ECN6540 Econometric Methods Autumn Semester 2023-2024
Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit
Department of Economics
Autumn Semester 2023-2024
ECN6540 Econometric Methods
Coursework
The answers to the questions must be type-written. The preference is that symbols and equations should be inserted into the document using the equation editor in Word. Alternatively, they can be scanned and inserted as an image (providing it is clear and readable). Maximum words 1,500 excluding any Stata output and commands.
The coursework comprises two questions where the second is a short Stata assignment. ANSWER ALL QUESTIONS. Both questions 1 and 2 carry equal weight and the marks shown within each question indicate the weighting given to component sections. Any calculations must show all workings otherwise full marks will not be awarded. In question 2 all Stata output should be shown.
Coursework must be submitted online through Blackboard using Turnitin and by no later than 12.00 noon on the deadline. Coursework submitted after the 12.00 noon deadline will have a late penalty applied. Details about the late penalty policy can be found in the Student Handbook.
You must attach a submission template coversheet to the front of your work when submitting it to Turnitin to avoid a 5% penalty. Full details of this policy can be found in the Student Handbook.
Please ensure that you have read the assessment guidelines provided in the Student Handbook, including the guidance about submission requirements, extension requests and extenuating circumstances and the use of unfair means.
Use of any generative AI tools in the preparation of the answer to this work is not permitted.
PLEASE WRITE YOUR STUDENT REGISTRATION NUMBER IN THE
SUBMISSION TITLE BOX AND USE THIS AS THE FILENAME YOU UPLOAD.
1. Using data from the U.S. Survey of Consumer Finances in 2019 the level of non-mortgage debt of 16,750 individuals was modelled as a function of: a cubic polynomial in age; whether the individual was married; their financial expectations about future income; whether they are very knowledgeable about their personal finances; and their log weekly wage rate, as shown in equation (1):
debtsi = β0 +β1 agei +β2 agei(2) +β3 agei(3) +β4maTTiedi +β5 expecti
+β6finknowi +β7 log(wagei ) +Ei
(eq. 1) where i is the unit of observation (individual) and log denotes the natural logarithm. Variables are defined as follows: maTTiedi is equal to 1 if married and 0 otherwise; expecti is equal to -1 if pessimistic (expect income to fall next year), 0 if no change in income expected, +1 if optimistic (expect income to increase next year); and finknowi equals 1 if the individual thinks they are very knowledgeable about their finances. The following Stata output shows the estimates.
regress debts c.age##c.age##c.age married expect finknow logwage
Source | SS df MS
-------------+----------------------------------
Model | 13695
Residual |
-------------+----------------------------------
Total | 4137700
Number of obs
F( , )
Prob > F
R-squared
Adj R-squared
= 16,750
=
=
=
=
debts | Coefficient Std. err. t P>|t| [95% conf. interval]
------------------+----------------------------------------------------------------
age |
| | |
-528.5081 |
130.2063 |
|
|
|
|
c.age#c.age |
| | |
12.46099 |
3.077809 |
|
|
|
|
c.age#c.age#c.age |
| | |
-.0924671 |
.0233559 |
|
|
|
|
married |
| |
-395.223 |
82.08281 |
-4.81 |
0.000 |
-556.1139 |
-234.332 |
expect |
| |
-80.52614 |
49.65763 |
-1.62 |
0.105 |
-177.8604 |
16.80807 |
finknow |
| |
280.6058 |
145.947 |
1.92 |
0.055 |
-5.465692 |
566.6773 |
logwage |
| |
35.97025 |
11.2247 |
3.20 |
0.001 |
13.96866 |
57.97184 |
_cons |
| |
7083.866 |
1764.3 |
4.02 |
0.000 |
3625.651 |
10542.08 |
sum debts age married expect finknow logwage
Variable | Obs Mean Std. dev. Min Max
-------------+---------------------------------------------------------
debts | 16,750 531.6209 4970.326 0 175000
age | 16,750 45.64818 11.65751 18 65
expect | 16,750 .0501493 .7748462 -1 1
finknow | 16,750 .9225672 .26728105 0 1
+---------------------------------------------------------
logwage | 16,750 10.25585 3.45366 0 17.62217
Interpret the results and test whether the age effects are individually statistically significant at the 1% level.
At what two values of age is the level of debt minimised?
For financial expectations and the log wage calculate the respective elasticity, based upon the sample mean.
Calculate the value of the RSS and the degrees of freedom associated with the ESS, RSS and TSS.
Test whether the parameters on the explanatory variables are jointly statistically significant at the 5% level.
Showing your calculation in full find the adjusted R-squared.
An alternative specification for equation (1) is to model the debt-to- income ratio as follows (note: incomei = wagei + benefitsi , where benefitsi represents benefit income, e.g. unemployment insurance):
= β0 +β1 agei +β2 agei(2) +β3 agei(3) +β4marriedi
+β5 expecti +β6finknowi +β7 log(wagei ) +Ei
What particular problems might be found with this model?
The initial model (equation 1) is now re-estimated changing the functional form by replacing the cubic function in age with binary indicators for different age bands, as shown in equation (2):
debtsi = β0 +Σk(3)=1βkageki + β4marriedi +β5 expecti +β6finknowi
+β7 log(wagei ) +Ei
(eq. 2) Age binary indicators are defined as follows: age1 = 1 if the individual is aged between 18-34, 0 otherwise; age2 = 1 if the individual is aged between 35-44, 0 otherwise; and age3 = 1 if the individual is aged between 45 and 54, 0 otherwise. The reference category is aged above 55. The following Stata output below shows the estimates.
i) Interpret the age coefficients explaining how they affect the level of debt. Are the results consistent with those found under the original functional form of equation (1)? Explain your answer.
ii) Which model is preferred (equation 1) or (equation 2), explain your answer? Show all calculations to reach your decision in full.
regress debts age1-age3 married expect finknow logwage
Source | SS df MS Number of obs = 16,750
-------------+---------------------------------- F( , ) = 6.18
Model | Prob > F =
Residual | 4127000 R-squared =
-------------+---------------------------------- Adj R-squared =
Total |
debts | Coefficient Std. err. t P>|t| [95% conf. interval]
-------------+----------------------------------------------------------------
age1 | 116.9167 -1.67 0.096 -423.9473 34.39083
age2 | 107.6893 -2.71 0.007 -502.494 -80.32902
age3 | 103.1392 -0.96 0.339 -300.687 103.6404
married | -416.1489 81.63058 -5.10 0.000 -576.1534 -256.1443
expect | -85.35759 49.65449 -1.72 0.086 -182.6856 11.97045
finknow | 276.0255 146.0629 1.89 0.059 -10.27332 562.3243
logwage | 37.13198 11.21791 3.31 0.001 15.1437 59.12026
_cons | 277.0043 186.8169 1.48 0.138 -89.17654 643.1852
Re-write the regression model shown in equation (2) to allow for differential wage effects with respect to age. The TSS and ESS from estimating this model are 4137700 and 12890 respectively. Based upon this information test whether there are differential wage effects across age groups at the 5% level. [15 marks]
STATA ASSIGNMENT
2. The data is given in the Stata file ECN6540_Assignment.dta, which is time series data for the U.S. over the period 1986 to 2001 (year is the time identifier in the data). Variables in the data are defined as follows: Y is sales of beer per capita; X1 is the tax rate on beer (%); and X2 is disposable income (in dollars, $).
Load the data into Stata. Then type the following commands where the number after "set seed" is your student registration number e.g. 200212232 (this ensures that each student has unique data):
set seed 200212232
replace X2=X2*abs(rnormal(0,1))
save “ECN6540_Assignment_mydata.dta”, replace
Load your unique data set into Stata:
use “ECN6540_Assignment_mydata.dta”, clear
To achieve full marks in the questions which follow all Stata output must be provided.
Estimate a model of beer sales per capita conditional on beer tax and income, i.e. yt = β0 +β1x1t +β2x2t +Et . Present the estimates of your model and interpret the results. For disposable income base the interpretation on a $10,000 increase in income.
Based upon the sample means calculate tax and income
elasticities.
Undertake the Durbin Watson test for autocorrelation at the 1% level and interpret your findings.
Undertake the White test for heteroscedasticity at the 5% level (without using any inbuilt Stata test commands).
Now re-estimate the model from part (a) to allow for the inclusion of a lagged dependent variable and a quadratic term in income,
i.e. yt = yyt-1 +β0 +β1x1t +β2x2t +β3x2(2)t +Et .
i) What does the quadratic term in income suggest?
ii) Test for autocorrelation at the 5% level (without using any inbuilt Stata test commands).
At the end of the coursework provide the text from your Stata *.do file in the word document.
2023-11-20