Bad for Rossi, Good for Gregoire

The hand recount in Washington State for the gubernatorial election of November 2, is underway, and is showing anomalies between the machine recount and the present check. As expected, the Libertarian candidate has been losing votes. Although the Republican Dino Rossi has (on paper) increased his lead to 43 over Democrat candidate Christine Gregoire, the signs are the King County will find sufficent extra votes to overturn the result. Some Republican activists have actually called on Rossi to concede in protest.

With 10 counties reporting, the discrepancies stand at: Democrat +28 votes, Republican +29 votes, Libertarian -2 votes. King County, which includes the city of Seattle, accounted for over 875,000 votes or over 30% of all votes cast. On current trends a discrepancy of over 800 would be consistent with 500 going to the Democrat candidate, sufficent to win the race for Governor. In the machine recount, King County found 971 extra votes of which 593 went to Gregoire.

Recounts in the US have a habit of boosting the Democrat score, whether in Washington, Ohio or Florida. Depending on one's political allegiance this is either cheating by Republicans who prevent good votes from counting, which is rectified in the recount, or cheating by Democrats who stuff the ballot boxes after the polls have closed. Either way, a bad taste is left by these lapses.


Was it the Guardian wot won it for Bush?

I just had a discussion about Operation Clark County, that fatuous attempt by the London-based Socialist-leaning Guardian newspaper, to persuade Ohio voters to back John Kerry in the November 2 election. The question being considered was: did the Guardian (nicknamed the Grauniad for its frequent typographical errors and its leftist obsessions) win the presidential election (accidentally) for George Bush? In 1992, the Sun newspaper proclaimed on its front page after the fourth consecutive defeat for the British Labour Party that "It Was the Sun Wot Won It". Did the Guardian's patronizing meddling produce a similar, if unintended effect?

The set-up was as follows. The Guardian calculated (correctly) that Ohio would be a key state in the presidential election. Ian Katz and Oliver Burkeman, the 'geniuses' behind the campaign selected Clark County on the basis that Al Gore had supposedly won that county by a mere 324 votes or 1%. The Guardian also claimed that George W Bush had won Ohio by 4%, but that polls were showing a tight race with a virtual dead heat expected.

The truth was slightly different. Clark County did go for Gore, but only by 0.56%, and Bush won Ohio by 3.51% in 2000. So the background information was slightly wrong to begin with.

Robin Grant of perfect.co.uk, went one better, not only claiming to have set up the campaign but actually collecting local press coverage, including expressions of outrage from Clark County residents. Grant's smug chuckles at the right-wing bloggers look funny now, but not the way that he had intended.

The result of thousands of Guardian readers sending letters to independent voters in Clark County was nothing less than dramatic. The campaign, despite being a "roaring success" was cancelled as soon as letters began to arrive in Ohio.

A typical report of the time reads:
The Springfield News-Sun also received about a dozen e-mails, starting early in the day, about the Guardian campaign, from places as diverse as New York, New Jersey, Georgia, Alaska and Switzerland, almost all of which expressed some degree of outrage.

The local paper ran a story with the headline "Butt out Brits, voters say". USA Today describing the Guardian as a "left-leaning newspaper", poured scorn on the campaign, portraying it as at best a publicity stunt.

Both local Republican and Democrat campaigns described the move as self-defeating. The scorecard shows the cost of Operation Clark County to John Kerry's hopes of winning Ohio.

Clark County20042000


Republican majority or deficit (%)
Ohio+2.12+3.510.7% to Democrat
Clark County+0.56-2.041.3% to Republican

So the Guardian swing was 2.0% from the Democrats to Bush. Clark County was the only Ohio county to switch from a Democrat majority in 2000 to a Republican one in 2004. In fact such a swing only occurred in 5% of the whole country's 3,113 counties according to USA Today. If we imagine what effect Operation Clark County might have had if the Guardian had run it across the whole of the USA we get some startling results.

Kerry would have failed to win the only State to switch his way from 2000, New Hampshire with 4 electoral college votes. Worse still, Michigan (17 votes), Minnesota (10), Oregon (7), Pennsylvania (21) and Wisconsin (10) would all have fallen to the Republican onslaught if the Guardian effect had been felt in those states. The final outcome would have been a 355 to 183 electoral college massacre for Senator Kerry.

It gets worse.

The Guardian ended up giving out contact details for 14,000 voters out of the 56,000 originally planned for. If these had been sent and they had achieved the same scale of effect then the swing would have been eight per cent more to Bush. So a nationwide Guardian campaign could have given a further eight states totalling 130 electoral college votes for Bush. At 485 to 53, the Democratic Party can only pray that next time round no progressive billionaire pays for such an operation.

The short answer to the original question is 'No', Operation Clark County cannot be credited with finding the 118,775 votes for George Bush that got him re-elected. In the event, the entire voting population of Clark County going for John Kerry would not have been enough. But the margin of victory, which encouraged Senator Kerry to concede gracefully on November 3, might not have been there without the efforts of Guardian readers. I understand that the local Republican campaign office wrote a letter of thanks to the Guardian.

One final question. Did the Guardian pay for the copy of the electoral roll, or do the taxpayers of Clark County have to find an extra $25? I hope Mr Katz pays up. It's the least he can do for the 'impoverished' masses of Ohio.



A federal judge has ruled that a full recount in the state of Ohio should go ahead, following lobbying by David Cobb, the Green Party candidate and Michael Badnarik, the Libertarian Party for US President.

To be honest the request makes no sense whatsoever for either candidate.

David Cobb wasn't even on the ballot and received 186 write-in votes, 11,721 less than the next placed candidate (Michael Peroutka, the Constitution Party candidate). He only got 72 votes more than Joe Schriner, a local independent. So for the Green candidate the only outcome that could change is that instead of finishing 5th, he could finish 6th.

For Michael Badnarik there is even less sense. As the Libertarian Party candidate, he might be expected to turn down a chance to waste taxpayers' money. Instead he demands a recount. As the LP candidate scored 14,695 votes in Ohio, or 2,725,357 fewer than John Kerry (Democrat), it does not seem plausible that Badnarik thinks he could finish second. On the contrary, the LP only finished 2,717 votes ahead of his nearest rival (Peroutka). So again there is only one plausible change, Badnarik could finish 4th instead of 3rd.

If there was any chance the result might give the minor parties automatic registration on the ballot next time, I could approve of the attempt.

Of course the real reason for the sudden enthusiasm of Libertarians and Greens for a full recount has more to do with the perception that if a recount lead to a Kerry win in Ohio, which overturned the presidential election, then both the Green and Libertarian parties could proclaim that they had achieved a massive impact.

However, unlike Florida in 2000, the gap between the Republican and Democrat candidates is greater than the total number of minor party votes combined (119,775). At 2.12%, the lead President Bush enjoyed over Senator Kerry was more than 230 times greater than that of Florida in 2000.

UPDATE: I found this news release from the Libertarian Party. It claims that the purpose of the recount is to "expose irregularities" in the election process.

More revealing of Michael Badnarik's agenda is the following quote:
In addition to other perceived fraud -- or at least irregularities -- in Ohio's presidential election, Badnarik pointed to the "disturbing" fact that exit polls in many states showed Kerry winning the race.

"The phenomenon that is most damaging, in my point of view, is that for 20 years now, the media have been doing exit polls and using that information to project who was going to win that election," he said. "And because people leaving the polling places are typically very candid, the exit polls have historically been within 1 percent of the actual vote totals.

"Now, all of a sudden in this election, the exit polls and the vote totals differ by a significant amount -- and the talking heads on television are looking at it and wondering why the exit polls were wrong in this election. From what I can see, there's no reason to believe the exit polls were wrong, and fairly good reasons to believe that it was the election process that was faulty."

So there we have it. Badnarik wants the John Kerry to win the presidency. And he is willing to waste $1.5 million of Ohio's taxpayers' money to try to achieve this aim. No amount of spin can dress this up as a vehicle for advancing Libertarian ideas.

A little overtime before Christmas

Latest from the Washington State governor election saga.

The hand recount order has been signed and sealed. The political parties are now trying to find volunteers to hang around for up to two weeks staring at piles of paper to a) deter the other side from voiding any of our votes and b) trying to spot void votes allocated to the other side. Details here.

Having done this sort of thing myself I can honestly say that I defy anyone to keep their mind on it for more than two hours, never mind ten days, knowing that the gap of 42 votes is 0.0015% of the total cast. Normally a scrutineer can comfort him or herself with the thought that a single error probably won't make a difference. In my experience we would watch the votes being counted for a couple of minutes, concentrating on small batches to watch for any discrepencies (votes for our candidate in the wrong pile, a stack that should have 50 votes with 51 instead). Added to the fun we can't touch the ballot papers in the UK, but have to attract the attention of the returning officer or a deputy to come over and check our query.

Having worked in banks with note counting machines, and hand counted £35,000 in small bills once, I would be amazed if a hand count failed to reveal more than 42 errors out of 2.8 million. Counting machines are built for speed. Ideally I would use them to double check a hand count. At a guess, the recount and the machines will have done a great job if fewer than 250 votes change hands across the state. Of course many of these errors would cancel each other out: there's no reason to suppose that Rossi votes stick together more than Gregoire's do.

Because the Libertarians may not have a full complement of volunteers for this hand recount, expect some of their votes to be challenged...


Cajun musical chairs

Results in Louisiana for the run-offs for the US House of Representatives, have left the Republicans winning a Democrat seat and the Democrats winning a Republican one. [Thanks to RealClearPolitics for the link.]

The elections are unique in the USA because Louisiana has the French electoral system of having no primaries, so all potential candidates stand in a first round of voting on general election day. If no candidate passes 50% of the vote, the top two candidates go to a run-off election the following month. This is the same election system used in successfully France and recently seen in the Ukraine (in less favourable circumstances).

With the final results for the US House of Representatives now in the score is:

2004Republican 232Democrat 202Independent 1
2002Republican 227Democrat 205Independent 1
ChangeRepublican +5Democrat -3Independent n/c

N.B. There were two vacant seats going into the 2004 election.

So the Republican Party has made its five gains after all, confirming that 2004 was a disaster for the Democratic Party across the board nationally. If the Washington State governor election re-count confirms a Republican gain there, then the Democrats will end 2004 with fewer presidential electoral college votes than 2000, fewer Senators and House Representatives than 2002 (already a bad year), and fewer Governors than in 2003 (which included the California recall fiasco for the Democrats).

And they probably outspent the Republicans this time, if we include outside supporters. Ouch!