Prof. Nathaniel Beck, 19 W. 4th St., Rm 407, x88535. TA is Alex Herzog (office hours Tuesday 2-4pm, Rm 318).
I will post the schedule, post data sets, course notes and other handouts here. Please note that the schedule will be updated as dates slip, etc., so please keep referring to this site.
The class meets in a seminar room so we can have discussion about issues unrelated to how do you do it in Stata. The Friday lab is in our computer lab, which is the appropriate place to discuss Stata issues.
Below are a set of topics and guesses about dates. This will be updated during the semester to reflect actual progress. Topics are keyed to readings in Cameron & Trivedi (CAT); handouts (pdf's) are also referenced. Data sets will be distributed in Stata format.
The general text is Cameron and Trivedi, Microeconometrics (Cambridge). For further work on event history analysis, I have ordered Box-Steffensmeier and Jones, Event History Modeling (Cambridge). For further work on causality, I have ordered Morgan and Winship, Counterfactuals and Causal Inference (Cambridge). Both of the latter books are available at a very reasonable price. Different people prefer the style of different serious econometrics texts. While each book covers somewhat different topics with somewhat different notation, two books similar to CAT are Greene, Econometric Methods, 6th Ed. (Prentice-Hall) and Wooldridge, Econometric Analysis of Cross-Section and Panel Data (MIT).
Note that there are now many web sites that deal with many of the issues we deal with. Amongst political scientists, you can check the web sites of (in alphabetical order, and just a subset): Jan Box-Steffensmeier (Ohio State), Charles Franklin (Wisconsin), Simon Jackman (Stanford), Brad Jones (Davis), Gary King (Harvard), Andrew D. Martin (Washington University), Greg Wawro (Columbia) and Chris Zorn (South Carolina).
I have listed some substantive applications. Where they are not keyed to a homework, feel free to think about other articles of interest that use the various methods. Class is much better if we can actually talk about how the methods are used by substantive reearchers. Substantive articles keyed to the homework are listed with the exercise.
Grading and such: There are weekly exercises. These will be discussed in the Friday section and are due the following Wednesday. Typically they are applications of what we have studied, and are done using Stata. The goal of the exercise, however, is NOT to show that you can type Stata commands, nor that you can cut and paste Stata output, but rather that you understand what is going on with the various models. There is a final project which involves your finding some extension of what we have done in class and applying it to some problem of interest to you. The exercises are worth 60 percent of your grade, the project is 30 percent and participation in class is worth 10 percent.
| Date | Topic | Reading in text | Handout (if any) | Substantive reading | Exercise (if any) | Data, do file, anything else of use |
|---|---|---|---|---|---|---|
| Week 1 (1/22 and 24)) | Some introductory materials and notation - OLS and particularly asymptotics | CAT ch. 4.1-4.5 and Appendix A | OLS - Intro and Asymptotics | Exercise due 1/29 Using Mata | Jacobson and Dimock article (look around Table 9) and Jacobson data and Example do file and OLS/MATA example | |
| Weeks 2 and 3 (1/29 and 31 and 2/ 5 and 7) | Maximum likelihood, including ols in ml and some simple applications (logit/probit) | CAT ch. 5. If you have King, Unifying Political Methodology, this would be a good time to review the first half, but this is not required. | Likelihood (intro), maximum likelihood for Poisson, Likelihood (maths) and Likelihood (OLS), Limited Dependent Variables (intro) | Abrajana, Alvarez and Nagler JOP article (look around Tables 3 and 4) | Week 2 exercise (due 2/5) and Week 3 exercise (due 2/12) | Data for week 3 homeworks and example Stata file |
| Weeks 4 and 5 (2/12, 14, 19 and 21) | Discrete Dependent Variables (mostly discrete choice | (Readings for specific days will be announced in class) CAT ch. 14 (through section 5), ch. 15 and 20 (though section 4, this chapter is on count data and is read for the last class in this section) | More advanced binary dv models, Multinomial overheads and Event count overheads. | Nagler, "Scobit," AJPS, 38(1994):230-55, Franklin and Kosaki "Republican Schoolmaster," AJPS (application of ordered probit), Alvarez and Nagler, "When Politics and Models Collide," AJPS, 42(1998), Alvarez and Bedolla "The Foundations of Latino Voter Partisanship: Evidence from the 2000 Election," JoP, 65(1): 31-49 (2003), Bercovitch and Schneider "Who Mediates? the Political Economy of International Conflict Management," 37(200) | More complicated ml exercise (due 2/20); Multinomial logit exercise and event count exercise (both due 2/27) | Stata data set of 1992 NES for binary dependent variable exercise and codebook for NES data. Stata data set of 1992 NES for ordered probit, Stata event count data. Dutch conditional data set, Dutch unconditional data set and Dutch codebook and a short file to help you with some Stata intricacies for clogit |
| Week 6 and 7.1 (2/28 and 3/4) | Endogeneity, simultaneity, instrumental variables | CAT ch. 2 and 4.7-4.11 | Identification and Instrumental Variables | Miguel, Satyanath and Sergenti "Economic Shocks and Civil Conflict: An Instrumental Variables Approach" (JPE. 112:725-53, 2004), Erikson and Palfrey "Campaign Spending and Incumbency" (JOP, 60:355-73, 1998), Bartels "Instrumental and 'Quasi-Instrumental' Variables" (AJPS, 35(3):777-800, 1991) | Event count exercise (part II) and exercise on identification (both due 3/5) | |
| Week 7.2 and 8.1 (3/6 and 3/11) | Selection Models | CAT ch. 16. | Overheads | Lemke and Reed, "War and Rivalry among Great Powers," AJPS, 45(2001), Geddes, "How the Cases You Choose Affect the Answers You Get: Selection Bias in Comparative Politics," PA, 2(1990), Romer and Snyder, "An Empirical Investigation of the Dynamics of PAC Contributions," AJPS, 38(1994), von Stein, "Do Treaties Constrain or Screen? Selection Bias and Treaty Compliance," APSR, 99(2005), Dubin and Rivers, "Selection bias in linear regression, logit and probit models," Sociological Methods & Research, 18(1989/1990) | IV exercise and
Tobit exercise (due 3/12)
Heckman excercise (due 3/26) |
IV data (MSS) and Tobit data |
| Week 8.2 (3/13) | Causal Inference | CAT ch. 25 (lightly), Morgan and Winship, Counterfactuals and Causal Inference | Overheads | Gordon and Huber "The effect of electoral competitiveness on incumbent behavior" (QJPS, 2007), Gilligan and Sergenti "Do UN Interventions Cause Peace?" (2007) | ||
| Week 9 (3/18 and 3/20) | Break - no class | |||||
| Week 10.1 (3/25) | Discussion of term paper topics | |||||
| Week 10.2 and 11 (3/27, 4/1 and 3) | Causal Inference (cont.) | CAT ch. 25 (lightly), Morgan and Winship, Counterfactuals and Causal Inference | Overheads | Gordon and Huber "The effect of electoral competitiveness on incumbent behavior" (QJPS, 2007), Gilligan and Sergenti "Do UN Interventions Cause Peace?" (2007) | Excercise on matching (due 4/9) | data from Gilligan and Sergenti |
| Week 12 (4/8 and 10) | Duration models | CAT, ch. 17-19; and Box-Steffensmeier and Jones, "Event History Modeling" | Overheads | Bennett "Testing Alternative Models of Alliance Duration" | Exercise on duration models (due 4/16) | King, Alt, Laver and Burns, "A Unified Model of Cabinet Dissolution" (AJPS),Coalition data |
| Weeks 13, 14 and 15.1 (4/15, 17, 22, 24 and 29) | Times-series--cross-section data | Continuous DV TSCS notes and Time series issues | Beck and Katz "Time-Series-Cross-Section Issues: Dynamics, 2004," Draft of July 24, 2004, Beck and Katz "Random Coefficient Models for Time-Series-Cross-Section Data: Monte Carlo Experiments," PA, 15:182-195, 2007, Plumper and Troeger "Efficient Estimation of Time-Invariant and Rarely Changing Variables in Finite Sample Panel Analyses with Unit Fixed Effects," PA, 15:124-139, 2007, Wilson and Butler "A Lot More to Do: The Sensitivity of Time-Series Cross-Section Analyses to Simple Alternative SpeciÞcations," PA, 15:101-123, 2007 | Exercise on TSCS data (due 4/23) and Exercise on binary TSCS data (due 4/30) | Garrett data and Data for binary exercise and Oneal and Russett preprint |