Formal Models in International Relations
Syllabus, with links to readings and summaries.
| Branislav L. Slantchev | Spring 2020 |
| Office: n/a | Tue 9a-11:50a |
| Hours: Slack #fmir | Zoom |
Description
This course is intended for advanced graduate students who are interested in the formal analysis of international relations. The goal is to familiarize students with the practical aspects of modeling various substantive ideas by exposing the advantages and limitations of mathematical formalization. To accomplish this, students will (i) derive or prove results from the selected readings, (ii) identify the contributions of their authors, and (iii) find ways to improve the research or extend the insights. This course concentrates on the theoretical aspects of formal research and complements POLI 247A: Quantitative Approaches to IR, where students investigate statistical modeling and empirical testing, and POLI 247C: History of International Relations, where students explore diplomatic and military history of Europe from 1618 to 1945. Students are expected to have taken courses in international relations (e.g. the core seminar), and game theory (e.g. POLI 204C).
Requirements
There are three requirements for this course:
- Formal presentation of one of the papers to the class (35%). Each student will be responsible for presenting and critiquing one of the articles he or she will be assigned during our organizational meeting. The presentation will consist of: (i) a summary of the article's main points, (ii) a careful exposition of the proofs and the intuition behind them, (iii) an evaluation of the substantive importance of the contribution, (iv) a critique of the paper, and (v) questions for discussion. Presenters must be prepared to lead discussion, which means that they should identify the broad themes underlying the reading, establish links to other literatures, suggest improvements for flaws in modeling or research design, derive testable hypotheses, and have ideas about their empirical testing. Presentations will use PowerPoint/PDF so we can all follow them online via Zoom.
- Written referee report (45%). Students will select one paper from the list of recommended/background readings (i.e., not a paper that will be discussed in class), and write a formal referee report. The paper must be cleared with me first. It is best to select one as early as possible because I will not allow more than one student to write on the same paper. Come with one or two alternative choices. The report must summarize (in one paragraph) the contributions of the paper, discuss their importance, and then provide a critique/discussion, in which the referee must analyze deficiencies in the model, possible extensions and improvements, and perhaps outline alternative approaches to the problem. Students must be aware that every model is deficient in some ways, and must not fall back on common, but vacuous criticisms of formal models. Instead, students must approach this task as if they are writing suggestions for a colleague or even for themselves (possible venues for research). The reports must be at least three single-spaced pages long (at 11pt font size). These are due to me by noon on Friday of finals week. You can read about writing a referee report from the excellent book by William Thomson. I have scanned the relevant chapter, but I highly recommend you buy this slim volume for reference.
- Participation in class discussions (20%). Everyone will be expected to have read the articles before coming to class. Presenters will provide the starting point for class discussion and other students will be expected to contribute actively.
Guidelines
Modern research in international relations tends to be methodologically sophisticated, and even though the formal models are not quite up there with the bleeding edge of the area, they are far from the toy examples of Prisoner's Dilemma and Chicken games that used to dominate the field. This means that you should expect to spend quite a bit of time reading each article, which is why we shall rarely discuss more than a handful each week. There is no set way to approach reading formal papers, but I strongly suggest you utilize the following strategy:
- Begin by reading the introduction and conclusion (and perhaps the literature review). What is the article about? What is its central point? Write down a list of claims (propositions) that you would like to see and that you would need to see to be persuaded of the validity of the article's argument. (And I mean, write these down!)
- Read the description of the formal model. Construct a dictionary for the notation: that is, write down a table with the symbols and their meanings. (As you will quickly see, there is no standard notation in game theory and that results in a lot of notational clutter, so keeping it straight is always a challenge.) Identify the players, their strategies, their payoffs, and the sequence of moves. Is this a game of complete, perfect, and symmetric information? If not, what is it? Why has the author chosen to model it that way? How closely does the formalization match the substantive motivation for the model? What are some of the simplifying assumptions that might be distorting? Why? What solution concept will the author use? Is it appropriate? Recall now your answers to the previous step: do you think this model will be able to address the puzzle adequately?
- Read the analysis. What propositions does the author present? How do they correspond to your wishlist? Are they stronger, deeper, more satisfying? Or trivial and disappointing? How do they correspond to the author's own claims? Does the author "stretch" their interpretation? How general are these results? Can you recall similar claims from different models? Or contradictory claims from similar models? Do you think there is a general underlying dynamic common to this and other models?
- (Difficult step.) Satisfy yourself that the claims are valid and the proofs work. Due to journal space constraints, many editors have now taken to insisting that authors leave our parts of (or sometimes even entire) proofs. This is most unfortunate and shows deep misunderstanding about what the value is of a formal article. The whole point of using mathematics is to verify the validity of conclusions and internal consistency, but without the proofs, one must essentially rely on the author to have done it correctly (bad idea, we all make mistakes, and not rarely either) or reconstruct the proofs themselves (a very time-consuming and often non-trivial task). If the article includes the proofs, then make sure you understand them fully. What is each step designed to accomplish? What is the intuition behind the sequence of steps taken? Are you satisfied that they produce the desired result? Draw pictures to illustrate claims, or to attempt to disprove them. Solve a simple case by assigning numbers to the parameters. If the article does not include the proofs, try to locate a working paper version online (sometimes these have all the math in them). If that fails, try to think of a strategy for a proof of the claim. Think of counter-examples, can you find one? Draw pictures! Look at similar models and observe how the authors have solved them.
- Evaluate the generalizability of the results. Which assumptions do you think are crucial for the results? Do you think the results are likely to extend if we relax them? Why or why not? Which part(s) of proofs will no longer hold? What do you think would happen instead? Are the implications serious for our substantive interpretations? What are some of the questions and new puzzles that this article raises (perhaps the author mentions some, but most often these will arise from your own reading). What type of article do you think would best handle these questions? Is it on the syllabus or does it need to be written? How does this model fit in the modeling enterprise?
- Step back and look at the conclusions: what have we learned that we did not know before? How does this relate to the other readings, both formal and non-formal that you have seen? Does it contradict some "received wisdom," modify and limit general claims, or support and extend others? What is the general abstract idea to take away? Is it applicable in other contexts, not necessarily IR-related? Which ones and how? If it needs slight modifications to be applicable, what would these be?
- What are some empirical hypotheses that this model suggests? Are they testable or is the model at such a high level of abstraction that it is difficult to see how to apply it? If they are testable, what sort of research design would you adopt? What variables would you need? Do the hypotheses differ from ones used in previous empirical studies? Do they suggest that the previous research design was flawed? In what ways? How would you improve upon it?
One thing I should make abundantly clear is that this course will deal with game-theoretic formal models. There are many other mathematical formalizations (differential equations, agent-based models, general equilibrium models) but they are not as useful for most of the topics we shall be dealing with because they are unable to account for strategic interdependence of decision-making.
Presentation Schedule
Readings
Rational Choice, Game Theory, and International Relations |
|
| Recommended and Background Readings |
|
Bargaining and the Inefficient Use of Power |
|
| Recommended and Background Readings |
|
Crisis Bargaining and Escalation |
|
| Recommended and Background Readings |
|
Conventional and Nuclear Deterrence |
|
| Recommended and Background Readings |
|
Arms Races and Power Transition |
|
| Recommended and Background Readings |
|
Alliances, Balance of Power, and Intervention |
|
| Recommended and Background Readings |
|
Diplomacy, Reputation, and Mediation |
|
| Recommended and Background Readings |
|
Domestic Politics: Constraints, Signaling, Guns versus Butter |
|
| Recommended and Background Readings |
|
Civil Wars, Ethnic Conflict, and Terrorism |
|
| Recommended and Background Readings |
|
Sanctions, Enforcement, and Cooperation |
|
| Recommended and Background Readings |
|
International Institutions |
|
| Recommended and Background Readings |
|
Trade and Trade Agreements |
|
| Recommended and Background Readings |
|
Exchange Rates, Debt Crises, and Size of Nations |
|
| Recommended and Background Readings |
|