In this podcast episode, we explore two real-life stories from the front lines of government performance improvement efforts that highlight the importance of replication and validation in evidence-based policy.
In the U.K., the department of revenue and customs, in conjunction with the U.K. Behavioral Insights Team (BIT), tested new versions of the letter sent to people who were late in paying their taxes. The new versions simply added ones additional sentence, a behavioral “nudge” that drew on the power of social norms. As the New York Times explained:
One nudge was a sentence telling recipients that a majority of people in their community had already paid their taxes. Another said that most people who owe a similar amount of tax had paid. Both messages bolstered tax collection, and combining them had an even stronger effect. Over the last financial year, the letters brought forward £210 million of revenue, Britain’s revenue and customs department says — money that otherwise would have had to be chased in costly court procedures and failed to earn interest for the government.
A few years later, the U.S. Treasury Department, in conjunction with the U.S. Social and Behavioral Sciences Team (SBST) — the Obama Administration’s version of the BIT — decided to test a similar approach. They updated the letters sent to people who owed non-tax debt to the Federal government, simplifying and personalizing the letters and adding similar behavioral nudges as in the U.K.. The results showed no effect on payment rates, underscoring how context matters.
To learn more, we are joined by Tammy Chang, a Senior Economist at U.S. Department of the Treasury within the Bureau of the Fiscal Service, and Nathaniel Higgins, a co-founder and Fellow on the former SBST.