Help future-self beat present-self
Commitment device
Commitment-device interventions let a person bind themselves in advance to a behavior their future self might rather skip — savings programs that lock in deposits, smoking-cessation deposits that are returned conditional on quitting, exercise pledges. The mechanism is temporal: it uses present preference to constrain future preference.
3
Experiments
1
Policy areas
2004–2008
Year span
3 / 3
Positive
When it works
When the person genuinely wants the long-run outcome but predictably wavers in the short run. Strongest evidence is for savings, smoking cessation, and exercise. The intervention requires self-aware participants who opt in.
Watch out for
Commitment devices that punish failure without supporting it can entrench shame for participants who relapse. The design needs to handle relapse gracefully — re-entry should be easy.
Commitment device across policy areas
Financial Services· 3 experiments
- Positive
Commitment Savings Accounts for Agricultural Workers
IPA / Equity Bank Kenya · Kenya · 2008
Effect: Commitment account holders: +82% savings relative to controls; agricultural input expenditures +37%; output +22%
- Positive
Save More Tomorrow (SMarT) Commitment Program
University of Chicago / Bernartzi & Thaler · United States · 2004
Effect: SMarT participants: savings rate increased from 3.5% to 13.6% over 4 years; immediate advice group: from 4.4% to 8.8%; advice refusers: 6.6% (no change). SMarT had 78% retention rate vs. 26% for immediate increase group.
- Positive
SEED Commitment Savings Account — Philippines
Yale University / IPA (Ashraf, Karlan, Yin) · Mindanao, Philippines · 2004
Effect: 28% of those offered SEED accounts opened one; savings balances increased by 82 pesos per week (approximately 81% of control mean) for account openers; no significant spillover to those offered but who declined