HARVARD UNIVERSITY
HARVARD KENNEDY SCHOOL OF GOVERNMENT
Course Syllabus for API-202 (Z-Section)
Empirical Methods II
Spring 2016
Class Time: Monday and Wednesday, 10:15-11:30am in LAND
Z-Section Review Sessions: Friday, 1:15 - 2.30pm in HKS 1BA 401
Professor Amitabh Chandra
Email: Amitabh_Chandra@harvard.edu
Office Hours: Taubman 344, Monday 3-6pm
To schedule an appointment please email Alexandra Kovalik
Faculty Assistant: Alexandra Kovalik (Alexandra_kovalik@hks.harvard.edu)
Teaching Fellow: Melissa Sands (sands@fas.harvard.edu)
Office Hours: Friday 2:30-3:30pm in computer lab
Course Assistants: Maurits Waardenburg (maurits_waardenburg@hks16.harvard.edu)
Audrey Baker (Audrey_baker@hks16.harvard.edu)
COURSE DESCRIPTION
The purpose of this course is to equip students with the quantitative tools necessary to tackle issues that
involve the empirical analysis of public policy problems of the sort they will encounter in a professional
environment. Specifically, the course introduces students to the use of multiple regression analysis for
analyzing data and evaluating social policy. The emphasis is on empirical applications.
The course is designed with two objectives in mind. The first is to provide students with the ability to
analyze critically the empirical analysis done by others at a level sufficient to make intelligent decisions
about how to use that analysis in the design of public policy. The second is to provide students with the
skills necessary to perform empirical policy analysis on their own or to participate on a team involved in
such an empirical analysis. An important segment of the course focuses on program evaluation. This
includes both the design and analysis of experiments that aim at measuring policy effectiveness and the
use of non-experimental data to evaluate policy effectiveness.
The Z section of API202 covers similar material as the other sections. The level of mathematical
sophistication is, however, substantially higher, and because of the use of calculus allows the course to
move through material quickly and cover more advanced topics. Students should enroll in the Z section
instead if they prefer a more formal treatment of the material, or if they are interested in a more fluid
discussion of the issues that arise in real-world program analysis. Assignments and exams differ
between the Z section and the other sections, but there is a considerable degree of overlap.
PREREQUISITE FOR THE Z-SECTION
Fluency with simple statistics, including hypothesis testing, at the level of API-201 (Section Z) is
assumed. Calculus (derivatives and integration) will be used, and you are strongly discouraged from
taking the Z section without fluency in basic calculus. The Z-section will also use the properties of the
Expectations, Correlation, and Variance operators. Students with less than an A- in API 201 are not
permitted to enroll in API 202 (Section Z).
TEXTBOOK
Stock, J. and Watson, M., Introduction to Econometrics, 2nd edition, Addison-Wesley (2007) is a highly
recommended reference. 10 copies are on reserve in the library. **Please note: you may also purchase
the 3rd edition**
Nate Silver. The Signal and the Noise. Penguin 2015. We will read the first 7 chapters but you are free
to read more.
GRADING
Problem Sets and quizzes:
Midterm Exam:
Final exercise:
Final Exam (Cumulative):
20%
30%
10%
40%
Pursuant to HKS recommended grading practices, 10-15% of the grades will be an A, 20-25% will be an
A-, 30-40% will be a B+, and 5-10% will be a B- or lower. All course activities, including class
meetings, homework assignments, and exams are subject to the HKS Academic Code and Code of
Conduct.
SWITCHING FROM Z-SECTION TO OTHER SECTIONS:
On February 10th, you will be given a comprehensive diagnostic quiz (multiple choice) that will give
you feedback on whether the Z-section is right for you. Scores will be returned to you quickly. You will
then have a midterm on 2/29 that will be returned quickly. This return date is also the last date that you
can switch into the non-Z sections. The School will not permit switching from the Z-section to the non-Z
section after this date because it is disruptive to the other sections, and more difficult for students to
catch up and succeed in the new section. If you switch over, we will send your homework grades to the
section that you have switched over to. Your quiz grade in the Z section will not affect your
performance in the non-Z sections because there is no corresponding quiz in those sections.
EXAMS
The midterm and final will be closed book and closed notes.
PROBLEM SETS
1) You are expected to turn in answers to the problem sets. Although problem sets will not be graded in
detail, they will be corrected by the course assistants. Detailed answers will be posted on the course
website for you to review.
2) To receive credit, problem sets must be submitted before the start of class on the day they are due.
3) The lowest problem set grade and lowest quiz grade will be dropped from the calculation of the
overall problem set score.
4) You may work on the problem sets in small groups. You must, however, write up your answers
individually, in your own words. Put the names of your study group member(s) on your problem set.
Duplicate answers will receive no credit and will be subject to disciplinary review.
5) STATA, a statistical software package, is available both in the computer lab and from the CMO.
CLASS PARTICIPATION AND ENGAGEMENT
We strongly believe that student participation can substantially enrich the learning experience for both
the students and the instructor. In this spirit, class participation is encouraged. Effective class
participation requires that you read any assigned readings before coming to class. You are encouraged to
ask questions, to share relevant insights you have from previous experiences, and to treat your
classmates’ participation with courtesy. We only ask that questions and comments be brief and related to
the topic at hand. Given that this is a large class, we will sometimes need to defer questions to a future
class or office hours.
Please do not bring food into the classroom. The use of computers, tablets, and phones is strongly
discouraged even for taking notes. Checking email, Facebook and Twitter during class is not
professional.
FINAL EXERCISE
The final exercise will require you to applying some of the empirical methods learned in class to readings of
empirical research on public policy. The topic will be announced after the midterm.
REGRADE POLICY
To submit a re-grade request for an exam, submit the exam along with a clear statement of your
concerns within two business days of the return of the graded exams. Your entire exam will be regraded.
Date
Stock & Watson,
2nd Edition
Topic
Assignment
1/25
Observational studies, RCTs and causality
1, 1.3, 9.1, 9.2
1/27
Bivariate regression
4.1-4.6, 5.1-5.2
Silver, Chapter 1
2/1
Multiple regression
6.2-6.6, 7.1
Silver, Chapter 2
2/3
Dummy variables
5.3
PS1
2/8
Omitted variable bias
6.1, (9.2)
Silver, Chapter 3
2/10
HRT Case
Read case
Quiz
2/15
HOLIDAY (President’s Day)
2/17
Interactions
8.1, 8.3, (8.4)
Silver, Chapter 4
2/22
R2 and joint hypothesis testing
7.2, 7.5
PS2
2/24
Logarithms and Quadratics
8.1, 8.2, (8.4)
Silver, Chapter 5
2/29
MIDTERM (in class)
3/2
Binary dependent variables
11.1-11.3
3/7
Binary dependent variables
11.1-11.3
3/9
Field experiments
3/14
HOLIDAY (Spring Break)
3/16
HOLIDAY (Spring Break)
3/21
Health Insurance Case
Read Case
3/23
Fixed effects
13.2
3/28
Difference-in-differences
10.1-10.7
3/30
Instrumental variables + Introduction to final
exercise
13.1, 13.3
4/4
Instrumental variables
12.1-12.6
4/6
Regression Discontinuity
12.1-12.6
4/11
Advanced Topics
5/6
FINAL EXAM (3-6pm)
Silver, Chapter 6
PS3
Silver, Chapter 7
PS4
PS5
Final exercise
Cumulative