Do Voters in Local Elections Prefer Campaign Promises About Attributable Policies?

66 Pages Posted: 11 Oct 2021 Last revised: 26 May 2022

See all articles by Tanushree Goyal

Tanushree Goyal

Harvard Weatherhead Center

Robin Harding

University of Oxford

Date Written: March 15, 2021

Abstract

It is widely accepted that policy attribution increases retrospective accountability. We extend this research by outlining the theoretical importance of policy attribution in electoral selection and investigating whether voters select candidates based on campaign promises for attributable policy outcomes. Selecting candidates on attributable campaign promises can increase political responsiveness and lower pandering, representing an overlooked channel that can increase accountability. We introduce novel measures of attribution and conduct comparative conjoint experiments in representative surveys in Accra, Ghana, and New Delhi, India. We find that in both settings the quality of attribution is high and policy promises are the strongest determinant of vote choice. Yet, we find no evidence that voters prefer attributable campaign promises. Instead, voters are just as likely to select candidates based on campaign promises for policies they do not attribute to them. Our findings have important implications for the operation of accountability, and for justifications of decentralization.

Keywords: Accountability, Attribution, Ghana, India, Candidate Choice conjoint experiments

Suggested Citation

Goyal, Tanushree and Harding, Robin, Do Voters in Local Elections Prefer Campaign Promises About Attributable Policies? (March 15, 2021). Available at SSRN: https://ssrn.com/abstract=3886411 or http://dx.doi.org/10.2139/ssrn.3886411

Tanushree Goyal (Contact Author)

Harvard Weatherhead Center ( email )

K239 1737 Cambridge St.
Cambridge, MA 02138
United States

Robin Harding

University of Oxford ( email )

Manor Road
Oxford, OX1 3UQ
United Kingdom

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
73
Abstract Views
397
rank
456,955
PlumX Metrics