« This Makes No Sense | Main | NH-02: We've Got a Race on Our Hands »

Wednesday, May 03, 2006

How Badly Do Losers Lose?

Posted by DavidNYC

This post is a response to some of the points raised by Chuck Todd, whose new piece on the 2006 elections I mention below. Let me elaborate on what I mean by the question I pose in the title.

What I wanted to know is, when the party in power loses seats in Congress, how bad do those losses tend to be? So I did a little research. I limited myself to the post-17th Amendment period of American politics - ie, 1914 onward, an era which covers 45 elections. During that period, the House switched hands seven times. Meanwhile, the Senate flipped eleven times. I don't think the frequency of flipping actually matters, though. I think the magnitude of losing is the important factor to look at.

With that in mind, the party in power ("PIP") in the House lost seats 25 times and gained seats 20 times. In the Senate, the PIP lost seats 26 times, gained seats 17 times, and remained unchanged twice. Those sets of numbers are pretty similar: In the House, 55.6% of the time, the PIP loses seats, while in the Senate, it's 57.8% of the time.

But here is where they are rather different: The median seat loss in the Senate (during losing years) is six. The median seat loss in the House, meanwhile, is 20. Put another way, 16 out of 25 losing years (64%) saw a seat change equal or greater to what we need this year (15) in the House, while only 14 of 26 losing years (54%) provided a margin equal or greater to what we need this year (6) in the Senate.

In other words, the Senate might change hands more often, but the delta of those changes is a lot less volatile than in the House. Of course, you can argue that the House is so much more gerrymandered these days and is likely to be less volatile today than it was in the past. If you're willing to consult a very small sample size, this is true: From Clinton on, the PIP lost 15+ in the House only once (1994), while it lost 6+ in the Senate twice (1994 and 2000). But if you extend the frame only as far back as Reagan, then it's four times in the Senate to three in the House.

This is all on a macro level. On the micro level, several commenters have pointed out that Dems would have to shoot the moon this year to take back the Senate. While it may be historically true that one party tends to do unusually well in Senate elections, the micro analysis just isn't that favorable to the Dems this year. Even if we draw a royal flush in PA, RI, MT, MO and OH, we'd still need a sixth seat. TN? AZ? VA? NV? With the possible exception of Tennessee, those are all extreme longshots.

In the House, on the other hand, we've got a lot more than 15 decent shots. There are (as of this moment) 23 GOP-held seats on CQ's list which are either no clear favorite or lean Republican, and Charlie Cook has 24. And on both of those lists, I think there are some hedges (for example, neither Sweeney nor Weldon are listed as toss-up or category 1 yet).

So for now, I'm still gonna say that I think the House is more likely to switch hands. I'm not saying I think we're gonna take back the House, but I could definitely see a House flip without a Senate flip as well.

Posted at 11:31 PM in Elections | Technorati

Trackback Pings

TrackBack URL for this entry:
http://www.swingstateproject.com/mt/mt-track-ssp.cgi/2325

Comments

Target list at least for me: 44 targets for me not measured in any way but all could go from Red to Blue. About a dozen seats or so could go Blue to red.

AZ-05, AZ-08, CA-04, CA-11, CA-50, CO-04, CO-07, CT-02, CT-04, CT-05, FL-09, FL-13, FL-22, IL-06, IL-10, IN-02, IN-08, IN-09, IA-01, KY-03, KY-04, MN-01, MN-06, NV-02, NV-03, NJ-07, NM-01, NY-20, NY-24, NY-29, NC-08, NC-11, OH-01, OH-15, OH-18, PA-06, PA-07, PA-08, PA-10, TX-22, VA-02, WA-08, WV-02, WI-08.

Posted by: D in FL. [TypeKey Profile Page] at May 4, 2006 01:09 AM | Permalink | Edit Comment | Delete Comment

your forgetting nh1 + nh2.

Posted by: yomoma2424 [TypeKey Profile Page] at May 4, 2006 11:07 AM | Permalink | Edit Comment | Delete Comment

In the US House. The Red Districts that can go Blue is
1)Arizona-5
2)Arizona-8
3)California-11
4)California-26
5)California-50
6)Colorado-4
7)Colorado-7
8)Connecticut-2
9)Connecticut-4
10)Connecticut-5
11)Florida-8
12)Florida-13
13)Florida-22
14)Illinios-6
15)Illinios-11
16)Indiana-2
17)Indiana-8
18)Indiana-9
19)Iowa-1
20)Iowa-2
21)Kentucky-3
22)Kentucky-4
23)Louisiana-7
24)Minnesota-1
25)Minnesota-2
26)Minnesota-6
27)Nebraska-1
28)Nevada-3
29)New Hampshire-1
30)New Hampshire-2
31)New Jersey-7
32)New Mexico-1
33)New York-19
34)New York-20
35)New York-24
36)New York-25
37)New York-26
38)New York-29
39)North Carolina-8
40)North Carolina-11
41)Ohio-1
42)Ohio-2
43)Ohio-15
44)Ohio-18
45)Pennsylvania-4
46)Pennsylvania-6
47)Pennsylvania-7
48)Pennsylvania-8
49)Pennsylvania-10
50)Texas-22
51)Virginia-2
52)Washington-8
53)West Virginia-2
54)Wisconsin-8
55)Wyoming

We will pick up
1)Arizona-8
2)California-50 Special Election
3)Colorado-4
4)Colorado-7
5)Connecticut-2
6)Connecticut-4
7)Florida-22
8)Indiana-2
9)Indiana-8
10)Indiana-9
11)Iowa-1
12)Kentucky-4
13)Minnesota-6
14)Nevada-3
15)New Mexico-1
16)New York-24
17)New York-29
18)North Carolina-11
19)Ohio-18
20)Pennsylvania-6
21)Pennsylvania-8
23)Texas-22
24)Virginia-2
25)Washington-8

Posted by: mrcharlesmburns [TypeKey Profile Page] at May 5, 2006 11:47 AM | Permalink | Edit Comment | Delete Comment