Anthony Fowler writes:
Despite extensive research on voting behavior, there is little evidence connecting turnout to tangible outcomes. Would election results and public policy be different if everyone voted? The adoption of compulsory voting in Australia provides a rare opportunity to address this question, because Australian states implemented compulsory voting policies at different times between 1914 and 1941. First, I [Fowler] collect two novel data sources to assess the extent of turnout inequality in Australia before compulsory voting. Overwhelmingly, wealthy citizens voted more than their working class counterparts. Then, exploiting the adoption of compulsory voting as a natural experiment, I find that the policy increased voter turnout by 24 percent which in turn increased the vote shares and seat shares of the Labor Party by 7 to 9 percent. Results suggest that democracies with voluntary voting do not accurately represent their citizens. Increased voter turnout can dramatically alter election outcomes and resulting public policies.
Fowler sent me this after seeing my recent online article. on voters and nonvoters in the New York Times. Lots of commenters (on the blogs and at the Times) mentioned Australia, but this is the first thing I’ve seen that compares different Australian states.