Date of Award

8-26-2022

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Public Administration

Advisor(s)

Leonard Lopoo

Keywords

experiment aversion, impact evaluation, public administration, public sector, randomized experiment

Subject Categories

Public Affairs, Public Policy and Public Administration | Social and Behavioral Sciences

Abstract

This dissertation examines whether public administrators are averse to randomized controlled trials, and why. More specifically, it explores whether public administrators are reluctant or opposed to conducting RCTs for the evaluation of social welfare programs in situations where they would help to answer an important and useful policy question and are feasible to conduct. While RCTs are widely used to evaluate clinical interventions and practices, they are used far less often in social policy. I first propose a conceptual framework for understanding the public administrator’s decision-making process as they contemplate whether to conduct an RCT, outlining nine factors that likely contribute to RCT aversion. Then, using a survey experiment with a nationally representative sample, I investigate three research questions. First, do people, on average, prefer a quasi-experiment to the RCT? Second, do features of the policy environment that create a greater perceived difference in treatment between groups contribute to more RCT aversion? And third, do preferences for the RCT differ for public administrators compared to their non-public administrator peers? I find that the majority of people demonstrate a strong preference for the quasi-experiment to the RCT. Public administrators are RCT averse on average, but less so compared to their general public peers. Additionally, I find that public administrators are likely to be more RCT averse when the intervention is perceived to be very different, and potentially superior, to the status-quo option available to members of the control group. People who are not public administrators are not sensitive to features of the policy environment. I conclude by outlining several avenues for future exploration of public administrator RCT aversion, and implications for social policy researchers, evidence-based policy advocates, and public policy and administration educators.

Access

Open Access

Download

Share

COinS