Reasons for policy experimentation that have nothing to do with selection bias

Cyrus Samii

    Research output: Contribution to journalComment/debate

    Abstract

    The conventional case in favor of policy experimentation focuses on how randomization controls selection bias. This is undoubtedly important. This essay focuses on additional benefits from experimentation that are completely distinct from controlling selection bias. These benefits derive directly from the fact that experimentation involves actively intervening to assign policy treatments. Experimentation “puts manipulability to the test” in ways that passive observation does not, and it allows for deeply engaged learning about policy formulation and implementation that ex post analyses miss.

    Original languageEnglish (US)
    Article number104825
    JournalWorld Development
    Volume127
    DOIs
    StatePublished - Mar 2020

    ASJC Scopus subject areas

    • Geography, Planning and Development
    • Development
    • Sociology and Political Science
    • Economics and Econometrics

    Fingerprint Dive into the research topics of 'Reasons for policy experimentation that have nothing to do with selection bias'. Together they form a unique fingerprint.

  • Cite this