Hillary's Predictions Revisited

If you haven't checked out Nate Silver's polling analysis site at FiveThirtyEight.com, I would highly recommend that you do so.  Silver, who has achieved a minor celebrity because of the thoroughness and originality of his work, essentially does a meta-analysis of every major poll done in the United States on the presidential and senatorial races, corrects sampling and interpretation errors, and runs simulated elections based on the results.  Under his FAQ section, he explains how the simulations are done to produce the "win" percentages:

“What is Win % or Win Probability? Simply, the number of times that a candidate wins a given state, or wins the general election, based on 10,000 daily simulation runs.

“How is Win Probability determined? By simulating the election 10,000 times each day by means of a Monte Carlo analysis, based on the current Projection in each state. The simulation accounts for the following properties:

(i) That the true margin of error of a poll is much higher than the sampling error, especially when the poll is taken long before the election.

(ii) That polling movement between different states tends to be correlated based on the demographics in those states.”

Using this approach, Silver estimates that Obama has about a 97% chance of winning the general election next week.  A prediction for every state individually is also given; for example, Obama has a 100% chance of winning New York and Massachusetts; he has a zero percent chance of winning Utah or Texas.  West Virginia was thought to be “in play” at one time, but not really, it turns out.  Obama has a 2% chance there.

So, to a certain extent, Hillary Clinton was right when she said that “hard working Americans, white Americans” preferred her.  They do indeed.  I imagine that West Virginia would poll very differently if Hillary had been the Democratic candidate.  On the other hand, FiveThirtyEight (named for the total number of electors in the American system) estimates that Obama’s current chances of carrying Ohio are about 84%; that was another state which Hillary won handily.  States such as Pennsylvania, another state carried by Clinton, will probably vote for Obama.  Silver estimates his chances at 99% currently.  538 also gives Barack the nod in Florida, to a 79% certainty.  The True Blue Bastions, such as New York, New Jersey, Massachusetts and California, of course, were always going to vote for the Democrat, so Hillary’s protests about being the more electable candidate did not make much sense with reference to these mortal locks.

 The most interesting states will be North Carolina and Virginia, come election night.  Because North Carolina has a Democratic governor (Mike Easley), and an ardent Obama supporter at that, it will be harder for the Republicans to cheat in the Tar Heel state, and cheating is the only practical way for the Republicans to steal this election.  They must cheat often, early and massively in order to compete at all, when you get right down to it.  Of the preferred sites for Republican cheating, Ohio and Florida, it has become more difficult in the Buckeye State because of a change in regimes at the Secretary of State level.  So the GOP effort will be aimed primarily at Florida, where voter ID laws of highly questionable constitutionality (picture ID with signature required, reminiscent of old poll taxes and “literacy” tests used to disenfranchise African-Americans in the Old South) are the latest Republican gambit for defeating the will of the majority.

 So 5,000 lawyers are flying into Florida just to keep an eye on things, among them your faithful correspondent.  It won’t stop all the cheating but it may help to keep things vaguely within the lines, a reasonable goal in the World’s Greatest Democracy, after all.