Bradley Effect Or McCain Effect?
Is Bradley Effect still there? How can we find out? If we check the last polls before the election, Obama was having double digit leads. However, his popular vote is 53% against 47%, 6% lead.
Here is a list of polls I gathered:
| Polling Company | Start Date | End Date | Obama | McCain | Total | Diff |
| Reuters/Zogby | 1-Nov | 3-Nov | 54 | 43 | 97 | 11 |
| Gallup National Polls | 31-Oct | 2-Nov | 55 | 44 | 99 | 11 |
| Fox News National Polls | 1-Nov | 2-Nov | 50 | 43 | 93 | 7 |
| CNN National Polls | 30-Oct | 1-Nov | 53 | 46 | 99 | 7 |
| Ipsos National Polls | 1-Nov | 1-Nov | 50 | 42 | 92 | 8 |
| ABC News National Polls | 30-Oct | 30-Oct | 54 | 41 | 95 | 13 |
| CBS News National Polls | 30-Oct | 30-Oct | 54 | 41 | 95 | 13 |
| Average | 52.86 | 42.86 | 96 | 10 |
CNN previously quoted polling experts that the Bradley Effect is worth about 6%.
Average of these last minute poll is about 10%. The actual margin is now 6%. Is the Bradley Effect 4%?
Bradley Effect is a voter claims to support an African American candidate and then votes differently. Did voters really change their minds this way?
The average of the polls for Obama is 53%. He gets 53% (52.86%). For this, the voters are consistent. However, the interesting part is this: the margin shrunk. What happened? All the “Other” category voters’ changed their minds. The Undecided voted for McCain. The Greens and Libertarians voted for McCain. So, instead of claiming for Obama and voting against Obama, they claimed something else and voted for the strongest candidate against Obama.
McCain gets 43% in the polls (average is the same as median and mode). However, the popular vote is 47%. So, if you are not an African America candidate and running against an African American candidate, you can count on getting extra votes on the election day, 4% of popular vote.
In the case of McCain, it is for every 100 votes committed to a pollster, he gets 9 extra votes.
For an African American candidate, he has to live by “You get what you see”. He probably cannot expect a measurable vote coming out from the Undecided category in the last days. This would be consistent with Obama’s strong push in the last few days of the campaign.
Can we now replace the Bradley Effect with this statement now: “if you are running against an African American candidate and you are not an African America candidate, you can count on getting extra votes coming from other third party candidates or previously uncommitted voters on the election day”.
If we can get more empirical results, then it is obvious that Democrats will have even more incentives to suffocate other third parties in the future. This is not about Nader anymore.
Instead of Bradley Effect, are we now moving into the era of McCain Effectt.
What Makes President Obama Not A Sure Thing
Are these undecided voters the key to this election?
As of today (2008.11.03), Obama has 228 electoral votes where the state has >10% polling margin and 50 electoral votes where the state has >5% polling margin. The difference less than 5% (either way) is 128 electoral votes.
Given the Bradley Effect, and the potential of undecided voters going for McCain, Obama may still get defeated. What are in play for this result?
1) Bradley Effect – voters misinform pollsters
2) Undecided voters for McCain – not willing to misinform, nor to be McCain voter out right
3) Polling methodology – too many people are not using land lines – greater and greater difficulty in forming a good sample to represent the population
4) First time voters – first time voters have lower voting participation rates; less responsive to GOTV. Given the huge ratio of first time voters this year, this factor is more important than before;
5) Voting machine – no more funny ballots. And are the votes auditable?
6) Voter suppression – prevent voters who fit your opponent’s voter profile to vote. Depriving sufficient resources or space to a few specific precinct or county maybe good enough, especially when coordinated with GOTV at other places.
7) Voting irregularities – fraudulent voter registrations, fraudulent voters. This list is not short.
These are all technical, minute issues. However, there are quite a few of factors for anyone to play with now, right? Any one of these can deliver us a dramatic event (or even traumatic) fit for Hollywood.
Typically, only the last 2 factors are in play. This time, we got seven factors!! Five of them have never been seen before. Too many unknowns. And there is no time to have an academic debate to discover the truth. So, there can only be offense for Obama since no number is completely good this time around.
Obama, 1 Up Over Karl Rove
Voting lines are long.problems/index.html. Not much news about voter intimidation. But then this is early voting, not the big day. Is America ready for the big news that is only 11 days away? Some people are preparing for the transitions to the next administration. Some people are already sensing potential presidential candidates for 2012: Hilary Clinton sensed the potential desire of Bloomberg for 2012.
Bloomberg got his way to run for the third term of NYC mayor. Less than two weeks ago, Clinton said Bloomberg’s third term is “disturbing”. Really, it would not have been disturbing if Clinton won the Democrat primary. It is disturbing because give the kind of difficulties the future administration will face, the next President may end up being a one term president. For that reason, Bloomberg’s presidential big would be a problem for her.
Given the turn out in the early voting, Democrats already see some success in their new and healthier operating model. Expanding the base is ideal. When it goes to the final stage, the operation is about turn out. Candidates locked up their supporters. How can they actually get them to the voting precinct?
Obama is experimenting something new: get them early and get them young. Voting is a habitual product, just like so many other products, such as carbonated drinks or auto insurance. We already saw how Obama got the young voters. We are now seeing how to get them early (early to the poll).
This is labor intensive. Given the labor cost in America is so high, this is an expensive operation. Expensive enough that Republicans never resorted to it before. Now, Obama does not only have a good certainty what votes they reached even though the ballots are not counted, but also what votes McCain got. Obama got a piece of insurance on these votes. These votes are of course the solid supporters. He can now have more flexibility to do his offense to McCain.
This is the kind of operation that shows to aspiring candidates that non profit groups are important. Or more to the point, candidates and non profit groups need each other. This kind of operation is labor intensive only people who got a commitment to a cause would be able to help.
Will he shift even more to the center? Will he attack McCain more aggressively? Or will he simply swamp him hardly in the battleground states, especially the ones that do not allow early voting?
Electioneering 102: A Lesson From Green Party of Canada
The federal election in Canada concluded on 2008.10.14. The incumbent party won another mandate. There are a lot of interesting content in this election to fill the media. However, for the purpose of Electioneering, this story may not make the cut to the paper, let alone to the headline. This case study lesson is experienced by the Leader of Green Party of Canada. This biggest contribution of this race is: how to choose a spot to run. The most applicable lesson of this is still city councilor.
Green Party of Canada has been increasing its vote share ever since the turn of the millennium. In 2006, they got 4.48% of the votes and no seats in the parliament. After that election, the Leader of the party then did not run in the leadership race again. It then became an open race and Elizabeth May won the leadership. She was an Executive Director of Sierra Club Canada, with government bureaucrat experience, a law degree and a recipient of Order of Canada. She ran a by-election earlier, against all parties and got the second largest votes in that race.
Green Party has long held the position that they are unable to win a seat because they are excluded from the televised debates. This election proves that a televised debate did not help.
May was looking for an epic race. This could have been the first sign of trouble. An election is not about making a statement. Election is about finding the most representative will of the people. By being the leader and possibly the first elected officer of a party, May’s election is more important than any other candidate. There is a lot of media attention, political resources, volunteer resources, and money involved in a leader’s race. Therefore, it is only prudent to maximum the vote, not to dramatize an epic.
If you are running in a city councilor race (or county board, school board, etc), you run to win. Any other objective is mischievous, and misleading the voters. It is true that people run elections for all kinds of reasons, some of them noble too. However, it is only when you aim to win would you be serving your voters, be honest to your voters, donors, volunteers and other kinds of supporters. Furthermore, if you do not run to win, your result tends to sink, even if you got some special interest groups’ backing because your true primary motives usually affect your plan, execution and result. For instance, since your real motive is a geography A, you divert more resources to that area instead. However, that area may be a contestable area. if the votes are already secured, no need to get vote there. However, you are doing it there to serve your personal interests. So, your volunteer hours are lost, your lawn signs are wasted. Alternatively, you may be interested at a specific donor group. Similarly you wasted your campaign.
Liberal party delivered their promise not to contest against May in order to maximize her chances to a seat, wherever her choosing. This promise is also unlikely to be offered in the next election since the leader of Liberal party is also in his trouble. And no one should plan a race with the expectation that this offer will be made twice anyway.
May placed this epic over at Central Nova, where the incumbent runs a dynasty there: 2 generations of incumbent, close to 10 elections. This is where the second problem is. One should contest in a place to win, not a place to dramatize. By being the leader, she can choose any leader she wanted. She should have picked a riding where
1) The Liberal candidate is not the incumbent, however strong enough that the votes actually matter in her race;
2) The Green votes are decent, say above the 2006’s record of 4.48% votes;
3) Incumbent votes are actually weak; below 50% is minimum requirement. The lower the better.
4) Since there are 3 major parties contesting in every riding in Canada, an ideal riding is where all three parties split their votes, i.e. around 30% each. Of course this is unrealistic.
Election is a contest of organization, stamina of the salesman (candidate), branding (party), money (fundraising) and product (platform). A leader got the luxury of choosing a riding, which most people cannot afford the infrastructure investment to do. Building up a local political network to support is not easy. However, this is fairly accomplishable if you were interested at a city race. Changing a house from this corner of the town to that corner is not too difficult. What is the ideal demographic for you? Ethnic group? Income class? Age group? Occupation? Family status?
If she spent 30 minutes to look around the >300 ridings in Canada, she maybe able to see that there are multiple ridings where Liberal is the incumbent, however with >25% of votes; Green votes are above 5%, and the incumbent got votes around mid 30%.
With 30 minutes of your time, you can see that Welland is one such riding where the incumbent is Conservative. Vancouver Kingsway is another one, where the incumbent is NDP. There are probably others. These 2 may not deliver an ideal environment for victory. However, the point is there are potential sites to choose from.
If May knew that it wasn’t going to be a victory, then dramatizing an epic is not a bad option. However, in your case, don’t run. An election is costly not only to you, but also to your supporters in forms other than money.
If you plan to remove a low performing incumbent, then get all the prospective candidates together. Gamble all resources in only one person. So, all your prospective candidates may want to have a quasi primary to determine who will have the best shot.
What Is The Price Per Vote To Pass The Bailout Bill?
CNN says more banks are likely to fall. It seems like that 700B will get plenty of asset to buy. Now, how much did the additional votes cost to get the bill?
Taxpayers For Common Sense compiles a list of 15 pet projects within the bailout bill (). It would be difficult to determine which provision actually got tucked into the bill to exchange for votes. However, we can see where the most direct beneficiary of the provisions and see the incentives of the votes, especially the switch votes.
Of the 15 listed there, 10 of them are costed out. They total price tag is $26,396 M.
There were 60 switch votes: 58 No-to-Yes, 1 Abstain-to-Yes, 1 No-to-Yes.
If you cross out the direct beneficiary of the provisions listed in the website and plus 2 guesses I have for the research provision (301) and racetrack provision (317), then you get the table below:
|
Provision |
Cost |
Votes |
Cost/Vote |
States |
|
601 |
3300 |
1 |
3300 |
OR, ID |
|
301 |
19000 |
9 |
2111 |
TX, IL, WA |
|
201 Sales Tax Deduction |
3300 |
7 |
471 |
TX, FL, WA, WY, NV |
|
325 |
148 |
1 |
148 |
NY |
|
502 |
478 |
7 |
68 |
CA |
|
317 |
100 |
3 |
33 |
MI |
|
211 |
10 |
1 |
10 |
OR |
|
503 |
2 |
1 |
2 |
OR |
|
602 |
9 |
7 |
1 |
OH, PA, KY, VA |
|
504 |
49 |
0 |
|
AK |
This is a total of $26.4B for 30 votes (some of these votes above are the same votes). This is a total of $890M/vote. These 30 votes cover 51% of all swtiched votes.
The famous archery provision is provision 503. That provision, jointly with 211, gave the bailout bill 1 vote. Together, those 2 provisions cost $12M. The interesting thing is that the House Representative who district covers the biggest beneficiary company Rose City Archery, which is located Mytle Point of Oregon, voted against the bill twice. The original House sponsor is actually Representative Kind from WI (Yes twice). By far, the most cost effective provision is 602 (Transfer to abandoned mine reclamation fund). $9M for 7 votes, $1.3M per vote.
The most expensive provision is 301, $19B. I only identified TX, IL, WA. However, I am sure there are more states which would be benefited because of it. This one so far got 9 votes (and very likely more).
The most cost ineffective is 601, $3.3B for 1 vote. It benefits OR and ID. However, ID completely voted against the bill. I guess they did not exactly want the money.






