Misvoting in the Ballot: Random Power to Parties and its Effect on Politicians and Policies

72 Pages Posted: 29 Jun 2020 Last revised: 14 May 2021

See all articles by Matteo Cervellati

Matteo Cervellati

Università di Bologna

Giorgio Gulino

Universitat Pompeu Fabra, Department of Economics and Business, Students

Paolo Roberti

Free University of Bozen-Bolzano - Faculty of Economics and Management

Date Written: June 2020

Abstract

A natural experiment allows isolating the role of parties for the selection of politicians and policies. We exploit lotteries used to set the position of party symbols on all ballot papers in Italian local elections for over a decade. We document and quantify the existence of systematic, but random, electoral noise in the allocation of votes across parties. We find that even small perturbations in the power of parties impact policies. Randomly treating a party with a boost in votes shifts coalition policies towards this party's platform, but only for the more divisive electoral issues. Changes in relative political power within majorities also lead to the appointment of cabinet members with significantly different socio-demographic characteristics, while no effect is detected for policy-maker level of education. The unintentional experiment also helps shedding light on the role of party ideologies.

Suggested Citation

Cervellati, Matteo and Gulino, Giorgio and Roberti, Paolo, Misvoting in the Ballot: Random Power to Parties and its Effect on Politicians and Policies (June 2020). CEPR Discussion Paper No. DP14906, Available at SSRN: https://ssrn.com/abstract=3638010

Matteo Cervellati (Contact Author)

Università di Bologna ( email )

Via Zamboni, 33
Bologna, 40126
Italy

Giorgio Gulino

Universitat Pompeu Fabra, Department of Economics and Business, Students ( email )

Barcelona
Spain

Paolo Roberti

Free University of Bozen-Bolzano - Faculty of Economics and Management ( email )

Via Sernesi 1
39100 Bozen-Bolzano (BZ), Bozen 39100
Italy

Do you have negative results from your research you’d like to share?

Paper statistics

Downloads
0
Abstract Views
264
PlumX Metrics