A few weeks ago, a Guardian leader stated that ‘Britain is about to experience its first direct-marketing campaign’, claiming in passing that the ‘result will diminish political debate’ and that ‘British politics is being shaped by the need to appeal to a certain type of voter in a certain type of constituency’.
Of course, political parties will always make special efforts to identify individual voters who are uncertain who they would vote for and concentrate on seeking their support. Nor was there ever a golden age when any political party did not concentrate its organisational effort in those seats that were likely to be most closely contested. Such strategies are necessary under most electoral systems and are perfectly legitimate and healthy.
The Guardian’s assumption that more sophisticated targeting is inevitably honing down that target audience to a homogenous core is, however, completely erroneous. If anything, the opposite is the case. Indeed, it is arguable that the next election will see a wider and more diverse electoral battleground than ever before, with burgeoning numbers of electors finding themselves identified as key swing voters living in marginal seats.
Such was the scale of victories for the Conservatives in the 1980s and Labour in 1997 and 2001 that the front line of the electoral contest has successively penetrated far into the historical heartlands of each party. In the forthcoming election, the Conservatives need to make 158 gains for a bare majority, and have openly declared that they have targeted 187 key seats. Add to these those that the Liberal Democrats are seeking to gain – let alone the nationalists, George Galloway and Robert Kilroy-Silk – by-election gains and defections, and well over 200 seats are theoretically up for grabs. Of the 628 constituencies to be contested in England, Wales and Scotland in the next election, 285 (45 per cent) have changed hands at least once in the last 25 years. The days of the two great immovable blocs of seats held by the major parties alongside a minority of perpetual marginals are long gone.
As for the assertion that parties now pursue ‘a certain type of voter in a certain type of seat’, that is immediately undermined by a glance at a selection of the ten most marginal Labour-held constituencies. What ‘type of voter’ is common to, for example, South Dorset, Braintree, Cardiff Central, Shipley and Dumfries and Galloway, comprising as they do elderly seaside towns, upmarket suburbia, commuter belt, market towns, sparsely populated rural districts and densely populated inner-city, spread across most parts of the country?
Even within those seats the old assumptions about elections being decided by a minority of ‘swing’ voters – mainly young to middle-aged homeowners in the usual stereotype – is, of course, now hopelessly outdated. Increasingly we are seeing unstable voting behaviour spreading into different voting groups – look for example at the huge swings in many seaside towns in 1997, evidence perhaps of the increasing importance of the ‘grey’ vote, of which Age Concern among others are constantly reminding us.
The 2001 election also introduced the new dimension of turnout into an already complex mix. No longer is it enough, if it ever was, for Labour to pocket the DEs, write off the ABs and throw everything at the C2s. The choice now may be not just between Labour and Tory but between Labour and nothing. The motivation of the reluctant voter is almost as important as the persuasion of the switcher.
With many of the old certainties transformed into new uncertainties, it is not surprising that political parties are turning to technology to try to introduce some ‘science’ into the art of campaigning.
In recent months, Conservative co-chairman Liam Fox has missed no opportunity to advertise the Tories’ use of ‘Voter Vault’, the latest in a series of software packages and databases that have been introduced supposedly to revolutionise the practice of identifying and targeting the message to swing voters. Their common characteristic is the presumption that it is possible to work out from extraneous data which voters are most likely to shift their vote without the need actually to speak to them. Thus various publicly and commercially available data are analysed and matched to political information in the hope that some telling and unforeseen pattern will emerge. Fox is fond of explaining how Voter Vault revealed that in the USA there are apparently very few Volvo drivers who vote Republican.
In the technological and information age, the search for a quick fix of this sort is irresistible. Surely out there somewhere there must be some political equivalent of the DNA code that explains why people choose to vote as they do, if only we could uncover it? Maybe there is. More likely there are a number of pointers, common patterns and trends that can be surrogates for political preference, many of which are obvious and others of which are undetectable. The problem with any project like Voter Vault is that it is entirely reliant on the content of the data that are available, much of which are either not matched to individuals or completely irrelevant to political behaviour. Eventually you still have to use political judgement to decide what may be significant and what is not.
Even if data can be shown to relate in general to political choice, its result is simply to raise the level of probabilities about a voter rather than to say anything definite. There are Republican Volvo drivers out there even if there aren’t many of them. Against this, the low-tech option of having actually spoken to that voter about his or her own views and intentions assumes a considerably higher value. Voter Vault and its ilk are essentially a poor substitute for old-fashioned canvassing.
It has never been the case that the outcome of elections could be reduced to the decisions of a small close-defined ‘type of voter’, but it is less true now than it has ever been. The breakdown of many of the traditional sources of partisan loyalty has inevitably led to some erratic and potentially unstable trends and to huge swings – between Conservatives and Labour in 1997 and in the 10 percentage point drop in turnout in 2001.
One of the features of 1997 and 2001 was the remarkable efficiency with which Labour and Liberal Democrat votes were distributed in a manner that maximised the number of seats the Conservatives lost. Some pundits describe this as ‘tactical voting’, and are now arguing that Labour is vulnerable in the next election to ‘tactical unwind’. So people who are really Lib Dems but who backed Labour to beat the Tories in the last two elections are no longer able to stomach the compromise with their real beliefs and stick with the Liberal Democrats.
It is plainly the case that eight years into government the Liberal Democrats may hold some appeal for those who disagree with some aspect of government policy and may be toying with the idea of casting a protest vote. There is, though, little evidence that they include significant numbers of ‘natural’ Liberal Democrats, as opposed to people who have no strong affiliation and often end up voting for them. The majority of potential Lib Dem supporters remain hard-pressed to name a single Lib Dem policy, let alone describe any philosophy that they ‘stand for’. Again, they comprise a fragmented, but electorally quite crucial, group.
So the Guardian’s picture of a sterile, technologically driven election campaign focused on a small minority of identikit swing voters in middle England is very wide of the mark. Instead, the next election promises to be a complex and responsive campaign in which, yes, political parties will try to present their message in a way that is relevant and appropriate to their audience, but in which they also try to reflect the diverse character of that audience.