PanelDataProblemSet1

PanelDataProblemSet1 - Department of Economics Econometric...

Info iconThis preview shows pages 1–3. Sign up to view the full content.

View Full Document Right Arrow Icon
Econometric Analysis of Panel Data Assignment 1 Part I. Mathematical Statistics The density f( y ) for a nonnegative random variable, y , is exponential with parameter λ , so f( y ) = 1/ λ exp(- y / λ ), y > 0, λ > 0. For this random variable, the mean is E[ y ] = λ . We make this a regression model by formulating the conditional mean (regression) function E[ y | x ] = λ ( x ) = exp( α + β x ). Suppose, further, that x is distributed uniformly with density f( x ) = 1, 0 < x < 1. Note that with this assumption, the joint density of y and x is f( y , x ) = f( y | x ) f( x ) = [1/exp( α + β x )] exp[- y /exp( α + β x )] × 1. 1. Derive the parameters of the linear projection, P ( x ) = δ 0 + δ 1 x , where δ 0 = E[ y ] – δ 1 E[ x ] and δ 1 = Cov[ x , y ]/Var[ x ]. Suppose α = 1/3 and β = 2. What are the values of δ 0 and δ 1 ? Hint: E[ y ] = E x E[ y|x ] = 1 1 0 0 exp( ) 1 exp( ) exp( ) x dx x dx α+β × = α β ∫∫ and Cov[ y , x ] = Cov[ x ,E[ y | x ]] = E x { x × E[ y | x ]} – E[ x ] E[ y ]= 1 0 exp( ) exp( ) x xdx αβ -E[ x ]E[ y ]. Find help at http://en.wikipedia.org/wiki/List_of_integrals_of_exponential_functions 2. Consider the linear Taylor series approximation to the conditional mean function. What are the values of θ 0 and θ 1 in the Taylor series: E*[y|x] = θ 0 + θ 1 x when the expansion point is E[ x ] = 1/2 and as before, α = 1/3 and β = 2. Department of Economics
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Part II. Linear Regression Analysis Data for this exercise are on the course website – please use the “Cornwell and Rupert Returns to Schooling Data.” We begin with the linear regression model (using the variable names in the data set) (*) LWAGE it = β 1 + β 2 OCC it + β 3 SMSA it + β 4 MS it + β 5 FEM i + β 6 ED i + β 7 EXP it + ε it The dependent variable is log wage. The RHS variables are defined in the data set. Although this is a panel data set, we are going to ignore that aspect and “pool” the data. 1. Compute the linear least squares regression results and report the coefficients, standard errors, ‘t-ratios,’ R 2 , adjusted R 2 , residual standard deviation, and F statistic for testing the joint significance of all the variables in the equation. 2. Test the hypothesis that neither education (ED) nor experience (EXP) is a significant determinant of the expected log wage. Use an F (Wald), likelihood ratio (assuming normality of ε ), and a Lagrange multiplier (also assuming normality) test. In each case, document in minute detail exactly how you are computing your results and what conclusion you reach.
Background image of page 2
Image of page 3
This is the end of the preview. Sign up to access the rest of the document.

This note was uploaded on 01/05/2012 for the course B 55.9912 taught by Professor Willamgreene during the Fall '11 term at NYU.

Page1 / 5

PanelDataProblemSet1 - Department of Economics Econometric...

This preview shows document pages 1 - 3. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online