Sie sind auf Seite 1von 8

ECON 141 UC Berkeley Fall 2011 Practice midterm 1

Please do not turn this page over until you are instructed otherwise.

Instructions. PLEASE READ CAREFULLY before starting.

This test has a total of 75 points. You have 1h 15m to solve it, that is, 75 minutes. Show your work! No credit will be given for correct answers if you are not justifying your argument. Please make sure that your handwriting is legible! In essay questions, be precise but brief. If a correct reply is hidden among wrong, or irrelevant, arguments, you will not get full credit. We will grade only what is written on your exam sheet. There should be plenty of space for all your calculations. Do not turn in anything besides your exam sheet. If time is running short, set up the problem without nal calculations.

Name:

1. (5 points) You know that income per head in France (denoted by y ), expressed in Euros, is normally distributed, that is, 2 y N y , y You also know that the mean y is Euros 16,000, and the standard deviation of y is 2000. What is the distribution of income per head in France, in US$, if the exchange rate is 1.50 US$ per one Euro?

2. Using a sample of 534 i.i.d. observations from the 1985 CPS (Current Population Survey), you have estimated that the relation between wages Wi and years of schooling Si is described by the following relation ln Wi = 0 + 1 Si = 1.06 + 0.077 Si

(.10743)

(.00809)

where heteroskedasticity-robust standard errors are reported in parenthesis, and ln Wi is the natural logarithm of the hourly wage. You can assume that all the usual OLS assumptions hold. (a) (4 points) What is the economic interpretation of the estimated slope in the above regression?

(b) (3 points) Does the relation between schooling and wage appear important, in economic terms?

(c) (3 points) What is the expected level of (ln Wi ) for an individual with 15 years of schooling?

(d) (4 points) In the above regression, is the slope signicantly dierent from zero, using a 1% signicance level?

(e) (5 points) You want to test the null hypothesis that 1 = 0.1, against a two-sided alternative. Calculate the p-value for this test.

(f) (5 points) Calculate a 95% condence interval for the predicted eect on (ln Wi ) of increasing schooling by 3 years.

(g) (4 points) Would you expect the point estimates of the coecients 0 and 1 to change if you did not assume homoskedasticity? Why, or why not?

(h) (4 points) You calculate the standard errors assuming homoskedasticity, and the estimated standard errors for 0 and 1 are respectively 1.05986 and 0.00807. Do you think heteroskedasticity is a serious concern in this regression?

Now you complicate the model, since you want to study whether the relation between wage and schooling changes if the individual is a Trade Union member. You estimate the following regression ln Wi = 0.91 + 0.75 Ui + 0.08 Si 0.035 Ui Si

(0.13) (0.25)

(0.01)

(0.019)

where Ui is a dummy variable equal to 1 if the worker is a Trade Union member, and zero otherwise. (i) (5 points) What is the predicted value of (ln Wi ) for a unionized individual with 12 years of schooling?

(j) (5 points) Does the eect of schooling on (logarithm of) wages dier importantly, in economic terms, between unionized and non-unionized individuals?

(k) (5 points) Test the null hypothesis that the predicted eect of increasing the level of schooling on (ln Wi ) is the same for unionized and non-unionized individuals, against a two-sided hypothesis, with a 10% signicance level.

3. You want to study what is the expected eect of years of education (denoted by e ) on income (denoted by Y ). Unfortunately, income is measured with error. You want to know what kind of problems this implies for your estimates. You have a sample of n i.i.d. observations (Yi , ei ) , where Yi is observed income, measured with error. You also know that the relation between observed income Yi and true income Yi is described by the following relation Yi = Yi + i where i is an i.i.d. reporting error. The relation between true income and education is described by the following equation (without the intercept, to keep things simpler): Yi = ei + ui E [ui | ei ] = 0 You also know that the reporting error is uncorrelated with the regression error ui , and that it is on average zero for every possible level of education ei , so that E [i | ei ] = 0 You want to estimate using OLS, but since you do not observe true income Yi , your OLS estimator will be the solution to the following problem
n

(1)

min
i=1

(Yi ei )2

(2)

(a) (5 points) Show that the OLS estimator for the slope is =
n i=1 ei Yi n 2 i=1 ei

(b) (5 points) Show that you can rewrite the estimator as =+


n i=1 ei ui n 2 i=1 ei

n i=1 ei i n 2 i=1 ei

(c) (3 points) Prove that E (i | e1 , ..., ei , ..., en ) = 0.

(d) (5 points) Using the results from the previous points, is an unbiased estimator of the true slope ? Justify your answer as rigorously as you can.

(e) (5 points) Do you think that the variance of the estimator will be aected by measurement error in reported income? Why? Or why not? Be as precise as you can, but an intuitive argument will earn you points as well!

Das könnte Ihnen auch gefallen