Correlation or causality?

Todays Aftenposten presents a study of obesity among kids. They find that the BMI of a three year old child is positively correlated with the mother’s BMI prior to, and during, birth.

They conclude that it is important to avoid overweight before and during pregnancy, in order to avoid obese children. Hence, it is interpret as something causal. However, as people living together tend to eat the same food and share lifestyle it is no wonder why different healt outcomes are correlated within families.

Seminar 3

The seminar problems for seminar 3 are found here.

You will need two datasets which can be downloaded here; FE_unions, Education_IV

Group E are supposed to hand in solutions to problem 1 (Flow and stock variables). Group F are supposed to hand in solutions to problem 2 (Estimation)

Suggested solutions by group C and D - Seminar 2.

Here is the suggested solution to Seminar 2 - Part 1, by group C, and here is the suggested solution to Seminar 2 - Part 2, by group D.

I will go through the exercises and comment on the suggested solutions tomorrow (20/9) at the seminar.

Problems for seminar 2

The seminar problems for the second seminar are found here.

Slides for lectures on empirical methods

The next two lectures will focus on empirical methods much used in labor market research and policy evaluation. You can download the slides for the lectures here.

For the first lecture I recommend reading Angrist and Pische, Mostly harmless econometrics, p. 3-68

Much of the same material is also covered in the textbook “labour economics”, but in a more condensed form, on pages 668-674.

Suggested solutions by group A and B - Seminar 1.

Here is the suggested solution to Seminar 1 - Part 1, by group A, and here is the suggested solution to Seminar 1 - Part 2, by group B.

I will go through the exercises and comment on the suggested solutions tomorrow (6/9) at the seminar.

Job search: Lecture notes and reading

The slides from today’s lecture are available here.

A background paper on the study of labor markets flows are The flow approach to labor markets: New data sources and micro-macro links.

You can read more on the Leibniz rule for differentiation of integrals can be found on Wikipedia.

One of the first papers studying the optimal path of unemployment benefits over time is The optimal payment of unemployment insurance benefits over time by Steven Shavell and Laurence Weiss (1979). You should at least read the introduction. I have written a short note which may help in understanding their argument, it is found here.

A more general paper assessing unemployment policy is: Unemployment insurance: Strenghtening the rrelationship between theory and policy.

Finally, I also recommend three papers that studies what I called the “benefit trap” empirically. These are Moral hazard and sickness insurance, The sick pay trap and Temporary disability insurance and labor suply: Evidence from a natural experiment.