OPTIMAL FISCAL POLICY WITH LABOR SELECTION FEBRUARY 20, 2019

Transcription

1 OPTIMAL FISCAL POLICY WITH LABOR SELECTION FEBRUARY 20, 2019

2 Inroducion Moivaion LABOR MARKETS STRUCTURE AND POLICY Labor marke srucure and analysis changing over he years Secular changes Cyclical changes Labor marke condiions a prime concern for policy Search and maching in labor markes Technological primiives Coss of posing job openings Maching funcion Bu many oher componens of hiring coss February 20,

3 Inroducion Moivaion LABOR MARKETS STRUCTURE AND POLICY Labor marke srucure and analysis changing over he years Secular changes Cyclical changes Labor marke condiions a prime concern for policy Search and maching in labor markes Technological primiives Coss of posing job openings Maching funcion Bu many oher componens of hiring coss Screening/selecion in labor markes Technological primiives Coss of inegraing new workers ino producion process Disribuion of idiosyncraic mach-qualiy shocks for new workers (Davis, Faberman, & Haliwanger (2013 QJE): 60% of hiring coss are vacancy coss) February 20,

4 Inroducion Quesions LABOR MARKETS STRUCTURE AND POLICY Benevolen policy desires efficien labor markes Characerizaion of efficien allocaions ( firs-bes ) Model-consisen efficiency Model-consisen disorions Builds on analysis of GE selecion efficiency in Chugh and Merkl (2016 IER) GE maching efficiency in Arseneau and Chugh (2012 JPE) February 20,

5 Inroducion Quesions LABOR MARKETS STRUCTURE AND POLICY Benevolen policy desires efficien labor markes Characerizaion of efficien allocaions ( firs-bes ) Model-consisen efficiency Model-consisen disorions Builds on analysis of GE selecion efficiency in Chugh and Merkl (2016 IER) GE maching efficiency in Arseneau and Chugh (2012 JPE) Characerizaion of opimal policy Ramsey (1927) ( second-bes ) Lucas and Sokey (1984 JME), CCK (1991 JMCB, 1999 Handbook of Macro), Werning (2007 QJE), many ohers Recen summary by Sigliz (2014 NBER WP) Pigovian correcive axaion Ramsey (1927) was response o quesion posed by Pigou w/o lump-sum T Pigou (1928, A Sudy in Public Finance) incorporaed Ramsey resuls w/ T February 20,

6 Inroducion Mehodology LABOR MARKETS STRUCTURE AND POLICY Calibrae exogenous policy economy o hi empirical volailiy of ue, lfp, and η(ε) Requires disorionary srucural parameers February 20,

7 Inroducion Mehodology LABOR MARKETS STRUCTURE AND POLICY Calibrae exogenous policy economy o hi empirical volailiy of ue, lfp, and η(ε) Requires disorionary srucural parameers Conduc opimal policy given srucural parameers February 20,

8 Inroducion Ramsey Resuls LABOR MARKETS STRUCTURE AND POLICY Calibrae exogenous policy economy o hi empirical volailiy of ue, lfp, and η(ε) Requires disorionary srucural parameers Conduc opimal policy given srucural parameers Labor ax rae smoohing no opimal in eiher selecion model Labor Selecion Walrasian Relaive SD (wr GDP) (CCK 1999, Werning 2007 QJE, many ohers) February 20,

9 Inroducion Ramsey Resuls LABOR MARKETS STRUCTURE AND POLICY Calibrae exogenous policy economy o hi empirical volailiy of ue, lfp, and η(ε) Requires disorionary srucural parameers Conduc opimal policy given srucural parameers Labor ax rae smoohing no opimal in eiher selecion model or in maching model Labor Selecion Labor Maching Walrasian Relaive SD (wr GDP) (Arseneau and Chugh 2012 JPE) 0 (CCK 1999, Werning 2007 QJE, many ohers) February 20,

10 Inroducion Ramsey Resuls LABOR MARKETS STRUCTURE AND POLICY Calibrae exogenous policy economy o hi empirical volailiy of ue, lfp, and η(ε) Requires disorionary srucural parameers Conduc opimal policy given srucural parameers Labor ax rae smoohing no opimal in eiher selecion model or in maching model Labor Selecion Labor Maching Walrasian Relaive SD (wr GDP) (Arseneau and Chugh 2012 JPE) 0 (CCK 1999, Werning 2007 QJE, many ohers) Hiring subsidies large in long-run h = 81% Hiring subsidies volaile in business cycles Economic difference vs. Walrasian labor markes? February 20,

11 Inroducion Inerpreaion LABOR MARKETS MODEL-CONSISTENT WEDGES Technological primiives Develop selecion-model consisen ransformaion funcion and MRTs Aggregae goods producion echnology Aggregae maching echnology model-consisen decenralized wedges Tax volailiy EFFICIENT flucuaions Selecion model wedge flucuaions EXACTLY = 0 February 20,

12 Inroducion Inerpreaion LABOR MARKETS MODEL-CONSISTENT WEDGES Technological primiives Develop selecion-model consisen ransformaion funcion and MRTs Aggregae goods producion echnology Aggregae maching echnology model-consisen decenralized wedges Tax volailiy EFFICIENT flucuaions Selecion model wedge flucuaions EXACTLY = 0 Analyically characerize source of exernaliies Cos gap = marginal hiring cos avg. hiring cos Selecion Marke Tighness Play highly similar role as marke ighness exernaliies in maching model Compare and conras wih search and maching model February 20,

13 Inroducion Inerpreaion LABOR MARKETS MATCHING VS. SELECTION Maching and selecion wo disinc componens of recruiing process mach qualiy February 20,

14 Inroducion Inerpreaion LABOR MARKETS MATCHING VS. SELECTION Maching and selecion wo disinc componens of recruiing process mach qualiy Componens of recruiing coss Vacancy posing coss Screening coss Training coss February 20,

15 Inroducion Conribuion CONTRIBUTION TO LABOR & POLICY LITERATURES Need o know appropriae wedges for designing opimal policy Fiscal policy Moneary policy Regulaory policy Recen lieraure on labor selecion Lechhaler, Merkl, and Snower (2010 JECD) Merkl and van Rens (2012) Brown, Merkl, and Snower (2015 MD) Chugh and Merkl (2016 IER) Faia, Lechhaler, and Merkl (2014 JMCB) (opimal moneary policy) Baydur (2017 AEJ: Macro) parial mach qualiy revelaion February 20,

16 Inroducion Conribuion CONTRIBUTION TO LABOR & POLICY LITERATURES Need o know appropriae wedges for designing opimal policy Fiscal policy Moneary policy Regulaory policy Recen lieraure on labor selecion Lechhaler, Merkl, and Snower (2010 JECD) Merkl and van Rens (2012) Brown, Merkl, and Snower (2015 MD) Chugh and Merkl (2016 IER) Faia, Lechhaler, and Merkl (2014 JMCB) (opimal moneary policy) Baydur (2017 AEJ: Macro) parial mach qualiy revelaion Previous work Arseneau, Chahrour, Chugh, Finkelsein Shapiro (2015 JMCB) (cusomer maching) Chugh and Ghironi (2015) (endogenous produc varieies) Arseneau and Chugh (2012 JPE) Aruoba and Chugh (2010 JET) (new monearis) Arseneau and Chugh (2008 JME) (nominal wage rigidiies in labor maching model) February 20,

17 Inroducion Plan OUTLINE Model Srucure of Labor Markes GE efficiency definiions of model-consisen wedges Exended model (labor search and maching + labor selecion) Focus on labor selecion model (his paper) Posiive analysis (non-ramsey policy) Ramsey equilibrium Calibraion Normaive analysis (Ramsey policy) wedge/disorion smoohing Compare and conras wih search and maching model Implicaions for Beveridge Curve Wage deerminaion Model-consisen wedges and marke ighness Conclusion February 20,

18 Plan OUTLINE Model Srucure of Labor Markes GE efficiency definiions of model-consisen wedges Exended model (labor search and maching + labor selecion) Focus on labor selecion model (his paper) Posiive analysis (non-ramsey policy) Ramsey equilibrium Calibraion Normaive analysis (Ramsey policy) wedge/disorion smoohing Compare and conras wih search and maching model Implicaions for Beveridge Curve Wage deerminaion Model-consisen wedges and marke ighness Conclusion February 20,

19 Selecion vs. Maching LABOR MARKET STRUCTURE Maching and selecion wo disinc conceps of fricional labor markes mach qualiy February 20,

20 Definiions LABOR MARKET STRUCTURE Disribuion of idiosyncraic hiring coss ε i beer workers worse workers February 20,

21 Definiions LABOR MARKET STRUCTURE ( ) H ( ) ( ) endogenous selecion hreshold endogenous selecion probabiliy average raining cos for all newly-seleced employees cross-secional SD of raining cos disribuion v h k m( s, v ) vacancies maching probabiliy for acively searching individual vacancy posing cos aggregae maching funcion February 20,

22 Plan OUTLINE Model Srucure of Labor Markes GE efficiency definiions of model-consisen wedges Exended model (labor search and maching + labor selecion) Focus on labor selecion model Posiive analysis (non-ramsey policy) Ramsey equilibrium Calibraion Normaive analysis (Ramsey policy) wedge/disorion smoohing Compare and conras wih search and maching model Implicaions for Beveridge Curve Wage deerminaion Model-consisen wedges and marke ighness Conclusion February 20,

23 Maching & Selecion: Model-Consisen Wedges MATCHING + SELECTION EFFICIENCY Social Planner s.. c cn,, s,, v = 0 ( ) max E0 u ( c) h lfp H ( ) + g + ( ) m( s, v ) + v = zn ( ) lfp h (1 ( ) k ) s + n Goods resource consrain Aggregae LOM for oal employmen February 20,

24 Maching & Selecion: Model-Consisen Wedges MATCHING + SELECTION EFFICIENCY Social Planner s.. c cn,, s,, v = 0 ( ) max E0 u ( c) h lfp H ( ) + g + ( ) m( s, v ) + v = zn ( ) lfp h (1 ( ) k ) s + n Goods resource consrain Aggregae LOM for oal employmen Efficien LFP Efficien vacancy creaion Ineremporal Efficiency rewrie in erms of modelconsisen wedges ( H ) '( ) ) 1 s ( 1, 1) + m s + v u c u( c ) + 1 (1 ( ) ( ) = z February 20,

25 Maching & Selecion: Model-Consisen Wedges MATCHING + SELECTION EFFICIENCY Social Planner s.. c cn,, s,, v = 0 ( ) max E0 u ( c) h lfp H ( ) + g + ( ) m( s, v ) + v = zn ( ) lfp h (1 ( ) k ) s + n Goods resource consrain Aggregae LOM for oal employmen Efficien LFP idenical o Arseneau and Chugh 2012 JPE (p. 949, eqn. 21) h'(lfp ) u'(c ) = g m (s,v ) s m v (s,v ) MRT beween nonparicipaion and oupu Ineremporal Efficiency ( H ) '( ) ) 1 s ( 1, 1) + m s + v u c u( c ) + 1 (1 ( ) ( ) = z IMRT: quaniy of c +1 ha can be produced by reducing c by one uni, all else equal, by accumulaing wealh in form of employmen February 20,

26 Selecion: Model-Consisen Wedges SELECTION EFFICIENCY Social Planner s.. c cn,, s,, v = 0 ( ) max E0 u ( c) h lfp H ( ) + g + ( ) m( s, v ) + v = zn ( ) lfp h (1 ( ) k ) s + n Goods resource consrain Aggregae LOM for oal employmen February 20,

27 Selecion: Model-Consisen Wedges SELECTION EFFICIENCY Social Planner s.. cn,, s, = 0 H ( ) c + g + ( ) s = zn ( ) n = (1 ) n 1 + ( ) s ( ) max E0 u ( c) h lfp lp f (1 ( )) s + n Goods resource consrain Aggregae LOM for oal employmen = 0 zero vacancy posing cos m( s, v ) = s v 1 if = 1 m( s, v ) = s rivial maching funcion Crucial obs. #1 for decenralized efficiency February 20,

28 Selecion: Model-Consisen Wedges SELECTION EFFICIENCY Social Planner s.. cn,, s, = 0 n = (1 ) n 1 + ( ) s ( ) max E0 u ( c) h lfp ( ) c + g + H ( ) s = zn ( ) lp f (1 ( )) s + n Goods resource consrain Aggregae LOM for oal employmen Efficien LFP h '( lfp ) u'( c ) = ( ) H ( ) MRT beween nonparicipaion and oupu Ineremporal Efficiency Derivaion in Appendix D based on ransformaion fronier u' ( c ) ( 1 ) = u( c ) + 1 ( 1 ( ( + 1) H ( + 1) )) z IMRT: quaniy of c +1 ha can be produced by reducing c by one uni, all else equal, by accumulaing wealh in form of employmen February 20,

29 Plan OUTLINE Model Srucure of Labor Markes GE efficiency definiions of model-consisen wedges Exended model (labor search and maching + labor selecion) Focus on labor selecion model (his paper) Posiive analysis (non-ramsey policy) Ramsey equilibrium Calibraion Normaive analysis (Ramsey policy) wedge/disorion smoohing Compare and conras wih search and maching model Implicaions for Beveridge Curve Wage deerminaion Model-consisen wedges and marke ighness Conclusion February 20,

30 Decenralized Economy OVERVIEW Infiniely-lived represenaive household, measure one of members Full consumpion insurance amongs i.e., risk-sharing Employed members Unemployed members Members ouside he labor force ( leisure ) Labor income axaion Governmen-provided hiring subsidies Guaranees compleeness of ax insrumens (Ramsey issue) Governmen-provided unemploymen benefis Individual-specific Nash bargaining for newly-hired worker wih ε i Only an exensive labor margin, no inensive labor margin Long-lasing labor-marke relaionships exogenous separaion rae ρ February 20,

31 Decenralized Economy FIRMS Producion New job i produces y i = z ε i (oupu hiring cos) Hiring coss subsidized a rae τ h (Each incumben job produces y = z ) New hires Avg. hiring coss Avg. wage for new workers Marg. wage for hreshold new worker Wage for incumben worker Tighness n =( ) s NEW H ( ) ( ) ( ) e ( ) w( ) w ( ) H ( ) I February 20,

32 Decenralized Economy FIRMS Producion New job i produces y i = z ε i (oupu hiring cos) Hiring coss subsidized a rae τ h (Each incumben job produces y = z ) New hires Avg. hiring coss Avg. wage for new workers Marg. wage for hreshold new worker Wage for incumben worker Tighness n =( ) s NEW H ( ) ( ) ( ) e ( ) w( ) I w ( ) H ( ) Selecion condiion (aka labor demand) condiion (1 ( + ( h ) ) (1 ) 1 1 h 1) 1 ( 1) I = z w E w + w + 1 Given individualisic wage seing, holds for every new worker (1 ( ( h ) ) (1 ) 1 1 h 1) 1 ( 1) I = i z w i + E w + w + 1 Cos of hiring = expeced payoff of hiring February 20,

33 Decenralized Economy HOUSEHOLD OPTIMIZATION Maximize expeced lifeime uiliy s.. c, n, s, b = 0 ( s ) max E0 u ( c ) h (1 ) + n i disuiliy of paricipaion aggregae flow dividends received lump sum 1 j 1 (1 ) 1 (1 n ) I (1 n e c ) (1 ) (1 pr + b n w s s b ) j + = j R wages for measure of incumben employees wages for measure of newly-hired employees ue benefis for unemployed View χ as fixed insiuional parameer, no a cyclical policy choice February 20,

34 Decenralized Economy HOUSEHOLD OPTIMIZATION Maximize expeced lifeime uiliy s.. c, n, s, b = 0 ( s ) max E0 u ( c ) h (1 ) + n i disuiliy of paricipaion aggregae flow dividends received lump sum 1 j 1 (1 ) 1 (1 n ) I (1 n e c ) (1 ) (1 pr + b n w s s b ) j + = j R wages for measure of incumben employees wages for measure of newly-hired employees ue benefis for unemployed View χ as fixed insiuional parameer, no a cyclical policy choice n = (1 ) n + s Perceived LOM for employmen 1 ( insananeous producion ) (exogenous) measure of pre-exising employmen relaionships erminae flow of new employmen relaionships = measure of searchers s x probabiliy a searcher successfully lands a job FOCs February 20,

35 Decenralized Economy HOUSEHOLD OPTIMIZATION Opimal labor force paricipaion (aka labor supply) condiion Envelope condiion of period +1 hh-level problem h '( lfp u'( c ) n e n I h+ 1 = (1 ) + (1 ) E + 1 (1 + 1) w (1 ) ) u'( c + 1) Equaes MRS cos wih expeced payoff of paricipaion ρ = 1 χ = 1 η = 1 h '( lfp ) u'( c ) n = (1 ) e Ness RBC labor supply funcion February 20,

36 Decenralized Economy WAGES Surplus sharing via individualisic Nash bargaining power Newly-hired employee bargaining power = α E Incumben employee bargaining power = α I February 20,

37 Decenralized Economy WAGES Surplus sharing via individualisic Nash bargaining power Newly-hired employee bargaining power = α E Incumben employee bargaining power = α I w( ) = + PDV n 1 E h w( i ) = + ( 1 )( i ) + PDV n 1 Wage for marginal new worker Wage for inframarginal new worker ε i February 20,

38 Decenralized Economy WAGES Surplus sharing via individualisic Nash bargaining power Newly-hired employee bargaining power = α E Incumben employee bargaining power = α I w( ) = + PDV n 1 E h w( i ) = + ( 1 )( i ) + PDV n 1 Wage for marginal new worker Wage for inframarginal new worker ε i e( ) E h H ( ) (1 ) PDV n ( ) = ( ) Condiional average wage for all new workers w Tighness = + (1 ) + PDV n 1 I I h Endogenous wage dispersion Wihin newly-hired employees Across newly-hired employees and incumben employees Wage for incumben worker February 20,

39 Decenralized Economy WAGES Disribuion of idiosyncraic hiring coss ε i Endogenous selecion hreshold i beer workers worse workers February 20,

40 Decenralized Economy GOVERNMENT AND RESOURCE FRONTIER Governmen Governmen spending Labor income ax Hiring subsidies Provision of ue benefis One-period sae coningen real deb Aggregae goods resource consrain H ( ) c + g + ( ) s = z n ( ) Aggregae LOM for labor n = + (1 ) n 1 ( ) s February 20,

41 Decenralized Economy PRIVATE-SECTOR EQUILIBRIUM Sae-coningen sochasic processes ha saisfy,,,, I j c n s w, w( ), e( ), R = 0 Household s bond Euler equaions LFP condiion Selecive hiring condiion Nash wage oucomes Law of moion for employmen n = (1 ) n + ) s 1 ( Governmen budge consrain (key condiion in Ramsey models) Goods resource consrain Given processes,, n, h g z = 0 H ( ) c + g + ( ) s = z n ( ) February 20,

42 Plan OUTLINE Model Srucure of Labor Markes GE efficiency definiions of model-consisen wedges Exended model (labor search and maching + labor selecion) Focus on labor selecion model Posiive analysis (non-ramsey policy) Ramsey equilibrium Calibraion Normaive analysis (Ramsey policy) wedge/disorion smoohing Compare and conras wih search and maching model Implicaions for Beveridge Curve Wage deerminaion Model-consisen wedges and marke ighness Conclusion February 20,

43 Plan OUTLINE Model Srucure of Labor Markes GE efficiency definiions of model-consisen wedges Exended model (labor search and maching + labor selecion) Focus on labor selecion model Posiive analysis (non-ramsey policy) n Ramsey equilibrium opimize on Calibraion Normaive analysis (Ramsey policy) wedge/disorion smoohing Compare and conras wih search and maching model Implicaions for Beveridge Curve Wage deerminaion Model-consisen wedges and marke ighness Conclusion h ( X ), ( X ) X is privaesecor decision funcions February 20,

44 Ramsey Problem RAMSEY PROBLEM s.. PVIC Ramsey problem Hybrid Formulaion c, n, s, 0 = 0 max E u( c ) h( lfp ) lfp (1-η )s + n c + g + ( H ( ) / ( )) ( ) s = z n d 0 '( ) c '( ) lfp '( ) (1 ) d = 0 = 0 E u c h lfp u c A Lagrange muliplier for each Single Lagrange muliplier μ February 20,

45 Ramsey Problem s.. PVIC RAMSEY PROBLEM Ramsey problem Hybrid Formulaion ma x E0 u( c ) ( ) I n h lfp c, n, s,, w( ), w,, h h '( lfp u'( c = 0 c + g + ( H ( ) / ( )) ( ) s = z n d 0 '( ) c '( ) lfp '( ) (1 ) d = 0 = 0 n = (1 ) n 1 + ( ) s E u c h lfp u c A ) Lagrange muliplier for each Single Lagrange muliplier μ n e n I h+ 1 = (1 ) + (1 ) E + 1 (1 + 1) w (1 ) ) u'( c + 1) lfp (1-η )s + n (1 ( + ( h ) ) (1 ) 1 1 h 1) 1 ( 1) I = z w E w + w + 1 February 20,

46 Ramsey Problem s.. PVIC RAMSEY PROBLEM NOT capured in PVIC Lagrange muliplier for each Ramsey problem Hybrid Formulaion ma x E0 u( c ) ( ) I n h lfp c, n, s,, w( ), w,, h h '( lfp u'( c = 0 c + g + ( H ( ) / ( )) ( ) s = z n d 0 '( ) c '( ) lfp '( ) (1 ) d = 0 = 0 n = (1 ) n 1 + ( ) s E u c h lfp u c A ) Lagrange muliplier for each Single Lagrange muliplier μ n e n I h+ 1 = (1 ) + (1 ) E + 1 (1 + 1) w (1 ) ) u'( c + 1) (1 ( + ( h ) ) (1 ) 1 1 h 1) 1 ( 1) I = z w E w + w + 1 w( ) = + PDV n 1 lfp (1-η )s + n e( ) E h H ( ) (1 ) PDV n ( ) = ( ) w = + (1 ) + PDV n 1 I I h February 20,

47 Plan OUTLINE Model Srucure of Labor Markes GE efficiency definiions of model-consisen wedges Exended model (labor search and maching + labor selecion) Focus on labor selecion model Posiive analysis (non-ramsey policy) n Ramsey equilibrium opimize on Calibraion Normaive analysis (Ramsey policy) wedge/disorion smoohing Compare and conras wih search and maching model Implicaions for Beveridge Curve Wage deerminaion Model-consisen wedges and marke ighness Conclusion h ( X ), ( X ) X is privaesecor decision funcions February 20,

48 Quaniaive Posiive Analysis CALIBRATION Parameerize so non-ramsey economy generaes empirically reasonable labor marke flucuaions in ue, lfp, and hiring rae η(ε) Condiional on exogenous flucuaions in Labor ax rae Produciviy Governmen purchases Crucial parameers Par I Disribuional form logisic (also have used normal disribuion) Mean μ ε = 0.30 Cross-secional sandard deviaion σ ε = 0.19 Value of σ ε disciplined by micro-level evidence SD σ ε across new hires of raining coss = 207 hours ( = 40% of MPN) Barron, Black, and Loewensein (1989 JLE) Firm-level coss of inerviewing/hiring/raining new workers Based on 1982 EOPP (Employmen Opporuniies Pilo Projec) February 20,

49 Quaniaive Posiive Analysis CALIBRATION Parameerize so non-ramsey economy generaes empirically reasonable labor marke flucuaions in ue, lfp, and hiring rae η(ε) Condiional on exogenous flucuaions in Labor ax rae Produciviy Governmen purchases Crucial parameers Par I Disribuional form logisic (also have used normal disribuion) Mean μ ε = 0.30 Cross-secional sandard deviaion σ ε = 0.19 Crucial parameers Par II Unemploymen benefis = 0.70 Worker Nash bargaining power For boh new hires and incumbens E I = = 0.50 Crucial obs. #2 for decenralized efficiency Oher parameers sandard (Table 3) February 20,

50 Quaniaive Posiive Analysis DYNAMIC RESULTS Non- Ramsey Policy (posiive) Daa Labor Tax Rae Mean 20% 20% Rel. SD Tighness ( εη(ε) H(ε) ) Rel. SD 5.58 Hiring rae η(ε) Rel. SD Unemploymen LFP Rel. SD Rel. SD February 20,

51 Plan OUTLINE Model Srucure of Labor Markes GE efficiency definiions of model-consisen wedges Exended model (labor search and maching + labor selecion) Focus on labor selecion model Posiive analysis (non-ramsey policy) n Ramsey equilibrium opimize on Calibraion Normaive analysis (Ramsey policy) wedge/disorion smoohing Compare and conras wih search and maching model Implicaions for Beveridge Curve Wage deerminaion Model-consisen wedges and marke ighness Conclusion h ( X ), ( X ) X is privaesecor decision funcions February 20,

52 Quaniaive Normaive Analysis DYNAMIC RESULTS Ramsey Non- Ramsey Policy (posiive) Daa Labor Tax Rae Mean 22% 20% 20% Rel. SD Tighness ( εη(ε) H(ε) ) Rel. SD Hiring rae η(ε) Rel. SD Unemploymen LFP Rel. SD Rel. SD Ramsey simulaions: shocks o z, g condiional on srucural parameers Observaions 1. Labor income ax rae smoohing NOT opimal 2. Volailiy of η(ε) much smaller han in daa February 20,

53 Quaniaive Normaive Analysis DYNAMIC RESULTS Ramsey Policy Baseline parameers Social Planner Labor Tax Rae Tighness ( εη(ε) H(ε) ) Mean 22% 0% Rel. SD Rel. SD Hiring rae η(ε) Rel. SD Unemploymen LFP Rel. SD Rel. SD Social Planner simulaions: shocks o z Ramsey simulaions: Inefficien srucural parameers DO appear in problem Efficien surplus sharing February 20,

54 Quaniaive Normaive Analysis DYNAMIC RESULTS Ramsey Policy Baseline parameers For ANY (α,χ) pair Social Planner Labor Tax Rae Tighness ( εη(ε) H(ε) ) Mean 22% --- 0% Rel. SD Rel. SD Hiring rae η(ε) Rel. SD Unemploymen LFP Rel. SD Rel. SD Social Planner simulaions: shocks o z Ramsey simulaions: Inefficien srucural parameers DO appear in problem Efficien surplus sharing for ANY (α, χ) pair February 20,

55 Quaniaive Normaive Analysis DYNAMIC RESULTS Impulse response o produciviy February 20,

56 Plan OUTLINE Model Srucure of Labor Markes GE efficiency definiions of model-consisen wedges Exended model (labor search and maching + labor selecion) Focus on labor selecion model Posiive analysis (non-ramsey policy) n Ramsey equilibrium opimize on Calibraion Normaive analysis (Ramsey policy) wedge/disorion smoohing Compare and conras wih search and maching model Implicaions for Beveridge Curve Wage deerminaion Model-consisen wedges and marke ighness Conclusion h ( X ), ( X ) X is privaesecor decision funcions February 20,

57 Maching & Selecion BEVERIDGE CURVE Ouwards shif of Beveridge Curve during Grea Recession Elsby, Michaels, and Raner (2015 JEL) Diamond and Sahin (2014 NY Fed Saff Repor) Reduced-form inerpreaion: maching efficiency Ω has declined m( s, v ) Cheremukhin and Resrepo-Echavarria (2014 EER) Pescaori and Tasci (2011) Chahrour, Chugh, and Poer (2018) Proposal: increase in primiive cross-secional SD σ ε Ges inside black-box Ω (Disribuional explanaion a hear of Leser (2010 JET)) February 20,

58 Maching & Selecion BEVERIDGE CURVE Ouwards shif of Beveridge Curve during Grea Recession February 20,

59 Plan OUTLINE Model Srucure of Labor Markes GE efficiency definiions of model-consisen wedges Exended model (labor search and maching + labor selecion) Focus on labor selecion model Posiive analysis (non-ramsey policy) n Ramsey equilibrium opimize on Calibraion Normaive analysis (Ramsey policy) wedge/disorion smoohing Compare and conras wih search and maching model Implicaions for Beveridge Curve Wage deerminaion Model-consisen wedges and marke ighness Conclusion h ( X ), ( X ) X is privaesecor decision funcions February 20,

60 Wage Deerminaion NASH BARGAINING & EFFICIENCY Suppose zero axes in decenralized economy Selecion model (no search and maching) (Trivial maching funcion) m( s, v ) = s v 1 (wih = 1) E Nash bargaining parameer = = 1 achieves efficien allocaions Efficien surplus sharing = s February 20,

61 Wage Deerminaion NASH BARGAINING & EFFICIENCY Suppose zero axes in decenralized economy Selecion model (no search and maching) (Trivial maching funcion) m( s, v ) = s v 1 (wih = 1) E Nash bargaining parameer = = 1 achieves efficien allocaions Efficien surplus sharing = s Search and maching model (no selecion) Maching funcion (fundamenal) m( s, v ) = s v (0.3, 0.7) 1 E Nash bargaining parameer = 1 achieves efficien allocaions Hosios (1990) condiion Efficien surplus sharing February 20,

62 Wage Deerminaion NASH BARGAINING & EFFICIENCY Suppose zero axes in decenralized economy Maching + Selecion E No value of can decenralize efficien surplus sharing Two compeing efficiency goals bu only ONE wage deerminaion mechanism February 20,

63 Plan OUTLINE Model Srucure of Labor Markes GE efficiency definiions of model-consisen wedges Exended model (labor search and maching + labor selecion) Focus on labor selecion model Posiive analysis (non-ramsey policy) n Ramsey equilibrium opimize on Calibraion Normaive analysis (Ramsey policy) wedge/disorion smoohing Compare and conras wih search and maching model Implicaions for Beveridge Curve Wage deerminaion Model-consisen wedges and marke ighness Conclusion h ( X ), ( X ) X is privaesecor decision funcions February 20,

64 Inroducion Inerpreaion LABOR MARKETS MODEL-CONSISTENT WEDGES Develop selecion-model consisen ransformaion funcion and MRTs Aggregae goods resource consrain Aggregae law of moion of employmen model-consisen decenralized wedges Tax volailiy EFFICIENT flucuaions Selecion model wedge flucuaions EXACTLY = 0 Analyically characerize source of exernaliies Cos gap = marginal hiring cos avg. hiring cos Selecion Marke Tighness Play highly similar role as marke ighness exernaliies in maching model Efficien labor supply in maching model (Arseneau and Chugh 2012 JPE) MRS c, n = 1 MRSc, ( ) ( ) n = H Efficien labor supply from Chugh and Merkl 2016 IER February 20,

65 Maching vs. Selecion: Wedges LABOR MARKETS Efficien labor supply in maching model (Arseneau and Chugh 2012 JPE) h( lfp ) = u ( c ) 1 Wihin-period MRT (1 ) u'( c ) (1 ) = u'( c 1) + z ( 1 ) + 1 Ineremporal MRT February 20,

66 Maching vs. Selecion: Wedges LABOR MARKETS Efficien labor supply in maching model (Arseneau and Chugh 2012 JPE) h( lfp ) = u ( c ) 1 Efficien labor supply in selecion model (Chugh and Merkl 2016 IER) Wihin-period MRT Wihin-period MRT (1 ) u'( c ) (1 ) = u'( c 1) + z ( 1 ) + 1 Ineremporal MRT Ineremporal MRT Subsiue inraemporal condiions ino ineremporal condiions February 20,

67 Maching vs. Selecion: Wedges LABOR MARKETS Marginal cos of hiring a new employee Marginal cos of hiring a new employee u'( c ) u'( c ) + 1 = (1 ) (1 ) (1 ) + 1 h( lfp 1) + u ( c + 1) z February 20,

68 Summary CONCLUSIONS Selecive hiring framework realisic Selecive hiring coss are disinc from vacancy posing coss Davis, Faberman, and Haliwanger (2013 QJE): 40% of hiring coss are NOT vacancy posing coss Smoohing (model-consisen) wedges he goal for opimal policy No smoohing policy insrumens Model-consisen wedges apply o Fiscal policy Moneary policy Regulaory policy February 20,

69 February 20,

70 Ramsey Problem PVIC 1 c + b + k = (1 ) w n (1 )( r ) k + b j n k j j R u '( c ) = R u '( c ) j j j + 1 h'( n ) u'( c ) n = (1 ) w k ( + 1)( ) u'( c ) = E u'( c ) 1 + (1 r ) February 20,