Data Digging III, Special Guest Edition...

I'm posting the following without comment of my own, but I'll probably have something to say about it later. A friend of mine was interested in gauging how much of an impact Bush Derangement Syndrome had on our little old county's elections -- specifically the race for County Executive -- and decided to take a look at the data. Below is how it was done and what was found.

-------------------------------------------------------------

"I don't think growth was an issue in this election. I think George Bush and Iraq were the issue." -Chris Merdon, The Baltimore Sun, November 12, 2006

Speculation in the local media and blogs suggests that the 2006 midterm elections were nothing more than a vehicle for voters to send a message of discontent to the national politicians and, specifically, Bush.

Is Ulman’s lead by 9,111 votes over Merdon attributable only to voter anger with the federal government and/or ignorance on local issues?

Looking at a Pew Research Center for the People & the Press poll conducted in the days leading up to the mid-term elections, we see that public opinion weighs heavily against Bush and the Iraq war. Poll results can be viewed here.

What if voters voiced their discontent through local elections on November 8? What if the votes were based only on dissatisfaction with national issues?

Three poll results were used to extrapolate possible county executive votes. They are: “If voters saw their vote for congress as a vote or against Bush,” “Opinion of how the military effort in Iraq is going” and “Approve or disapprove of the way George W. Bush is handling his job as president.”

Poll respondents were registered voters, grouped as registered Republicans/Republican leaning, registered Democrats/Democrat leaning, and undecided. The results were also tabulated to reflect the responses of all registered voters. For the purposes of this exercise, only respondents grouped into the first two groups were counted.

Calculations were run to determine probable county election outcomes based on three scenarios. The first scenario assumes that voters saw their local county executive vote as a vote for or against President Bush. The second scenario assumes that a vote for the county executive candidate followed sentiment regarding the status of the Iraq war. The third scenario assumes voters chose the county executive based solely on their approval of President Bush’s job as President.

In each scenario, it was assumed that public opinion in favor of Iraq or the President would equal a vote for the Republican candidate for county executive. Where public opinion was not in favor of the President or the war, a vote was tallied in favor of the Democratic candidate for executive. In instances where a material percentage of poll respondents indicated that Bush or the war were not factors or they were undecided, votes were split according to respondents’ party lines. Lastly, results from Hayduke’s Votes Stolen by Wallis calculation were used to remove votes from the both candidates’ calculated results and attribute them to Wallis.

The results if our county residents were truly trying to send a message to the federal government?



The calculation shows that if voters actually voted based solely on their dissatisfaction with national issues, Ulman would have won with a significantly larger margin.

In a county where registered Democrats outnumber registered Republicans by nearly 22,000 voters, is it any surprise that a Democratic candidate won? If anything, the margin of only 9,000 votes speak volumes to the fact that voters were willing to jump across party lines to vote for the candidate that they trusted with the next 4 years at the helm of Howard County’s government. That the margin of victory for Ulman was not wider shows that despite strong public sentiment against Bush and the Iraq War, voters were willing to vote Republican.

In every election, there are voters, both Republican and Democrat, who vote strictly according to party affiliation. However, in 2006, Ulman’s victory cannot be attributed solely to the anger and dissatisfaction of voters.