Tuesday, May 31, 2011

The Evils of the Cook PVI

In 1997, Charlie Cook came up with a method of analyzing congressional districts voting patterns that'd enable people to compare districts across the country, the Cook PVI. The people rejoiced. The number used only Presidential data, thus taking out local candidates who might have partisan followings that'd skew a district in a way that wouldn't continue after he or she left office.

This was a good way to compare congressional districts across the country. Like the RBI in baseball the PVI tells a part of the story but not always the best story. So when Republicans in Illinois win districts with Democratic PVIs but Republicans in Tennessee don't, people assume it's because Republican congressmen in Illinois overachieve.

The problem is that the PVI's perceived strength, not letting one local politician skew the number, is also a great weakness. Chances are that Republicans in Illinois aren't the outlier. It's Barack Obama and John McCain in Illinois. McCain underachieved what other Republican statewide candidates usually get. The Republicans in Illinois aren't overachieving at all. Obama-McCain numbers don't accurately reflect local voting patterns.

A better way to judge how a party will do statewide is to look at how candidates for statewide office do. A better way to judge how a party will do in a congressional district is to look at how the party performs in that congressional district. Sure, one good candidate or a bad one can skew the numbers but chances are that this candidate is the one running again next year. Barack Obama won't be running for a congressional seat any time soon. Yet people cling to his performance in a district as being indicative.

Cook PVI isn't a bad stat. A Democrat is more likely to win a D+6 seat than an R+5. When looking at a district or a state you need to look at how Obama and McCain did compared to statewide candidates and how the congressional candidates do. As noted below, every incumbent Republican beat Barack Obama by 9.6 points or more in 2008. They can't all be good candidates running against bad ones. Chances are the number you're comparing their performance to, Obama-McCain 2008, is inaccurate.

Sunday, May 29, 2011

Actual Illinois Map

Last Monday I talked about Illinois redistricting and sure enough a map has come out.

The media immediately jumped all over it. I've seen predictions on the web that this is a 12-6 Democratic map, possibly even 13-5.

So let's see how these changes would've impacted Republicans in 2008:

At first glance it doesn't look good. Four districts Republicans currently hold would flip over, although one that's expected to flip IL-10 would not.

The problem here is that Democrats won those four districts in 2008. So increasing Democrats in their district won't change things. We need to look at those five freshmen Republicans. How will they do in 2012? The first question is how much the 6 Republican congressmen who ran for re-election improved.


So 2010 was worth an average of 6.7%. Now 2012 is unlikely to be as tough as 2008, even in Illinois, so this is probably worst case scenario. Adjusting for the drop off and the change in their districts we end up with:

Argh! It is ugly. Maybe we should write off IL-8, 10, 11, and 17.

Not so fast. Let's go back up one chart. There was one dramatically larger improvement, Aaron Schock. Schock was the only one who was a new representative. Just like these five. This isn't unexpected. Intuitively we know that an incumbent does better than someone running for an open seat. But how much better?

Let's go back to 2002-2004. They had very similar national congressional results. Republicans got 52.4% of the vote in 2002 and 51.4% in 2004. The President was running for re-election in 2004, same as he is this year. I'm going to stick with races that were competitive in 2002, since the races we're looking at all were. I've excluded a few that had unusual circumstances (e.g. the candidate ran unopposed in 2004, Texas redistricting). Here's what we come up with:


Well, that's all over the board, isn't it? We do have 36% of the elections where the incumbent goes up by 9.0% or more, but also 3 elections where the incumbent went down. On average, however, incumbents went up 6.3%. So let's take this +6.3% and deduct the 5.7% decrease that the Republicans (without Schock) got.

That's certainly much more interesting, isn't it? Here, every race but one is competitive.

There are a lot of things we don't know about 2012. Will Obama once again win 63.8% of the vote in Illinois? If he does worse statewide, he'll likely do worse in each district, resulting in the district being more Republican. Will these freshmen all run in their current districts? Will they have quality challengers? Are they the incumbent who does a lot better in their second election or one that does worse?

There are a lot of factors that'll go into 2012 results. These districts might be the easy wins for Democrats people are predicting. That's just not set in stone. The Democrats will have to hit something that right now looks like it has a 50%-60% chance of happening all four times. That's statistically very difficult.

Wednesday, May 25, 2011

Medicare Won't Factor into 2012

Medicare was an issue with yesterday's special election, causing the media and the parties to run with the idea that it'll be a big issue in 2012. It won't.

In the spring of 2005 Republicans took on Social Security reform. The effort generated a lot of opposition and was very unpopular. Yet it wasn't an issue in the 2006 elections. The issues were the Iraq War, Republican ethics scandals, and Hurricane Katrina.

In early 2007 the Iraq War was going badly and was a huge albatross around the neck of George W. Bush and the Republicans. Yet it wasn't an issue in the 2008 elections. The issues were the economy, the fiscal crisis, and George W. Bush's unpopularity.

In early 2009 there was huge opposition to the Democrats' stimulus plan. Yet it was only a minor issue in the 2010 elections. The issues were the healthcare bill and the back room shenanigans involved with it, the deficit, and the economy.

The factors that influenced elections were pretty much long term issues, usually those that became a law. Once Social Security reform was dead it ceased to be an issue, because no one felt the impact of a bill that didn't happen, and Republicans abandoned it. The Iraq War was badly mismanaged but by the time the 2008 election came around the war was going well. The stimulus drew a lot of opposition, but the bill's expenditures didn't make the news after that.

The healthcare bill was in the news for nearly a year and once it was enacted into law the Republicans could use the bill as a weapon, even if their claims weren't true. People knew something was coming with the healthcare bill. Nothing was coming with Social Security reform.

Other big issues, Iraq, the economy, President Bush's unpopularity, lasted a long time and were issues that wouldn't go away all the way up to the election.

In 2010 Democrats tried to use Social Security as a weapon against the Republicans ("they're going to take your Social Security away!"), but it was an issue the Republicans weren't pushing and wasn't in the news. It likely didn't register as an issue, especially compared to things that the public could tangibly see every day.

If the Republicans abandon Medicare reform or go with a smaller change that isn't controversial, there'll be nothing to use against them. The 2012 election will be fought over issues that people will experience throughout 2012. The issue was big now because the Ryan budget is fresh. By November 2012 there'll be another budget and other big issues. They can only ride this one out if Republicans keep shooting themselves in the foot.

There is an issue that might have legs, however. The bills in several states that limit union collective bargaining rights could be a rallying point. The differences here are that these are laws that passed that people will actually feel or anticipate feeling in the future.

CA-36 and NY-26 Insights

We've had two special congressional elections in the last two weeks, neither of which tell us that much about things we want to know.

California-36 doesn't give a big clue to the political environment because the 56.1%D-41.4%R is about what we'd expect in this district. We did learn that Republicans will vote for a Republican candidate in California's new top two primary system, even in a district they are perceived to not have much a chance. One theory is that Republicans would vote for the more centerist Democrat, giving them the better choice in the final result.

It's not surprising that they didn't. People don't come to the polls with clever strategies. They vote for who they like, even if other people think they have no shot. This is very evident in that so many candidates received votes.

The authors of Prop 14 envisioned moderates making the final vote. In the final vote the moderates would vote for the moderate, regardless of party, and the party who didn't have a more party line candidate would also. The winners, Hahn and Huey, are seen as a liberal Democrat and a conservative Republican. This insight may prove to be true, but we won't find out now.

The right leaning Libertarian Party got 1.4%, the left leaning Peace and Freedom Party got 0.5%, and independents got 0.6%. These totals are insignificant and well within the percentage that could shift from a primary to a general election. California has produced a number of elections where the right leaning or left leaning party got 4% or more of the vote. A much higher total will tell us what the third party voters will do.

If the final vote varies much from an expected 57.5%D-42.5%R it won't tell us a lot other than one candidate ran a stronger campaign.

NY-26 produced an upset victory for the Democrats. It was, however, only a few percentage points above what I'd expect a Democrat to get. While the "Tea Party" candidate certainly didn't siphon all his votes from the Republican, the Republican could've won without him. Corwin was unpopular and ran the weaker campaign.

One conclusion people are drawing is that it's not 2010 any more and the Democrats aren't still unpopular. That shouldn't be a surprise to anyone familiar with elections. A party is punished when they are in the majority. The people did that last November. Democrats are no longer in the majority and voters no longer are voting with anger. Anyone thinking that America turned sharply to the right was kidding themselves.

These elections don't mean the Republicans will be punished in 2012 and they'll lose the House. There remains only about a 20% chance they'll lose the House. Republicans control only the House and voters can't put all the blame for whatever anger they have on them. The President attracts most of the anger when there is some and the President isn't Republican.

When a President runs for re-election it rarely produces a significant congressional change, even when the House is in the hands of the opposing party. The MSM story is that 1996 was a rebuke of the Republicans due to Clinton's win. Yet the GOP only went from 237 House seats before the election to 228 after it. That was most likely a regression to the mean. Republicans averaged a 228-207 edge during this period and they got that.

Likewise neither 1984 or 1972, Republican Presidential landslides, produced much movement in the House. In both cases Republicans rose a little above the mean, as the President had short coat tails. Republicans should lose 10-15 seats in 2012 to regress toward the mean. In such a case a 25, now 24, seat loss would be unexpected but not that unlikely. Thus, my 20% figure.

Redistricting may put the mean higher, however, meaning that instead of a possible 215 Republicans with a mean of 228 there might be 228 Republicans with a mean of 240. And that's if Obama wins. Of course, a Republican winning the Presidency would likely push Republican totals higher.

There were several special elections in 2008 that confirmed what we'd already learned from 2006 and George Bush was even more unpopular. Neither total is more than a few points from what you'd expect, not the major swing in those elections. Democrats lost two special congressional elections in 2009 and still got clobbered in 2010. So setbacks in New York don't have much meaning.

The Democrats know their voters are no longer asleep, but again that was expected either.

Monday, May 23, 2011

When Gerry Meets Dummy in Illinois

The Illinois congressional redistricting maps are expected to be finalized in the next week and Democrats are licking their chops expecting big gains. They have to be careful, however, as a district they may think is Democratic might not actually be. Republicans won 7 congressional seats in 2008, the year Obama was taking 63.8% of the vote in Illinois. Of those 7, Obama won 4 of them.

Six of those were won by incumbent Republicans, while the 7th was an open seat. Democrats took one open Republican seat. In fact, Democrats didn't beat an incumbent Republican in either 2006 or 2008, while Republicans knocked off four Democratic incumbents in 2010.

The chart above shows that Obama only beat the average Illinois Democrat by 2 points. The problem with that is that it includes districts where the incumbent democrat won over 85% of the vote four times. It'd be difficult for Obama to beat those. It also includes an estimate for what Phil Hare would've gotten had he not been unopposed.

If we exclude districts Democrats hold now, and ones Republicans likely won't compete in, the average margin a Republican did better than McCain did was 6.4%. If we further narrow it down and eliminate the three districts Republicans won in 2010 but didn't win in 2008. This includes CD-17, where they didn't even offer a candidate. The average Republican did 12.7% better than McCain did, with the worst of the eight doing 9.8% better. On average McCain got 47.2% in their districts, while they got 59.9%.

If an incumbent Republican does at least 9.8% better than the Republican nominee, then it's likely that any district the GOP currently holds and Obama gets less than 59.9% is one the Republicans will win. Peter Roskam is in a suburban Chicago district that the Democrats would like to get. Since he finished 14.2% ahead of McCain he might be favored in any district Obama got 64% or less of the votes.

IMO, Republican incumbents should be safe in 2012 in any district that's drawn that Obama got 56% or less, feel confident in districts he got 57-59% and be competitive in districts Obama got 60%-64%. If the Democrats make a lot of districts where Obama won 58% of the vote, they will likely lose them. People will split their tickets in Illinois and vote for the congressional candidate they like who is a Republican and the President from their home state.

Wednesday, May 18, 2011

California 36th Congressional Election.

Everyone is shocked by Craig Huey's performance yesterday.

Republicans were beaten 57%-41%. That's about what I expected. It was just a question of how it'd shake out. Huey went hard at the convention and at the RPLAC meeting. He was the only one seriously spending on the GOP side. He had endorsements from everyone from David Dreier, who talked about Huey on the Hugh Hewitt Show, to Tony Strickland. He's well known in the Torrance business community and Torrance is the business center of this district. I was expecting Huey to lead the GOP with about 17-19% and finish no worse than 2nd with 13-14%. He exceeded my expectations, but not by much.

Saturday, May 14, 2011

Huckabee's Out

Mike Huckabee's exit makes things a lot easier for the other candidates. I still maintain Mitt Romney is the frontrunner.

All the polls have him way up in NH. It's his backyard. He blew it in 2008 because McCain is very popular there, but he's learned the lesson. Romney had a great organization last time around and it may be better this time. Romney and Scott Brown are old friends and Brown will campaign hard for him there. It won't be close.

He wins Nevada in a landslide because the caucus voters will likely be 25% Mormon. If Romney can get a second again in Iowa, he can afford a weak finish in South Carolina.

Then it's a question of what states are next. There are still 13 states that violate party rules in January and February, although it looks like Alabama, California, Connecticut, Delaware, Hawaii, Wisconsin, and New Jersey are going to move. New York hasn't introduced a bill to move, but they haven't said they want to go early.

That leaves Minnesota, Florida, Arizona, and Michigan before March 6. Mitt was born and raised in Michigan and should have a lock on that primary as well. He'll push hard for Michigan to come before Super Tuesday.

With the way the calendar is likely to be, Romney will be the front runner.

Tuesday, May 10, 2011

Generic Ballot

A new CNN/Opinion Research poll has the Democrats with a 4 point edge in the generic ballot. Already a few Democratic friends are imagining their party taking back the House.

The problem is that this assumes that a 4 point lead in the generic ballot will produce a 4 point win on election day. That's unlikely.

The poll is with registered voters, not actual voters. Democrats tend to do better with registered voters than likely/actual voters. Blacks, Hispanics, Asians, and young people vote below their share of the electorate, even in Presidential years. The registered voter polls will likely overstate Democratic performance. Democrats were leading in the generic ballot by 12-14 points for most of 2007 into 2008. That number dropped when they went to the likely voter model, before going up after the fiscal crisis.

Even after the likely voter model is used Democrats tend to underperform it. In 2004, the generic ballot was a tie, but Republicans won by 2.5 points. In 2006 Democrats were ahead by 11.5 points in the generic ballot and won by 7.9.

The next factor is that the Democratic percentage of the vote hasn't been reflected in their seat total. After each election from 1996 to 2006 Democrats had a lower percentage of seats than their total House vote. In 1996 they even took a majority of the vote, but still managed only 206 seats.

One explanation for this is that Democrats have been packed in a smaller number of congressional districts during redistricting. Currently there are 234 Republican leaning districts, but there are only 192 Democratic leaning districts. With Republicans controlling redistricting, there likely will be more than 234 Republican leaning districts. Republicans won't have to do much to retain at least 218 seats.

Sunday, May 8, 2011

California Congressional Races 2012

The conventional wisdom seems to be that the Democrats will pick up several congressional districts, even if the districts are created fairly.

Bull dung.

These people are looking at the McCain-Obama 2008 race and plugging in those numbers into new districts. The average California statewide race breaks down as 55% Democratic - 45% Republican. Obama beat McCain 62%-38%. McCain was 12% below 50%, while you can expect a Republican to only be 5% below, however, and that's without adjusting for the advantage Republican incumbents have.

If McCain-Obama isn't representative, what is? I chose Whitman-Brown. Brown won the race 58%-42%. That's still above what Republicans will likely do, but it's closer. So these numbers are likely conservative estimates.

Orange County - Whitman won the county 60%-40%. Whitman won 32 of the 35 Orange County cities and won nearly 70% of the unincorporated areas. The only significant Democratic stronghold is Santa Ana. That means that you'd have to gerrymander Orange County heavily for the Democrats in order to get a Democratic leaning district. Republicans should pick up a seat here. Loretta Sanchez is a goner.

San Diego County - San Diego could go a bit better for the Democrats. Republicans now have a 3-2 edge in the congressional delegation, but there likely will be two safe Republican districts, two likely Democratic districts, and one that could be a toss-up.

Inland Empire - Democrats have only one seat in San Bernardino-Riverside and they should have a shot at several districts. The good news for Democrats is that they lost Riverside County by 8 points, but San Bernardino by only 1. The problem could be that Democratic strongholds (e.g. Ontario, San Bernardino, Fontana) are all close to each other. Redistricting could break badly if, as I suspect, too many of these cities are in one district.

Los Angeles County - The good news for Democrats is that every Republican seat could be in play. David Dreier might have it especially tough, but as long as San Dimas, La Verne, and Claremont are in his district he'll have a shot. If both Pomona and West Covina don't end up in his district he may have more than a shot. Buck McKeon and Elton Gallegly could be seriously challenged too. On the Democratic side, they could have trouble in a district that spans San Gabriel Valley cities like Hacienda Heights and Diamond Bar, especially if they're paired with Orange County cities immediately to the south.

Central California - The story is better for Republicans here. The Central Valley is so Republican that it takes a gerrymander to get them seats. In a more balanced redistricting both Jim Costa and Dennis Cordoza would be in trouble. Like Dreier they'd still have a shot. On the coast Lois Capps should end up in a competitive district, albeit one she's favored in.

Northern California - Not much change should be expected here. Dan Lungren should once again be in a competitive district for the Republicans while Jerry McNerney should be in one for the Democrats.

My mapping gave Democrats 24 safe seats and 4 likely seats. Republicans had 7 safe seats and 3 likely seats. That puts 15 seats as being very competitive, although I expect more to lean Republican than lean Democratic. In a good year the Democrats could certainly pick up 8 or 9 seats, but in a good Republican year the GOP could pick up 6. In a neutral year I'd put the breakdown at 34-19 or 35-18, a pick up of 1 seat for the Democrats.

We won't know until the districts are finalized but don't count on a map that'll let the Democrats easily pick up 3 or 4 seats.