Home Hubdub logo
 
Guest account   Cash: hd$1,000   Predictions: hd$0
You currently have hd$1,000 (Hubdub dollars), Hubdub's play money, to stake on your predictions. Your predictions are currently worth hd$0
Home
Leaderboards
Forums
My Predictions

Will race upset the 2008 US Presidential Election?

Voided

Market voided. This market seems to assume that any difference between polling numbers and vote totals will be race related, and I don't think that is a fair way to judge that. If you want to recreate this market, and take race out of the title, and then just see what the difference between polling of gallup daily and actual results that is fine. But I don't like idea of attributing the entire difference to just race.

Background:

This question attempts to quantify the extent that racial effects play in actual voting compared with polls in the present election.

There is an effect, called the Bradley effect, named after Tom Bradley, mayor of LA, who was running in 1982 for governor of California. He lost, despite being more than 10 points ahead in the polls.

They think that change was due to a racial effect, wherein people will say one thing in a poll, and then vote another when in a voting booth. So, we will answer this question by comparing the Gallup poll on November 3rd, with the recorded popular vote on Nov. 4th. By taking the difference between McCain and Obama, calling this M-O(V-O). This is a measure of how much McCain gains compared with Obama in the vote compared with the polls, in percentage points.
If the value is near 0, or even less than 0, there is no apparent effect. In Bradley's case, it was > 10%, so the effect was large. It now is a different time, and a different place, but can the effects of racism be so easily erased?


Settlement details: As reported by a major mainstream news source, using the Gallup poll before the election and the election results afterwards.

 
Forecast history, %
   Zoom in

Voided

M-O(V-O) less than 0
20%
M-O(V-O) less than 5% (and greater than 0)
35%
M-O(V-O) less than 10% (and greater than 5)
35%
M-O(V-O) greater than 10%
10%
Voided Sun 12th Oct 7:11pm PDT

Suspend date: Mon 3rd Nov 11:59pm PST

Initial likelihoods: M-O(V-O) less than 0: 20%, M-O(V-O) less than 5% (and greater than 0): 35%, M-O(V-O) less than 10% (and greater than 5): 35%, M-O(V-O) greater than 10%: 10%

Action history:

Created Sun 12th Oct 4:10pm PDT by bigken1
Voided Sun 12th Oct 7:11pm PDT by destry[Admin]: Market voided. This market seems to assume that any difference between polling numbers and vote totals will be race related, and I don't think that is a fair way to judge that. If you want to recreate this market, and take race out of the title, and then just see what the difference between polling of gallup daily and actual results that is fine. But I don't like idea of attributing the entire difference to just race.

Suspend date: Mon 3rd Nov 11:59pm PST
more info...

 

Predictions (1)

1 prediction

7 weeks ago
onedave[Power User] predicted M-O(V-O) less than 5% (and greater than 0) (H$20 at 35%)

Comments (1)

Although I think this is an interesting question in general, it seems rather overstepping to necessarily call the discrepancy between polls and the final result as being due to a race/Bradley effect. In general Democrats running for President, regardless of race, tend to do better in the final polls than they do in the actual election. Just look at the past two elections. Depending on who you ask this can be explained by a variety of factors, e.g. poor youth vote turnout, electoral fraud, poll over/undersampling, tactical use of ballot initiative (e.g. gay marriage), etc.
posted 7 weeks ago

Please log in or join to add a comment

What is Hubdub?

Hubdub makes news more exciting by letting you stake virtual dollars on the outcomes of real running news stories.

Join now   or   learn more