By Jan Corbett
At the 1996 election, 37 per cent of us voted for an electorate candidate from a different party than the one we nominated on our party vote.
At first, the researchers on Waikato University's New Zealand Election Study wondered if it showed a high degree of voter confusion, given
that this was our first try at MMP. In Germany, from where we take the MMP model, only around 16 per cent split their votes.
Instead, the study found a high degree of voter sophistication.
The more educated the electorate the more likely it is to happen. In Wellington Central, said to be the smartest electorate in the country, 54.6 per cent split their vote in 1996 - 46 per cent opting for National on the party vote and 34 per cent deciding they wanted Act leader Richard Prebble as their MP.
Only in Ohariu-Belmont was there a higher level of vote-splitting (69 per cent), but there National, which took 44 per cent of the party vote, did not stand a candidate against United's Peter Dunne, who took 48 per cent of the electorate vote. Because the party vote is the most important for determining which party wins the majority, it is also the sincerest expression of preference.
It is with the electorate vote that strategic games get played, especially when the candidate from the party of your choice has no hope of winning the seat and his or her own party is not investing in that candidate's campaign.
Generally, those who abandon their preferred party candidate for one more likely to win do not stray far along the left-right spectrum. Alliance supporters will go with a Labour candidate, Act people with National.
But in some instances split-voters may simply look for the lesser of two evils. For instance, a traditional left-wing voter living in a conservative seat might vote Labour or the Alliance on the list, but National for the electorate if the National candidate is liberal and personable and his or her win would prevent Act from getting the seat.
