The impact of constituency campaigning in the 2001 general election

Among the most interesting developments in British elections during the 1990s was the increased attention given to constituency campaigning by the major parties.Having been eclipsed by television and the national campaign since the 1960s, and consistently downgraded in importance by academic analysts, local campaigning was revitalized.There were a number of reasons for this, including the development of personal computers, the campaigning possibilities opened up by telephone canvassing and an increasingly professional approach to campaigning on the part of staff at party headquarters.The revival of the parties' enthusiasm for constituency campaigning was matched by, and in part also encouraged by, renewed interest on the part of academics whose work significantly revised the prevailing orthodoxy about the impact of local campaigning.Far from being a 'ritual', undertaken by party workers out of habit, a series of studies by three separate groups of researchers provided clear and persuasive evidence that local campaigning affected election outcomes.Measuring the strength of campaigns in various ways, Seyd and Whiteley, Johnston, Pattie and colleagues, and Denver and Hands demonstrated that in recent elections variations across constituencies in the intensity of campaigns mounted by the parties were associated with variations in their electoral performance (see, for example, Denver and Hands, 1997, 1998; Denver, Hands and Henig, 1998; Johnston, Pattie and Fieldhouse, 1995; Whiteley and Seyd, 1994).There are some areas of disagreement between these research teams -on whether Conservative campaigning has been as effective as that of the other major parties, for example, and on whether campaigns by incumbent parties were as effective as those of challengers -but there is substantial agreement on the central point: at least as far as Labour and Liberal Democrats are concerned, constituencies which mount strong campaigns generally achieve better results than those whose campaigns are weaker.It is worth emphasising, however, that most of the evidence supporting this conclusion has come from the general elections of 1992 and 1997.In both of these elections the electoral context was similar -the Conservatives were in government and on the defensive.In advance of the elections it was widely anticipated that Labour would improve its position and, on past form, the Liberal Democrats would also have been expected to benefit with an unpopular Conservative government in office.Denver and Hands' work suggested that in 1992 and 1997 more effective Labour and Liberal Democrat campaigning clearly improved their performance but that this was not the case for the Conservatives.Although their results for 1997 were more ambiguous than those for 1992, it appeared that in both elections Conservative local campaigning did not significantly improve Conservative performance. 1 In 2001, however, the electoral context was different.Labour was now in government and on the defensive -concerned mainly to hold on to seats rather than make new gains -while the Conservatives hoped, and probably expected, to make some advance from the low point of 1997.The Liberal Democrats were also in a novel situation.Having made an unexpectedly large number of gains in 1997, they had more than the usual handful of seats to defend.This changed context clearly affected the targeting strategies of the parties -the Conservatives and Labour, in particular, targeted many more constituencies than normal.2 Whether the changed context would affect the ability of local campaigns to mobilize voters remained to be seen, however, and that is the question addressed here.All parties now attempt to focus their local campaigning efforts on 'key' or 'target' seats and an initial general indication of the effects of campaigning can be obtained by comparing the parties' performances in their targets as opposed to other seats.Table 1 shows the overall change in vote shares for each of the parties in three categories of seat. 3 At first sight, targeting by the Conservatives does not appear to have been very successful -the increase in their share of the vote in target seats was smaller than in the seats that they already held and almost the same as in those that they were unlikely to gain.On the other hand, Labour's worst performance by far was in the seats that they already held and did not target, while the decline in their vote share was smallest in their targets.The Liberal Democrats also did considerably better in their relatively small number of targets than in their safe or unlikely seats.This is at best indirect evidence of the effects of campaigning, however.In what follows, we discuss problems in measuring the impact of campaigning on electoral outcomes and present more direct evidence about the impact of constituency campaigning in the 2001 general election.