FiveThirtyEight: Politics Done Right

For the Soccer Fans Among You

noreply@blogger.com (Nate Silver) — Thu, 12 Nov 2009 01:13:00 +0000

Much of my summer was spent on a consulting project that I did in conjunction with ESPN, in which I helped them a design a soccer ratings system known as the Soccer Power Index (SPI). SPI launched today and we're pretty proud of the results, which feature a combination of intuitive (Brazil and Spain are #1 and 2, natch) and somewhat bolder rankings (SPI is fond of 'second-tier' South American teams like Chile, Uruguay, and Paraguay, as well as African up-and-comers Ivory Coast). The United States is ranked 14th.

Unlike other soccer ratings systems, SPI is explicitly designed to be predictive -- so a team like Argentina, which in fact struggled to qualify for the World Cup, won't be penalized that much provided the system is convinced that the talent is still there. The two main innovations in the SPI are to incorporate results from club play -- if Cameroonian striker Samuel Eto'o scores a goal for Inter Milan, it will (marginally) help Cameroon's rating -- as well as to incorporate a "competitiveness coefficient" based on the actual lineups that each team used in each match. The latter is important because international soccer clubs play a lot of matches -- friendlies, some second-tier international tournaments -- in which they're essentially sending their taxi squads in, which tell us very little about the teams that will actually be on the field in South Africa next year.

Anyway, this is a politics blog -- not a sports one -- so I'll direct you over to the very, very, long article on methodology I did at ESPN.com if you're curious about the details -- or check out the Wall Street Journal's Carl Bialak for a good capsule summary.

House Handicapping One Year Out

noreply@blogger.com (Tom Schaller) — Wed, 11 Nov 2009 16:57:00 +0000

One year out, how many losses might we project for Nancy Pelosi’s House majority? Which Democrats are most vulnerable? Piggybacking on the post I wrote last week after attending the latest Cook Political Report political briefing at their Watergate offices, I'll take a look the analysis by CPR’s David Wasserman, and perhaps tweak it a bit.

Wasserman identified five factors--in chart above, the five columns starting with "CT" for "cap-n-trade"--that might be indicators of vulnerability for House Democratic incumbents next year:

1. Did they vote “yes” on the so-called cap-and-trade bill?
2. Is their district’s Partisan Voting Index—a measure produced for CPR by Polidata which reflects the partisan performance of the major parties’ presidential candidates during the 2004 and 2008 cycles—scored at D+5 or less?
3. Is their district’s PVI R+5 or more?
4. Was their 2008 popular vote 55 percent or less?
5. Does their opponent have at least $100,000 cash on hand as of 9/30/09?

Based on those criteria, there are 28 Democrats who match on at least four factors, including one--Maryland 1st District rookie Congressman Frank Kratovil--who matches on all five. Though these are not necessarily the 28 Democrats Wasserman rates as most endangered--there is some but not perfect overlap between these 28 and the 12 Democratic-held districts he rates as "toss-up" races and another 21 he rates as "leaning Democratic"--many could be in serious trouble next November. (Kratovil, I should note, has at least one compensatory advantage over the others: The Democratic Campaign Committee Chairman, Rep. Chris Van Hollen (MD-8), has a vested interest in defending a seat in his home state.)

Wasserman included the cap-and-trade vote, he says, because voting for it may be particularly dangerous for Democrats from districts containing significant chunks of those white working-class areas depicted in that now-famous map of the 22 percent of counties nationally where John Kerry in 2004 outperformed Barack Obama in 2008. "In the rural areas of the South—the parts of the South were Barack Obama did worse than John Kerry in 2004—there is an open revolt against Democrats and particularly Barack Obama," said Wasserman. He cited as specific evidence of this revolt from last week the coal-heavy and thus highly cap-and-trade resistant 3rd Delegate District in the southwest corner Virginia, the Appalachian county that uniquely borders both Kentucky and West Virginia. Mark Warner got 61% of the vote in 2001 in District 3, Tim Kaine dropped to 49% in 2005, Obama slipped further to 40% last year, and Creigh Deeds plummeted to just 32% last week. "In places like this all across the South," Wasserman quipped, "as Toby Keith would say, ‘the fit’s gonna hit the shan.’ ”

For kicks, I decided to add two factors/columns to Wasserman's analysis. The first is for the recent vote on the House health care bill ("HC"), on which 39 Democrats voted nay. (A good list and analysis of the 39 can be found in the New York Times.) Those Democrats are explaining their votes in a variety of ways, and of course a "nay" vote may be the potentially more electorally damaging for many Democrats. But, for the sake of argument, let's presume that voting for the House bill is an added risk factor for 16 of the 28 who voted "aye."

I also added a column showing the comparative cash-on-hand advantage (or in some cases disadvantage) for each of these 28 incumbent Democrats ("Ratio"). I calculated this as cash-on-hand minus outstanding debts as of September 30, 2009 (data courtesy of the Center for Responsive Politics here), and then created an incumbent-to-challenger COH ratio by dividing the incumbent's COH by the challenger's. This relative money advantage is useful because (a) it contextualizes the financial competitiveness of challengers who have cleared the $100K bar relative to the incumbent they aim to unseat, as least as of 9/30/09; and, relatedly, (b) because the party committees (DCCC, NRCC) look at fundraising competitiveness as one indicator of challenger strength, the smaller this ratio is for the incumbent the smaller still it could become, as the RNCC and strategic givers gravitate to races they believe will yield better returns on their investments. For lack of a better cut point, I counted as another risk point any Democrat who does not presently enjoy a 3:1 or better COH ratio. (I rounded up OH's Boccieri ratio to 3; all others below that gained a point.)

I'm not quibbling with Wasserman's analyses. Maybe the health care vote will not be very predictive next November, and surely there will be some well-funded incumbents who lose and others who win despite well-funded challengers. I'm just updating the analysis to include one new policy factor and one updated political factor. One could also set different (wider?) thresholds for the PVI ratings, and so on.

In any case, the way I've done it stratifies these 28 into three groups, including those who may be most at risk among the at-risk: the ones score a "6." (Kratovil was the only one with a chance for a "7," but he voted "nay" on the health care bill.) There are not a lot of members from those Appalachian-area districts who voted for the health care bill--which is ironic, given that some of them represent districts with high uninsured rates, but let's not even go there right now. NH's Carol Shea-Porter and northern Virginia's Gerry Connolly are good examples members who may be helped by voting "aye." On the other hand, those who are from such areas and voted "nay" may still face money problems, like AL's Bobby Bright and MS's Travis Childers.

We'll just have to wait a year to see how much these key votes and the candidates' finances matter come 2010 midterm Election Day.

Is the Prognosis For Health Care Getting Better or Worse? The Market Weighs In.

noreply@blogger.com (Nate Silver) — Tue, 10 Nov 2009 19:36:00 +0000

It's a bit hard to assess where we are in the health care debate. On the one hand, the Democrats pushed through and passed a bill in the late hours of Saturday evening, clearing a hurdle that meaningful health care reform has never before cleared. On the other hand, the vote in the House perilously close, there were new complications introduced by the Stupak amendment, and the bad jobs numbers and arguably the outcomes of the elections last Tuesday Virginia and New Jersey last will give nervous lawmakers plenty to worry about.

One leading indicator of the prospects for health care reform so far has been the performance of the half-dozen or so publicly-traded health insurance company stocks. Favorable developments for health care reform have been met with decreases in the prices of these stocks, and unfavorable developments with improved valuations. So what's happened to these stocks since market close on Friday -- before the House passed its health care bill?

Well, not much. Actually, that's not true: they've gained an average of 1.8 percent so far, weighted for market capitalization. But the S&P 500 has gained 2.1 percent over the same day-and-a-half of trading, meaning that the performance of the health insurance stocks is right on par with the market. The news has not been good or bad so much as indifferent -- which feels, intuitively to me, like about the right assessment.

Interestingly, the markets had a much less ambiguous assessment of the elections, with health insurance stocks gaining an average of 3.7 percent over the course of the day on Wednesday while the S&P -- after a volatile session -- was basically unchanged:

I don't necessarily think that the stock market is particularly adept at forecasting political risk, but to the extent that you're looking for an objective assessment of the consequences of lats week's results for the Democratic agenda, this is a pretty decent one.

Many Previously Pro-Choice Dems Voted for Stupak Amendment

noreply@blogger.com (Nate Silver) — Mon, 09 Nov 2009 22:52:00 +0000

When I first learned that the Stupak Amendment, which would prevent abortion from being covered under health care plans included in the health care insurance exchanges, had passed by a 240-194 margin in the House, I assumed that something like the following happened: anybody who was either pro-life or who disapproved of the health care bill in general had voted for it.

Certainly, that description is apt for the half-dozen or so generally pro-choice Republicans, all of whom voted both for the Stupak Amendement and against the health care bill. But it doesn't really hold for the Democrats.

Rather, I was surprised at the number of Democrats who have solid pro-choice voting records but who nevertheless voted for Stupak Amendment. And the vast majority of these Democrats voted for, not against, passage of the underlying health care bill.

The below chart lists the 'yea' votes on Stupak among those representatives who had a rating of 67 in 2007-08 according to Planned Parenthood, and a rating of 33 or lower according to the National Right to Life Committee. (Note: no freshmen representatives are listed on this chart as they have not been rated yet.)

As you can see, 17 of the 20 Democrats who fell into this category voted for final passage of the health care bill. So what gives?

I'm sure there are idiosyncratic explanations in a number of cases, but I take this as a sign that they're worried about the re-election environment they'll face in 2010. 11 of the 20 pro-choice Democrats who voted for Stupak reside in districts that are rated as vulnerable according to Cook Political (note: candidates who are leaving the House to run for Senate or governor are rated based on those races instead). And, interestingly, they seem to think that a pro-choice vote would render them more vulnerable than a pro-health care vote, even though the pro-choice position is generally more popular than the health care bill on the table at the moment (although some recent polls have shown the pro-choice position losing ground).

Certainly, on health care, some of this may be a consequence of the logic that James Carville and others have espoused: Democrats know -- or believe -- that they'll be damned if don't pass a health care bill, so why not take the chance that things will turn out OK if they do? But there may also be something more here. Whereas the pro-life (anti-choice) movement is very well organized and has a long history of delivering votes, the anti-health care movement is somewhat disjointed, seemed to be limited in its electoral reach in NY-23, and carries a lot of baggage -- Glenn Beck, Michelle Bachmann, town hall screamers, and the like. And it may also be revealing of how they perceive their own base: whereas health care is a sine qua non for most Democratic base voters, they seem to be betting that the pro-choice position might no longer be.

Republicans Far Behind on Women Legislators

noreply@blogger.com (Renard Sexton) — Mon, 09 Nov 2009 17:30:00 +0000

On the front page of Politico this morning, a message of warning to congressional Republicans:

The gist of the article was as follows -- as the Republican party has pushed out moderates, it has tended to appear more hostile to women candidates and congresswomen. Debbie Wasserman Schultz (D-FL) was quoted as saying that the GOP "is a party that doesn’t respect women, a party that doesn’t believe women are equal to men," while former Republican conference chair Deborah Pryce bemoaned the loss of moderates in general -- many, she maintained, being women.

The article cites the major difference in proportion between Republican and Democratic women, with women making up nearly 23 percent of the Democratic caucus in the House of Representatives, while less than ten percent of Republican congresspeople are women. In the U.S. Senate, the proportions are just about the same.

However, truth be told, the problem identified in the Politico article is quite a bit broader than simply the GOP's turn to the right in recent years.

The US trails behind OECD allies other than Japan and Turkey, often by a significant margin.

However, when we compare the Republican Party with other conservative parties and coalitions in the OECD DAC (wealthiest democracies in Europe, North America and Japan) -- the case studies are bolded above -- there is a fascinating result.

Let's start with Sweden, the highest OECD country on the list.

Both the governing centre-right coalition in Sweden, led by the Moderate party, and the left side of Sweden's political spectrum have extremely high representation by women in the Parliament. The range is fairly high, with the farthest left party ("the Left") having more women than men, while the conservative Centre party has just 38 percent women.

In Germany's Bundestag (using 2005 data, since current data on gender and party was hard to come by), again the break between left and right is quite significant, with a 17 point gap separating the two coalitions, though at this time the government was run by a grand coalition between the SPD and CDU/CSU.

Japan's Diet reflects the recent strength of the Democratic Party of Japan, which just months ago thrashed the traditionally powerful Liberal Democratic Party. The rather low percentages of women in both camps reflect Japan's low position on the global chart -- 97th overall.

In order to compare apples to apples in this analysis, the real question is how a party fares within its own political system, as compared to other parties within their systems. So, for example, comparing U.S. Republicans to the Swedish Liberal Party or the Moderates would tell us more about the difference between Sweden and the U.S. rather than how close each party is to the center of their electorate.

Comparing U.S. Republicans to U.S. Democrats, it turns out that Republicans (the right of the U.S. in this model) are by far the lowest in terms of proportional representation of women legislators within their political system, as compared to other conservative parties and coalitions

In other words, as a proportion of the Democrat's percentage, the Republican Party has the lowest percentage of women legislators, as compared to the shares of the conservatives in the other three countries. In fact, as the percentage of women parliamentarians in Germany rose in the 2009 election (from 31.8 percent to 32.8 percent, with 9 new women) in an election with a victory from the center-right, the gap is likely even larger than this chart suggests.

The U.S. political right, namely Republicans, has a long way to go to modernize and moderate towards the center of the American political spectrum on gender issues, something we have know for quite some time. A turn towards a hard conservative line since 2006 has only amplified the gap. Perhaps the rising influence of moderate Republican women like Olympia Snowe and Susan Collins, signals the importance of this change for the future of the party, but no such high profile individuals have yet been spotted on the House side. Indeed, the flogging of Dede Scozzafova in NY-23 indicates that those who might want to step into this role could be quickly pushed out by conservative activists.
---
* A star signifies that a party is in the majority or in the governing coalition

Renard Sexton is FiveThirtyEight's international columnist and is based in Geneva, Switzerland. He can be contacted at sexton538@gmail.com

Real Oklahoma Students Ace Citizenship Exam; Strategic Vision Survey Was Likely Fabricated

noreply@blogger.com (Nate Silver) — Sun, 08 Nov 2009 12:30:00 +0000

In detailing some of the evidence against Strategic Vision LLC, a pollster I am now almost certain is disreputable and fraudulent, I pointed in particular to a poll that they conducted on behalf of the Oklahoma Council of Public Affairs, an conservative-leaning educational thinktank. The poll purported to show that Oklahoma's high school citizens were deficient in some of the most basic aspects of citizenship. Only 23 percent of them knew that George Washington was the first president, the poll claimed! Just 43 percent knew that the Democrats and Republicans are the two major political parties!

These conclusions seemed dubious to me on their face. Several years ago, at my old consulting job, I participated in a project for the State of Ohio's public schools which involved sitting down in a third or fifth grade classroom for the better part of a day and seeing how the students were learning. Most of these observations took place in poor, post-industrial towns, which were still suffering the effects of the steel mill or the axle plant that had long ago left town. What struck me, most of all, was how smart the kids were, relative to my expectations. These kids might not have been the highest achievers -- but I'm pretty sure that more than 90 percent of them would have known who George Washington was. And these were third and fifth graders.

There were other hints too, that Strategic Vision's poll may have been fake. The scores that Strategic Vision claimed the kids had gotten, for instance, were strangely underdispersed. And they seemed to contradict results from Oklahoma's own standardized testing, which asked much more difficult citizenship questions and found most of the students doing just fine.

It turns out that I was not the only person who had doubts about the survey. So did Ed Cannaday, the State Representative from Oklahoma's 15 House District.

In a telephone interview yesterday, Cannaday told me he was shocked when he heard of the results, which had received widespread media attention. "When I saw the statistics, I was just flabbergasted and said it cannot be true," he told me.

There were two items in particular that sent up warning flags for him: the one claiming that only 23 percent of the students knew the identity of George Washington, and another that claimed that about one in every ten students had listed the two major political parties as "Republican and Communist".

"Given the dialog of today, if they had said Republican and socialist, then maybe," Cannaday told me. "But communist -- that's just not something that you throw out there any more. I don't think Sarah Palin even used that term."

Cannaday, age 69, would be in a position to know. Before entering the State Legislature three years ago, he had spent decades in education, first as a teacher in a large public school in Kenosha, Wisconsin, and then in Oklahoma where he set up an alternative school. After a stint in private business, Cannaday returned to classroom, first as a teacher and then as a principal, and then -- finding he missed the one-on-one interaction with his students -- as a teaching principal at a small school in House District 15. He now serves on the House's education committee in Oklahoma City, and continues to pay regular visits to the schools in his district. "Most schools like to have me once a month," he says, to talk about legislation pending before the state.

Cannaday therefore had little difficulty setting up an experiment: he arranged to have all the seniors in the 10 secondary schools in his district take the Strategic Vision/OCPA survey. Cannaday tried to replicate the Strategic Vision survey to the greatest extent possible. The same exact questions were used, and as in the case of the original survey, the answers were open-ended rather than multiple choice. The survey was administered to a total of 325 seniors, including special education students.

Cannaday's survey however, found his students doing just fine: They answered an average of 7.8 out of the 10 questions correctly. By comparison, the high school students that were purportedly surveyed by Strategic Vision had gotten just 2.8 out of the items correct. 98 percent of the students on Cannaday's survey -- not 23 percent -- knew that George Washington was the first President. 81 percent -- not 14 percent -- knew that Thomas Jefferson had written the Declaration of Independence. 95 percent -- not 43 percent -- knew that the Democrats and Republicans are the major political parties. There was just no comparison between the two.

Cannaday distributed his results via e-mail to the constituents on his mailing list, including Karina Henderson, who published his findings in a dairy at Daily Kos. He also sent hard copies to each of the schools in his district, as well as all of Oklahoma's state legislators. The reaction so far has been entirely positive -- "even from the Republicans," said Cannaday, a Democrat.

Cannaday also sent his results to OCPA, the thinktank that had commissioned the survey, but has yet to receive a response. In October, before the results of Cannaday's survey had surfaced, OCPA had told the Oklahoma Gazzette that they were taking "a closer look at the raw data and the methodology,” behind the Strategic Vision survey but were not yet ready to "toss out" the results.

House District 15 is generally quite representative of Oklahoma, especially its Eastern portion, but is somewhat poorer than the state as a whole. "Rural" was the first adjective that came to mind when I asked Cannaday to describe his district -- no town has more than 3,000 people. Most of the residents make their living in the natural gas industry, commute to service-sector jobs in the comparatively large towns of Muskogee, Oklahoma or Fort Smith, Arkansas, or are engaged in what Cannaday calls "cow/calf operations". The five counties that make up the district range from middle-class to impoverished. Haskell County, for instance, where the town of Stigler is located, has one of the highest unemployment rates in the state and one of the largest proportions of its students on free and reduced lunch programs, the preferred benchmark of socioeconomic status in public education. House District 15 has no private schools.

Cannaday is proud of the achievements of his students -- particularly their low drop-out rate, which is about five percent, and their success in the state's mock trial tournaments, where they've frequently finished in the top 5 in the state competing against much larger schools. He has seen his students become doctors, attorneys, optometrists and accountants, he told me. "Any time you can have one of your former students in your district who's on speed dial in Oklahoma City as a physician, that's not too bad," he said.

But the schools in House District 15, which sends 40-50 percent of its students to college and sees 20-25 percent compete it -- are not exceptional in any obvious way. The students at Haskell High School, for instance, received below-average scores in 5 of the 7 categories tested by Oklahoma's standard exam, including in U.S. History.

There is no reason to think, in other words, that the students in House District 15 should have gotten such profoundly superior results to the "students" in Strategic Vision's survey. Nor could Strategic Vision's results have been the result of any sort of mathematical or methodological oddity. Consider their claim that literally none of the 1,000 students they surveyed were able to answer more than 7 of the 10 questions correctly -- lower than the average score achieved in Cannaday's test.

There are, rather, only two possibilities. Either the Strategic Vision survey was entirely fabricated -- or Cannaday's was.

I would put every dollar to my name on Cannaday, who has kept the surveys and is happy to show them to them to anyone who comes asking.

Next week is Celebrate Freedom Week in Oklahoma, with public schools students to be taught from a special curriculum highlighting the Declaration of Independence. "If were going to be pass education reform then we need to be out in the classroom demonstrating it," Cannaday said of his fellow legislators. "I will be in a classroom Friday," he told me. "I enjoy it."

I e-mailed to David E. Johnson, the CEO of Strategic Vision, a draft of this article and asked for any comments. The entirety of his comments were as follows:

"Thank you for the opportunity to respond. Our company did survey the Oklahoma students grades 9-12 for the Oklahoma Council of Public Affairs. Our client has all of the raw data, cross tabs, methodology from the survey."

Johnson did not reply to a second e-mail asking for his interpretation of the substantial difference between his results and those found by Cannaday.

House Likely to Pass Health Care Bill Tonight

noreply@blogger.com (Nate Silver) — Sun, 08 Nov 2009 00:31:00 +0000

I don't like to make non-substantive posts such as this one. But the comments in the previous article are getting a bit unwieldy, so I'll leave this up as a placeholder for anyone spending their Saturday night watching C-SPAN. The U.S. House is debating health care tonight and is expected to pass it. A previous 538 analysis had forecast that a bill much like the one under debate tonight would get about 222 votes. It looks like Pelosi may do a bit better than that, thanks to a sort of Faustian bargain on abortion. On the other hand, if she has votes to spare, she may will release some of her more vulnerable members, so anything from the bare minimum of 218 up to the mid 230s seems possible.

UPDATE: (12:54 AM) The bill passes, 220-215; roll call can be found here. Our original projection wound up being really good -- we'd projected 220 Democratic and 2 Republican yeas; in fact there was 219 Democratic and 1 Republican yea. Although, for a variety of reasons, this was a relatively easy projection to make.

Unemployment Hits 10.2 Percent

noreply@blogger.com (Nate Silver) — Fri, 06 Nov 2009 15:11:00 +0000

Oops. Back in August, I said that I thought there was only about a 1 in 3 chance that unemployment would break the 10 percent barrier. I think it was a well-reasoned analysis, but sure did turn out wrong.

The data from the establishment survey -- which is completely separate from the household survey by which the unemployment rate is calculated -- wasn't quite as bad. This is the figure that the markets tend to trust more, which is one reason the Dow is slightly up today (although I do think it's somewhat overvalued and has been for a month or so). Still, even the "good" report had the economy shedding another 190,000 jobs.

Not that it necessarily means anything, but there's an odd symmetry to that graph: a period from January through August last year where the payroll deltas were bad, but fairly steady from month to month, and then an inverted peak where things were really awful, and now what looks to be another sort of plateau. January 2008 was the month when the payrolls figure turned negative, and they were at their worst in January 2009 -- if things indeed turn out to be symmetric, that would mean that January 2010 is the last month when the economy is still shedding jobs, and February when it starts creating them.

This stuff ought to be much more of a reason for Democrats to worry than whatever happened on Tuesday. Even if the jobs come back a little faster than expected once the employment picture in fact turns the corner, which I think is possible, voters are liable to be looking at an unemployment rate on the order of 9.5 percent as they go to the polls for the midterms.

Skipping Elections, Strategic Vision Has Not Polled Since Controversy Arose

noreply@blogger.com (Nate Silver) — Fri, 06 Nov 2009 11:55:00 +0000

Several weeks ago, I ignited a controversy by pointing toward statistical evidence that Strategic Vision, LLC, a Blairsville, Georgia based public relations firm that until recently had issued political polls, may have been faking its results. Strategic Vision vehemently denied my interpretation of the evidence and made public threats to sue me. But no lawyer has contacted me, and in fact, Strategic Vision has not conducted any further public polling since that time.

The firm's last poll was issued on September 24th -- one day before the controversy arose. It was a poll of the governor's race and other contests in New Jersey, a state which Strategic Vision has polled frequently for many years, including on 4/22, 6/24, 7/22 and 9/24 of this year. And yet, even as the race drew closer and began to receive widespread national attention, Strategic Vision did not issue any fresh polling. This contrasts with previous patterns in which they had accelerated their polling schedule prior to elections, including the previous gubernatorial election in New Jersey in 2005 when Strategic Vision issued its final poll of the contest on November 2nd of that year.

Strategic Vision's CEO, David E. Johnson, was interviewed by the Washington Times about the Virginia gubernatorial race in late October. And Strategic Vision has issued a couple of press releases on matters unrelated to politics; on October 16th, for instance, they issued a press release to announce that they would be "offering people within the toy industry free thirty-minute consultations to jumpstart their marketing and publicity efforts for the holidays and Toy Fair 2010." But in general, they've had very little public presence over the past several weeks.

One of the facts that may be significant here is that it appears that polling has never been Strategic Vision's main source of income. A search of Congressional Quarterly's Moneyline database sent to me by DavidNYC of Swing State Project turned up just $5,795 in disbursements to Strategic Vision from committees and candidates for federal office since 2004. All came in 2004 from Mike Crotts, a former candidate in Georgia's 8th Congressional District, and none were for polling -- instead, the expenditures were marked as being for website design and advertising. By contrast, the prolific and well-regarded Republican polling firm Public Opinion Strategies had more than $20 million in disbursements over the same period, covering more than 220 clients.

Now, Strategic Vision has clearly gotten some business from polling clients -- the aforementioned series of surveys they did for the Friedman Foundation, for instance. And the CQ Moneyline database will not cover gubernatorial candidates or candidates for other state and local offices. Strategic Vision claims on its website to have conducted polling on behalf of candidates for the U.S. Senate in Kansas and Florida, and candidates for the U.S. House in FL-13 and GA-8, but the identities of the candidates are not specified. An e-mail sent to David E. Johnson inquiring whether the CQ Moneyline database accurately reflected the limited scope of their polling for federal candidates was not returned. In any event, their client list appears to be rather limited, especially when compared with the volume of public polling that Strategic Vision has released, which by its own estimation would have cost it a couple million dollars had it actually been conducted.

The reason this might be relevant is that it may give Strategic Vision more incentive to essentially adopt a "duck-and-cover" strategy and make a quiet exit from the polling business. If Strategic Vision were more dependent on polling clients for its revenues, then it would probably have wanted to make a more vigorous effort to defend its reputation. But in light of their unwillingness or inability to do so, it appears they may have concluded that releasing additional public polling would only invite renewed scrutiny and further damage their reputation. In other words, they may have decided to cut their losses and focus on their original line of business in public relations, presumably hoping that prospective clients in the toy manufacturing or literary services businesses are far enough removed from the political world that they won't care about the possibility that Strategic Vision has faked some or all of their polls.

The Big 2010 Question

noreply@blogger.com (Tom Schaller) — Thu, 05 Nov 2009 21:05:00 +0000

Just attended a morning-long presentation by the team over at the Cook Political Report--namesake Charlie Cook, affiliated pollster Tom Riehle, gubernatorial/Senate specialist Jennifer Duffy, and House specialist David Wasserman. I have lots of details, plenty of stats and trends and other goodies to share and unpack for you, and will do so in a series of posts in the coming days.

But as I walked back from the Watergate to my Logan Circle neighborhood, I became increasingly convinced that the big question for both parties--and particularly the Democrats--is one I raised this morning on MSNBC's Dylan Ratigan show: How replicable is Barack Obama's precedent-setting presidential coalition in an off-year election?

Visit msnbc.com for Breaking News, World News, and News about the Economy

It's easy to just say, well, it's not replicable. Of course it isn't exactly replicable. The so-called "Obama surge" voters clearly will not turn out at the same rates, and thus not constitute the same proportion of the electorate a year from now that they did a year ago. So the question really is, To what degree, along some continuum between the 2008 presidential electorate and the ones from the 2009 elections this week, will 2010 look like one or other other? And looking backward may provide poor guidance: Because there's never been an electorate assembled like the one Obama did in 2008, we've also never had a post-Obama midterm cycle.

Yes, issues and the economic-political environment and the resources that candidates and parties--money, quality of candidates, messaging, field and contacting operations--will all be contributing factors next November. I will come back in a future posts to talk about which of these factors might buffer the expected Democratic losses, and which might exacerbate them.

But all of these factors are ultimately mediated to some, significant degree by the electorate and its composition. That said, I want to start this series of posts with a very simple question that is, more or less, directed at the Obama White House political operation, and can be rather simply stated: One year out, what are you planning to do in order to safeguard your newly-acquired congressional, gubernatorial and even state legislative majorities?

This question in turn begets a variety of sub- and even sub-sub-questions, for which the following is hardly an exhaustive list:

On agenda-setting, do you need to constrict the national policy conversation to fewer agenda items, presumably those more finely attuned to the national economic situation, and how can you do that? Even David Frum admits that rising health care costs the past decade consumed potential income gains...but how many Americans truly understand this and, even for those who do, how long are they willing to wait for their incomes to rise again as a result of savings on premiums? Relatedly, for those without an earned income right now, what messages and themes are you planning to deploy if unemployment a year from now is not significantly below 10 percent?
On candidates--and this question also lands squarely in the laps of DSCC chair Bob Menendez, DCCC chair Chris Van Hollen and DGA chair Brian Schweitzer--are you thinking about where the president will be a co-campaigning, coattail asset and where he will be a drag? Which races will you attempt to localize and which will you attempt to nationalize? Given rising frustration with incumbents, do you need a different strategy for incumbent, challenger and open-seat Democrats? Speaking of potentially open seats, how do you make sure that worried Democratic incumbents do not retire this cycle?
On contacting and turnout, for downballot Democrats, how much are you going to make available the types of voter and fundraising lists, as well as volunteer tools and tactics generated between late 2007 and Election Day 2008 to elect Obama? Indeed, which sorts of resources are even transferable, which not, and how do you utilize the ones that are? Are you at all worried that the technology chasm you created between the parties in 2008 will narrow or even disappear by 2010, and if so, how worried and what will you do about that?
On messaging, what will it take to mobilize the "Obama surge" voters? Do they need to have the 2010 midterms contextualized as a safeguard of their votes cast for the president in 2008? How much of their turnout and support is a function of Obama-mania, or whatever you want to call the specific attachment to the president as a political identity/commodity, and how much of it represents a medium- to longer-term established political identity and partisan attachment? Are you surveying these people about their post-election attitudes and concerns? In order not to trigger an older and/or white voter backlash, might you need to use dog-whistle signaling to the younger and more multi-racial "surge" voters to get them to turn out in 2010, and if so, how will you do that?

I realize it's far easier to pose questions than to answer them. But these questions need to be asked, and the Obama political team--which I suspect is thinking about them this week if they weren't already--had better start generating some answers.

Independent Voters and Empty Explanations

noreply@blogger.com (Nate Silver) — Thu, 05 Nov 2009 16:05:00 +0000

Karl Rove:

The trend here is that suburban and independent voters moved into the GOP column. The overall shift away from Democrats was 13 points in Virginia, 12 points in New Jersey, and eight points in Pennsylvania. [...]

Looking ahead, the bad news for Democrats is that the legislation that helped lead to the collapse of support for their party on Tuesday could yet inflict more pain on those foolish enough to support it.

The Washington Post:

But moderate and conservative Democrats took a clear signal from Tuesday's voting, warning that the results prove that independent voters are wary of Obama's far-reaching proposals and mounting spending, as well as the growing federal debt.

Jim Cooper (Blue Dog-TN)

"Lesser mortals need to be worried about their independent voters because they have shifted strongly against Democrats in recent months. Independent voters tend to look at the issue, not the party, and they don't like a lot of what Congress has done."

This is what passes for analysis nowadays.

Why did Democrats lose in Virginia and New Jersey on Tuesday? Because independent voters moved against them, say the pundits.

This is true, insofar as it goes; Democrats lost independents nearly 2:1 in the gubernatorial race in Virginia, and by a 25-point margin in New Jersey.

But it doesn't really tell us very much. It's a lot like saying: the Yankees won the Game 6 last night because they scored more runs than the Phillies. Or: the unemployment rate went up because there were fewer jobs.

In almost every competitive general election, the party that loses the contest has also lost independent voters. This is because most people (although less so in gubernatorial elections) vote strictly along party lines: the Democrat might be all but guaranteed 80 to 90 percent of the Democratic vote, and the Republican 80 to 90 percent of the Republican vote. Except in certain regions of the country where one or another party encompasses a particularly wide range of ideologies (such as NY-23's Republicans or vestigial "Solid South" Democrats), it's independents who swing the vote, since they represent the overwhelming majority of the votes which are up-for-grabs. This must necessarily be the case.

But in politics, it's not the proximate cause we're interested in but the ultimate one. Yes: independents went mostly for Republicans in New Jersey and Virginia (we could have inferred this without having to look at the exit poll). Yes, this "caused" the Democratic defeats. But what caused the independents to move against the Democrats? That's what we're really interested in, since that's what will have implications for future elections.

Too often in "mainstream" political analysis, once it is pointed out that independents have swung in one or another direction, the analysis stops. The pundit inserts his own opinion about what caused the independent vote to shift ("Obama's far-reaching proposals and mounting spending", says the Washington Post), without citing any evidence. It's a neat trick, and someone who isn't paying attention is liable to conclude that the pundit has actually said something interesting.

But in New Jersey, there's literally almost no evidence that the Democrats' agenda had anything to do with Jon Corzine's defeat. Voters who cited a national issue were more likely to vote for Corzine, and voters who cited a local one, the Republican Chris Christie.

In Virginia, the evidence is certainly a little stronger, insofar as the national agenda may have affected the lopsided turnout (the electorate which turned out Tuesday had voted for John McCain by 8 points, a near-reversal of the actual results). Even there, however, the quarter of the electorate that cited health care as their main issue went for the Democrat Deeds 51-49. And in NY-23, which was supposed to have been the ultimate smackdown of the Democrats' agenda, the ~~Republican~~ Conservative candidate unexpectedly lost.

Part of the problem is that 'independents' are not a particularly coherent group. At a minimum, the category of ‘independents’ includes:

1) People who are mainline Democrats or Republicans for all intents and purposes, but who reject the formality of being labeled as such;
2) People who have a mix of conservative and liberal views that don’t fit neatly onto the one-dimensional political spectrum, such as libertarians;
3) People to the extreme left or the extreme right of the political spectrum, who consider the Democratic and Republican parties to be equally contemptible;
4) People who are extremely disengaged from politics and who may not have fully-formed political views;
5) True-blue moderates;
6) Members of organized third parties.

These voters have almost nothing to do with each other and yet they all get grouped under the same umbrella as 'independents'.

But that's getting away from the point. Independent voters are treated as a cause, when all that they really are is a symptom. The key is in figuring out what ails the patient.

Congressional Foreign Politics

noreply@blogger.com (Renard Sexton) — Thu, 05 Nov 2009 12:00:00 +0000

One vote that occurred this election day did not pit Democrats and Republicans against one another in an effort to win seats or change policy. Instead, it was a widely bi-partisan effort, meant to show solidarity with a U.S. ally. In a vote under suspended rules (requiring 2/3 majority) the U.S. House of Representatives voted Tuesday in favor of a resolution that called on "the President and the Secretary of State to oppose unequivocally any endorsement or further consideration" of the UN-endorsed Goldstone Report which investigated whether violations of human rights law were committed by Israel and Hamas during the Israeli assault on Gaza a year ago.

Regardless of the merits of the report or the resolution, the 344-36 final vote with 22 voting present (30 not voting) signified widespread support in both parties for the House's rejection of the document.

In the House Committee on Foreign Relations, from which the bill originated, there was an almost identical vote share in favor (on the final bill), though the "not voting" contingent was a higher percentage and included key members. Committee Vice-Chair Gary Ackerman, who is also the Chair of the Sub-committee on the Middle East affairs and Donald Payne, Chair of the Sub-committee on Africa and Global Health, both chose not to vote on the bill. On the Republican side of the committee, only Ron Paul voted against.

For all of its bluster and bi-partisanship, the non-binding resolution does not have a direct practical impact on either the consideration of the Goldstone report at the UN General Assembly or Security Council, or the Obama Administration's overall strategy in Israel-Palestine. Indeed, like many other resolutions that express the "sense of the House," the legal and policy implications are tenuous at best.

The role of the legislative branch in the development of foreign policy is constantly changing, based in part on the interest and timeliness of the Congress and in part of the relevance of Congress' main tools to issues at hand. The political aspect plays a major role as well, with members of Congress eager to vote on resolutions that bolster their standing in their home districts, without bearing the brunt of responsibility for the impacts.

For making a serious impact, however, veteran congress-people identify a few key tools that the legislative branch has to adjust the foreign policy priorities and actions by the executive. The last two, treaties and appointment holds, are restricted to the Senate.

1. Power of the purse: On large scale initiatives, such as the wars in Afghanistan and Iraq, the use and deployment of weapons systems, troops or warships, Congress can delay, deny or put guidelines using appropriations measures. The most recent major example was a slew of amendments and bills with regard to the Iraq war, which was largely unsuccessful at actually changing US activities in the country. This strategy requires constant pressure on the President by Congress as well as strict spending guidelines, and therefore requires a great deal of political capital and commitment.

2. War powers: Congress has the power to "declare war," and "call forth the militia," though in modern practice, this authority is rather unclear. In conventional activities of the US military against another country, such as the 1991 Gulf War and the 2003 Invasion of Iraq, Congress' purview of the main operation was clear. However, for many military activities of the US abroad, conventional war authorization is not particularly relevant. For example, after September 11th, Congress passed a resolution authorizing the " Use of Military Force Against Terrorists," which has been interpreted in quite broad terms.

3. Individual lobbying: Congress(wo)men and Senators, particularly those on the Foreign Relations committees, are able to travel around the globe and bring back to the President pet causes, whether localized (hunger in a rural developing country) or more broad (e.g. human rights violations against the Tibetans).

4. Senatorial holds on Presidential appointments: In the Senate, preventing executive appointments to sensitive positions, such as a senior staff to key departments or ambassadors abroad, can be used for leverage.

5. Senate's treaty powers: The U.S. Constitution requires that all treaties be approved by the Senate by 2/3 majority before they become law. Ignoring that the legal question of what qualifies as a "treaty" has had a long and circuitous history, the fact is that many agreements negotiated with foreign governments include consultation with the Senate, whether an official vote is held or not.

Since the President often plays the dominant initiating role on most foreign policy issues, he or she therefore holds the majority of political responsibility (read: risk) for the success or failure of the initiatives. As the leading edge, the executive has to deal with the nuances of diplomacy and foreign engagement, while in the background, Congress can deal in more black and white absolutes that are more linked to politics than policy. For example, the debate over restricting funding for the Iraq war in the 110th Congress (2006-2007) was framed in terms of withdrawal or surge, with little gradation in between.

As a result, there is often little risk for Congress to pass aggressive yet myopic policy resolutions or amendments, even in cases where the impact on U.S. actions is real. But when the time comes for re-election, there is adequate ammunition for touting strong commitment to ideals, defense and allies.

In the case of the Goldstone report, it is possible that the actual contents of the report did not matter much to most voting members. Except for the relative few who are closely engaged on the Goldstone report itself, for example Minnesota Democrat Keith Ellison, the roll call was merely a "Pro-Israel or not" vote, supported by both the Democratic and Republican leadership as a point of bipartisan cooperation. In fact, apparently the initial mark-up of the bill had a number of factual errors in it, such as characterizing the Goldstone mission as investigating only the Israeli conduct in the operation rather than both Israel and Hamas, which were only corrected after the author wrote a letter to the bill's sponsors.

Given that the House's effort on this is largely symbolic, using none of the tools discussed above, it is clear that the resolution is largely rhetorical. But the question remains about what kind of action the current Congress would take if they were intent on solidly changing the Obama administration's approach on something broader, such as the Iran negotiations or the Afghanistan conflict. Would they follow a similar path as was taken in 2006 and 2007, which failed to change the Bush administration's conduct of the Iraq war? Or would a different track be taken, perhaps threatening to reduce or change the 2001 anti-terrorism war resolution or more informal techniques such as Senatorial blockage?

Nonetheless, with many incumbents looking vulnerable going into the 2010 midterms, it could be that Congress does not really want to engage in foreign policy making in the short term, instead focusing on the soft-edge political votes like the one from Tuesday. For a case like Afghanistan or Iran, there may be simply too much risk for little reward, particularly in a political environment where domestic issues continue to dominate.
---
Renard Sexton is FiveThirtyEight's international columnist and is based in Geneva, Switzerland. He can be contacted at sexton538@gmail.com

DGA Chief Says Dems "Carry Burden of Proof"

noreply@blogger.com (Tom Schaller) — Wed, 04 Nov 2009 21:17:00 +0000

Just listened to a 3 p.m. post-election conference call hosted by Nathan Daschle, executive director of the Democratic Governors Association. (538 regulars will recall an interview Daschle conducted with the site late this summer.)

Here is a quick summary of his comments today, as best and as quickly as I could type them:

“As you can imagine, there are about a 1,000 things I’d rather be doing today than discussing last night’s results," he opened. "One thing you learn in politics is that you have good nights and bad nights."

He said that the DGA is trying to “learn what you can from results, but not dwell on them,” adding that they are “disappointed, but not discouraged.” He twice pointed out that New Jersey and Virginia are historically counter-trending states—that the two states have elected governors from the opposite party controlling the White House for the past 24 years, arguing that it “would have been an upset for us to win.” He said they were turning the focus immediately to the 37 governors races on the ballot 2010.

In what sounded like a prepared statement, Daschle went on to say it was “important to draw lessons but not over-analyze” the results, which he said did not support: (1) the conclusion that there’s a Republican resurgence, again noting the counter-trending of the two governors’ seats relative to the party controlling the White House; nor (2) the idea that President Obama “is politically wounded,” saying it was not a political referendum on the White House; nor (3) that “Democrats are in trouble for 2010,” citing polls in New Jersey showing that Jon Corzine won on the “economy” yesterday but lost because “other issues superseded the economy.”

There are three fair conclusions to draw, however, he said:

It’s a reminder that “Democrats carry a burden of proof” to deliver real results, specifically in “creating and saving jobs, and restoring prosperity.” He added later, during the question-and-answer period that he believes it a tough time for incumbents, but that that “cuts both ways” for Democratic and Republican governors or gubernatorial candidates next year, especially given that 48 states are presently running budget deficits.
Second, it’s a reminder that "federal issues are a matter that, as the majority party, we need to show that we are a party of actions, not rhetoric." Later, during the question and answer, he expanded on this theme in response from a question from US News reporter Dan Gilgoff, saying that the DGA has been advising their candidates that "if the Republican opponent is trying to box you in on federal issues it is because they have no policy or platform of their own on state issues."
Finally, it showed that the “Republican Party is still in disarray and not ready to take over. They are very much still a house divided,” he said, dropping references to Rush Limbaugh and Michael Steele and reiterating the common Democratic talking point that the GOP is the “Party of No.”

“If there’s a silver lining, it’s that we conserved our resources. We spent half what the RGA spent,” said Daschle. “It was the right call because neither race would have been helped by more spending.” He said the DGA spent $4 million in VA to the RGA’s $5M, and about $3.5 million in New Jersey to the RGA’s $7 million.

I asked Daschle whether he thought Bob McDonnell’s strong performance in the Northern Virginia counties was a byproduct of his local ties there or some larger change in partisan support in these counties which helped Democrats Jim Webb, Mark Warner, Tim Kaine and Barack Obama all carry the state. “I think it was always something of a misnomer that Virginia had become a Democratic state,” he said. As for NoVa voters, he specifically added that “those voters respond to optimistic messages and positive messages and care about how public officials are going to their lives better. The fact that [McDonnell] came from Northern Virginia and reminded people of that in all his signs didn’t hurt. Voters in that part of the state respond to positive, optimistic economic messages.”

Turning to the House, Democrats were more upbeat after winning the two special elections. Democratic Congressional Campaign Committee chairman Chris Van Hollen (MD-8) said this about the New York 23rd district race that received nationwide attention:

"Congratulations to Congressman-elect Bill Owens on his remarkable victory. Voters in New York's 23rd District responded to Bill Owens' message and track record of creating jobs and attracting economic development to Upstate New York.

"This election represents a double-blow for National Republicans and their hopes of translating this summer's ‘tea party' energy into victories at the ballot box. Not only did eight extreme right-wing groups spend more than $1 million to drive the moderate Republican - and the NRCC's chosen candidate - out of the race. Now, after losing a seat that was held by Republicans for nearly 120 years, they have to deal with an emboldened and well-funded far right-wing that refuses to tolerate moderate Republicans with differing opinions."

Finally, the RSLC's Carrie Cantrell, who spoke with 538 Friday, was enjoying a partisan victory lap today. The RSLC issued a press release which reads in part:

Voters across the country last night from New Hampshire to Washington state sent a loud and clear message--Americans want a limited, less overreaching government. Not only did Republicans win all three statewide offices in Virginia for only the second time in history, but gains were made by Republicans in state legislatures in Virginia (+6), New Hampshire (+1), New Jersey (+1), Michigan (+1) and Washington state (+1). The Republican State Leadership Committee, the nation’s largest caucus of Republican state leaders, contributed more than $2.2 million to state races across the country this year and focused on a new media strategy to heighten awareness of these races.

“There will be a lot of talk about what last night’s election means. One thing is crystal clear--the American people have had enough. These overwhelming victories in ‘blue states’ are the result of strong Republican candidates providing commonsense solutions to the real problems facing all Americans,” said Republican State Leadership Committee (RSLC) President Scott Ward. “Our candidates listened to the voices of the people, and offered a clear vision of fiscal responsibility and economic growth that attracted not only Republicans, but Independents and Democrats as well, to our winning cause. They understand taxpayers are the owners of the government – and that the government ought to let individuals determine their destiny rather than government gimmicks that reduce their freedoms.”

So that's a roundup of how the parties are spinning yesterday's results.

New Jersey ≠ Virginia

noreply@blogger.com (Nate Silver) — Wed, 04 Nov 2009 17:28:00 +0000

New Jersey and Virginia are similar states in many respects. Both are wealthy and suburban. Both live in the shadow of large cities -- New York, of course, in New Jersey's case, and Washington in Virginia's. Each has its "white ethnic" population -- the conservative-leaning Scots-Irish lineage that dots rural Virginia, and the working-class-Democrat Italian and Portuguese-American families that populate New Jersey (herein lies the difference between the two states electorally). Both have their share of minorities, who may or may not turn out in midterm elections.

But the outcome of the gubernatorial contests in the two states yesterday was very different -- even though both were won by Republicans.

In New Jersey, you had an extremely unpopular incumbent -- Jon Corzine -- who had never particularly clicked as a governor, and who's approval ratings crossed into the red in February, 2008. In Virginia, you had an open seat race to replace an outgoing, and in fact fairly popular Democratic governor.

In New Jersey, you had an electorate that gave Barack Obama a 57 percent approval rating -- the identical fraction to the 57 percent that elected Obama last November. In Virginia, Obama's approval rating was 47 percent, a significant drop from the 53 percent of the vote that he earned.

In New Jersey, it was Jon Corzine who tried to nationalize the race, making sure that everyone knew that Chris Christie was a Republican. And insofar as this went, it worked: voters who said their main issue was health care went for Corzine 78-19 (!), according to exit polls, and he won voters focused on the economy and jobs 58-36. Christie won because he focused on two local issues that are very important to New Jerseyans -- corruption and property taxes, and won overwhelmingly among voters who keyed in on these issues. In Virginia, meanwhile, it was Bob McDonnell who won the economy voters -- 57-42, and the candidates split the vote among those most concerned about health care.

In New Jersey, 27 percent of the electorate was nonwhite, exactly identical to the proportion from a year earlier. In Virginia, the percentage of nonwhite voters dropped from 30 percent to 22.

In Virginia, the Republicans swept the downballot races, and are projected to have made a net gain of five seats in the House of Delegates. In New Jersey, Democrats are only expected to lose one seat in the state's legislature.

Most obviously of all, in New Jersey you had a candidate losing by 4 points, while in Virginia you had one losing by 18. In New Jersey, you do have to pick and choose an explanation for Corzine's defeat -- and the evidence points overwhelmingly toward local factors. In Virginia, it would be irresponsible not to pick all-of-the-above: absolutely, Deeds was a poor candidate and McDonnell a good one; absolutely, Democrats had huge turnout problems; absolutely, Virginia has a tendency to be counter-cyclical in its gubernatorial elections. But absolutely also, national issues played a role, and damaged the Democrats.

Ironically, though, because the outcome in Virginia was not close and the one in New Jersey was, it's the latter race that has received more attention. And many of the explanations that actually do apply to Virginia -- where Creigh Deeds got clobbered -- have been superimposed onto the Garden State, with no basis in the evidence. I live in Brooklyn and watch sports every Sunday and saw every freakin' one of Chris Christie's commercials a dozen times over, the only redeeming facet of which was that there were moderately less putrid than Corzine's. Not once did Christie mention Nancy Pelosi or Harry Reid or the Democrats' health care plan or the stimulus package -- and a good thing he didn't, since this wasn't what New Jerseyans were concerned about.

If you're a Chief-of-Staff for a conservative Democrat in Congress, you have two relevant data points to look at from Tuesday: Virginia, which is undoubtedly a little scary, and NY-23, which ought to be rather relieving. Certainly, there were a whole host of local factors and unusual contingencies on the ground in NY-23. But it also spoke to the limitations of conservative populism (CP) as an electoral instrument. (You can call the CP's 'teabaggers' if you want, but my term is both more neutral and more descriptive.) There's not really any evidence that the CP movement is yet anything more than an isolated and regional one. It will almost certainly have some implications in the South -- and if I were a Democratic Congressman there, I'd be very nervous. But only 18 of the 52 Blue Dogs in fact come from the South, and if I were a conservative Democrat in California, or South Dakota, or Michigan, I'd be feeling rather relieved.

All Politics is Local, Ctd.

noreply@blogger.com (Nate Silver) — Wed, 04 Nov 2009 14:22:00 +0000

In a piece over at CNN.com, I do my best to take a coherent message out of a night that didn't particularly lend itself to one:

Corzine, for his part, ran a polarizing campaign; every time Christie's name appeared in one of his commercials, it came with a scarlet (R) -- for Republican -- attached. Republicans are not popular in New Jersey, but local issues drove the race.

Whereas three-quarters of Corzine's voters cited a national issue -- health care or the economy -- as their primary reason for voting for him, two-thirds of Christie's picked a local one (property taxes and corruption).

What Happened and Why?

noreply@blogger.com (Nate Silver) — Wed, 04 Nov 2009 07:43:00 +0000

The outcome of all seven contests that we were tracking tonight appears settled, or very nearly so:

Virginia Governor: Republican Bob McDonnell wins by 17 points, toward the upper end of the range predicted by the pollsters, although not to anybody's great surprise. Democrats had major turnout problems here; exit polls show that the electorate which turned out in Virginia supported McCain in last year's election 51-43, almost exactly the opposite of the actual margin. But Deeds also appears to have been the weaker candidate. The electorate was roughly spit on approval of Obama, but 20 percent of those who approved Obama nevertheless voted for McDonnell, while just 5 percent of those who disapproved Obama voted for Deeds.

New Jersey Governor: Republican Chris Christie wins 49-45. We had (somewhat tentatively) characterized the race as leaning Christie on the basis of superior enthusiasm and the incumbent rule. Corzine never polled at better than 44 percent in any individual poll of the race. It looked for a time like 44 or 45 percent might nevertheless have been enough to win him the election, but support for the third party candidate Chris Daggett collapsed, leaving him exposed.

Obama approval was actually pretty strong in New Jersey, at 57 percent, but 27 percent of those who approved of Obama nevertheless voted for someone other than Corzine. This one really does appear to be mostly about Corzine being an unappealing candidate, as the Democrats look like they'll lose just one or two seats in the state legislature in Trenton. Corzine compounded his problems by staying negative until the bitter end of the campaign rather than rounding out his portfolio after having closed the margin with Christie.

NY-23: Democrat Bill Owens prevails in a result that will be regarded as surprising; the final tally isn't in yet but it appears as though it will be something on the order of 50-45 over Conservative Doug Hoffman. I don't think I've ever hedged more on predicting the outcome of a race; the main issue is that there was a rather large discrepancy between the polling, which heavily favored Hoffman, and what I perceived to be the facts on the ground. NY-23 is solidly Republican but not especially conservative (it voted for Barack Obama last year), and Hoffman was a relatively uncharismatic candidate with poor command of the local issues.

If New Jersey was a win for the incumbent rule, then NY-23 may have ben a win for the Median voter theorem, as Owens -- a conservative Democrat -- was actually much closer to the average ideology of the district than the capital-C Conservative Hoffman. It was also a reminder that all politics is local (sometimes). More than 95 percent of Hoffman's contributions came from out-of-district, and the conservative activists who tried to brand him as a modern-day Jefferson Smith never bothered to check whether he resonated particularly well with the zeitgeist of the district. In any event, this is a Democratic takeover of a GOP-held seat and they expand by one their majority in the House.

CA-10: California Lieutenant Governor John Garamendi has been declared the winner. His lead as of this writing is 10 points, almost exactly matching the margin in the only poll of the race, but smaller than the margins by which retiring incumbent Ellen Tauscher had grown accustomed to winning. Both sides will breathe a bit of a sigh of relief here: Democrats for avoiding an embarrassingly close result -- although 10 points is closer than it "should" have been -- and Republicans for not having to second-guess themselves for their decision not to put money into the race.

Maine -- Question 1. Maine votes Yes on Question 1 -- which means no on gay marriage -- by a margin of about 52-48. Turnout was extremely high and should eventually surpass 500,000 voters, about where it was during the 2006 midterms. This fact was initially thought to favor the pro-gay marriage side -- but, obviously, it didn't. The results showed a very strong urban-rural divide, with the initiative being rejected by a margin of about 2:1 in Portland but racking up big margins in smaller towns and rural areas, especially in the north of the state.

We had given Question 1 about a 70 percent chance of being defeated based on a combination of an analysis of the polling and a statistical model. I don't know how much time I'm supposed to spend defending being on the wrong side of a 70:30 bet -- we build in a hedge for a reason -- but here comes a little self-reflection. As for the polling, I think we have to seriously consider whether there is some sort of a Bradley Effect in the polling on gay rights issues, although one of the pollsters (PPP, which had a very bad night in NY-23) got it exactly right. As for the model, I think I'll need to look whether the urban-rural divide is a significant factor in a state in addition to its religiosity: Maine is secular, but rural. At the end of the day, it may have been too much to ask of a state to vote to approve gay marriage in an election where gay marriage itself was the headline issue on the ballot. Although the enthusiasm gap is very probably narrowing, feelings about gay marriage have traditionally been much stronger on the right than the left, and that's what gets people up off the couch in off-year elections.

I certainly don't think the No on 1 campaign can be blamed; by every indication, they ran a tip-top operation whereas the Yes on 1 folks were amateurish. But this may not be an issue where the campaign itself matters very much; people have pretty strong feelings about the gay marriage issue and are not typically open to persuasion. There's going to be an effort by many on the left to blame Barack Obama for his lack of leadership on gay rights issues; I think the criticism is correct on its face, but I don't know how much it has to do with the defeat in Maine. A more popular Democratic governor, for instance, who had been a bit quicker on the trigger in his support of gay marriage, might have helped more.

Washington -- Referendum 71: Moreover, it actually wasn't all that bad an election for gay folks. Sources differ as to whether the race has officially been called or not, but it appears that Referendum 71, which expands domestic partner rights to an everything-except-marriage standard in Washington, will be Approved. The initiative leads by only about 22,000 votes right now, but about a third of the outstanding vote is from Seattle's King County, which supports it heavily. I wonder to what extent measures like Referendum 71, which is sort of a separate-but-equal compromise on the gay marriage question, will come to be seen as an acceptable alternative by either side.

New York City -- Mayor: Mike Bloomberg wins a third term, but the margin is stunningly close -- just five points over Democratic rival Bill Thompson. Looking at the exit polling, however, I actually think the interpretation of this race is relatively straightforward. Voters in the exit poll approved of Bloomberg's performance 70-29, but a quarter of those who approved of Bloomberg voted against him anyway. Why? Because 58 percent of voters said that Bloomberg's decision to change the city's term limits law to enable him to seek a third term was a significant factor in their vote, and those voters broke against him 2:1.

Everybody's talking about reverberations here in New York City, although it's not exactly clear what they'll be, since most people are perfectly satisfied with Bloomberg's job performance. One of the more obvious ones is that you may see a high profile Democrat, perhaps someone from the Congress, running for this seat in 2013.

NY-23 Results by County; Democrat Owens Appears Poised for Victory

noreply@blogger.com (Nate Silver) — Wed, 04 Nov 2009 04:41:00 +0000

As of about ~~11:35~~ 12:05 New York time. Source.

Note that the major outstanding stash of votes is in St. Lawrence County, where Owens has done very well all evening. There are also a goodly percentage of the vote left in Lewis County, where Hoffman leads 52-43, but there just aren't that many voters there.

Most of the 5,800 or so absentee ballots throughout the district also haven't been counted; I don't know who that favors. But it looks to me like Hoffman is going to fall further behind when the rest of St. Lawrence comes in (which may not be until tomorrow; they've had some machine problems there) and the absentees -- of which Scozzafava will have a decent share -- won't be enough to help him.

With mucho caveats, as always. I'm just one guy here trying to cover seven elections at once and call 'em as I see 'em in real time.

EDIT: For instance, Watertown is not in St. Lawrence County, as I previously reported. Hence, the caveatting.

UPDATE: Fox News has called it for Owens.

UPDATE: There's been some confusion about the number of absentee ballots in the district. 10,000 were sent out, but 5,800 were returned.

Still Too Close to Call in New Jersey (Update: Apparently Not?)

noreply@blogger.com (Nate Silver) — Wed, 04 Nov 2009 02:36:00 +0000

Here's a comparison of Jon Corzine's vote share in 2005 versus today, in the counties that have represented a decent fraction of their vote.

Taking an unweighted average in these counties, Corzine is running about 8 points off his pace from four years ago, when he won 53.5 percent of the vote. That would put him at 45.5 percent -- probably not quite enough to win -- but this is an imprecise analysis for a number of reasons and Corzine's numbers are holding up a bit better in the more vote-heavy counties in North Jersey. And I don't know to what extent absentee ballots are included or not included in this count, which were thought to favor Corzine.

~~Still too close to call. Too close to even think about calling.~~

EDIT: Err... apparently not, as the AP has called the race for Christie. Other networks have not so far. But I'm looking at this one-dimensionally -- comparing percentages, but not turnout -- whereas hopefully they're looking at both factors.

Deeds v. Obama in Virginia

noreply@blogger.com (Nate Silver) — Wed, 04 Nov 2009 00:58:00 +0000

A comparison in seven counties where all votes have been counted.

County_______ Deeds Obama
King & Queen_ 41.0  51.8
Prince Edward 44.9  54.3
Westmoreland_ 41.4  54.6
Wythe________ 27.3  32.9
Lexington____ 60.5  62.2
Williamsburg_ 54.6  63.8
Winchester___ 39.2  52.0
AVERAGE______ 44.1  53.1

These counties are actually fairly representative -- Obama won 52.6 percent of the vote in Virginia overall -- so it looks like Deeds is going down to defeat on the order of 56-44. This is why some of the networks are starting to call the race.

Election Night Overview

noreply@blogger.com (Nate Silver) — Tue, 03 Nov 2009 23:20:00 +0000

Here are the seven races that we'll be making at least some effort to track for you tonight:

Polling averages are from Pollster.com, except for CA-10 and Referendum 71 where there's only one recent independent poll available and I'm just listing that verbatim. Our previews of the races can be found at the links below:
Virginia
Maine Question 1
New Jersey
NYC
NY-23: (original; revised; re-revised)
CA-10
Washington Referendum 71

Consider this your 'homethread' for the night. There will of course be frequent updates, although some of the shorter stuff I'm going to post to the sleeping giant that we call the 538 Twitter feed. You don't have to go offsite to see the tweets as they're also contained in the box in the top RH side of the page.

I will be on WNYC periodically throughout the night, and on Morning Meeting (MSNBC) again tomorrow from 9-11 AM.

Have fun and play nice -- this is turning out to be a much more interesting election night than it looked like it would be even a few weeks ago.

2009 Elections Preview: Washington Referendum 71 (Domestic Partnership)

noreply@blogger.com (Nate Silver) — Tue, 03 Nov 2009 21:19:00 +0000

Washington (state) -- Referendum 71 -- "This bill would expand the rights, responsibilities, and obligations accorded state-registered same-sex and senior domestic partners to be equivalent to those of married spouses, except that a domestic partnership is not a marriage. Should this bill be Approved or Rejected?"

The Positions: An Approve (Yes) vote upholds the expanded domestic partnership rights approved by the state legislature. A Reject (No) vote withdraws those rights and benefits, although it does not overturn domestic partnership itself.

The thing to notice is that the language is the functional opposite here of what it was in Maine or California. An affirmative (approve) vote is good news for same-sex couples, and a negative (reject) vote is bad for them. This has caused some confusion; one pollster has stated that as much as 10 percent of the electorate might vote in a way opposite to their true intentions.

The Polls: The only independent polling in the race is from SurveyUSA, which shows the Approve side winning 50-43, an improvement from 45-42 a month ago. GQR also polled the race on behalf of the pro-domestic partnership group Washington Families Standing Together and showed it passing 53-36, although the usual caveats apply as this is a nonindependent poll.

Analysis: Washington is similar to Maine in certain respects, being white and fairly secular, and since I think the pro-gay marriage side is more likely than not to prevail in Maine, you might think I feel the same way about the initiative in Washington state. Indeed I do feel that way, although the initiatives are not directly comparable. On the one hand, Referendum 71 does not go as far as Maine's Question 1 or California's Proposition 8 since it seeks to reaffirm an "everything but marriage" bill that does not formally bestow the title of marriage upon same-sex couples. On the other hand, a rejection of the referendum would not overturn Washington's 2007 domestic partnership law, but instead only the expanded, marriage-like benefits that were afforded to those couples this year.

Were Washington to vote on a measure to ban domestic partnership outright, it would almost certainly fail and fail badly: by a 58-42 margin, according to my statistical model. A measure to ban gay marriage but not domestic partnership would be much closer; I have such a measure failing 52.5-47.5, but there is a good deal of uncertainty there, and in an off-year election the numbers might be closer to 50:50. Referendum 71 appears to be polling somewhere in between those two goalposts, which makes sense, since it takes Washington somewhere in between domestic partnership and full-blown marriage.

There is also arguably less uncertainty about the outcome in Washington than in Maine. This is because, as in California, most Washingtonians vote by mail, and SurveyUSA has the Approve side leading 53-42 among those who have already voted. A small bit of good fortune for the Approve side is that there is a highly competitive mayoral race in Seattle, which might encourage turnout in that obviously very liberal corner of the state.

The Odds: Although there is a lot of uncertainty in both the polling and the statistical model because of the ambiguity of the measure under consideration, they do tend to point toward the same result: Referendum 71 passing by a margin on the order of 7-10 points. Coupled with what appears to be movement toward the pro-domestic partnership side -- which may reflect voters familiarizing themselves with the language of the ballot -- and what also appears to be an advantage for the Approve side in votes collected thus far, Referendum 71 appears to be fairly safe. I would give about 10-1 odds against its being rejected.

Isn't 2009 More About the GOP?

noreply@blogger.com (Tom Schaller) — Tue, 03 Nov 2009 20:36:00 +0000

Perhaps it's my imagination, but there seems to be a lot of media discussion of how the results today in two governor's races and one rather bizarre House race serve somehow as a referendum on Barack Obama, the Democrats, their majority, the Democratic agenda and platform, the bank bailout, the stimulus package, "HOPE," your haircut, my haircut, the undefeated New Orleans Saints, and just about anything else that can be piled on.

OK, OK, I'm exaggerating for effect. But if the 2009 cycle is a referendum on either party, isn't more of a referendum on the Republicans? Or perhaps, more precisely, a referendum for Republicans about the meaning of Republicanism?

Yes, elections are typically--and rightly--a referendum on the policy performance of the in-power majority party. But during an election cycle, and particularly in the primaries leading up to the general, it is the out party that is working out its issues and kinks, trying on its new or reconfigured identity. This was certainly the storyline when liberal Democrats were backing people like Ned Lamont back in 2006.

That said, consider that the GOP nominees in the three highest-profile races today--Virginia's Bob McDonnell, New Jersey's Chris Christie and New York 23's Dede Scozzafava--represent three variations on the question of what ails the GOP and how to fix it. Moving from center-right to right, lets' take a look at each model:

The McDonnell model: reform the establishment GOP from within as an insider and consensus-builder who moves to the middle. McDonnell is the most conventional and establishmentarian of the three. He came up through the normal channels as a law-and-order attorney general, touted his Fairfax roots in an effort to appeal to the key swing region of Northern Virginia, avoided culture war appeals and focused on meat-and-potato issues like taxes and transportation. Granted, his avoidance of culture war politics was as much a necessity as a choice, given the potentially toxic issue of his graduate thesis controversy. But he didn't run on a divide-and-conquer, base-oriented, Bush-era strategy. He proved able to appeal to the party's base and middle as well as independents.
The Christie model: reform the establishment GOP by running as a solid conservative. Next comes Christie, who also ran as a party man. He's not the non-traditional conservative firebrand that Steve Lonergan, whom he beat in the GOP primary, is. But Christie isn't Christine Todd Whitman either, and he doesn't strike me as any less conservative than McDonnell, although he's running in a more liberal state than McDonnell. Christie may still beat Corzine, whose background as a Goldman-Sachs millionaire is a tough sell right now. But Corzine may hold on.
The Hoffman model: revolt against the establishment GOP from the outside with a third-party candidate, consequences for the GOP be damned. Hoffman may win, too, but he ran outside the GOP. Defenders will say the party's choice of Dede Scozzafava gave him and his conservative backers no choice. But that's not entirely true: They could have supported her and worked to move her toward their positions to earn that support. Whatever the case, the establishment choice was viewed as such an affront--an apostate--that something radical had to be done...and if that meant doing it outside the confines of the party, so be it.

No matter what, because of the Hoffman-Scozzfava dustup, Republicans will not win all three of these races. And if Christie loses, the GOP will go 1 for 3. That may be the most telling lesson of all in terms of how the party needs to re-build. Whatever the results tonight, it seems to me that 2009 is the year of the Republicans almost by definition because these contests represent a broader, intramural contest to define GOP's future strategy and identity.

I'm not saying the the Democratic nominees, the way they ran their campaigns, or state and national issues are or were meaningless. They're not, of course. And I'm not saying voters in these elections have no opinions about the president, or are not voicing those opinions through their votes. But 2009 is about the GOP.

2009 Elections Preview: CA-10

noreply@blogger.com (Nate Silver) — Tue, 03 Nov 2009 17:30:00 +0000

California 10th Congressional District -- Special Election to replace Ellen Tauscher.

The Candidates: California Lieutenant Governor John Garamendi, Democrat
David Harmer, Republican
Various minor party candidates

The Polling: The only poll of the race, from SurveyUSA, shows Garamendi with a 50-40 lead.

Analysis: The ballyhooed race in New York's 23rd Congressional District is not the only special election today; voters will also go to the polls to select a successor to Ellen Tauscher in California's 10th Congressional District in Northern California's inland East Bay. Actually, they've already been going to the polls, or at least to their driveways, since a majority of the voters in the district are expected to use California's vote-by-mail option.

On paper, this race ought not to be close. Against Republican opponents in 2004, 2006 and 2008, Tauscher was re-elected with margins ranging from 65.2 to 66.5 percent of the ballots. In the special "jungle" primary in September, five Democratic candidates earned a collective 64.4 percent of the vote (closely matching Tauscher's historical figure), with Garamendi taking the fairly narrow plurality at 25.7 percent. The district went for Barack Obama nearly 2:1 last November, and Democrats have a 47-29 registration edge in the district. Garamendi, being the state's Lieutenant Governor, is quite well known, and had previously been considering a gubernatorial bid.

Still, the Republicans have made it a somewhat competitive campaign, with Harmer collecting a respectable $670,000 toward the race, not that far behind Garamendi's $942,000. The SurveyUSA poll shows them within 10 points. The district did vote for Arnold Schwarzenegger by an 18-point margin in 2006. And Garamendi has one major liability: he is associated with S-a-c-r-a-m-e-n-t-o, that horrible place where budgets and incumbents go to die. Could Harmer actually pull off the upset?

It's not very likely. The most immediate problem is that 51 percent of the district has already voted, according to SurveyUSA, with those votes going to Garamendi by 10 points. Now, to be clear, these aren't "hard" numbers -- this is still just a poll, with all the usual inexactitudes owing to sampling error and the like. But the Garamendi campaign claims that there have been 25,000 Democratic absentee ballots returned so far to 19,000 Republican ones -- not, actually, all that impressive a margin for the Democrats, but probably enough to keep them out of trouble. Indeed, the fact that voting is comparatively easy in California because of mail balloting ought to hedge against the possibility of hugely lopsided turnout in this heavily Democratic district.

Meanwhile, independents who are dissatisfied with Garamendi will have the option of voting for the Peace and Freedom or Green Party candidates, which may hold down Garamendi's margin but also take potential votes off the table from Harmer.

A final point: I'm not a huge fan of this inference-by-absence stuff, but the fact that the NRCC is not putting much effort into the race suggests that they don't think it's winnable. If Harmer had some internal polling showing himself within striking distance, you'd certainly think they'd have thrown a few hundred thousand in.

The Odds: I might take about a 15-1 flier on Harmer -- and if Republicans do win here, or perhaps even pull close enough that the outcome will be uncertain for several days as California finishes counting mail ballots, they'll really have something to crow about. With that said, I suspect that Garamendi will more likely than not win by larger than the 10-point margin that SurveyUSA projects. By the way, this race is not without upside to the Democrats, as Garamendi should be significantly more liberal than Tauscher, who has not always been a reliable vote for her party on issues like national security.

DLCC: Worried About Low Turnout Today

noreply@blogger.com (Tom Schaller) — Tue, 03 Nov 2009 16:21:00 +0000

In the second of my two interviews with top staffers from each party's state legislative campaign committee, below is the transcript of my conversation Monday with Democratic Legislative Campaign Committee executive director Michael Sargeant. Prior to his elevation to executive director, Sargeant served as the political director and then national political director of the DLCC. Before joining the DLCC, he was the Tennessee House caucus director.

538: Can you give our readers a sense of how much the DLCC has invested in the state legislative races in New Jersey and Virginia for the election tomorrow?

Michael Sargeant: The reports haven’t all come out yet, but we’ve been a major donor to the effort in Virginia this year, just as we were two years ago.

538: And in New Jersey?

Sargeant: We’ve been a major donor in Virginia but not in New Jersey because we believe we’re going to hold the majority in the [New Jersey] assembly.

538: Are you tinkering with any new technological tools or tactics in the field campaign and for voter contacting generally, and if so what?

Sargeant: It’s an interesting question. We’ve all been taking a look from the last election on at a variety of things regarding microtargeting, and making sure we’re also doing more polling and just being more aggressive as well at the doors doing GOTV.

One of the things we’re very proud of is our DLCCweb program, which we’re making available on our website for legislative candidates around the country using this. We have around 350 to 400 candidates around the country using this and they were in full force last election cycle in 2008.

538: So you are using it beyond these races in 2010 and beyond?

Sargeant: We first used it in the ’08 races, again this year, and the program is just growing by leaps and bounds.

538: The conventional wisdom is that there is not going to be the “Obama surge” voters turnout in either state tomrrow, and you’ll have an older, whiter cohort turn out. Are you worried about that?

Sargeant: Yes. Look, I think it’s only natural that the electorate is going to be different in an off-year after a presidential election. I can’t speak enough that the strength of the campaign [VA House minority leader] Ward Armstong has organized. But it’s only natural that the turnout model is going to be different in 2009 than it was in 2008.

538: Do you think candidates Creigh Deeds and Jon Corzine have any coattails—or alternatively that Bob McDonnell and Chris Christie will coattails or downballot effects—on the state legislative races, or statewide will it be a series of localized contests?

Sargeant: That’s a great question. I will say this. Our program has been to make an emphasis to make sure our candidates run localized, personalized campaigns that fit their districts—that they’re talking about that matter to their district, not necessarily a national message or a statewide message. That’s the strategy we’ve been working with our candidates since 2004, 2006, 2006, and the off-years since I’ve been here at the DLCC. So that’s been our strategy.

As for statewide campaigns, I think there will be places in the state where those campaigns help us and places where they will hurt.

538: You’re talking about Virginia? And which places will it hurt? The conventional wisdom is that Democrats do well in the NoVa counties near DC, but McDonnell is touting his strength there.

Sargeant: I think Bob McDonnell so far has run a strong campaign. I think our candidates have done a good job in Northern Virginia and have done a good job in the rest of the state. But McDonnell has run a strong campaign.

538: Shifting away from New Jersey and Virginia for a moment, can you comment on the situation in the New York State Senate?

Sargeant: Our main objective is working with Democratic leaders to put together the best campaign operation possible to win majorities. We did work with the New York Senate last cycle to help them gain the majority for the first time in many, many years. What’s happening now in New York, let’s just say it’s interesting.

538: Do you think that significant gains the Democrats have made in state legislatures the past three cycles—2004, 2006 and 2008—that these gains, in terms of total legislatures or chambers controlled, has reached a limit? Have you guys capped out?

Sargeant: I think the map is interesting and it looks different state to state. We’ve moved from 41 Democratic majorities to 60 since the 2002 election. That’s a seachange at our level of campaigns and politics. And there are still some places where Democrats can make serious gains, such as the Texas house and senate, and the Tennessee house. Those are places where we feel Democrats can make serious gains.

I think in large part, instead of it being a national issue, these campaigns are run locally and the dynamics, district to district, are going to be different, as well as what happens in each of the states. It’s not so much that Democrats have capped as much as what’s going to be the strength of the campaigns on the ground and what are going to be the issues that matter most to the voters who are there.

Democrats, you know, we moved 10 chambers toward Democratic control in 2004. And most people would say that was not a very strong year for the entire Democratic ticket. Democrats running for state legislature were running in a lot of strong campaign programs and by personalizing these campaigns.

538: Going back to New Jersey and Virginia, can you identify a rising legislative star or legislative candidate this cycle who is a rising star in that state?

Sargeant: I think Dave Englin in Virginia is very smart and works hard for his issues. He’s a real rising star. He’s a sitting house member from Northern Virginia and a very strong candidate.

538: What will be the headlines out of Richmond and Trenton on Wednesday morning as far as overall gains or losses for Democrats in those two state legislatures?

Sargeant: I think Democrats hold our majority in New Jersey Assembly. And in Virginia, they’ve gained 11 seats over the last two house election cycles in the Virginia House. So that’s been very tough math they’re battling in Virginia.

I think we’re in a position where there may be some losses but there just as easily may be some gains. There are some tossups that could go either way. I wouldn’t be surprised if we had a net gain in Virginia.

538: So do you think, in terms of the overall storyline nationally, if the governors races split, do you think Republicans will claim a victory if they do better in the state legislative races and statewide races? Is it a foreshadowing of a correction, or a pushback or Republican surge—what’s the bigger takeaway?

Sargeant: I think the bigger takeaway is that the Virginia house Dems have put themselves in the position to get close to a majority. And they have a very difficult map that they’re toiling under. With the turnout projections we’re looking at, it’s going to be difficult to win those last couple of seats to win that majority.

I guess the larger point I’m trying to make in Virginia is that it’s a very tough map and that redistricting matters.

Gubernatorial Races a Poor Yardstick

noreply@blogger.com (Nate Silver) — Tue, 03 Nov 2009 13:17:00 +0000

Here is the problem with using the results from New Jersey or Virginia tonight to judge the status of the national political environment. It's not so much that the races are "meaningless" in the abstract, but that ticket-splitting is so common in gubernatorial races that the noise swamps the signal.

Of the eleven states with a PVI of D+7 or bluer, five (Hawaii, Vermont, Rhode Island, California and Connecticut) currently have a Republican governor. Of the ten states with a PVI of R+10 or redder, meanwhile, four (Kentucky, Kansas, Oklahoma, Wyoming) have a Republican governor. The correlation between gubernatorial elections and elections to the House, Senate and Presidency has been very weak, at least recently. In fact, if you compare the share of the vote that the Democratic candidate got in the most recent gubernatorial election in each state to the share that Barack Obama got last November, it is almost literally zero:

Another way to expose the same concept: here are the percentage of ticket-splitters -- Obama voters voting for a Republican governor, or McCain voters voting for a Democratic governor -- in the 12 states which held gubernatorial elections last year:

As you can see, the percentages vary radically from state to state. In North Dakota, 57 percent of Obama's voters voted Republican for governor, in Utah 48 percent, and in Vermont, 38 percent. Meanwhile, in West Virginia, 59 percent of McCain's voters went with a Democrat for governor, 41 percent in New Hampshire, and 36 percent in Montana. On average, about 22 percent of voters for both candidates split their gubernatorial and Presidential ticket in these states. This compares with 8 percent of Obama voters who voted for a Republican for the U.S. House, and 12 percent of McCain voters who voted Democratic for the House. Ticket splitting is about twice as common in gubernatorial races as in races for the Congress.

Public Policy Polling thinks that, in both New Jersey and Virginia, about 15 percent of Obama voters (the ones who bother to turn out) will vote Republican for governor, whereas about 5 percent of McCain voters will vote Democratic for governor. Those percentages, coupled with depressed turnout among their base, will almost certainly cost the Democrats Virginia and possibly New Jersey. But they're also well within the range of "normal" given the local circumstances intrinsic to gubernatorial campaigns -- similar, for instance, to what happened in Washington State last year.

To be clear, the fact that gubernatorial races are not a reliable benchmark does not mean the Democrats are not in trouble in 2010 -- whoa, too many double negatives there -- nor that Democrats might not have done better if Obama's approval rating was 62 percent instead of 52 percent. It just means that New Jersey and Virginia don't have particularly much informational value -- we won't become very much smarter about the future based on what happens there. To the extent that we do learn something, it will probably be hints about turnout, motivation and enthusiasm, rather than something about the electorate's policy preferences.

NY-23 could potentially be more telling, except that the dynamics of the race are so unusual that everyone will have plenty of (good) excuses for whatever outcome might ensue. Still, I'll bet you that NY-23 -- not New Jersey or Virginia -- is actually the race monitored a little more carefully on Capitol Hill.