Moretti's (2004) study on productivity spillovers uses a Cobb-Douglas production function to describe the productivity of manufacturing plants. Specifically, he estimates a regression model: In(ypict ) = In(Apict ) + OHIn(Hpjet ) + alln( Lpict) + akin(Kpict) where Upjet is the production of plant p in industry j in city c in year t and Hpict, Lpict, and Kpict are hours of skilled labor, hours of unskilled labor, and capital stock, respectively. Apjet is a measure of a plant's productivity and he models it as In ( Apjet ) = YS-jet + Epjet, where S_jet is the percent of workers in city c outside of industry j with college degree and Epjet is a residual. Short Answer 2a Explain what the residual Epjet is in this setup. Short Answer 2b What type of productivity spillover does this approach NOT capture? Please be specific with your response by giving an example. Short Answer 2c How would you use estimates from this regression to learn about productivity spillovers? Specifically what does it mean if y > 0, y