Why wonâ€™t the pre-1997 polls comparison die?
Michael Thrasher gave an interview for Sky News yesterday afternoon in which he built upon the article written for their website, inferring that the election is heading towards a hung parliament. In it, he made various claims and inferences that really shouldnâ€™t be allowed to stand and which I find quite astonishing from such a senior expert in the subject. To summarise:
He claimed that the ToriesÂ are struggling toÂ break through 40%. In fact, of the 19 polls since the end of the conference season, only two have the Tories sub-40. The rest are in the 40-45 band.
He claim that the Tories arenâ€™t as far ahead in the polls as Labour were pre-97. This is such a basic howler that it is the sort of thing which undermines credibility as an expert. As has been frequently pointed out, not least on this site, the methodology used in polling has changed significantly since 1997. Only ICM can be compared on a like-for-like basis, and using that sequence, the Toriesâ€™ lead is very close to where Labour was at the same time in the parliament – in the mid- to high-teens. Laughably, Skyâ€™s Poll Tracker – linked in the article – gives the May 1997 Labour figure as 59%. Labour actually polled about 43% at the general election that month.
He claimed the Glasgow NE result showed little enthusiasm for the Tories. There may be something in this but all it showed was minimal enthusiasm in Glasgow NE – a trend weâ€™ve already seen in two other similar Scottish seats, one next door and another in Glenrothes. Even so, in government-held seats where the principal challenge comes from a third party, the main opposition often gets squeezed. Labour lost their deposit in both theÂ Christchurch and Newbury by-electionsÂ during the 1992-7 parliament when the Lib Dems powered toÂ gains from the Tories; Blair still won his landslide.
He claimed that the Tories arenâ€™t doing well enough in â€˜realâ€™ elections. Well enough is a relative game. In marginals like Crewe & Nantwich or Norwich North – the sort of seats that will determine the election result – Labour received a resounding raspberry in contrast to at least qualified support for the Tories. Itâ€™s notable how badly the Lib Dems are doing: if it were all down to Labour unpopularity, the Lib Dems would have been serious challengers in both by-elections. Besides, the real story of the last year has not been Tory strength but Labour weakness, most notably their sub-16% share in the European elections.
There was a strong inference that the Tories need to be matching Labourâ€™s polling pre-97. However, Cameron doesnâ€™t need a 177-seat majority (though he no doubt wouldnâ€™t turn it down if offered). 50 seats is enough. For that matter, Labour didnâ€™t need what they got: itâ€™s not a meaningful benchmark. Of more relevance is thatÂ because of the distortions in the system, the Conservatives do need to be further ahead to achieve the same. Even so, Thrasherâ€™s claim in his article that a 12% lead would produce a majority of just six seems bizarre unless he expects turnouts in all of Labourâ€™s safe seats to fall as in Glasgow NE. Also, if anti-Tory tactical voting weakens, so the lead the Tories need to reach any particular milestone falls.
He repeats the â€˜Economy stupidâ€™ myth In 1997, the Conservatives were ahead on the Economic question but still got trounced. Likewise, fluctuations in Labourâ€™s deficit over the last two years have borne little relation to the state of the economy. Both facts counter the suggestion that an economic recovery will of itself have much impactÂ on Labourâ€™s fortunes.
Glasgow North East was a very good result for Labour – to hold the seat by more thanÂ 8000 after losing the near-identical Glasgow East last year was a fine achievement. However, while it may point to a firming of the Labour vote in their heartlands, thatâ€™s not where the election battle will be won and lost. I for one will continue to place more faith in the opinion polls.