To what extent does social class continue to affect voting behaviour in Britain?

Authors such as Heath, Jowell, and Curtice advocate their work around the work of the sociologist, John Goldthorpe. These authors argue that given the appropriate classification and measurement, the relationship between social class and party vote has not weakened. However, others such as Dunleavy and Husbands argue that following the growth of the welfare state the division between public-sector and private sector workers has become increasingly important in determining voting behaviour. The first systematic evidence of class voting came from Gallup polls published just after the Second World War. Several measures of social class can be used to monitor trends of voting behaviour over time

Heath et al contended that the Goldthorpe model is “wholly inadequate for studying the social bases of politics since it ignores important divisions which have little to do with the colour of a man’s or woman’s collar”, nevertheless, the Goldthorpe model is believed to be best fit to the analysis of class and the effects it has on voting behaviour. This is due to the simplicity of its two class model; the white collar and blue collar.

An indication of how class sizes have changed is needed, and this can be achieved by comparing the class profile of 2001 and 2005. This will be done with the use of statistical software to analyse and study British Survey data. This is will help determine the extent to which class affect voting behaviour.

Evaluating this univariate analysis of social class in 2001 suggests that most workers in Britain have a relative good income, coming from beneficial employment. The level of job security is high, as well independence. The table also shows that over half of the workers (57.3%) are in the top two groups (Salariat and Routine nonmanual).

If we contrast the class profile of 2001 to that of 2005, the most helpful and utmost increase is from the Routine non-manual worker, which has an increase of 4.2%. The strongest decrease comes from the Salariat, where there is a 3% decrease. On the other hand, the working class have increased, but by only 0.3%. This evidence suggests that the upper and lower class are decreasing, which results in a growth of the middle class.

This kind of data analysis, known as bivariate analysis, explores the concept of association between two variables. Association is based on how two variables simultaneously change together -- the notion of co-variation.  A bivariate analysis is needed to analyse whether class appeared to affect voters’ behaviour in both 2001 and 2005.

The table shows that the Labour Party have attracted the working class, as a substantial portion of support is from their class, and from foremen and technicians. It also indicates that the conservatives did not attract the votes needed from the two upper groups. In fact, the data shows that more workers from the salariat group preferred Labour over the Conservatives, even though Labours policies would affect many of those in the class, such as high taxation. Most of Labours votes, indicates that they are coming from the working class. Reason being a Labour government would produce security in their employment. Also to note, a labour government would provide those in the working class welfare and benefits for those who are not able to work, or produce such a high incomes.

For the data, from the bivariate crosstabulation, to be statically significant, the range of the Pearson Chi-Square must be between 0.00 and 0.05. Our crosstabulation between vote and social class is significantly significant as it falls within this range.

Data from 2001 showed us most of Labours votes were from the Working class, however in 2005 labour lost 7% from the Working class. This, oddly enough, shifted over to the Liberal democrats, who took pleasure in a 6% increase from the working class. Concentrating on the upper two classes, there still seems to be a fair share of votes between the Labour and Conservatives. This shows that Conservatives are still finding it hard to pull attraction to their traditional class group. On the other hand, the petty Bourgeoisie show an increase in their support for Conservatism.

It can be noticed again, that the range in within 0.00 and 0.05, making social class and party vote statistically significant according to the Pearson Chi-square tests.

Multivariate analysis has been carried out to assess whether men and women acted on class interest. This analysis is the concept of association between multiple variables.

Evaluating this crosstabulation, it proves that there is no real difference between the class voting of men and women in 2001.  

As the data from 2001 clearly showed not much significant difference, 2005’s data is very similar. Not much difference lies between men and women votes. Nevertheless, the only clear changes are that of the petty bourgeoisie women votes. Evidence shows that more women in this group favoured the Conservatives than men.

The data above shows that there is a decline in class voting for both men and women. Men still have a stronger trend; nevertheless, there is a decrease from 2001.

Heath et al, How Britain Votes, (Pergamon Press, 1985) pp.14-15

• R. Andersen and A. Heath, ‘Social Class and Voting: A Multi-Level Analysis of Individual And Constituency Differences’,  Centre For Research Into Elections And Social Trends, Working Paper Number 83 September 2000 [www]

---