June 22, 2012
I just got a call from AP saying that they are calling the race for Big Tobacco.
I was a little surprised that they did this at this point in time. There are wide differences in the responses by county, which makes predictions of the outcome complicated. I have developed a statistical model that predicts the votes since the election from those on election night. (The post-election counts are systematically more for Prop 29 than the election day returns.) The last projection I did (based on returns as of June 20) showed Prop 29 losing by .31%, with 49.85% yes vs 50.15% no) or 15,861 out of 5,129,712 votes.
This difference was, however, well within the statistical margin of error.
Here is what I told AP by way of reaction:
1. Both the Secretary of State and the health groups should carefully consider whether or not a recount is in order. (It the Secretary of State orders it, the state pays; if the health groups request it, they have to pay.) This is a complicated question, but given how close it is, the cost of a recount might be worth it. (If, at the last minute, Philip Morris and Reynolds lose, you can be sure they will demand a recount.)
June 17, 2012
June 11, 2012
At 4:43 am June 7, the morning after the election, the Secretary of State showed Philip Morris and Reynolds leading on Prop 29, with 50.7% no vs. 49.2% yes on Prop 29, a spread of only 63,176 votes, with about 1 million uncounted mail-in votes.
As of 5:55 pm on June 11, the Secretary of State showed that, with about half the votes counted, the cigarette companies' lead had narrowed to 37,096 votes, with 50.4% no to 49.6% yes.
I did a county-by-county analysis, using the election day returns to predict post-election returns (using linear regression), then estimated the number of yes votes remaining based on the number of uncounted votes by county. The result shows a very narrow loss of 29, but the difference is well below the "margin of error" of the statistical estimates.
It's still too close to call.
June 8, 2012
I have been giving the outcome of the Prop 29 election a lot of thought and talking to others involved in and watching the campaign.
There is a broad consensus that the single largest boost that Philip Morris and Reynolds American got for their campaign against Prop 29 was the fact that the LA Times urged a "no" vote.
The cigarette companies heavily promoted the Times' position in their $47 million ad blitz, especially in Southern California and it worked. Everyone expected Prop 29 to do worse in LA than Northern California because Southern California is more conservative than Northern California, but there was an unprecedented difference: The election day returns showed Prop 29 winning by 73%-27% in San Francisco County but narrowly losing in LA. The only thing that can explain such a huge split was the LA Times endorsement of the tobacco industry's position.
While the Times tried to distance itself from Big Tobacco, they parroted its argument: