Fox National Poll: Clinton +3/+5

[cinyc]

How are they defining urban, suburban, and rural?  What %age of the country do they put in each basket?  If it was an even split between the three, then the average would be (+22 +7 -17) / 3 = +4, for a 4 point Clinton lead.

It might be self-reported.  I was recently polled on a state race and asked the question.  But it could imputed from the area code and telephone prefix (at least for landlines), like Google Consumer Surveys does in its Internet polling.  Without asking the Fox News pollsters, we don't know.

[cinyc]

Here are the MoE's for the urban/suburban/rural subsamples:
Urban: +/- 5.5%
Suburban: +/- 4%
Rural: +/- 5%

So there are more suburbanites than rurals than urbans, however the Fox pollster defines them.

[cinyc]

I thought this was discussed at the time of the last Fox News national poll...

Regardless, even if Trump is up 17% among "Rural" voters, however that is defined while losing by large margins with urban/suburban voters, it would be hard to find a silver lining even if one were a Trump supporter.

But yeah, we can't easily extrapolate date from this one particular poll and compare/contrast based upon 2012 exit polls, but regardless of how one chooses to slice and dice, these numbers don't look at all good for Trump unless there are a ton of urban/suburban voters based upon area codes/ zip codes/ self-ID or what have you that fall into a "rural" category considering that over 80% of Americans live in urban/suburban areas by any measure whatsoever.

Well, we know from the subsample MoEs in this poll that Suburban>Rural>Urban.  So there are more rural residents (however defined) than urban residents.  Whether that's because self-reporting residents of, say, Staten Island or the less dense areas of northern San Antonio, claiming they live in a suburban rather than urban area despite technically living in a city, or exurban residents claiming they live in a rural area, or the Fox pollsters have some other criteria like density or whether someone lives in an urban cluster we don't know.  But we do know that there are fewer urbans in this poll than any other type.