Very interesting, and quite surprising that (as many as) 39% of Americans say they are working class, given the cues built into the poll which would encourage respondents to interpret class as status. Look at the question on the necessary elements of a "middle class life". This is obviously an extremely leading question, in that it specifically authorised respondents to treat class in terms of relations of consumption rather than production.
I'd be interested to see what mixture of motivations and interests lies behind this self-identification. Vanneman and Cannon, whom you cited recently, argued that there was a strong correlation between Americans' class self-identification and their actual position in productive relations, and I would expect this 'class consciousness' to have informed the responses in this poll, despite its format. I expect that many of those who identified as working class have all of the "elements" essential to a "middle class life", and thus defied the poll's implicit bias. Pierre Bourdieu's critique of opinion polls (Sociology in Question, 1993) is relevant here, in that this surveys presumes what it in fact produces: a consensus on the nature of problems supposedly being addressed, on the relevant questions, on the appropriate framing of the issues.
-- Richard Seymour Writer and blogger Email: leninstombblog at googlemail.com Website: http://www.leninology.blogspot.com Twitter: http://www.twitter.com/leninology Wiki: http://en.wikipedia.org/wiki/Richard_Seymour_(writer) Book: http://www.versobooks.com/books/nopqrs/s-titles/seymour_r_the_liberal_defense_of_murder.shtml