Bart Taub - AgEcon Search

<strong>Bart</strong> <strong>Taub</strong>HWWA DISCUSSION PAPER151Hamburgisches Welt-Wirtschafts-Archiv (HWWA)Hamburg Institute of International Economics2001ISSN 1616-4814

---

The HWWA is a member of:• Wissenschaftsgemeinschaft Gottfried Wilhelm Leibniz (WGL)• Arbeitsgemeinschaft deutscher wirtschaftswissenschaftlicher Forschungsinstitute(ARGE)• Association d’Instituts Européens de Conjoncture Economique (AIECE)

---

<strong>Bart</strong> <strong>Taub</strong>This research is extremely preliminary. This paper subsumes and extends (8). The subjectof this paper is assigned to HWWA’s research program, “Internationalization ofFirms and Labor Markets”. I am grateful for comments from HWWA seminar participants.

---

HWWA DISCUSSION PAPEREdited by the DepartmentEUROPEAN INTEGRATIONHead: Dr. Konrad LammersHamburgisches Welt-Wirtschafts-Archiv (HWWA)Hamburg Institute of International Economics (HWWA)ÖffentlichkeitsarbeitNeuer Jungfernstieg 21 - 20347 HamburgTelefon: 040/428 34 355Telefax: 040/428 34 451E-mail: hwwa@hwwa.deInternet: http://www.hwwa.de/<strong>Bart</strong> <strong>Taub</strong>University of Illinois at Urbana-ChampaignDepartment of Economics1206 S. Sixth StreetChampaign, Illinois 61820E-mail: b-taub@uiuc.edu

---

ContentsPagesAbstractVI1 Introduction 72 <strong>Search</strong> without replacement 83 The firm’s problem 124 The gain from filtering 205 A more formal mechanism 236 Discussion 26Appendix 28References 29FiguresFigure 4.1 20Figure 4.2 21Figure 4.3 22V

---

AbstractA firm monopsonistically hires labor from a pool containing both skilled and unskilledworkers. The marginal value of a worker depends on the match between the job and theworker’s skill level. Unskilled workers can have negative productivity if they are placedin a skilled job. The firm cannot distinguish the two types.The workers are initially dispersed and search for the high wage jobs from the firm. Theworkers’ skill levels are correlated with their patience; equivalently, they obtain indirectbenefits, such as non-firm-specific career capital, from jobs that use their skill appropriately.By judiciously choosing different wages for different types of jobs, the firm canpartially filter the appropriate worker types and match them with the appropriate jobs.This mechanism works because the probability structure of the job offers changes assearchers accept jobs. This entails a delay in hiring workers who search, but the benefitsfrom filtering that are requisite with the delay can exceed the benefits of hiring allworkers immediately without filtering. The wage differentials assumed in standardsearch models are therefore motivated.JEL classification: C73, D83, E24, J64VI

---

¢¡¤£¦¥¨§©§£¥#"%$'&¢()+*,"-/.) 0)1.-2¤.-&¢(3(345+$63.- #"- 0"- 7$'89:;)+$=?&:@A)

---

¡ 2E.).-(HYW$[dYG¦&:".-‰@¤+8.rdW.-&V*+;$#"-+UY+(¤.- ¡$

---

©[£¦§£¥ ¨ € Zq[@WFUYZ [^]¦_badcaefpa_h adY„[¢¡}nka¥[oT¦{p‡©€ e GR:(0+(*1sE+*2¤:84:.- '*uC.-#") t4&¢28*0@! Y:(02(Ev¤8+8+ #*=?&:@’

---

6 @5A@5A,J 10,J I0¡©£> ¨ ¡ Z4^:H¢ B\„&:q ,"M"-&:@@¤8+1.|d’$

---

¢¡ £¥§¤£ k d© oLsL¡`§ §B ¨ |

---

&z& `…". UZ f¢¡¤£¦¥§©¨ 0 ¢!…¡ ? Go7 0 ¢`…¡ ? Gm¢7 0 ¡`?K 0 ¡¦?7 G& 0 ¡`? 0 ¡¦? …0h;dIUW@!&¢8+"- J:LG¦&¢88+&[4L'€^…B€{s:8+2 OG¦2($#.-&¢(p&¢G¨.- Nv"-U ¡a4&:")ST ,"J&¢GŠ.rdI! /45&/4&:";SEQ+(g=?&:@ ¡ ^i2'C! ‘Q.-£¦¥T€5UZ"-H¢+(¤:8"-&I*2$,.J&¢GŠ.-'*F4&:")ST,"545&M4&:";SE+(Xt;SE8+8+#*a=?&:@ ¡UZ"-H¢(¤:8!"-&c*2$#.D&¢G:(%2(;SE8+8'D:+*X.-&Z(7.) a2(;SE8+8+ #*0:(*x;SI+8+8'*M=?&:@¤ ¡ ^ V4^:H¢ 'N"- V.- W-:UY a:‰C¢›€4:H¢¡ “:(*D¢©£g$' '(i.LG¦&:"L.- Nv"-U ¡45+$6p++(#"- J+Dt$'&¢;. G¦&:"D2(Ev¤8+8 '*Fs:$

---

5"- ';)2E"- D4&¢28*gH¢&h.- 5&:.- #"‰4[d ¡P 8)&Cc;2E")¤8+2›=?&:@¤{.-¤.R8

---

=?&:@¤L+G.- VH¢I\_4^:H¢ ^=?&:@¤DH¢ #.D.-&c&Y*#¤8#.-'* ¡ j 2E"- ');+(H.- hs:$

---

'*M=?&:@¤'C.-i2NG:+8+(Hg.-&gv¤81.-#"N&¢2E.Q.- a2(;SI+88+ '*X.rdI!' ¡ … j .6:H¢ e)e $':(7@! a$'&¢(;.)")2$#.- '*7.-&;SI+8+8g #")+&I*uC&[4#sT,".) tv"-Uw*&c 'O(&:.J8&¢) tUt2$67@cdx¤

---

I0ƒ2;¢ 1032 ¢ 0 2¨¢ 0 … 2(’.- g&:.- #"V¤:(*uC¨.- g;SI+88+ '* œ:(*’.-+ '(T.B…Q4&:")ST#"-V:$'$'#.a&¢(81d0.) g+H¢ˆ&:q! #"-'C;&p.-£#sT,"-) h( 'bI2¤:8.rdUW2;.@ h).-+;v¤ #*"-¦;q¢ 0 2›¨¢ 0 Š… 2(p;.6:H¢ e)e)e .-+5-:Ug N+( #bI2¤:8+1.rdUt2?.L;.)+8+8u@! J-.)+;v¤ '*uš+(%;.6:H¢ e£¢ ."- #sT,"-)' ¡e'(F.) O"-&:@@¤8+1.-+ '#C¤.) h 'UY¤8&‹dIUg'(T.^*EdE(¤:Ug$'$

---

¡ ?7 . ¡7 £ #i…¡ ? . 10 ? …# 0 ¡7 0 ¡ !… G ¡ £ ¨© …0 ¢¡ ?7 G ¡ ? …0 }G `… 0 ¢¡ ?7 G ¡ ? …9¢ 0 2^¨¢ 0 … 2`¢ 0 H 2›¨¢ 0 H… 2Wv");.O+( 'bc2¤:8+.rdF+N(''*'*0#sT'(0.Q.-+N;.-:H¢ V@ '$

---

,"-'CT.-'(g8#.).-(HJ.- 4&:")ST #"-{v¤8+8:8+8E.) Š=?&:@¤‰UgUg '*+.-'81dTC.6SE(HO45¤.-#sT,"‰4^:{&:q#")'* ¡4&:")STR"-&:&:").)+&¢(a&¢G¤

---

#W.-(ESF&¢G.-D:Q:(p:*EsT#").-)+(Hg$#&¢;.B….-#(p.-#"- h45+88u@ a:(X&:.-UZ:8uUgŸfF&¢G¤=?&:@¤ ¡s:$

---

Firm60005500value 50004500400010201510 Skilled jobsUnskilled jobs20305Unskilled searchers: 700 Skilled searchers: 300Optimal unskilled jobs: 1750Optimal skilled jobs: 2050Unskilled wage: 1. Skilled wage: 1.34326Marginal Productivity: fij, i is job, j is worker type: fUS: 0.2 fSS: 4. fSU: -0.8 fUU: 2.5UVacancyCost: 0.025 SVacancyCost: 0.027Firmr: 0.9 bU: 0.8 bS: 0.85Optimal firm value: 6031.54Check threshhold for U-searchers: PStar for U-types =0.271682 < 0.460526 necessary for single-period cull of U-types<strong>Search</strong> duration for high types after t=1:PStar: 0.485893 LowProbability: 0.588944 DurationCount: 1813.43 Second, Null

---

Mv"-U 4&¢28+*’8ST W.-&FUg f+UY§¦' V.- Y(c2Ua@ ,"V&¢G=?&:@¤a:(*0.-&F$6&I&¢; Z%"6.-&F&¢G=?&:@n.|dc '#C^'UW28+.- :(K*

---

‰v"-Uk4L8+8T¤[sT >.- ‰*;.)"-1@¤2E.-+&¢(a&¢GIs:8+2‰;.6.-'›&¢G) s"-9@¤8 ' ¡ ^&¢)^'.-'(p#sT&¢81sT h:$#$'&:"-*+(Ht.-&X ©I¡+ …:5(XM?.6:(*¤"-*p$'&¢(T.)"6:$,.-+(HMUg&I*'8 ¡ J) '$'&¢(*FST,ds:8+2h*EdI(¤:Ug+$

---

£ £¦¥ ¢ & Hz&¢ & , £¥ & !…".6K ¡60 ? …¤£ ?7Go¢7 0 ? Go¢7£¡ § K¥¥£§ £¤££?£¥ … …KK ¡¨ ¢¢£¦¥ 0 ¢ … £ £¥ ¢¢… ©¤£'$#.5.-&M.) h)

---

#.-2EX8

---

'UZ:(*“:a$'&¢()2UY#"-¥¡A(.-+:8+8dn$6¤"-H¢ 8+&[4 "-$' 'W+(†&:"-*#"W.-&7)2E@¤;+*§¦#%,fI#")+Ug'(T.6.-&¢(*#"- ¡ ZUg&c* '8{#"- Z+J+(’.-a;¤")..-&0$#2;.-&¢Ug #")'CR+(?.- UW281.-¤8&¢GŠ4^:H¢‹Q2Ug ,"-&¢2 ,fE.- #()+&¢()2H¢H¢ '?./.) 'Ug) '81sT ' ¡ 7(i2Ua@ #"/&¢GR=?&:@¤F:(*3)

---

!¥›£ #(*1fp")&‹sI+*'N:(7&¢2E.-8+( V&¢G.- h")&I&¢GŠ&¢G ,f;.-'($# W:(*x2(bI2 '(#)NG¦&:"Q.- Vv"-Up• ^+N,fE@¤8+ B\_4^:H¢ "-&:@¤8+ 'U ¡ D1.- #"6.)+&¢(Y$#&¢(;.)"6:(i.-‰G¦&:"‰.- 5 'UY¤8+&[dEUY'(i.>G9"6:$#.)+&¢( £ £¦¥ ¢ :(*ZG¦&:"¡¤‰")&:@@¤+8+1.rdM"- R?.)"6:+H¢T.-G¦&:")4"-* ¡ R")&:@@¤+8+1.rdh8+ '&¢(V.) R"-&:@@¤+8.rdh;+UY¤8 ,fh45+$6.-'*X:(*F@!&¢2(*'*uš¤.- h 'UM¤8+&[dEUg'(T.-"- N@&¢2(* '*p@id.- h(.-+:8u(c2Ua@! #"D&¢G›)

---

z #")H¢'UZ:((uC Š 1")SY:(* ¢22)& ¢ :8+UZSI …,C›Ž fI #")+Ug'(T.6.-&¢(Y(gUZ")ST#.)'C–O4&:")SE(H#"