long duration of the episodes, the cumulative negative output effect is massive.

Third, the authors show that in 11 of the 26 public debt overhang episodes, real interest rates were, on average, comparable to or lower than rates at other times. Public indebtedness thus does not necessarily involve a rise in real interest rates or a financial crisis. Given the observed negative correlation between debt overhang and output, factors other than higher interest rates must play a role in the transmission mechanism. One possibility is the downward adjustment in private spending in anticipation of future increases in taxation.

Applied to the current situation, a protracted continuation of the uncomfortably high levels of public indebtedness is very likely, with possibly large cumulative adverse effects on growth. The authors take issue with the view that soaring government debt does not matter as long as markets and central banks seem willing to absorb it at low interest rates.

More broadly, they suggest that any policy analysis should include a comparison of the cumulative long-term costs of high indebtedness and the short-term benefits of fiscal stimulus.

The analysis draws on a cross-country database of central government debt that covers 22 advanced economies. The length of the time series varies by country and depends on the availability of data. Although the time series for the United States starts in 1791, for example, the first datapoint for Ireland is 1924. All data end in 2011.

The authors identify public debt overhang episodes, which they define as gross public debt levels in excess of 90% of nominal GDP for five years or more. The 26 debt overhang episodes they identify are concentrated in 13 economies. For these economies, the authors provide summary statistics about average real GDP growth and real interest rates for the two regimes—debt overhang and no debt overhang—and calculate the share of years that the countries have spent in a state of debt overhang. To get a better understanding of the debt overhang episodes, they also cluster the episodes according to major types or causes.

The historical evidence the authors provide about past periods of fiscal overhang is both convincing and sobering. As they demonstrate in the introduction, other forms of debt are also a reason for concern. These include private debt, external debt, and the actuarial debt implicit in underfunded public pension and medical care programs. Because all of these are also at record levels, the historical record of fiscal overhang is likely to underestimate the true extent of the current problem.