Home › What the research says

Commitment Devices: What Field Experiments Found About Saving

The Philippines commitment account

Nava Ashraf, Dean Karlan, and Wesley Yin published "Tying Odysseus to the Mast" in the Quarterly Journal of Economics in 2006. Using a randomized design with existing and former clients of a bank, they offered a commitment savings product (SEED) to a randomly chosen subset of 710 clients; 202 of them, 28.4 percent, opened the account. After twelve months, average savings balances increased by 81 percentage points for clients assigned to the treatment group relative to the control group. They also report that women who exhibited a lower discount rate for future relative to current tradeoffs, and therefore potentially a preference for commitment, were significantly more likely to open the account.

The simplest possible device: a box

Pascaline Dupas and Jonathan Robinson reported in the American Economic Review in 2013 on health savings experiments in Kenya. One of their interventions was a lockbox, similar to a safe box, where the key was held by a program officer so participants could not easily access the money. Simply providing a safe place to keep money was sufficient to increase health savings by 66 percent. Participants themselves reported two reasons: it made it easy to save small change (33 percent) and it increased the barriers to spending because the money was not immediately at hand (32 percent).

Commitment through future income

Richard Thaler and Shlomo Benartzi's Save More Tomorrow (SMarT) program, published in the Journal of Political Economy in 2004, took a different route: participants commit in advance to allocating a portion of future salary increases to retirement savings. At one company, employees who joined saw savings rates rise from 3.5 percent of income before the program to 13.6 percent by the fourth pay raise after joining. At a U.S. manufacturer, average savings rates rose from 4.4 to 10.6 percent.

What these studies have in common

In every case, the intervention changed the structure surrounding the money rather than the person's knowledge or motivation. A locked box, a restricted account, an automatic escalation schedule. That is the same family of ideas envelope budgeting draws on: constrain the money in advance so fewer decisions have to be made under pressure later.

The important caveats

These are strong designs, but they are specific. Ashraf, Karlan, and Yin studied bank clients in the Philippines; Dupas and Robinson studied health savings in rural Kenya; Thaler and Benartzi studied employer retirement plans in the United States. Take-up was also far from universal: only 28.4 percent of those offered the commitment account accepted it. Generalizing to any particular US household, or to any consumer budgeting app, is an extrapolation, not a finding.

Sources

Every source below was retrieved and checked. Findings are reported as the source states them.

• Tying Odysseus to the Mast: Evidence from a Commitment Savings Product in the Philippines — The Quarterly Journal of Economics, 2006
  In a randomized evaluation, 202 of 710 clients (28.4 percent) offered a commitment savings product opened one, and after twelve months average savings balances increased by 81 percentage points for the treatment group relative to control.
• Why Don't the Poor Save More? Evidence from Health Savings Experiments — American Economic Review, 2013
  Providing a lockbox whose key was held by a program officer increased health savings by 66 percent; participants cited easier saving of small change (33 percent) and increased barriers to spending because money was not immediately at hand (32 percent).
• Save More Tomorrow: Using Behavioral Economics to Increase Employee Saving — Journal of Political Economy, 2004
  Employees who joined the SMarT plan at one company saw savings rates rise from 3.5 percent of income to 13.6 percent by the fourth pay raise after joining; at a U.S. manufacturer average savings rates rose from 4.4 to 10.6 percent.

Common questions

Do commitment devices work for everyone?

The evidence suggests not. In the Philippines study only 28.4 percent of clients offered the commitment account opened one, and take-up was higher among those whose measured preferences suggested a demand for commitment. Commitment devices appear to be most useful for people who already sense they need them.

Does the device have to lock money away?

Not necessarily. Dupas and Robinson found that a lockbox increased health savings by 66 percent, and participants cited both easier small-change saving and increased barriers to spending. Both a designated place and a restricted place appear to contribute.

Do these studies apply to budgeting apps?

Not directly. None of them tested a consumer budgeting application. They tested savings accounts, physical boxes, and payroll-deduction programs.

Related