c. Suppose we estimate impact of the policy change using regression techniques. The difference in differences estimates of the impact of a 1% in capital and labor on log of life expectancy are shown in the table below: Source SS df MS 34 Model Residual .02472673 025898491 2 .012363365 31.000835435 Number of obs F(2, 31) Prob > F R-squared Adj R-squared Root MSE 14.80 0.0000 0.4884 0.4554 .0289 Total ..050625221 33 .001534098 inlifeexp Coef. Std. Err. t P>|t| (95% Conf. Intervall .051859 Inphys 1000 Inhlthpc -.0006698 .0534656 3.913638 0257556 0107318 . 0786268 -0.03 4.98 49.77 0.979 0.000 0.000 -.0531986 .031578 3.753277 .0753532 4.073998 _cons Where Inlifeexp is log life expectancy, Inhlthpc is our measure of log capital, Inphys1000 is log physicians per 1000 people (labor). The model also holds constant the breast cancer survival rate. Use this regression output to answer the two parts of question c below c.i. Use the above estimates to write out a cobb douglas production function with the estimated parameters. [10 points] C ii. Now assume labor is fixed at 6 and we have 10 units of capital: as measured in expenditures on health capital (hospitals, ambulances, MRI machines, etc) per capita. Those units can either be spent on improving life expectancy or exported abroad for 10 million units of consumer good each unit of capital exported. Graph a production function with millions in consumer goods on the X-axis and life expectancy on the Y- axis. [10 points] d. Garber and Skinner use the idea of the production function to make their point. Draw a graph of a production function (and societal indifference curve) to illustrate the difference between productive and allocative inefficiency. Briefly discuss Garber and Skinner: give the best reasons as to why we might think that the US is either allocatively or productively inefficient. [10 points] c. Suppose we estimate impact of the policy change using regression techniques. The difference in differences estimates of the impact of a 1% in capital and labor on log of life expectancy are shown in the table below: Source SS df MS 34 Model Residual .02472673 025898491 2 .012363365 31.000835435 Number of obs F(2, 31) Prob > F R-squared Adj R-squared Root MSE 14.80 0.0000 0.4884 0.4554 .0289 Total ..050625221 33 .001534098 inlifeexp Coef. Std. Err. t P>|t| (95% Conf. Intervall .051859 Inphys 1000 Inhlthpc -.0006698 .0534656 3.913638 0257556 0107318 . 0786268 -0.03 4.98 49.77 0.979 0.000 0.000 -.0531986 .031578 3.753277 .0753532 4.073998 _cons Where Inlifeexp is log life expectancy, Inhlthpc is our measure of log capital, Inphys1000 is log physicians per 1000 people (labor). The model also holds constant the breast cancer survival rate. Use this regression output to answer the two parts of question c below c.i. Use the above estimates to write out a cobb douglas production function with the estimated parameters. [10 points] C ii. Now assume labor is fixed at 6 and we have 10 units of capital: as measured in expenditures on health capital (hospitals, ambulances, MRI machines, etc) per capita. Those units can either be spent on improving life expectancy or exported abroad for 10 million units of consumer good each unit of capital exported. Graph a production function with millions in consumer goods on the X-axis and life expectancy on the Y- axis. [10 points] d. Garber and Skinner use the idea of the production function to make their point. Draw a graph of a production function (and societal indifference curve) to illustrate the difference between productive and allocative inefficiency. Briefly discuss Garber and Skinner: give the best reasons as to why we might think that the US is either allocatively or productively inefficient. [10 points]
