ABSTRACT
In the preceding chapter we reviewed the evidence regarding the effects on the voter of the opinions of others among whom he lives and works. This is but one aspect of the inter-relationships between electors and their environments which might influence their voting decisions. It portrays the individuals discussing political issues with each other as voters, when many people in society are more than just ‘other voters’; a few are candidates for office; many more are party workers, committed to winning the election for their organization. And so we turn now to an investi gation of these particular individuals, on where they live and campaign, and how they influence the voting decisions of the electorate. Much voting in partisan elections is habitual; the electors
always cast their preference for the same party, on a continuing belief that only that party can best serve their personal selfinterests. A considerable proportion re-determine their preference during each election campaign, however, surveying the policies of the competitors and passing judgement on the actions of the current government (and perhaps of past ones too). Where parti san politics are not as strong as in Britain, say, and campaigns are more personalized, as they frequently are in the United States, then the elector may be more swayed by the characteristics of the individual candidates and their positions on what the elector per ceives as the key issues (interestingly, Ennis, 1962, pp. 196-205, has shown how candidates are often able to determine those key issues). Voting decisions are based on a variety of criteria. As part of a
survey conducted in four American states (Colorado, Iowa, Min nesota and Washington) prior to and after the 1960 presidential election, respondents were asked (Ennis, 1962, p. 189), ‘when you vote for Senator, which is most important to you: the man himself, the stand he takes on public issues, or his political party ?’ With the Iowa respondents, who were not atypical, several fea tures stand out from the answers (Table 6.1):
Table 6.1 Important determinants in voting for senator: Iowa, 1960 Percentage giving the important determinant as: The man The issues His party
Voters with high educational levels living in Large towns 23 71 6 Small towns 36 61 3 Rural areas 42 54 4 Total 38 58 4
Voters with low educational levels living in Large towns 33 61 6 Small towns 60 37 3 Rural areas 50 43 7 Total 49 45 6
(1) Voting for party, irrespective of the issues or of who the candidate is, is rare. (2) Better-educated voters are more concerned with issues, and the senatorial candidates’ positions on them. (3) Small-town and rural residents are more likely to vote for the candidate as a person than are their contemporaries in the larger settlements. This could in part be related to the problems of mobilizing the periphery that we discussed in Chapters 3 and 4. Ennis (1962, p. 191) argues that: ‘where social class cleavages are
pronounced in a community, class perspectives will ramify into political choices to a greater extent than in communities less sharply polarized . . . traditional or personalistic forms of reference are more dominant in communities where a class orientation is minimized’ - with the latter being the small towns and the rural areas. Of these three findings, the last is of the most general interest. That voters were not very party-oriented may be a peculiar feature of the American party system; that rural voters are more ‘personal’ in their politics may not. Voters may use different criteria when voting for different
offices, or the same criteria may lead them to vote for candidates from different parties because of personal characteristics or issue orientations. In many American elections, for example, people are voting every four years not only for a president and a congress man, but probably also for a senator and a state governor, as well as for a variety of state and local government legislators and officials. Most candidates are allied with a particular party and many electors vote, say, for a full Democrat slate, a decision which is particularly easy if voting machines are in use, since a single lever can then be pulled to cast all of their votes. Exactly how people make the decision to vote for a full slate, unless they habitually vote entirely for the one party, is not clear. A ‘coat tails’ procedure has been suggested, by which voters first make their decision on which candidate to support for a major office, such as president, and then decide to vote for all of the candidates for other offices from the same party. The 1960 four-state study, referred to above, found that as the election came closer voters were more polarized in giving all of their choices to either one party or the other, but there was no evidence that the major decision was taken on the important offices and all the others were consequent upon that (Meyer, 1962). A recent marked change in American voting behaviour has
been the shift away from the decision to vote a full party slate. Between 1940 and 1972, for example, Burnham (1975) found a considerable reduction in the correlations between the per centage of the relevant votes obtained by the Democrat candi dates for four major offices in West Virginia (his observation
units were the 55 counties). Clearly, West Virginians have become less partisan. A vote for Nixon in 1972 was probably associated with one for the Republican gubernatorial candidate, as the cor relations show (Table 6.2), but not necessarily for that party’s
Table 6.2 Democrat voting for four offices: West Virginia, 1940 and 1972
Correlations between votes for* President Senator Congressman Governor
and votes for President +•988 +•980 + •991 Senator +•734 +•992 + •994 Congressman + •536 +*680 +•998 Governor +•910 + •718 +•513
Senate and House candidates as well. This suggests that the state’s voters did not believe that their interests in the Congress would be best served by members of the same party as that of their preferred White House incumbent. One reason advanced to account for this is the immense power of senior (long-serving) members of both houses of Congress, in the committee positions they can obtain and the rewards this can bring, such as defence contracts, to their states. Voters are thus loath to remove an incumbent senator or representative without very good reason, leading to a growing institutionalization of Congress (Polsby, 1968) and ‘split ticket’ voting when a candidate from a different party seems a better prospect for the presidency. Such split voting may have been particularly marked in the southern states in 1948 and 1968 when third-party candidates - Thurmond and Wallace respectively - stood against the Democrat strength there (Burnham, 1975), but it was also a general feature in 1956 with
Eisenhower being returned to the White House when there was a strong swing to the Democrats in the House of Representatives (Campbell and Miller, 1957; Cummings, 1966). The extent to which voters do ‘split their ticket’ cannot be discerned easily from aggregate voting data, but inspection of individual ballots in Milwaukee has shown that whereas in 1960 some 68 per cent of all voters selected candidates from a single party, in 1970 the corresponding figure was only 32 per cent (Burnham, 1975, p. 319). ‘Split ticket’ voting is encouraged by the American electoral
system in which many contests are decided simultaneously. Else where it occurs over periods of time, which may account for the relatively large swings in opinion recorded at the Australian and New Zealand general elections of 1975. Both phenomena suggest that the electorate is responsive to a variety of stimuli in making its voting decisions. In some cases, more so perhaps in certain types of contest than in others, it may be the personal charac teristics of the candidates, irrespective of party, which determine their success or failure and, as we suggested in Chapter 1, such personalized voting might have a very clear pattern in its geog raphy. Other decisions may be made on specific issues; such issues may be national in scope, but others may be very localized, with candidates or parties taking up positions relative to local interests and receiving votes accordingly. Finally, contestants may decide that it is more beneficial to campaign for votes in some parts of the constituency than others. All of these deter minants suggest a clear geography of influence on voting patterns and election results, and they form the foci of the remainder of this chapter.
