AU kursuskatalog arkiv

4630: Microeconometrics ( efterår 2010 - 10 ECTS )

Rammer for udbud

• Uddannelsessprog: engelsk
• Niveau:   COURSE LEVEL:  Optional MSc course  
• Semester/kvarter:  SEMESTER FOR WHICH THE COURSE DESCRIPTION APPLIES: Fall 2010  
• Timer per uge:   NUMBER OF HOURS PER WEEK:  4(2+2), 3 lectures and 1 exercise on average in 12 weeks.  
• Deltagerbegrænsning:
• Undervisningssted: Århus
• Hovedområde: Det Samfundsvidenskabelige Fakultet
• Udbud ID: 23408

Formål

After having followed the course, the goal is that the student should be able to

• formulate the statistical concepts of convergence in probability and convergence in distribution and analyze the asymptotic behavior of estimators of parameters of interest
• describe features of panel data and compare the assumptions behind and features of linear panel data models to those of other linear panel data model as well as to those of linear cross-sectional models
• derive m-estimation and apply the results from m-estimation to maximum likelihood estimation and nonlinear regression models
• describe features of discrete response data, censored and truncated data and compare the assumptions behind and features of different models for these types of data
• evaluate the principles behind and problems associated with estimating panel data models for discrete response, censored, and truncated data
• analyse problems associated with endogeneity in linear models as well as discrete response models, censored regressions, and sample selection models and explain solutions to this type of problem
• describe heterogenous treatment effects and causes hereof, compare these to constant treatment effects, and explain the challenges involved in identifying parameters of interest in parametric models when treatment effects are heterogenous
• analyse the identification of average treatment effects using instrumental variables and propensity score matching when treatment effects are heterogenous
• for all models mentioned above, reflect on the applicability of a type of model in a given empirical setting, evaluate the principles behind estimation, and test hypotheses about parameters of interest

Indhold

COURSE DESCRIPTION:

The purpose of the course is to provide the student with a rigorous introduction to advanced econometric analysis of cross section and panel data. We will discuss the concept of causality, identification of parameters of interest, and estimation. After this general introduction we will focus on a number of topics. We will see how access to panel data - under certain assumptions - may help us identify parameters of interest. We will also analyze the modelling of limited dependent variables, a type of data often encountered within economics. Here we do not have full information about the outcome of interest. For example, we only observe whether an individual chooses a short or long further education, not to which degree he prefers one over the other. Similarly, we may be interested in modelling whether length of education causes higher public sector wages for a random individual in the workforce. The problem is that we only observe public sector wages for individuals who actually choose to work in this sector, not for individuals who work in the private sector and individuals working in the public sector may be fundamentally different from individuals working in the private sector. Finally, we acknowledge that individuals may not react in the same way when exposed to some sort of treatment or policy. Some students may, for example, have a higher gain from following Microeconometrics than others even though the two groups are observationally equivalent. This presents a challenge for our standard set-up and we will discuss both identification and estimation when we loosen the assumption of constant treatment effects.

COURSE SUBJECT AREAS

M-estimation
Nonlinear regression models
Models for panel data
Models with discrete dependent variables
Censored regression models
Sample selection models
Average treatment effects

Faglige forudsætninger

REQUIRED COURSES: 3620: Econometrics

Underviser

LECTURER: Marianne Simonsen

Undervisnings- og arbejdsform

TEACHING METHOD:

3 lectures and 1 exercise, theoretical as well as empirical, on average per week

TEACHING LANGUAGE:  English

Litteratur

LITERATURE:

A. Colin Cameron and Pravin K. Trivedi (2005), Microeconometrics: Methods and Applications . Chapters 1-3, 5, 14-16, 21-23, 25.

TOTAL NUMBER OF PAGES: Approx. 351 p., notes approx. 50 p.

Bedømmelse

• Mundtlig, bedømt efter 7-skala med intern censur
5 xxxxx

FORM OF ASSESSMENT:  20 minutes' oral exam with 20 minutes' preparation

EXAMINATION AIDS ALLOWED:  All - except for any kind of electronic communication including calculators, mobile phones and PCs.