In the last few posts of this series we've seen how some of the numbers for minorities in the PPP polls are unreliable. But their toplines are right in line with other polls this year, and, on average, very good in prior years compared to actual results. So how could this be? Well, obviously, if minority demographics are skewed towards Republicans, majority demographics must be skewed towards Democrats. And that is indeed the case, or at least it was for 2010, when comparing to exit polls:
We see something similar with the 'minority' age groups, those under 45, that have the worst response rates:
A few days ago, I explained the deviation in geographic distribution of Asians and Hispanics with a simulation of about 9% of respondents pushing the wrong button for race. That simulation also suggests that around 2% of 'white' respondents are actually minorities who pushed the wrong button for whatever reason. That would lead to about a 2 point Democratic bias for the margin in the white numbers, assuming these minority respondents answered the ballot test questions correctly and in the same proportion as other minority respondents. What we actually see in 2010 is an average 3.6 point Democratic bias in the margin among respondents saying they are white, so a good portion of this bias could be explained by incorrect responses.
It does not even out completely, as PPP's 2010 polls had a slight skew to the right (by about a point in the margin). That is likely where the lack of cell phones comes directly into play.
So, despite everything, the toplines can still be trusted (at least they could in 2010). In other words, if you try to 'adjust' a PPP poll by 'correcting' the low ballot test numbers for minorities or youth, you will likely be sorely disappointed come November. This is very likely to be true for other automated pollsters as well, and possibly to some extent polls that use live interviewers too.
Remember, though, that the ballot test numbers for youth and minorities are, indeed, incorrect. And although in theory we can still watch for trends, we must use extreme caution, at least in the Daily Kos polls, as the number of respondents (50-80 per week for ages 18-29, for example) is too small to be able to spot a typically slight trend in amongst all the noise. Massive changes can still be detected, however, such as the huge swing in opinion among African-Americans regarding marriage equality.
Beyond the Margin of Error is a series exploring problems in polling other than random error, which is the only type of error the margin of error deals with.
The Curious Incident of the Young Republican Minorities. Only a little over half of respondents in the category of African-Americans age 18-29 said they approved of Obama - but only because many of those respondents weren't actually African-American or age 18-29. The numbers for the 18-29 age group are inaccurate as well.
This Is Why We Can't Have Nice Things. A small number of respondents press the wrong button when answering the DailyKos poll question on race, leading to inaccurate numbers for racial minorities in the crosstabs.
Why Don't People Know Where They Live in the DKos Poll? A small number of respondents - around 5-9% - press the wrong button when answering the geography question on the Daily Kos poll. This is far greater than than can be explained by observed rates of misunderstandings or data entry errors.
Why State Polls Look More Favorable For Obama than National Polls. In the spring and summer, lack of support in Blue States was bringing down Obama's performance in national polls, while Swing States and Red States were polling about the same as 2008.
Presidential Polls Are Almost Always Right, Even When They're Wrong. How the presidential polls in red and blue states are off, sometimes way off, and how to predict how far off they'll be.
When Polls Fail, or Why Elizabeth Warren Will Dash GOP Hopes. Why polls for close races for Governor and Senate are sometimes way off, and how to predict how far off they will be.