An election outcome reflects institutional, behavioural and attitudinal influences. We set out a model showing it is a function of the electoral system, the offices at stake and the number of parties competing as well as the choices of voters and the level of turnout. Therefore, any attempt to estimate the impact of increased turnout on an election outcome must go beyond a comparison of the party preferences of voters and non-voters. This paper presents a model which integrates six different types of influences that collectively determine election outcomes. It demonstrates empirically that maximum turnout falls well short of 100 percent turnout. It also shows the effect of proportional representation and multiple parties in reducing the net benefit that any one party could expect from increased turnout and the inadequacy of using shares of the popular vote to predict increased turnout effects in the United States. It leaves open the normative debate between advocates of civic participation and the libertarian value of being free not to vote.