About the Australian Candidate Study
The Australian Candidate Study (ACS) surveys are conducted in parallel with the AES surveys of voters, and have been conducted at each federal election since 1987, with the exception of the 1998 election. All but the 1987 study have been funded by the Australian Research Council.
In 1987, all candidates for the House of Representatives and Senate were sampled. Since 1990, the surveys have been restricted to samples of all major party candidates, identiﬁable Green and other environmental candidates, and candidates of parties that were expected to win 10% or more of the ﬁrst preference lower house vote. This restriction is designed to cut ﬁeldwork costs, since about half of the total number of candidates at any federal election are minor party or independent candidates, almost all of whom lose their deposits.
The ACS methodology
The questionnaires are mailed to candidates about one week after the election. The mail-out/mail-back methodology is similar to that used for the voters' survey, with the exception that a letter of introduction from the candidate's political party is usually included. The questions are based on the voters' survey, but an extensive range of additional questions are included about the candidate's political background and experience, and about the election campaign itself.
In reporting results for the candidates as a whole, it is necessary to restrict the estimates to major party candidates so that the views of minor party candidates (who heavily outnumber major party candidates in our surveys) do not bias the results. In reporting party differences, results are shown for Liberal-National, ALP, Australian Democrat and Green candidates.
|Australian Candidate Study Sample Details, 1987-2016|
|Election candidates||Australian Candidate Study|
|House of Representatives||Senate||Total||Total||Valid responses||Effective response (%)|
The 1987 and 2016 elections were double dissolution elections for the Senate. Other elections are half-Senate.
The response rate is estimated as valid responses/total contacted.