Econ 205

b. The table below summarizes the level of severance pay, at different years of service before and after the reforms. Suppose you are interested in estimating the effect of the reforms en the extent of employment volatility over the business cycle in Aussie You have access to firm-level data that gives total employment for each month; the total number of hires each month; and the total number of workers tired each month by the workers' years of service with the rm. Data are available for each marsh from 1999 to soon The reforms began in 2000, and were phased in gradually through 2004. Outline the empirical approach you would take to estimate the eect of the refonns on employment volatility. Provide as much detail as possible on the specific econometric specication you would use, and how the reform effects would be identied. Tears of service Old Severance New Severance Payment {amount at interval Pavment {constant beginnlng of veers of service intervall" thre ughout veers of service Interval] _w months :3 at5 ears] months [rev at 10 veers} months a at 15 am months pa at zvears months _-a_ at 25 are maximum of 12 months pav at 3? veers _ Q4. Pislaferri opens his Eil3 paper on the intertemporal labor supply effects of wage changes with the following: ' \"A long-standing question in labor economics is whether and to what assent individual labor supply responds to anticipated wage changes {also known as evolutionary wage changes}. This eet is measured by the intertemporal elasticity of substitution. . ." . a. Why is it important that wage changes used to identify the intertentporal labor supply elasticity use only anticipated changes in wages? Discuss the expected sign of the labor supply response to anticipated wage changes. If some of the wage changes are unanticipated, what will be the likely effect on the estimated labor supply response? Wind it. Pistai'erri {ZEUS} has data on individual's subjective expectations oftheir future wage 31'th which he utilizes to decompose observed wages into three components: a. Anticipated, evolutionary movements in the wage over individuals\" lifecyclcs h. Unantioipated permanent shifts in the future prole of wages c. Unanticipated transitory changes in wages Q. Use this question to deinonstrate your loiowledge about empirical methods in labor Economics. For each part, be as precise and comprehensive as possible You are trying to measure the causal impact of college education on labor market outcomes. As an initial specication. you estimate with 0L3, including a dummy 1rariahle for \"College Graduate" asweli as screw} control variables. Your focus is on the coefcieu on College Graduate. You have data on many sets of siblings; where for each person you know their education level, labor market outcomes, who is related to whom, and each individual's covariates. A. Describe the \"standard reasons" 1why We should be concerned about hmsed coefficients from estimating by DLS. E. 1What problems with GL5 will propensity-score matching help to solve? What problems will it not help to sol're'iI A colleague suggests adding \"family xed effects\" toyour specication. In what key 1|stays does this change the nature of the identifying yariati on? What types of biases do family sed eects help to remove? What are some ways that this strategy could still lead to biased results? Are there reasons why this strategy could increase the bias? Describe why or why not. What are the reasons why this strategy might not lead to biased estimates but still give drfferent results than ULS estimates? .mfig 1. [10) Each period. an innitelylived agent divides his endowment of 1 unit of time between human capital production (a nonmarket activity} and work in order to maximize the present discounted value of lifetime earnings. Let h: denote the human capital stock at time t and let 1 It be time spent working. Income each period is given by h; (1 It} as where w is the rental rate of human capital. New human capital [i.e. investment in human capital} is produced via the production function Unt)\" ; o: E (:0, 1} - Note that human capital depreciates at the rate 5 and the real interest rate is given by the constantT r- [a] Express the agent's maximization problem as a dynamic programming problem and identify the states and controls- [b} Derive and interpret the Euler equations associated with this problem- {c} Assume that a steadystate exists so that h; = E and 1 = f. Solve for these steadystate values. [(1) \"'hat is the impact of the two prices ['11:- T} on the steadystate values? Explain- 2. [20) Consider a representative agent economy in which preferences are given by: no 11; Eu (2;? [12:1, sh]) J E {mm} o t=D where E denotes the expectations operator and E; is an iii-d. random variable which affects the disutility of labor supply. Output in the economy is produced by rms that use labor and an inelastically supplied unit of non depreciating capital [owned by rms you can think of this as land]. The choice of labor is made to maximize prots each period: maxm y: sh: '9'! Illa: a E [0, ll where yt denotes output and mg is the wage. The prots are returned to the households. In addition to labor supply, agents also trade one period bonds [riskfree) that cost 3): at time t and return