Political Marketing

Marketing politics to voters: late deciders in the 1992 British election Bernadette C. Hayes Department of Sociology, Queen’s University, Belfast, and

Marketing politics to voters 127 Received June 1996

Ian McAllister Department of Government, University of Manchester, Manchester, UK. Elections and their accompanying political campaigns may be considered a form of bloodless but serious marketing warfare. Not only must political parties determine the scope and nature of their product, but they must also identify the most effective way of communicating, or selling, its benefits to a target audience. Thus, as in other marketing contexts, political campaigns, or the process of getting elected, is primarily one of marketing or product positioning. As Reid explains: “Political parties, either wittingly or unwittingly, produce a political product for the consumption of voters” (Reid, 1988). Despite the increasing realiZation of the importance and relevance of marketing principles and procedures to the study of British electoral politics, the application and evaluation of marketing techniques had been somewhat amateurish (Butler and Collins, 1994; Kavanagh, 1995; Niffenegger, 1989). In contrast to the USA, British politicians and the general public have remained deeply suspicious of the perceived manipulative effects of political marketing (Clemente, 1992; O’Shaughnessy, 1990). Until recently, not only did all the major political parties shun the widespread use of such practices, but also, despite record-breaking levels in party political expenditure, no systematic marketing strategy has emerged to clearly differentiate either the market product ( political party) or the specific target market (differing segments of the electorate) to which the product is addressed (Scammell, 1995). As far as political parties are concerned, traditional party allegiances and the existence of long-standing ideological differences are still seen as the primary forces shaping electoral opinion. While this simplistic understanding of the electorate may have worked in the past, recent research suggests that it is inadequate (Webb, 1992). In the first place, traditional party loyalties have declined consistently in Britain since the 1960s, and this relationship holds regardless of whether Conservative or The 1992 British Election Study was collected by Anthony Heath, Roger Jowell, John Curtice, Jack Brand and James Mitchell and funded by the Economic and Social Research Council. The data were made available by the ESRC Data Archive at the University of Essex. Our thanks to two anonymous reviewers for this journal for their constructive comments; the usual disclaimer applies.

European Journal of Marketing, Vol. 30 No. 10/11, 1996, pp. 127-139. © MCB University Press, 0309-0566

European Journal of Marketing 30,10\11 128

Labour identifiers are considered ,(Crewe, 1992; Smith and Saunders, 1990). Second, political competition has become increasingly devoid of economic ideology and all the major parties have become indistinguishable on important policy issues. This trend was most marked during the 1992 election (Butler and Collins, 1994). Finally, one important by-product of this phenomenon has been the rise of a new political animal, the unaligned, or “floating voter”. In fact, more so than any other segment of the electorate, it is this group of individuals to which political marketeers have devoted most attention. As Reid puts it: “The battle for ‘share of mind’ among floating voters is as ferocious as any which is sought by consumer good manufacturers for space on supermarket shelves” (Reid, 1988). Despite this concentration of effort and expenditure, the potential impact of political marketing on non-aligned British voters in terms of their eventual electoral choices has rarely been assessed. In this article we examine the background characteristics and voting decisions of these undecided or floating voters. More specifically, adopting a market segmentation approach, both the demographic profile of floating voters as well as their ultimate electoral decisions are empirically examined. The concept of market segmentation was first introduced by Smith (1956), almost four decades ago. Since then, it has become the dominant concept in consumer marketing, although it has had little impact on political marketing. Market segmentation is based on the overt recognition that customers within the market for a given product are not homogeneous. Segmentation rests on three basic premisses: customers are different; these differences influence demand; and segments of the customers can be isolated within the overall market. In other words, market segmentation is the process of dividing a market into distinct groups of buyers with similar requirements (Engel et al., 1972; Wing, 1978). There are two major approaches for market segmentation (Rao and Wang, 1995; Green et al., 1988). One is an a priori approach in which the segmentation and their categories are predetermined or decided prior to the data collection or analysis stage. Traditional segmentation approaches based on geographic or socio-demographic variables fall into this category. For example, in consumer markets, geographic, demographic and socio-economic variables such as age, income, occupation, race, religion, family size, and education, have traditionally been used as the bases for segmentation. The other approach is a clustering-based segmentation design in which the segments are investigated a posteriori by a range of sophisticated statistical techniques such as cluster analysis. In this approach, respondents are placed into groups according to their similarity with those in the same group, and dissimilarity with those in other groups. Much of the political advertising conducted in Britain rests on a simplistic application of the traditional model. Not only are undecided voters specifically targeted as the relevant segment for persuasion, but because of their perceived volatility, their electoral impact through brand switching and product

conversion are taken for granted (Butler and Collins, 1994; Kavanagh, 1995). More specifically, the underlying assumption of this approach is that not only do undecided voters form a single homogenous block, but by concentrating their marketing efforts on these voters and devoting less attention to the other (brand loyal) voters, they are targeting a group of individuals who are most persuaded by campaign pleas and whose predispositions are consequently the most susceptible to modification. We examine some of these assumptions in this paper. Political marketing in the 1992 election The decade of the 1980s is generally considered to be pivotal in the development of political marketing in Britain. When the Conservative Party commissioned the advertising agency of Saatchi and Saatchi in 1978 it made headline news. By the end of the 1980s, it would have been just as noteworthy if a major party had chosen not to engage a professional marketing company. The turning point in this transformation was the highly praised, although ultimately unsuccessful, communications campaign mounted by the Labour Party in the 1987 election. This ensured that by the end of the decade the use of marketing techniques by parties had attained widespread application across the political spectrum (Harrop, 1990; Scammell and Semetko, 1995). The 1992 election campaign represented the logical progression of this trend. Not only did all three major parties (Conservative, Labour and Liberal Democrat) engage private advertising agencies, but many of the techniques and tactics associated with political marketing were much in evidence even before the formal election campaign had commenced (Butler and Kavanagh, 1992; Newton, 1992). Precampaigns (that is, before the start of the official four-week election campaign period) were waged on all sides, with differing effects. Advertising, especially on television via party election broadcasts, became the most prominent weapon in the political communications arsenal, as it had been since the early 1970s (McAllister, 1985). In addition, the vast array of media specialists, advertisers, pollsters and public relations consultants attached to the major parties swelled in number and influence (Kavanagh, 1995; Scammell and Semetko, 1995). All of this activity, of course, had a price. Even by the record-breaking standard of the 1987 election, the estimated expenditure in the 1992 election campaign was a record (Butler and Kavanagh, 1992; Kavanagh, 1995). The Conservatives spent £10.1 million during the actual campaign (compared with £9.0 million in 1987); the equivalent figures for Labour and the Liberal Democrats were £7.1 million (compared with £4.2 million in 1987) and £2.1 million (compared with £1.9 million in 1987), respectively. Thus, for all three political parties, the 1992 general election campaign was by far the most expensive in history. Why did the parties feel the need to expend such vast sums on political advertising during the election? More important, what influence did it have on the election outcome? We address these questions using

Marketing politics to voters 129

European Journal of Marketing 30,10\11 130

the 1992 British Election Survey (BES), a national cross-sectional survey conducted immediately after the general election (Taylor et al., 1992). The growth in electoral volatility The main explanation for the expansion in political advertising is the growth in electoral volatility. Not only has the number of individuals who switch their voting preference from election to election dramatically increased since the 1970s, but this is also true during the election campaign itself (Farrell et al., 1995; Miller et al., 1990). Throughout the 1980s, only half of the electorate who were eligible to vote at two successive elections cast the same vote twice (Miller et al., 1990). A similar, though less dramatic, pattern of turnover is evident among campaign switchers (Clifford and Heath, 1992; Miller et al., 1990). Recent estimates suggest that not only do 24 per cent of the British public leave their voting decision until the start of the election campaign (as compared to 17 per cent in 1964), but swinging voters – individuals who change their intended vote during the final stages of the election campaign – now constitute about one quarter of the British electorate. The data in Figure 1 illustrate the extent of these changes. Although the proportion of voters who decided on their vote more than two years before the election campaign has declined steadily since the 1960s (from 77 per cent in Per cent 80 77

77 70 64 60

60

60

62 60

57

40 25 23

22

22

21

18

17

18

20 12

12

18

11

11

12

1964

1966

1970

14

15

1974

1974 1979 1983

24 16

0 General election

February October

Key Figure 1. When voters decided how to vote, 1964-92

Up to two years During campaign Long time ago Sources: Political Change in Britain Survey (1964-70), British Election Surveys (February 1974-April 1992)

1987 1992

1964 to only 60 per cent in 1992), there has been a dramatic rise in late deciders, or individuals who decide on their vote during the election campaign. Thus, while only one in every ten voters made their decision during an election campaign prior to 1970, by February 1974 this had reached almost one in four. A similar, though less dramatic increase, occurs among individuals who made their decision up to two years before the election campaign, climbing from 11 per cent in 1964 to its current level of 18 per cent in 1992. As far as the British electorate is concerned, then, not only has the proportion of late deciders increased significantly during the past 20 years, but they are now sufficiently numerous to have the potential to determine the outcome of the election. While a number of explanations have been advanced to account for this rise in late deciders, with few exceptions (Clifford and Heath, 1992; Miller et al., 1990), most have emphasized the importance of declining party attachments. Put simply, British voters have become less partisan. In 1964, 93 per cent of the electorate volunteered a party identification; by 1987 only 86 per cent did so. The strength as well as the incidence of party loyalty has also diminished. In the 1960s nearly half of the electorate were self-declared “very strong” identifiers; by the mid-1980s the proportion had fallen to one-fifth, while the number of “not very strong” identifiers had more than doubled, from only 11 per cent in 1964 to 29 per cent in 1987 (Crewe, 1992). It is to this group of voters – the non-aligned or undecided voters – that political advertising is specifically aimed. In contrast to the politically committed, who rarely, if ever, change their voting intentions and for whom political advertising simply reinforces their choice, campaign strategies have been developed to stir the hopes and the fears of the undecided voter. From a marketing perspective, floating voters may be considered individuals with no brand loyalty. As in other marketing contexts, the purpose of political advertising is not only specifically to target this segment of the electoral population, but also to persuade them about the benefits of choosing their particular product. It is important to note, however, at least as far as political marketing in Britain is concerned, that the underlying assumption of this approach is that not only do floating voters represent a single homogenous target audience, but, because of their absence of brand loyalty, they are the most susceptible segment of the electorate to political persuasion. To what extent are these assumptions correct? Do floating voters constitute a single homogenous block and what is the empirical evidence concerning their greater susceptibility to political persuasion? Identifying late deciders Although the number of undecided British voters has increased dramatically since the mid-1970s, with few exceptions (Heath et al., 1985, 1991; Miller, 1991), the origins of this phenomenon has received little research attention. In line with recent American research (Granberg and Holmberg, 1990; 1991), the few available British studies suggest that not only are late deciders less knowledgeable politically and less politically involved, but they are also more

Marketing politics to voters 131

European Journal of Marketing 30,10\11 132

likely to be ideologically moderate and young (Miller et al., 1990). Our results from the 1992 British election confirm this finding. For the purposes of identifying late deciders – those who made their voting decision while the 1992 election campaign was under way – age and strength of partisanship are the most reliable indicators. In terms of demographic background, later deciders tended to be much younger than their more stable, or prior decision-making, counterparts (Figure 2). For example, whereas 37 per cent of males and 41 per cent of females aged 18 to 24 years decided on how to vote during the campaign, for the 45 to 54 year old age groups these figures are almost half, at 22 and 23 per cent respectively. There is also a slight tendency for late deciders to be female rather than male, although this difference is only of marginal importance across all age groups, with the notable exception of voters aged 25 to 34 years. Late deciders also tend to be non-partisan or partisans who hold weak identifications (Figure 3). Among voters with a very strong party identification only 6 per cent delayed their voting decision until the election campaign was under way; among non-partisans the same figure was 47 per cent. A similar pattern is evident for individuals who had decided on their vote more than two years before the election campaign. While an overwhelming majority of very strong identifiers (89 per cent) had decided how to vote prior to the election campaign, slightly less than one-third of non-partisans had done so. Only in relation to weak partisans are the proportions evenly distributed across both groups. Thus, 37 per cent of weak identifiers made their voting decision at least Per cent decided during campaign 50 41

40 34* 37

30

30

28 24*

20

23 22

18 15 17

10

11

0 18-24 25-34 Age group (years)

35-44

45-54

55-64

65+

Key Figure 2. LAte vote deciders by age and gender, 1992

Male Female Notes: *Gender difference statistically significant at p <0.01, two-tailed Figures are percentage of survey respondents who said that they decided how to vote during the election campaign

Per cent 100 89 90 80 64 70 60 50 40 30 20 19 10 6 0 Very strong Fairly strong Partisanship

Marketing politics to voters

47

133

40 37 32

Weak

None

Key During election campaign Long time ago Source: British Election Study (1992)

two years before the election campaign compared with 40 per cent of late deciders. Although late deciders may be considered a relatively homogeneous segment of the electoral population in terms of age and lack of partisanship, previous research suggests that it is not these factors per se that make them indecisive, but rather the characteristics that are associated with them (Heath et al., 1985, 1991; Miller et al., 1990). The young tend to be less politically aware, to have accumulated little political information in their everyday lives, and to have fewer social networks from which they can acquire such information. Similarly, weak partisans lack the conceptual tools to use political information in a meaningful way, and as a consequence, often do not hold any strong ideological feelings. Furthermore, although late deciders are less knowledgeable politically and lack political involvement, this does not necessarily mean that they are easier to influence. As Philip Converse demonstrated more than three decades ago, less politically aware voters also pay less attention to political information and consequently are as difficult to convert as the most committed and knowledgeable voters (Converse, 1962). The most likely outcome is that even when these groups are targeted in party political advertising, it will not result in conversion in sufficient numbers to influence the eventual result of the election. The electoral consequences The impact of political advertising in persuading voters remains controversial. While party strategists and media specialists underline the major shifts in

Figure 3. Copy to follow

European Journal of Marketing 30,10\11 134

party support that often occur during an election campaign, academic researchers have remained sceptical. Arguing that political advertising simply reinforces voters’ existing predispositions, studies of the mass media have found that the impact of television and newspapers is limited to the better informed and already committed section of the electorate (Glaser and Salmon, 1991). Furthermore, this also seems to be the case even among the undecided, or floating, voters (Miller, 1991; Miller, et al., 1990). While political advertising may influence undecided voters on some election issues and on public feelings of warmth towards the party leaders, it has little overall effect in shaping the final voting choice. Even when political advertising does result in conversion, it rarely affects the election outcome since there is a mutually cancelling pattern of voter movement between the parties. As Miller et al. (1990) conclude in their comprehensive examination of voting patterns in the 1987 election: “That these [final campaign] changes did not have a greater effect upon the eventual outcome was in part a product of mutually cancelling patterns of voter movement between the parties. In 1987, much of the change in voters’ political beliefs, attitudes, and opinions cancelled out in the aggregate”. Because later deciders are less committed voters, the factors that shape their ultimate choice often tend to be random; the absence of any patterned response means that no one party is likely to make a significant net gain from their movement. If there is any vote advantage to be derived from late deciders, we would expect that the minor parties would be more likely to benefit. Partisan dealignment creates the conditions under which voters may defect from the major parties, but not cross the psychologically-important dividing line between Conservative and Labour. Thus, over the past quarter of a century the Liberal-Democrat vote has risen significantly. Although the Liberal-Democrat vote in 1992 slipped to 17.8 per cent, in the two elections of the 1980s it averaged 24.0 per cent, compared to less than half that figure in the elections of the 1960s. The extent to which the Liberal-Democrats benefited from late deciders is shown in Table I. About four in every ten Liberal-Democrat voters were late deciders, compared with 17 per cent of Labour voters and 22 per cent of Conservatives. Indeed, Liberal-Democrat voters who were late deciders actually Vote

Table I. Vote by when decided, 1992

A long time ago In 1991 In 1992, before the campaign During the election campaign Total (n)

Con

Lab

Lib-Dem

Other/DK

62 7 9 22 100 (1,103)

71 6 17 17 101 (826)

36 8 39 39 100 (413)

48 7 36 36 99 (147)

Source: British Election Study (1992)

outnumbered their counterparts who said that they decided a long time ago. Although the aggregate vote totals tend to disguise it, the effect of this volatility on Liberal-Democrat support is a high vote turnover from election to election (Studlar and McAllister, 1987). However, the electoral benefits of increased proportions of late deciders to the Liberal-Democrats is limited, as we noted before, if defections and conversions occur in approximately equal proportions. This is, in fact, substantially the case, at least as far as inter-election shifts are concerned. The figures in Table II are the proportion of electors who voted for a particular party in the 1987 and/or 1992 elections, followed by the proportion who defected from that party between the two elections, and the proportion who were converted. The figures are disaggregated by the timing of the vote decision. Overall, the LiberalDemocrats experienced a net vote gain of 3 per cent, the product of 4 per cent defection and 7 per cent conversion. Labour gained 1 per cent based on these shifts, while the Conservatives had a net loss of 4 per cent. So although the Liberal-Democrats did experience a net gain in votes, their advantage would have been much more substantial if they had been able to defray defections among their 1987 supporters. The figures also illustrate the extent to which a disproportionate amount of defection and conversion took place among late deciders, and among those who had decided how to vote in the previous year. For example, of the 32 per cent of 1987 and 1992 Conservative voters who had decided how to vote a long time ago, only 3 per cent – or about one in ten – changed their vote between the

Con

(Per cent) Lab

Marketing politics to voters 135

Lib-Dem

Decided long time ago Voters in 1987, 1992 32 26 6 Defections –2 –1 –1 Conversions +1 +2 +1 Net vote 31 27 6 Decided in last year Voters in 1987, 1992 8 3 3 Defections –3 –2 –1 Conversions +1 +2 +3 Net vote 6 3 5 Late deciders Voters in 1987, 1992 9 5 4 Defections –3 –2 –2 Conversions +2 +2 +3 Net vote 8 5 8 Total vote 45 35 16 Note: Estimates exclude voters who were too young to vote in 1987 or who did not vote in either 1987 or 1992 Source: British Election Study 1992

Table II. Net vote gains and losses 1992

European Journal of Marketing 30,10\11 136

two elections. By contrast, among 1987 and 1992 Conservative voters in the late deciders group, more than half had changed their vote during the same period. Clearly, then, there is considerable potential for parties to benefit electorally from late deciders – if they can maximize conversions and minimize defections. Although these results relate specifically to inter-election shifts, the data also enable us to make some estimates of the voter shifts within the campaign period. The 1992 BES contained a question asking respondents if they had seriously thought of voting for another party during the election campaign and, if so, what that party was. The results showed that about one quarter of voters had considered changing their vote, substantially more (60 per cent) in the late deciders group than among those who had decided in the previous year (26 per cent) or a long time ago (12 per cent). Of the late deciders who had finally opted to vote Conservative or Labour and who had considered voting for other parties, by far the largest proportions said that they had thought about the LiberalDemocrats, rather than the other major party. These results lend some support to previous market research findings on brand switching and product choice. As is also the case among consumers, voters who considered changing their vote were more likely to switch between parties or brands that were their closest substitute (Carpenter and Lehmann, 1985; Grover and Srinivasan, 1987; Keller, 1993). In other words, similar to commercial markets, the political market may also be intuitively understood in terms of a set of hierarchical submarkets in which a party (brand) competes more directly with other parties (brands) within its submarket. The positioning of the party within the political spectrum will thus determine the extent of voter loyalty, and the party’s ability to compete with other parties for votes. Conclusion Since the 1980s, British election campaigns have become increasingly organized and professionally-managed media events. All of the major political parties use advertising consultants to plan their campaigns and the tools of professional marketing – photo opportunities, rallies, walkabouts and even the use of film directors and actors – are commonplace. Yet, despite the activity and the cost, there is little empirical evidence to suggest that political advertising has any significant impact on electoral outcomes. Past research suggests that, although the use of political campaigns does result in a better informed voting public, its overall impact is limited to nothing more than a reinforcement of the voter’s existing predispositions. In terms of their actual electoral behaviour, comparatively few voters are influenced by these events and the net effect on aggregate vote is negligible (McAllister, 1985; Miller, 1991; Miller et al., 1990). Our results from the 1992 election broadly confirm this finding. Even among late deciders, where there are proportionately greater inter-election and campaign shifts, the net impact on the votes cast for the three parties is comparatively small.

From a marketing point of view, the central problem is that political advertising is currently targeted to the wrong group of individuals. By concentrating their efforts on undecided voters, the least aware or least partisan group, political campaigns have chosen a segment of the electorate that is the least likely to be persuaded by either their product or marketing efforts. Furthermore, because the undecided are also most likely to be determined by random factors, even when effective, the overall benefits of this approach remain highly questionable; swings in support in the direction of one party are likely to be cancelled out by swings towards the opposing party. As we have shown, the Liberal-Democrats made a net gain among late deciders at the expense of the major parties, but much of the gain was lost because of defections. The figures for the Conservative and Labour parties are similar. From a marketing perspective, not only will the continuing targeting of this segment of the electorate fail to result in any short-term gains in productpositioning or dominance, but it will also impede the development of any longterm party or brand loyalty among the wider electorate. Would a re-orientation towards an alternative segment of the electorate – the middle group of voters, or those who decided in the previous year – reap greater electoral rewards for the parties? The evidence presented here suggests that it might. This group remains substantial – accounting for 16 per cent of the electorate in 1992 – and is both better politically informed and more cognitively skilled than late deciders. Moreover, research in the USA has suggested that the images of the party leaders have a substantial impact on the party preferences of this group, mainly via voter perceptions of the leaders’ integrity, reliability and competence (Kessel, 1988; Miller et al., 1986). Voters with average levels of political awareness are now more likely to use such information as an electoral cue rather than partisanship, social position or their peer group. If British political parties and their campaign managers targeted their products in terms of leadership qualities to this group of voters using precampaign strategies, perhaps they would find it a more rewarding and efficient use of their resources. It is to this group of voters, and not the undecided, that political advertisers should be directing their campaigns. References British Election Surveys (1974-92), February. British Election Study (1992). Butler, D. and Kavanagh, D. (1992), The British General Election of 1992, St Martin’s Press, New York, NY. Butler, P. and Collins, N. (1994), “Political marketing: structure and process”, European Journal of Marketing, Vol. 28, pp. 19-34. Carpenter, G.S. and Lehmann, D.R. (1985), “A model of marketing mix, brand switching, and competition”, Journal of Marketing, Vol. 22, pp. 318-29. Clemente, M.N. (1992), The Marketing Glossary, Amacon, New York, NY. Clifford, P. and Heath, A. (1992), “The election campaign”, In Heath, A., Jowell, R., Curtice, J. and Taylor, B. (Eds), Labour’s Last Chance? The 1992 Election and Beyond, Darmouth Publishing, Aldershot.

Marketing politics to voters 137

European Journal of Marketing 30,10\11 138

Converse, P. (1962), “Information flow and the stability of political partisan attitudes”, Public Opinion Quarterly, Vol. 26, pp. 578-99. Crewe, I. (1992), “Partisan dealignment 10 years on”, in Denver, D. and Hands, G. (Eds), Issues and Controversies in British Electoral Behaviour, Harvester Wheatsheaf, London. Engel, J.F., Fiorillo, H.F. and Cayley, M.A. (1972), Market Segmentation, Concepts and Applications, Holt, Rinehart and Winston, New York, NY. Farrell, D.M., McAllister, I. and Broughton, D. (1995), “The changing British voter revisited: patterns of electoral campaign volatility since 1964”, in Broughton, D. et al (Eds), British Elections and Parties Yearbook, Frank Cass, London. Glaser, T. and Salmon, C. (1991), Public Opinion and the Communication of Consent, Guildford, New York, NY. Granberg, D. and Holmberg, S. (1990), “The Berelson Paradox reconsidered: intention-behavior changers in US and Swedish election campaigns”, Public Opinion Quarterly, Vol. 54, pp. 530-50. Granberg, D. and Holmberg, S. (1991), “Election campaign volatility in Sweden and the United States”, Electoral Studies, Vol. 10, pp. 208-30. Green, P.E., Tull, D.S. and Albaum, G. (1988), Research for Marketing Decisions, Prentice-Hall, Englewood Cliffs, NJ. Grover, R. and Srinivasan, V. (1987), “A simultaneous approach to market segmentation and market structuring”, Journal of Marketing Research, Vol. 24, pp. 139-53. Harrop, M. (1990), “Political marketing”, Parliamentary Affairs, Vol. 43, pp. 277-91. Heath, A., Jowell, R. and Curtice, J. (1985), How Britain Votes, Pergamon Press, Oxford. Heath, A., Jowell, R., Curtice, J., Evans, G., Field, J. and Witherspoon, S. (1991), Understanding Political Change, Pergamon Press, Oxford. Kavanagh, D. (1995), Election Campaigning: The New Marketing of Politics, Blackwell, Oxford. Keller, K.L. (1993), “Conceptualising, measuring, and managing customer-based brand equity”, Journal of Marketing, Vol. 57, pp. 1-22. Kessel, J.H. (1988), Presidential Campaign Politics, 3rd ed., Dorsey, Chicago, IL. McAllister, I. (1985), “Campaign activity and electoral outcomes in Britain, 1979 and 1983”, Public Opinion Quarterly, Vol. 49, pp. 300-15. Miller, A.E., Wattenberg, M.P. and Malanchuk, O. (1986), “Schematic assessments of presidential candidates”, American Political Science Review, Vol. 80, pp. 521-40. Miller, W.L. (1991), Media and Voters, Clarendon Press, Oxford. Miller, W.L. et al. (1990), How Voters Change, Clarendon Press, Oxford. Newton, K. (1992), “Caring and competence: the long, long campaign”, in King, A. et al. (Eds), Britain at the Polls, Chatham House, Chatham, NJ. Niffenegger, P.B. (1989), “Strategies for success from the political marketeers”, Journal of Consumer Marketing, Vol. 6, pp. 45-51. O’Shaughnessy, N.J. (1990), The Phenomenon of Political Marketing, Macmillan, Basingstoke. Political Change in Britain Survey (1964-70). Rao, C.P. and Wang, Z. (1995), “Evaluating alternative segmentation strategies in standard industrial markets”, European Journal of Marketing, Vol. 29, pp. 58-75. Reid, D. M. (1988), “Marketing the political product”, European Journal of Marketing, Vol. 22, pp. 34-47. Scammell, M. and Semetko, H.A. (1995), “Political advertising on television: the British experience”, in Kaid, L. Lee and Holtz-Bacha, C. (Eds), Political Advertising in Western Democracies, Sage, London. Scammell, M. (1995), Designer Politics: How Elections Are Won, St Martin’s Press, New York, NY.

Smith, G. and Saunders, J. (1990), “The application of marketing to British politics”, Journal of Marketing Management, Vol. 5, pp. 295-306. Smith, W.R. (1956), “Product differentiation and market segmentation as alternative marketing strategies”, Journal of Marketing, Vol. 21, pp. 3-8. Studlar, D.T. and McAllister, I. (1987), “Protest and survive? Alliance support in the 1983 British general election”, Political Studies, Vol. 35, pp. 39-60. Taylor, B., Brook, L. and Prior, G. (1992), “Appendix: the 1992 cross-section and panel surveys”, in Heath, A., Jowell, R., Curtice, J. and Taylor, B. (Eds), Labour’s Last Chance? The 1992 Election and Beyond, Darmouth Publishing, Aldershot. Webb, P. (1992), “Britain: the 1987 campaign”, in Bowler, S. and Farrell, D. (Eds), Electoral Strategies and Political Marketing, Macmillan, London. Wing, Y. (1978), “Issues and advances in segmentation research”, Journal of Marketing Research, Vol. 15, pp. 317-37.

Marketing politics to voters 139