[...] One of the participants pointed out that the use of this procedure meant that it was less likely that a vote would be wasted than in a conventional majority vote (FPTP), as had been used previously, and therefore the voters had less reason to liaise among themselves on voting for a particular candidate. This applies even more in the case of approval voting than to this narrow variant of it. Approval voting also makes it more difficult for voters to liaise among themselves on voting for a particular candidate because it is necessary not only to arrange to put a cross beside the name of one specific candidate, but also to agree not to put a cross beside the names of others.

In a certain sense, approval voting lies midway between a conventional majority vote and sequential choice. The voter is able to express his preferences more accurately than in a conventional vote, but not with as much precision as in sequential choice. Whatever method is used, the question may arise as to whether the voter should follow his conviction or vote to achieve the best result. In a conventional election, people are often warned not to waste their vote on a hopeless candidate, but rather to vote for someone who is more likely to have a chance of winning and so keeping out a third candidate who is regarded as the worst possible choice. In such a case, a voter who ranks the candidates in the order A,B,C will then vote for B. With approval voting, the voter can also put a cross against the name of a second candidate, even if he regards him as being a much poorer choice than his first preference, if it seems likely that by doing so it will prevent the election of a third candidate whom the voter finds even less attractive. A voter ranking the candidates in the order A,B,C would then put crosses against A and B. Voting in this way can, in fact, result in the election of B by one vote more than A, a situation that can also arise with sequential choice.

In this connection, it may be instructive to examine the appeal made to the voters in the presidential elections in Iceland in 1980, in which Vigdis Finnbogadottir was elected with 33.8% of the vote, Gudlaugur Thorvaldsson received 32.3%, Albert Gudmundsson 19.8% and Petur Thorsteinsson 14.1%. People had been urged not to waste their votes on outsiders. A poll taken just before the election showed that Albert and Petur had the smallest support; nevertheless, a third of the voters voted for them.

Whether it is realistic to expect to achieve one’s aims by voting against one’s convictions depends on the circumstances. It is certainly harder to be sure in advance about the relative standing of the candidates under approval voting than with a majority vote, since it is likely that there will be more candidates, and if sequential choice is used it is even harder because rankings can vary so greatly from voter to voter.[...]