Exposing the ‘Millbank myth’ of political marketing success

One powerful legacy of the 1997 general election campaign, and of the reinvention of the Labour Party which preceded it, has been what might be called the “Millbank Myth”.

According to the conventional wisdom, the Labour Party’s staggering success in the polls was partly the result of its wholesale adoption of the techniques of “strategic communications” – of focus groups, market research, positioning statements, direct mail and advertising, as well as traditional news management – familiar from the world of commercial marketing but hitherto unknown, or at least under-exploited, in politics.

The party was introduced to such techniques initially by that modern Machiavelli, Peter Mandelson, and then by the pollster Philip Gould and his colleagues in the Shadow Communications Agency. The plans were co-ordinated in the run-up to the election from Labour’s campaign headquarters (now, indeed, the party HQ as well) in Millbank Tower.

In the words of a new academic study of the election campaign, “New Labour’s victory in 1997 was widely regarded as a textbook triumph of packaging over politics, spin over substance and image-building over ideology”. It wasn’t “The Sun wot won it” (though the paper’s decision to back Labour played a significant part, in the conventional reading of events), but the spindoctors wot swung it. Blair’s triumph was also a victory for political marketing.

But now it seems the conventional wisdom may be wrong. That same academic study (On Message: Communicating the Campaign, Sage Publications) concludes that, in reality, Labour’s campaign was not the most effective of the main political parties’, that the efforts of all parties to set the news agenda were subverted by the media’s insistence on pursuing a quite different agenda in their coverage of the campaign, that huge amounts spent on outdoor advertising apparently had little effect and that no newspaper – not even The Sun – managed to change the way its readers planned to vote.

The authors’ academic credentials are impeccable. They include Dr Pippa Norris of the Kennedy School of Government at Harvard, Professor John Curtice of Strathclyde University, and Dr Maggie Scammell of the Department of Politics at the LSE. Their findings are bolstered by a great many charts and much statistical number-crunching, by a “content analysis” of the media during the campaign, and by an experimental study in which groups of voters were shown videos of television election coverage and then questioned about them.

Yet in some ways their conclusions are remarkable. For one thing, the 1997 election was by a considerable margin the most expensive ever. Between them the parties spent an estimated &£56.4m on the campaign. In real terms, that was more than double what they spent (at 1997 prices) in 1992. A lot of the money in 1997 went on advertising and direct mail. The Conservatives spent an estimated &£11.1m on posters (far too much, Cecil Parkinson later thought), Labour &£4.8m. The Tories spent &£3.2m on newspaper ads, Labour &£900,000. Both spent about &£2m on direct mail and &£500,000 on party election broadcasts.

The Liberal Democrats, by contrast, spent a paltry &£2.3m in total – less than a tenth of their larger rivals. Yet the Lib Dems were the only party to see their support, as measured by the opinion polls, actually rise in the run-up to the election. And it was the Lib Dems, in the view of the study’s authors, who ran the most effective campaign – defined as the one which stuck most closely to the party’s ideal agenda of social welfare subjects like education, as well as being the most positive and the least prone to negative attacks on opponents.

It was of course easier for them: they were almost completely ignored in the cut and thrust between Labour and Conservative, making it simpler for them to press on regardless. But then, like the two big parties, they suffered from the fact that the news media largely ignored what they were trying to say.

Comparing what the parties were actually saying in press releases about substantive issues such as the economy, welfare and education with what actually appeared in newspapers and television, the study found that two-thirds of all stories weren’t about the issues at all but about the conduct of the campaign, opinion polls and the party leaders. The media, the study concludes, followed its own “media logic”, running stories they thought would grab viewers and readers, rather than following the priorities of any party.

This is disputed by David Draper, a Millbank insider during the election, who maintains that press releases give little indication of what’s really important to the parties, and that actually Labour was very successful in getting the news media to run with one of its most important messages, namely that the Tories were “sleazy, sleazy, sleazy”. (In which case, it is tempting to ask why they bothered with so much of the other stuff.)

Overall the study will make sobering reading for the spindoctors. It suggests that half of their effort and expenditure in 1997 was wasted – though, like Lord Leverhulme, it may not be immediately clear which half. And it suggests that, however successful the techniques of advertising and marketing may be in the commercial world, it’s far from clear that they work as well when applied to political brands.