Social-Norm Tax Reminder Letters: The HMRC Natural Field Experiments
Edited by Paweł Raja, PhD · Published · Updated
In a natural field experiment with 119,527 UK taxpayers who owed Self Assessment tax, 33.6% of those sent the standard HMRC letter paid within 23 days; adding a local-and-debt social norm raised payment by 5.0 percentage points, a 14.9% relative increase (Hallsworth et al., 2017).
Taxpayers paying within 23 days (Experiment Two) · Natural field experiment
Standard reminder letter: 33.6% → Letter + local & debt norm: 38.6% (+5.0 pts, +15% relative)
38.6% = the 33.6% control rate plus the reported +5.0-point effect (a 14.9% relative increase).
Study details
- Study design
- Natural field experiment
- Problem
- HM Revenue & Customs (HMRC) sends reminder letters to Self Assessment taxpayers who have declared their income but not paid the tax they owe. Every further letter, phone call and enforcement step costs money, so the question was whether one extra sentence in the standard letter could get more people to pay on time.
- Intervention
- In two randomized trials run with HMRC and the Behavioural Insights Team, a single short message was added to the standard reminder letter. Messages included social norms ("Nine out of ten people in the UK pay their tax on time"), a minority framing ("You are currently in the very small minority of people who have not paid us yet") and, in Experiment Two, a local-and-debt norm: "The great majority of people in your local area pay their tax on time. Most people with a debt like yours have paid it by now."
- Outcome
- Taxpayers paying within 23 days (Experiment Two): 33.6% (Standard reminder letter) → 38.6% (Letter + local & debt norm); +5.0 pts, +15% relative
- Sample
- 119,527 taxpayers in Experiment Two; 101,471 in Experiment One
- Key takeaway
- Descriptive norms that are specific to the reader (their area, their kind of debt, or their status as one of the few who have not paid) raised payment rates by several percentage points at practically zero cost, and they outperformed injunctive "you should pay" messages.
- Biases involved
- Social Proof
- Source
- Hallsworth, M., List, J. A., Metcalfe, R. D., & Vlaev, I. (2017). The behavioralist as tax collector: Using natural field experiments to enhance tax compliance. Journal of Public Economics, 148, 14-31. doi:10.1016/j.jpubeco.2017.02.003 ↗ full text (PDF)Numbers checked against: NBER WP 20007: Table 4 (Experiment One); Table 7 and pp. 24-25 (Experiment Two); Section 7.2 (external validity)
What the study did
The paper reports two natural field experiments run inside HMRC's normal debt-collection process. Taxpayers did not know they were in a study: each one received one version of a reminder letter HMRC was sending anyway. The outcome comes from HMRC's administrative records. It is whether the taxpayer paid, or started to pay, the outstanding balance before the next reminder could arrive, which the authors put at 23 days after the first letter.
Experiment One (2011) covered 101,471 taxpayers with debts between £400 and £100,000. They were split into a control letter and five message variants. Every message raised payment. The basic norm added about 1.1 to 1.3 percentage points and the country norm 1.7 to 2.1 points, depending on the regression specification. The minority-norm message had the largest effect, 3.5 to 5.1 points, significantly larger than any other message. Public-goods messages ("Paying tax means we all gain from vital public services like the NHS, roads, and schools") added about 1.3 to 1.6 points, and framing them as a loss instead of a gain made no extra difference.
Experiment Two followed with 119,527 taxpayers in fourteen groups, designed to find out which part of the message mattered. With the standard letter, 33.6% paid within 23 days. A general "great majority of people in the UK" norm added 1.4 points, a local-area norm 2.2 points and a "debt like yours" norm 3 points. Combining the local and debt norms added 5.0 points. The minority result replicated: the minority-status sentence alone added 4.7 points and the full minority norm 4.2 points. Descriptive norms beat injunctive ones, and messages about interest charges and how to pay added 3.2 to 3.9 points.
The authors estimate that the messages brought in more than £3 million within the 23-day window in Experiment One and more than £9 million in Experiment Two. The marginal cost of changing a sentence in a letter that is mailed anyway is practically zero.
The often-quoted figures "67.5% to 83%" and "140,000 taxpayers" are not from this paper. They come from an earlier, separate trial reported in the Behavioural Insights Team's 2012 paper Applying behavioural insights to reduce fraud, error and debt. That trial compared HMRC's old-style letter with whole new letters, rather than changing a single sentence, and its outcome was the share of late payers who had responded after three months. It was reported in a government paper, not peer reviewed, so we use the journal article's estimates above.
Limitations and caveats
- The sample is taxpayers who had already missed the deadline and ignored a reminder, so it is not representative of all taxpayers. The authors argue (Section 7.2) that the messages would likely work better on the wider population.
- Effects varied by subgroup. For taxpayers with liabilities of £30,000 or more, the minority norm reduced the likelihood of payment, while the public-goods messages worked well (footnote 16).
- The outcome is payment within 23 days, before the next reminder. The trials measure how quickly people pay, not total long-run compliance.
- The figures here are from the NBER working-paper version (No. 20007, 2014), Tables 4 and 7. The peer-reviewed version (Journal of Public Economics, 2017) reports the same two experiments; small differences between versions are possible.
Further sources
- Our research summary: The Behavioralist as Tax Collector: Using Natural Field Experiments to Enhance Tax Compliance
- Behavioural Insights Team (2012). Applying behavioural insights to reduce fraud, error and debt. Cabinet Office, London (Trial 1a, p. 23). ↗ link
Cite this page
Raja, P. (Ed.). (2026, 16 Oct 2026). Social-Norm Tax Reminder Letters: The HMRC Natural Field Experiments. Behavioral Economics Lab. https://www.behavioraleconomicslab.com/findings/uk-tax-compliance
Primary source: Hallsworth, M., List, J. A., Metcalfe, R. D., & Vlaev, I. (2017). The behavioralist as tax collector: Using natural field experiments to enhance tax compliance. Journal of Public Economics, 148, 14-31.