Franklin, Mark. "How the Decline of Class Voting Opened the Way to Radical Change in British Politics". British Journal of Political Science, Vol.14, No.4 (1984): 483-508.

Franklin, Mark. "How the Decline of Class Voting Opened the Way to Radical Change in British Politics". British Journal of Political Science, Vol.14, No.4 (1984): 483-508.

• In the decade between 1959 and 1970, the importance of socio-economic class -- generally considered to be the critical identifier in British politics -- as the deciding factor in British politics was drastically declined, losing between a quarter and half its predictive power in elections (483).
• When class voting was at its height in 1964, two factors: class of parents, and occupation of respondent, were the most important indicators of political choice. The ability of these variables to explain voting choices was the primary decline in class-based voting (485).
• While the Conservative and Labour parties are still dominant, this is because of a number of other advantages and not a result of class voting. The success of the Scottish National Party has demonstrated this fact in the 1984 elections (485-486, 507).
• The data used for this study comes from nationwide surveys of the British population, excluding Northern Ireland, following each general election between 1964 and 1979. Six variables are recorded: parents' class, parents' party, respondent's education, and  occupation, union membership and type of housing (486). 
• Unlike previous studies, this article records the results on this survey for non-voters as well as voters. The author believes that non-response is an important variable and should be studied (488).
• The variables measured obviously related to one another, like parent's class affecting the education received by their children. This is explained on page 491 and 494.
• The data demonstrates the strength of class voting in the 1964, as working-class characteristics, such as minimal education, rented housing, union membership, and working occupation, were all correlated with politics. Only 2% of people without these characteristics voted for Labour, and only 1% of those with all of them voted for Conservative (489-490).
• The decline in correlation between these class characteristics was part of the large decline in class voting overall. The ability of major variables like parent's class to predict other class variables, like minimum education, declined by more than half from 1964 to the contemporary election (495).
• Occupation and parent's class were the two core variables underpinning the system of class voting, with other class characteristics supporting the institution (496, 498).
• Although the Labour and Conservative parties of 1984 have a number of serious policy differences, these divisions were much more muted 20 years ago. At the height of class voting, party stances on issues such as unions and housing were identical (496).
• The decline of class voting comes at the same time as a decay in the stability of its underpinning variables. Whereas class was set a few decades ago, social class mobility is now common. Social class is now one of the most volatile social characteristics (499).
• "Not only have social characteristics declined in their ability to structure party choice therefore, but such social constraints as remain are more readily altered during adult life" (501).
• Of the variables linked to class voting as explored earlier, only the political preferences of parents still has a significant effect on voting patterns. This is true, however, for all political parties, not just Labour and the Conservative party (501).
• While the effect of social class on individual voting patterns has been declining, contemporary studies have shown that average class divides in a region may be able to predict voting patterns within that area. Accordingly, a mostly middle-class area will have more Conservative voters, even among lower-class populations than a majority lower-class area (503).
• The author contests this hypothesis, arguing that other factors beyond class are causing these new voting trends. Dr. Franklin particularly points towards rates of home ownership, which are an indicator on partisan policy issues coming to the fore in voting when class voting subsided (504).
• When factors not directed associated with class are controlled for, the area voting affects of class are not replicated in individual voter choices, which are instead based on new policy factors like housing or education, which tend to be geographically clustered (504-505).
• The contemporary correlation between class and political affiliation still holds in 1984, but it does not have any durability. Currently, the strong class voting patterns of previous generations are maintaining the balance through the power of inherited political loyalty, but this balance could change in the next election cycle as British politics approaches a period of unpredictability (507-508).