[Dean's World] Andrew Cory: An overview of the 2006 midterms (House).

[notify at powerblogs.com]

Posted by Andrew Cory:
An overview of the 2006 midterms (House).
http://www.deanesmay.com/posts/1148447782.shtml


   Iâve seen a couple polls of the 2006 race, and the Democrats seem to
   have an edge here. For the first time in about a decade, their
   favorable rating is higher than the Republicansâand people are pissed
   at the Republican Party. Yet national polls are not all that important
   when analyzing a congressional election. With no Presidential
   candidate to act as a party standard bearer, and more importantly with
   no presidential coattails to ride on, each candidate must stand on
   their own merit. Each race must be looked at as itâs own creature...

   Yet there are trends; in America (from about 1950 onward) incumbents
   have something like a 99+% reelection rate. Part of the reason (a
   large part) is that incumbents who feel they will probably lose tend
   to retire rather than invest in a doomed campaign. Thus the best
   indication that a seat up for changing partisan hands is itâs vacancy.
   Thus, I will develop some formulas with that in mind...

   The first formula I will give is for seats with an incumbent running.
   We expect 1% of all these seats to change hands. I will calculate
   these separatelyâas if each Republican and each Democrat had an equal
   chance of winning or losing their seat. The formula is: (Number of
   incumbent Republicans running)*01 (Number of incumbent Democrats
   running)*01

   I also think that Democrats will gain roughly 75-80(*)% of all open
   seats. Therefore our formula shall be: (open seats)*.75

   And now we look at the makeup of Congress. Currently the House is
   split 231 Republicans and 202 Democrats (1 of those is an independent
   who votes Democratic) with a pair of vacancies. There are 30 seats in
   2006 in which no incumbent will be running(**). Of these, 8 were held
   by Democrats, and 22 by Republicans. Therefore we can plug in our
   numbers: (231-22)*.01=2.09 (201-8)*.01=1.93 So we expect 4 incumbents
   to lose their job, and they ought to be evenly split between the
   parties.

   Which brings up the non-incumbent seats. Democrats will win between 22
   (rounding down) and 24 seats.

   Under the scenario with the smallest Democratic gains, this would
   leave the House with Republicans in charge: 215 (D) vs. 220 (R). Under
   assumptions with the largest gains, Republicans are still charge, but
   will have only a 3 person majority. This result seems to be inline
   with current speculation. The message to Democrats is obvious: donât
   get your hopes up...

   ____ (*) thereâs nothing particularly super scientific about these
   numbers. But they feel about right to me. You may feel differently...

   (**)the two vacancies will have an incumbent with less than a year on
   the job

   NB: Open Seat numbers from [1]This Wikipedia article

   (crossposted on [2]Punningpundit.com, a site you should read more
   often)

References

   1. http://en.wikipedia.org/wiki/U.S._House_election,_2006
   2. http://www.punningpundit.com/archives/2006/05/an_overview_of.html