So far, Scott Brown, the Republican, looks likely to win in a landslide. A big landslide.
Elected Democrats are predictably freaking out. Well that they should. This race may be a bellwether for the next ten years.
If the establishment Democrat, Martha Coakley, loses big to a Republican in Massachusetts, then no Democrat will feel that their seat is “safe” anywhere. If a Democrat can lose “Ted Kennedy's seat,” then all bets are off on every Democrat “safe” seat.
If Coakley loses, and loses big, watch for a rebellion among elected Democrats who will realize
that they could lose their cushy sinecures by being associated too closely with Nancy Pelosi's and Harry Reid's flavor of politics.
The polls are all over the place. There are many reasons for this. The whole polling industry is struggling with some huge changes that they must come to grips with.
The primary problem is cellphones and cellphone only households.
If you would like a first hand anecdotal feel for how big of a problem cellphones present to the polling industry, ask the following question in the next large group that you are in. “How many of you no longer have a land-line phone and only have cellphones in your household?” If your group is made up largely of people under forty years of age, odds are that half to over half of them will be cellphone only households. This is very significant. It is a game changer in the polling business. It is also a number that is very likely to increase over the next ten years. In twenty years, land-line phones may be
as anachronistic as Telegraph machines.
There are rules for calling cellphone sample that make it very expensive to work with. Mainly, you cannot use any kind of automated dialing method. If you are knowingly dialing on cellphone sample, you must hand dial the phone numbers.
Robo polls will miss cellphone only respondents. The results of a robo poll are ever more doubtful due to the fact that the robot dialed poll cannot account for the opinions of respondents that live in cellphone only households. (Texting may be a way around this problem. So far as I know, there are no rules against using automated systems to send out text messages to cellphones.[Where there is a will, there is a way - especially if there is money involved.])
When it comes down to the wire in regards to elections, I find that I am much more trustful of the punters then I am of the pollsters.
The punters ask a different question. They do not ask “Who will you vote for?” they ask “Who are you willing to bet good money, your money, on to win?” Its a question of knowledge, not of opinion. Which makes its a very different question. The results can be significantly different and potentially far more accurate.
Predicting elections was once something that the odds-makers dominated. Scientific telephone polling changed the game and allowed telephone pollsters to take that role away from the bookies. Now the game may be changing again, giving the advantage in predicting election outcomes back to the gamblers.
Take a look at Intrade and watch the election numbers as they come in Tuesday. Election Markets may be the way of things to come. Time will tell.
And these are very exciting and interesting times indeed.