Can behavioral nudges and incentives help lower-income households build emergency savings with tax refunds? Evidence from field and survey experiments

Mathieu Despard, Stephen Roll, Michal Grinstein-Weiss, Bradley Hardy, Jane Oliphant

Research output: Contribution to journalArticlepeer-review

Abstract

Tax refunds are an opportunity for lower-income households to accumulate emergency savings so they have cash on hand to cover expenses when income is insufficient. Our field experiments testing different behavioral interventions to encourage refund saving via online tax filing show small effect sizes (0.12–0.14) and a low aggregate savings rate (12%) that might be increased were filers to receive financial incentives. We test a key provision of the Refund to Rainy Day Saving and Financial Security Credit Acts using a survey experiment, finding that hypothetical refund saving jumps from 16% with no financial incentive, to 71% and 80% with 25% and 50% matches, respectively, findings which are mostly insensitive to refund size. Our results suggest that public policies to provide greater financial support—including stronger income supports—will better prepare lower-income households for financial emergencies than behavioral interventions to nudge refund saving.

Original languageEnglish
Pages (from-to)245-263
Number of pages19
JournalJournal of Consumer Affairs
Volume57
Issue number1
DOIs
StatePublished - 1 Jan 2023
Externally publishedYes

Bibliographical note

Publisher Copyright:
© 2022 The Authors. Journal of Consumer Affairs published by Wiley Periodicals LLC on behalf of American Council on Consumer Interests.

Keywords

  • lower-income
  • saving
  • tax refunds

ASJC Scopus subject areas

  • Sociology and Political Science
  • General Economics, Econometrics and Finance

Fingerprint

Dive into the research topics of 'Can behavioral nudges and incentives help lower-income households build emergency savings with tax refunds? Evidence from field and survey experiments'. Together they form a unique fingerprint.

Cite this