Assessing vote trends in the Mayo by-election

By-elections offer a tempting opportunity to track voter trends in politics. However there are variables which make it dangerous to compare raw data from previous elections with the raw results from a by-election. One of these variables is that voter turnout tends to be much lower at by-elections. Another variable is that there will tend to be more candidates at a by-election.

Both of the above variables applied to the 2008 Mayo by-election. Voter turnout was down to 74 per cent from a turnout of 96 per cent in the previous 2007 federal election. The number of candidates rose from 6 to 11.

The significance of more numerous candidates at a by-election is that there are simply more individuals competing for a finite number of first preference votes. Thus if the number of candidates doubles, then it would be reasonable to expect that the first preference vote share for candidates would also be lower than in a contest with fewer candidates.

How does one therefore compare voter trends as evidenced in two very different electoral contests?

One solution is to compare the actual vote for an individual candidate against the mean first preference vote share for all candidates in that election. A comparative index for a candidate or political party can be derived from the following equation: the actual vote for a candidate divided by the mean first preference vote share for candidates in that election.

This really is an index of how parties/candidates are performing against other candidates, rather than against their own record. Thus this can be used to track voter trends, or rather, the comparative performance of a candidate or party.

In the 2007 federal election, the mean vote share for candidates for Mayo was 13,199. However in the 2008 Mayo by-election, with many more candidates, the mean vote share for candidates was 6,055. An analysis of voting trends, as suggested by the results of the Mayo by-election, follows.

The Liberal Party candidate in the Mayo by-election received 487 per cent of the mean first preference vote share. This compared to 364 per cent of the mean first preference vote share for the previous 2007 federal election.

On these figures the result was a resounding success for the Liberal Party. Yet one could obviously argue that this was due to the fact that the ALP was not standing; although there is no evidence that ALP voters crossed over to the Liberals.

Although the ALP did not stand a candidate, the former ALP candidate did stand as an independent and it is useful for this comparative exercise to treat her as an ALP candidate. The candidate received 220 per cent of the mean first preference vote share in 2007, although in the Mayo by-election this dropped to 29 per cent of the mean first preference vote share.

The above difference is perhaps an indication of how much the ALP brand-name is worth to a candidate. The 29 per cent result is perhaps also an indication of the personal following which the former ALP candidate enjoyed.

The dominant success story of the by-election was the Greens. In 2007 the Greens candidate gained 77 per cent of the mean first preference vote share. However in the 2008 Mayo by-election this jumped to 263 per cent of the mean first preference vote share.