Weekly Update: September a Month of Progress for Republicans

Summary

Republicans entered September on a virtual role;  Republican electoral prospects improved across the board during the summer months.  September could best be characterized as a month of consolidation for Republicans.  To many observers the upset primary wins of Christine O'Donnell and Joe Miller did not lead to a reversal of Republican momentum.  It is true that Republican prospects for victory in Delaware are priced significantly lower than at the beginning of the month, but every time pundits suggested that the Republican momentum was shifting into reverse other races popped onto the radar.

Republicans are positioned well heading into the campaigns last month.  Markets clearly expect a Republican House takeover.  While markets still price the prospects of Republican Senate control at around 25%, the number of seats they expect Republicans to control in the next Congress continues to increase incrementally.



House Outlook

According to our analysis of InTrade House Control contract data, Republicans are roughly a 3-1 favorite to gain control of the House of Representatives.  The midpoint probability estimate for Republican control of the House stands at 72.6%.  The range of estimates is fairly narrow -- the upper bound is presently 73% and the lover bound 72.2%.   To put this into context,  while this is a slight decline from the 75% registered September 1, it is significantly above the June 1 reading of 44%,

Our estimate of the number of seats Republicans will hold in the House come January stands at 225 down 1 from our September 1 estimate of 226.  This compares to our June 1 estimate of 212 seats.  It will be interesting to watch the pricing of the contract on projected Republican gains of more than 60 seats over the next couple of weeks; the InTrade data presently prices that probability at just over 25%.

The charts below show the upward progression of Republian House prospects as well as offering a perspective on the pricing curve for various levels of seat pickups compared with September 1.

Senate Outlook

The following chart illustrates the dilemna Democrats face with one month to go before the election.  The chart shows the change in the probability of Republican chances since September 1. While there have been some major declines for Republicans the number of states they have gained in far outweighs the number of states in which their prospects have declined.

Alaska and Deleware are the states with the biggest declines for Republicans.  In the case of Alaska, that reflects our model's philosophy of only counting a Joe Miller victory as a Republican victory.  A win by Senator Murkowski is counted as a Republican loss in our model.  If a Murkowski victory were also considered a Republican victory, there would have been little change since September 1.

The macro level model suggests that Republicans have a 23% chance of taking the Senate -- a level that is virtually identical to the September 1 pricing.   The InTrade contracts do not count anyone elected as an independent for determining control under the contract.  So 49 Democrats plus Senators Lieberman and Sanders would not constitute control under the contracts.  As the chart below shows, expectations of Democratic control with out the Independents has declined recently.

Our macro level model midpoint estimate stands at 48.9 Republican seats compared to the micro level model's estimate of 48.0 seats.     Upper and lower bounds for the macro model are 49.2 and 48.6 seats; this compared to corresponding micro model estimates of 48.5 and 47.6 for the upper and lower bounds.  The following chart shows the progression of the micro model's pricing range over the past month.

The following table shows the midpoint pricing estimates for all 37 Senate races on October 1 compared to a month earlier.

The two following charts show just how dramatically the micro model's outcome distributions have changed over the past month.

How Does ElectionMarkets.com Probabilities Compare to FiveThirtyEight.com Probabilities?

The following table shows the difference between our October estimates and FiveThirtyEight's most recent estimates.  There is no clear pattern apparent in the differences.  It will be interesting to observe any convergence or divergence in the estimates over the next two weeks.  Please note that only races with different estimated probabilities were included.

Macro vs. Micro Price Curve for Senate Races

The following charts outlines the pricing curve differences between the macro and micro level models as well as the macro price curve compared to September 1.