Political cleavages have long been a central concept used by political science and sociology to understand and predict outcomes that are relevant to the political system and society more generally. In the literature the debate of the character and development of these political cleavages has been sharp and a source of claims ranging from the proclamation of the death of the old traditional cleavages; the testification of the birth of new cleavages; to methodological disputes. What we know is that the class cleavage has been subject to observations of declining relevance and explanatory power. Entering into this mosaic of polemics, I try to contribute to this long debate by subjecting the apparent decline of class cleavage politics to an analysis through the lenses of political generations and the unique social and historic socialisation of cohort groups as a potential source of this explanatory decline. I do this with logistic regression to analyse the tendency of Swedish left-wing voting and electoral turnout through the use of longitudinal quantitative data over a 42-year period and 14 election cycles from 1960 to 2002. By categorizing the data depending on the respondent’s birth cohorts and occupation, I find that there exists a generational difference in voting preferences and election turnout depending on cohort identification. These trends and tendencies are different depending on occupation, and the decline in class as an explanatory variable for political and social outcome seems to be in part explained by birth cohort, and possible also by the subsequent alienation of people from the political system in general, indicated by a lower generational tendency for election turnout. In addition to these results, I further try to expand upon the knowledge of political generation by examining the influential powers of macroeconomic forces during respondents’ formative years as a possible explanation for the generational differences. I give a restrained possibility for this to be the case, having in mind possible limitations with the data.