Applied Economics for Business BEE3062

UNIVERSITY OF EXETER BUSINESS SCHOOL
Homework Assignment
Applied Economics for Business
BEE3062
Maximum Marks: 100
Deadline: November 6, 2019
Before the end of deadline you must submit on ELE:
1. A document containing your answers to the questions, including any tables and
graphs.
2. A Stata do-file containing the commands you used to obtain the answers.
Materials to be supplied: Data file.
Page 1 of 5 { Applied Economics for Business (BEE3062)
Detailed Instructions:
Answer all questions. Feel free to use all course material, textbook, etc. to complete
this homework.
Submission is online i.e. you have to upload both the files on the ELE. The submitted
documents will contain your answers to the questions, including any output, tables and
graphs. Also, attach the do-file containing the commands you used to obtain your answers.
You must indicate which question your commands relate to. You may only submit once;
it is not possible to alter your documents and re-submit at a later time.
The data corresponding to this homework is uploaded on the ELE page (ref.dta).
Marking will follow the university’s assessment guidelines. Emphasis is placed on the
correct interpretation of summary statistics, regression coefficients and hypothesis tests.
For graphical analysis, emphasis is placed on the conclusions drawn from a graph, as well
as a mention of the limitations, if any.
Unless mentioned otherwise, assume a significance level of 5% (i.e. 95% confidence
level) for all statistical tests. Always state the null hypothesis and alternative hypothesis
of the test, as well as your decision at the stated level of significance.
Page 2 of 5 { Applied Economics for Business (BEE3062)
Introduction
People in the U.K. voted for EU referendum on 23 June, 2016. Each voter had a choice
between two options: Remain (U.K. remains a part of EU) and Leave (U.K. leaves the
EU). You are given a data set (ref.dta) which contains information on 267 cities across
England to study the factors that may have driven voters’ decision. A city is considered
to have voted for Remain if the number of votes for Remain is larger than the number of
votes for Leave; otherwise the city voted for Leave. The data contains following variables
for each city:

Variable Description
city Name of the city
region Geographical region of the city
pct turnout % of eligible voters who voted
remain Number of votes for Remain
leave Number of votes for Leave
rejected ballots Number of invalid votes
pct remain % of votes for Remain
pct leave % of votes for Leave
youngerpop % of population aged 16-29 (Younger Population)
midpop % of population aged 30-44
olderpop % of population aged 45+ (Older Population)
meanage Mean age
medianage Median age
totalpop Total population
ukpop Population of U.K. citizens
unemp Unemployment Rate
deprivationindex A measure of income, education, health and crime.
(a higher number implies more deprivation)
immipop % of immigrant population
vote =1 if a city has a larger number of votes for Leave as compared to Remain;
0 otherwise
turnout =1 if pct turnout is larger than 77; 0 otherwise

Page 3 of 5 { Applied Economics for Business (BEE3062)
I Basic Descriptive Statistics [35 marks]
1. [ 3 marks ] How many cities voted for Remain i.e. they received a larger number of
votes for Remain as compared to Leave?
2. [5 marks ] What fraction of cities that voted for Leave have a lower than average
unemployment rate and also have a lower than average deprivation index? Interpret
your answer.
3. [ 8 marks ] Plot a graph of your choice to distinguish the unemployment rate distribution between cities which voted for Remain and Leave. Interpret the graph.
4. [ 7 marks ] How many geographical regions in the data have the mean unemployment
rate larger than their median unemployment rate? Interpret your answer.
5. [ 6 marks ] Calculate the total number of invalid votes for each geographical region.
6. [ 6 marks ] Claim: Cities that voted for Leave (vote=1) have a higher % of eligible
voters who voted (pct turnout) than the cities that voted for Remain (vote=0), on
average. True or False.
II Linear Regression Analysis [65 marks]
pct leave = β0+β1immipop+β2deprivationindex+β3olderpop+β4olderpop×unemp+error
(1)
1. [ 10 marks ] Estimate a linear regression model with pct leave as the dependent
variable and immipop, deprivationindex, olderpop and (olderpop × unemp) as the
independent variables (as shown in equation (1)). Interpret the meaning of the
coefficients on immipop and (olderpop × unemp).
2. [ 10 marks ] Plot a graph of your choice to comment on whether the residuals of the
above regression model in equation (1) are normally distributed. Conduct a test to
confirm your findings.
3. [ 10 marks ] What is the average elasticity of pct leave with respect to immipop in
the above regression model in equation (1)? Construct a 95% confidence interval
around the average elasticity.
4. [ 12 marks ] Use the regression model in equation (1) to predict the outcome of voting
in the Edinburgh City. The characteristics of Edinburgh City are: olderpop=38.2%
; unemp=5.4% ; deprivationindex=28,000 and immipop=15.8%. Based upon the
prediction, did Edinburgh City vote for Leave?
Page 4 of 5 { Applied Economics for Business (BEE3062)
5. [ 8 marks ] Plot a scatterplot graph of the residuals from the regression model in
equation (1) versus the deprivationindex. Discuss it.
pct leave = β0 + β1immipop + β2deprivationindex + β3olderpop + β4olderpop × unemp
+β5olderpop2 + error
(2)
6. [ 15 marks ] Estimate the regression model above in equation (2). Answer the
following questions based on this new regression model:
(a) [ 5 marks ] In terms of the goodness-of-fit, is this new regression model in
equation (2) preferred over the model in equation (1) ?
(b) [ 10 marks ] The marginal effect of olderpop on pct leave at an unemployment
rate of 7% in a city is zero. What is the percentage of older population in this
city?
||| End of Assignment |||
Page 5 of 5 { Applied Economics for Business (BEE3062)

Leave a Reply

Your email address will not be published. Required fields are marked *