BZÖ and I together at last (or: Dewey Defeats Truman)
by Yussi Pick
Late Jörg Haiders almost late party, the Alliance for the Future (of?) Austria (that really sounds even more pretencious in English), demanded on Tuesday that it should be forbidden to publish polls three weeks before election day. While the argumentation for that new stance is darn stupid, the general idea isn’t. BZÖ argues that the polls in last Sundays elections in Carinthia mobilized for the social democrats (SPÖ), that for polls/their publishing should be forbidden. The BZÖ won the elections by 45ish over 28ish percent. Polls in the last weeks before the election showed a close horserace between both parties at around 30%. BZÖ is arguing I don’t see how that close race couldn’t help both parties – and obviously helped only one after all, but the “evil liberal media” argument works in Austria as well.
But back to the proposition: Polls are influencing voting behaviour to an extend where it is necessary to intervene in media rights. I think so. Since I am eligible to vote, there was not one election where the pollsters hit the target within MOE. In 2001 in the elections for Vienna the polls showed the SPÖ at around 42% far away from the absolute majority at 47% that they reached on election day. In the national elections of 2002, the closer Sunday got, the further away the polls went from the actual result: At a sample of 1000 people, which is about double the usual sampling size, IMAS missed the target by 6%, claiming that ÖVP and SPÖ were in a MOE race at 37:33. ÖVP won the election with 42 over 36%.It’s not that those examples show polls way off their margin of error, but they are still a couple of points off (1) and much more important: they couldn’t predict a general trend in the elections. Should they? No, but since they pretend they do, someone should call their bluff.
In the 2008 national elections, a lot of people were thinking of voting Liberales Forum in order to have a 6th party (or rather 3rd center-social liberal) party in parliament. Polls showed LiF up at around 4%, which is the make-it-point into parliament. Those people wouldn’t have thought of committing to this pragmatic compromise, if it hadn’t been for polls who showed LiF could make it. I’m convinced they would’ve made it with those votes from unlikely LiF voters, attracted by the poll (2).
Polls influence voting behaviour. Voters vote, don’t vote or vote something else, because parties look well/not so good in polls. Of course it’s not a mainstream phenomenon, but it is significant. It’s not the only reason for strategic (ergo: not idea driven) voting, but it encourages it. It’s not the reason for bad (horse race) media coverage, but it helps distracting journalists.
Banning polls from newspapers in the last 2-3 weeks of the elections really doesn’t hurt anyone. Parties can still have internal polls, that are usually much more significant anyways. Newspapers have to find something else to write about, bad case scenario: it’s gonna be the chancellor’s dog; good case scenario: something that matters. And voters can follow their gut feeling without being distracted by voting strategy. Or was it issues they were voting on?
(1) Actually 2 points off their 4 point MOE is A LOT.
(2) That they didn’t in the end was “only” because their chairperson at the time was involved in a lobbying affair.