(cross-posted at WMR, BFM, and ML-pb)
A few weeks ago I examined the underlying political conditions of the 110 Michigan State House districts, which can be found [http://www.michiganliberal.com/diary/16421/state-of-the-house-part-i here]. This second part performs the same analysis on the State Senate. As stated previously, the forthcoming third and fourth parts will examine the candidates who have filed to run for seats in the State House and State Senate. As mentioned previously, I maintain a listing of State House and State Senate candidates that includes the financial filing statements that can be obtained via a subscription by emailing email@example.com.
While in previous posts I had made some guesses as to which seats are the most vulnerable for a potential takeover by the opposing party, I want to quantify this estimation. Using a variation of the reputable House Vulnerability Index developed by Crisitunity at the Swing State Project, I have created my own District Vulnerability Index (DVI) which is shown in Table 1 below.
Some of the columns are pretty self-explanatory. Column 1 provides the senate district number, 2 provides the current senator's name, column 3 the senator's party, and column 4 the senator's current term. Columns 5-7 provide the Democratic baseline number from the 2004, 2006, and 2008 elections, with column 8 the three cycle average for the district. I tested this model out using the Presidential/Gubernatorial numbers from 2004 to 2008 instead of the baseline numbers, but went with the baseline numbers given that the higher ticket numbers distorts the DVI by giving a Democratic percentage that is much higher on the top of the ticket than for lower level races (like the state house and state senate). Column 9 is the Partisan Voting Index, which is calculated by subtracting the statewide partisan average of 50% from the district's average Democratic baseline found in column 8. Column 12 provides my previous classification of the race in February 2010, and is included here to test its validity while using the DVI. Column 10 ranks the PVI for each district from 1 to 110, with 1 being most Republican and 110 being the most Democratic.
Column 11 is the perhaps the most important column in the entire analysis. As stated in previous analysis, incumbency is a wonderful advantage, as it provides funds, name recognition, and built in party support. In Michigan, incumbent candidates have a win percentage of nearly 95% every cycle in both the State House and State Senate since term limits were established in 1992, and despite the electorate being anti-incumbent in the 2010 cycle, the voters hate every incumbent expect their own. Column 11 is the Democratic candidate's margin of victory from the 2008 cycle, and open seats (whether from term-limits or from candidates leaving the State House to pursue other opportunities) have no margin given. Columns 14 to 17 show the data used to determine column 11, providing the votes for the Republican candidate (Column 14) and the Democratic candidate (Column 15), the Democratic candidate's two-party vote percentage (Column 16) and the Republican candidate's two-party vote percentage (Column 17).
The DVI is shown in Column 13. The DVI is determined by multiplying Column 11 by 50 and adding the PVI rank in Column 10. Districts with no incumbent thus are at a greater risk for being taken over by the opposing party. Table 1 ranks the districts from the smallest to largest DVI.
Table 1: State Senate DVI
Table 1 largely confirms that there are a number of solid Republican and Democratic seats that the opposing party will not attempt to take over in 2010. I'm guessing that 12 Democratic districts and 14 Republican districts will face minimal opposition. While the nine incumbents are all likely to return in in 2011, I was surprised how vulnerable Roger Kahn's district is, despite his incumbency advantage. Indeed, the 32nd District has as solid as a Democratic base as Glenn Anderson's 6th State Senate district, a seat that no Republican will seriously consider as a possible flip. Likewise, the usual suspects of open swing seats appear- the 7th District in western Wayne County, the 20th District (Kalamazoo County), the 29th (Grand Rapids and Kentwood), the 34th (Muskegon County) and the 36th District (northeast lower peninsula). However, Democratic candidates for the 36th (Joel Sheltrown) and 37th (Gary McDowell) are considering running for the open 1st Congressional District seat. Losing either of these candidates would make flipping these districts a much steeper challenge for the Senate Democrats in November.
One question that I haven't be able to answer yet is the role of money in winning an open seat. Generally, the candidate who spends the most (including resources from the state party and PAC money) wins. However, this does not include candidates who self-finance, who have a harder time convincing voters that they are not simply buying the seat. With the sheer number of open seats in the State Senate, both parties need to make sacrifices on where to send their dollars. This might hinder a candidate like Republican Tom Casperson in the 38th District, who will likely need ample funding from the Michigan Republican Party to compete in a traditionally Democratic district. Likewise, a candidate such as Democrat Mary Valentine, who is struggling to raise money in a competitive open seat with an increasingly Democratic lean, might require much more in state party resources, forcing the Michigan Democratic Party to reduce resources to more "long-shot" districts. Be sure to keep your eyes peeled for the forthcoming Parts 3 and 4 that examine the candidates running for the State House and State Senate seats.