Saturday, 19 May 2012

cnlmisc 0.2 for R

The oft-promised update of the cnlmisc package for R is now posted. New in this release is a convenience method, sepplot, that produces separation plots using the separationplot package; this method works directly on model fit objects as a post-estimation call, and works with both binary and ordinal models at present. In addition, epcp now works with clm2 objects from the ordinal package.

Most of this was motivated by continued work on the economic voting paper, which has also been updated. cnlmisc still has a long way to go before I submit it to CRAN, but at least it’s progress, right?

Monday, 5 March 2012

An aside comment on the "Best Star Wars Film" debate

While I don’t want to wade too deeply into the argument, seemingly initiated by Kevin Drum’s rather absurd notion that Return of the Jedi is the best of the six* Star Wars films, regarding the relative merits of the various films in the series (see also: Doug Mataconis at OTB and Seth Masket), I do want to raise a minor point in response to Dan Drezner on the politics (or lack thereof) in the triology:

The conundrum that political scientists face is that even though the original trilogy contains the better films, the second trilogy has the better politics. There are no politics in Episodes IV-VI, unless one counts Vader and the Emperor’s wooing of Luke. In the prequel trilogy, however, there are lots of parliamentary machinations, tussles between the Jedi Council and the Chancellor, Anakin’s lust for power, and Darth Sidious’ grand strategy for converting the Republic into an Empire.

To a political scientist, that’s good stuff. To human beings interested in enjoying a film, it’s tissue paper without things like strong characters, a good screenplay, and decent plotting.

While I’m slightly sympathetic to Dan’s argument here, the reality is that the politics of the prequel trilogy are, in a word, silly, even leaving aside arguments about whether one would plausibly construct an elective, term-limited monarchy in which the only valid candidates for office are teenage girls, or what sane society would elect the likes of Jar Jar Binks to high office (ok, maybe that one is more credible). Sure, there are depictions of politics, but only within the context of political structures that make no sense, such as the Senate of the Republic (there’s a reason that real legislatures don’t have membership sizes in excess of the population of a mid-sized city) and the Jedi Council (there’s also a reason that real legislatures governing groups of people in the millions have more than a half-dozen, self-selected members).

Slathering on a layer of thinly-veiled BusHitler allegory doesn’t exactly help matters either, if only because in 20 years nobody will get the point Lucas was belaboring—to illustrate the point, imagine if Lucas had taken a 20-minute detour during Empire Strikes Back to establish some boring parallel between the political ascents of “black mayors” Walter Washington and Lando Calrissian, perhaps by giving Lando a bunch of long-winded, boring speeches that paralleled the racial politics of the early 1980s, and then imagine how that would play today.

The other problem of course is that the politics depicted in the prequels is boring. Politics of course need not be boring (for example, the writers of Parks and Recreation manage to make politics entertaining on a weekly basis), but in the hands of Lucas—who’s obviously more interested in the prequels in advancing plot only to serve as a scaffolding for spectacle rather than having the CGI elements there in service of a sensible plot—most of the politics gets reduced to tedious speeches and arguments in what seem to be shot-for-shot remakes of scenes from academic department meetings. In the hands of a skilled writer (or, perhaps more charitably, a writer who cared) I have no doubt the political machinations promised in the prequels might have been interesting; as presented, the Wikipedia summaries of them are positively life-like by comparison.

* Part of me wishes there were only three, but that might edge into the territory of Frequent Commenter Scott’s denial that the sport that is played in the American League qualifies as “baseball.”

Wednesday, 19 May 2010

Another year bites the dust

My sixth year of full-time teaching is now at an end. Overall I think it went well, although I missed my target grade distributions in both of my upper division classes (too tough in Congress & The Presidency; too easy in Political System of the USA). One of these decades I’ll get it right.

I’m now looking forward to a very busy summer, including a conference, AP exam reading, two summer courses to teach, and three or four research projects in various stages from completely unwritten (my APSA paper) to on the verge of journal submission (my Midwest paper with Scott and Adolphus). After all of that, I’ll probably be looking forward to a relatively restful three-course semester with only one totally-new-to-me course, the first semester of graduate research methods.

Sunday, 9 May 2010

STV is high-threshold PR

Contra Simon Jackman, the single transferable vote is a form of proportional representation, albeit one with a very high effective electoral threshold (the share of the vote a party needs to gain representation)—in the worst case, something on the order of (but not quite) 100 percent divided by the average district magnitude + 1 (number of seats per STV constituency).

Of course, the motivation for this discussion is the British election and the Liberal Democrats’ demand for a more proportional electoral system, specifically STV. Labour seem rather more enthused about electoral reform than the Tories at present, but one suspects Labour’s newfound sponsorship of the idea had more to do with pre-election positioning than a genuine interest in reform—Labour certainly didn’t complain with the 2005 election awarded them a healthy Commons majority on essentially the same share of the vote the Tories got this week.

Labour’s pre-election offer was the alternative vote, better known in the United States as instant runoff voting, or IRV. IRV effectively is a simplified form of STV in single-member districts, e.g. STV with a district magnitude of 1. I doubt the LibDems would be willing to settle for IRV, as it probably wouldn’t net them many additional seats, even if their supporters would have fewer wasted votes under IRV (as their second preferences would be allocated rather than discarded). IRV and other similar SMD systems (like the French two-round arrangement) are generally regarded as majoritarian rather than proportional.

In the British context at least, STV makes a lot of sense as a preferred electoral reform. Any proportional system will somewhat disadvantage the two leading parties (the Conservatives and Labour) compared to plurality (first-past-the-post/winner takes all) voting, but STV is less proportional at sane district magnitudes (3–6 seats per district) than virtually all PR systems, so the damage to leading parties is smaller. The major beneficiaries are the regional parties, regionally-weak parties (such as the Scottish Tories), and of course the Liberal Democrats; it should also have the salutary effect of somewhat depoliticizing the constituency boundary-drawing process in Northern Ireland in particular.

Fringe parties and those whose platforms can easily be co-opted by larger parties don’t come out ahead under STV, but that would seem to be a feature, rather than a bug—Parliament doesn’t need the BNP around, UKIP is a party without a purpose in a world with the Tories still in it, and the Greens are effectively Liberal Democrats who just don’t want to call themselves LibDems. Denying these groups 25 or so seats in the Commons between them doesn’t seem like any great loss for British democracy.

Saturday, 8 May 2010

Once more into the cesspool

P.J. O’Rourke once said that giving money and power to politicians was akin to “giving whiskey and car keys to teenage boys.” But that pales in comparison to the effects of giving an anonymous forum to mentally-teenaged political science graduate students and their hangers-on.

There was a point a few years ago—perhaps even a few months ago—when I believed having a job rumors forum was a necessary corrective to the fundamentally broken hiring process in our discipline. I firmly believe that if we are going to share a discipline of a few thousand people, and if we’re going to work with these people for decades in the future as peers, we ought to treat those starting out on the tenure track with the basic standards of decency we would expect from our own colleagues—and that requires honest, up-front information about the job market and search process as it happens, rather than a few summary statistics a year or two down the road from the hiring season. It is a principle I tried to uphold when we successfully searched for a colleague last year—and given that I still have a job, it was a pretty costless one. Although not one that many of my fellow political scientists have decided to follow, alas.

But whatever the hell is going on over at the rumor site has very little to do with fostering collegiality and openness today. Instead, the site seems to have been captured by an element of jealous, petty individuals who resent the success—or, seemingly more often, revel in the apparent lack thereof—of a small number of graduate students from leading political science programs. Perhaps these students are, to borrow a phrase from a former American president, major-league assholes. Maybe they pick on little kids at playgrounds. I suspect not, but I really don’t know these people (with the exception of Facebook inexplicably offering some of them as suggested friends to me on a regular basis—even though I’ve never met them); it’s rather beside the point regardless.

I freely concede that I am a minnow. I am a threat to no one in the discipline. I get interviews when there’s 13 applicants for a job, not 130. I don’t neatly fit any of the little boxes that define political science as a discipline either—being an “applied methodologist” who studies political behavior seems about as popular as being an H1N1 carrier. On paper, my position is probably just one or two steps above a community college job in the political science hierarchy; in practice, some days it feels like one (albeit without the fun paintball fights). I aspire to jobs that many of these snot-nosed brats wouldn’t even deign to apply for. So maybe I just don’t get why some graduate student’s success at an Ivy would be so personally threatening to anyone else.

I don’t know what the solution is here. Required registration drove down traffic, but it also drove up the level of discourse substantially. Perhaps the only solution is an economic recovery that lessens the perception of the market as being a totally zero-sum game. All I’m certain of is that a website like PSJR as currently constituted that makes me feel the need to shower after every visit isn’t one that’s doing our discipline—or anyone else, for that matter—any good.

Saturday, 17 April 2010

Conference paper done for now

You can now download a copy of my upcoming conference paper with Scott Huffmon and Adolphus Belk, “The Truth is Never Black and White: An Examination of Race-Related Interviewer Effects in the Contemporary South,” at the usual place. Both Scott’s and Adolphus’ contributions immensely improved this version over the previous iteration; of course, any remaining problems are clearly my fault alone, since I’m the only untenured co-author!

Monday, 12 April 2010

epcp for R

I finally have packaged up a very rough port of my epcp routine from Stata to R as part of a package unimaginatively called cnlmisc; you can download it here. In addition to the diagnostics that the Stata routine provides, the glm method includes a bunch of R-square-like measures from various sources (including Greene and Long).

The only part I’m sure works at the moment is the epcp for glm objects (including survey’s and Zelig’s wrappers thereof); the others that are coded (for polr and VGAM) are probably half-working or totally broken, and some wrappers aren’t there yet at all. The error bounds suggested by Herron aren’t there either. The print routines need a lot of work too; eventually it will have a nice toLatex() wrapper as well. But it beats having it sit on my hard drive gathering dust; plus I may eventually get motivated to write a JSS piece or something based on it.

epcp for Stata is still available at my site. For more information on the measure, see Michael C. Herron (1999), “Postestimation Uncertainty in Limited Dependent Variable Models” Political Analysis 8(1): 83–98 or Moshe Ben-Akiva and Steven Lerman (1985), Discrete Choice Analysis, MIT Press.

Thursday, 18 March 2010

Doin' it wrong

A mildly bemusing job ad that came across the wire today:

The Department of Government and Sociology invites applications as Course Redesign Coordinator. This is a non-tenure track, limited term, faculty position with the rank of Lecturer. The term is for a period of two years subject to re-approval and budget in year two. The successful applicant will lead a pilot study to redesign the introductory course in Political Science which is a required course in the university’s core curriculum. The position is responsible for producing an initial design for offering the course to larger sections while remaining consistent with the university’s public liberal arts mission; teaching one large (150 minimum) section of POLS 1150, Politics and Society, each semester; collecting and analyzing comparative data on student satisfaction and performance in larger course settings; supervising a graduate assistant and undergraduate student mentors ; preparing recommendation s for final redesign and implementation; conducting a required Freshman Seminar for departmental majors.

To review: this institution prides itself on its “public liberal arts mission” and excellent classroom instruction. So it is going to hire a non-tenure-eligible faculty member (who may not even have a doctorate) to come in to figure out some way to cram 150 students into an introductory course without any loss of quality. And once they’ve done this favor for the existing faculty, since they aren’t on the tenure track, they will be summarily kicked to the curb.

Somehow I do not expect this experiment to end in a rousing success.

Wednesday, 24 February 2010

Discussing discussants

Mike Allison and Greg Weeks are discussing the value (or lack thereof) of discussants on panels. Given that one of my major problems with the rising challenge to panels in our discipline, the similarly-poorly-attended poster session, is the lack of discussants, I can’t really concur in whole with Greg’s position that discussants aren’t helpful. I do mostly concur with his advice for discussants, however:

1. Do not try to tie the papers together artificially. There is no point.

2. Keep your comments as brief and focused as possible. No preambles or tangents. The audience did not come to listen to you, unless you are very clearly an expert on the panel’s topic.

3. Don’t whine about how long it took someone to get their paper to you. We’re all busy.

4. If time is short after the last presentation, give it up to the audience Q&A and give the authors your comments privately. Interested audience members very often have better insights.

That said, when I have discussed papers I usually try to see if I can identify common themes and ways the papers speak to each other, in part because I think scholars at the pre-publication stage can often strengthen their papers by looking beyond the literature they’ve embedded themselves in during the drafting process. Sometimes, though, that is futile on “potpourri” panels that often get titles like “New Directions in Research on X.”

Once upon a time (I can’t remember where; possibly at one of the iterations of the job rumors site) I saw a suggestion that took things to the opposite extreme—that panels might be better organized by having the discussant briefly present all of the papers, followed by feedback and discussion from the authors and the audience. It might be an interesting experiment to try, and I think it would certainly be a good test of whether or not the papers communicate their ideas clearly enough to their readers, although I think for it to work effectively you’d need to organize the conference in a way that completed papers would be due much sooner than is the norm in political science—where usually the “deadline” is enforced about as rigidly as most undergraduates would like their assignments’ deadlines to be.

Wednesday, 13 January 2010

QotD, deadweight loss of dead trees edition

John Sides on open access in political science:

Every political scientist should have a webpage where ungated copies of their papers and articles are available. Period.

(Alas, mine needs work in this regard, as most of my pubs aren't there in final form, but it will be better soon.)

Thursday, 7 January 2010

Like SPSA, but cheaper and less snowy

The slides for tomorrow’s presentation of my paper with Frequent Commenter Scott are now available online—not that they will make much sense without my allegedly-engaging patter attached.

Wednesday, 23 December 2009

Working syllabi for spring 2010

Preliminary syllabi are now posted at the usual place, although there’s a good chance the assignments may change—requirements that seemed appropriate for a 30-student senior-level class no longer appear quite so reasonable with 45 students and counting.

Tuesday, 22 December 2009

Lee Sigelman, RIP

Political scientist Lee Sigelman, probably best known as the past editor of one of the flagship journals of the discipline, passed away last evening. I never met Lee myself (the closest I got was hearing him speak at an SPSA luncheon keynote about a decade ago in Atlanta) but I was well aware of his contributions to our field and to helping to legitimize blogging among political scientists by helping launch The Monkey Cage with several of his GWU colleagues. His contributions to our discipline will surely be missed.

Monday, 21 December 2009

Yet another conference paper

My paper with Frequent Commenter Scott™ entitled “Can We Really Have a Conversation about Race? Investigating Race-of-Interviewer Effects in the Contemporary South” is now online for your perusal at the usual place.

Wednesday, 4 November 2009

On excessive moderation

As Steven Taylor notes, the third-party candidacy by Doug Hoffman in New York’s 23rd congressional district seems to have backfired, delivering a solid Republican seat for generations to Democratic candidate Bill Owens.

While some conservatives like my Twitter pal (and OG blogger) Jayvie Canono have suggested that Republican nominee Dede “Scozzafava would’ve been a vote for the Dems,” one of the iron laws of contemporary politics in the House is that the vast majority of the time, even the most liberal Republicans and conservative Democrats vote with their party. Would Hoffman have been a more reliable Republican vote than Scozzafava? Probably. But Owens, if he’s anything like the vast majority of his future colleagues, will almost certainly vote with the Democrats more than 90% of the time; even the most “disloyal” Republicans only break from their party around 35% of the time while the vast majority only defect less than 10% of the time. In other words, conservatives have probably traded a reasonably Republican vote in the House for a reliably Democratic one, which in the grand scheme of things is not likely to be smart politics.

Tuesday, 27 October 2009

Your spring 2010 textbook lists

I don’t have syllabi yet—the plan is to make a few changes to the spring 2009 syllabi but nothing radical—but here’s the list:

For PSCI 2306, Texas government (aka American State Government): Government and Politics in the Lone Star State, 6th ed. We had to pick the same book for all our sections due to Early College High School… and then they put all the kids from ECHS in the same section, defeating the purpose of picking a common book. I was using CQ’s Lone Star Politics before, which I liked and tried to sell my colleagues on using, but the lack of a test bank was the deal-breaker.

For PSCI 3320, Congress and the Presidency: The American Congress, 6th ed.; The American Congress Reader; and The Politics of the Presidency. I’ve been using the CQ Congress books for years but felt like trying something different for Congress, hence the switch to Smith, Roberts, and Vander Wielen’s books. New edition of Pika and Maltese, but otherwise no change on the presidency end, where I only spend a few weeks anyway—really I teach the class more focused on “Congress and Interbranch Relations.”

For PSCI 4320, Political System of the U.S. (which I basically treat as a political behavior class, since we don’t have anything on the books at the undergrad level that covers that stuff): Polling and the Public, 7th ed.; Public Opinion: Democratic Ideals, Democratic Practice; Citizen Politics, 5th ed., and Party Politics in America, 13th ed. I’ve decided to try to cut back on the voting behavior and interest groups material I was including before and focus more on parties and public opinion. This class was the least successful one last time around, in part because I was too ambitious in what I planned to cover. I’m also going to replace the research paper requirement with a couple of shorter papers, which hopefully will work better.

And that’s it; thankfully I’ll be doing my 3-class semester in the spring so I might actually be a bit saner and more prolific here and elsewhere (e.g. OTB) then.

Tuesday, 13 October 2009

Your well-deserved-Nobel roundup

Congratulations to political scientist Lin Ostrom and economist Oliver Williamson on sharing this year’s Bank of Sweden prize in economics in memory of Alfred Nobel for their work on understanding non-governmental and non-market mechanisms for overcoming collective action problems; Ostrom has garnered the greater attention in my circles, due to her being a political scientist by training (the second to win the Nobel, following Herbert Simon), being one of the pioneering female scholars in the social sciences, and being associated with Indiana University (with which I have what Mark Granovetter might call “weak ties”).

Those with further praise and discussion of Ostrom and her work include Steven Taylor, James Hanley, Mike Munger (who also links a lecture by Ostrom on sustainable development), Ilya Somin, Dan Drezner, Virginia Postrel, Lynne Kiesling, Margaret Soltan, Julian Sanchez, and Alex Tabarrok. Not to neglect the economists, Tabarrok also discusses Williamson’s equally valuable contributions.

Wednesday, 7 October 2009

NSF: Non-Sufficient Funds for Political Science

I find myself in total agreement with Dan Drezner’s thoughts on Sen. Tom Coburn’s deeply asinine proposal to save less than $10 million per year by eliminating the National Science Foundation’s funding for political science research.

Of course, you can probably take this as self-interested pleading since I have been the direct beneficiary of approximately $1500 in (taxable) NSF-funded stipend money and an indirect beneficiary of its funding by having access to the American National Election Studies. And I’ll freely concede that in my ideal universe, the federal government wouldn’t spend $10 million/year on political science research, but in that same universe the government wouldn’t spend hundreds of times that amount on Medicare prescription drug benefits, enforcing about 90% of the regulations of OSHA, building “infrastructure to nowhere,” bailing out every dying domestic industry, and subsidizing the activities of the world’s best-paid farmers.

And, I’d be rather more impressed if Coburn (or his staff) actually understood what the ANES did in the first place; the funding (all of about $1 million per fielded survey, which includes several thousand face-to-face interviews with voters across the United States) doesn’t fund data analysis but the collection of original data that nobody else collects because pollsters—unlike social scientists—don’t really care why people hold the opinions they have. If the ANES simply duplicated the work of Gallup, Harris Interactive, Zogby, Research 2000, and the dozens of other polling houses doing work for political candidates and the media I’d gladly agree that the spending was misplaced. And if the Census Bureau weren’t legally prohibited from collecting much of this information (for good reason, I might add) a case could be made that using their resources would be less costly.

But the reality is that the “basic science” that Coburn thinks is having resources diverted away from it for frivolous research on understanding our political system is the type of research with practical applications that has the least necessity for governmental subsidy. Materials science research on “bone that blends into tendons,” “next-generation biofuels,” and “microchip-sized fans” are examples of applied research that can be easily commercialized where the private sector is essentially freeloading on the taxpayer—I see no obvious reason why medical implants companies, major energy concerns, and Intel or IBM or Texas Instruments (respectively) couldn’t fund these research projects themselves since there are fairly obvious financial benefits to them in the short-to-medium term. Certainly there’s a better case to be made for “market failure” in providing most social scientific research than there is for “hard” science research—which still receives the lion’s share of funding and is often supported not just by NSF, but also the Department of Energy and other federal research funds that dwarf the $10 million/year spent on the study of political phenomena.*

If there is a valid critique to be made here, it is that the NSF has strayed from being focused on grants for “basic” science into the applied and pedagogical realms that are beyond the NSF‘s core mission and are best left to private industry and other government agencies such as the Department of Education, respectively.

More on this theme from Henry Farrell, Andrew Gelman, and Joshua Tucker.

Update: There are further thoughts in this vein from Steven Taylor and Charli Carpenter. Farrell points out that this isn't the first time the NSF political science program has been a target. And "Miss Self-Important" takes an ambivalent view.

Wednesday, 16 September 2009

RIP Neal Tate

Via a friend on Facebook comes the sad news of the passing of C. Neal Tate, who was a prominent political scientist at the University of North Texas (rising to Dean of Graduate Studies) before taking on the rather unenviable task of rebuilding Vanderbilt’s political science department in the wake of their bout with receivership earlier in this decade.

I only had the opportunity to meet Neal once, in the context of an APSA meat market interview for a position at Vandy, but in that interaction he was most cordial even though I probably had absolutely no business being interviewed for that position. Even based on that brief interaction, however, I am certain that he will be widely missed by colleagues and former students alike.

Wednesday, 9 September 2009

APSA recap

I really don’t have that much to say about my visit to Toronto for APSA; I was a rather bad political scientist when it came to attending panels, so I can’t report on much of the doings at those. Judging from the panels written up at IHE, I can’t say it seems like I missed much anyway. But I would be remiss if I didn’t say that the panel I attended discussing the future of the American National Election Studies was very informative and reassuring, considering that whether or not I get tenure is likely to ride in large part on the quality of the 2008–12 surveys.

Since all three of my official conference activities were, to borrow the colorful phrasing of IHE writer Scott Jaschik, conducted “in the faux privacy of a large room with tables, off limits to journalists,” I suppose I shouldn’t really spill the beans here about them. Suffice it to say I’ve learned enough in the past six years to know that reading the tea leaves of the interview room is virtually impossible—some meetings that have gone “well” in my opinion went nowhere, while some awkward meetings eventually ended with job offers. Hence the vibe that the discussion regarding the position I was most interested went the best is pretty much meaningless.

More likely of interest to readers: here are my photos from the trip on Flickr.

Wednesday, 2 September 2009

Your day-before-conference APSA links

A brief “I have three classes to teach today” roundup:

I have nothing in particular to add, except to say that most of my conference activities will be off-the-radar in one way or another. But any readers of more-than-passing acquaintance who are interested in coming to a Friday evening “recession-beating reception” may contact me via email for an invite, with the caveat that it’s a BYOB event.

Wednesday, 19 August 2009

Extrapolate this

It took me two years to finally get back to it, but the paper from my 2007 PolMeth poster is reasonably close to done after throwing out virtually everything I did for the previous iteration of the research. Now on to the year-old projects.

Thursday, 13 August 2009

Counterintuitive finding of the day

Andrew Gelman shows 2008 figures that indicate that there is virtually no relationship between income and self-reported ideology in the U.S. electorate.

This finding is subject to the usual caveats: namely, that everyone thinks they’re moderate, that most people really don’t think about politics in ideological terms (at least in the broad categories of “liberal” and “conservative”), and that these ideological terms themselves are fuzzy categories capturing multiple underlying political orientations to begin with. But at first blush it does contradict the accepted wisdom.

Monday, 20 July 2009

QotD, I need more "real" pubs edition

From today’s Inside Higher Ed, reporting on a survey of department chairs in The Discipline™:

[T]he survey found that the “scholarship of teaching” ideas of Ernest L. Boyer—in which colleges would see research and publication related to pedagogy or teaching as “counting”—has not been embraced by a majority of departments in any sector, and by relatively few at doctoral institutions. Asked if they agreed that “teaching publications and substantive publications are equal” in tenure reviews evaluating research, only 11 percent of chairs at doctoral universities agreed. (The figures were 32 percent for master’s institutions and 43 percent for bachelor’s institutions).

I guess I’d better get back to that “substantive” paper I’ve been cogitating on…

Monday, 22 June 2009

Psephological = pretentious political phraseology

Amber Taylor’s word of the day is one I’ve never used and hardly ever encountered. You’d think that was odd, since “psephology” is another name for my field of research, but I doubt most political behavior scholars could define (or even pronounce) the term. From a position of ignorance, I’d probably think it referred to reading bumps on people’s heads or something. (A Google News search suggests the term is reasonably common in India of all places, and gets some play in Britain and Ireland, but is rare elsewhere.)

In short: don’t expect me to order new business cards describing me as a psephologist any time soon.