UK Voting Behaviour

This essay will examine the topic of voting behaviour in the United Kingdom. It will investigate the social structures of Britain, and to the extent at which social factors such as class determine the voting habits of a nation. It will then go on to outline the ‘dominant ideology’ model of voting behaviour, wherein political groups determine through the media, the political agenda, through such things as opinion polls and newspaper editorials. Next, the ‘party identification’ model will be investigated, asking the question, how important is party loyalty important to elections?

The essay will then seek to find out, with balanced evidence, which one of these models of examining voting behaviour is the most accurate. The essay will then be summed up with a conclusion of all the points raised. Firstly, we’ll examine the influence of class in the electoral process. One explanation for why people vote in a certain way could be that most people vote along class lines, or more accurately, along the lines of their profession.

Traditionally, it is expected that the working class (low wage, annual workers), vote for the Labour Party, whilst those in non-manual work, such as the middle and upper classes, vote for the Conservative Party. However, it is accepted that between 1945 and the 1970, the working class made up the majority of the British population. If the method of measuring voting behaviour of class has any credibility, the Conservatives would have been out of power for the whole of this time.

In reality, time in government was evenly distributed between the Labour and Conservative Parties. Some records from this period did in fact show that more than a third of the working classes voted Conservative. It is the theory of embourgeoisement that attempts to explain the phenomenon of the working class Conservative. It is the idea that it is due to factors such as rising wage levels and living standards that the working classes started to employ more middle class behaviour, such as voting Conservative.

However, a study in the 1960s (Goldthorpe, et al. 1969), found no evidence to corroborate this theory. In fact, it found that non-Labour support of manual workers was higher amongst those who had friends and family who were non-manual workers. Also, the sociologist Frank Parkin argues that manual workers living outside the more traditional, single-industry, working class communities are more exposed to what he describes as the ‘dominant value system’, and are therefore less likely to vote Labour (Parkin, 1972).

It was the electoral period between 1979 and 1992 that led to the revival of the embourgeoisement theory as a way of explaining the decline in the working class vote for the Labour Party. However, due to the failure of the embourgeoisement theory to explain the fall in the middle class Conservative support, the theory has very few contemporary supporters. In another study of cross-class voting, McKenzie and Silver distinguished between ‘deferential’ and secular working class Conservative voters. The former defer to the traditional authority represented by Conservative leaders who are seen as Britain’s ‘natural rulers’.

The latter vote Conservative not because they are enthusiastic supporters of Conservative values, but because they believe that they will be better off financially with a conservative government. They are Conservative because they are not true believers in Conservative ideology (McKenzie and Silver 1968). Now, the essay will go on to explain what the ‘dominant ideology’ model of voting behaviour means. During elections, various news outlets, such as newspapers and television news-gathers, attempt to gauge the outcome of the election by asking a random amount of people for their voting intentions.

These are published by them to show the level of support for the main parties. These are known to most people as opinion polls. Strictly speaking, opinion poll results are not predictions of election results. Rather, they provide a snapshot of voting intentions on a certain day. But, the media often presents poll data as forecasts. Opinion polls are not only carried out at elections, nor are their questions bound to political topics. It is during election time, however, that opinion polls gain the greatest attention and scrutiny.

The first British general election campaign, which featured opinion polling, was the 1945 general election. That was at the time when only a single polling organisation existed. Since the 1970s, the number of polls published in the media has increased sharply. Indeed, the media commissions most polls. During the 45-day election campaign in 1997, 44 sets of national poll results were published in the national press (Butler and Kavanagh 1997, p. 123 and Crewe 1997) This was a smaller number than the number published during the much shorter general election campaign of 1992.

According to Kavanagh (1992), the polls had a reasonably good record of accuracy at general elections up to 1970, but, after 1970, polls were inaccurate in five of the following seven elections. Even so, Eatwell (1993) has calculated that the average margin of error for the final polls between 1945 and 1987 was only 1. 3%. This is an impressive record, especially given that the polling organisations themselves stress that in general elections with a sample of around 1000 respondents there is a margin of error of plus or minus 3%.

It should be noted that polls conducted during by-elections are usually less accurate because there is often a greater degree of volatility in voting behaviour compared to that in a general election. According to Sanders (1993), the question of the accuracy of the polls during the 1992 general election boiled down to one or both of the following – either the polls were correct at the time they were conducted (and late switching took place) or the polls unconsciously misrepresented ‘true’ public opinion.

It is the latter idea, which presented pollsters with a major problem. Much of the work of psephology (the statistical study of elections and trends in voting) involves the construction of ideas and theories about what people think and what people do and think. But, opinion polls only measure what people say. They cannot be expected, therefore, to take account of discrepancies that might arise between these different aspects of human behaviour. That people might say one thing yet think or do another is a possibility in an election.

The 1992 election campaign was marked by the high moral tone taken by the opposition parties and by an appeal of self-interest by the Conservatives. If, as some studies suggest (Kavanagh 1992), people tend to give what they feel are politically correct answers to opinion poll questions whilst voting according to what they perceive to b their economic self interest, this might account for the discrepancy between the polls and the actual election result.

In addition to the question of polling accuracy, there is a long-running debate about whether or not the publication of opinion poll results influences voting behaviour. One view is that polls have a ‘bandwagon’ effect and encourage some voters to vote, which appears to be the most popular. An opposite view is the idea of a ‘boomerang’ effect – a party trailing in the polls picks up sympathy votes as the ‘underdog’ or supporters of the leading party become complacent and fail to turn out the vote. According to Denver (1994), however, research has found little evidence to support either view.

Now, this essay will seek to find out which is the most accurate system of determining the voting behaviour of the British public, between the ‘dominant ideology’ model and the ‘class interest’ model. Firstly, when examining the argument that the ideology model is a poor method of examining voting behaviour, the main argument that arises is that opinion polls can be susceptible to being inaccurate, especially when the most common type of elections are usually local elections and by-elections, which are notoriously unpredictable.

Also, it is due to this reputation that some, such as the Speakers’ Conference on Electoral Reform in 1967, had suggested that polling operations should cease over the final three days of an election campaign. Even during the historic Conservative defeat in 1997, not one of the polling agents returned a result that reflected the result on polling day. However, where class-aligned voting is concerned, the supporters of the class interest model would argue that class has been and still is the single most important social factor in determining voting behaviour.

However in recent times, there has been the phenomenon of class-dealignment. After the government of Margaret Thatcher, the economic gap between the working and middle classes widened, meaning that those on the higher end of the working class became upwardly mobile, owning their own property for the first time, earning higher wages than they did in the 1970s, and with this, they started to vote according to their new-found wealth, such as Conservative candidates who promised to keep taxes down, and preserve private wealth.

Equally, during the period between the 1992 and 1997 general elections, the Labour vote in the middle class Conservative heartlands of the South East went up by more than 14%, while the Conservative vote went down by over 12%. One way of explaining this rapid increase in the Labour vote is that it is in this area that much of the then new, hi-tech, non-unionised industries were based, and it was this group of skilled working class voters who deserted Labour in droves in 1979 and throughout the 1980s (Dorey 1998, p. 5-6). It is this example of class dealignment that has led to the revision of the old consensus that working class voters vote Labour and middle class voters vote Conservative. In conclusion, one explanation for voting behaviour is that most, but definitely not all people vote accordingly to their class interests. From 1945 to 1970, nearly two-thirds of all people voted for their natural class party. Since the 1980s there is open discussion between political scientists whether ‘class dealignment’ has taken place.

In respect to the ‘dominant ideology’ model, that pollsters have an almost completely unreliable reputation for accuracy, and it became clear during the 1997 general election that the changes made by the polling organisations between this time had not solved the problem of faulty sample designs, late swings and non-registration. To sum up, it is un-controversial to say that there is no conclusively ‘perfect’ system of determining voting behaviour.