13NFL1-Compulsory Voting Page 66 of 163 www.victorybriefs.com PREVIOUS STUDIES SUFFER FROM SEVERE LIMITATIONS. Anthony Fowler
– 2013
Department of Government, Harvard University. Electoral and Policy Consequences of Voter Turnout Evidence from Compulsory Voting in Australia Quarterly
Journal of Political Science, 2013, 8: 159
–182. As a result
of methodological limitations, no previous study adequately addresses the primary question of interest.
Confounding variables,
reverse causation, and model misspecification may bias the correlational and survey studies. The most compelling causal evidence on the effects of turnout comes from studies of small shocks to turnout. These studies address a separate question about the effects of marginal
changes to voter turnout, but they do not assess the effects of near-universal turnout. If marginal voters are unrepresentative of the
entire population of nonvoters, then these studies do not speak to the counterfactual question at hand. To determine what
would happen if everyone voted, we need a policy change that closely mimics
the ideal counterfactual — one where almost everyone is brought to the polls.
