Econometrics-I

BSC- III
Spring 2020
Lahore School of Economics
Final-Term Examination
Name: __________________ Section: ______
Student ID: ______________ Total Marks: 80

Section II

Question 1 (20 marks):

The following graph depicts the number of deaths in Beijing due to the Coronavirus starting
from January 22nd up to 10th May for the year 2020. The vertical axis represents the number of
deaths in Beijing due to the coronavirus and the horizontal axis represents the time frame
(starting from 22nd January, 2020 up to 10th May, 2020).
600
Corona Deaths in Beijing

400
200
0

01feb2020 01mar2020 01apr2020 01may2020

date

a) We estimated the following regression using the above data.

. reg coronadeathsinbeijing date aprildummy date_aprildummy

Source SS df MS Number of obs = 110

F(3, 106) = 197.87
Model 2423034.44 3 807678.146 Prob > F = 0.0000
Residual 432668.117 106 4081.77469 R-squared = 0.8485
Adj R-squared = 0.8442
Total 2855702.55 109 26199.106 Root MSE = 63.889

coronadeath~ing Coef. Std. Err. t P>|t| [95% Conf. Interval]

date 4.701919 .238872 19.68 0.000 4.228332 5.175506

aprildummy 95016.93 30136.62 3.15 0.002 35268.16 154765.7
date_aprildummy -4.314822 1.368649 -3.15 0.002 -7.028301 -1.601342
_cons -102951.1 5250.232 -19.61 0.000 -113360.2 -92542.04

Page 1 of 5
 Variables Description
date Time span starting from January 22nd till May 10th, 2020.
aprildummy =1 for days in the month of April and 0 otherwise
date_aprildummy date*aprildummy (interaction term)
t Time variable
tsq t*t= time squared
Looking at the regression results above, what can you conclude about the coronavirus
deaths in Beijing? Explain carefully. (6 marks)

b) It has been theorized that higher temperatures may reduce the incidences of coronavirus
(and thus deaths). Therefore, it could be very useful to take the daily data on deaths in
Beijing due to coronavirus (coronadeathsinbeijing) and regress it against daily high
temperatures(temp_beijing) and a time trend (t). This relationship is tested in the
regression below.
. reg coronadeathsinbeijing temp_beijing t

Source SS df MS Number of obs = 110

F(2, 107) = 270.58
Model 2384281.94 2 1192140.97 Prob > F = 0.0000
Residual 471420.615 107 4405.80014 R-squared = 0.8349
Adj R-squared = 0.8318
Total 2855702.55 109 26199.106 Root MSE = 66.376

coronade~ing Coef. Std. Err. t P>|t| [95% Conf. Interval]

temp_beijing -1.102413 1.40715 -0.78 0.435 -3.891925 1.687099

t 4.881535 .3736194 13.07 0.000 4.140878 5.622192
_cons 192.6245 13.21727 14.57 0.000 166.4228 218.8262

What can you say about the relationship between the coronavirus deaths in Beijing and
daily high temperatures? (6 marks)

c) For the validity of regression analysis, it is crucial to know whether or not dependent or
independent variables are highly persistent. We ran the Dickey-Fuller test on both of our
variables. Interpet the results given below and clearly state the null and the alternative
hypothesis while intrepreting your results. (8 marks)

. dfuller coronadeathsinbeijing

Dickey-Fuller test for unit root Number of obs = 109

Interpolated Dickey-Fuller
Test 1% Critical 5% Critical 10% Critical
Statistic Value Value Value

Z(t) -9.295 -3.507 -2.889 -2.579

MacKinnon approximate p-value for Z(t) = 0.0000

. dfuller temp_beijing

Dickey-Fuller test for unit root Number of obs = 109

Interpolated Dickey-Fuller
Test 1% Critical 5% Critical 10% Critical
Statistic Value Value Value

Z(t) -2.536 -3.507 -2.889 -2.579

MacKinnon approximate p-value for Z(t) = 0.1070

Page 2 of 5
Question 2 (15 marks): For each of the following estimated equations of weekly wage (in $) on
experience (in years), interpret in a sentence the relationship between experience and wage:
̂
(𝑎) log(𝑤𝑎𝑔𝑒) = 2 + 1.4log⁡(𝑒𝑥𝑝𝑒𝑟)

(𝑏)𝑤𝑎𝑔𝑒
̂ = 100 + 100 log(𝑒𝑥𝑝𝑒𝑟)

̂
(𝑐) log(𝑤𝑎𝑔𝑒) = 4 + 0.06𝑒𝑥𝑝𝑒𝑟

Question 3 (26 marks): Using data from Pakistan Demographic and Health Survey (2012-13) of
11,274 women—which includes a measure of autonomy of the women surveyed (a scale from 0-
10, 10 being the most autonomous), a dummy for husband’s education greater than primary school
(husb_edu), and a dummy for a Rural location (rural)—we estimate the following model:
. reg autonomy husb_edu Rural

Source SS df MS Number of obs = 11,274

F(2, 11271) = 99.93
Model 586.476453 2 293.238227 Prob > F = 0.0000
Residual 33075.3934 11,271 2.93455713 R-squared = 0.0174
Adj R-squared = 0.0172
Total 33661.8699 11,273 2.98606137 Root MSE = 1.7131

autonomy Coef. Std. Err. t P>|t| [95% Conf. Interval]

husb_edu .1597918 .0344256 4.64 0.000 .0923116 .2272719

Rural -.4001932 .0330415 -12.11 0.000 -.4649604 -.3354261
_cons 1.986886 .0349405 56.86 0.000 1.918396 2.055375

a) Fully interpret the result of location and the role of husband’s education on women’s
autonomy. Use p-values to show how confident you are of each variable’s impact on autonomy.
(6 marks)
b) Find the predicted autonomy scores for the following women: (10 marks)

Live in Urban Areas

Husband has primary education Part b (i)

Husband has more than

Part b (ii)
primary education

Page 3 of 5
c) From your model in (d), assume woman’s autonomy only depends on education and region.
Now, using a graph with autonomy on the vertical axis and education on the horizontal axis,
show the relationship between autonomy and education separately for urban and rural areas.
Clearly label your intercepts and slopes. (10 marks)
Question 4 (10 marks): Consider the following model of car sales:

ln(𝑞𝑛𝑐) = 𝛽0 + 𝛽𝑝 ln(𝑝𝑟𝑖𝑐𝑒) + 𝛽𝑖𝑛𝑐 ln(𝑖𝑛𝑐𝑜𝑚𝑒) + 𝑢

where qnc is the number of cars sold (in thousands), price is the average price of new cars (in
$1,000), and income is the per capita income (in $1,000).
The estimation of the model with quarterly car sales and seasonal dummy variables in the U.S.
from 1975 to 1990 gives:

ln(𝑞𝑛𝑐) = 𝛽0 + 𝛽𝑝 ln(𝑝𝑟𝑖𝑐𝑒) + 𝛽𝑖𝑛𝑐 ln(𝑖𝑛𝑐𝑜𝑚𝑒) + 𝛽𝑊 𝑤𝑖𝑛𝑡𝑒𝑟 + 𝛽𝑠𝑝 𝑠𝑝𝑟𝑖𝑛𝑔 + 𝛽𝑠 𝑆𝑢𝑚𝑚𝑒𝑟 + 𝑢

where winter = 1 for the winter quarter and 0 otherwise, spring = 1 for the spring quarter and 0
otherwise, and summer = 1 for the summer quarter and 0 otherwise.

Using results from model 1 and 2, perform a joint test of significance on the parameters on the
seasonal variables. Do you find evidence of seasonality in car sales? (hint: the null in this case
would be that of no seasonality). [F-critical= 2.76] (10 marks)

Page 4 of 5
Question 5 (9 marks): Suppose an econometrician gathers information on the number of
cigarettes smoked by students in local universities in Lahore. The econometrician gathers data
from 100 students on the number of cigarettes smoked per day (smoke), the education of the
person in years (educ), the income of the person (income) in rupees thousands, and two dummy
variables: if the person lives in an area in which restaurants ban smoking (restban) and if the
person is a female (female).

a) Write a regression equation that would help you predict the effect of your independent
variables on the number of cigarettes smoked per day. (4 marks)

b) Write a regression equation that would help you predict the effect of your independent
variables on the number of cigarettes smoked per day if you knew that the effect of
income was dependent on whether an individual lived in an area where smoking was
banned in restaurants. (5 marks)

T-Critical values

Critical values for the t-distribution

t-crit Significance level
1.64 10%
1.96 5%
2.58 1%

Page 5 of 5