What I teach
I teach development economics,
applied econometrics and
industrial organization. I teach at the following levels:
Follow these links if you want more information, or
if you want to access my lecture notes.
One day, perhaps a book...
Francis
Teal and I have a rather long-term plan to write a book on
empirical development economics, suitable for MSc students in
development economics and development studies. Whether this is
a good idea or not remains to be seen. In any case, our current
thoughts on this project, and a very preliminary table of
contents, can be found below. Comments on this enterprise would
be most welcome.
Empirical Development Economics
Synopsis
This book has two objectives. The first is to give some
insight into how development economics can be viewed from an
empirical perspective. The second is to introduce the tools that
will enable the student to carry out empirical work in development.
It may seem strange to stress the empirical in a subject such as
development which is the study of the processes which have generated
the enormous range of outcomes we observe in the world at the moment
between rich and poor countries. However there are many other
approaches to development than ones which stress the need to analyse
data, indeed such approaches are the dominant ones in the subject.
One reason for this divorce between studying development and the
analysis of data is that the latter is typically taught in a course
on quantitative methods where the student is asked to address a
range of apparently esoteric questions: Are the errors in the
regression heteroscedastic? Do the data co-integrate? Should
an instrumental variable estimation approach be adopted? The student
then turns to a course which they actually want to do on development
where the topics addressed will typically include: Who are the poor
and how is poverty measured? Does globalisation impoverish the poor?
What is the role of human capital in growth and poverty reduction?
Are neo-liberal polices increasing poverty? How does gender impact
on poverty? These questions can be, indeed usually are, taught
without any mention of the esoteric terms to which the student has
been exposed in the course on quantitative methods. The courses
become part of parallel universes in which the successful student
learns to speak different languages and in which the connection
between the universes is a puzzle.
This book seeks to link
these parallel universes. It will attempt this by going over much of
the ground taught in basic statistics course but rather than
focusing on the statistical issues it will focus on how they inform
our understanding of development questions. There is no point in
pretending that data can be analysed without certain basic
statistical techniques. Equally the issues that are of concern to
many areas of econometrics are irrelevant to understanding much of
development as the data is simply unavailable. The guiding principle
in this book is that if there is no data then there is no
interesting question about the processes of development.
This is a book about development economics and most
students of development are not economists. Is this book not for
them? We hope they will not think so. Nobody we think would wish to
argue that one can understand any aspect of the developing world
without a knowledge of how their economies have changed over
the last decade or longer. This book is going to focus on the data
and while non-economists may well wish to skip over the bits on how
economists wish to analyse the data we hope there is enough data
presented here, and enough argument about how the data should be
used and what it means, to encourage non-economists to allow some of
this thinking into their analysis.
It is probably true to say
that development economists are not very well regarded by many who
study development. This is partly because they are likely to have
spent more time in the QM universe rendering them unable to
communicate with inhabitants from the other universe. But the
problem goes deeper than that. Economists are viewed as social
scientists who think theory is more important than data. This book
is definitely for those who think that economists should think the
opposite.
Book
Outline
Part 1
Linking Models to Data for Development
Chapter 1
An Introduction to Empirical Development Economics
Chapter 2
The Simple Linear Regression Model
Chapter 3
Multiple Regression Analysis:
estimation
Chapter 4
Multiple Regression Analysis: inference
Chapter 5
Maximum Likelihood Estimation
Chapter 6
Heteroskedasticity
Chapter 7
Modeling Choice: LPM, Probit and Logit
Models
Chapter 8
Logit and Probit Models: Inference and Diagnostics
Chapter 9
An Introduction to Time Series
Chapter 10
Serial Correlation in Time Series
Models
Chapter 11
Cointegration
Chapter 12
Panel Data: An Introduction
Chapter 13
Panel Estimates: POLS, RE, FE, FD
Chapter 14
Instrumental Variable Estimation
Chapter 15
Program Evaluation: the Basics
Chapter 16
Program Evaluation: Imperfect compliance and
heterogeneity
Part 2
Determinants of Income and
Growth Chapter 17 Principles of Modeling: Endogeneity and Instruments Chapter 18 Structural Models Chapter 19 Econometric Analysis of Dynamic Panel Data Chapter 20 Estimating the Burnside and Dollar and the MRW Growth Models Chapter 21 Panel Data and Endogeneity Chapter 22 Sample Selection Chapter 23 The Tobit Model Chapter 24 Multinomial choice Chapter 25 Long-T Panel Data Analysis: An Introduction Chapter 26 Nonstationarity and Cointegration in Panel Time Series Chapter 27 Cross-section Dependence in Panel Time Series Chapter 28 Omitted Variable Bias, Measurement Error and IV: A Review Chapter 29 Reduced Form Evaluation Methods Chapter 30 Evaluation with Structural Models Chapter 31 Modeling: An overview Chapter 32 What does determine development?
Class Based Work (Each class based
exercise follows after two
chapters)
Exercise 1
Describing and Understanding
Data
Exercise 2
Cross Section Data and Issues of Functional
Form
Exercise 3
Introducing Maximum Likelihood in STATA
Exercise 4
Normality, Heteroskedasticity and the
LPM
Exercise 5
Interpreting Marginal Effects in Probits and
Logits
Exercise 6
Time Series
Applications
Exercise 7
Using Panel Data
Sets
Exercise 8
Using STATA for
Modeling
Exercise 9
Program Evaluation using Progresa Data
Exercise 10
Using Structural Models
Exercise 11
Using XTABOND2
Exercise 12
Selection in Indian Schools: Do bossy parents have bolshie
kids?
Exercise 13
Multinomial and Censored
Choice
Exercise 14
Mean Group
Estimators
Exercise 15
Panel Unit-root testing, Testing for CSD
Exercise 16
Evaluation Methods
|