As a political junkie, elections are the Superbowls of the year for me. Literally. So, it should come as no surprise that I have been glued to the television and computer for most of the night crunching the numbers on the "Big 3" races that are supposed to magically tell the future of President Obama's administration, according to the media pundits. One year on from his historic victory and speech in Grant Park, the gubernatorial elections in Virginia and New Jersey and the congressional special election in New York's 23rd congressional district are the equivalent of the Oracle at Delphi - forecasting fortune or doom going into the midterm elections of 2010. Maybe. The media may be wrong on this one and their incessant reliance on the election horse race.

First, the exit polls. Voters in Virginia and New Jersey (two states out of 50) gave President Obama approval ratings over 50%. These races were not about Obama, but very local issues - transportation and the economy in Virginia, the economy/taxes and corruption in New Jersey. Therefore, the Oracle is not so much of a truth teller as much as a mirage. The media can stop looking at it.

Second, the pre-election polls. The media make a living off the horse race. Bringing in pollsters and pundits draws in politicos (ok, I'll admit it - me too) and ratings. RealClearPolitics' polling averages in the Virginia gubernatorial race had Republican candidate Bob McDonnell ahead by 13.4%. He won the race by 18%. In New Jersey, RCP polling averages had Republican candidate Chris Christie ahead by 1%. He won by 4%. And finally in NY-23, the final Public Policy Polling showed Conservative candidate Doug Hoffman with a 17 point lead. He went down to defeat by 4%. In all of these cases, the media latched on to sparse data regarding a handful of election contests. Additionally, media organizations sponsored exit polls in New Jersey and Virginia that were reported shortly before the polls closed. They made up for a lack of information and generalizability by paying for numbers to report on for several hours before the polls officially closed. These exit polls, a surprise for an off year election, are used to "test the waters" for future election projections, according to Georgetown professor Diana Owen.

Third, the circus. In the absence of more races to cover, the media descended and perpetuated a story of Republican chaos in New York's North Country. The media coverage then drew in an ideological army of "out of towners" who drowned a far right candidate in funds. This caused the Republican nominee to withdrawal, further perpetuating the chaos. Polling data was skewed as reported on Twitter tonight by Public Policy Polling (PPP). PPP said they learned a valuable lesson from all of this: if a candidate withdrawals, ditch the poll. Instead, they went ahead with a poll showing Hoffman ahead by 17 points. Therefore, the few legitimate methods of gauging these races were quickly rendered useless by the media coverage.

The media coverage of these races, it seems by preliminary examination, had little impact on the final result in Virginia, some impact in New Jersey, and a great deal of impact in the New York congressional election. New Jersey was covered more frequently because the polls were closer and President Obama made more frequent trips. The New York race was covered because of the bizarre nature of the contest. Once Virginia became a lock for the Republican ticket several weeks ago, media coverage dropped off the radar. The media aggregator at RealClearPolitics listed eight stories related to NY-23 on the Monday before the election and four "general" stories regarding the elections as the all important Obama Oracle. Obviously, media coverage on rural, upstate New York paralleled that of rural Iowa before its presidential caucuses.

Wednesday November 4, 2009 will be remembered as the day the campaign for the 2010 midterm elections began. On the heels of strong Republican victories in Virginia and New Jersey, a surprising Democratic victory in New York, and a narrow victory for Mayor Michael Bloomberg of New York City (the polls were wrong there too), the pundits will crunch the numbers about suburban voters in Fairfax County, VA who supported Obama last year and Republican McDonnell this year or Bergen County, NJ voters who voted for Obama last year and Republican Christie this year. I will keep in mind that generalizing these results is risky business (and not in the Tom Cruise sense).

The electorate that elected Obama last year was decidedly younger and more diverse than the older, whiter electorate this year in a handful of contests. Not to take the wind out of the Republican sails. They had a very good night and elections are about who shows up. But that is why you cannot read this year's results into next year's: elections are about who shows up. Republican efforts to reach the voters who did not show up are not the subject of this post. It is about the numbers and tonight they favored the Republicans. Next year is yet to be determined.