## Examination Paper

Examination Session: May/June

Year:2021

Exam Code:ECON2061-WE01

Title: ECONOMIC DATA ANALYSIS

Release Date/Time  13 th May 09:30

Latest Submission Date/Time   14th May 09:30

Format of Exam  Take home exam

Duration:2 Hours

Word/Page Limit:3000 words

Additional Material provided:Statistical tables are attached to the end of the paper    经济数据分析代考

Expected form of Submission Word Document using template provided or PDF of Handwritten Work

Your uploaded file should be named with your Anonymous Code and the Exam Code e.g Z0123456 ECON1111-WE01

Submission method   Turnitin

Instructions to Candidates: Answer ONE question.

The marks available for each sub-question are given in brackets.

### 1.A researcher wishes to estimate the causal effect of years of education on wages so collects data on 900 individuals aged between 30 and 40 in 2020. 经济数据分析代考

Variable descriptions are provided in Table 1. The researcher estimates five models and obtains the results in Table 2 below.The dependent variable in all models is 𝑙𝑤𝑎𝑔𝑒. Heteroskedasticity-robust standard errors are in brackets. 𝑅 2 is the coefficient of determination. OLS denotes ordinary least squares and TSLS denotes two stage least squares.

a) Interpret the estimated coefficient on 𝑒𝑑𝑢𝑐 in column 1 of Table 2.(10 marks)

b) The researcher now substitutes four binary variables, 𝑒𝑑𝑢𝑐_13, 𝑒𝑑𝑢𝑐_14, 𝑒𝑑𝑢𝑐_15 and 𝑒𝑑𝑢𝑐_16, in place of 𝑒𝑑𝑢𝑐. Outline any advantages of this specification of the model.(10 marks)

### c) The results from estimation of this model are presented in column 2. Interpret the estimated coefficients on 𝑒𝑑𝑢𝑐_13, 𝑒𝑑𝑢𝑐_14, 𝑒𝑑𝑢𝑐_15 and 𝑒𝑑𝑢𝑐_16.(10 marks)

d) The 𝑅 2 in column 2 is higher than the 𝑅 2 in column 1. Would this also be true of the adjusted 𝑅 2? Which measure of fit do you consider more useful? Explain.(10 marks)

e) Comment on and provide an explanation for the difference between the estimated coefficients on 𝑒𝑑𝑢𝑐 in columns 1 and 3.(10 marks)

f) The researcher now decides to use 𝑚𝑒𝑑𝑢𝑐 as an instrumental variable for 𝑒𝑑𝑢𝑐. The results from two stage least squares estimation are given in column 4.Discuss whether 𝑚𝑒𝑑𝑢𝑐 will satisfy the conditions to be a valid instrumental variable.(15 marks)

g) Consider the results in column 5. What do the estimated coefficients suggest about the effect of education on wages?(10 marks)  经济数据分析代考

h) An F test of the null hypothesis that the coefficients on 𝑠𝑜𝑢𝑡ℎ, 𝑠𝑜𝑢𝑡ℎ × 𝑒𝑑𝑢𝑐 and 𝑠𝑜𝑢𝑡ℎ × 𝐼𝑄 are jointly equal to zero yields a test statistic of 7.33. What is the conclusion from the test(use a 5% significance level)?(5 marks)

i) Using the results in column 5, predict the value of 𝑙𝑤𝑎𝑔𝑒 for an individual with 12 years of education and an IQ score of 100 that lives in:

i.The south of the country(5 marks)

ii.The north of the country(5 marks)

j) The researcher is considering collecting information on the same variables for 2021 to allow the use of fixed effects estimation. Would you advise the researcher to do this? (10 marks)

### 2.A researcher wishes to estimate the causal effect of exporting on productivity. The datasetcontains information on a random sample of 10,000 firms for 2010, 2011 and 2012. Variabledescriptions are provided in Table 1.   经济数据分析代考

The researcher estimates five models and obtains the results in the table below. The dependent variable in all models is 𝑝𝑟𝑜𝑑𝑢𝑐𝑡𝑖𝑣𝑖𝑡𝑦. Clustered standard errors are in brackets. 𝑅 2 is the coefficient of determination. OLS denotes ordinary least squares, TSLS denotes two stage least squares and FE denotes fixed effects(within) estimation.

a) Interpret the estimated coefficient on 𝑒𝑑𝑢𝑐 in column 1 of Table 2.(10 marks)

b) Comment on and provide an explanation for the difference between the estimated coefficients on 𝑒𝑥𝑝𝑜𝑟𝑡 in columns 1 and 2.(10 marks)

c) Interpret the value of the 𝑅 2 in column 2. Does it indicate that the model is not appropriate for estimating the causal effect of exporting on productivity?(10 marks)

### d) Using the results in column 2, predict the value of 𝑝𝑟𝑜𝑑𝑢𝑐𝑡𝑖𝑣𝑖𝑡𝑦 for an exporting firm with £40,000 of capital stock and 200 employees.(5 marks)  经济数据分析代考

e) Column 3 presents results from two stage least squares estimation. The instrumental variable used by the researcher is the materials-labour ratio (𝑚𝑙_𝑟𝑎𝑡𝑖𝑜). Discuss whether this variable will satisfy the conditions to be a valid instrumental variable.(20 marks)

f) Explain why it is not possible to conduct a statistical test of whether the instrumental variable used to obtain the results in column 3 is exogenous.(10 marks)

g) Comment on and provide an explanation for the difference between the estimated coefficients on 𝑒𝑥𝑝𝑜𝑟𝑡 in columns 2 and 4.(10 marks)

h) Explain why no estimate of the intercept is provided in columns 4 or 5.(10 marks)

i) Interpret the estimated coefficients on 𝑒𝑥𝑝𝑜𝑟𝑡 and 𝑒𝑥𝑝𝑜𝑟𝑡 × 𝑓𝑜𝑟𝑒𝑖𝑔𝑛 in column 5. (10 marks)

j) The researcher conducts an F test of the null hypothesis that the coefficients on 𝑒𝑥𝑝𝑜𝑟𝑡 and 𝑒𝑥𝑝𝑜𝑟𝑡 × 𝑓𝑜𝑟𝑒𝑖𝑔𝑛 are jointly equal to zero and obtains a test statistic of 𝐹 = 2.69. What is the conclusion from the test (use a 5% significance level)?(5 marks)  