Demographics and the Electorate

[Mr. Morden]

Is there a compelling hypothesis as to *why* LV screens might be doing particularly poorly this time?  If a demographic group is being systematically over or under-estimated as LVs, then why is that more of an issue this year than other years?  Are NCWs so psyched by Trump that they're more likely to tell pollsters that they're enthusiastic about voting, or what?

And I know that each individual pollster doesn't like to divulge this kind of info, but has anyone written a good article that describes the most common methods that pollsters use to rate respondents as LVs?

[Mr. Morden]

Aren't there better polls than Reuters to use for this?

I can't find any other giving projected turnout rates by different groups. There's a lot of smoke and mirrors when it comes to pollsters stating what the LV screen actually is.

Even without turnout rates don't a lot of pollsters publish crosstabs with nr of respondents included? I'm guessing that would allow backing out what they think the shares of the electorate are.

Are you after the raw number of respondents in each demographic group or the %age of the electorate that they are assumed to make up (after weighting)?  I would think the latter would be the more relevant number, and I'm pretty sure that PPP for example always includes those percentages, for example.

[Mr. Morden]

Is there a compelling hypothesis as to *why* LV screens might be doing particularly poorly this time?  If a demographic group is being systematically over or under-estimated as LVs, then why is that more of an issue this year than other years?  Are NCWs so psyched by Trump that they're more likely to tell pollsters that they're enthusiastic about voting, or what?

And I know that each individual pollster doesn't like to divulge this kind of info, but has anyone written a good article that describes the most common methods that pollsters use to rate respondents as LVs?

I don't think it's about LV screens being a problem this time. It's about LV screens being a problem. The polling in 2012 was off, especially for the Romney campaign. Gallup and others looked at the problem and concluded that LV screens were overreliant on behavior from 2008 that didn't apply to 2012.

HuffPo has all the LV national polls leading up to election day in 2012, and it has an Obama lead of 1.5%:

http://elections.huffingtonpost.com/pollster/2012-general-election-romney-vs-obama

Obama won by, what?  3.9%?  So that's a 2.4% "error" in the margin between the two candidates.  Nothing too unusual, if memory serves.  I thought the whole premise of this thread is that people are questioning whether the LV screens might be off by a lot more than 2.4% this time?  In the first page of the discussion, there's talk about how Clinton's lead "should be" in the double digits if the demographics are what they're expecting, when it's actually about 1 or 2% at the moment.

[Mr. Morden]

Well for example if we take the Bloomberg poll, they don't even provide demographic breakdowns at all. Monmouth gives me that, but not raw numbers to calculate turnout and so on. A few polls give what they assume the makeup of the electorate is, but not differential turnout. Indeed some of them appear to be doing exactly what they did in 2012 with respect to weighting.

If they give you what % of the electorate is white, black, young, old etc. in their poll release, then can't you work out the *relative* turnout of those groups from that?  Not an absolute number, but just relative turnout of the groups compared to each other, using the 2012 exit poll to compare to.

Or is the whole idea that we want absolute turnout more than relative turnout?  I've lost the thread as to what the original question being asked was. 

[Mr. Morden]

I think the original question being asked was where Trump is making up ground in the polls relative to what we'd expect from demographics. Is white turnout being driven up like crazy relative to non-white turnout? Is college white turnout depressed? Is male turnout up relative to female turnout? Demographics may not be destiny, but some explanation is in order when the demographics from last time, combined with what the polling tells us of the demographics now, give us such a starkly different result from the polls.

Well, there are two different sets of parameters, right?  1) What %age of the vote does each demographic make up?  and 2) Within that demographic group, what %age of them are voting for Clinton, Trump, or someone else?

If a pollster tells you in their poll release what %age they have for whites, blacks, etc., and they tell you how each of those demos is voting, then you can use that info to figure out what demos are most out of whack with your a priori expectations, no?  Or am I missing something?

[Mr. Morden]

This last point seems, to me, to go against pretty much everything we know about elections.  Of course, 2016 is a weird year with weird candidates.  It's possible that this is indeed the case, and presumably that is what Likely Voter screens are showing.  But I've got a pretty large Bayesian prior against that, and personally I'm more willing to believe that the Likely Voter screens are wrong than that the bolded point above is true.

OK, but that goes back to my earlier question: Why are they wrong?  What are the LV screens selecting on that makes them pick up more uneducateds this time?

And also, going back to my last post, aren't there some poll releases that actually tell you what % of the electorate they have for college-educated whites, non-college-educated whites, etc.?  Can we look at them to see what they're showing, and if the fraction of the electorate for each group is indeed ridiculous?

[Mr. Morden]

Of course, census data shows different results (see e.g. this US Census report on the 2000 election), though I'm more inclined to believe exit polling (where you know you got voters) rather than retrospective polling where people may say they voted when they did not.  That isn't to say exit polling doesn't have its own issues, however.

With exit polls, you know you've got voters, but do you know that people are telling the truth about their education levels?  Isn't it pretty widely assumed that there's some lying going on when people are asked about their educational attainment in polls, inflating the number of those with college degrees above what we realistically think it must be?  I don't suppose there's any reason why people might be lying about their educational level less now than they were a few years ago?  Maybe Trump's made it cool to be "uneducated"? 

[Mr. Morden]

This last point seems, to me, to go against pretty much everything we know about elections.  Of course, 2016 is a weird year with weird candidates.  It's possible that this is indeed the case, and presumably that is what Likely Voter screens are showing.  But I've got a pretty large Bayesian prior against that, and personally I'm more willing to believe that the Likely Voter screens are wrong than that the bolded point above is true.

OK, but that goes back to my earlier question: Why are they wrong?  What are the LV screens selecting on that makes them pick up more uneducateds this time?

And also, going back to my last post, aren't there some poll releases that actually tell you what % of the electorate they have for college-educated whites, non-college-educated whites, etc.?  Can we look at them to see what they're showing, and if the fraction of the electorate for each group is indeed ridiculous?

Here's an example from the latest Bloomberg / Selzer poll.  Likely Voters are those who say they will "definitely vote" or have already voted.

The latest Bloomberg / Selzer poll had 62 non-college - 38 college in their latest poll, very similarly to ABC/WaPo. Again, this is a huge difference from the 2012 exits (53 - 47).  

Though it should be noted that it's less of a departure from previous years (56 - 44 in 2008 and 58-42 in 2004, and I believe 58-42 in 2000).

I don't suppose anyone with a lot of time on their hands would be willing to go through all the national polls one by one to see if they have crosstab info on the %age of the sample with college degrees, and then we could look at whether there's really a systematic deviation from what we would expect?  Also could see if there's a strong correlation between the %age of the sample that has college degrees and the overall support for Clinton.

[Mr. Morden]

How does the census dataset work?  The census bureau asks people some demographic questions, and also includes the question "Did you vote in the last presidential election?", then releases the demographic breakdown of people who answered "yes" to that question?

[Mr. Morden]

In 2012:
Actual Turnout: 129.2 m
CPS Turnout: 132.9 m

That's only a 3% difference, which is actually much less of a mismatch than I would have guessed.  It seems odd to me that a 3% discrepancy in turnout #s would translate into a 10% discrepancy in the fraction of voters with a college degree.  The best explanation I can think of is that people are more likely to lie to exit pollsters about their educational status than they are to lie to the Census Bureau.

What alternative explanation is there?  Did either the Census Bureau or the exit pollsters do a poor job of selecting a representative sample of voters?

[Mr. Morden]

In 2012:
Actual Turnout: 129.2 m
CPS Turnout: 132.9 m

That's only a 3% difference, which is actually much less of a mismatch than I would have guessed.  It seems odd to me that a 3% discrepancy in turnout #s would translate into a 10% discrepancy in the fraction of voters with a college degree.  The best explanation I can think of is that people are more likely to lie to exit pollsters about their educational status than they are to lie to the Census Bureau.

What alternative explanation is there?  Did either the Census Bureau or the exit pollsters do a poor job of selecting a representative sample of voters?

My personal guess is that exit polls are just oversampling college educated voters.  College educated voters are more enthusiastic about voting / civics in the first place (as evidenced by their higher turnout, by any measure), and perhaps are more interested in spending time with an exit pollster.

There are of course also the usual geographic sampling concerns with exit polls, but it's hard to say a priori what sort of effect that would have.

OK, here's my stupid question: What kind of demographic weighting, if any, do the exit pollsters do?  Do they just pick precincts that are demographically representative, and then poll every third person there, and let the chips fall where they may?  Or do they do anything to account for the possibility of different response rates from different demographic groups....other than weighting the eventual "topline" numbers (what % voted for Obama vs. what % voted for Romney) to match the election result?

[Mr. Morden]

And then there's the question...even if the exit poll #s are biased in a way that the Census #s are not, does that really mean that we should be weighting the current round of telephone polls to match the Census #s?  The telephone polls could have the same bias that the exit polls have.

[Mr. Morden]

And then there's the question...even if the exit poll #s are biased in a way that the Census #s are not, does that really mean that we should be weighting the current round of telephone polls to match the Census #s?  The telephone polls could have the same bias that the exit polls have.

If it's a response bias that's an issue in the exit polls, then we should certainly be weighting them to match the census; that's why pollsters do demographic weights.

If the bias is due to people lying about their education, then they shouldn't.

Good point.  If you're right that it's a response bias in the exit polls, then I think that also means that the exit polls from past elections have also gotten it wrong wrt what % of the electorate with a college degree was voting Democratic vs. what % were voting Republican....since they are weighting the toplines to match the election outcome.

I guess one thing to do if someone has a lot of free time on their hands is to go back to pre-election polls from 2012, and look at what they were estimating *before* the election both for the fraction of the electorate with a college degree, and the fraction of each educational group supporting Obama or Romney.  Do those #s line up more with the exit polls or the census #s?

[Mr. Morden]

*bump*

So supposedly the demographic crosstabs were supposed to be telling us that the topline #s in the polls were too Trump-friendly, when in fact it was the opposite.

So what happened?  Which demographics did the pre-election polls get right and wrong?

[Mr. Morden]

I'd really like to see someone compare the exit poll demographic breakdown of 1) who showed up and 2) how each group's support split against what the polls were saying....and against the predictions in this thread of what was "realistic".

But I don't have time to do that myself.