2008 Metrics: National Polls and Name ID
by Chris Bowers, Sun Nov 19, 2006 at 11:35:07 AM EST
For me, there is a bright side to Russ Feingold not running for President. Now, I can go through numbers and conduct analysis on the state of the campaign without anyone accusing me of bias. And so, I happily bring you my first edition of 2008 metrics. In this edition, I will look at the relationship between name ID and national Democratic primary polls.
Yesterday, there was a diary on the latest national primary poll from Pew. Many of the comments centered around the impact of national name recognition on the imaginary trial heat numbers (I write "imaginary" both because the field has not been set yet and because there is no national primary). Some commenters argued that the poll was all name ID, while others argued that name ID played a role, but since not all high name ID potential candidates were performing equally, other factors were at play besides name ID. I think the latter commenters are definitely correct, as I believe an analysis of the five national primary polls over the past month, Pew, Gallup, CNN, McLaughlin, and Rasmussen, reveals.
Yesterday, there was a diary on the latest national primary poll from Pew. Many of the comments centered around the impact of national name recognition on the imaginary trial heat numbers (I write "imaginary" both because the field has not been set yet and because there is no national primary). Some commenters argued that the poll was all name ID, while others argued that name ID played a role, but since not all high name ID potential candidates were performing equally, other factors were at play besides name ID. I think the latter commenters are definitely correct, as I believe an analysis of the five national primary polls over the past month, Pew, Gallup, CNN, McLaughlin, and Rasmussen, reveals.
Across the five polls, there were only five candidates included in every question: Clinton, Obama, Edwards, Gore and Kerry. Looking at the numbers for Democrats and Democratic leaners for each of the five polls, you receive the following average for every candidate:
National Trial Heat Mean
Clinton: 30%
Obama: 20%
Gore: 11%
Edwards: 10%
Kerry: 8%
(Note: the Pew numbers used for this average were two parts Democrats, and one part Independents)
In January of 2008, all of the major candidates will have equal name ID in places like Iowa and New Hampshire. However, right now, that is not the case nationally. According to Qunnipiac's latest national thermometer poll of leading political figures from both parties, here was the national name ID for these five candidates among Democrats and Independents:
National Name ID Average
Clinton: 99%
Gore: 97%
Kerry: 94%
Edwards: 75%
Obama: 47%
Clearly, some candidates are over-performing compared to their current national name ID, and other candidates are under-performing. If the mean trial heat number for each candidate was divided by his or her name recognition number, the results would be as follows:
Known Universe Metric
Obama: 42%
Clinton: 30%
Edwards: 13%
Gore: 11%
Kerry: 9%
I call this the "known universe" metric because this is the percentage of supporters each candidate is winning within the universe of people who know the candidate well enough to form an opinion about him or her. It is a crude means of normalizing trial heat numbers for name recognition discrepancies. Right now, 42% of Democrats and Democratic leaners who know about Barack Obama are willing to give him their preliminary support for the Democratic nomination. That is clearly a very good position to be in right now.
I do not think that this metric means Obama should be considered the favorite. What it means is that he has done an extremely good job appealing to people so far. However, rising from a 47% name recognition to a 99% name recognition is a path filled with many dangers, and there is absolutely no guarantee he could maintain this current pace. Clinton, Kerry and Gore can probably tell Obama about what happens to your national image in the age of the Republican Noise Machine during the time period when your name ID rises from 40-45% all the way to 95%-100%. It rarely ends well for Democrats.
I also think that this analysis shows that Clinton's current leads in the imaginary national primary is not based just on name recognition. Clearly, she is well ahead of both Kerry and Gore, who both have name ID's near Clinton's level. Right now, she is winning a much higher percentage of her known universe than are either Gore or Kerry.
Now, does that mean it will stay like this? Of course not. Support at this point is extremely soft. Further, the field is not even set yet. Yet further, there is over a year of campaigning between now and the first primary contest. Perhaps most importantly, there is no national primary. Right now, at this very early date, all I can note is that Clinton's support is currently not just based just on name ID, and Barack Obama appears to have the most potential for upward movement. However, over the next fourteen months, everything can change.
National Trial Heat Mean
Clinton: 30%
Obama: 20%
Gore: 11%
Edwards: 10%
Kerry: 8%
(Note: the Pew numbers used for this average were two parts Democrats, and one part Independents)
In January of 2008, all of the major candidates will have equal name ID in places like Iowa and New Hampshire. However, right now, that is not the case nationally. According to Qunnipiac's latest national thermometer poll of leading political figures from both parties, here was the national name ID for these five candidates among Democrats and Independents:
National Name ID Average
Clinton: 99%
Gore: 97%
Kerry: 94%
Edwards: 75%
Obama: 47%
Clearly, some candidates are over-performing compared to their current national name ID, and other candidates are under-performing. If the mean trial heat number for each candidate was divided by his or her name recognition number, the results would be as follows:
Known Universe Metric
Obama: 42%
Clinton: 30%
Edwards: 13%
Gore: 11%
Kerry: 9%
I call this the "known universe" metric because this is the percentage of supporters each candidate is winning within the universe of people who know the candidate well enough to form an opinion about him or her. It is a crude means of normalizing trial heat numbers for name recognition discrepancies. Right now, 42% of Democrats and Democratic leaners who know about Barack Obama are willing to give him their preliminary support for the Democratic nomination. That is clearly a very good position to be in right now.
I do not think that this metric means Obama should be considered the favorite. What it means is that he has done an extremely good job appealing to people so far. However, rising from a 47% name recognition to a 99% name recognition is a path filled with many dangers, and there is absolutely no guarantee he could maintain this current pace. Clinton, Kerry and Gore can probably tell Obama about what happens to your national image in the age of the Republican Noise Machine during the time period when your name ID rises from 40-45% all the way to 95%-100%. It rarely ends well for Democrats.
I also think that this analysis shows that Clinton's current leads in the imaginary national primary is not based just on name recognition. Clearly, she is well ahead of both Kerry and Gore, who both have name ID's near Clinton's level. Right now, she is winning a much higher percentage of her known universe than are either Gore or Kerry.
Now, does that mean it will stay like this? Of course not. Support at this point is extremely soft. Further, the field is not even set yet. Yet further, there is over a year of campaigning between now and the first primary contest. Perhaps most importantly, there is no national primary. Right now, at this very early date, all I can note is that Clinton's support is currently not just based just on name ID, and Barack Obama appears to have the most potential for upward movement. However, over the next fourteen months, everything can change.
Tags: Democrats, polls, President 2008, Primary Elections (all tags)
-----------
myDD - skin
-----------
78 Comments