Nudge Unit and HMRC moved £2.4m by telling late payers most locals had paid

HM Revenue & Customs sent the same standard reminder letter to taxpayers who had declared but not yet paid their tax bill for years, essentially restating the amount owed and the consequences of nonpayment, with response rates that had plateaued — the default lever tax authorities reach for to improve on-time payment is escalating threat: bigger fines, more explicit enforcement language, faster referral to collections.

The UK Cabinet Office's Behavioural Insights Team, established in 2010 and popularly known as the 'Nudge Unit,' set out to test something different with HMRC: rather than making the penalty more frightening, tell taxpayers an accurate fact about what their peers were actually doing, on the theory that people calibrate their own compliance partly against a belief about how many others are complying.

Around 100,000 taxpayers who owed declared but unpaid tax were randomized across several reminder-letter variants, the most effective of which added a specific social-norm statement — that the great majority of people in the recipient's own local area or town had already paid their tax on time — rather than repeating the amount owed or the penalty for nonpayment. No fine was increased, no new enforcement mechanism was introduced; the only change was one added sentence stating an accurate local norm.

Replaced the threat of a bigger penalty with an accurate social norm: telling late payers most locals had already paid corrected a mistaken belief and moved behavior.

Late payers likely assumed nonpayment was common, so stating the true local norm corrected that belief.

People calibrate their own compliance against what peers do, so a credible norm shifts what feels expected.

Local reference groups increase identification, strengthening the norm's influence.

The change cost one sentence, so the £2.4m gain vastly outweighed the testing cost.

Before escalating threats, test whether correcting a misperceived social norm with an accurate, specific reference group changes behavior more cheaply.

Payment rates rose from roughly 67.5% in the standard-letter control group to as high as about 83% with the most localized 'home town' norm message — pulling forward an estimated £2.4 million in tax payments within a 23-day trial window, with the Behavioural Insights Team projecting the approach could advance roughly £160 million of tax debt collection over a six-week period if applied at scale.