Reading the Political Tea Leaves

As most followers of politics know, in midterm elections, the party not controlling the Presidency tends to pick up a bunch of seats. While there are exceptions to this rule like 1998 and 2002 were, the average pick-up is 28 seats over the last 17 midterm elections. Sometimes the shifts are even more dramatic. The Republicans lost 48 seats in both 1958 and 1974 and the Democrats lost 54 seats in both 1946 and 1994. Most people would admit that it would appear that 2010 is going to look a lot more like 1994 when the Democrats lost 54 seats than it will 1978 when they only lost 15 seats. At the very least, 2010 will be another 1950 when they lost the average of 28 seats.

There seems to be quite a disconnect between what is happening in politics and what the analysts are predicting will be the results of the November elections. This is particularly true in the 435 races for the House of Representatives. There are four reports for these results: CQ Politics, the Cook Report, the Rothenberg Report and our own, the Marston Report. Thus one would expect all four of them to show at least 28 Democrat seats going Republican and perhaps one showing as high as 54 seats in view of what happened in Massachusetts with the election of Scott Brown. None of the first three show other than minimal gains for the Republicans.

CQ Politics predicts a net gain of 3 seats with 18 toss-ups. Charlie Cook predicts a net gain of 6 seats and 23 toss-ups. Stu Rothenberg predicts a net gain of 10 seats and 10 toss-ups. Only our own report shows a net gain of 63 seats for the Republicans. There are 9 polls taken in House races in recent months showing the Republican candidate ahead or tied with the incumbent Democrat. One would expect those races with the Republican ahead to be favoring the Republicans and the tied one to be a toss-up but that is not the case as the following table shows:

RACE	POLL	CQ	COOK	ROTHENBERG	MARSTON
Indiana 9	R +8	Leans D	D toss-up	Leans D	Leans R
Maryland 1	R +13	Toss-up	D toss-up	R toss-up	Likely R
Massachusetts 10	R +3	Likely D	Likely D	Safe D	R toss-up
Michigan 7	R +10	Toss-up	D toss-up	Toss-up	Likely R
Nevada 3	Tied	Leans D	D toss-up	Leans D	Leans R
New Hampshire 1	R +10	Leans D	D toss-up	Toss-up	Leans R
New Hampshire 2	R +7	Toss-up	D toss-up	Toss-up	Leans R
North Dakota AL	R +6	Leans D	Leans D	D favored	D toss-up
Ohio 1	R +17	Toss-up	D toss-up	R toss-up	Leans R

As you can see, CQ Politics is the least realistic listing 5 races as favoring Democrats when the polls have a Republican from tied to 10 points ahead. The Cook Report shows 2 races as favoring Democrats when the polls have a Republican up by from 3 to 6 points and even the 8 toss-ups as slightly favoring the Democrats. The Rothenberg Report shows 4 races for the Democrats when the polls range from tied to a Republican up by 8 points but at least it shows that 2 of the races lean slightly to a Republican candidate when the polls have him up from 13 to 17 points. On the other hand, Rothenberg lists MA-10 as not even in play and hence is a safe Democrat district when a poll has a Republican candidate up by 3 points even before the Amy Bishop scandal.

We make no pretense about our not being Republican analysts and that opens us up to being biased in favor of the Republican Party. In light of all of the above, we feel that our own report is the most realistic given the known facts, but we will let the public come to its own conclusion on that score.