Microeconometrics 2400-ICU1MIR

Introduction[1]
- Ceteris paribus assumption and control variables
- Random vs. fixed regressors assumption
- Random sample assumption and its role in econometrics
- Model defined in conditional expectations
- Properties of conditional expectations
- Partial effects, elasticities and semielasticities
- Average partial effects
- Basic asymptotics

Ordinary least squares and instrumental variable estimation [2,3]
- Omitted variable bias
- Properties of OLS with omitted variables
- Proxy variable solution
- Instrumental variable solution (IV and 2SLS)
- Assumptions of IV
- Asymptotics of IV
- Endogeneity testing
- Generated regressors and instruments
- Pooled cross sections
- Clustered samples

Unobserved effects and panel analysis [4-6]
- Linear unobserved effects model - assumptions
- Random effects estimator
- Fixed effects estimator
- First difference estimator
- Panel estimators - comparison (Hausman test)
- Panel estimation of dynamic models - sequential exogeneity
- Individual specific slopes
- Panel models with endogenous explanatory variables
- Hausman and Taylor type estimators

Nonlinear models [7,8]
- M estimators - assumptions and properties
- Hypothesis testing
- Optimization
- Simulation and bootstrapping
- Maximum and quasi maximum likelihood estimators (ML) - asymptotic properties
- Least absolute distance (LAD) estimators and Generalized Methods of Moments (GMM) - assumptions and properties
- Efficiency of estimators

Count data models [9]
- Poisson model - assumptions
- Negative binominal models - assumptions
- Advanced models for count data (ZIP)

Discrete dependent variable [10,11,12]
- Latent dependent variable problem
- Index models for binary data (linear, probit and logit)
- Dynamic models for binary data - initial condition problem
- Multinomial choice models - extreme value type I distribution
- Multinomial logit - independence of irrelevant alternatives (IIA) assumption
- Conditional logit and hierarchical logit
- Ordered choice models - ordered logit and probit
- Censored data models - tobit, 2 stage model
- Advanced topics

Nonrandom sample selection, attrition and stratified samples [13]
- Sample selection problem
- OLS and sample selection bias
- Truncated samples and II type tobit (Heckman model)
- Binary choice models with sample selection (Heckman probit)
- Panel data models for rotated panels and attrition problem
- Stratified samples - sampling schemes and weighted estimators

Policy response analysis [14]
- Regression methods
- Methods based on probability of response
- Instrumental variable methods