Austerity: a spreadsheet error?

Episode Summary

Executive Summary: The episode examines how Reinhart and Rogoff’s influential 2010 paper linking public debt above 90% of GDP to sharply lower growth shaped austerity politics worldwide, and how a student replication effort found spreadsheet and methodology errors that weakened—but did not erase—the correlation. The segment argues that consequential economic research needs scrutiny, because one study can strongly influence policy debates despite being incomplete or flawed.

Main Topics: Reinhart and Rogoff’s 90% debt threshold claim (Priority: 5/5): The episode explains the original paper’s finding that countries with debt-to-GDP ratios above 90% experienced dramatically lower growth, and why it quickly became a powerful idea in policy debates. Replication failure and student investigation (Priority: 5/5): Thomas Herndon, assigned to replicate the paper, could not reproduce the results, prompting further investigation with his professors and direct contact with the authors. Spreadsheet and data-handling errors (Priority: 5/5): The investigation uncovered an Excel averaging error that excluded five countries, missing data issues, and a weighting choice that gave a single year equal weight to many years of another country’s experience. Austerity politics and policy influence (Priority: 4/5): The paper was cited in debates over austerity in the US and Europe, influencing lawmakers and reinforcing deficit-reduction arguments during the post-crisis period. Scientific uncertainty and peer review (Priority: 4/5): The discussion contrasts the paper’s lack of peer review and public scrutiny with the need for caution when using empirical studies to justify major economic policy. Limits of the correction (Priority: 3/5): Even though the dramatic threshold result weakened, the guests note that high debt can still correlate with slower growth and that the austerity debate remains unresolved.

Key Arguments: The original paper’s headline claim of a 90% debt-to-GDP tipping point was highly influential but not robust to replication. A spreadsheet mistake excluded five countries from an average, affecting the results. Missing postwar data for some countries and the authors’ averaging method materially changed the conclusion. High debt still correlates with lower growth, but not with the dramatic cliff effect originally presented. Because the paper was not peer-reviewed and had major policy consequences, it should have faced more scrutiny before being used in political arguments. A single empirical study should not be treated as conclusive in a complex policy environment.

Data Points: Debt-to-GDP threshold: 90% - The headline level in Reinhart and Rogoff’s original paper associated with sharply lower growth Countries excluded by spreadsheet error: 5 - Australia, Austria, Belgium, Canada, and Denmark were accidentally left out of an average Countries included in the average: 15 of 20 - A spreadsheet oversight caused only 15 countries to be counted instead of 20 New Zealand growth year: 1951, -8% - Used as an example of a single year being given equal weight to many years in other countries Britain high-debt period growth: Nearly 20 years at 2.5% - Illustrates how the averaging method treated many years of experience as equal to one year from New Zealand Greece high-debt period growth: Nearly 20 years at 2.5% - Another example used to critique the weighting approach Publication date of replication draft: 15 April - Thomas Herndon and colleagues published their draft working paper on this date Policy meeting size: 40 senators - Reinhart and Rogoff’s results were presented to U.S. senators, amplifying their influence Conference date: January 4, 2010 - The annual American Economic Association meeting where the original paper was presented

Pivotal Quotes: "It set this image of a threshold or cliff at 90% public debt to GDP, and beyond that, woe unto the country that passes." — Michael Ash: Describing the original paper’s powerful interpretation of the debt threshold "We had this puzzle that we were unable to replicate the results as Reinhardt and Rogoff published them, and that really got under our skin." — Michael Ash: Explaining why the replication effort continued after the student project "This research has not been peer-reviewed. It has not been opened up to public scrutiny." — Mike Konzal: Critiquing the influence of the paper on austerity policymaking

Implications: The episode shows how fragile empirical findings can shape major fiscal policy. For listeners, it’s a caution to treat striking economic claims carefully, demand transparency, and remember that policy debates rarely turn on a single study.