The Brown Bear Bares All!
In this and subsequent posts, the Brown Bear will reveal exactly how he apportions the electoral vote to each candidate. He will start with a brief description of the process.
1. The Brown Bear takes recent national poll results, both the "who would you vote for" result and President Bush's Job Approval, and calculates a National Margin (currently Bush +2.76%). By assigning third party candidates 1.8% of the vote, the Brown Bear calculates the percentages for President Bush and Senator Kerry that appear on each morning update.
2. The Brown Bear examines the previous four presidential elections (1988, 1992, 1996, and 2000) and calculates the Degree of Partisanship of each of the fifty states. The Degree of Partisanship of a state is the number of percentage points the margin of victory (or defeat) in that state exceeds the national average. (The Brown Bear ignores the District of Columbia and arbitrarily assigns its 3 electoral votes to John Kerry.) The Brown Bear adds the Degree of Partisanship for each state to the National Margin, calculated in step 1. The Brown Bear has now calculated a preliminary result for each state.
3. The Brown Bear takes recent poll results in each state and calculates a state margin for each state. The Brown Bear combines this result with the similar result calculated in step 2 to calculate a final State Margin. If the State Margin is greater than 10% points, the Brown Bear designates that state Solid. If the State Margin is less than 10% points, yet greater than 5% points, the Brown Bear designates that state Likely. If the State Margin is less than 5% points, yet greater than 2% points, the Brown Bear designates that state Close. If the State Margin is less than 2% points, the Brown Bear designates that state a Tossup.
This is how the Brown Bear calculates the electoral votes for each candidate. Following this post will be other posts explaining the process in detail.
posted by Trinian | 11:00 AM
4 Comments:
1. As discussed elsewhere, the validity of the "job approval" numbers is a little suspect to my mind. I approve of the "Who would you vote for" question, but I don't see job approval as directly translating into votes. I also see it as a more volitile measure, and susceptible to media influences. Concerted media effort can drive down "job approval" by focussing on negatives (ex. in Iraq, the public may be discouraged by a continual barrage of negative reporting on reconstruction in Iraq and yet continue to favor Bush as the best man to move that effort forward. Conversely, the media can artificially buoy the outlook on the economy, with cheerleading in their reportage, to artificially elevate "job performance" numbers even when a period of economic recession is already well underway).
It is also an interesting question because it focusses entirely on one half of the electoral decision. It specifically does not include Kerry in the equation. That question promotes the assumption that poll respondant who don't "approve" of the job Bush is doing will vote against him when given the opportunity.
1. As discussed elsewhere, the validity of the "job approval" numbers is a little suspect to my mind. I approve of the "Who would you vote for" question, but I don't see job approval as directly translating into votes. I also see it as a more volitile measure, and susceptible to media influences. Concerted media effort can drive down "job approval" by focussing on negatives (ex. in Iraq, the public may be discouraged by a continual barrage of negative reporting on reconstruction in Iraq and yet continue to favor Bush as the best man to move that effort forward. Conversely, the media can artificially buoy the outlook on the economy, with cheerleading in their reportage, to artificially elevate "job performance" numbers even when a period of economic recession is already well underway).
It is also an interesting question because it focusses entirely on one half of the electoral decision. It specifically does not include Kerry in the equation. That question promotes the assumption that poll respondant who don't "approve" of the job Bush is doing will vote against him when given the opportunity.
2. The Degree of Partisanship is a very useful metric. It explores the historic predisposition of a state to vote in a particular way. Conceptually, I think I understand this. However, one question for clarification:
"the degree of partisanship of a state is the number of percentage points the margin of victory in that state exceeds the national average"
Does this mean you calculated the average margin of victory for all 50 states (call that delta RD and say it is 5%); If a state voted for Bush over Gore by 8%, then the degree of partisanship is 3. If the state voted for Gore over Bush by 15%, then what is the degree of partisanship?
2 (cont.)
Now, we take our example state with a 3% degPartisan, and add it to the 2.76% of the national margin and we have a 5.76% call for Bush. In my previous question, when we had a state that went 15% for Gore, its deg of partisanship is 20% Democrat, or -20% Republican added to the national margin gives -17.24 for Bush.
Post a Comment
<< Home