I came across the term, which seems to refer to a theory initially introduced by Lipset & Rokkan in 1967. It appears to refer to a way of categorising voters into 'frozen' blocs that don't shift between parties easily; rural & urban voters, for example.

Is my understanding of the theory correct, and is this theory still useful and applicable today? I'm thinking predominantly about the shift of voters from traditionally Labour voting blocs (northern coal towns/working class) to Conservative candidates in the 2019 UK General Election which would seem to break down traditional voting patterns, however I'd be interested in answers that apply to any country.


The 'freezing hypothesis' is more of an observation than an outright theory: something that researchers have demonstrated to be 'real' and in need of a proper theoretical explanation. All it really says is that the constituencies of political parties seem to be resistant to change over long periods of time, even despite large changes in social and cultural contexts. That makes sense in a generic sort of way: parties both capture and reinforce particular worldviews and ideologies, and worldviews and ideologies are largely passed down across generations, so that the generational stability of a worldview will correlate with the generational stability of a party.

People still use this theory today, though not necessarily directly. For instance, a lot of political planning — voting projections, issue choice, base solidification, respondent choice for opinion polls and focus groups, even noxious elements like gerrymandering or voter suppression — stem from the idea that voting preferences can be predicted by demographics, which implicitly invokes the idea that political identification is 'frozen' across broad swathes of social groups. This may more often be 'seat of the pants' odds-making than true scientific investigation, but the scientific assumption is still lurking there in the background.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .