Friday, December 31, 2010

2012: New York

Senate: Kristin Gillibrand (D)
Legislature: Republicans +7 (+17%)
State Senate: Republicans +2 (+7%)
House: Democrats 21 Republicans 8
Redistricting: Bi-partisan

I wouldn’t put much stock in Republicans beating Gillibrand in 2012 unless Pitaki or Guilani were to run. On the House side, New York was the land of opportunity for Republicans in 2010, winning 6 seats. Yet there will be opportunities for pick-ups in 2012. Republicans lost 3 seats by a total of 6 points, even after winning those six. Their 3 smallest margins of victory added up to 9 points, so they were closer to gaining than losing.

Redistricting can be split into two parts, the 17 downstate districts, which will lose 1, and the 12 upstate districts that’ll also lose 1. As I showed in the post about the VRA, Republicans should have a fighting chance in at least 3 downstate districts, including Steve Israel’s 2nd district. Democrats could try to pack in Democratic votes to protect Israel, since he’s the head of the DCCC, but that would leave Tim Bishop and Gary Ackerman vulnerable.

Since a Democratic district will be eliminated downstate, the trade will likely be that a Republican district will be eliminated upstate. The question is whether they pit two Republicans against each other or an incumbent Republican against a Democrat. It isn’t easy to create safe districts upstate. There are heavily Democratic strongholds like Buffalo, Rochester, and Albany but the Democratic cities aren’t as Democratic as New York City. While it’s possible to make two reasonably safe Democratic and two reasonably safe Republican districts in Western New York, the five Eastern districts will likely be swing districts. Republicans Nan Heyworth (19th), Chris Gibson (20th), Richard Hanna (24th), and Ann Marie Buerkle (25th) and Democrats Paul Tonko (21st), Maurice Hinchey (22nd) and Bill Owens (23rd) will be squeezed from six districts to five. My speculation is that Owens will go up against Hanna. Since all these districts will be swing districts I don’t imagine two Republicans will run against each other in a primary rather than running in two districts.

Thursday, December 30, 2010

VRA and Redistricting

The 1965 Voting Rights Act has provisions in it to make sure minorities aren’t disenfranchised. Prior to this, White legislators would take an African-American community of 250,000 and split it into 2-3 districts to dilute African-American voting power. The VRA was designed to keep these communities together. It’s gone a bit further than that. Legislatures have gerrymandered districts to take that 250,000 and find another 100,000 people somewhere in the state to create a Black majority district. That’s what they do now.

It works great for African-American and Latino representatives. With a majority/plurality from their ethnic group their re-election is virtually guaranteed. In some states this is a big positive for Democrats. Right now the only districts Democrats have in Louisiana, Alabama, Mississippi, and South Carolina are minority majority. Democrats are having trouble appealing to Whites and don’t have to worry about trying in these states in order to win some districts.

The VRA also helps Republicans. By combining Blacks in Selma, Tuscaloosa, and Birmingham the remaining districts become very difficult for the Democrats to win. If AL-7 were 40% African-American instead of 62%, there could be enough Democrats in another district to make them more competitive.

Missouri is going to lose a congressional seat. Right now there are three Democratic districts, one of which is majority minority and a second one that is 30% Black/Hispanic. These districts will need to add people in redistricting. The 1st district is currently heavily Democratic and could easily add Republicans and still be an easy win for the Democrats. If they did, however, there’s no guarantee the person elected would be Black. In order to make sure the Republicans give the 1st and 5th as many African-Americans as possible, Black Democrats will sacrifice Russ Carnahan’s seat. Republicans will take that trade.

New York will lose two seats in redistricting, one of which will be in the New York metropolitan area. There are currently 17 districts, 15 of which are Democratic and only 2 Republican. You’d think that after redistricting it’d be 14-2 Democratic. Maybe. Maybe not.

There are currently 8 majority minority districts in the metropolitan area and 2 districts that are largely in Manhattan. Since Republicans hold a majority in the state senate they’ll probably tell the Democrats they can gerrymander New York City and Long Island however they want as long as they a) Keep as many Republicans in Peter King’s district as possible b) make sure all of Staten Island is in the 13th.

Even if they try to put as many Republicans as they can in these two districts, they’ll still run into a problem. They’ll have to create 8 majority minority districts. These districts, on average, will need to increase in size by 65,000 voters. In order to maintain majority minority status, they’ll have to grab as many minority voters from surrounding Democratic districts. That’ll make the other districts less Democratic.

I used Dave’s redistricting App. I eliminated the 4th district since a Democratic district will need to be eliminated. By maximizing the minority make-up in the 8 majority-minority districts, this is what I came up with:

I wasn’t attempting to make the districts more Republican. By gerrymandering the districts to keep minorities in the other eight districts, almost all of these districts go more Republican. The 8th and 14th are the New York City districts. There may be a way to have the 8th grab less Democrats, but there are four majority minority districts in Queens and Brooklyn bordering Manhattan. They could grab some Republicans from the 9th or 13th, moving them a little bit toward the Democrats. Any way you cut it, however, the Democratic 1st, 2nd, 5th, and 9th will likely have to become more Republican. This is after eliminating the Democratic 4th and maximizing Republicans in the GOP held 13th and 3rd. Democrats failed to get to 60% in 3 of the 4 remaining districts. None of the 2010 candidates were part of the GOP’s Young Guns program. So they were neither a top candidate nor got help from the NRCC. If these districts become more Republican, however, they could be in play in 2012.

Wednesday, December 29, 2010

2012: Texas

Senate: Hutchinson (R)
Legislature: Republicans +22 (+29%)
State Senate: No change
House: Democrats 9 Republicans 23
Redistricting: Republicans

Texas gains four electoral votes in 2012, so that’s four more electoral votes for the Republican candidate. Democrats have no chance of taking this state.

Texas will gain four congressional districts in 2012 and that’ll be a huge Republican advantage. Democrats have a fantasy that the influx of Hispanics into Texas will eventually turn the state blue, or at least a little less red. Once again the future will belong to the Democrats. The way things have changed every two years this decade I don’t see how anybody knows how things will be in 10 years.

There are some problems with this idea. While Hispanics are 36% of the Texas population, they were only 20% and 17% of the electorate in 2008 and 2010. Many Hispanics aren’t eligible to vote and those that are vote at a much lower rate than other groups. This isn’t California. While Republicans don’t do well with Mexican-Americans elsewhere they do okay here. This is George W. Bush’s home state and he pulled 45% of the Hispanic vote in 2004.

That’s not to say Hispanics don’t matter. There are currently 11 Congressional districts in Texas that are majority minority or nearly so. The Democrats hold 9 of them and lost the 2 others by narrow margins this year. There should be 13 or 14 majority minority districts in 2012 and the Democrats should be heavy favorites in 11 of them.

That’s the bad news. The good news is that Republicans know how to gerrymander in Texas. They can likely turn the majority minority seats that Canseco and Farenthold have into safe seats, while still retaining a minority majority.

Like the rest of the south, Democrats get clobbered with White voters. White voters will be in the majority in 22 or 23 congressional districts. The GOP should easily be able to end up at least with a 24-12 advantage.

The senate race should be interesting, but not because Democrats have a good shot at winning it. Kay Bailey Hutchinson was retiring then kinda retiring and now we don’t know. She isn’t terribly popular with conservatives and will get a tea party challenge if she runs for re-election. Well, I expect every sitting Republican senator to get a tea party challenge, simply because any challenger will declare him or herself to be a tea party candidate. Maybe John Barrasso can dismiss the possibility but the other nine Republican incumbents better take it seriously.

Tuesday, December 28, 2010

Red State Blue State Part II

If Democrats are turning red states bluer, then the states they’re leaving behind should get redder. The chart below show legislative gains/losses in 15 states the GOP hasn’t won for the Presidency since the 1980s. The 7 states on top are ones where Republicans didn’t win any congressional seats in 2010. Republicans cleaned up in the ones on the bottom, gaining 22 seats total. California is excluded because it’s so heavily gerrymandered with almost no seats flipping either way all decade. New Jersey didn’t have elections this year.

The group at the top has been held up to show how the Republican wave bypassed the northeast and west, but that was only at a congressional level. Republicans made net gains in almost every state, and took two houses in Maine and now have the same number as the Democrats in the Oregon legislature. While gaining 14 in the Connecticut legislature and 15 in Massachusetts isn’t a sign the states will swing GOP in 2012, it is a sign that the wave didn’t miss these states.

The second group may be another story. Before the election Democrats controlled 14 of the 16 chambers in this group. Now they control 7. There’s no chance Maryland, New York, or Illinois will go GOP, but the other 5 could be up for grabs.

Red State Blue State Part I

Democrats have come up with the idea that the people are moving from blue states to red states must be Democrats. So they’re going to turn red states, well, a little less red. What you see is what you get. These people are already voting and how they vote can be seen in congressional and Presidential elections. We’re seeing the results. Nevada and Colorado look like they have flipped, but the only other state that looks like it’s moving Democratic is Montana.

Increasing the number of districts in a state just divides the Republican and Democratic votes in the state a little bit differently. With Republicans controlling the redistricting in some of these they’ll just divide them up a little bit differently. The more districts you have, the easier it is to gerrymander.

The chart below shows where each state was in 2004, before the switches, 2008, at the Democratic height, and where I expect them to be in 2012. The top group are the states gaining seats, while the bottom group has the states losing seats.

My projections have the Democrats gaining 8 seats and losing 6. The Republicans gain 4 seats and lose 6. “Aha!” The liberal says, “I see! The Democrats will gain with redistricting!” Yes, they will and part of it is the blue state people moving to red states. More of it has to do with the bloated number of seats the Republicans hold. Even without redistricting they would be bound to lose seats. They simply hold to many marginal districts.

If the Democrats net 2 seats out of the states that are redistricting, that would mean they’d have a 240-195 deficit. I think the GOP can live with that. They’ll have to rely on the states that are retaining their seats. That’ll be difficult. The only states where Republicans currently have seats and the Democrats control redistricting are Arkansas, West Virginia, and Maryland. Arkansas and West Virginia actually gave John McCain a higher share than they gave George Bush. So these states are moving heavily in the other direction. Maryland is already heavily gerrymandered. It’s conceivable the Democrats could find a way to go from a 6-2 advantage to a 7-1 edge, but that’d only net them 1 seat.

There aren’t a lot of opportunities out there beyond that. They could gain in Colorado, Minnesota, Virginia, Tennessee, New Hampshire, or the Dakotas but I wouldn’t expect much. They’ll need a favorable redistricting from the California commission, which again can’t be counted on.

It’s easy to see that the Democrats should pick up 5-10 seats in 2012, but they’d need people to hate Republicans as much as they did in 2006 and 2008. With George Bush in Texas that won’t happen.

Monday, December 27, 2010

Redistricting: Congress

A lot of people think the Republicans will make big congressional gains because they control so many state legislatures. That’s unlikely because the GOP made such huge gains in the 2010 election. The good news is that Republicans hold only two districts where John McCain got less than 42% of the vote, and in those he got 41% and 38%. They hold 13 seats John Kerry won, while the Democrats hold 12 seats McCain won.

There are, however, 63 Republicans in districts that are between R +4 or less while only 39 Democrats in D +4 or less. If each party won half of those, the Democrats would pick up 12 seats.

With control of redistricting I’d expect there to be very little if any net change in 2012. Here are the states gaining seats:

Texas: The Democrats will likely take 3 of the 4 new seats, but that’s only because Republicans will want to strengthen incumbent seats. It’s certainly possible that the GOP could win 26 or even 27 seats in a good year, but I think 24 of the 36 districts is the best the Republicans can gerrymander. If they try to take too many seats they run the risk of losing more if things break a little bit wrong.

Florida: Florida was heavily gerrymandered for Republicans in 2002. This year all the Florida balls bounced their way. Right now it’s 19-6 Republican. In a fair redistrict, Republicans would likely get a 15-12 advantage. Yeah, that bad. I don’t think that’ll happen, but 17-10 seems likely.

Arizona: While Republicans control government, redistricting is non-partisan. The state could go from 5-3 Republican to 6-3, but 5-4 is more likely.

Georgia: There are three minority majority Black districts, but after that every district is on the table. I could easily see Georgia going from 8-5 Republicans to 10-4 with another GOP redistricting.

Nevada: This state is leaning Democratic now. The GOP has a 2-1 advantage and probably can expect a 2-2 split.

South Carolina: Unless the Justice Department requires the legislature to create a second majority minority district, Republicans will pick up the new seat.

Utah: Easy GOP pick-up. An additional district enables them to spread the Democrats over more districts. They may be able to finally unseat Jim Matheson.

Washington: Another non-partisan redistricting. The population growth is in the Seattle area, a Democratic stronghold. A new Seattle district will mean that the 4 Republican districts will recede further from that city and be safer for the GOP.

Here are the states that are losing seats:

Illinois – A potential GOP disaster. Republicans control 11 of the 19 congressional districts in a Democratic state. Expect the Dems to combine the down state districts represented by Bobby Schilling and Aaron Schock. Democrats are likely to make the districts represented by Bob Dold and Joe Walsh more Democratic too. Obama won 8 of the 11 districts now held by Republicans, so a 10-8 Democratic advantage with Republicans losing 3 seats might be a good outcome.

Iowa: As I’ve said before, this is a good one for Republicans. Tom Latham (R) will likely face off against Leonard Boswell (D) in a district favorable to Latham. Boswell will be 78 years old and may opt to retire. The other two seats could also be vulnerable. Democrats will almost definitely lose 1 seat here, if not 2 with a Republican pick-up.

Louisiana: Republicans control 6 of the 7 seats. The Democratic seat is majority minority. So there’s no way Republicans don’t lose a seat. The 2nd may stretch into Baton Rouge in order to maintain its minority majority status. The 4th and 5th will need to move south, adding counties from the 6th and 7th. If Lafayette moves into the 6th, the 3rd and 4th would be combined.

Massachusetts: Democrats have all 10 seats. After 2012 they’ll have all 9. They’ll likely eliminate the 7th, 8th, or 9th, depending on who challenges Scott Brown for the Senate.

Michigan: Republicans have a big advantage controlling redistricting, since there is a huge population loss in Democratic southeast Michigan. John Dingell’s district will be eliminated and gerrymandered in a way the Democrats will lose a seat.

Missouri: The Republicans are lucky that the St. Louis area lost the most population. Russ Carnahan (D) could be the big loser, although Todd Akin (R) could also be in trouble. It’s a question of how much of each of their districts Bill Clay’s district takes.

New Jersey: Another commission, this one bi-partisan. Two incumbents will have to run against each other. With population loss in the center of the state, either it’ll be Leonard Lance or Chris Smith on the Republican side and Frank Pallone or Rush Holt on the Democratic.

Pennsylvania: They know how to gerrymander in Pennsylvania and they’ll eliminate Mark Critz’s 12th. They’ll put more Republicans in the 6th, 7th, and 11th, to push them slightly in favor of Republicans. Pennsylvania is such a swing state, however, that it could go anywhere from 14-4 Republican to 12-6 Democratic even with a Republican gerrymander. I expect the Democrats to lose a seat in 2012 to make it 12-6 Republican.

New York: Two seats are going, but it might not be so bad for Republicans. Upstate is Republican enough that the GOP might not lose a seat. Freshmen Anne Marie Burkle or Richard Hanna will likely face off against either Democrats Bill Owens or Maurice Hinchey. The other loss is likely Democratic in Long Island. Republicans hold only two seats downstate. Michael Grimm’s seat is largely Staten Island, so that should retain its integrity. There’s no way Republicans will let Peter King be in a district that’s any less Republican. He’s the only Republican who has been in Washington more than two terms and has high level positions on Republican committees.

Ohio: A 13-5 Republican advantage likely means that there’s no way the GOP doesn’t lose one of the two seats. Democrat Betty Sutton will see her district eliminated. One Republican congressman will likely challenge Sherrod Brown for the senate. His district will be eliminated.

California: Who knows? I think it’ll favor the Demcorats but you could get some “crazy” redistricting. Actually it’s crazy now. We’ll get common sense redistricting. And that could create havoc. Darrell Issa represents the 49th. He’s from Vista, whichis in the southwest corner of his district. Brian Bilbray represents the 50th. He’s from Carlsbad, at the northern edge of his district. Vista and Carlsbad are neighboring cities in North County. In a common redistricting these two cities would be in the same district. It certainly isn’t unusual for two sitting Republican congressmen to be pitted against each other, but there may just be a new district right next to this one that covers no congressmen’s homes. So you may see a big scramble and end up with open seats.

I don’t see redistricting costing any other seats. It’s possible Republicans could squeeze out a Democrat in Indiana, Oklahoma, Tennessee, or Wisconsin. Democrats in Maryland are creative, but it won’t be easy to squeeze out another district.

Thursday, December 23, 2010

Redistricting Advantage

Chris Cillizza is the first mainstream media member to observe what I did last month. The GOP currently has a 62-29 advantage in districts that are between R+4 and D+4. The middle right now is a Republican 228-207 advantage. Republicans are 14 above that. If they're lucky, the Republicans can use redistricting to keep their net losses at 5-7 instead of 15-20. Republicans would need an electorate like 2010 and the same advantage with independents. They're not going to get that.

Tuesday, December 21, 2010

Redistricting: The Presidential Race

The U.S. Census numbers came out today. Those are conservative cheers you’re hearing today. Assuming the Republicans win all the states they won in 2008, the GOP has picked up 6 electoral votes. If Republicans lose any of those states, Obama will have won. Of course that only changes the gap from 192 to 180, not exactly close to a win. The path to victory will have to include Indiana, Virginia, and North Carolina, all close losses.

The likely 2012 swing states will be Iowa, Wisconsin, Michigan, Ohio, Pennsylvania, Florida, and New Hampshire. All these states went to Obama, but all of them will have a Republican governor and Republican control of both houses of the legislature. If one candidate takes three of Pennsylvania, Ohio, Florida, and Michigan, he’ll win. If the candidates split those four states, 20 electoral votes will be left with New Hampshire, Iowa, and Wisconsin. The candidate winning Florida might only need win the smallest of the states, New Hampshire, to get to 270 electoral votes.

I don’t think Nevada, Colorado, Washington, New Jersey, Maine, Minnesota, Missouri, or New Mexico will be swing states, because the 2010 elections were more split. Seven of those were won by Obama in 2008, with Minnesota, Maine, and New Mexico were the best for the GOP in 2010.

2012: Colorado

Senate: None
Legislature: Republicans +6 (+22%)
State Senate: Republicans +1 (+7%)
House: Democrats 3 Republicans 4
Redistricting: Split

Republicans made significant gains in Colorado in 2010, picking up the two most rural Colorado congressional seats and enough legislature seats to give the GOP the majority in that house. Don’t be fooled, however, into thinking this is a Republican comeback state. The electorate was 33% Democratic/28% Republican/39% independent. This marks the first time there were more Democrats than Republicans in a Colorado electorate. Of the 28 states with exit polls either this November or November 2009, Colorado had the biggest shift away from Republicans compared to both 2008 and 2004.

The only reason Republicans did so well was because they took independents by a 53%-37% spread and independents are the largest group in the state. If independents fall more equally the GOP will get beat in this state. They may still be able to hold onto the state’s rural congressional districts, both of which voted for McCain in 2008. I see little chance of putting this state back in the GOP presidential column. Colorado is the least likely of all the ten states Bush won in 2004 or 2000 to flip back to the Republicans. Even Wisconsin, Minnesota, Michigan, Pennsylvania, and New Jersey, none of which Bush won either time, are more likely to flip Republican. The GOP would be wise to put its resources into another state.

Monday, December 20, 2010

2012: Michigan

Senate: Stabenow (D)
Legislature: Republicans +20 (+47%)
State Senate: Republicans +5 (+24%)
House: Democrats 6 Republicans 9
Redistricting: Republican

The GOP hasn’t won Michigan in a Presidential race since 1988. They’ve lost 12 of the last 13 senatorial elections since 1972. Yet there’s reason for optimism for the GOP. Michigan currently has roughly the same number of people it had 10 years ago. It may end up being the only state in the union with no population growth during that period. The reason why that’s good news for the GOP is that the Democratic areas in and around Detroit have seen a mass exodus of people, while the Republican areas in the rest of the state have actually had population growth.

Republicans control redistricting due to their big 2010 gains and it shouldn’t be hard to make a Democratic seat the one that the state loses. The black majority 13th and 14th congressional districts have shed a lot of people and will have to take in the more ethnic areas in the 15th and 9th to remain majority minority. This will likely mean that the 15th district is split up with Dearborn and Romulus going into the 14th, Ann Arbor into the 7th, and the rest of the district into the 11th. John Dingell is the longest serving member of the House of Representatives, but he’ll likely have an uphill fight if he runs for re-election again. He makes his home in Dearborn. If it is in a majority minority district, Dingel would have little choice but to run against Republicans Thad McCotter in the 11th or Tim Walberg in the 7th.

Both of these districts represent opportunities for the Democrats. Obama won both of these districts. Bush won them in 2004 and both have a current Republican PVI. Even after adding some of the most Democratic areas from the 15th to the 14th, the 7th and 11th figure to be more Democratic in 2012. Republicans should have a 9-5 congressional advantage after 2012, but it’s not hard to see a more favorable Democratic delegation in a more Democratic year. Obama’s worst 2008 showing in any Michigan district was 48%.

Stabenow is vulnerable to a strong candidate and polled as a toss-up recently. The Republicans’ best hope for a strong showing in Michigan would be Michigan native Mitt Romney on the Presidential ticket in 2012.

Republicans don’t need a swing to the right as great as 2010, but with a rightward swing in the electorate they could win the state for the Presidential and Senate elections, and a big majority for congress. Of course, it could go all Democratic. That’s why they call it a swing state.

Thursday, December 16, 2010

House Voting

The charts below cover 111 House districts that have been competitive in the last four elections.

Wednesday, December 15, 2010

Swing States

The term "swing states" implies that a state could go either way in the Presidential election. Some of the states that were considered swing states during the Bush years certainly weren't in 2008. Bush won Iowa, but Obama won it by 9 points. Bush nearly won Michigan and Pennsylvania, while McCain wasn't competitive in either. Republicans romped in Michigan, Maine, North Carolina, Pennsylvania, Minnesota, Wisconsin, Indiana, Ohio, and New Hampshire this year. Obama won all nine in 2008. If they're swing states in 2012, the election is a toss-up. Below I have red states, blue states, and toss-ups. There are 16 toss-ups here. Whoever wins the majority likely will win the election.

Monday, December 13, 2010

Why Mitt Romney Already Has The Republican Nomination

It really is over. Some people haven't accepted it. But it is. I'm not a Mitt Romney fan, but I can't deny the facts. The important states are Iowa, New Hampshire, Nevada, and South Carolina, because they are going to happen in February. Iowa and Nevada are limited to those who participate in caucuses. History tells us that if a candidate doesn't win at least one of these, they're likely to drop out.

Iowa - There's a potential issue for Mitt with Republican activists, but his organization excelled at caucuses. He lost only two, Iowa and West Virginia. Despite having little appeal in Iowa, Romney managed to finish a strong second to Huckabee. I'd anticipate him doing at least that well. He knows how to do caucuses.

New Hampshire - Last time he ran up against McCain here. This time there won't be anyone who is popular locally. In a poll taken last month Romney got 40% of the vote, more than Palin, Gingrich, and Huckabee got combined. This is New England and should be a slam dunk.

Nevada - Mitt's win wasn't nearly as big here, but he was still ahead of Gingrich 34%-21%. Nevada is 7% Mormon. Nearly all of those are Republicans and they turn up heavier in a caucus. Romney could get more than 20% of the vote just with Mormons. So he really doesn't need a lot more help that that.

South Carolina may be troubling for Romney, but he may already have two firsts and a second or possibly three firsts. The field will narrow and he'll have momentum. He'll be playing downhill.

The Truman Projection

When I looked at the House last week, I looked at Presidents who were elected and then ran for re-election in order to compare apples to apples with Obama. New Presidents average 16.6 House seats above normal, while re-elected Presidents average 6.4. Since "normal" for Democrats right now is 208 seats, regaining the House would be less likely than Republicans maintaining control. It is, however, the best metric for Democrats, even though one of the 7 Presidents was enough above normal for the Democrats to regain the House.

There were several Presidents I excluded, because they didn't win an election before being re-elected.

Gerald Ford - His predecessor resigned amid scandal, hurting his party. That's not going to happen.

Lyndon Johnson - His predecessor was assassinated. That's not going to happen either.

Franklin Delano Roosevelt - Due to the uniqueness of the times, it's difficult to pin down what normal was. In 1936 he gained 12 seats to go to 334. In 1940 he gained 5 seats to go to 267. In 1944 he gained 20 seats to go to 242. This was the Great Depression, of course, and things didn't get that bad.

Harry Truman is the only President not initially elected who could be indicative. He served only a few months shy of a full term. He experienced mid-term elections and got slaughtered, giving the Republicans control of the House and Senate. Obama's mid-term drop was a bit larger than any of the others listed Thursday, but Truman's was the bottom falling out. Of course, Truman came roaring back and gained 75 seats, giving the Democrats 263. That would be a total more in line with a new President, not re-election.

Why Truman is a good projection for Obama:
1. He couldn't have been more popular than Obama was at his initial election. We know he was more unpopular than Obama was at the mid-term. So if Harry Truman can do it, surely Obama can.

2. He lost control of the House and was able to frame the "do nothing" Republicans as the cause for the country's problems. Obama also lost the House in his mid-term. Of course, Eisenhower and Clinton did as well. Neither regained it when they won re-election.

Why Truman is a poor projection for Obama:
1. In this era, normal was a significant Democratic majority. So while 263 was more than could be expected, the Democrats had a lot of gimmes which were going to come back to them no matter what. Republicans had 4 seats in West Virginia, 3 in Kentucky, 28 in Pennsylvania, 20 in Illinois, 19 in Ohio, and 14 in California. This was unsustainable. Normal now is Republicans in a majority.

2. Republicans control redistricting in much of the country. 1948 wasn't a redistricting year, so the GOP couldn't shore up their 1946 gains.

3. Republicans had been shellacked in the last 4 Presidential elections and hadn't had above 209 reps in 18 years. Now a GOP majority is the norm.

4. The Democrats had a "solid south" in those days. Republicans didn't have any seats in Texas, Arkansas, Louisiana, Mississippi, Alabama, South Carolina, North Carolina, Georgia, Virginia, or Florida. The Democrats had few seats that potentially could be competitive and the Republicans had to defend huge delegations in a few states.

A Truman projection is possible, but I consider it highly unlikely.

Wednesday, December 8, 2010

2012: Iowa (Second Look)

Last month I mentioned that despite getting 56% of the congressional vote the Republicans only had 2 of the 5 seats. I decided to draw a new map to find out what the impact would be. Iowa is drawn by an independent commission with importance placed on counties not being split between congressional districts. The map below attempts to draw the districts the way the commission might.

1st District - The new 1st retains almost all of its districts, losing only Jones to the 2nd. It grabs 11 lightly populated counties from the 4th and 1 from third.

2nd District - The 2nd district picks up seven 3rd district counties and one from the 1st.

3rd District - The 3rd retains Des Moines as its anchor but has to move west and pick five 4th district counties, including Story County that has 4th District congressman Tom Latham's Ames home.

4th District - The new 4th has all the current 5th's counties and picks up the remaining 11 from the 4th.

I took each county's congressional vote and re-assigned it as Republican and Democratic votes in the new configuration. These are the changes:

The 5th district is mostly the new 4th district, with all of Steve King's voters. He'd likely run in the 4th and actually expand his margin with new Republican voters coming in.

If the counties had been set up this way in 2010 Republicans would've won 3 of the 4 districts and come close on the 4th. Of course this was a Republican year. It's unlikely 2012 will favor the Republicans as much. The more important thing to look at is that the 1st, 2nd, and 3rd absorb the remainder of the 50,000 Republican edge, so each will be more Republican than it is now. They'll be harder for the Democrats to win than they would be if the map was untouched. Leonard Boswell and Tom Latham would both have their homes in the new 3rd, setting up a likely match-up. While this new district had more Republican votes in 2008, a majority of those will not have been in Latham's district. The "new Republicans" who voted for Brad Zaun should move to Latham.

I see this as being 3-1 Republican after 2012, possibly 4-0. At worst it'll be 2-2. In all those scenarios the Democrats lose at least one seat.

Tuesday, December 7, 2010

Rating the Pollsters

People on the left spent the year yelling that Rasmussen was really a Republican pollster that was influencing the election. I doubt the polls had any effect. People don't make up their mind based on a poll. There were races where all the pollsters were wrong. Did people not listen to those polls?

I took all the gubernatorial and senate polls published within three weeks of the election. That's close enough to be a pollster's best effort, considering early voting had started. There were 10 pollsters that polled at least 7 races, with Rasmussen leading the pack with 45 polls. They only failed to poll the Minnesota gubernatorial contest.

When it came to accuracy, the poll being within 4 points of the final result, the pollsters didn't do very well. This should be within the margin of error. I'd expect the pollsters to be within 4 points one way or another. Yet all of the pollsters except Quinnipiac and Mason Dixon were within the range of 47%-61%. Quinnipiac was better. Mason Dixon was worse. I wouldn't trust any of these pollsters for the next election.

When we set aside the spread and only look at the pollster getting the correct winner, the pollsters did better. Here 8 of the 10 pollsters were above 80%, with 5 of them being between 86%-90%.

When it comes to bias, the critics were correct. Rasmussen had a Republican bias, with their polls being 3.3 points more Republican than the final results. They were no more Republican biased than McClatchy/Marist, Reuters/Ipsos, or Fox News. Fox used a subsidiary of Rasmussen for their polls, however. Suffolk, on the other hand, had a 3.0 bias in favor of the Democrats. I'd normally expect the pollsters to have similar numbers on each side of the bias curve, but it seems they were the ones who were influenced by the narrative and adjusted their polls too far to the Republicans.

Mason Dixon and CNN/Time had the least bias. It's worth noting, however, that Mason Dixon was the worst pollster. Their polls were inaccurate, but they were as likely to be biased in favor of the Republicans as they were the Democrats. Three of their polls were biased an average of 6 points to the Republicans, while four were biased an average of 5.3 points in favor of the Democrats. I'll take a biased pollster who is more accurate over one that could be way off in either direction. At least I know how to adjust Rasmussen's polls.

Who was the best? Quinnipiac was accurate most often, did fairly well picking the correct winner, and didn't have a strong bias.

Thursday, December 2, 2010

2012: House

The big question is whether the Republicans will retain the House in 2012. The answer is almost certainly yes. Remember that I said it was definitely flipping this year and that the Senate wouldn't. Here's by the numbers:

Redistricting net impact: No net seat change
McCain PVI+: Republicans 228 Democrats 207
Bush 04 PVI+: Republicans 237 Democrats 198
Era Average: Republicans 227 Democrats 208
Obama wins Re-election: Republicans 220 Democrats 215
Obama loses Re-election: Republicans 237 Democrats 198

I don't see redistricting helping either party too much. The Republicans control a number of redistricting plans, but they have so many seats in those states that any redistricting will just make sure they don't lose seats. Republicans have a 19-6 advantage in Florida. The state will gain one or two seats. No one knows how Fair Districts Florida will impact the districts but it's safe to say that the GOP will have a difficult time retaining such an advantage. Even in a good year I could see Republicans ending up 17-10.

Texas is similar. The GOP has a 23-9 advantage. They'll have to help their most vulnerable members. To do so, they'll create a map that'll likely result in a 25-11 or 24-12 GOP lead. Despite control of the redistricting process in Michigan, Ohio, and Pennsylvania, Republicans are so overextended that they're likely to lose two seats in those three states. Republicans simply have too many seats concentrated in certain states to benefit even if they're drawing the lines.

McCain and Bush PVI are the number of districts that Republicans did better than +2.5 points with Bush and -6.4 points with McCain. Since 1994 the Republicans have averaged 227 seats.

People talk a lot about coattails. They aren't as strong in a re-election.

Since World War II new Presidents have averaged about a +16.6 from the expected number of seats. Obama's coattails were longer than anybody else has had. That could indicate he'll bust out of the expected number in re-election.

The coattails disappear in mid-terms. While it's been depicted that a President gets clobbered in mid-terms it's really just him losing the seats he won when he was elected. A return to normal. Not only did Obama do better when he was elected but he was the most below expected in mid-terms. So maybe the Obama magic isn't ever lasting.

When a President runs for re-election his coattails aren't nearly as long. If he wins his party ends up around 6.4 seats over expected. Reagan, Nixon, and Eisenhower all won re-election in landslides, doing significantly better than they did win originally elected. Their congressional totals averaged 8 above expected, compared to the 19 they had in the original election.

I'm almost certain the Democrats will gain seats in 2012. In worst case scenario above they're +5. In best case, they're +22. It's conceivable Obama is on the high end. If he equals Nixon the Democrats take the majority. Of course he also had the worst mid-terms too.

Wednesday, December 1, 2010

2012: Senate

Any way you slice it, 2012 looks like a tough year for Senate Democrats. They have 21 Democratic and 2 independents who caucus with them up for re-election. Even excluding the 2 independents, Democrats have a -11 margin of seats to defend. That's never been easy. Here are all the times since World War II one party has had to defend 8 or more seats than the other party has had to defend.

At first glance this doesn't look so bad. While there have been times when a party defending a significant number of seats has lost a lot, it doesn't happen all the time and the average is only 5.9. Six of these elections are mid-terms. Comparing a mid-term to a Presidential election is apples and oranges because the election and electorate are different.

The six Presidential years are actually better for the party with more seats than the mid-terms. Excluding the mid-terms favors the Democrats. Of course there are two Presidential elections which have circumstances that won't be repeated. In 1963, a popular Democratic President was assassinated. Strong sympathy for him prevented his party's losses. In 1976 the Republicans were coming off Watergate and couldn't take advantage of the differential. The other four Presidential elections have one thing in common, the Presidency changed parties. In 2008 and 1968 the Presidents were far more unpopular than Obama is today. The other two years, 1960 and 1980, the opposition candidate proved to be far more popular than he was thought to be even a year before. His strength led to his party's strong showing.

So let's approach this from a different perspective. How did the President's party do when the President faced re-election. Here I've excluded 1964 and 1976 because of circumstances and because that President had no initial election baseline.

This contains an interesting mix of Presidents. Eisenhower, Reagan, and Nixon were re-elected in landslides. Bush 43 and Clinton won closer elections. Carter and Bush 41 lost. It's difficult to say that Obama, Nixon, and Reagan underperformed in their election campaigns. While they didn't take more than the differential, they all made significant gains. If this were a new election, such a big differential wouldn't be as important.

As expected, almost all the Presidents underperformed two years later. It's tough to regard Clinton's +2 very highly because he lost 8 seats.

Reagan, Nixon, and Eisenhower didn't have the huge differential Obama has in 2012. Even their landslides were unable to prevent some losses due to a negative differential. Jimmy Carter and Bush 41 lost re-election and that hurt their party in the senate. Obama's coattails in 2008 only resulted in 20 Democratic wins vs. 15 Republican wins. If re-elections are at all telling, even a strong performance in re-election likely won't prevent a loss of 7-11 seats.

If you're looking for a more pleasant answer for Democrats, vulnerability is a good way to look. States were grouped based on how Republican or Democratic the state is. In 2010 Democrats only lost 75% of the most vulnerable states and half of the moderately vulnerable states. Repeating this they'd lose 8.2 seats. The Republicans, on the other hand, had 4 moderately vulnerable and 14 low vulnerability seats in 2010. In 2012, Maine and Massachusetts will be much better opportunities for Democrats than anything this year. Picking them up would limit the losses to 5 or 6.

My current analysis has the Democrats at a net loss of 5 seats. That'd be lower than these other historical indicators but still enough to cause them to lose the Senate.

Wait a second, a Democrat might say. If you count Lieberman and Sanders as Democrats the Democratic party went 24-9 in 2006. If these guys were good enough to win in 2006 why couldn't they be good enough again? Circumstances are far different. In 2006 you had an unpopular Republican President, an unpopular scandal ridden Republican congress, a disillusioned Republican base, and a war that energized Democrats. Obviously an unpopular President won't be a problem for Republicans. They only control one house of congress. At worst they're not likely to be that unpopular again. The Republican base seems unlikely to be disillusioned this time around. Democrats could once again be energized, but without the other factors winning anything close to 24 races is extremely unlikely.