Economics 123B代做、代写R编程设计、R代写、代做data留学生-代写Algorithm 算法作业

Economics 123B

Econometrics II Winter 2019

Homework 3

Due on 2/28/2019

Problems and Derivations

Problem 1: Suppose that we want to evaluate the effect of several variables on

annual saving and that we have a panel data set on individuals collected in January

1990 and January 1992. If we include a year dummy variable for 1992 and use first

differencing, can we also include age in the original model? Please explain.

Problem 2: Consider a data set consisting of observations on i = 1, . . . , n units

over t = 1, . . . , T time periods and suppose that n is large and T is small (e.g., you

have n = 5000 patients observed over T = 10 years). Write down an econometric

model that can be used to analyze the data and explain its components. What

problems arise if in your regression you ignore the fact that patients are potentially

heterogeneous? Show an example of this graphically. Discuss one approach to

estimating the parameters of this model, listing the necessary steps.

Problem 3: The mock dataset hw3 data.csv on the course website includes observations

on i = 1, . . . , 100 California counties over t = 1, . . . , 4 time periods. The

dependent variable y, stored in column 1, is a measure of pollution. The next three

columns include determinants of y, and a county indicator appears in column 5.

1. Load the data into your favorite statistics software. Without using panel data

packages or toolboxes, implement the mean-differncing and first-differcing fixed

effects estimators. Show your code and report your point estimates. Are the

estimates under mean-differencing and first-differencing comparable? (Note: to

confirm the accuracy of your code, you may wish to compare your results to

those obtained from the panel data analysis tools in your software, e.g., if using

R, install package plm and enable it with the command library(plm) at the top

of your script).

2. Provide a forecast of pollution if x1 = 1.1, x2 = 1.5, and x3 = .2. Because you

have removed the county specific intercepts, what is the proper interpretation

of your forecast?