>
>The more you elaborate, the more that seems true -- the respondents hear
>that key phrase and they know immediately that this a political
>referendum, not a value free query. And they respond accordingly, not to
>the specific request for facts as much as to the underlying question: have
>Clinton & Gore given us a good economy, yes or no?
>
>You hear a lot about pollsters duping people with leading questions.
>This seems a good example of lazy pollsters allowing people to outwit them
>by leaving in flag words.
>
>Michael
i don't see how so. sure, there's the "compared to four years ago" aspect. well, if you want to get a sense of whether there's a real perception or not, i suppose you could look back at all the opinion polls taken during an election year, since that will give you a sense of whether there are any distinct differences this year.
it's just not clear to me why folks think the average person is going to hear that question and have it register. ever been polled? having done some polling, i'll tell you that most people want to get the hell off the phone! people don't psych the question out altogether that much, imnsho.
at any rate, what stake does anyone have in keeping the dems in office as opposed to the repugs? did folks all of a sudden change their minds about the meaninglessness of the distinction (referring back to doug's post about people not thinking it matters esp. much). perhaps folks really are doing better --or, at least, perceive that they are doing better.
their answer may well be "true". remember, the question is asking about perceptions so, as long as you use if for that then this is what it means: blacks perceive themselves as doing better compared to 4 yrs ago.
americans typically say, btw, that they are doing well. when asked back to back: "what do you think the state of the health care system is.." and "how do you rate your own health care" people will rate their own health care better than the US as a whole.
i would argue, given the above, that some ethnographies comparing poor/working class blacks and whites often show that blacks often are more inclined to believe in the system when younger. so, disaggregating the data might be a way to answer the questions you have -- by age and income would seem most appropriate. it would be interesting to know just who among blacks feels this way -- if it's an evenly distributed response or if there are differences depending on age, gender, income. but, this is probably a 1000 N survey which means we're working with about 140 blacks and that N is too small to do a whole lot with unless they oversampled.
kelley