Politicians as Persuaders: A Field Experiment on Messenger Effects and Social Program Uptake

Abstract

Accessing public sector programs often involves onerous administrative processes for citizens, which can demotivate program applicants and discourage uptake. In this study, I worked with the Philadelphia Housing Development Corporation (PHDC) to test an intervention to boost uptake of PHDC’s Basic System Repair Program (BSRP) by using local political leaders to persuade citizens to complete the program’s onboarding process. Specifically, I randomly assigned 2,575 BSRP applicants to either receive a motivational mailer from PHDC directly, or from the office of that citizen’s representative on the Philadelphia City Council. While I find that the average response rate for individuals who received councilmember mailers was roughly 2 percentage points higher than for the PHDC mailers, this result is largely driven by the effectiveness of letters sent by one particular councilmember’s office. Furthermore, I find little evidence to suggest that councilmembers are more effective messengers for constituents who share their race or gender, relative to others in their districts. These findings suggest that “messenger effects” can vary significantly case-by-case, and underscore the importance of analyzing heterogeneity when evaluating behavioral interventions.