Mapping the expressed ideology of the Senate candidates is interesting on its own - but we can also use this information to start thinking about the upcoming race. One of the easiest concepts to grasp is the concept of the median voter theorem. Now about 60 years old, the median voter theorem suggests that when preferences are single peaked (effectively analyzing one's choices along a single dimension), the candidate who appeals to the middle voter will win an election. This is one of the reasons why the two-major parties tend to move toward the center in a general election, as both parties are focused on the median voter. This also gives rise to what political scientists call the primary paradox (those things one does to win a primary often lead to negative outcomes in the general election); during a primary and a general election, there are two different median voters to consider.
What this means is that in the general election, the candidate closest to Florida's median voter will likely win the Senate seat. So what is the ideology of Florida's median voter? Well one idea would be to simply look at partisan identification: in Florida, there are 4.7 million Democrats, almost 4.1 million Republicans, and about 2.4 "other" voters. Some of those "others" are to the likely to the left of most Democrats (e.g., the Greens and the Socialists), some are likely to the right of most Republicans (e.g., the Objectivists and the Constitutionists), but the vast majority of those "others" are not registered with a party (or registered with the Independent Party of Florida). Those individuals could be distributed randomly, or exactly in the ideological center. This method of adducing Florida's median voter is not necessarily fruitful.
Another method of determining Florida's median voter would be to look at the ideology of Florida's elected officials. Using DW-NOMINATE scores, this would make Vern Buchanan's .447 as the 13th out of the 25 House of Representative members. Adding in the two Senators, this would make Senator Mel Martinez's .412 as the exact center of the Florida delegation. This does not mean that the median Florida voter would be comfortably placed in the same ideological space as Martinez - but does mean that Martinez's ideological estimate is likely closer to the median Florida voter than was Martinez's opponent in 2004 (Betty Castor). Castor's ideological score is not easily discernible, as she has not served in a legislative position (Fl State Senate) since 1986. I have been unable to find interest group scores for her from that time period. The most promising option is to look at the median Republican House member and the median Democratic House member - and finding the midpoint. That would mean that Florida's median voter would be about .042. That would be the "most liberal" interpretation of the median voter (the "true" median could be closer to the Republican median). One could also consider the 2006 Bill Nelson versus Katherine Harris race - where Nelson (.-320) defeated Harris (.561) quite easily. That would give a conservative boundary for the median voter at .121.
If we look at the varying pairs of Democratic and Republican candidates, we can adduce which candidates would be closer to that voter.
Assuming the "liberal" interpretation, this means that among the following six candidates for office: Gelber, Meek, and Klein, and then Buchanan, Mack, and Rubio, we would examine the pairs and figure out quickly which candidates would be favored in a head-to-head match.
Gelber would be closer to the median voter than Mack or Rubio.
Meek would be closer to the median voter than Mack or Rubio.
Klein would be closer to the median voter than Buchanan, Mack, or Rubio.
Buchanan would be closer to the median voter than Meek or Gelber.
Alternatively if we assume the "conservative" interpretation, this means the following winning combinations:
Gelber would be closer to the median voter than Rubio.
Meek would be closer to the median voter than Mack or Rubio.
Klein would be closer to the median voter than Mack, or Rubio.
Buchanan would be closer to the median voter than Klein, Meek, or Gelber.
Mack would be closer to the median voter than Gelber.
All of this neglects campaign effects and exogenous shocks to the system (scandals) - to the extent that those notions are not already factored into the original numbers. Now whether each party's voters are strategic in terms of maximizing their party's chances for winning (right now Klein and Buchanan), remains to be seen.
