Polarizing information and support for reform

Nicholas Haas, Mazen Hassan, Sarah Mansour, Rebecca B. Morton

    Research output: Contribution to journalArticlepeer-review


    We examine whether political polarization is an obstacle to common value reforms. We conduct experiments in two ideologically polarized countries, the United States and Egypt. Subjects vote between enacting a reform which yields higher expected financial payoffs than the costs of implementation for all (but has indirect differential benefits for supporters of only one group of voters) versus not enacting the reform and everyone receiving lower payoffs. We find that when the groups are polarized ideologically, subjects are less likely to vote for reform when informed that another political group would differentially benefit, and more likely to support reform should their own group benefit more. When subjects are told that one group will differentially benefit from reform, they are significantly more likely to explain their vote as being influenced by their own group membership. In contrast, we find that when subjects are organized into nonpolarized groups, group membership predicts reform support less, and when there are no differential benefits for a particular group, the effect of membership on support is significantly reduced. Hence, we find that the effect of polarization on support for common value reform is contingent on the existence of indirect differential benefits and the degree of ideological polarization of the groups who receive those benefits.

    Original languageEnglish (US)
    Pages (from-to)883-901
    Number of pages19
    JournalJournal of Economic Behavior and Organization
    StatePublished - May 2021


    • Experiment
    • Political polarization
    • Reform
    • Voting

    ASJC Scopus subject areas

    • Economics and Econometrics
    • Organizational Behavior and Human Resource Management


    Dive into the research topics of 'Polarizing information and support for reform'. Together they form a unique fingerprint.

    Cite this