How the count works
The alternative vote — known in Australia as preferential voting and in some other jurisdictions as instant-runoff voting — fills a single seat by asking voters to rank the candidates in order of preference. A voter marks 1 against their first choice, 2 against their second, and so on, either to all candidates or to as many as the rules require.
On the first count, only the first preferences are tallied. If any candidate holds an absolute majority — more than half of all valid votes — that candidate wins outright. Where no majority exists, the candidate with the fewest first-preference votes is eliminated, and every paper that named them first is transferred to the next-highest-ranked candidate still in the race. The new totals are checked again; if a majority has formed, the count ends, and if not, the next-weakest candidate is eliminated in the same way. This continues until one candidate crosses the majority threshold.
The method has a structural consequence that distinguishes it sharply from first past the post: a winner must clear fifty percent of votes remaining in the final round, rather than merely outrun a divided field. That requirement is the mechanism's whole point, and almost everything else about how campaigns behave follows from it.
What it does to strategy
In a plurality system, a vote cast for a minor candidate risks being wasted entirely if that candidate finishes third or lower. Voters who worry about this often abandon their genuine preference and vote for the lesser of two major evils — what analysts call "strategic" or "tactical" voting. The alternative vote largely neutralises this pressure, because a vote for a minor candidate is not discarded when that candidate is eliminated: it travels to the next preference. A voter can rank a candidate they prefer without forfeiting influence over the final choice between the two strongest contestants.
This changes what a rational campaign looks like. A candidate strong enough to win but not strong enough to lead on first preferences can still accumulate transferred votes and win overall. Australia's House of Representatives elections have repeatedly illustrated the pattern: a candidate who runs second on first-preference totals can overtake the leader once transfers are distributed, without either candidate needing a majority of first preferences. The incentive structure shifts: a campaign that aggressively alienates supporters of minor parties to prevent them from transferring is punishing itself.
There is, however, a documented anomaly. Under some preference distributions the alternative vote violates what is called monotonicity — the intuitive expectation that gaining more first-preference support can only help a candidate. In theory, a candidate can lose by attracting votes that, in a different preference configuration, would have changed the elimination sequence in their favour. This occurs only under specific conditions and is uncommon in practice, but it is real, and worth naming clearly: the alternative vote is not monotone.
Scope and limits
The mechanism is used for House of Representatives constituencies in Australia and for most state lower houses, for Irish presidential and certain other single-seat elections in Ireland, and for the election of the leaders of several political parties in the United Kingdom and elsewhere. A 2011 United Kingdom referendum on replacing the House of Commons plurality system with the alternative vote produced a substantial majority against the change. The vote settled the institutional question in that country for a generation, but it did not affect the mechanics.
Two limits are worth keeping distinct. First, the alternative vote is a single-member method: it allocates one seat per district and cannot itself produce proportional representation across a legislature, because proportionality depends on district magnitude. A party whose votes are geographically dispersed may win many second-preference transfers and still carry few seats. Second, the count is genuinely more complex than a plurality count — it requires tracking paper-by-paper transfers rather than simple totals, which demands more from counting staff and takes longer. Australia manages this through experienced returning officers and well-established procedures; newer adopters have had variable results in managing the administration.
The alternative vote answers a specific problem — ensuring that a winner commands majority support in their own seat — and does so reliably. What it cannot do is convert that seat-level majority mandate into a proportional share of seats for any party across an entire chamber.
In theory, a candidate can lose by attracting votes that, in a different preference configuration, would have changed the elimination sequence in their favour.
