ECON8331f2022 Econometrics II

Fall 2022       Instructor: Bent E. Sorensen

Classes will be in person and on Zoom. I expect you to attend most classes in person, though. Classes September 21 and TBA will be Zoom only as I am out of town.

TA: William Bennett.    email:

William have office hours TuTHu 11.30-1.00, or by appointment.

TA Session will be in person Friday in SR2 134 from 11:30am to 1:00pm.

No class: TBA if needed

Make-up classes: Monday Oct 31 in SR2 134 4pm-5:30pm

Exam Friday December 2nd.


Two midterms, Monday September 26th and Monday October 24th.

2021 Midterm 1

2021 Midterm 2

2021 Final

NOTE: There will be computer exercises using Matlab as part of the homeworks. NOTE: the computer exercises are at the heart of this course.

Material Covered 2020 (similar for 2021)


Review of Maximum Likelihood (updated 2022).

Binary Choice Models Binary Choice (version 2022).

Short Introduction to Time Series. (How to work with time series model, unchanged from Macro II notes.)

Estimation of AR and MA models (revised 2021).

Truncation, censoring and selection. (Similar to the coverage in the Davidson-MacKinnon book, but with more details. Small corrections 2022)

My 1992 JBES article on credit rationing. (Example of an ordered-sequential logit model.)

After covering the selectivity model, I will talk about my 2000 Journal of Econometrics article on portfolio demand. (Example of an multinomial discrete-continuous logit model.)

The following notes may be adjusted during the semester.

Note on Panel Data (important addition for unbalanced panels 2021)

Literature on clustered standard errors:


Bertrand, Dufflo and Mullainathan

Cameron and Miller: Guide to Cluster Robust Inference

(link to Cameron's WEB page which has the paper as well as Stata code and datasets)

Weak Instruments:

Class Notes on Weak Instruments (summarizes some of the surveys below, updates a little 2021)

Know the striking example in Nelson-Starz Journal of Business (1990)

Michael Murray's survey in Journal of Economic Perspectives 2006, (you should know the formulas on pp. 123-124 and the Stock- Yogo (2005) rule of thumb in footnote 8)

Most up-to-date survey on weak instruments Andrews, Stock, Sun (2018)

Local Average Treatment Effects (LATE) if we get to it (not 2019):

Derivation of simplest case in my paper in Quantitative Economics

GMM Notes part 1 (updated 2022 for more consistent notation)

GMM Notes part 2 (updated 2022)

GMM Notes part 3 (The first two pages are the most important. The theory is not going to be on exam.)

Class Handouts 2021 (I expect to use version of these, with small updates)

Statistics Notes     Aug 24

Newton Algorithm     Aug 26

Information Matrix Identity     Aug 26

Likelihood Dependent Variables     Sep 2

Short intro to duration models     (Updated Sep 2022)

Short intro to SURE estimation      (small corrections 10/6/21)

Short intro to Multivariate, Multinomial, Ordered and Sequential Probit/Logit Models     (updated Oct 21)

Short intro to identification in multivariate linear model    

Short intro to bootstrapping of critical values    

Short intro to robust cluster estimation of standard errors (some corrections 2022)    

Homework #       Matlab Code           Due

Homework 1               PdfCode          Wed Auguest 31

Homework 2                           Wed September 7

Homework 3          PdfMatlabAR-MA Code    Wed September 14

Homework 4                Program code         Wed September 21

Homework 5             Program code                  Wed October 5

Homework 6             Program code                Wed October 12

Homework 7       Program code for 2SLS, LIML..            Wed October 19

           Program code panel estimation       

Homework 8       Program code for bootstrapping        Wed November 2

            Program code for clustering       

Homework 9       Program code for weak IV             Wed November 9

            GMM Program code     

Homework 10                              Wed November 16

Homework 11                              Mon November 28