MOD. 1 ECONOMETRICS OF INDUSTRIAL ORGANIZATIONS
The most trivial competence that the student should obtain from the course is to master the interpretation of the results of OLS, IV and 2SLS estimation. More ambitiously, the course aims at teaching something about the scientific method, namely how statistical inference can help in testing hypotheses and which arrangements have to be done to this procedure in a complex context such as the economic environment.
The objective of the course is to provide the students with some econometric tools
for the analysis of the behavior of industries and markets: the tools covered are the linear regression model, as estimated through ordinary least squares or through instrumental variables, and the simultaneous equations model. In particular, we study the impact of policy interventions on economic agents in regulated markets.
We aim at estimating not simply a purely descriptive relationship among the variables appearing in the economic problem, but the causal relationship linking them. This because, when the objective is to study the effect of a policy change on the behavior of individuals, we would like to use the estimated relation to simulate the effect of alternative circumstances on the economic reality. For this it is important to be sure that the regression estimates obtained through ordinary least squares or instrumental variables techniques can be interpretable as direct estimates of the parameters of interest.
In natural sciences, this problem is generally solved through a careful design of the experiment in which data are collected. In social sciences, this is generally impossible as classical controlled experiments are often unfeasible because of economic or ethical constraints. Therefore, we are compelled to use more refined inferential techniques (in general instrumental variables estimation in its several variants) to obtain causal relations from observational data.
In this part of the course, we will deal with these problems. In particular, we will describe the design of controlled experiments and we will see in which sense economic data do not comply with the standards of this class of experiments. Then we will describe the techniques that can be used to infer causal relationships in the absence of experimental data.
Each topic will be accompanied by several practical examples and by a worked-out paper-length application illustrating the method on real data. Econometrics of Regulated Markets
Lecture 1 - Causal Relationships and Counterfactuals
Lecture 2 - Further Results on Linear Regression
Lecture 3 - Controlled Experiments
Lecture 4 - Further Results on the Instrumental Variables Estimator
Lecture 5 - Natural Experiments and Identification Strategies
Students will be evaluated on the basis of a written essay in which they discuss a published paper.
The course will be based on a set of slides written by the lecturer. A list of academic papers addressing more advanced topics will be provided to the students as additional readings. Mathematical derivations will be dealt with in handouts written by the lecturer.