Variation in Incentive Effects across Neighbourhoods

Abstract

Small monetary incentives increase survey cooperation rates, however evidence suggests that the appeal of incentives may vary across sample subgroups. Fieldwork budgets can be most effectively distributed by targeting those subgroups where incentives will have the strongest appeal. We examine data from a randomised experiment implemented in the pilot phase of the Irish Longitudinal Study of Ageing, which randomly assigned households to receive a higher (€25) or lower (€10) incentive amount. Using a random effects logistic regression model, we observe a variable effect of the higher incentive across geographic neighbourhoods. The higher incentive has the largest impact in neighbourhoods where baseline cooperation is low, as predicted by Leverage-Saliency theory. Auxiliary neighbourhood-level variables are linked to the sample frame to explore this variation further, however none of these moderate the incentive effect, suggesting that richer information is needed to identify sample subgroups where incentive budgets should be directed.