DSOC 25A32 - Quantiative methods for Policy Evaluation

Over the course of the last four years societies have witnessed one of the most dramatic returns of state interventionism since the Second World War. Governments have sought to mitigate the consequences of multiple crises – be that the 2008 financial crash, the COVID-19 pandemic, or the inflationary shock post the Ukrainian war – through a raft of new policy measures be that financial regulation, price controls, lockdowns, furlough programs, basic income packages, and infrastructure initiatives. However, it is often very hard to say with confidence what the consequences of these polices were. Using this crisis context as an overarching frame, this class seeks to introduce students to quantitative methods employed in evaluating the efficacy of public policy. In the first two weeks, the course will focus on the basics of working with data, emphasizing the importance of how data is designed (longitudinal or cross-section) as well as the level of data collection (ecological or individual). In weeks 3 to 8, the focus will turn to empirical methodology: how researchers design natural experiments to estimate the average treatment effect of a policy. Beginning with simple OLS regression, the course will advance to studying three basic approaches; difference-in-differences, regression discontinuity design, and instrumental-variables regression. Finally, during weeks 9 to 12 the course will concentrate on examples of applied research. In this final part of the course, an examination will be made as to how researchers connect these methods with important debates in public policy and social theory. Throughout the course there will be a strong emphasis on adopting a hands on approach, using the R programming language. This course will seek to give students the basic tools to be able to: manipulate and clean data, perform simple analyses both descriptive and analytical, as well as visualize their results in an aesthetically pleasing way. But more than that, this course will try and give students the confidence to explore R on their own, understanding that programming is a skill which scholars develop over many years through persistent practice.

Bartholomew KONECHNI

Séminaire

English

This class expects students to have a certain mathematical confidence but does not expect students to know anything about econometrics. Students worried about the level of mathematics in this class should feel free to get in touch. That being said, students who want to come to this class prepared would do well to review certain basic topics which should have been covered in the first year: • Measures of Centrality (i.e. mean, median, and mode) • Measures of Dispersion (i.e. variance and standard deviation) • The Equation of a Line • Polynomials (e.g. quadratic functions, cubic functions, etc) The class does not expect students to have any experience of using the R programming language, although any previous exposure is clearly advantageous. Students wanting to get a bit of a head start on using the R programming language should have a look at Garrett Grolemund's Hands-On Programming with R, a very helpful introduction to the basics. I recommend you start with Appendix A (https://rstudio-education.github.io/hopr/starting.html#starting) which tells you how to download R and R-Studio. Students can also get in touch (before, during or even after the course) if they feel completely confused by R.

Spring 2024-2025

To validate the course, the student is expected to pass the following assignments: 1°) Complete 2 short problem sets (30%). -Problem Set 1: Cleaning & Merging /Applying OLS Regression. -Problem Set 2: Differences-in-Differences and Regression Discontinuity Design. 2°) Deliver a group presentation on one of the readings (in the final four weeks of the class) (25%). This presentation should be done in groups of at most 4 students and last no more than 20 minutes (with an additional 20 minutes for class discussion). Each group will be assigned a paper and tasked with presenting it to the rest of the class. The key criteria on which students' presentations will be marked is in their ability to explain how the authors mobilize evidence to respond to a key debate in the social science/policy literature. It is important that students demonstrate an understanding of both the strengths and weaknesses of the paper they are presenting and are able to evaluate the underlying assumptions behind each article. 3°) Write a 2500-word research note (35%). This research note should try to evaluate a pandemic policy, modelled on the short research notes published in leading journals. See Van Winkle & Konechni (2022) as an example. Students should be able to ask and motivate a clear research question. Students should be able to define a clear identification strategy using one of the methods covered in the class. They should be able to present their results in a clear manner. They should also be able to describe the limitations of their approach, be that in their identification strategy or intrinsic in the data available to them. 4°) Attendance and Participation (10%) Students will be evaluated on their contribution to the class, their ability to generate meaningful discussions, and engage with others in a respectful way.