Synthesis
What we know about behavioral nudges in tax administration
Letter wording, social-norm framing, deterrence messaging, and form simplification — what raises tax compliance and what doesn't.
3 experiments synthesized · 3 positive, 0 mixed, 0 null, 0 negative
Tax authorities have been some of the most enthusiastic adopters of randomized behavioral pilots, in part because compliance is easy to measure and the cost of a letter is trivial. The accumulated evidence across the UK, US, Indonesia, Dominican Republic, and Scandinavia is now substantial.
The reliable findings: deterrence framing (referencing penalties or audit probability) consistently outperforms social-norm framing in the immediate term, and personalization to the recipient's situation outperforms generic letters. Form simplification — reducing the steps required to comply — typically produces larger and more durable effects than any message change.
What is less well studied is durability: most trials measure one filing cycle. The few that measure multiple cycles find that effects from message-based nudges decay quickly, while structural changes to the form persist.
Takeaway
Before rewriting a tax letter, audit the form. Most large compliance gains come from removing steps, not from changing words.
The underlying experiments
Positive findings
3 experiments- Positive
IRS Direct File Pilot
Internal Revenue Service / US Digital Service · United States (12 pilot states) · 2024
Effect: 90%+ user-satisfaction rating on post-filing survey; users reported saving an average of $160 in filing fees and approximately 30 minutes vs. commercial alternatives; ~78% completion rate among users who started a return. Independent analysis by the Government Accountability Office estimated annual potential savings of $11 billion if scaled nationally.
- Positive
Simplified Tax Penalty Payment
Indonesia Directorate General of Taxes · Indonesia · 2022
Effect: Simplification arm: +32% timely settlement vs. control; Deterrence: +27%
- Positive
Social Norm Tax Letters
Dominican Tax Authority (DGII) · Dominican Republic · 2019
Effect: Treated firms paid $184 million more than controls (0.22% of GDP)
Related policy areas