Can hypothetical measures of time preference predict actual and incentivised behaviour? Evidence from Senegal.

• Our paper measures time preference for hypothetical rewards via the standard Multiple Price list (MPL) format. We employ an incentivised experiment using unique reference numbers of banknotes to determine individual’s willingness save money. Hypothetically derived parameters do not predict either choices in our or actual saving decisions. results remain robust confounding factors within (fungibility and reputation effects). Time preferences are important determinant decision-making widely measured through survey questions. However, extent which they offer a good representation discounting remains largely unexplored. This estimates commonly-applied elicitation method. explore whether estimated correlate with incentivized behaviours related preferences. First, we consider correlation between result as means determining Individuals given banknote informed that if chose retain this specific note randomly assigned period (2, 7 14 days) will receive second banknote, effect doubling their initial endowment. Second, observable behaviour, including ownership savings account participation Rotating Credit Savings Association (ROSCA). Overall, show hypothetically-derived significantly correlated behaviour. due limited power find version comparable relevant literature can detect sizes line similar studies. Furthermore, recognise be noisy reflection subject factors, such inherent fungibility provide ancillary evidence suggesting main these considerations others.