8. Using regression adjustments to estimate the effects of policy intervention

Suppose you want to examine the effects of a job training program, and estimate the following regression model to do so:

earn98=5+2.8train+0.36earn96+0.37educ0.13age+2.7marriedearn98^=5+2.8train+0.36earn96+0.37educ0.13age+2.7married

where

	earn98earn98 = labor market earnings in 1998, in thousands of dollars
	traintrain =1 if the employee participated in the training program, and =0 otherwise
	earn96earn96 = labor market earnings in 1996 (the year before the training program), in thousands of dollars
	educeduc = years of schooling
	ageage = age, in years
	marriedmarried =1 if the employee is married, and =0 otherwise
	nn = 1,080
	R2R2 = 0.45

Using unrestricted regression adjustment, which of the following best represents the estimated equation with full interactions?

earn98=5+3.1train+0.38earn96+0.4educ0.16age+3.3married+0.11train(earn96earn96)+0.031train(educeduc)+0.055train(ageage)+1.05train(marriedmarried)earn98^=5+3.1train+0.38earn96+0.4educ0.16age+3.3married+0.11trainearn96earn96+0.031traineduceduc+0.055trainageage+1.05trainmarriedmarried

earn98=5+3.1train+0.38earn96+0.4educ0.16age+3.3married+0.11earn96(traintrain)+0.031educ(traintrain)+0.055age(traintrain)+1.05married(traintrain)earn98^=5+3.1train+0.38earn96+0.4educ0.16age+3.3married+0.11earn96traintrain+0.031eductraintrain+0.055agetraintrain+1.05marriedtraintrain

earn98=5+3.1train+0.38earn96+0.4educ0.16age+3.3married+0.11train(earn96earn96)+0.031train(earn96earn96)+0.055train(earn96earn96)+1.05train(earn96earn96)earn98^=5+3.1train+0.38earn96+0.4educ0.16age+3.3married+0.11trainearn96earn96+0.031trainearn96earn96+0.055trainearn96earn96+1.05trainearn96earn96

earn98=5+3.1train+0.38earn96+0.4educ0.16age+3.3married+0.11train(educeduc)+0.031train(educeduc)+0.055train(educeduc)+1.05train(educeduc)earn98^=5+3.1train+0.38earn96+0.4educ0.16age+3.3married+0.11traineduceduc+0.031traineduceduc+0.055traineduceduc+1.05traineduceduc

Using your answer from the preceding question, the estimated average treatment effect of the training program is uraura^ =

.

Which of the following best represents the meaning of the estimated average effect of the training program?

The training program led to an increase in 1998 earnings of $400.

The training program led to an increase in 1998 earnings of $160.

The training program led to an increase in 1998 earnings of $110.

The training program led to an increase in 1998 earnings of $3,100.

8. Using regression adjustments to estimate the effects of policy intervention

Suppose you want to examine the effects of a job training program, and estimate the following regression model to do so:

earn98=5+2.8train+0.36earn96+0.37educ0.13age+2.7marriedearn98^=5+2.8train+0.36earn96+0.37educ0.13age+2.7married

where

	earn98earn98 = labor market earnings in 1998, in thousands of dollars
	traintrain =1 if the employee participated in the training program, and =0 otherwise
	earn96earn96 = labor market earnings in 1996 (the year before the training program), in thousands of dollars
	educeduc = years of schooling
	ageage = age, in years
	marriedmarried =1 if the employee is married, and =0 otherwise
	nn = 1,080
	R2R2 = 0.45

Using unrestricted regression adjustment, which of the following best represents the estimated equation with full interactions?

earn98=5+3.1train+0.38earn96+0.4educ0.16age+3.3married+0.11train(earn96earn96)+0.031train(educeduc)+0.055train(ageage)+1.05train(marriedmarried)earn98^=5+3.1train+0.38earn96+0.4educ0.16age+3.3married+0.11trainearn96earn96+0.031traineduceduc+0.055trainageage+1.05trainmarriedmarried

earn98=5+3.1train+0.38earn96+0.4educ0.16age+3.3married+0.11earn96(traintrain)+0.031educ(traintrain)+0.055age(traintrain)+1.05married(traintrain)earn98^=5+3.1train+0.38earn96+0.4educ0.16age+3.3married+0.11earn96traintrain+0.031eductraintrain+0.055agetraintrain+1.05marriedtraintrain

earn98=5+3.1train+0.38earn96+0.4educ0.16age+3.3married+0.11train(earn96earn96)+0.031train(earn96earn96)+0.055train(earn96earn96)+1.05train(earn96earn96)earn98^=5+3.1train+0.38earn96+0.4educ0.16age+3.3married+0.11trainearn96earn96+0.031trainearn96earn96+0.055trainearn96earn96+1.05trainearn96earn96

earn98=5+3.1train+0.38earn96+0.4educ0.16age+3.3married+0.11train(educeduc)+0.031train(educeduc)+0.055train(educeduc)+1.05train(educeduc)earn98^=5+3.1train+0.38earn96+0.4educ0.16age+3.3married+0.11traineduceduc+0.031traineduceduc+0.055traineduceduc+1.05traineduceduc

Using your answer from the preceding question, the estimated average treatment effect of the training program is uraura^ =

.

Which of the following best represents the meaning of the estimated average effect of the training program?

The training program led to an increase in 1998 earnings of $400.

The training program led to an increase in 1998 earnings of $160.

The training program led to an increase in 1998 earnings of $110.

The training program led to an increase in 1998 earnings of $3,100.