Vote, Alternative

BIBLIOGRAPHY

The alternative vote is an electoral method wherein voters rank candidates, and the winner is determined by a sequential count in which the weakest candidate is repeatedly eliminated until one candidate has secured a majority of the vote. Electoral-law terminology is unfortunately nonuniform, but the alternative vote can be regarded as a special case of the “single-transferable vote” or the “preferential vote,” with the conditions that only one candidate is elected per district, and election requires an absolute majority. Most real-world examples of this system come from Australia, but it has also seen use elsewhere.

An example from the 1952 election in the Canadian province of British Columbia (Table 1) clarifies how it works.

Voters could rank as many of the four candidates as they liked, and the “first count” column shows how many placed each candidate in first place. None of the four had the necessary majority (5,771) immediately, so the last-place candidate, the Progressive Conservative (P.C.), was eliminated. The second-count “transfer” column shows how the 1,690 ballots on which Wright was ranked first were then redistributed: 169 of those ballots had no further rankings, and were thus discarded; 937 of them had Turnbull ranked second; and so on. Following elimination of the candidate from the Co-operative Commonwealth Federation (C.C.F), the Social Credit (S.C.) candidate, Robert Sommers, ultimately won the seat on the third count.

Advocates of the alternative vote often stress that it tends to be more proportional than simple plurality election, wherein voters pick one preferred candidate and whichever candidate wins the most votes wins the seat. In other words, the proportions of votes and seats won by the parties are usually more nearly equal under the alternative vote system. Again, British Columbia illustrates. The 1952 and 1953 elections were held using the alternative vote (“vote” in Table 2 refers to first count), whereas later elections awarded seats by plurality.

Unlike proportional-representation electoral rules, which ensure very close equality between seat and vote shares, the alternative vote merely tends not to exaggerate the support for large parties when turning votes into seats as drastically as does plurality.

One criticism of the alternative vote is that getting higher rankings can actually cause a candidate to lose a seat. The example in Table 3, devised by Steven Brams and Peter Fishburn, demonstrates this surprising trait. Candidates A, B, C, and D are competing for the seat, and the voters fall into four types, according to their preference rankings.

The outcome is a win by candidate A, as shown in Table 4 (since all voters ranked all candidates, exhausted ballots play no role).

Now suppose that the Type-IV voters move A up from last place to first place, so that their rank ordering is A, D, C, B. If all other voters remain the same, the outcome is now shown in Table 5.

Candidate B wins. Candidate A did worse (losing, not winning) because some voters gave him a higher ranking. If voters recognize the possibility of outcomes like this, they have incentives to misrepresent their preferences. Surprisingly, however, all electoral systems are vulnerable to such strategic voting.

SEE ALSO Elections; Electoral Systems; Voting; Voting Schemes

BIBLIOGRAPHY

Farrell, David M. 1997. Comparing Electoral Systems. London: Prentice Hall.

Lijphart, Arend, and Bernard Grofman, eds. 1984. Choosing an Electoral System: Issues and Alternatives. Westport, CT: Praeger.

Brian J. Gaines