Econometrics I Econ 403 – Fall Semester 2018-2019
Econometrics代写 nswer all parts of Questions 1, 2 and 3 below.The datasets that you are required to use for this examination are available
Final Examination: 100 points
Due: 11:59 PM on Tuesday, December 18, 2018
- Answerall parts of Questions 1, 2 and 3
- The datasets that you are required to use for this examination are available as csv files in the ”Examination Final – TakeHome” myCourses content module.
- You must use R to produce the statistics that are asked for in this examination. Results produced using any other software will not be accepted.
- You must submit both the R code and the corresponding R output along with any written responses that are asked for in each question of this examination.
- Please paste the R code, the R output, and your written responses into one MS Word or PDF document. Please label each set of code, output and the written responses so that it is clear which specific question it is associated with.
- Post that document to the Examination Final – TakeHome dropbox on my- Courses by the due date given above.Econometrics代写
- Please see RIT’s Student Academic Integrity Policy at the link below. In par- ticular, see section IV. ACADEMIC INTEGRITY points out that cheating in- cludes using unapproved sources. Other students’ exam work qualifies as an un- approved source. Thus, work on this examination on your own. A student will earn zero points on any question response that has been copied from an- other student. The student who allows their work to be copied will also earn zero points on that question.
- I will be available to answer clarifying questions and to provide any help that you might need with the R code. Please see myCourses news for the times of availability.
Question 1: 35 points Econometrics代写
This question requires you to analyze data that are available in the voucher1.csv data file. The file includes the variable names in the first row.
The overall objective of the analysis is to estimate the effect of school choice on stu- dents’ academic achievement. The schools in question are for grades Kindergarten through 12th grade.Econometrics代写
These data are from the Milwaukee public school district for the school years 1991 to 1994. In this time period, the school district had a school choice program. This allowed students to apply for a school voucher. From the applicants, students were chosen by a lottery to receive a voucher, which paid for attendance at a school of the student’s own choice. (That is, the students who received vouchers were randomly selected from amongst those students who applied.)
The voucher allowed the student to attend their school of choice in the remaining school years of the 1990 to 1994 time period.Econometrics代写
That is, if the student applied for the program in 1990 and received a voucher in that year, then the student would be able to select a school of their choice for the four years covered by the data (school years 1991, 1992, 1993, and 1994). Similarly, if he or she applied in 1991 and received a voucher in that year, then the student would be able to select a school of their choice for the years 1992, 1993, and 1994; and so on. A student could not choose to attend their school of choice if they were not selected (by lottery) to receive a voucher.
The observations are students who applied to the voucher program and were selectedto receive a voucher, and students who applied but who were not selected to receive a voucher.
The variables from this dataset that you will need to use are the following:
mnce: Student percentile score on a standardized math test that was ad- ministered in 1994.Econometrics代写
selectyrs:The number of years that the student was selected (by lottery) to receive a voucher that allowed the student to attend their school of choice. This takes on integer values from 0 to 4.
choiceyrs:The number of years that the student received a voucher and that the student chose to attend their school of choice. This takes on integer values from 0 to 4.
selectyrs1: = 1 if selectyrs == 1. (selectyrs1 = 1 if a voucher was awarded in 1993). = 0 otherwise.
selectyrs2: = 1 if selectyrs == 2. (selectyrs2 = 1 if a voucher was awarded in 1992). = 0 otherwise.
selectyrs3: = 1 if selectyrs == 3. (selectyrs3 = 1 if a voucher was awarded in 1991). = 0 otherwise.Econometrics代写
selectyrs4: = 1 if selectyrs == 4. (selectyrs4 = 1 if a voucher was awarded in 1990). = 0 otherwise.
black: = 1 if the student is African American. = 0 otherwise.
hispanic: = 1 if the student is Hispanic. = 0 otherwise.
f emale: = 1 if the student is female. = 0 otherwise. Econometrics代写
a.UseOLS to estimate a regression model with mnce as the dependent variable and with the following as the explanatory variables: choiceyrs, black, hispanic, and f emale. Precisely interpret the parameter estimates for statistically significant variables. What does this analysis suggest about the effectiveness of this school choice program? 5 pointsEconometrics代写
b.Give at least one specific reason for why choiceyrs might be an endogenous ex- planatoryvariable in the model referred to in part 5 points
c.Use OLS to estimate a regression model with choiceyrs as the dependent vari- able and with the following as the explanatory variables: selectyrs1, selectyrs2, selectyrs3, and selectyrs 5pointsEconometrics代写
d.Usethe results from part c and your intuition to determine if the selectyrs Boolean variables are valid instruments for choiceyrs in the mnce regression model from part a. 5 points
e.Use 2SLS to estimate the regression model from part a. Use the 4 selectyrs Boolean variables as instruments for choiceyrs. Contrast the 2SLS and the OLS results. 5pointsEconometrics代写
f.Run the following tests and precisely interpret the resulting test statistics: 10 points
i.The test for weak
ii.Sargan’s test of over-identifying
Question 2: 45 points Econometrics代写
Use the dataset in guns.csv to answer the questions below. The data are for the 50 US states (plus the District of Columbia) during the 1980’s and the 1990’s. The variables that you will use are as follows:
mur: murder rate (number of murders per 100,000 members of popula- tion)
shall: = 1 if the state has a shall-carry law in effect in that year; = 0 otherwise
incarc rate: incarceration rate in the state in the previous year (sentenced pris- oners per 100,000 residents)Econometrics代写
density: population per square mile of land area, divided by 1,000
avginc: real per capita personal income in the state, in thousands of dollars
pb1064: percent of state population that is African American, for those aged between 10 and 64 yearsEconometrics代写
pw1064: percent of state population that is Caucasian, for those aged be- tween 10 and 64 years
pm1029: percent of state population that is male, for those aged between 10 and 29 years Econometrics代写
The purpose of the statistical analysis here is to examine the effect of ”shall-issue” laws on crime rates. ”Shallissue” laws allow citizens to carry concealed weapons. The laws are named this way because they instruct local authorities to issue a concealed weapons permit to all applicants who are citizens, are mentally competent, and have not been convicted of a felony. (Some states impose additional restrictions.) Proponents of these ”shall-issue” laws argue that, if more people carry concealed weapons, crime will decline because criminals are less likely to attack other people out of fear of retaliation.
Opponents argue that ”shall-issue” laws will increase crime because of accidental or spontaneous (unplanned) use of the concealed weapon.Econometrics代写
a.UseOLS to estimate a model with the natural log of mur as the dependent variable and the following explanatory variables: shall, incarc rate, density, the natural log of avginc, pb1064, pw1064, pm1029, and the following interaction variable: shall pm1029. Precisely interpret all of the statistically significant parameter estimates. Use the 5 percent significance level. 15 points
b.Forthe regression model in part a, run an F test of the linear restriction that the parameter for pb1064 equals the parameter for pm 9 points
i.Using the 5 percent significance level, interpret the resulting test Econometrics代写
ii.Within the context of this particular regression model and the data thatare used to estimate it, clearly explain the meaning of the hypothesis that is being tested
iii.Forthis F-statistic test, what is the difference between the restricted regression model and the unrestricted regression model?Econometrics代写
c.Consider a version of this regression model that has mur (in ”level” form) as the dependent variable and the same set of explanatory variables from part a. Determine if this model or the model in question part a provides a better fit to the data. Explain clearly how you make this determination. 6points
d.Runthe Breusch-Pagan test of homoskedasticity for the data and the regression model in part a. 9 pointsEconometrics代写
i.Interpretthe resulting Breush-Pagan test Use the 5 percent signifi- cance level.
ii.Describethe regression model that underlies this Breush-Pagan That is, to produced the Breush-Pagan test result, a regression model is run: what is the dependent variable in this regression model and what are the explanatory variables?
iii.Clearly explain the concept of heteroskedasticity.Econometrics代写
e.Referringback to the purpose of this analysis (on page 4), can the results from this regression analysis contribute to this debate about ”shall-issue” laws? (This will be a somewhat subjective response: there are plausible reasons for responding ”yes” and there are also ones for responding ”no”.) If your response is ”yes”, explain clearly how the regression results can contribute to this If your response is ”no”, provide at least one reason why we should be cautious about using the relevant regression model estimates in discussing the effect of ”shall-issue” laws. (For example, you might refer to possible omitted variable bias in the regression model.) 6 points
Question 3: 20 points
Use the dataset in htv.csv to answer the questions below. The data are from the Na- tional Longitudinal Survey of Youth (NLSY); this surveys young adult male workers. The survey was carried out in the early 1990’s. The variables that you will use are as follows:
educ: the individual’s years of education completed by 1991
motheduc: the number of years of education completed by the individual’s mother
f atheduc: the number of years of education completed by the individual’s father
abil:a measure of the individual’s ”ability” that is constructed fromperformance in the ASVAB (Armed Services Vocational Aptitude Battery) standardized tests. The units are ”points scored” on these tests.Econometrics代写
tuit17: state-level college tuition rates when the individual was 17 years old
tuit18: state-level college tuition rates when the individual was 18 years old
a.Estimate the following linear regression modelby
educi = β0+β1motheduci+β2f atheduci+β3abili+β4abil2+β5tuit17i+β6tuit18i+ui (1)
Precisely interpret the statistically significant parameter estimates, using the 5 percent significance level. (When interpreting the abil and abil2 parameter esti- mates, describe the shape of the estimated relationship between abil and educ in as much detail as you can.) 10 pointsEconometrics代写
b.Run an F-test of the joint significance of some parameters from the regression model in part a. Choose these parameters so that this F-test result provides informationthat cannot be determined from the regression statistics that resulted from completing part Explain why you are applying this test to these particular parameters and precisely interpret the resulting F statistic, using the 5 percent significance level. 5 points
c.Preciselyexplain the concept of multicollinearity in the context of multiple regres- sion analysis. 5 points