19 October 2013


It is even suspected by some that Col. Peraza himself was in fact the agent of Santa Ana, instead of Yucatan. Should this suspicion turn out to be well founded (and I believe that servile and faithless people capable of the basest acts of national perfidy as well as of individual treachery), we will present to the world the unenviable and humiliating spectacle of a nation of anglo-Americans, gulled, entrapped, circumvented, and in her vital interests crippled by the perfidy of the imbecile Mexican.
--Congressman Darnell of the Republic of Texas, on the floor of the Texian House of Representatives, 14 December 1841

08 October 2013

The splendor of democracy

A nugget from the proceedings of the Third Congress of the Republic of Texas.
[Senator] Barnet submitted the following resolution, viz.: That the seat of the member from Harrisburg county [Robert Wilson] be vacated. Mr Barnet stated the ground upon which the report had been made.
  1. The repeated clamorous oaths made by the Senator from Harrisburg and invoking high Heaven to strike dead in their tracks all those who voted against him on the bill under consideration of the Senate on Monday.
  2. His peremptory refusal to come to order, saying he would be God damned if he would come to order.
  3. His refusing to seat himself or be seated by the Sergeant-at-arms; and ordered the Sergeant-at-arms to stand off, and not touch him, swearing that no power but God's could seat him.
  4. Disclosing secrecy, contrary to the meaning and intention of the 17th section of the constitution.
Even at the beginning, Texas was Texas.

22 September 2013

"That Bitch-Goddess Quantification"

The response of the historical profession's elite was rapid, but by no means single-minded. To the sometimes strident demands of the devotees of the new history that traditionally trained historians "retool, rethink, reform, or be plowed under," as one older economic historian caricatured the new program, some historians at first reacted with fright, irrationality, and something close to panic. Arthur Schlesinger, Jr., whose description of Whig and Jacksonian electoral coalitions had failed Benson's systematic numerical tests, retreated behind a hastily erected wall of dogma. "Almost all important questions," Schlesinger proclaimed, "are important precisely because they are not susceptible to quantitative answers." In a presidential address to the American Historical Association, Carl Bridenbaugh issued a jeremiad against the infiltrating priests of the new religion, warning his fellow historians never to "worship at the shrine of that Bitch-goddess, QUANTIFICATION."
From J. Morgan Kousser, "Quantitative Social-Scientific History," in The Past Before Us: Contemporary Historical Writing in the United States, ed. Michael G. Kammen, p.434.

18 July 2013

Don't let your technology dictate your theory

[T]he electronic media circulate[s] opinions rapidly, widely and effectively [but also fosters] opinions which are necessarily attenuated and transitory. The spoken word takes much longer to communicate than does the written word, so that there is not time enough for extensive and thorough coverage of a situation by electronic media. Even hourlong accounts, attempting to provide background and detail, are inevitably truncated. Thus instead of being presented as a complex issue, with intricate sources and ramifications, the situation is reduced to a problem with one or two major aspects stemming from one or two causes and involving one or two potential consequences. In addition, electronically distributed opinions do not remain very long in the circulatory system. Except for tape recordings in the possession of broadcasting companies, electronically circulated opinions cease to exist as soon as they have been transmitted. That means that if an opinion-holder's mind wanders during the broadcast, or if he is confused by the commentary, he cannot rectify his absent-mindedness or his confusion. There is no way of turning back the radio or television dial and listening to the account again from the beginning.
--James Rosenau, Public Opinion and Foreign Policy, 1961 (6th Ed., 1968) p. 80 Rosenau is brilliant and I am not. But one can't help but ask the question that didn't occur to him: What if any of these things had changed? What if, for instance, broadcast technology DID involve a level of permanence similar to that of text? Would that change things? (I actually doubt it has, all that much at least.) Too science-fictional, you say? Well, Aristotle discussed machines running themselves in the Politics, in a passage that reads as fresh as if it were written yesterday. The moral, for myself: Think through the implications of shifts in your assumptions.

05 July 2013

Thoughts on Teaching Introduction to International Relations

Some thoughts about three years of having taught Introduction to International Relations. These are notes for myself before the next time I teach the course.

When I was hired to teach the class for the first time, I had a clear vision: I wanted to teach theories that were relevant to helping undergraduates explain the world as it is and as it is likely to become. Instead of recounting the details of the First World War or expounding upon the design of the post-Second World War institutions, I would only invoke those events sparingly. Instead, I would focus on unipolarity and its consequences, structuring the course around the implicit question of what would happen once the current era was over.

First, I felt an emotional intuition that courses that took the Cold War and "bipolarity" as their baseline were missing the point for most students, for whom even the world before September 11, 2001, was distant. For students now, the 20th century is as remote intellectually and emotionally as the 19th century was for me as an undergraduate. Consequently, drawing on examples from before the fall of the Berlin Wall or even of the Twin Towers was leaving students cold and confused.

Moreover, I was motivated by an intellectual conviction that the incumbent courses at my institution, as productive as they were, were too focused on explaining theories that were comparatively irrelevant to the decisions that states make now--especially as regards their lack of emphasis on topics in political economy--and that were too focused on the world of the past. And in any event I knew that I could no more teach a course that required me to be an expert on World War I historiography (or whatever) than I could teach a convincing course about democratization; the training and the background reading that I would have to undertake in order to feel like I was truly an expert in such matters vastly outpaced the budget of time that I had for course prep. I would go to lecture with the knowledge base I had, not the knowledge base I might want or wish to have at a later time.

Finally, I thought that the pedagogical content of most such courses was divorced from the empirical and theoretical content of "real" political science, and that we were doing a disservice to students by presenting many of these theories as faits accompli instead of showing how models could be applied to and derived from real-world situations--and also demonstrating the advantages and the drawbacks of thinking theoretically. A greater engagement with abstraction would, I believed, lead to better development in thinking critically not only about international relations but about the world (the philosophic "world", in this case, not just the international system) more generally among my students.

Having taught my version of this course two and a half times (twice to classes in the 20 to 30 range with adequate prep time and once to an overflow session of 10 students on a week's notice), it's time to evaluate how well my intuitions described the reality I've faced, how realistic my vision was, and how well my strategies for implementing it have fared.

Assessing My Intuitions

History and Intro to IR

My first intuition was that students knew little and cared less about the past (defined for my purposes as the world before the end of the Cold War). I'll talk later about why I feel that political scientists should not see part of their mission as inciting a love of the past as such; right now, I only want to assess how well this intuition described reality.

The answer? It described reality almost perfectly. I have never had a student wish that we had spent more time talking about the Cold War. The Soviet Union is as remote as the Achaemenid Empire to today's undergrads (and I teach at a top-25 institution, so I'm pretty sure that this generalizes very well). I should note, by the way, that this breaks my heart; about twice a week I wish that I could teach a course in the style I imagine that IR was taught in during the 1970s, where we'd all get together, talk about Bismarck and Thucydides, and then, presumably, repair to the sitting room for brandy and cigars.

I want to forestall all misunderstandings: I don't like this state of affairs and I wish that this weren't so. I wish my students could reliably distinguish Gorbachev from Brezhnev, not to mention Walesa from Jaruzelski or Honecker from Hoxha. But they can't. Nor, in some sense, should they be expected to. Not only do students infamously never get beyond World War II in their high school history classes (which are usually hideously taught, anyway), but when I was an undergrad and high school student (the latter only five years past the fall of the USSR!) I didn't much care about American foreign policy before Pearl Harbor.

Political Economy and IR

My intellectual critique of the content of introductory courses was that they were too focused on theories applicable to a world of multipolarity and bipolarity, especially in their security concerns, while simultaneously giving short shrift to theories about trade and globalization. Teaching students about alliance politics is all well and good, but alliances between superordinate and subordinate states don't work like alliances among relatively equal-status great powers. (Does NATO look like the Triple Alliance? In fact, do any international formations look like their nineteenth-century equivalents? If not, then why should we bother teaching about them?)

Similarly, there was a weird and unstated assumption that the world after the Second World War was simultaneously both a bipolar and a unipolar world. Bipolar, because of the Soviets, but unipolar, because the United States was held to be underwriting the liberal order among the "free world". To my recollection, this basic tension--the fact that the eras that we were studying were held to be explained by two extremely disjoint theories!--was never recognized or addressed. (And let's not get into the somewhat tenuous notion that the USSR actually posed a threat to US hegemony after 1945, which increasingly seems like a relic of Cold War hysteria than an accurate description of the world system circa 1960 or even 1980.)

This critique also had a positive component: that an IR course should describe what states do now and that most of what "great powers" do nowadays is economic stuff. (I use "stuff" broadly because that includes more than trade or capital but also all of the knock-on effects, from migration and remittances to illicit smuggling to domestic grievances exacerbated by trade.) In other words, I would excise a lot of the guns-and-bombs (but by no means all!) and spend more time talking about dollars and cargo containers.

This intuition is hard to assess. The "right" amount of time to spend on any subject is, of course, a matter of professional discretion and, ultimately, taste. But have students responded well to this? The answer is much more mixed. Students are deeply confused about why we spend so much time on trade politics within states (especially rich countries). This might reflect on the structure of the course or my own failure in tying the content of domestic contestation to the content of international trading relations. (The shift from states-as-actors to actors-within-states is pretty analytically ambitious.)

But students are also more engaged with the spinoff topics of the international economics section than they are with any other part of the course. A section I introduced this year on the Mexican drug war and the dark side of globalization was so successful that I intend to rework and expand it next year. (Anyone know any good IR scholarship on nonstate actors who are normatively bad?) As a footnote: Given my deep skepticism of the influence of transnational activists and my belief that their influence is much more due to the liberal properties of the incumbent unipolar order than to any intrinsic property of transnational lobbying, my discussion on nonstate actors is much different than many instructors'; to put it another way, the fundamental text for the NGO lecture is more Cooley and Ron than Keck and Sikkink (which is discussed and assigned, of course).

In other words: sexycool crime and NGOs? Awesome! Meat-and-potatoes trade stuff? Not so much! But that might be a consequence of poor exposition. And in any case, given my expectations about the way in which the world is likely to develop, I am not yet ready to fundamentally rethink my weighting of the course, even though (as I discuss further below) I am rethinking how I present the course material.

Teaching Critical Thinking Via International Relations

I think about a number of models when I teach and think about pedagogy--people who I'd like to lecture like, or write exams like, or choose readings like, or, most of all, educate students like. Some of them--Dan Nexon, PTJ, Tim Burke (at a distance)--are well known in the blogosphere; others are not. But rarely (ever?) do I teach a section or evaluate a reading without thinking about how one or many of these people would approach the same issue or problem.

One principle that all of these folks agree on is the importance of using the classroom to inculcate critical thinking. The paradoxical point is that science classes in undergrad may be the least critically-aware courses I know of (you may have to think logically but there is almost always a "textbook answer". By contrast, my beau ideal--the (very) idealized versions of computer science classes that I carry around in my head--is of something like an extremely rigorous poetry workshop, in which creating is melded with logical rigor. (Thinking about it explicitly, this is pretty well the model that Paul Graham espouses in "Hackers and Painters".)

There's no way to duplicate that kind of Nirvana in my courses. Can't be done (given any feasible investment of time and effort): Even a student-to-teacher ratio of 25 to 1 exceeds the amount of time that I could devote to sitting with each student to craft a useful, rigorous final project. Especially, I should add, for beginners: I don't know how to convey a relatively fixed amount of knowledge to students while also allowing them seminar-style freedom to pursue an independent project. (For upper-division courses, I'm much happier conveying some skills but spending more time helping students acquire the skills their ambitions suggest.)

With that said, there's still a lot of ways to begin to get students to think theoretically. The easiest way to do this is to begin with increasing the writing content in the course. In this, I'm inspired by a lot of folks, but principally Derek Bok in Our Underachieving Colleges, which makes the argument inter alia that students don't write enough, that they should write a lot more, and that the core of developing good writing habits is having your composition exposed to editing. In response to that, I added four 2-4 page papers to this semester's class; four was too many (due to a last-minute and unavoidable rescheduling of the midterm, we ended up with a paper due on the same day as a midterm). On the other hand, the students got much, much better in writing over the course of the term, especially the high schoolers (the contemporary university summer course seems much more full of high schoolers than they were when I was taking them). Moreover, they began to make more connections across the readings--and, I suspect, to do more of the readings--than students had in previous years.

But I didn't just want to introduce more critical writing in the abstract. I wanted to present political science as she is done, which means understanding some theoretical model in at least some detail. (I stole this idea from Phil Arena's blog posts breaking down popular models and building them back up again.) Again, this was very much a strategy I adopted based on what I could teach; if I could teach feminist or critical IR, then I would do so, but that wasn't in the cards. Hence, my lectures on Fearon's bargaining model of war.

The bargaining model of war! You do not know how inaptly named this model is until you have watched undergraduates--who have been exposed to the model via the original article (edited, but lightly, by Mingst and Snyder), the course textbook (Frieden, Lake, and Schultz), and a lecture (by me)--completely fail to understand it. If something is named a model of something--let us say the Bargaining Model of Parcheesi--then it should involve that something as a natural outcome of the model. Yet in this case the bargaining model of Parcheesi should always end with the parties never playing parcheesi. Partly, this is due to language difficulties--ever consider that the English phrase "prefers X to war" might be confusing to someone who is reading [[to war]] as an infinitive verb and not a comparison?--but partly it is because the only way war emerges from the Fearon model is via a violation of the model! It is actually a bargaining model of not-war, and I have finally figured out why this is so counterintuitive to undergraduates after only three progressively more frustrating attempts.

It is hard to get to "critical thinking" when you are stuck trying to explain the basics of spatial modeling. (This isn't limited to IR, either; I once watched a brilliant Americanist explain the Hotelling model of voting to advanced undergrads, in a smaller class, and then watched those students mostly fail to get it as well. They got it, eventually, but I should have learned from the incredible investment of time and creativity that this required.) The concepts of the ideal point and a utility function increasing in negative distances are no more intuitive than the IS-LM model was when I learned it as an undergrad myself. After three attempts, I think I have finally hit on a good way to get this across--it involves a fake war between Colorado and Missouri for the (nicely rectangular) state of Kansas and a whole lot of transparent overlays--but I will have to wait until the next time I teach (that is, if there is a next time) to see if it works better. (We also roll through the typical game-theory topics, the security dilemma, basic collective action, H-O and R-V models, and so on, but rationalist explanations for war is the signature model we discuss.)

In general, students are resistant to theoretical explanations. Recall that this is not a course for majors--the majors or proto-majors or coulda-been-majors are always pretty able to roll with the punches. My concern is with students at the 75th or 80th percentile of interest, who are interested enough to show up and take notes but not so interested that they will teach themselves what I fail to convey. By the end of the class, when we roll back into questions about what the future will look like, they have normally come around to the point that distinguishing realist or liberal perspectives is pretty well second-nature, but for the first few weeks it is hard to explain why they should care about ideas like the security dilemma.

Of course, the whole reason that I wanted to teach them modelling and the basics of working through theory is that thinking in abstract terms is not natural. It is not easy. And it demands a kind of engagement with the actual logic of arguments that many of them have never had to develop before. (This is why I think that crits could do as well as formalists on this dimension; anything that requires axioms, corollaries, and conclusions clearly stated would serve the pedagogical task well.) But this is, I've found, an uphill battle.

Implementing the Vision


Lecturing is hard.

Oddly, graduate school doesn't teach you much about lecturing. I mean "lecturing" as distinct from "pedagogy"; frankly, nobody in academia seems to know much about pedagogical research, but having skimmed that research I am not sure there is much to learn. (What seems to matter, from the longitudinal studies I've read, are having good teachers and adequate prep time, but the "good" in "good teacher" seems to spring forth fully formed, like Athena, from the brows of wise deans.) No, I mean specifically lecturing--that thing that we do in class.

This is a puzzling gap. We're trained on how to give conference talks, which we'll give perhaps four times a year (and the median performance there is still pretty bad), and extensively workshopped in giving job talks, which my generation will probably give (at the median) about three times. But lecturing, which is our most common vehicle for expressing scholarship, is never clearly discussed. What makes a good lecture? What is an appropriate amount of time to write a lecture about a subject you know well--or, as happens to us all, about which you know almost nothing?

In part, this is because the "apprenticeship" model of graduate school--the legal fictions that mean that schools like mine can get away with often irritating and occasionally abhorrent treatment of grad students (I hasten to add, entirely by midlevel administrators)--is terribly designed. There's no true apprenticeship here, in the sense that would be recognizable to a silversmith or even a contemporary doctor. I've only ever twice had a faculty member observe my teaching (once in my classroom and once as a guest lecturer) in the course of having taught four solo courses and TA'd (with sections) another three. I've had lots of discussions with faculty, of course, and those have helped immensely. Nevertheless, that's pretty far from the immersive engagement that "apprenticeship" suggests. Again, I don't blame my institution for this---the structural conditions of graduate school are pretty well the same most places, so why single out an individual institution? Yet we shouldn't let the structural conditions off the hook, either; there really is no good reason why (for instance) job postings outside of the research-only universe shouldn't require a teaching evaluation from a tenured or at least tenure-line professor (perhaps even from a different institution?).

I've learned a lot about lecturing. Mostly, I've learned that they are pretty awful, but that they can sometimes be good. Charisma doesn't matter, thankfully--except for student evaluations, which is another way to say that charisma matters a great deal. Similarly, lecture may actually be easier on students than sleeping (humorous and serious links embedded).

With that said, I'd still like to become a better lecturer. I find myself looking at other people's lecture notes--everyone from Simon Jackman and Cosma Shalizi down to more local luminaries--with envy: how do they do so much in 60 or 90 minutes? And I'd like for my lectures to be the same model of erudition, serving both the students by surveying vast areas of human inquiry and boiling them down to the most important principles, a couple of good illustrative cases, maybe a data visualization or two, and then the remaining areas of contestation---and myself, by serving as lit reviews for the articles I'd like to publish. But that takes so much work that I find myself in even greater awe of how deep the comprehension of the field that my elders possess actually is. The gap between my learning and theirs makes me feel like my junking of the First World War, however justifiable (and I think it is!), is just laziness. (It isn't. It really isn't.) And I wonder just how much of my time I should be assuming will be devoted to lecture prep when/if I get "a job."

With all that nerve-wracking, though, I do know that I can move about three or four lectures a year into a much higher state of revision; I now have three or four lectures I'm pretty proud of and would be happy to give to an audience more or less anywhere. And I also know that many lectures are rarely updated once written. But for all that ...


Assessment is a critical part of any course. Indeed, in some classes, you can solve the lecture dilemma--they are expected but rarely effective--by basing a course on problem sets (essentially, the "flipped classroom". Problem sets are great. The literature I've skimmed from science pedagogy suggests that they are basically the only way to actually teach anybody anything. I am not sure if essays are the precise equivalent of problem sets or not; I incline toward "no", but I've never seen a problem-set based intro poli sci course. That's no reason why there shouldn't be one, of course, except that when I tried to do one in a quant methods course I received nothing but complaints from the students. (That the students in question learned rather more in their semester of stats than I did in my first year of graduate stats seemed beside the point, as did the fact that the only section of the class for which they were not assigned a problem set was exactly the one that they learned least satisfactorily.) Humanities and social science students are, in my experience, comparatively adverse to a model of pedagogy and assessment that requires a constant stream of work, preferring instead to binge before exams.

Ideally, I'd actually like to do something like Tyler Cowen's final exam. (The aforementioned stats class, with most of its grade based on a final project, came close to this--a harder project executed to a higher level of skill received a better grade than a safer project executed competently, without any need for me to design a "final project" for the students.) In fact, I'm still not quite sure why it's not "done" to simply hand out some bluebooks with two questions:

  1. What is the most important unsolved issue in this subject?
  2. How could we solve it?
OK, that's not true. I do know why it's not done. Thirty to forty percent of the class would spontaneously combust and the remainder would mark heavily against the instructor on the "Was the grading fair?" feedback form.

As a body, students hate assessment, and they will hate any form that is minimally stringent. Perhaps that is too much of my post-exam-writing stress being released, but it really is the case that I've found it difficult to write tests. For convenience (and to keep the fair-grading numbers up), in the summer I use multiple-choice exams, but they are not nice multiple choice exams---the form "Which of the following is the BEST explanation of X" or "Which of the following is NOT an explanation of Y" seem to lend themselves very easily to discriminating between mastery and mere competence. Similarly, my Eric Mazur-style free-response questions---employing word problems or transpositions that are isomorphic or nearly so to examples to class, but without textbook-style key phrases---are unpopular with a large mass of students, because they think it is unfair to ask questions that are not explicitly covered in the text. The clever students, by contrast, seem to greatly enjoy those questions, but the circularity of applying "clever" to the successful students in that last phrase is, of course, fairly obvious and perhaps self-congratulatory. With that said, though, the Mazur questions have been great so far for me, in showing more precisely where students' understandings are breaking down and, gradually, helping me change my presentation.


I like to think I'm a good teacher. Of course, all of us do, and the lack of any sort of long-term evaluations of our actual effectiveness in the classroom allows us all to persist in our happy--and tactfully private--delusion that we're better than our colleagues. ("Ah, yes, she gets higher rankings than I do, but do her students really understand the Fashoda crisis?") Regardless of my comparative ability, however, it seems pretty clear that my Intro to IR course needs a bit of an overhaul.

What will that look like? Well, I'm unwilling to let go of my core philosophical and pedagogical wagers. In fact, I have been confirmed in some of them--I think that there is an even greater need to get students writing, thinking, and through that independently engaging with the course material. In others, I am less sure; one of my original strategies was to try to use as much original political science material as possible, but the intersection of "quality political science" and "material 18 year olds can understand" is pretty slim. This might, in fact, require developing some of my own materials (e.g.) because a lot of what is out there, even in compiled form, is pretty bad. Rogowski, for instance, is great on IPE and distributive politics--but I'm deeply unsatisfied with having my principal text on trade politics, even in edited form, rely so heavily on comparative nineteenth-century history, or, for that matter, on a comparatively difficult application of the Stolper-Samuelson model. Land-labor ratio--yes, that's really intuitive in a two-good, two-factor world! And every year, I create a lot of offensive realists simply because John Mearsheimer is a better (= more persuasive, more accessible) reader than either Doyle on liberalism or Wendt on constructivism. So it is time to go hunting for newer, better readings. As always, it is also time to consider dropping the textbook--but the text is so good at conveying background information that doing so would quite possibly cause more harm than good. (The Drezner Zombies text is safe indefinitely.)

I made a bad choice this year in dropping the Nexon and Musgrave (and Motyl and ...) lecture on empire and imperialism in favor of keeping the Christensen-and-Snyder reading on alliance dynamics. The latter is a great piece of theory (especially for undergrads) but if I don't think that the alliance conditions of the pre-war period are likely to return, then why am I teaching it? That's all the more true if I do believe that the structure and conduct of unipolarity does matter. That's also a strong argument for doubling-down on power transition theory (at least in an expositional sense).

As I mentioned before, the section on political economy needs to be reframed, at least to reinforce the division between why state-centric theories should lead to free trade (or for realists protection) as an unproblematic outcome and why the current liberal order is in some sense puzzling. Students greatly enjoy learning about the "Third World", especially the dark side of globalization; more attention to this, and to the scholarship in this area, would be a good idea. Similarly, I need to have better explanations about the degree to which any given economic ill really is attributable to globalization. And my globalization story needs to truly integrate discussion of capital flows and labor migration instead of leaning so heavily on the textbook; perhaps using the Frieden, Lake, and Broz reader as a required text would assist better.

I do poorly on international law, but that is unlikely to change without a brain transplant. I could do much better at presenting descriptions of rising powers' foreign policies from the rising power's point of view; doing so, in fact, could make the foreign-policy lecture less of a throwaway and more of an integral part of the course. (Perhaps Goddard on Prussia and Zakaria on the United States? At least everyone is equally ignorant about those cases.) And I could do better at re-introducing constructivism via a discussion of AIJ and similar works.

09 February 2013

Sadistic Grading (Or, How Apparent Sadism Is Tough Love)

Let's imagine we have a professor, P, and a student, S. The final grade for the class is the result of a test consisting of two questions, Q1 and Q2, such that G = Q1 + Q2 and 0 ≤ G ≤ 100. Each question is objective and the student's performance on each is given by Gi = Ei + Bi, such that Ei is the effort the student has invested in the question (a variable) and Bi represents the question-specific baseline score the student will receive regardless of any effort invested (B is thus a constant). In the absence of any studying, the student will thus receive a grade equal to B1 + B2.

The student is crafty but lazy. He knows his baseline levels of competence with certainty. The student has a fixed budget of time T in which to allocate E (total effort) and R (recreation). The student's utility is maximized by selecting an effort level E = E1 + E2 that minimizes E while guaranteeing that he receives a grade Gpass, the passing grade in the class (0<Gpass<100). (In other words, the student's utility is increasing in R, decreasing in E, and increasing in G up to a point.) Consequently, if the student receives a maximum score on one question, he knows he must only invest enough effort into studying for the second one such that Gpass= (G1+G2)/2.

The professor is altruistic but bad with people; she wants to maximize the student's total learning but is both indifferent to the share of time the student invests in recreation (which she views as wasted anyhow) and does not know which question the student is better at answering. If the professor believes that learning is a nondecreasing function only of effort invested, and if the professor derives utility only from the amount her students learn in the class, how can she design an incentive to maximize the student's time spent studying?

The professor announces that the final exam will continue to consist of two questions, but that the grade will now be based on only one question to be chosen with probability p, 0 < p < 1. (The credibility of this threat may be enforced through several ways, including the use of a binding agreement with a colleague or a reputation for whimsical or arbitrary behavior.) In expectation, the student's grade is now G = p*Q1 + (1-p)Q2.

Let us assume that under the original policy the student at baseline levels would have received a score of 90 on the first question and 60 on the second, for an average of 75. If the passing mark is 70, then the student need not study any further. But if there is a 50-50 shot that each question will be graded, then the expected grade remains 75 but the expected utility payoff is much different. He retains the benefits of his time spent in recreation (R) but he now has a 50 percent chance of suffering disutility from failing the class. Assuming the total utility from R is less than the utility from passing the course, he should instead invest a sufficient amount of studying in his second question.

This scheme has apparently been tried at least once, although the student seemed unappreciative of the professor's altruism. In real life, I suspect it would prove unworkable--it might well lead to too many false negatives. On the other hand, its logic seems no less arbitrary than the act of actually designing an exam in which the questions are unknown to the student until the moment he or she opens the blue book.

26 December 2012

Structure Induced Equilibria in Everything, Peloponnesian War Edition

From The History of the Peloponnesian War:
The territory of Athens was being ravaged before the very eyes of the Athenians, a sight which the young men had never seen before and the old only in the Median wars; and it was naturally thought a grievous insult, and the determination was universal, especially among the young men, to sally forth and stop it. Knots were formed in the streets and engaged in hot discussion; for if the proposed sally was warmly recommended, it was also in some cases opposed. Oracles of the most various import were recited by the collectors, and found eager listeners in one or other of the disputants. Foremost in pressing for the sally were the Acharnians, as constituting no small part of the army of the state, and as it was their land that was being ravaged. In short, the whole city was in a most excited state; Pericles was the object of general indignation; his previous counsels were totally forgotten; he was abused for not leading out the army which he commanded, and was made responsible for the whole of the public suffering. He, meanwhile, seeing anger and infatuation just now in the ascendant, and of his wisdom in refusing a sally, would not call either assembly or meeting of the people, fearing the fatal results of a debate inspired by passion and not by prudence.
Let's imagine the range of policies that the Athenians could have adopted. Trivially, they range from "arm everyone and sally immediately" to "surrender now;" Thucydides implies that the relevant policy range is from the Archarnians' ideal point ("send as many men as necessary to secure the immediate neighborhood") to Pericles' ("wait until the situation has resolved itself"). Under the Athenian constitution, presumably, Pericles has the right to call the assembly to adopt a new policy, but he knows that in the heat of the moment they will adopt a policy much closer to the Acharnians' ideal point than to his. Pericles suspects that over time Athens' allies will reinforce him, and at the same time that the Spartans will tire of offering battle without a response. The greater strategic flexibility of the Athenian navy also (he believes) offers him the ability to choose when and where to strike at the Peloponnesians and their allies, whereas giving battle to the Spartans outside of Athens risks everything. The ability to control the agenda of policy is crucial but it must be nerve-wracking to exercise. Thucydides captures the dilemma here well. Over time, Pericles's policy (in this instance) will be proven to be the correct one, but implementing it both requires bearing immediate costs and a reliance on the formal institutions of Athens. Had the policy failed, it would have failed catastrophically, with Pericles removed by irregular means and Athens itself in jeopardy.

26 September 2012

I'm Happy To Bash Woodrow Wilson

America's worst postbellum president.
From Will Riker, Liberalism against populism:

"In a sense, political science and political events have passed the adherents of 'responsible parties' by. It may seem foolish to waste time now on their mistake of expecting a clear binary choice. Still, we can understand their mistake a little better today. It is not simply an empirical matter--as Dahl, for example, argued--that the American (or any other) system does not work in the neat way the Wilsonians wanted. More important is the fact that what they wanted was itself morally wrong from a liberal democratic view, because to get binary choice one must enforce some method of reducing options. This is precisely the coercion involved in populist liberty. The populist expectation of a coherent program is not attainable by weaving together individual judgments, either for a small group or for a large nation. To attribute human coherence to any group is an anthropomorphic delusion. Worse, however, the populists expected to calcify politics on their own issues, progressivism or the New Deal. That is fundamentally unfair. Winning politicians are, of course, happy to stick with the issues on which they are winning. Losing politicians are not so entranced with that future. The losers of the last election become the winners of the next by changing issues, revealing new dimensions of choice, and uncovering covert values in the electorate."

Will Riker was not only a damn fine first officer, he was a damn fine political scientist. And passages such as this one should both refute the quantoid delusion that our work is value-free and the normativist's delusion that he may be uninformed about the work his formalist colleagues engage in.

30 July 2012

My lectures gone by, I miss them so

I'm changing this year's IPE lectures a bit to downplay the details of negotiation and to bring interest-group explanations for protectionism to the fore. But that meant that I had to lose half of one lecture, which was a lot fun--not least because the last time I taught it, I delivered the lecture just as the budget shutdown was about to take place. And so no more prize fights...

I should note that this was my favorite slide of the semester largely because it broke every rule I have for my slides. (Well, all but one: I only use hi-res jpgs in designing my slides.) Garish colors? Check. Weird fonts? Check. More than two fonts in one slide? Check. Absolutely zero information conveyed? Check. And although you can't tell from the screen grab, the various elements of the slide enter in via every terrible transition feature that Apple has to offer--explosions, dropping with a thud, and maybe even the lens flare.

Given that I created something like 500 slides for the course, it was a lot of fun to have one intentionally terrible moment ...

29 July 2012

Q. Are we not researchers? A. We are professors

I've been thinking about 80s music a lot lately.
 What follows is a digressive and only partly thought-through exploration of a tough topic. I'm posting it as an excuse to work through some of my own thinking about the responsibility of academia to the public, the scholar to the student, and the discipline to all other parties. I hereby reserve the right to admit that I'm completely wrong about everything.

Why should the public pay taxes to support our research?

This used to be an academic question, but officeholders like Tom Coburn and Jeffrey Flake have put the question squarely before political scientists. Other disciplines have faced this moment before, as physics did in the late 1980s and early 1990s, and still more will face it in the future. My reasoning is plenty motivated, so I think that the case for funding basic academic research is pretty good. Like a lot of social scientists, I'm frustrated by the often self-inflicted lack of solidarity among academics ("Which social science should die?") and the professoriate's inability to sustain a campaign to make our arguments. Even the political science blogosphere's concerted arguments in favor of NSF arguments probably did less to move the needle than one foolish New York Times op-ed.

The outlook for funding for the social sciences and academia more generally is a lot dimmer for my generation than it was for the two or three generations of academics before us. At worst, they had to look forward to long periods of stagnation (unless you were a Sovietologist); at best, you could go from nothing--literal non-existence--to a seat at the grown-up table in the time it takes some disciplines' doctoral students to finish a dissertation. As the funding pool dries up, there's bound to be a lot more friction and fighting among academics than there was when the only question was how to divide up the increasing pie.

The situation won't get better on its own. I'm unconvinced by the idea that state governments' spending will naturally pick back up as the economy does, since it relies both on the idea that the economy will pick up--green shoots!--and also that politicians will refrain from directing that savings from their structural adjustments toward other priorities. And even if it does, higher education is under increasing pressure to justify its business model in every dimension:
  • Why should college be a physical location, and not an MMORPG-cum-TEDtalk?
  • Why should professors be hired and promoted based on their research skills--most without ever having had any practical or theoretical training in pedagogy--when the public's understanding of funding universities rests on their teaching productivity?
  • Why should there be an expectation that tuition should rise faster than inflation?
  • Why should an English degree take as long to complete and cost the student as much as a physics or an engineering degree?
All of these are different ways of asking the "why-pay-for-my-research" question. But historically we've only had to answer one or two of them at a time. And now we have to respond to all of them in a way that recognizes that answers to questions that were serviceable ten years ago--in 2002!--may now be unpersuasive or unrealistic.

At root, we largely rely on one of two arguments to answer all of these questions. The first is that the knowledge produce is a public good, and that institutions are organized to support our knowledge production for the maximal good of society (or, at least, that reform would cost more than it would gain--that we are in a stable local maximum). The second is that our research leads to better learning outcomes for our students. Winning the future, then, requires academics to have maximum time for research, even if there are short-term tradeoffs for the students.

I think the first answer is better, since it does at least appeal to the innate beauty of knowledge, but it tends to obscure the third reason that we voice silently--that learning, like art, is a luxury good. We can't say that aloud, since it carries with it the implication that  opposition to learning is opposition to civilization itself, and that the superior culture is the more refined culture. That is a line of argument that has never played well with the citizenry, and these days would not play well with the mythical one percent.

The second answer--better research, better students, better jobs--is, to put it bluntly, not overly persuasive. It is, in fact, close to being faith-based. We expect better evidence and stronger argumentation from our students when they turn in term papers than when we produce public justifications of our budgets. Why should we ask students who want to be lawyers to sit through four years of an undergrad degree and then three years of law school when the British and others manage to produce serviceable advocates after three years of law-focused undergraduate studies? And much of the other justifications for our teaching--our students learn critical thinking! statistics! data visualization! discipline!--only beg the question of why students should learn those skills by majoring in our subject instead of majoring in something more practical in which those skills are foregrounded.

There are answers, of course. One part is that it's that the ivory-tower academics are not the ones who are out of touch with the real world. Rather, frequently it's our critics who deride us for not taking "real-world" problems seriously enough who have no idea what the real world is actually like. This sounds counterintuitive, but sober reflection or residence in Washington (or Jefferson City, or Albany, or wherever legislators gather) will quickly remind you that politicians and activists who make these claims frequently have no idea about what either academics or businesspeople or bureaucrats actually do. The "soft skills" matter, and those are hard to impart at a distance or by paying professors less than a living wage. As Tim Burke, the charge of know-nothingism applies with especial force to fact-free discussions about information technology and teaching.

A second part would be to forcefully stress how social science matters by discussing our research. This is easy in the abstract but harder concretely. Yes, individual scholars and individual projects make tangible contributions, but how many of us are willing to say there are scholarly consensuses about issues in social science as there are in comparably observational sciences like geology? Where we are most consensual, we are least interesting; where we are most interesting, we are least consensual--and often (as with the democracy-promotion agenda of the Bush administration, or the debate over whether there is a resource curse) most dangerous. So we're left with the unsatisfying choices of either "teaching the debate" or picking sides among competing findings where there is real and lasting disagreement. (One way of squaring the circle is to find and promote popularizers of our research--popularizers in the real Carl Sagan or even Stephen Jay Gould tradition--and let them handle the rhetorical aspects.)

But the third part is to recognize that we should probably systematically invest in and require better teaching and measurements of teaching effectiveness. We measure research productivity decently well--by no means as well as we should, or as we as we ought to given the importance of the measurement and the time we've spent on it--but we rely on ... student surveys to monitor teaching effectiveness. We lack a well-defined progression of introductory, intermediate, and advanced courses comparable to what other disciplines have. (What should students learn in Introduction to Comparative Government? We don't have an answer to this question that we could explain to a congressman in the same way that economists could say what Intro Micro is about.)

Achieving comparability, however, would require some conscious moves toward standardization. Such measures would reduce the autonomy of the individual instructor. And it would also prompt us to begin to assess what the discipline is and what it should become. In one sense, that means picking winners and losers, at least at the level of undergraduate instruction. And that likely would require the formalization of much political-science instruction (graduating students who don't know the median-voter theorem or the differences in free-trade voting preferences of workers and capitalists should be acutely embarrassing). It might even mean dropping one subfield--or, better, prompting that subfield to more clearly engage with the questions and methods the rest of the field has adopted (something that would benefit both sides of the debate). The relationship between international studies and political science would be redefined, perhaps drastically.

At the end of such a process, though, the relationship between our instruction and our research would be much clearer. The risk of intellectual monoculture would be blunted by the fact that different institutions would naturally choose to focus on different parts of the field. And our students would likely have a clearer idea both of what to expect in graduate school (an IR Ph.D. is not Foreign Affairs with more footnotes) and also of what the key insights of political science really are.