ECON0004: APPLIED ECONOMICS
Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit
ECON0004: APPLIED ECONOMICS
FINAL COURSEWORK ASSESSMENT
Answer ALL questions. [Word limit: 2000 words]
Each of questions one and two carries 30 per cent of the total mark. Question three carries 40 per cent of the total mark.
Academic Misconduct is strictly prohibited, including the use of essay mills, homework help sites, plagiarism, collusion,falsification, impersonation or any other action which might give me an unfair advantage. By submitting this assessment, I pledge my honour that I have not violated UCL's Assessment Regulations which are detailed in https://www.ucl.ac.uk/academic-manual/chapters/chapter-6-student-casework-framework/section-9-student-academic-misconduct-procedure.
Q1 [30 MARKS]. A researcher investigates the effect of youth unemployment on later wages in the UK. He uses data that follows people born in May 1970 since their birth and runs the regression as follows:
lnwi = β0 + β1 Unempi, 1-2+ β2 Unempi, 3-4+ β3 Unempi, 5-6+ β4 Unempi, 7- 12+ β5 Unempi, 13++εi
where lnwi is the log wage of individual i at age 35 and Unempi, a-b is a dummy variable indicating whether individual i experienced a period of unemployment of a to b months between the ages of 16 and 23. For example, Unempi, 1-2 = 1 if individual i was unemployed between 1 and 2 months when she/he was between the ages of 16 and 23 (and 0 otherwise). Unempi, 13+ = 1 if individual i was unemployed for more than 13 months when she/he was between 16 and 23 years old (and 0 otherwise). εi is the error term. The measures of youth unemployment are based on questions asked retrospectively to the individuals about their
experience in the labour market between 16 and 23 years old and their wages at age 35. The results from this regression are reported in the table below:
|
Coefficient |
Standard Error |
Constant |
1.36 |
0.87 |
Unempi, 1-2 |
-.015 |
0.04 |
Unempi, 3-4 |
-0.13 |
0.03 |
Unempi, 5-6 |
-0.14 |
0.05 |
Unempi, 7- 12 |
-0.25 |
0.06 |
Unempi, 13+ |
-0.48 |
0.09 |
Sample size= 16,500; R2=0. 13
(a) Based on the regression results, explain the effects of unemployment for more than one month between 16 and 23 years old on later wages. Use literature to support your arguments.
(b) Discuss two problems with this empirical model estimating the effects of youth unemployment on later wages and propose solutions to these problems.
(c) Suppose the researcher wants to understand whether the effects of youth unemployment on later wages differ between female and male. Propose a regression model to investigate this. Carefully explain what variables are needed in the regression and which regression coefficients measure the differences. Show the steps in detail.
Q2 [30 MARKS]. A researcher is investigating the differences in consumption across different counties in the UK. He assumes that a household is choosing how to allocate consumption across periods of its life by maximising lifetime utility V, such that:
V = ut (ct ) where ut (ct ) = ct(a) ,
where ct is consumption in the t-th period and ut (.) is a within-period utility function for the t-th period. T is end of life and a < 1.
The researcher has data from a series of monthly household consumption surveys in the UK at the county level for a period of 40 years. He uses these data to run the following OLS regression:
ΔlnCit =β0+β 1 rt + µit
where Cit is average consumption expenditure in the i-th county of the t-th period, ΔlnCit is the change in the logarithm of Cit across two periods (Δ ln Cit = ln Ci,t+1 − ln Cit ), and
rt is the real interest rate between periods t and t+1.
He gets the following estimates:
|
Coefficient |
t-ratio |
Constant |
-0.2 |
- 1.5 |
r |
2.0 |
2.4 |
(a) Use the estimated results to calculate the estimates of 6 and a. Show the derivations in detail.
(b) Can we reject that a = 0.75 at the 5% level of significance? Can we reject that 6 = 0 at the 5% level of significance?
(c) In order to investigate this further, the researcher estimates the following model ΔlnCit =β0+β 1 rt + β2ΔlnYit + µit
where Yit is income per person in the i-th county of the t-th period, ΔlnYit is the expected change in the logarithm of Yit between period t and t+1.
The researcher splits the sample into two groups: renters and homeowners, and estimates the model separately for each group and gets the following estimates:
|
Group 1: Renters |
Group 2: Homeowners |
||
|
Coefficient |
Standard Error |
Coefficient |
Standard Error |
Constant |
-0.23 |
0.14 |
-0.18 |
0.12 |
r |
2.17 |
0.98 |
2.28 |
1. 12 |
ΔlnY |
0.06 |
0.03 |
0.02 |
0.18 |
Do the above results suggest that there is one particular group for whom the prediction of the life-cycle model does not hold? If so, identify this group and explain why this might be the case.
Q3 [40 MARKS]. The government in a country started a large welfare program targeting poor rural households. One major aim of the program was to alleviate child labour by giving a monthly cash grant to mothers provided their children showed up at school every day. Economists have shown that the program has had a major impact on reducing child labour and increasing educational attainment of children. They are now interested in whether the programme affected mother's labour supply.
You are given data on weekly working hours for a sample of women living in the rural areas of the country. You get the following estimated OLS regression:
li = 9.78 + 9. 12 wi - 0.76 wi2 – 0.05 hi + 0.87 Ni - 6.12 ni + 1.82 gi + (2.12) (3. 18) (2.56) ( 1.98) ( 1.02) (3.03) ( 1.03)
where li is weekly hours of work, wi the mother's hourly wage in £, hi is husband's income in £ per week, Ni is the number of children aged 6- 16 living in the household, ni is the number of children aged 1-5 living in the household, gi is a dummy variable that takes the value 1 if a grandparent lives in the household, is the residual term. The numbers in parentheses are absolute t ratios.
(a) Give one economic reason why the cash grant program could affect the mothers' labour supply. Does the theory predict an unambiguous effect of the cash allowance on mothers' labour supply? Explain the reasoning. Use literature to support your arguments.
(b) Consider a household with 2 children (aged 5 and aged 10), in which the mother earns £4 per hour, the father earns £180 per week and the mother's mother lives at home.
(i) How many hours would you expect the mother in this family to work?
(ii) Would an increase in this mother's wage lead her to choose more or fewer hours of work? What does this imply about the relative size of income and substitution effects? Use a diagram to illustrate your points.
(c) You ran this regression using the sub-sample of women who work, since you do not observe wages of women who do not work. Do you think this might be problematic? Why or why not?
(d) The cash grant program was randomized across villages. Your dataset contains a dummy variable Di that takes the value 1 if woman i was in a village that received the cash grant and 0 otherwise.
(i) Do you think if this randomization helps you with identifying the causal effect of the program on mothers' labour supply? If so, in what way? Be specific about the assumption it would allow you to validate.
(ii) How would you specify the regression you would run to estimate the effect of the
program on the labour supply of the women who received the cash grant?
2023-05-04