Thursday, April 29, 2010

Networks analysis on Stata

Someone finally put together a networks analysis package for Stata.  It's pretty limited so far -- it can only do circle or MDS plots -- but it has a lot of potential, and it's a lot easier to use than R (although considerably slower).  The plot at left is the hookup network for the original 90210 cast.  I made it in about two minutes.

Wednesday, April 28, 2010

Does party predict sports team affiliation?

A few weeks ago I asked whether liberals were more likely than conservatives to root for Butler over Duke in the men's NCAA basketballs finals, given that Butler was clearly the underdog.  Faithful reader BK, who teaches sports marketing at a business school, writes in to say, basically, no:
After spending years in sports marketing, and after kicking the question around with colleagues, we were relieved to conclude (unscientifically) that political ID was an almost undetectable and rarely reliable indicator of fan support. Interestingly, fans of teams may gravitate to a choice of media (CNN vs. Fox), but it may be a function as much of age or other factors than political lean. Other elements that may form a political identification -- geographic considerations, socioeconomics, race -- may influence ticket purchases, but fans just don't like their politics mixing with their fun time. A friend with the St. Louis Blues said they greatly regretted having hockey mom Sarah Palin drop the ceremonial first puck at a game in the fall of 2008. The backlash against the team, not from Democrats, but instead from fans who wanted a break from the political season, was intense - although the ProgressNow "player cards" handed out pre-game were priceless. Same response to the Scott Brown/Martha Coakley/Curt Schilling silliness earlier this year, and to some degree, John Elway's endorsement of John McCain.
Your own musings about the Duke/Butler contest likely confirm your gut suspicions about fans in general.... The good news, from my perspective, is that sports is still largely devoid of the political opinion that occasionally touches film and music stars. Human nature usually forces us to actually unite - on the side of the hometown team and/or the underdog!

Tuesday, April 27, 2010

"We were this close to going out forever"

Humanity nearly died out about a million years ago:
Early humans living about one million years ago were extremely close to extinction. Evidence from a novel genetic approach, one that probes ancient DNA regions, suggests that the population of early human species back then, including Homo erectus, H. ergaster and archaic H. sapiens, was 55,500 individuals, tops.
I assume we can credit a proto-John Connor for organizing us and teaching us how to fight.

Colorado Republicans counter-organize

The first rule of the Vast Right Wing Conspiracy -- you do not talk about the Vast Right Wing Conspiracy.

I wrote a while back about Colorado's Gang of Four -- a shadowy organization that seeks to get around campaign finance rules and is, in a sense, a new version of the state's Democratic Party.  The process by which that happened is nicely described in Schrager and Witwer's The Blueprint, which I recommend highly.

Well, it appears that the state's Republicans are trying to follow the blueprint and are building their own counter-organization in the form of something called Common Sense Colorado.  This organization has allegedly raised around $800,000 and plans to raise around $10 million, which it plans to distribute to Republicans who are challenging vulnerable Democrats in swing statehouse districts.

The Denver Post obtained much of the information for this story from a Common Sense PowerPoint presentation.  Given that one of its organizers, Jon Andersen, provides quotes for the stories, it doesn't seem like the Post obtained this PowerPoint by accident.  Frankly, it sounds like a lot of bluster.

For one thing, yeah, $10 million is a lot of money for a handful of state legislative races, but given that Democrats have shown an ability to spend that kind of money on these races in the last three cycles, it doesn't look like Common Sense is going to overwhelm the Democrats.  Second, Common Sense has less than a tenth of the money it's promising to spend, and it's almost May.  Now, maybe they have a very aggressive fundraising agenda for the summer (and maybe the Post article is part of that), but they still have a long way to go.

Third, one of the brilliant parts of the Democratic blueprint was secrecy.  Republicans were massively outspent in 2004 in large part because they didn't see it coming.  The Gang of Four didn't give interviews bragging about their fundraising strategy.  They just built this bizarre web of 527s and independent expenditure committees and started pumping tons of money into it.  This made it hard to track contributions and campaign activity.  And then they disbanded those organizations and built new ones for the 2006 cycle, and then did the same thing in 2008.  By contrast, Common Sense Colorado is now a big fat target.  They will be sued for anything that vaguely looks like coordinations with candidates and will spend hundreds of man hours and hundreds of thousands of dollars dealing with litigation.

Copying the Democratic blueprint seems like a smart idea.  But I get the impression that the founders of Common Sense only skimmed the book.

Monday, April 26, 2010

"A past-its-prime institution trading on its name"

More bad news for Berkeley: a lousy review of Chez Panisse.

For what it's worth, I ate at Chez Panisse a few times during college when my parents would come to visit.  I always thought it was excellent.  I recall eating some sort of calzone with goat cheese that was wonderful.  My positive reviews, however, should be discounted by the fact that this was two decades ago and I was around 20 at the time and did not have as refined a palette as I do today.

Now where'd I leave my candy corn and Yoo Hoo?

On the perils of Internet voting

I tend not to be too concerned about the dangers associated with Internet voting.  It doesn't strike me as having much more potential for fraud than absentee balloting or even in-person voting.  Meanwhile, I can move thousands of dollars around online through my Wells Fargo account or on E-Trade, and that doesn't seem to cause many problems.

Okay, that said, check out this story from the Daily Californian.  Turns out Noah Stern, the new president-elect of the UC Berkeley student government, is being prosecuted for voter fraud.  Among the allegations:
Stern approached sophomore Axel Prompt and senior Roy Pfaffman on the last night of voting, offering to cast a ballot on behalf of the two students. A reporter for The Daily Californian observed the exchange between Pfaffman and Stern in which Stern offered the use of his BlackBerry so that Pfaffman could log into the elections website with his CalNet ID. Pfaffman then returned the phone to Stern, who voted on Pfaffman's behalf.
Now, it's not obvious to me here which party is more guilty.  Unless Stern was somehow coercing the students to vote, they're just as guilty for letting someone vote for them as he is for stuffing the virtual ballot box.  But obviously the crime is more costly for Stern, who could end up being disqualified for office.

Yoko

This is not nearly as exciting as I think it is.

MPSA roundup

Well, I'm back from the Midwest Political Science Association.  While I blogged nothing, I Tweeted quite a bit, enjoying the challenge of summarizing conference papers in 140 characters or less.  Just thought I'd mention a few papers I saw -- hopefully they'll be up on the conference website shortly.
  • Eitan Hersh analyzed the differences between caucus attendees and primary voters and found basically none.  There are no differences in terms of ideological extremism or any other political variable he examined, and his N (using CCES data) was quite high, meaning that if a difference was there, he should have detected it.  This seems pretty surprising, given that caucus turnout is usually around a tenth that of primary turnout and given the higher participation barriers for caucuses.  Hersh suggests that the sorts of people who go to caucuses simply have a higher threshold for meetings, but aren't extreme in other dimensions.
  • Robert Boatright is following up on his research on the primarying of incumbents with an interesting paper on the effects of getting primaried.  The paper is a fascinating read if for no other reason than its portrayal of the complexity of studying the effects of primaries.  Do primaries weaken incumbents in the general election, or do weak incumbents just invite primary challenges?  Does a primary challenge cause an incumbent to change her roll call votes, or does shifting roll call behavior lead to primary challenges?  There are huge causality and endogeneity issues here.  Boatright makes a valiant attempt to wade through some of this, dividing up hundreds of congressional primary challenges by type: being challenged from the extremes, being challenged from the center, being challenged due to scandal or old age, etc.  He does find some modest effects, and generally in the expected directions.  He also finds that Democratic incumbents seem to suffer the effects of primary challenges more than Republican incumbents do.  However, basically none of these results is statistically significant.  I sympathize with this work -- Boatright is doing us a real service here, but it looks frustrating.
  • Jordan Ragusa shows us a few interesting things about the career paths of senators.  First, the percentage of senators with prior experience in the House has been increasing steadily over the past half century.  Second, senators with former House experience tend to be more partisan than "indigenous" senators.  Ragusa argues that the career socialization in the more partisan House tends to train politicians as partisan, and they take that training with them into the Senate.
Beyond that, I had a bunch of productive meetings and saw many old friends.  I finally met Brendan Nyhan -- we enjoyed a limo ride and some deep dish pizza along with a large group of devastatingly cool political scientists.

Wednesday, April 21, 2010

Sweet home Chicago

I'm on my way to the annual conference of the Midwest Political Science Association in Chicago.  Looking forward to four nights of revelry in the Palmer House Hilton and a few nights of fried smelt at Miller's Pub.  I hope to blog some of the more interesting things I see and hear.  I'll probably be doing some live-tweeting, as well, so be sure to follow me if political science conferences interest you.  (The official hashtag is #mpsa.  It's official in the sense that a few other people and I have started using it.)

Could a Perot campaign happen today?

Over at 538, Tom Schaller nicely takes apart the notion that the modern Tea Party movement is the heir to the Ross Perot campaign of 1992.  The main reason, of course, is that the Perot supporters weren't necessarily conservative.  For an historical contrast, think about what happened during the Democratic convention of 1992.  Perot decided to withdraw, suggesting that the Democrats seemed to have their act together that year, and many of his supporters and campaign aides went to work for the Clinton campaign.  Now flash forward to 2012, when Sarah Palin, the Tea Party nominee, withdraws, leaving the field a two-man race between President Obama and GOP nominee Mitt Romney.  Any chance she praises Obama or that her supporters go to work for the Democrats?  Basically none.

But I think it's a fair question whether a Perot candidacy -- a fundamentally centrist political movement -- could exist today.  Note Ron Rappoport's description of Perot supporters:
Perot callers were slightly right of center on the liberal-conservative scale, but on specific issues they were were not consistently conservative. They strongly favored abortion rights, national health insurance, and government controls on pollution, while strongly opposing affirmative action, gun control and the revocation of the death penalty.
[...]
In terms of demographics, Perot supporters were unobservant religiously. Only 38% attended services every week and 16% never attended services-both very different from the American public. And while 33% did not identify as either Protestant, Catholic or Jewish among Perot callers, such was the case for only 15% of tea party supporters. On the other hand the tea party movement and the Perot supporters were both about 60% male and over 90% white.
What a mix!  I get the impression that 1992 was about the last time it was possible for a politically active group of people to hold such a blend of views.  Supporters of abortion rights and health insurance have steadily embraced the Democrats, as have the religiously unobservant.  Opponents of affirmative action and gun control have increasingly become Republicans.  These people were hugely cross-pressured at the time, but that tension has largely resolved one way or another.  The early 90s was the time that the parties in the electorate polarized on issues of race, abortion, and church attendance, among other things. Putting together a group of voters that could unite around a candidacy despite embracing such a diversity of partisan stances and still capture a fifth of the vote strikes me as impossible today.

Giving money to senators not actually all that weird

The NYT had a front page story yesterday in which they suggest that financial firms are trying to buy off the support of the Senate Committee on Agriculture.
A main weapon being wielded to fight the battle, of course, is money. Agriculture Committee members have received $22.8 million in this election cycle from people and organizations affiliated with financial, insurance and real estate companies — two and a half times what they received from agricultural donors, according to the Center for Responsive Politics.
Much of that lobbying has centered on Senator Blanche Lincoln, the Arkansas Democrat who is the committee’s chairwoman and who last week introduced the bill that would prevent banks from trading derivatives directly.
The daughter of a sixth-generation rice farmer, she has found herself navigating a dangerous channel between Wall Street firms, which raised $60,000 at two fund-raisers for her re-election campaign so far this year, and her constituents, many of whom want a crackdown on the speculation that led to the financial crisis.
Other committee members, on both sides of the aisle, also have reaped donations from people and companies in the derivatives business, including Senator Saxby Chambliss of Georgia, who is the committee’s ranking Republican member; Kent Conrad, the North Dakota Democrat; and Charles E. Grassley, the Iowa Republican.
Sounds curious, all right!  Except that Lincoln, Conrad, and Grassley also sit on the Senate Finance Committee.  The Times might have mentioned that.

(Thanks to Jennifer Victor for noticing.)

Tuesday, April 20, 2010

Simgov

I am in the middle of teaching an American government simulation course, or Simgov.  Every student is portraying a current member of the House of Representatives, with the exception of three students who comprise the executive branch.  We're about halfway through the quarter, but so far this is proving to be one of the most fascinating and useful courses I've ever taught.

The first few weeks were a bit of a slog, in which I lectured extensively on parliamentary procedure and a few basics of executive/legislation relations.  But as of last week, they've been meeting in committees.  Each student is required to author six pieces of legislation during the course, and the first bills have been working their way through the different committees.  I have never seen such buy-in from such a large percentage of students before.  I've observed several committee hearings and so far every student has been engaged in the work, participating in the discussions, and learning the material.  I only needed to assign reading material for the first few weeks -- a few classics like Mayhew's The Electoral Connection and Sinclair's Unorthodox Lawmaking.  The students are generating reading requirements on their own now by doing research on bills.

I've heard of several professors doing American politics courses that incorporate a week or two of congressional simulation (see this article by Sands and Shelton), but doing it for an entire course strikes me as novel.  Not that it was my invention.  I inherited this course from my former colleague Tom Knecht, who helped develop it in grad school based on the work of other professors going back to the early 80s.  I've been relying heavily on this earlier work.  One of the newer updates to the Simgov course is the addition of Blackboard, which allows students to post bills and communicate electronically.  I'm also using clickers to electronically record roll call votes.  If I get enough votes, I might try to calculate ideal points for the students.  (If I feel really geeky, I'll compare students' ideal points to those of the members they're portraying, and then deduct points if there's too big a difference.)

Anyway, one obviously can't teach every course this way.  It's a deeply atheoretical class.  Students are focused on very pragmatic issues, such as how to put a bill together, how to assemble a winning coalition, how to speed up or slow down legislative procedures, how to run a hearing, etc.  But the learning is dramatic and incredibly widespread.  I'd certainly make my course materials available to anyone who wants to adopt this class.

Thursday, April 15, 2010

Regression on the iPhone

Ever been at a presentation and seen some data thrown up on the screen and wanted to analyze it yourself right there on the spot?  If you're as geeky as I am, and you own an iPhone, there's now an app for you.  I've been waiting for some decent statistics apps for a while, and it looks like a few have recently shown up on the market.

I downloaded one called Regression Calculator for $.99.  Unlike with spreadsheet apps, data entry is very easy.  You just enter the x variable, with all the numbers separated by a comma, and then enter the y variable.  It'll then give you a plot, R-squared, and even make a predicted y based on an inputted x.  You can do linear, logarithmic, exponential, or power regression.  The plot at left shows growth in real disposable income predicting the incumbent party's share of the presidential vote since 1948.  I inputted the data in about 3 minutes.

The program is very simple and could probably use a few features (adjusting the axis scale, saving data, etc.).  But for what it is, it works quite well.  I'm sure this will come in handy at MPSA next week.

Wednesday, April 14, 2010

The echoes of field organization

I did a post a while back on the power of field offices to affect elections.  Specifically, I showed that the presence of an Obama field office was associated with a modest but detectable bump in the Democratic vote share in the 2008 general presidential election.  (The results were published here.)  I also have a chapter coming out (hopefully) that shows the power of field offices to affect delegate selection during a party nomination campaign.  I'm now finding some evidence that such field offices have a persistent effect in later contests.

I've been looking at the results of the March 2010 Democratic caucuses in Colorado's U.S. Senate race.  Using a multi-variate analysis of the caucus-night vote, I find that Michael Bennet did, on average, twenty-one percentage points better than Andrew Romanoff in counties where Obama had a pre-caucus campaign office two years ago.  This is statistically significant (≤ .05), even after controlling for virtually every major demographic characteristic of the counties.

I don't fully understand the mechanism here.  It may be that the early Obama office counties just built strong organizations, which translated into strong OFA groups, which are active in the current Romanoff/Bennet contest.  Or it may be that the early Obama offices developed more lasting connections with active Democrats in those counties, so that the Obama endorsement of Bennet this year was more influential in those counties.

Another hypothesis might be that there is some peculiar facet of the Obama field office counties that causes them to prefer certain candidates in Democratic nominations contests but has nothing, in fact, to do with any effort exerted by those offices.  Obama put his offices in those counties, but the offices didn't change the vote -- they would have preferred him, and Bennet, anyway.

One way to test this hypothesis is to look at a Democratic statewide primary contest that occurred prior to the establishment of these field offices.  Probably the best one is the 2004 Senate contest between Mike Miles and Ken Salazar -- this was the last time there was a major Democratic nomination contest for statewide office in Colorado.  In fact, there was no relationship between that primary vote and the location of Obama's field offices.  Now, this isn't an overwhelming piece of evidence for my case -- it was six years ago, and people do move in and out of counties.  But, in theory, if there was something peculiar about the Obama field office counties in '08, we might have seen some evidence of it in '04, and we don't.  This is at least suggestive that the offices Obama built in '08 did cause the vote boost for him and are affecting the current Democratic Senate nomination contest.

More conservative counties voted for Ken Buck

This is a followup to my previous post analyzing the votes on caucus night in Colorado last month.  I'm still trying to gather results on the county assemblies, something that's proving very difficult to do (although thank you Dan Willis).  For now, though, some interesting patterns have emerged from the caucus votes.

One pattern, which is perhaps not so Earth-shattering, is that Ken Buck did better than Jane Norton in more conservative counties during the March caucuses.  McCain's presidential vote in 2008 is a reliable predictor of Buck's successes.  This result is statistically significant (p ≤ .05) and holds even with all the control variables I could throw at it.
Interestingly, county ideology played no role in the Democratic contest.  Liberal counties were just as likely to go for Romanoff as for Bennet.

Tuesday, April 13, 2010

Taibbi on Brooks

I'm pretty ambivalent on David Brooks these days.  I like that fact that he draws from actual social science research once in a while, and his columns are rarely dull.  But this latest bit about Duke winning the men's NCAA basketball playoffs was monstrous:
The rich are not always spoiled. Their success does not always derive from privilege. The Duke players — to the extent that they are paragons of privilege, which I dispute — won through hard work on defense. ...For the first time in human history, rich people work longer hours than middle class or poor people. How do you construct a rich versus poor narrative when the rich are more industrious?
I could go off on this for hours, but Matt Taibbi beat me to it:
Only a person who has never actually held a real job could say something like this. There is, of course, a huge difference between working 80 hours a week in a profession that you love and which promises you vast financial rewards, and working 80 hours a week digging ditches for a septic-tank company, or listening to impatient assholes scream at you at some airport ticket counter all day long, or even teaching disinterested, uncontrollable kids in some crappy school district with metal detectors on every door.
Most of the work in this world completely sucks balls and the only reward most people get for their work is just barely enough money to survive, if that. The 95% of people out there who spend all day long shoveling the dogshit of life for subsistence wages are basically keeping things running just well enough so that David Brooks, me and the rest of that lucky 5% of mostly college-educated yuppies can live embarrassingly rewarding and interesting lives in which society throws gobs of money at us for pushing ideas around on paper (frequently, not even good ideas) and taking mutual-admiration-society business lunches in London and Paris and Las Vegas with our overpaid peers.
Meanwhile, with regards to the NCAA series, does anyone have a sense of how Duke-Butler rooting broke down by party ID?  I found myself instinctively rooting for Butler, despite a) knowing nothing about the school; b) having never been to Indiana; c) having several friends on faculty at Duke; and d) caring very little about March Madness.  Is this a liberal preference for the underdog?

Sunday, April 11, 2010

Timing

Did the Tea Partiers really pick April 19th, the 15th anniversary of the Oklahoma City bombing, as a date to gather in state capitals to inveigh against the federal government? Really? I hope they're just completely ignorant of the date's significance, because the alternative is far worse.




The Denver County Democratic Convention

I spent yesterday at the county Democratic convention. The main item of controversy there was, of course, the Bennet/Romanoff Senate contest. So it was sort of like the Obama/Clinton split at the county convention two years ago, only without the acrimony. I just saw very little tension between the candidates' supporters. Pretty much everyone there likes Romanoff and also thinks Bennet is doing a fine job as senator. But even in terms of seating, there didn't seem to be the kind of self-segregation I encountered two years ago.

There was one interesting moment of tension. Former mayor Wellington Webb got a few boos when he praised Bennet's handling of Manual High back when Bennet was DPS superintendant. Webb yelled back, telling the booers to go to East High where the county Republicans were meeting.

Romanoff took the county convention 65-35, echoing a pattern in just about every county I've seen in which Romanoff has improved his numbers over caucus night.

I got picked as a delegate to next month's state convention, so I'll report more on that.

Friday, April 9, 2010

The most polarized cake in America

I went to the UC Davis Political Science Department today to give a talk, only to find that my confectionary arch-nemesis/mentor Sherry Zaks had provided the refreshments.  I can really think of no greater honor than this delectable doppelgänger.

Thursday, April 8, 2010

KCRA interview

I'm in the middle of my northern California tour right now.  Things are going really well, and right now I'm sitting outside a La Bou cafe across from the state capitol drinking coffee and taking advantage of free wi-fi and gorgeous weather.  So far things are going very well.  I had a really nice interview this morning on Sacramento's NBC affiliate.  They gave me three and a half minutes, which I think is more time than they devoted to national politics in total.  And the reporter had actually read some of my book, which was much appreciated.

Tuesday, April 6, 2010

I walk these streets, a loaded laptop on my back...

I begin my big Northern California book tour today.  If you're within range of any of these events, please drop by!

  • Wed., 4/7, 12-1:30PM, UC Berkeley's Institute of Governmental Studies, 118 Moses Hall
  • Thu., 4/8, 12-1:30PM, Center for California Studies, 1130 K Street, Suite LL22, Sacramento
  • Thu., 4/8, 7-8PM, The Avid Reader Bookstore, 1600 Broadway, Sacramento (flyer here)
  • Friday, 4/9, 12-2PM, UC Davis Dept. of Political Science, 693 Kerr Hall
Also, if you live in Sacramento, watch for me on NBC news Thursday morning at 7:40AM.

Saturday, April 3, 2010

The Ballet Star Cake


I made this one for my daughter's fifth birthday.  The basic design comes from Lindy Smith's book.  The pleated skirt is my favorite part.

Bennet collecting petitions

According to the Pueblo Chieftain, Sen. Michael Bennet is starting a petition drive to get on the August primary ballot.  This sounds like a smart move to me.  Yes, he got around 43% of the caucus night vote, but what will ultimately determine his spot on the ballot is the vote at the state assembly in May.  Early reports from the county assemblies suggest that Romanoff is picking up big numbers there.  I'm hoping to get some hard numbers there, but my guess is that the Romanoff caucus/assembly-goers are just more hard core and more willing to show up at the multiple stages of this whole nomination process.  It's certainly conceivable that Bennet could end up below 30% by the state assembly, which would require him to petition onto the ballot.

Bad blogger

Apologies for the light blogging.  Between class prep, visiting in-laws, and a frenzied attempt to clean out the house, life's been a bit overwhelming.  I'll try to get back on the ball soon.