konometri

Hjemmeopgave 1
Baseret p
a Wooldridge (2016):
Introductory Econometrics, 6th edition, kap. 1-2
Afleveringsfrist: Mandag d. 15. februar, 2016 kl. 8:30.

Opgaven skal besvares med jeres egne ord, enten p

a engelsk eller p
a dansk. R output skal ikke
inkluderes. Figurer, tabeller og R-kode vedlgges som et appendiks. Anfr navnene p
a samtlige
medlemmer af gruppen p
a forsiden. Opgaven uploades samlet til Absalon i en enkelt PDF fil.

1 Job training I
A justification for job training programmes is that they improve worker productivity. Suppose
that you are asked to evaluate whether more job training makes workers more productive. However, rather than having data on individual workers, you have access to data on manufacturing
firms in Demark. In particular, for each firm, you have information on hours of job training per
worker (training) and number of nondefective items produced per worker hour (output).
1a. Carefully state the ceteris paribus thought experiment underlying this policy question.
1b. Does it seem likely that a firms decision to train its workers will be independent of worker
characteristics? What are some of those measurable and unmeasurable worker characteristics?
1c. Name a factor other than worker characteristics that can affect worker productivity.
1d. Suppose that you find a positive correlation between output and training, would you have
convincingly established that job training makes workers more productive? Explain.

2 Job training II
Load the dataset jtrain2.RData related to a job training experiment in the 1970s.1 Look at
the description of the variables and answer the following questions.
1

This data was used in the paper by R.J. Lalonde (1986), Evaluating the Econometric Evaluations of Training
Programs with Experimental Data, American Economic Review 76, 604-620. Professor Lalonde obtained the

2a. What is the structure of this dataset (see Wooldridge, section 1.3)? Are the data observational or experimental? What do the observations (rows) in this dataset represent?
2b. Use the indicator variable train to determine the fraction of men receiving job training.
Visualize and compare the number of men receiving job training and the number of men
not receiving job training using a bar plot.
2c. The variable re78 is earnings from 1978, measured in thousands of 1982 dollars. What
is the unconditional sample mean of re78 ? What is the conditional sample mean of re78
among those receiving job training? What is the conditional sample mean of re78 among
those not receiving job training? Are the salaries in the two groups very different? Use a
box plot to visualize and compare the distribution of re78 for the two groups.
2d. The variable umem78 is an indicator of whether a man is unemployed or not in 1978. What
fraction of the men who received job training are unemployed? What about the men who
did not receive job training? Use a spine plot to illustrate the relationship between train
and umem78. Comment on the differences.
2e. Based on your answers to questions 2c and 2d, does it seem that the job training program
was effective? What would make your conclusions more convincing?

3 Birth weight
Please load the dataset bwght.RData related to the birth weight of a sample of children and
mothers smoking habits.2 Look at the description of the variables and answer the following
questions.
3a. What is the structure of this data set (see Wooldridge, section 1.3)? Are the data observational or experimental? What do the observations (rows) in this data set represent.
3b. How many women are in the sample and how many report smoking during pregnancy?
3c. What is the average number of cigarettes smoked per day? Illustrate the distribution of
this variable graphically using a histogram. Is the average a good measure of the typical
woman in this case? Explain.
3d. Among women who smoked during pregnancy, what is the average number of cigarettes
smoked per day? How does this compare with your answer from question 3c, and why?

data from the National Supported Work Demonstration job-training program conducted by the Manpower
Demonstration Research Corporation in the mid 1970s. Training status was randomly assigned.
2
This data was used in the paper by J. Mullahy (1997), Instrumental-Variable Estimation of Count Data Models:
Applications to Models of Cigarette Smoking Behavior, Review of Economics and Statistics 79, 596-593.
Professor Mullahy obtained the data from the 1988 National Health Interview Survey.

3e. Estimate the simple regression model

bwght = 0 + 1 cigs + u
What are the estimates of 0 and 1 ?
3f. Illustrate the relationship between cigs and bwght in a scatter plot. Add the estimated
regression line.
3g. What is the predicted birth weight when cigs = 0? What about when cigs = 20 (one
package per day)? Comment on the difference.
3h. Does this simple regression necessarily capture a causal relationship between the childs
birth weight and the mothers smoking habits? Explain.
3i. To predict a birth weight of 125 ounces, what would cigs have to be? Comment.
3j. Create a new variable cigsPkg that indicates the smoking habits of the mother during pregnancy, measured as the average number of cigarette packages per day, where it is assumed
that a package always contains exactly 20 cigarettes. Estimate the simple regression model
bwght = 0 + 1 cigsP kg + u,
interpret the estimates of 0 and 1 and calculate the predicted birth weight when cigsP kg =
0 and when cigsP kg = 1 (20 cigarettes per day). What are the differences between this
model and the original model with cigs as the explanatory variable?