Sunday, October 21, 2012

Another poll with 2008 voting breakdowns

RCP has included another national poll that asks their sample who they voted for in 2008, from the Hartford Currant/University of Connecticut. They give Obama a 3 point edge, 48-45.

I've maintained that the Republicans and Romney are in a much improved position over 2008. This poll actually gives support to this contention, showing that McCain voters are going for Romney 95-5 compared to 2008 Obama voters, who are only going to Obama by 86-10 margin.

In fact, if you compared these percentages to the actual 2008 votes, you would get:

Romney 10% of 69,496,897 plus 95% of 59,934,814 equals 63,887,763 voters.(.1*69496897+.95*59934814=63,887,763)

Obama 86% of 69,496,897 plus 5% of 59,934,814 equals 62,764,072 voters.

So if this poll is correct, and the 2012 election consisted only of 2008 voters, Romney would win by over a million votes!

So how can this poll say that Obama is leading by 3 points? By oversampling Obama voters in comparison to McCain voters. Their sample includes 856 people who say they voted in 2008. Of those, 469 voted for Obama vs 370 for McCain. This indicates an 11.6 Obama margin ((469-370)/856). Obama, of course, only won by 7.2%.

Given that it is instead highly probable that a smaller percentage of Obama voters will vote this year, if this poll had a realistic sample it would show Romney in the lead, probably by a 2-4 point margin.

Tuesday, October 16, 2012

Favorable/Unfavorable ratings

I've seen a couple of people notice that the RCP average of favorability ratings now favor Romney +.4 (Romney +5.4 compared to Obama's +5.0). That's even more remarkable considering it includes a huge outlier from ABC/WaPo. This poll gives Obama a 15 point edge (+11 vs -5) in favorability. Almost every other poll has Romney with the edge to a greater or lesser extent:

Pew:                Romney net +3   (+4 vs +1)
Fox:                 Romney net +4   (+10 vs +6)
Monmouth:      Romney net +6   (+7 vs +1)
Politico:           Romney net -1   (+7 vs +8)
Gallup:            Romney net +5   (+8 vs +3)

Throw out the outliar (heh), and Romney would actually lead this measure by 3.4 points.

PPP, Rasmussen, and Todd Akin

After Todd Akin's moronic comments on rape and pregnancy were first published, both PPP and Rasmussen released polls. They were ... at odds with each other, shall we say. Republicans accused PPP of fudging their (Akin +1) results and Democrats accused Rasmussen of fudging his (McCaskill +10).

Well, that was August. The most recent polling of the race, coincidentally, was also by PPP and Rasmussen, and the dates are almost identical. Interestingly, they each now agree! Pretty clearly, one or the other WAS fudging their data. What makes more sense, that a month and a half after making extremely damaging comments, that a candidate would see a smallish 4 point recovery? Or would they 'recover' by losing another 7 points?

In any event, with the McCaskill corruption scandal that has emerged after both of those polls were released, the race should be getting pretty close.


1. In spite of all the pollster protests that their samples "just come out that way" with D+9 and D+11 samples - miraculously, none of their final national polls will have samples with a bigger slant than D+4.

2. None of the final polls from the 'true' swing states (i.e. Ohio, Florida, Colorado, Iowa, Virginia) will have samples larger than D+4 either. There won't be any that bear any resemblance to the D+11 fantasy Ohio poll that NBC/Marist recently put out. Wisconsin, Pennsylvania, and Michigan poll samples will all be D+single digits.

3. Due to the remarkable decrease in the "Garbage In" factor, Nate Silver's final model will show Romney's chances of winning being greater than 50%, but it will still overstate Obama's position.

4. Any national poll that DOES have a D+4 or D+3 sample will understate Romney's winning margin by at least 2 points (including Rasmussen, if he is still insisting on using that weighting system at that time).

5. Republicans will have 50 or more Senate seats, and there will be very little change in the House.

Sunday, October 14, 2012

Saturday, October 13, 2012

The Best Measure of Sample Bias

There are two competing interpretations of all of the polling from the last couple of months with samples containing many more Democrats than Republicans. The Republicans (including me) contend that these samples are not representative of the actual population of American voters, and so are not showing results that will agree with the election. The pollsters (and Democrats) maintain that they don't weight  their samples this way on purpose, they weight according to the Census reports by categories of age, sex, race, and assorted other criteria, and the results are the results.

Other than looking at the evidence outside of polling (as I've done) like voter registration, historic patterns, etc. - none of which indicate a big Democratic advantage this year - it didn't seem like there was any way to test which of these two interpretations was correct within the actual polling results. However, there IS a poll question that I believe can definitively establish whether the polls are biased toward Obama and the Democrats or not. The only problem is that so few pollsters ask the question and/or report the results. In any event, I have found one national pollster that asks the question (Pew), and one state level pollster who asks the question (WMUR/UNH Granite State Poll).

What's the question? It is, "Who did you vote for in the previous presidential election?"

It is a fact that Obama won nationally by 7.3 points, 52.9 to 45.6. It is also a fact that Obama won in New Hampshire by 9.6 points, 54.1 to 44.5. ANY FAIR SAMPLE should have respondents who match these numbers, within a couple of points. In fact, I'd argue that a correct sample should have a somewhat smaller margin of Obama voters given the depressed Democratic registrations and Republican enthusiasm advantage across the country.

So, have the samples been biased?


Pew is in fact a terrific example. In September they released a poll showing Obama up by 8 points (the "Who did you vote for in 2008?" question is on the top of page 81), and then this week released a poll which caused much wailing and consternation among the Democrats showing Romney up by 4 (This is on page 37). Was this mostly due to Romney's sterling debate performance, or was it mostly due to a shift in the sample?

The September sample included 86% of respondents who stated that they had voted in '08. Of this number, 54.7% (47/86) voted for Obama. Only 37.2% (32/86) of their sample of 2008 voters had voted for McCain! This is a +17.5 Obama sample, from an election that produced a 7.3 point Obama win. In other words (in those who had previously voted), this sample was biased towards Obama by a more than 10 points! There is no spinning this as 'this is just what the numbers are', this is proof positive that this poll strongly over-sampled Obama voters.

In October, the sample included a nearly identical percentage of 2008 voters, 85%. This time, however, only 49.4% (42/85) of these reported voting for Obama vs 43.5% (37/85) McCain voters. This is a much more representative sample, with a 5.9 point Obama advantage vs the 7.3 points he won by, a difference of only 1.4 points.

As I said earlier, I believe that given falling Democratic registrations across the country - a symptom of lower Democratic enthusiasm - that a small change away from Obama voters makes perfect sense. So this Pew poll's sample looks to accurately reflect the race as it currently sits, while the September poll's sample did not. Pretty much the entire 12 point shift from Obama to Romney was a function of fixing the sample.


The University of New Hampshire released a poll on October 1st showing Obama leading Romney by 15 points, 52-37. They don't label their pages, but about three quarters down they show their 'likely voter' statistics, including the number of respondents for each category. If you just looked at their Democratic vs Republican numbers, it actually looks reasonable - 166 Democrats vs 161 Republicans. However, when you look at the breakdown of Obama/McCain voters, the Obama bias pops out. Of the 543 'likely voters' (215+314+14) who voted in 2008, 57.8% of them (314/543) voted for Obama vs 39.6% (215/543) McCain voters. This is a 18.2 point edge in 2008 Obama voters, even though his edge was only 9.6 points - an 8.6 point skew towards Obama.

On Oct 9th, after the debate, they released another poll showing a big shift toward Romney, with Obama leading 50-44 among 'likely voters, with leaners' (again, about 3/4 down). Was this also a result of a shift in sample? Actually, no. The swing appears to be entirely a function of changed minds, because the sample remains just as skewed toward Obama as the previous one. There were 484 reported 2008 voters (289+195), and 59.7% (289/484) were Obama voters while only 40.3% (195/484) were McCain voters. Again, Obama only won by 9.6 points, so this sample is also skewed towards Obama by almost 10 points.

Keep in mind that New Hampshire Republicans have gone from a tiny deficit in registration in 2008 to a comparatively large advantage in the most recent data for 2012, 32.8% of registered voters vs 28.3% for the Democrats.

The Granite State Poll sample simply does not reflect reality. If it was weighted to reflect the actual number of 2008 voters, Romney would be leading. If it was further weighted to reflect the strong trend in registration toward the Republicans, Romney would be leading comfortably.


The evidence from the handful of polls I have found that include the question, "Who did you vote for in 2008?" indicates that those which show Obama doing well have samples that are strongly biased toward 2008 Obama voters. The single poll that did not have this issue showed Romney leading by 4 nationally.

Biden - The Modified Clinton Defense.

In the VP debate, Biden infamously stated, “We weren’t told they wanted more security [in Benghazi] ". When questioned on this bald faced lie, the Obama team's post debate defense has consisted of essentially saying, "That depends on what the meaning of the word 'we' is".

Monday, October 08, 2012

Gallup's Sample Part 1 - Education Demographics

There have been numerous complaints over the samples used by polling organisations this year. They have been reporting an unlikely (to be generous) number of Democrats in their samples. The response has been more or less uniform, something on the order of, "We don't weight our samples to political orientation, because those have a history of change, we only weight by other characteristics like sex, race, age that aren't subject to change to the same degree.

So far, I haven't seen anyone look into how reasonable the breakdown for those things is. This is my first post doing so. I chose Gallup, because they have been quite reputable for a long time. As it turns out, their samples also show major problems in other demographics.

The first issue I will go over  is in the area of education, because not only did the numbers look obviously wrong, they may be the easiest to demonstrate. Gallup's sample over the dates September 10th-September 30th includes 22% (2162 of 9842) people with a 'Post-Graduate' education level. Nearly a quarter of registered voters hold advanced degrees?! Impossible. According to 2011 Census tables (the top file), the percentage of the adult population with better than a bachelor degrees is less than 10% (22 million out of 231 million). While I'm sure that those with advanced degrees are more likely to be registered than the average American, we had approximately 187 million registered voters in 2008. Even if every single one of those 22 million people with post-graduate degrees were registered, if the number of registered voters is similar this year, they would account for less than 12% of registered voters (22/187).

This hugely over-sampled group goes to Obama by almost 3:2 - 57% prefer Obama to 39% for Romney.

Another major issue with the educational demo is that the 'College Graduate' demographic is under-sampled. There are over 62 million Americans with Bachelor and/or Associate degrees, according to that same Census table - nearly 3 times as many as hold advanced degrees. Yet they only make up about 24% of the sample. According to statistics posted by George Mason University, 87.9% of eligible Americans registered to vote in 2008. If anything, college graduates should be more likely to register than the average American. However, using that percentage would give us about 55 million registered college graduates (.879 times 62.5 million). That would, in turn, indicate that a representative polling sample should include at least 29% college graduates (55/187).

This under-sampled demographic leans toward Romney, 50-44.

Adjust for both of these discrepancies, and Obama's lead goes poof. When you consider that this is a registered voter poll, and actual voters almost always turn out to favor Republicans by several points when compared to registered voter polls (especially this year), and Romney is actually leading.