Is it possible for John Kerry to win Ohio?

Dirty Harry says it ain't, or, if not impossible, it's bloody well unlikely even if several things broke John Kerry's way with both provisional ballots and the undercount (punched ballots that, for whatever reason, did not register in the machine recount).

Let's look.

Give this to Harry -- he takes the provisional totals from this blog that's sympathetic to the efforts to explain and reform the irregularities and suppression incidents of November 2. I'm not sure where Ohio Vote Suppression News -- which has done a terrific job compiling county by county news and information -- gets their totals, but as of Saturday, with 65 of 88 counties reporting -- but several larger, urban, Democratic counties not having reported -- we have the following totals:

Bush +36,050
Kerry +32,490

Bush began with a margin of about 130,000 votes, so figure now that he has a lead of 133,500.

The current percentage of provisional ballots which have been accepted by the authorities has been 77%. Note, though, that, since Secretary of State Blackwell did not institute any across-the-state standards for acceptance of provisional ballots, you have a county like Lorain County where only 32% of provisional ballots were accepted. (In Lorain, 702 of the 1,264 provisional ballots rejected were rejected for "improper registration" reasons -- while in nearly all other counties, "the largest percentage of rejected provisionals is due to ballots cast in the wrong precinct. Here, however, these amount to 160 and are apparently not part of the 702." FWIW.) Also Cuyahoga County, home of Cleveland, has been accepting about 66% of their provisional ballots.

How many provisional ballots, total, were issued in Ohio? According to this website from Blackwell's office, 135,149. But according to this article from the AP, and several other ones I found on Google, there were 155,337 provisional ballots. Given that the rest of the Blackwell site has not been updated -- the results listed here from election night still only reflect 8% of precincts! -- let's take the number 155,337 as the total.

If we take 77% of 155,337, to represent the number of the provisional ballots that were accepted, that gives us 119,609 total accepted provisional ballots. When we subtract the provisional ballots already counted (68,540), that gives us 51,069 ballots yet to be counted.

Let's say that Kerry wins a huge majority of those 51,069 ballots. Let's say he wins 75% of them. (Unlikely, but not impossible, given that many of those outstanding ballots come from urban areas.) Kerry wins 38,302, Bush wins 12,767.

That would then make the provisional ballot results total:

48,817 Bush
70,792 Kerry

Which would give Kerry a margin of 21,975 in the provisional ballot total.

Let's subtract that from Bush's previous margin of 133,500 in the overall vote. That would leave Bush still ahead of Kerry by 111,525 votes after the provisional ballots have been factored into the mix.

But what about the undercount, involving those ballots that do not register a vote for a particular race, often due to machines being unable to read the punch on a punchcard ballot? (Two Dayton precincts, for example, had the strange occurrence of over 27% of their voters not voting for a Presidential candidate; surely there are some voters who don't vote for any Presidential candidate, but 27%?)

If the recount that Badnarik and Cobb have filed for goes through, as I think it will, will a hand recount reveal an undercount that can drag Kerry across the finish line?

The Cleveland Plain-Dealer has reported that there were 92,000 spoiled ballots. Let's say that two thirds of those spoiled ballots go for Kerry, with a third going for Bush. (Surely, a great number of those spoiled ballots will show preference for neither candidate, but let's just play with this for a second.) That gives Kerry 61,333 votes, and Bush 30,667 votes from the undercount, which gives Kerry a margin of about 30,666 within the undercount vote.

That would bring Bush's margin of victory in Ohio to about 80,859 votes. And that's with generously assuming that 75% of the remaining provisional ballots break for Kerry, and generously assuming that Kerry wins two-thirds of the 92,000 spoiled ballots.

Let's push plausibility even further. Let's say that Kerry wins all of the spoiled ballots. Bush would still have a margin of over 20,000 votes.

(Yes, I said in a previous post that I thought Kerry would chop the margin down to 30,000. I was pulling a number out of an orifice whose statistics should not be trusted.)

What does this mean?

It means that if the Plain Dealer's estimate of 92,000 spoiled ballots is accurate, the only way that Ohio is going to end up in John Kerry's column is if a hand recount reveals that there were significant other glitches or irregularities -- accidental or intentional -- in the already tabulated totals that create a shift of 40,000 votes from Bush to Kerry, or where 80,000 votes for Kerry somehow were not tabulated, or where 80,000 votes for Bush that were tabulated did not exist.

How you look at this depends on how much confidence you look upon the systems used to tabulate the votes, and how much confidence you have that a situation like Warren County's lockdown resulted in no funny business at all.

We know that the tabulation system can be hacked, and easily; but did someone actually hack it?

We know that it's strange that Franklin County's glitch on one machine added 4,000 votes to Bush's totals, but do we think that there's a possibility that if one machine had that glitch, there could be other machines that could have done the same thing in the state? (Many have theorized that anyone wanting to commit effective electoral fraud would not focus on individual machines, but the systems that tabulate the vote totals from hundreds of those machines.) All it would take is twenty other systems to have the identical glitch as that Franklin County one for Bush to lose that 80,000 vote margin.

Obviously, those of us pinning any hopes to John Kerry somehow staging the ultimate comeback in December of 2004 have to come to grips with the sheer unlikelihood of that happening. The Red Sox winning the World Series probably filled the quota for the amazing miracles of 2004.

But just as obvious is the fact that should the recount not occur, there were too many strange glitches and irregularities that will go unexplained, and will severly mar the confidence with which we approach the computer systems and technologies with which we conduct the basic execution of democracy.

Have there been enough accidental glitches across the country, consistently favoring George W. Bush, to give one skeptical pause regarding the accuracy of the totals? I believe there have.

Has the Republican Party earned the benefit of the doubt in an election year where they back-channelled money to pay for an advertising campaign that lied in painting a war hero as a war criminal...

where they took RNC dollars to pay for leaflets announcing that the Democrats intended on banning the bible...

where after complaining about activist judges and touting tort reform they sued to have the right to place partisan poll watchers in urban polling places that already had pollwatchers present...

where the wife of the Vice President publicly described the Democratic presidential candidate as "not a good man"...

where the blatant disregard for church and state separation resulted in ignoring churches' tax free status by using churches as political outreach centers far, far, far outdid any outreach the Democrats ever made with black churches in the city...

where within a few months we watched as the rationale for a war shifted from allegations of involvement in 9/11 to funding terrorist activities to producing weapons of mass destruction to controlling the energy supply to simply taking a dictator down...

-- Has the Republican party earned the benefit of the doubt that without a recount in Ohio, given the irregularities present, we can be confident that everything was copasetic, that, yes, nothing was rotten in the state of Denmark?

I believe we cannot.

And that's why we the Ohio recount should go forward.

I ask my Republican friends to support a recount, so as to let George W. Bush serve his second term with no Roger Maris asterisk like that which hovered over his head during his firs termt.

But more importantly, I ask them to support a recount for the love of this country, and the recognition of how fractured it remains. When 20% of the voting population things Bush was not legitimately elected President this time around, let's have a recount to assauge and affirm that nothing went wrong in Ohio -- y'know, beyond the ridiculously long lines in urban polling centers and Republican pollwatchers hovering within the ballotplaces like a relic of Jim Crow.

Let us have the recount so that people can feel sure and confident that our democracy works.

That's what's happened so far in New Hampshire; they've been having the recount, and it's been dispelling worries and affirming the effectiveness of the Diebold systems.

We've watched in the Ukraine this week as people took to the streets in the name of assuring the sanctity of their democracy.

Here in our country, we do not need to take to the streets. All we have to do is support the right of Michael Badnarik and David Cobb's campaigns to register and pay the fees, as dictated by Ohio law, to see a recount take place and earn the assurance that every vote was counted in that state.

Our democracy will only be stronger for it. I certainly support putting up a million dollars for that, given how freely this administration has spent billions in far, far worse pursuits. In a year where the two political campaigns spent half a billion dollars or more, a million dollars is not a huge price to assure Americans the certainty and sanctity of the franchise.