30 The nature of American politics
tion of 1936, a majority of non-manual workers voted for the Republicans.
Lipset has pointed out that, if these crude occupational categories are broken
down, we find that, the further down the social scale a group is placed, the
greater the percentage preference for the Democratic Party. In 1948 nearly
80 per cent of American manual workers voted for the Democrats, a higher
percentage than most left-wing parties in European countries could achieve.
In the 2004 election, the Democratic candidate, John Kerry, received 63 per
cent of the votes of those electors with a family income of less than $15,000
per annum, but only 41 per cent of the votes of those with a family income
of over $100,000 per annum. Thus a connection between social class, income
and voting behaviour clearly exists, but it is by no means a simple one, and
there is a considerable variation from one election to another in the extent
of ‘class voting’. In 1948, for example, the issues before the electorate were
largely of an economic kind, and the voting patterns reflected class interests
to a high degree. The elections of 1956 and 1972, however, present a different
picture. In 1956 almost as many manual workers voted for the Republican,
Dwight Eisenhower, as for the Democratic candidate, and in 1972 the Re-
publican candidate, Richard Nixon, received the support of 57 per cent of the
manual workers, as against only 43 per cent voting for his Democratic rival,
George McGovern. In 2004 61 per cent of members of trades unions voted for
John Kerry, the Democrat, while only 38 per cent voted for his Republican
rival, George W. Bush.
There are therefore wide variations from election to election in the extent
to which voters are influenced by their perceptions of their class interests
in their voting behaviour. Of course, the attribution of sectional or class
motivations on the basis of the sort of statistics quoted above is a very dif-
ficult exercise. What is apparently class voting may be motivated in quite
different ways, because religious, ethnic and regional groupings all over-
lap with class to a very considerable extent. Exactly why a low-paid Irish
Catholic industrial worker in the North-East votes Democratic rather than
Republican can hardly be explained by any simple formula. Angus Campbell
and his co-authors in their study The American Voter found that a third of the
American population was ‘unaware’ of its class position, and that social class
played a significant role at a conscious level in the political behaviour of only
a fairly restricted and sophisticated portion of the population. The relative
volatility of the American electorate, the readiness to switch votes from one
party’s candidate to another at successive elections, is also a measure of the
limitations upon appeals to class orientation as a source of voting behaviour.
American elections can produce ‘landslide’ results that would be unthinkable
in a system where stable class voting is the norm. The figures in Table 2.2
give the percentage of the total votes cast that have gone to the Democratic
and Republican Party candidates in presidential elections since 1936.
These figures reveal that, from 1936 to1984 at any rate, there was a startling
propensity for large sections of the electorate to switch their allegiance from