Is thre a Risk-Return Trade-off accross Occupations? Evidence from Spain

Is he e a Risk-Re u n T ade-o ac oss Occupa ions? E idence om Spain Luis Diaz-Se ano♣ Na ional Uni e si y o I eland, Maynoo h IZA, Bonn CREB, Uni e si a de Ba celona Joop Ha og♦ FEE-SCHOLAR, Uni e si ei an Ams e dam Abs ac : We use da a om Spain o es o an e ec o ea nings isk and skewness on indi idual wages. We ca y ou sepa a e es ima ion o men, women, public and p i a e sec o employees. In acco dance wi h p e ious e idence o he US we show he exis ence o a isk- e u n ade-o ac oss occupa ions in he Spanish labou ma ke . These esul s a e in con o mi y wi h p e e ences o isk-a e se indi iduals wi h dec easing absolu e isk a e sion. JEL code: J3, D8 Keywo ds: Risk-a e sion, skewness a ec ion, occupa ional choices, compensa ing wage di e en ials. ♣ Depa men o Economics, Na ional Uni e si y o I eland Maynoo h, Co. Kilda e, I eland. E-mail: [email p o ec ed]. ♦ Depa men o Economics. Uni e si ei an Ams e dam. Roe e ss aa , 11. 1018WB Ams e dam. (The Ne he lands). E-mail: J.Ha [email p o ec ed]. 1 1. In oduc ion In a iskless wo ld, choosing an occupa ion should be, undoub edly, an easy ask. In such a case, indi iduals can make hei choice jus maximizing he ea nings ac oss occupa ions. Howe e , in a isky wo ld like ou s, whe e he ac ha wo ke s a e a e se o luc ua ions in hei incomes is uni e sally ecognised, such a decision becomes much mo e complica ed. Such a luc ua ions in ea nings is wha we call ea nings isk. Economic heo y sugges s ha i hese isks a e o eseeable, hey should be compensa ed o . In he empi ical li e a u e on compensa ing wage di e en ials he e exis a wide a ie y o s udies analysing he e ec o di e en sou ces o isk in he job place on wages. Howe e , his li e a u e is mainly ocused on inju y o a ali y isks. As Adam Smi h we claim ha in a compe i i e labou ma ke he e mus exis a way o compensa ion in hose occupa ions ha en ail a highe p obabili y o ailu e (highe a iance in ea nings) in o de o a ac su icien supply. While unce ain y in labou income is accoun ed o in mos heo e ical models i has been a ely es ed apa om some excep ions in he US. Addi ionally, he e is also a g owing li e a u e ha shows he ele ance o he skewness o he e u ns in many economic decisions. Fo example, Ga e and Sobel (1999) and Golec and Tama king (1998) ind e idence ha isk-a e se indi iduals playing lo e y games and be ing in ho se aces in he US base hei pa icipa ion decision on he skewness o he p ize dis ibu ions, espec i ely. P ackash e al. (2003) ound empi ical e idence om La in Ame ican, US and Eu opean capi al ma ke s ha in es o s do ade expec ed e u n o he po olio o skewness. Diaz- Se ano (2004) empi ically suppo s ha posi i e skewness a ou homeowne ship in Ge many and Spain. And Ha og and Vij e be g (2002) in he US and Diaz-Se ano, 2 Ha og and Nielsen (2004) in Denma k shows ha indi iduals app ecia e posi i ely skewed income dis ibu ions, and hey inco po a e his in o ma ion in o hei educa ional choices. In his pape , we empi ically es o he Spanish labo ma ke whe he isk-a e se wo ke s a e compensa ed o ea nings isks in hei occupa ional choices, and whe he he e is a willingness o pay o posi i e skewness in hei incomes. The emainde o he pape is s uc u ed as ollows. In sec ion 2 he exis ing li e a u e in his issue is ex ensi ely e iewed. Sec ion 3 p esen s he heo e ical backg ound. In sec ion 4 we desc ibe he empi ical amewo k. In sec ion 5 we desc ibe he da ase . We ca y ou he empi ical analysis in sec ion 6. And sec ion 7 summa izes and concludes. 2. Li e a u e e iew In labou economics he e is a g owing li e a u e dealing wi h ea nings isk and i s e ec s on indi iduals’ beha iou in he labou ma ke . This li e a u e conside s h ee di e en app oaches, he e ec o isk on he human capi al in es men , on he occupa ional choices, and on he compensa ing wage di e en ials. 2.1. Ea nings isk and schooling choices Le ha i and Weiss (1974) buil a wo pe iod model wi h he choice be ween wo king o s udying du ing he i s pe iod, and s ochas ic ea nings du ing he second pe iod o hose who a ended educa ion in he i s pe iod. In hei heo e ical model hey ob ain ha inc easing ea nings isk educes he in es men in educa ion. Kodde (1986) buil on Le ha i and Weiss model’: in his heo e ical model he inds an ambiguous e ec , bu his empi ical es ima ion sugges s inc easing demand o 3 educa ion wi h inc easing isk. Snow and Wa en (1990) ound ha human capi al dec eases wi h inc easing isk in i s u u e e u ns i such in es men is an in e io ac i i y and indi iduals exhibi dec easing isk a e sion. Williams (1979) applied dynamic p og amming o he educa ion decisions, whe e p oduc ion and dep ecia ion o human capi al and u u e wages a e all s ochas ic. He concludes, unde condi ions, ha highe isk in he p oduc ion o human capi al educes in es men in schooling. Belzil and Hansen (2002) and Hogan and Walke (2002) also used he dynamic p og amming amewo k, bo h models ind ha indi iduals p e e o s ay a school longe wi h inc easing isk in u u e ea nings. The i s a gue ha i is because while being a school hey ecei e non isky pa en al income, he second a ibu e his esul o he inc eased alue o wai ing o a good ea nings d aw. The wo k o Ha og and Diaz-Se ano (2002) cons i u es one o he ew empi ical s udies on his issue. They de elop a human capi al model o op imum schooling leng h wi h s ochas ic ea nings and highligh he pi o al ole o isk a i udes and he schooling g adien o ea nings isk. Thei empi ical es ima ion uses Spanish da a on high school g adua es deciding on a ending uni e si y educa ion. They ind ha he basic esponse o inc easing ea nings isk is nega i e bu ha in households wi h lowe isk a e sion, he esponse will be dampened subs an ially and may e en be e e sed o posi i e. 2.2. Occupa ional choices O azem and Ma ila (1986) de eloped an empi ical model o occupa ional choice unde unce ain y. They conside ed he i s wo momen s o he ea nings dis ibu ion (mean and a iance) wi hin a ious al e na i e occupa ions. Thei 4 empi ical esul s con i m ha inc easing mean and dec easing a iance o he ea nings dis ibu ion inc ease he p obabili y o choosing a gi en occupa ion. Siow (1984) es ima ed supply cu es assuming unce ain y in he u u e wages and enu e o he occupa ional choice by a coho o en an s in o an occupa ion. This au ho used a sample o Ame ican lawye s. De Meza (1984) also analysed occupa ional choices unde wage unce ain y, his au ho inds ha inc easing isk dec eases indi idual’s u ili y. 2.3. Risk compensa ion in wages This li e a u e s a s wi h King (1974) using agg ega ed da a eg essed he a iance and he skewness by occupa ion cells on a e age ea nings, obse ing a posi i e e ec o a iance and nega i e o skewness in he US labou ma ke . McGold ick (1995) buil on King’s wo k and using US mic oda a ob ained he same esul s. McGold ick and Robs (1996) s udied he e ec o wo ke s mobili y on compensa ing di e en ials o ea nings isk. They obse e ha wo ke s wi h high mobili y ecei e smalle compensa ions o ea nings isk, since hey show p e e ence o unce ain si ua ions. Also in he con ex o he US labou ma ke , Feinbe g (1981a) es ima ed signi ican wage compensa ion o ea nings isk using a 6-yea s panel. He es ima ed ea nings- isk as he coe icien o a ia ion o indi idual yea ly ea nings du ing he sample pe iod. And Feinbe g (1981b) also inds ha wo ke s expe iencing employmen ins abili y in hei occupa ions a e compensa ed o . Ha og and Vij e be g (2002) we e he i s o p o ide a o mal modelling on isk compensa ion in wages. Using US da a hey es ima ed s uc u al equa ions measu ing isk and skewness by occupa ion-educa ion cells, and hey obse e a posi i e compensa ion o isk and a penal y o skewness in wages. Diaz-Se ano, 5 Ha og and Nielsen (2003) also es ima ed posi i e compensa ion o isk using Danish panel da a. They es ima ed isk using jus educa ion cells and p o ide new es ima es by expe imen ing wi h se e al dynamic measu es o ea nings isk and skewness. They also ound a signi ican posi i e e ec o he a iance and nega i e o he skewness on wages. 3. Theo e ical backg ound As we men ion in he p e ious sec ion, he e is a wide numbe o s udies ha conside s an e ec o ea nings unce ain y in he educa ional an occupa ional choices. Howe e , a li le has been said ye abou skewness. Tsiang (1974) ound heo e ical suppo ha isk-a e se indi iduals display p e e ence o skewness, in addi ion o a e sion o dispe sion ( isk), o he p obabili y dis ibu ion o he e u ns in economic decisions ha en ail an unce ain ou come. This esul sugges s ha since inc easing absolu e isk a e sion is absu d, dec easing absolu e isk a e sion equi es ha indi iduals app ecia e highe momen s as e.g. skewness. O he well es ablished heo ies also emphasize a ound his inding. Fo example, p ospec heo y (Kahneman and T e sky, 1979, 1991) s a es ha he indi idual’s disu ili y caused by a loss is g ea e han he u ili y caused by a gain o he same size, which goes in he same di ec ion as Tsiang’s indings. These a gumen s sugges ha i we assume u u e ea nings a ached o an occupa ional/educa ional choice o be unce ain, bo h he a iance and skewness o ea nings, in addi ion o he mean, should be conside ed when analyzing how hese choices a e planned and achie ed. To unde s and how such a compensa ion mechanism in wages may a ise we ollow Ha og and Vij enbe g (2002) and Diaz-Se ano e al. (2004). Assume ha a isk- a e se indi idual has o choose be ween wo occupa ional op ions ha only di e in isk. 6 In he iskless al e na i e, annual ea nings a e gi en as Y , gene a ing u ili y U(Y ), whe e U( ) is a conca e u ili y unc ion wi h U’ > 0, U” < 0 and U’” > 0 ( he la e condi ion is necessa y o declining absolu e isk a e sion, see Tsiang, 1974 o Ha og and Vij e be g, 2002). In he isky op ion, income is a single d aw o he es o wo king li e, w i en as Y + ε . Equal expec ed li e ime u ili y equi es 00 () ( ) TT UY e d E UY e d ρρ ε −− =+ ∫∫ (1) whe e T is he leng h o wo king li e and ρ he ime discoun a e. We can w i e he le - hand side as () 0 1 () 1 () T T UY e d e UY ρρ ρ −− =− ∫ (2) Fo he s ochas ic e m on he igh -hand side we apply a hi d-o de Taylo expansion a ound he expec ed alue Y, one o de up om P a ’s o iginal con ibu ion (P a , 1964), o () 23 0 111 ( ) 1 ( ) ''( ) '''( ) 26 T T p p UY e d e UY U Y U Y ρρ ε σκ ρ −− ⎡ ⎤ +=− + + ⎢ ⎥ ⎣ ⎦ ∫ (3) whe e 2 p σ is he second momen ( isk) and 2 p κ is he hi d momen (skewness) o ε a ound he expec ed alue ze o. Equa ing (2) and (3) and ew i ing a li le, a e applying a i s -o de Taylo expansion a ound Y o (2), we ge 23 23 23 23 1 '' 1 ''' '' 1 1 2'6'''2 6 p p p p s YY UUU YYYVVV YUUU YY YY σκ σκ −=− − = − (4) whe e V is A ow-P a ’s ela i e isk a e sion and Vs is he simila de ini ion o ela i e skewness a ec ion (we call i a ec ion, because indi iduals like skewness; see Ha og and Vij e be g, 2002). Wi h V and Vs posi i e by de ini ion, we no e om (4) ha 7 indi iduals only en e an occupa ion i he pe manen e ec om an unknown occupa ional ou come is ma ched by a posi i e p emium o he isk ( a iance), while hey allow an ea nings d op o skewness. 3. Empi ical amewo k Assume ha ea nings o an indi idual can be exp essed as ollows isj s ij YY η = , (5) whe e Yisj a e he obse ed ea nings o indi idual i wi h educa ion S in occupa ion j, Ys= µ sY0 a e he expec ed ea nings o an indi idual wi h educa ion le el S, wi h Y0=exp(X β ), and η ij=exp(uij) is a e m picking up speci ic e ec s on ea nings o indi iduals’ occupa ion. This speci ica ion allows indi iduals wi h he same schooling o ea n di e en wages. Appling loga i hms in bo h sides o (5) we ge log log ij s i ij YXu µ β = ++ , (6) whe e X a e he obse able de e minan s o Yij, log µ s is he ixed e ec o educa ion on wages, and uij is he e ec on wages o occupa ion j o indi idual i. Exp ession (6) p o ides he amilia Mince ma k-up model. I wages include a compensa ion o isk in labou ea nings, uij can be decomposed as ollows 23 ij j j j i uu λασ γκ =+ + +, (7) whe e λ j a e he occupa ion ixed-e ec s, 2 j σ is he a iance o ea nings wi hin occupa ions, 3 j κ is he skewness, and ui is a andom e m no mally dis ibu ed wi h 0 mean and cons an a iance ac oss indi iduals. By subs i u ing (7) in (6), he comple e speci ica ion o indi iduals’ ea nings can be exp essed as 8 23 log log isjijji YX u µλ βασγκ =+++++ , (8) The exis ence o a isk compensa ion equi es α >0, whe eas he so called “skewness a ec ion” ha en ails a penal y in wages equi es γ <0. We ake as sui able measu es o 2 j σ (he ea e Rj) and 3 j κ (he ea e Kj) he a iance and he skewness o he obse ed ea nings by occupa ions cells, espec i ely. We decompose ea nings acco ding o he sou ce o a ia ion, i.e. sys ema ic and unsys ema ic componen . Sys ema ic luc ua ions in ea nings a e caused by supply a iables (e.g. human capi al), which a e usually known by indi iduals, and he e o e, ha e no hing o do wi h isk. Howe e , unsys ema ic a ia ions in ea nings ca ch a ia ions which a e unknown by indi iduals when hey ha e o make hei choice o educa ion and ensuing occupa ion. They e lec indeed he isk o he indi idual: hei as ye unknown abili ies, sui abili y o he job, and hence ela i e posi ion in he occupa ions’ ea nings dis ibu ion. They also e lec demand ac o s (e.g. business cycle o shocks in ou pu demand) and hey a e expec ed o gene a e compensa ing wage di e en ials. Hence, sui able measu es o isk and skewness in wages equi e ha sys ema ic a ia ion in wages shall be pu ged om obse ed ea nings, since o he wise he ue ela ionship be ween isk and wages migh be obscu ed. We use a wo-s ep me hod o es o isk compensa ion in he ea nings equa ion (8). Fi s ly, we es ima e equa ion (6), whe e we assume ha log s i X µ β + collec s he sys ema ic a ia ion in ea nings, and hence he es ima ed esiduals a e used o calcula e R and K by occupa ions as ollows 15 Re e ences Bellan e, D. and A.N. Link (1981): “A e public sec o wo ke s mo e isk a e se han p i a e sec o wo ke s?”, Indus ial Labo Rela ions Re iew 34, pp. 408-412. Belzil, C. and J. Hansen (2002): “Ea nings dispe sion, isk a e sion and educa ion” IZA Discussion Pape #513, IZA, Bonn. Ch is iansen, C. and H.S. Nielsen (2002): “The educa ional asse ma ke : a inance pe spec i e on human capi al in es men ”, ASB Wo king Pape #02-10, ASB, Aa hus. De Meza, D. (1984): “Wage unce ain y, expec ed u ili y and occupa ional choice”, In e na ional Economic Re iew 25, pp. 757-762. Diaz-Se ano, L., J. Ha og and H.S. Nielsen (2003): “Compensa ing wage di e en ials o schooling isk in Denma k”, IZA discussion pape #963, IZA, Bonn. Diaz-Se ano, L. (2004), “Labo income unce ain y, skewnes and homeowne ship: a panel da a s udy o Ge many and Spain”, IZA discussion pape #1008, IZA, Bonn. Feinbe g, R. (1981a): “Ea nings- isk as a compensa ing di e en ial”, Sou he n Economic Jou nal 48, pp. 156-163. Feinbe g, R. (1981b): “Employmen ins abili y, ea nings, and ma ke s uc u e”, Applied Economics 13, pp. 257-265. Ga e T.A. and R.S. Sobel (1999): “Gamble s a o skewness, no isk: u he e idence om he Uni ed S a es’ Lo e y games”, Economics Le e s 63, pp. 85-90. Golec, J. and M. Tama king (1998): “Be o s lo e skewness, no isk, a he ho se ack”, Jou nal o Poli ical Economy 106, pp. 205-225. Ha og, J. and L. Diaz-Se ano (2002): “Ea nings isk and demand o highe educa ion: a c oss-sec ion es o Spain”, IZA Discussion Pape #641, IZA, Bonn. Ha og, J. and W. Vij e be g (2002): “Do wages eally compensa e o isk a e sion and 16 skewness a ec ion”, IZA discussion pape s #426, IZA, Bonn. Ha og, J., E.J.S. Plug, L. Diaz-Se ano and A.J.C. Viei a (2003): “Risk compensa ion in wages: a eplica ion”, Empi ical Economics 28, pp. 639-647. Heckman, J.J. (1976): “The common s uc u e o s a is ical models o unca ion, sample selec ion, and limi ed dependen a iables and a simple es ima o o such models”, Annals o Economic and Social Measu emen 5, pp. 475-492. Hogan, V. and I. Walke (2001): “Educa ion choice unde unce ain y”, Wo king Pape UCD, Dublin. Kahneman D. and A. T e sky (1979): “P ospec heo y: an analysis o decision unde isk”, Econome ica 47, pp. 263-291. Kahneman D. and A. T e sky (1991): “Loss a e sion in iskless choice: a e e ence- dependen model”, Qua e ly Jou nal o Economics 106, pp. 1039-1061. King, A.G. (1974): “Occupa ional choice, isk a e sion, and weal h”, Indus ial and Labo Rela ions Re iew 27, pp. 586-96. Kodde, D.A. (1986): “Unce ain y and he demand o educa ion”, Re iew o Economics and S a is ics 68, pp. 460-67. Koenke , R. and R. Basse (1978): “Reg ession quan iles”, Econome ica 46, pp. 33-50. Lee, L. (1983): “Gene alised econome ic models wi h selec i i y”, Econome ica 51, pp. 507-512. Le ha i, D. and Y. Weiss (1974): “The e ec o isk on he in es men in human capi al”, Ame ican Economic Re iew 64, pp. 950-63. McGold ick, K. (1995): “Do women ecei e compensa ing wages o ea nings unce ain y?”, Sou he n Economic Jou nal 62, pp. 210-222. McGold ick, K. and J. Robs (1996): “The e ec o wo ke mobili y on compensa ing wages o ea nings isk”, Applied Economics 28, pp. 221-232. 17 O azem, P.F. and J.P. Ma ila (1986): “Occupa ional en y and unce ain y: males lea ing high school”, Re iew o Economics and S a is ics 67, pp. 265-273. P akash, A. J., C.H. Chang and T.E. Pac wa (2003): “Selec ing a po olio wi h skewness: ecen e idence om US, Eu opean, and La in Ame ican equi y ma ke s”, Jou nal o Banking and Finance 27, pp. 1375-1390. Schube , R., M. B own, M. Gysle ; H.W. B achinge (1999): “Financial Decision- Making: A e Women Really Mo e Risk-A e se?”, Ame ican Economic Re iew 89, pp. 381-385. Siow, A. (1984): “Occupa ional choice unde unce ain y”, Econome ica 52, pp. 631-645. Snow, A. and R.S. Wa en (1990): “Human capi al in es men and labo supply unde unce ain y”, In e na ional Economic Re iew 31, pp. 195-206. Tsiang, S.C. (1974): “The a ionale o he mean-s anda d de ia ion analysis, skewness p e e ence, and he demand o money”, Ame ican Economic Re iew 64, pp. 354-371. Williams, J. (1979): “Unce ain y and he accumula ion o human capi al o e he li ecycle”, Jou nal o Business 52, pp. 521-48. 18 Table 1: Sample desc ip i e s a is ics Full sample Women Men Public sec o P i a e sec o Mean STD Mean STD Mean STD Mean STD Mean STD Age 37,14 12,09 34,03 11,40 38,54 12,14 39,10 11,48 36,46 12,23 Women 0,31 0,46 0,38 0,49 0,29 0,45 Ma ied 0,36 0,48 0,31 0,46 0,38 0,48 0,41 0,49 0,34 0,47 Household size 4,14 1,54 4,02 1,57 4,19 1,52 3,93 1,51 4,22 1,54 # o child en 0,57 0,50 0,52 0,50 0,59 0,49 0,55 0,50 0,58 0,49 Yea s o schooling 8,39 4,29 9,21 4,40 8,03 4,19 10,78 4,65 7,56 3,83 Log(yea ly wages) 13,87 0,83 13,55 0,89 14,01 0,75 14,19 0,74 13,76 0,83 Dummies indus y Ag icul u e 0,08 0,26 0,05 0,21 0,09 0,28 0,01 0,12 0,10 0,30 Ene gy 0,02 0,14 0,00 0,07 0,03 0,16 0,02 0,14 0,02 0,14 Chemical 0,03 0,17 0,01 0,11 0,04 0,20 0,02 0,13 0,04 0,19 Mechanics 0,08 0,26 0,02 0,14 0,10 0,30 0,02 0,13 0,10 0,29 O he manu ac u es 0,13 0,33 0,13 0,34 0,12 0,33 0,01 0,11 0,16 0,37 Cons uc ion 0,10 0,30 0,01 0,11 0,14 0,35 0,02 0,14 0,13 0,34 Comme ce 0,16 0,37 0,19 0,39 0,15 0,35 0,02 0,12 0,21 0,41 T anspo 0,06 0,23 0,03 0,16 0,07 0,26 0,08 0,28 0,05 0,22 Banking 0,05 0,22 0,05 0,21 0,05 0,22 0,01 0,10 0,06 0,24 O he se ices 0,30 0,46 0,51 0,50 0,21 0,41 0,79 0,40 0,13 0,34 Dummies occupa ion Manage ial 0,01 0,10 0,00 0,05 0,01 0,12 0,01 0,10 0,01 0,10 P o essionals 0,09 0,29 0,13 0,33 0,08 0,27 0,23 0,42 0,04 0,20 Scien is s 0,07 0,26 0,11 0,31 0,05 0,23 0,12 0,33 0,05 0,22 Technicians 0,13 0,33 0,20 0,40 0,09 0,29 0,19 0,39 0,10 0,31 Cle ical 0,21 0,41 0,37 0,48 0,14 0,34 0,16 0,37 0,23 0,42 Quali ied wo ke s 0,06 0,24 0,00 0,07 0,08 0,28 0,04 0,18 0,07 0,25 Cle ical 0,08 0,28 0,01 0,11 0,12 0,32 0,05 0,22 0,10 0,30 Manual wo ke s 0,33 0,47 0,18 0,38 0,39 0,49 0,14 0,35 0,39 0,49 O he 0,02 0,14 0,00 0,07 0,03 0,16 0,06 0,24 0,01 0,08 Public wo ke s 0,26 0,44 0,32 0,46 0,23 0,42 Dummies loca ion U ban a ea 0,58 0,49 0,62 0,49 0,56 0,50 0,70 0,46 0,53 0,50 Region 1 0,40 0,49 0,39 0,49 0,40 0,49 0,40 0,49 0,40 0,49 Region 2 0,37 0,48 0,37 0,48 0,37 0,48 0,40 0,49 0,36 0,48 Region 3 0,23 0,42 0,24 0,43 0,23 0,42 0,20 0,40 0,24 0,43 Sample size 18132 5620 12512 4675 13457 19 Table 2: P obi es ima es o emale pa icipa ion Coe icien Z-s a Cons an P ima y educa ion Seconda y educa ion Highe educa ion Age Age squa e Child en Ma ied Numbe o wage ea ne s U ban a ea -3.6399 (-18.6) 0.1565 (2.5) 0.3684 (8.4) 1.0198 (4.3) 0.1098 (8.7) -0.0016 (-6.8) -0.2909 (-4.7) 0.3725 (7.1) 0.5089 (3.8) 0.1374 (4.2) 0 – Non-pa icipan 1 – Pa icipan 15569 5620 Sample size 21189 No es: Endogenous a iable is emale pa icipa ion. Es ima es include dummies o egion. 20 Table 3: P obi es ima es o public-p i a e sec o choice Coe icien Z-s a In e cep P ima y educa ion Seconda y educa ion Highe educa ion Ac ually s udying Age Age squa e Female Child en Ma ied U ban Region 1 Region 3 Scien is s P o essionals Ope a o s Adminis a i e Whi e-colla s Blue-colla s -4.1242 (-16.1) 0.3698 (4.6) 0.9197 (10.1) 1.6174 (16.0) 0.4602 (6.5) 0.0998 (7.9) -0.0008 (-7.3) 0.2351 (7.4) -0.1505 (-3.4) 0.1630 (3.7) 0.2127 (7.1) 0.0728 (1.6) -0.4938 (-9.7) 0.5552 (6.1) 0.1760 (2.1) 0.1258 (1.6) -0.3535 (-4.8) -0.6774 (-7.4) -0.9382 (-13.0) Public sec o P i a e sec o 4675 13417 Sample size 18092 No e: Endogenous a iable is wo king in he p i a e sec o . 21 Table 4: OLS and wo-s ep es ima es o equa ion (2) used o pu ge sys ema ic ea nings. Full sample Males Females Females1Public2 P i a e2 Cons an Schooling Exp. Exp. squa e Exp. cube Female U ban Region 1 Region 3 Co ec ion e m 12.2383 (397.8) 0.0969 (71.1) 0.0757 (18.2) -0.0014 (-7.2) 6.4·10-6 (2.5) -0.4507 (-41.2) 0.1319 (12.6) -0.0937 (-8.2) 0.1044 (7.9) 12.1515 (334.6) 0.0849 (57.8) 0.0937 (18.9) -0.0018 (-8.1) 8.9·10-6 (3.1) 0.1374 (11.9) -0.0963 (-7.6) 0.1105 (7.5) 11.6686 (212.2) 0.1182 (44.3) 0.0843 (11.3) -0.0025 (-6.7) 2.6·10-5 (4.9) 0.1274 (6.1) -0.0872 (-3.8) 0.1052 (4.1) 11.6906 (103.7) 0.1198 (41.1) 0.0835 (11.1) -0.0025 (-6.7) 2.5·10-5 (4.9) 0.1293 (6.2) -0.0860 (-3.7) 0.1079 (4.2) -0.0403 (-1.3) 13.0405 (110.1) 0.0658 (17.5) 0.0736 (10.4) -0.0017 (-7.4) 1.4·10-5 (3.4) -0.3178 (-17.9) 0.1263 (6.2) -0.0829 (-4.2) 0.1147 (4.6) -0.2856 (-7.1) 12.4335 (328.0) 0.0675 (24.8) 0.0613 (12.3) -0.0011 (-4.8) 2.6·10-6 (0.8) -0.5673 (-42.4) 0.1268 (10.4) -0.1006 (-7.4) 0.2034 (12.2) 0.3992 (9.9) Sample size 17919 12512 5251 5251 4675 13457 No es: (1) Heckman’s selec i i y co ec ion (2) Lee’s selec i i y co ec ion 22 Table 5: Summa y s a is ics o es ima ed R ( isk) and K (skewness). Risk Skewness Mean STD Mean STD To al 0.631 0.867 3.502 3.140 Indus y Ag icul u e 0.804 0.897 5.608 1.421 Ene gy 0.561 0.333 2.681 1.838 Chemical 0.588 0.887 2.255 1.999 Mechanics 0.500 0.935 2.653 1.952 O he manu ac u es 0.660 0.774 3.493 3.331 Cons uc ion 0.801 0.424 8.096 4.493 Comme ce 0.809 1.506 2.976 1.810 T anspo 0.791 0.549 4.100 2.959 Banking 0.492 0.694 2.529 1.811 O he se ices 0.457 0.505 2.143 2.070 Educa ion Incomple e p ima y 0.666 0.572 4.973 3.738 P ima y 0.678 0.835 3.955 3.421 Seconda y 0.602 1.057 2.971 2.247 Uni e si y 0.484 0.816 1.690 1.361 Job le el Manage ial 0.782 0.463 1.380 0.974 P o essionals 0.421 0.094 1.284 0.453 Scien is s 0.600 0.346 1.907 1.750 Technicians 0.419 0.024 3.081 1.230 Cle ical 0.597 0.663 2.601 2.051 Quali ied wo ke s 1.813 2.807 5.840 2.683 Cle ical 0.455 0.117 3.280 1.516 Manual wo ke s 0.649 0.482 5.046 4.215 O he 0.240 0.151 1.153 0.487 Age 18 o 25 0.600 0.541 3.424 3.009 25 o 35 0.624 0.846 3.386 3.045 35 o 45 0.645 1.003 3.461 3.157 45 o 55 0.643 0.898 3.751 3.376 55 o 65 0.665 1.062 3.699 3.213 Sec o P i a e 0.685 0.952 3.853 3.281 Public 0.476 0.525 2.492 2.427 Gende Men 0.676 0.974 3.993 3.418 Women 0.532 0.550 2.409 2.019 23 Table 6: Es ima ion o e u ns o schooling, e u ns o isk and skewness penal y o se e al wo k o ce g oups1. ( -s a is ics in pa en hesis) Full sample Males Females Public2 P i a e2 In e cep Schooling Risk Skewness Co ec ion e m 12.4281 (227.6) 0.0686 (37.3) 12.4140 (227.5) 0.0682 (37.1) 0.0377 (6.3) 12.4250 (226.0) 0.0679 (36.8) 0.0390 (6.5) 0.0199 -0.0027 (-2.7) -0.0096 12.4033 (212.8) 0.0610 (30.3) 0.0401 (6.8) 12.4221 (213.1) 0.0601 (29.7) 0.0424 (7.4) 0.0246 -0.0045 (-2.8) -0.0155 11.5004 (154.3) 0.0784 (19.4) 0.0140 (1.5) 11.8744 (110.9) 0.0783 (19.4) 0.0528 (4.2) 0.0359 -0.0527 (-4.8) -0.1012 13.5375 (67.4) 0.0531 (8.8) 0.0127 (0.4) -0.3270 (-3.8) 13.5005 (61.5) 0.0513 (8.5) 0.2330 (3.2) 0.0862 -0.0890 (-3.4) -0.0898 -0.3403 (-4.0) 12.2944 (116.5) 0.0525 (16.9) 0.0240 (7.3) 0.3288 (4.5) 12.3326 (116.1) 0.0521 (16.8) 0.0278 (7.9) 0.0181 -0.0063 (-3.0) -0.0119 0.3273 (4.5) Sample size 17919 12512 5251 4675 13457 No es: (1) Full model speci ica ion includes a cubic polynomial on yea s o expe ience, and dummy con ols o gende , amily s a us, geog aphical, indus y and occupa ions (2) Lee’s selec i i y co ec ion