Price and Control Elasticities of Demand for Savings in the Philippines

The researchers collaborated with a for-profit bank in the Philippines on a randomized field experiment to study the determinants of demand for a savings commitment account. After a baseline survey, 9,992 individuals were solicited to open a savings commitment account. Potential clients were randomly assigned interest rates and (conditional on being married)account ownership requirements. The researchers used an OLS specification to estimate the impacts of the two treatments on take-up and saving, controlling for neighborhood, marketer, and week-of-offer fixed effects. The researchers used the results to estimate the price and control elasticities of demand for the savings account.

Of the 9,992 individuals offered, 2,265 opened accounts. 1,523 accounts were opened by married individuals. The researchers found no statistically significant impacts of either interest rate or account ownership on take-up or average savings balance. Estimated demand elasticities suggest that demand was insensitive to both interest rate variation and account ownership requirements. Demand is strongly correlated with client characteristics and bank sales efforts, and there is no evidence of effect heterogeneity.

2007-04-01

2008-11-24

Take-up; several average account balance measures; whether an individual reached their saving goal; if more than one deposit was made

The field experiment was carried out by door-to-door marketers who approached more than 10,000 people with solicitations for a commitment savings account. Using a PDA, marketers helped prospective clients complete a brief baseline survey, which was also used to screen out unpromising prospects. After the survey, marketers offered prospective clients the opportunity to open the commitment savings account. Clients set their own savings goal and time frame with no fixed deposit schedule. Marketers independently randomized potential clients into the different treatment arms using PDAs.

The interest rate treatment had three experimental arms with equal probability of assignment: a "normal" rate of 1.5% APY, a "high" rate of 3% APY; and a "reward" rate where the client would receive the high rate if they met their savings goal or the normal rate if they did not. The account ownership treatment had three experimental arms with equal probability of assignment, conditional on being married: individual ownership, joint ownership with spouse, and optional ownership where client could choose individual or joint ownership.

Individuals were randomized independently on personal PDAs after being surveyed

Individual

No

9,992 individuals

9,992 individuals

Interest rate treatment: 3,329 assigned to the regular rate; 3,367 assigned to the high rate; 3,296 assigned to the reward rate

Account ownership treatment: 3,275 assigned to individual ownership; 3,283 assigned to joint ownership; 3,434 assigned to chose individual or joint ownership

Assumptions about individual demand for savings underlie workhorse models of intertemporal choice and intra-household bargaining, banking strategy, and financial inclusion policy. A Philippine bank tested sensitivity to interest rates and account ownership requirements in 10,000 randomized door-to-door solicitations for a commitment savings account. Take-up is substantial (23%), but price elasticity of saving in this account is not significantly different from zero in either the full sample or sub-groups of plausibly marginal savers. The upper bound is less than 0.5 in the full sample, and exceeds 1.0 in only 1 of 22 sub-groups. Nor do we find sensitivity to ownership requirements.

Karlan, Dean, and Jonathan Zinman. "Price and Control of Demand for Savings." Working Paper, January 2014.

https://www.povertyactionlab.org/sites/default/files/publications/63_49%20SavingsElasticities%20Jan2014.pdf