Instant-runoff Voting - Resistance To Tactical Voting

Resistance To Tactical Voting

The Gibbard–Satterthwaite theorem demonstrates that no voting system using only the preference rankings of the voters can be entirely immune from tactical voting unless it is dictatorial (there is only one person who is able to choose the winner) or incorporates an element of chance. This implies that IRV is susceptible to tactical voting in some circumstances.

Nonetheless, IRV is considered one of the less-manipulable voting systems, with theorist Nicolaus Tideman noting that, "alternative vote is quite resistant to strategy" and Australian political analyst Antony Green dismissing suggestions of tactical voting. James Green-Armytage finds the alternative vote to be second most resistant to tactical voting among the methods tested, only beaten by a class of AV-Condorcet hybrids, although the alternative vote resists strategic withdrawal by candidates less well.

By not meeting the monotonicity, Condorcet winner, and participation criteria, IRV permits forms of tactical voting when voters have sufficient information about other voters' preferences, such as from accurate pre-election polling. FairVote mentions that monotonicity failure can lead to situations where "Having more voters rank candidate first, can cause to switch from being a winner to being a loser." That assessment is accurate, although it only happens in particular situations. The change in lower candidates is important: whether votes are shifted to the leading candidate, shifted to a fringe candidate, or discarded altogether is of no importance.

Tactical voting in IRV seeks to alter the order of eliminations in early rounds, to ensure that the original winner is challenged by a stronger opponent in the final round. For example, in a three-party election where voters for both the left and right prefer the centrist candidate to stop the "enemy" candidate winning, those voters who care more about defeating the "enemy" than electing their own candidate may cast a tactical first preference vote for the centrist candidate.

The 2009 mayoral election in Burlington, Vermont provides an example in which strategy theoretically could have worked but would have been unlikely in practice. In that election, most supporters of the candidate who came in second (a Republican who led in first choices) preferred the Condorcet winner, a Democrat, to the IRV winner, the Progressive Party nominee. If about 20% of the backers of the Republican candidate had insincerely raised the Democrat from their second choice to their first, the Republican would have dropped from first to third in first choices, and the Democrat would then have won the instant runoff. But given that the Republican was a strong candidate who nearly won in the instant runoff, few of his backers would have risked giving up on his candidacy based on a chance, unknown before the fact, to elect the compromise Condorcet winner.