|
Headlines
This
workshop is designed to as an introduction to methods and strategies for
researchers and practitioners in empirical corporate finance. Although
the focus is in corporate finance studies, students and researchers
interested in econometrics, banking, development economics, and general
policy analysis can benefit from it.
In this
workshop, we introduce a series of challenges in regression studies, and
introduce possible solutions. We focus primarily on propensity-score
matching and regression discontinuity. (In next workshops, we can extend
topics to include the use of natural and quasi-natural experiments,
dynamic panel regressions, falsification tests, etc.). This workshop
comes with an extensive Stata programming supplement so, participants
are encouraged to bring their personal laptops where they can access
Stata. Stata codes and datasets will be provided.
Some knowledge of basic
econometrics, and some basic stata familiarity is required.
Part 1: Review of
basic econometric methods
-
Linear models, non-linear models
(mostly binary-choice models).
-
Linear and non-linear models in Stata:
how to run an actual study
-
Diagnostics: cross-sectional
heterogeneity, endogeneity issues
Part 2: Addressing
endogeneity- Part 1: econometric remedies
-
Heckman two-step
and self-selection
-
Propensity-score
matching (direct application of binary models with diff.
distributions
-
Applications with
Stata
-
Remaining
endogeneity issues (effects from unobservable)
-
Examples from the
literature
Part 3: Addressing
endogeneity- Part 2: causal inference
-
Regression discontinuity settings
-
Examples from the literature
-
Regression discontinuity with
propensity score matching
-
Stata applications
-
More examples from literature
|
