Media Briefings


  • Published Date: February 2014


Voter turnout is generally higher under proportional representation (PR) when the election is relatively uncompetitive. But in highly competitive elections, it is a ‘first-past-the-post’ or ‘winner-takes-all’ voting system that generates a higher turnout. These are the conclusions of research by Professors Helios Herrera, Massimo Morelli and Thomas Palfrey, published in the February 2014 issue of the Economic Journal.

Does proportional representation (PR) lead to higher voter turnout compared with winner-takes-all (WTA) elections? This question has been debated for years among academics and politicians, as well as in the press, with no clear resolution.

The new study demonstrates that in fact there is no simple yes-no answer. Voter turnout will actually be lower in PR systems if the electorate’s support of the competing parties is relatively equally split. But the result flips and PR leads to higher turnout for elections with a clear favourite or a dominant party.

The researchers establish this comparative result using a mathematical theory of voting that can predict turnout across a wide range of different electoral systems. They then corroborate the finding in a laboratory economics experiment.

The study’s main result implies that the effect of institutional differences on voters’ participation between majority and minority parties depends crucially on the distribution of voters’ preferences. The effect cannot be unambiguously predicted without additional contextual knowledge about the specific election or party system.

This is the most important of several findings reported in the study about the differences between PR and WTA voting systems. The mathematical theory develops a number of additional results that provide fresh insights into exactly why we should expect these particular differences in voter turnout between the two electoral systems.

One key finding is that the incentives for voters to turn out or for parties to try to mobilise voters are extremely sensitive to the competitiveness of the race in WTA systems. But the incentives are much less sensitive to the competitiveness of the race in PR systems.

The reason for the difference in sensitivity is that in WTA systems, the value of an additional vote for a party or candidate only matters when that vote changes the outcome of the election. That value is extremely high, because it would actually change which party wins power.

But the likelihood that a single vote will tip the outcome from one side to the other is extremely low, and in lopsided contests, it is essentially zero. Therefore, the ‘expected value’ of a vote in WTA elections is relatively high only if the race is expected to be very close.

In PR systems, an additional vote for a party is always valuable to the party because it makes a difference in the party’s share of the vote even when the election is not close. Of course, this marginal value of one vote is always very small, but the key is that it is always a factor regardless of the competitiveness of the election.

The figure below is an illustration of the different effect of competitiveness under the two systems. The figure also illustrates what the authors call ‘the competition effect’, which is present in both systems: turnout is always higher in more competitive races.

The study also reports the results of a laboratory economics experiment that was specifically designed and conducted to explore the theoretical predictions of the mathematical model. The results are broadly supportive of the main theoretical hypotheses about comparative voter turnout:

· Turnout generally is higher in PR systems when the election is relatively uncompetitive; and the reverse true in highly competitive elections.

· Voter turnout is more sensitive to competitiveness in winner-take-all elections than in PR elections.

· The competition effect is present in both systems.


Notes for editors: ‘Turnout and Power Sharing’ by Helios Herrera, Massimo Morelli and Thomas Palfrey is published in the February 2014 issue of the Economic Journal.

Helios Herrera is at HEC Montreal. Massimo Morelli is at Columbia University. Thomas Palfrey is at the California Institute of Technology.

For further information: contact Massimo Morelli via email:; or Romesh Vaitilingam on +44-7768-661095 (email:; Twitter: @econromesh).