ronpaulyourmom
11-20-2007, 12:54 AM
I did a little research, and here are my half-educated conclusions.
Myth #1: The phone polls are way off because they don't call cell-phones
Answer: Half-true
It's true that phone polls don't call cell-phones for the most part (some exceptions but none meaningful). It's also true that something like 30% of all young adults (under 30) are cell-only. That said, it doesn't necessarily mean that the young demographic is completely missed. Although they tend to get fewer young people on the phone, pollsters add weight to the data as a way to compensate.
The real issue here is whether or not a young adult reached via land-line is accurately representative of young adults who support Ron Paul. For example are Ron Paul supporters also early adopters? (our internet presence suggests we are) If so, they might be more inclined to be cell-only or VOIP. Cell-only users are also more likely to be minorities (by a small margin), slightly more likely to be male (13% vs 11%), and less likely to be homeowners. It's hard to say to what extent these variables miss the Ron Paul support base, but just based on what I've learned about most supporters I would say that both Ron Paul and Barack Obama are under-represented in this area.
Myth #2: The phone polls are heavily biased to candidates with high name recognition and do not measure enthusiasm / turnout
Answer: True, but...
Even though polls tend to over-inflate candidates with high name recognition at this stage of the game, the bottom line is that a good size chunk of the people who were included in the poll results are people who will also vote on voting day. Whenever you see a poll that has responses from 400 or so people, you should understand that something like 1000 were called and the 400 included in the results were put there because they indicated their intention to participate in the primary voting season. Plenty of un-informed voters participate in polls, and plenty of them show up on election day as well. Ron Paul supporters can certainly brag about enthusiasm, and I have no doubt we'll turn-out more than any other candidate, but do understand that many less enthusiastic supporters of other candidates still consider voting their civic duty and will participate. That is especially true this year, given the low approval ratings for the establishment on both sides. The closer we get to voting day, the more accurate these results become in the real world.
The real thing to watch here is the trends. Most voters claim to be undecided at this point, so a careful analysis of the overall trends can reveal a lot. For example, the more people find out about Rudy, the more they don't like him. The opposite is true for Romney, and McCain seems to be stuck where he is. Ron Paul's trend is very good, but his starting point is very bad because of the media hurdles he's having to overcome. Our name recognition is still below 70% nationally, compared with 85%-98% for all our competitors, so we'll get some extra mileage as we close the gap, but the truth is we still need to win over hearts and minds. Be on your best behavior!
Myth #3: The phone polls measure male/female support inaccurately
Answer: True.
As you may or may not know, Ron Paul gets a minimum of 70% of his support from men. Huckabee is dead even, Rudy and McCain have slight preference among men, and Romney gets a lot of support from women. Ron Paul is the most disproportionately male supported candidate in the race.
The problem with phone polls is that they assume an equal number of men and women will show up on voting day. Traditionally, slightly more women vote than do men (democrat and republican combined). However the data seems to suggest that a huge portion of women are turning out for democratic candidates this year, and so it may just be the case that more men will show up to the republican voting booths this year than women. If that turns out to be the case, then Ron Paul's numbers are off.
My own conclusions:
- Romney's IA/NH numbers are inflated. (female vote)
- Rudy's national numbers are inflated. (name recog.)
- McCain's national numbers are inflated. (name recog.)
- Hillary's national numbers are inflated. (name recog.)
- Huckabee's numbers are probably close to true.
- Obama's numbers are under-inflated.
- Ron Paul's numbers are under-inflated. (to what extent we can't really say)
Myth #1: The phone polls are way off because they don't call cell-phones
Answer: Half-true
It's true that phone polls don't call cell-phones for the most part (some exceptions but none meaningful). It's also true that something like 30% of all young adults (under 30) are cell-only. That said, it doesn't necessarily mean that the young demographic is completely missed. Although they tend to get fewer young people on the phone, pollsters add weight to the data as a way to compensate.
The real issue here is whether or not a young adult reached via land-line is accurately representative of young adults who support Ron Paul. For example are Ron Paul supporters also early adopters? (our internet presence suggests we are) If so, they might be more inclined to be cell-only or VOIP. Cell-only users are also more likely to be minorities (by a small margin), slightly more likely to be male (13% vs 11%), and less likely to be homeowners. It's hard to say to what extent these variables miss the Ron Paul support base, but just based on what I've learned about most supporters I would say that both Ron Paul and Barack Obama are under-represented in this area.
Myth #2: The phone polls are heavily biased to candidates with high name recognition and do not measure enthusiasm / turnout
Answer: True, but...
Even though polls tend to over-inflate candidates with high name recognition at this stage of the game, the bottom line is that a good size chunk of the people who were included in the poll results are people who will also vote on voting day. Whenever you see a poll that has responses from 400 or so people, you should understand that something like 1000 were called and the 400 included in the results were put there because they indicated their intention to participate in the primary voting season. Plenty of un-informed voters participate in polls, and plenty of them show up on election day as well. Ron Paul supporters can certainly brag about enthusiasm, and I have no doubt we'll turn-out more than any other candidate, but do understand that many less enthusiastic supporters of other candidates still consider voting their civic duty and will participate. That is especially true this year, given the low approval ratings for the establishment on both sides. The closer we get to voting day, the more accurate these results become in the real world.
The real thing to watch here is the trends. Most voters claim to be undecided at this point, so a careful analysis of the overall trends can reveal a lot. For example, the more people find out about Rudy, the more they don't like him. The opposite is true for Romney, and McCain seems to be stuck where he is. Ron Paul's trend is very good, but his starting point is very bad because of the media hurdles he's having to overcome. Our name recognition is still below 70% nationally, compared with 85%-98% for all our competitors, so we'll get some extra mileage as we close the gap, but the truth is we still need to win over hearts and minds. Be on your best behavior!
Myth #3: The phone polls measure male/female support inaccurately
Answer: True.
As you may or may not know, Ron Paul gets a minimum of 70% of his support from men. Huckabee is dead even, Rudy and McCain have slight preference among men, and Romney gets a lot of support from women. Ron Paul is the most disproportionately male supported candidate in the race.
The problem with phone polls is that they assume an equal number of men and women will show up on voting day. Traditionally, slightly more women vote than do men (democrat and republican combined). However the data seems to suggest that a huge portion of women are turning out for democratic candidates this year, and so it may just be the case that more men will show up to the republican voting booths this year than women. If that turns out to be the case, then Ron Paul's numbers are off.
My own conclusions:
- Romney's IA/NH numbers are inflated. (female vote)
- Rudy's national numbers are inflated. (name recog.)
- McCain's national numbers are inflated. (name recog.)
- Hillary's national numbers are inflated. (name recog.)
- Huckabee's numbers are probably close to true.
- Obama's numbers are under-inflated.
- Ron Paul's numbers are under-inflated. (to what extent we can't really say)