« Poll Positions II | Main | My talk as reported by The Hindu »
Poll Positions III
It's past midnight for me, but we have some reasonable data so far: Bush will likely win the popular vote because Kerry did worse in the Gore states and Bush will either tie or win the Electoral college with about 275-285 votes. Ohio might tighten as the morning wears on and absentee ballots/early votes are counted. The massive number of absentee/early ballots in New Mexico might flip the state to Kerry. So I'll wait till tommorrow to see the final EC count, but I can safely say the following...
The following folks were right:
1. The final round of pre-election polls by major news organizations and the reputable polling organizations. If you average them, you got a Bush lead - not huge - but it was there. Individual polls were wrong, but the batch was right as a whole.
2. The Iowa Electronic Markets up till Monday. The Bush vote share was selling at $.51 - probably close to final number.
3. My sociology undergraduates (scroll down to the "Poll Positions II" post below) collectively predicted a clear Bush popular vote win. The students predicted about 49.7% for Bush. They might have done better if they had not overestimated the Nader vote.
The following folks were wrong:
1. Fabio Rojas. Following the incumbent rule, I believed that the undecideds would go for Kerry. I was wrong. Plain wrong. [Note: There is a slim chance that my electoral college prediction might come true.]
2. John Zogby - Early on election day, polling world demi-god Zogby predicted a Kerry blow out. Even if Kerry wins the electoral college, it might be because of slim Kerry win in Ohio and the New Mexico lead evaporating for Bush. Kerry could take the day, but not because he won the popular vote.
3. Exit polls. I tell people that I have a very low opinion of exit polls. They are often poorly executed and are easy to misread. Today is yet another peice of evidence against exit polls.
4. Iowa electronic market on tuesday. There was a short panic when Kerry contracts sold for more than Bush contracts. I guess some traders panicked when they heard about the exit polls. Shame on them!
That's it for now. Now back to our normal blogging at Marginal Revolution...
Posted by Fabio Rojas on November 3, 2004 at 01:04 AM | Permalink
TrackBack
TrackBack URL for this entry:
http://www.typepad.com/services/trackback/6a00d8341c66b253ef00e5508347ed8833
Listed below are links to weblogs that reference Poll Positions III:
» Tooting My Own Horn from voluntaryXchange
I predicted Bush would win almost a year ago. I am not the only one to do this, or the only one to use economic data to make this prediction. But, my method was simple enough to do as an [Read More]
Tracked on Nov 3, 2004 12:54:56 PM
» Tradesports Accuracy from voluntaryXchange
Tradesports state by state prices on Monday afternoon made a perfect call of the electoral vote total. That was the data set I used to give Bush a 57% probability of winning in an earlier post. And, according to those [Read More]
Tracked on Nov 3, 2004 12:55:17 PM
» Predicting Elections from voluntaryXchange
"Either the exit polls are completely wrong or George Bush loses," FOX News analyst Susan Estrich said. Now we know ... pay attention to trading markets next time around. [Read More]
Tracked on Nov 3, 2004 12:59:31 PM
» Election Simulations from voluntaryXchange
So, I made two posts today about how to interpret prices in prediction markets. I would be remiss if I didn't do my own simulations. These only incorporated my earlier criticism - that prices are typically closer to 50 than [Read More]
Tracked on Nov 3, 2004 12:59:31 PM
