Saturday, August 28, 2010

Ten in Ten

Wherein we explore the prospects for restoring divided government by means of the GOP winning the Senate

NOTE: Prognostications in this post updated on 10-10-10 post [LINKED HERE]

Labor Day is fast approaching, bringing the unofficial end of a long hot summer and the official start of a short hot fall political election season. As a political blogger, the sun never sets on the election season. Time to take another stroll down the Dividist beach to see if that beautiful object of our affection, Divided Government, is drawing any closer.

Last time we looked for her the answer was "no", the same answer we found shortly after the election in 2008, and when we looked again in 2009. Conventional wisdom still says "no", but conventional wisdom has taken some surprising turns in 2010.

Conventional Wisdom
In January, the expectation was that the GOP would make gains in both houses of Congress, but fall short of retaking a majority in either. It just looked like the GOP was buried too deep in the sand to dig themselves out in one cycle. The Scott Brown "Massachusetts Miracle" eclipsed that particular ray of conventional wisdom, and since then CW has cautiously has settled on a partly cloudy forecast with a chance of heavy Republican rain. The current political weather report gives the GOP a good chance to retake the majority in the House of Representatives, but the Senate is still considered by most to be out of reach. Conventional wisdom is not unanimity, so you can find some grasping at straws, others fearing the worst, and a few wondering how bad it could get. To the center right, it looks like a done deal. We'll start our analysis by narrowing down the range of possibilities.

Every Possible Scenario
The entire universe of possibilities can be distilled to these four outcomes - listed in order of Current Conventional Wisdom:
  1. Democrats retain Senate, Republicans win House
  2. Democrats narrowly retain House and Senate
  3. Republicans win House and Senate
  4. Republicans win Senate, Democrats retain House
The best way to evaluate this would be a bottoms-up analysis looking at detailed polls and statistically correlating demographics and voting history on a district by district, state by state, and election by election basis. I'm not going to do any of that. For one thing, it is beyond my ken, for another, I can get all that from the usual suspects doing the polling and Nate Silver's blog doing the quant work. Instead, I’m going to look at the election through the prism of two "rules of thumb" and look for similarities and differences to historically analogous elections. And steal from Nate.

Maxims and Thumbs
The first rule of thumb does not get much publicity, but is an interesting fact that I've dubbed "The 100 Year Rule". In the almost 100 years since we have been been electing Senators directly (only since the 17th Amendment was ratified in 1913) the House of Representatives has never flipped majorities unless the Senate flipped first or at the same time. If conventional wisdom is correct and the Republicans take the House but not the Senate, it would be an historic first. So my first prediction is to go out on a limb and say this is not going to happen. Conventional wisdom is wrong and this scenario is the least likely of the four.

The second rule of thumb is Tip O'Neil's maxim "All politics is local." To the degree that O'Neills maxim is true, it is true about the House. This is just another way of saying (as is the first rule) that it extremely difficult to flip majorities in the House of Representatives. House incumbents, (frequently aided by gerrymandered districts) enjoy extraordinarily high re-election rates. Even when voters tell pollsters they despise Congress in general, they'll say they love their specific representative who is often the conduit by which federal services are delivered or expedited to individuals, municipalities, and businesses in the district. House elections are almost always "local." Almost.

Looking Back
1994 and 2006 were two midterm election cycles where elections were decidedly not local. They turned on national issues and the House of Representatives flipped majorities simultaneously with a flip in the Senate. These two mid-term elections shared several characteristics: We were under One Party Rule (Democrats in '94 - Republicans in '06); There was widespread dissatisfaction with the party in power; The opposition party was energized; The base of the incumbent party was disillusioned with a palpable lack of enthusiasm; There was a widespread perception that the party in power was arrogantly pursuing policies opposed by a majority of Americans; Major corruption scandals were in the headlines for the party in power throughout the election year (Rostenkowski in '94, Abramoff and Foley in '06).

Now, without a doubt, all of these elements are present in 2010. However, I don't believe the 2010 corruption stars (Maxine Waters and Charlie Rangel) rise to the level of the corruption superstars we had in '94 and '06. In both of those elections, the corruption scandals were the last straw and triggered the "throw the bums out" gag reflex in the voters. Unless there is an October surprise and more corrupt Democratic pols make into the headlines, I just don't believe there is enough animus to overcome the huge House of Representatives incumbent advantage to get the massive 40 seat shift. Plus, one should not underestimate Nancy Pelosi. My conclusion on the House: Close, but no cigar. 2010 will not be quite like 1994 or even like 2006.

Miracles Happen
So if we are to see divided government restored in 2010, the best chance will be the Senate. In January this looked like an impossible hill to climb. The Democrats held a 60-40 super majority and the tie-breaker in the person of Joe Biden. To gain the majority the Republicans would have to win 11 seats. Nobody in either party considered that realistic. But - then something remarkable happened. Republican Scott Brown won Ted Kennedy's Senate seat. You might not think that one seat would change the complexion dramatically, but it does.

Thanks Nate
Time to rip Nate Silver's work. His chart on the left is remarkable. It shows Nate's stack ranking of the Senate seats most likely to change parties. Of the top 12 seats most likely to switch parties, 11 of them are currently held by Democrats. All either have the Republicans leading in the polls or are within the margin of error. The one seat of the top 12 currently held by a Republican is the Florida Senate, and it is only there because Independent Crist is in a dead heat with Republican Rubio. The Democrat has no chance in Florida. And if Crist were to win, he would likely caucus Republican for reasons that I'll outline shortly. Now - this still appears to be a very tough climb as the Republicans need 10 of the 11 Dem seats in play to secure an outright majority. But wait! - there is another scenario - they may need to win only 8 or 9 of the 11 seats to take control of the Senate. How? the answer can be discerned by looking to the 2012 election.

2012 effect on 2010
This year the structural playing field is even. There are 37 Senate seats yet to be decided, with 19 currently held by Democrats and 18 held be Republicans. In 2012 the Republicans will have a huge structural advantage in the Senate elections. Of the 33 seats contested, 23 are held by Democrats and 10 by Republicans. The Democrats will be on defense with many more seats to defend, the Republicans will have a target rich environment. If they don't already have the majority, it it is a lock the GOP will take the majority in 2012.

Why is this important in 2010? Because Senators Ben Nelson and Joe Lieberman can count. If the GOP gets within 1 or 2 seats of an outright majority, Nelson and Lieberman will be in play. They'll have one shot to cut a deal to guarantee their committee chairmanships for at least another 4-6 years (if re-elected), whereas they will be out as Committee chairs after two years if they continue to caucus Democratic. This also applies to Crist should he knock off Rubio in Florida. My take - these guys like the power and perks that come with committee chairmanships and will not be inclined to give them up too quickly. It just would not be as much fun for them, being in the Senate without that chair. And lets be honest - its not like you liberals have been particularly nice to either of them over the last couple of years.

The Dividist Prognosticates
The Dividist 2010 election prediction: The GOP wins 8 or 9 more Senate seats outright, then takes majority control by flipping Lieberman and/or Nelson. They fall a few seats short in the House and Nancy Pelosi continues as Speaker of the House.

Stack ranking of all possible election scenarios in order of likelihood:
  1. Republicans win 8-9 seats flip Lieberman,/Nelson take Senate, Democrats narrowly retain House
  2. Democrats narrowly retain House and Senate
  3. Republicans win House and Senate
  4. Republicans win House, Democrats retain Senate
We'll be tracking this dirty dozen of Senate races in posts between now and the election to monitor our last best chance of restoring a perfect "10" in '10 and once again gaze upon a beautiful, desirable, smoking hot divided government in 2011.

Takeover Chances

Race

Margin
N. Dakota
Hoeven v. Potter +40
Arkansas
Boozman v. Lincoln +32
Indiana
Coats v. Ellsweorth +14
Delaware
Castle v. Cook +9
Pa.
Toomey v. Sestak +8
Colorado
Buck v. Bennet +5
Nevada
Angle v. Reid +1
Florida
Rubio v. Crist +1
Illinois
Kirk v. Giannoulias +0
Wash.
Rossi v.Murray -1
Calif.
Fiorina v. Boxer -2
Wisconsin
Johnson v. Feingold -3
Kentucky
Paul v. Conway +4
(chart from Nate Silver's 538)

Ok - I actually listed a baker's dozen. Rand Paul's race in Kentucky is included in the watch list, as his lead is barely out of the margin of error. It'll be an interesting race to watch. Like Sharon Angle in Nevada, Rand Paul was a tea party favorite that knocked off the GOP establishment candidate. In both cases, media gaffes prompted unfavorable polls and a lot of chortling from the left. The left-o-sphere was confident that the Tea Party had torpedoed GOP chances in Kentucky and Nevada. I checked in with Moonage, a Kentucky blogger who always has his finger on the pulse of local politics. He is calling Kentucky for Rand Paul. Take it FWIW, but I think Moon's got it right.

Nevada? Another matter. Frankly I am astonished that this race remains a virtual dead heat, and it is prompting signs of panic on the left. The Nevada race may be the single clearest indicator that this may be a bigger GOP tsunami than Conventional Wisdom has yet to acknowledge.

Divided and Balanced.™
Now that is fair.



6 comments:

Tully said...

Leiberman's not really in play. His loyalty to the Dems has been proven over and over again -- he will oppose them when he thinks they're dead wrong, but he will not leave them, even when they try to shoo him away by beating him with a stick.

Nelson, on the other hand ... switching to GOP is about the ONLY thing that could help him get re-elected at this point. That, or the full frontal mea culpa in public as regards his ObamaCare vote.

Current House/Senate seat-flip over/under is now +GOP 48/7, and your friendly neighborhood political bookie will most happily sell you the under in either house of Congress, while being most reluctant to sell you the over.

Hitting the "push" would give the GOP a majority of 226 to 209 in the House, and a filibuster-sustaining-even-with-Snowe-and-Collins-and-Brown 48 seats in the Senate.

In the House, well, of the 35 seats that look like toss-ups and aren't already leaning strongly GOP in the polling, only ONE is currently GOP. There are only about 4 GOP seats really in play for potential switching, but 65 or more Dem seats in serious play -- or already lost.

In the Senate, the GOP is pretty certain of picking up at least three seats before toss-ups are considered, bringing them to 44. Of the seven seats in the toss-up category right now, six are DEM seats and the other is the Rubio/Crist race, where even if Crist wins he'll caucus GOP. So the GOP can reasonably expect to get to at least 47 seats anyway.

The GOP could well match or exceed the 1994 switch numbers of 54/8 this cycle.

Tully said...

Oh, and the GOP looks to pick up a net of 8 governorships this cycle as well.

mw said...

Sabato has a new prediction out today that are close to mine on the Senate and yours on the house and govs.

I'm just glad I got my post and predictions out ahead of Sabato, as he is making many of the same points. Now I can claim he cribs from my blog.

I think Lieberman is loyal to Lieberman first. He was ready to run on the McCain ticket. McCain backed out. If it means keeping his chairmanship, he'll flip. OTOH, if Sabato is right and the GOP wins nine more outright, he may not be needed, Nelson will give them the majority.

Tully said...

Leiberman may be loyal only to Leiberman, but he knows he would find it VERY difficult to get re-elected as a Republican, and he's up again in '12. His constituency and his own consistent centrist/liberal philosophy are such that he would get savaged trying to be a Republican in a "Tea Party Revolution" Congress. He's safer staying indie and swinging as he sees fit, caucusing mostly with the Dems.

Of course, I could be wrong, but I like to stick with the odds ... ;-)

My take is based on a fairly conservative "odds" approach that takes into account the time remaining and the fact that while it's only two months, it still ain't over until it's over and future events can swing the margins. If I were betting on what the elections would produce if held today, my numbers would be somewhat higher for the GOP. But for now I'm sticking with the bookie numbers -- and trying to find one that will sell me the over.

Anonymous said...

At least you didn't include Christine O'Donnell in this list.

mw said...

Anon,
No - I said all I had to say about Christine O'Donnell in this post. However, since she has no chance of winning, I am perfectly happy with the Dems and the Media focused on her, and taking their eye off the ball on the Senate Races that matter.