KECD 2110 - ECONOMETRICS 3 : Treatment effects

Course Objectives: Students should have some basic knowledge in cross-sectional econometrics (ordinary least squares, heteroskedasticity, instrumental variables, probit, logit and tobit models). These concepts and notions are introduced in the probability, statistics and econometrics courses taken in the M1. The course covers my lecture notes, see outline Useful complementary readings include:

Clément de CHAISEMARTIN,Viktor VETERINAROV

Cours magistral et conférences

English

Problem sets Riddhi's TA sessions will take place from week 2 to 13 of the semester: there will not be a TA session in week 1. Riddhi will post one new problem set on the Moodle page of the class almost each week, one week before the problem set is due. Only 5 of those problem sets will be graded (see list below). The others will not be graded, though you are strongly encouraged to work on them as well: midterm and final exams tend to be similar to problem sets. For graded problem sets, it is assumed that study groups will collaborate on problem sets, though each student needs to hand back one problem set and no problem set should be the identical copy of that of another student. Graded problem sets are due to Riddhi, on Tuesday at 6pm of the week the problem set is due. Problem sets are graded on a scale from 0 to 2. Riddhi has been instructed to spend no more than 3 minutes grading each problem set. A subset of 2 to 3 questions of the problem set will be randomly selected and she will only grade those questions. Those questions may not pertain to the exercises you will review with Riddhi during TA sessions (see outline): unless otherwise indicated, you need to hand to Riddhi solutions for all the exercises in the problem set. Following each problem set, you will receive detailed solutions.

Students should have some basic knowledge in cross-sectional econometrics (ordinary least squares, heteroskedasticity, instrumental variables, probit, logit and tobit models). These concepts and notions are introduced in the probability, statistics and econometrics courses taken in the M1.

Autumn 2024-2025

Grading will be based on problem sets (20%), a midterm exam (40%), and a final exam (40%).

List of graded problem sets: PS Week 1, due Tuesday of Week 2 PS Week 2-3, Exercises 3 and 4, due Tuesday of Week 4 PS Week 5, due Tuesday of Week 6 PS Week 7 question 3), due Tuesday of Week 9 PS Week 9, due Tuesday of Week 11 Exams : In the midterm and final, there will be mathematical exercises requiring students to perform derivations. To get full credit, each step in the derivations should be carefully justified. E.g.: when invoking a theorem to go from one step to the next in a derivation, students should explain why the assumptions under which the theorem holds are satisfied in the situation they are considering. The midterm exam will take place in week 8. If you miss the midterm exam because you were sick, you will need to provide a signed letter on letterhead from a medical professional as evidence, and there will be no catch-up exam: the weight of the midterm will just be transferred to the final exam. Catch-up exam : There is a catchup written exam after the final, for students whose average grade (computed as explained above) is strictly below 10/20. After the catchup exam, students' grade is recalculated, replacing 0.4*midterm+0.4*final by 0.4*max(midterm,catchup)+0.4*max(final,catchup). In other words, a good performance in the catchup can offset a poor performance in both the midterm and the final exam. After this catchup exam, in line with the rules of the school of research, there cannot be a second catchup exam. The only exception is if a student had to miss the catchup exam due to extreme medical or personal circumstances that will have to be documented within 48 hours after the catchup exam (by a medical report or a report from the relevant authority)."