Manasse, Paolo; S anca, Luca
Wo king Pape
Wo king on he T ain? The ole o echnical P og ess and
he T ade in Explaining Wage Di e en ials in I alian Fi ms
Quade ni - Wo king Pape DSE, No. 482
P o ided in Coope a ion wi h:
Uni e si y o Bologna, Depa men o Economics
Sugges ed Ci a ion: Manasse, Paolo; S anca, Luca (2003) : Wo king on he T ain? The ole o echnical
P og ess and he T ade in Explaining Wage Di e en ials in I alian Fi ms, Quade ni - Wo king Pape
DSE, No. 482, Alma Ma e S udio um - Uni e si à di Bologna, Dipa imen o di Scienze Economiche
(DSE), Bologna,
h ps://doi.o g/10.6092/unibo/amsac a/4812
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/159323
S anda d-Nu zungsbedingungen:
Die Dokumen e au EconS o dü en zu eigenen wissenscha lichen
Zwecken und zum P i a geb auch gespeiche und kopie we den.
Sie dü en die Dokumen e nich ü ö en liche ode komme zielle
Zwecke e iel äl igen, ö en lich auss ellen, ö en lich zugänglich
machen, e eiben ode ande wei ig nu zen.
So e n die Ve asse die Dokumen e un e Open-Con en -Lizenzen
(insbesonde e CC-Lizenzen) zu Ve ügung ges ell haben soll en,
gel en abweichend on diesen Nu zungsbedingungen die in de do
genann en Lizenz gewäh en Nu zungs ech e.
Te ms o use:
Documen s in EconS o may be sa ed and copied o you pe sonal
and schola ly pu poses.
You a e no o copy documen s o public o comme cial pu poses, o
exhibi he documen s publicly, o make hem publicly a ailable on he
in e ne , o o dis ibu e o o he wise use he documen s in public.
I he documen s ha e been made a ailable unde an Open Con en
Licence (especially C ea i e Commons Licences), you may exe cise
u he usage igh s as speci ied in he indica ed licence.
h ps://c ea i ecommons.o g/licenses/by/3.0/
Wo king on he T ain? The Role o Technical
P og ess and T ade in Explaining Wage
Di e en ials in I alian Fi ms∗
Paolo Manasse†
, Luca S anca‡
June 2003
Abs ac
This pape p esen s i m-le el e idence on he dynamics o he el-
a i e demand o non-manual wo ke s in I alian manu ac u ing du -
ing he 1990s. The analysis p o ides a numbe o in e es ing esul s.
Fi s , he ise wi hin i ms in he sha e o non manual wo ke s in
bo h employmen and hou s wo ked (wi hin- i m skill upg ading) is
he main de e minan o he inc ease in he ela i e demand o skilled
wo ke s. By con as , demand changes associa ed o ade ha e mi -
iga ed such a ise by shi ing employmen away om skill-in ensi e
i ms. Second, while he ela i e numbe o hou s wo ked by skilled
wo ke s wi hin i ms has isen, he hou ly wage p emium has allen.
Thi d, wi hin- i m skill upg ading is s ongly and signi ican ly ela ed
o in es men in compu e s and R&D. Fou h, we ind ha echnical
p og ess has aised he ela i e p oduc i i y o skilled wo ke s ( he
skill-bias o echnical p og ess is posi i e). Finally we show ha he
s anda d app oach ha measu es annual, a he han hou ly ela i e
∗We hank ISTAT o kindly p o iding he da a o his s udy. We a e indeb ed o
Sil ia P ina o excellen esea ch assis ance. Gio gio Base i and Paolo Epi ani p o ided
use ul commen s.
†Co esponding au ho . Depa men o Economics, Uni e si y o Bologna, S ada
Maggio e 45, 40100, Bologna, I aly. Telephone: #39 51 209 2613. E-mail: man-
[email protected]
‡Uni e si y o Milan-Bicocca.Pza A eneo Nuo o 1, 20100 Milano, I aly. Telephone:
#39 2 6449 8877. E-mail: [email p o ec ed]
1
wages, p oduces a downwa d bias in he es ima e o he skill-bias o
echnical p og ess.
JEL Classi ica ion: F1, F16, J31, O3
Keywo ds: wage di e en ials, skill bias, echnical p og ess, global-
iza ion.
1 In oduc ion
Once upon a ime, be o e he e a o po able compu e s and cellula phones,
commu e s on he Milan-Rome ain ou e b oadly ell in o wo ca ego ies:
i s class a elle s, mainly business people and academics, usually spending
hei ime eading he inancial and gene al p ess, o aking naps ( he la e );
economy class a elle s, mainly amilies, young people and ou is s, o en
in ol ed in anima ed con e sa ions wi h ellow a elle s, ypically abou soc-
ce o poli ics. Nowadays, i s -class a elle s can be seen silen ly hunched
o e hei lap ops, o hea d noisily alking business o e hei cellula phones.
Mos second-class a elle s s ill cha hei way o hei des ina ion, al hough
now o e cellula phones, and some wa ch DVD’s on hei lap- ops. Aca-
demics, now a elling in economy class, ei he ead newspape s o wo k on
hei lap ops (o ake naps).1
This anecdo al e idence sugges s h ee wo king hypo heses: 1. echnical
p og ess in I aly, as in many o he coun ies, has been skill-biased, ha is,
i has aised he ela i e p oduc i i y o mo e educa ed wo ke s ( i s -class
a elle s p esumably make a mo e p oduc i e use o pe sonal compu e s) as
well as he ela i e numbe o hou s wo ked by skilled wo ke s ( i s class
a elle s now wo k ins ead o elaxing); 2. ela i e wages in I aly ha e
no ( ully) adjus ed o he change in ela i e p oduc i i y and hou s (as a
consequence, academics can lo longe a o d o a el – and ake naps – in i s
class); 3. possibly as a esul , i ms ha e conside ably aised he p opo ion
o non-manual wo ke s in employmen .
This pape explo es hese conjec u es by in es iga ing he dynamics o
manual and non-manual employmen and wages in I alian manu ac u ing
du ing he 1990s. We p esen i m-le el e idence on he sou ces and de e -
minan s o he inc ease in he demand o non-manual wo ke s, based on a
new da a se , p e iously una ailable o esea ch, ha co e s a la ge panel o
1We a e g a e ul o Gio gio Base i o his example.
2
manu ac u ing i ms be ween 1989 and 1995. The analysis p o ides a numbe
o esul s suppo ing hese conjec u es.
Fi s , I alian i ms ha e subs i u ed unskilled o skilled wo ke s a a a e
compa able o hose expe ienced in o he indus ialized coun ies, wi h high-
ech i ms playing a leading ole in his p ocess (wi hin- i m skill upg ading
is he main de e minan o he shi in ela i e labo demand in he nine ies).
Second, he ela i e s abili y o annual wage di e en ials wi hin i ms hides
an impo an composi ion e ec : a i m le el he ela i e numbe o hou s
wo ked by skilled wo ke s has isen, whe eas ela i e hou ly wages ha e allen.
Thus subs i u ion owa d skills has occu ed no only in e ms o employmen
bu also in e ms o hou s wo ked. By con as , demand changes associa ed o
ade ha e mo ed employmen away om skill-in ensi e i ms, con ibu ing
o mode a e he change in ela i e ac o p ices: be ween- i m employmen
shi s ha e educed he ela i e demand o skills. Thi d, wi hin- i m skill
upg ading, measu ed by changes in bo h ela i e employmen and numbe o
hou s, is s ongly and signi ican ly ela ed o in es men in compu e s and
R&D. Fou h, we ind ha echnical p og ess has signi ican ly aised he
ela i e p oduc i i y o skilled wo ke s (we es ima e a posi i e skill-bias o
echnical p og ess). Finally, we show ha he con en ional app oach ha
measu es annual, a he han hou ly ela i e wages, p oduces a downwa d
bias in he es ima e o he skill-bias o echnical p og ess. The eason is
ha changes in ela i e hou s wo ked a e inco ec ly a ibu ed o changes
in ac o p ices a he han quan i ies.
The pape is s uc u ed as ollows. Sec ion 2 b ie ly discusses he heo e -
ical backg ound o he analysis and ela es he p esen wo k o he li e a u e.
Sec ion 3 p o ides a desc ip ion o he da a se and p esen s some s ylized
ac s o wage and employmen dynamics in I aly in he las decade. In sec ion
4 we p esen a decomposi ion o he agg ega e changes in he ela i e wage
bill, employmen and wages, in o hei espec i e wi hin- i m and be ween-
i m componen s. Sec ion 5 akes a close look a he beha io o wages,
and shows he implica ions o disagg ega ing annual wages in o he num-
be o hou s wo ked and hou ly wages. In sec ion 6 we p esen e idence
om i m-le el eg essions o p o ide an in e p e a ion o he obse ed wage
and employmen dynamics, and sec ion 7 ocuses on he bias o skill-biased
echnical change. Sec ion 8 concludes wi h a discussion o he main esul s.
3
2 Technology, ade and wages
In he las wo decades labo ma ke s in OECD coun ies ha e wi nessed a
signi ican change in he s uc u e o employmen and wages o skilled and
unskilled wo ke s. In he Uni ed S a es and he Uni ed Kingdom, bo h he
sha e o non-manual employmen and he wage di e en ial be ween manual
and non-manual wo ke s ha e g own conside ably since he ea ly 1980s.2
In con inen al Eu ope, wage di e en ials ha e been s able, and mos o he
adjus men has aken place on he quan i y side, wi h ising non-manual
wo ke s’ employmen a es and manual wo ke s’ unemploymen a es.3The
con en ional wisdom o Eu ope is ha he lack o adjus men in ela i e
wages is due o mo e igid labo ma ke ins i u ions (minimum wages, hi ing
and i ing cos s, cen alized ba gaining and union powe , e c.), wi h unem-
ploymen a es adjus ing o he alling demand o unskilled wo ke s.
A la ge body o li e a u e has a emp ed o p o ide an in e p e a ion o
hese de elopmen s,4wi h mos s udies concen a ing on he de e minan s
o he ela i e demand o skilled labo .5In pa icula , ade in eg a ion and
echnological change ha e been conside ed he main ac o s behind he ise
in demand o skilled wo ke s.6The “ echnology” iew a gues ha echnical
p og ess has been skill-biased: new p oduc ion p ac ices associa ed o he
in oduc ion o compu e s ha e inc eased he ela i e p oduc i i y o skilled
wo ke s. This has led o highe ela i e demand, and in u n o highe
employmen sha e and wage p emia o skilled wo ke s. Empi ically, skill-
biased echnical change is consis en wi h inc eased employmen sha es o
skilled labo wi hin indi idual sec o s (o i ms/plan s, depending on he
2See e.g. Ka z and Mu phy (1992)), Bound and Johnson (1992), Law ence and Slaugh-
e (1993), Be man, Bound, and G iliches (1994)) o he Uni ed S a es, and Haskel (1998),
Haskel and Slaugh e (2001b)) o he Uni ed Kingdom.
3See e.g. F eeman and Ka z (1996), OECD (1997), Be man, Bound and Machin (1998),
Machin and Van Reenen (1998), Ca d, K ama z, and Lemieux (1998).
4Fo ecen su eys o his li e a u e see Haskel (2000) and Slaugh e (1999).
5As o supply, Ka z and Mu phy (1992) a gue ha lowe ela i e supply o skills could
accoun only o a small pa o he obse ed changes in ela i e wages in he Uni ed
S a es be ween 1963 and 1987. See also Topel (1997) o an analysis o he supply-side
de e minan s o wage inequali y.
6O he expana ions o en p oposed a e ou so cing (see e.g. Haskel (1996), Feens a and
Hanson (1999)), and changes in ins i u ional ac o s such as he decline o he in luence
o unions, collec i e ba gaining, and lowe minimum wages (see e.g. Gosling and Machin
(1993) and Fo in and Lemieux (1997)).
4

le el o agg ega ion and he speci ic way new echnologies a e adop ed). The
“ ade” iew poin s o S olpe -Samuelson e ec s o inc eased exposu e o
in e na ional ade.7Acco ding o ad oca es o his explana ion, compe i ion
om de eloping coun ies has lowe ed he ela i e p ice o unskilled-in ensi e
goods. As esou ces ha e shi ed o sec o s p oducing mo e p o i able skill-
in ensi e p oduc s, he ela i e demand o manual wo ke s has allen. This
a gumen hus “blames” he g ow h o ade in goods, se ices and ac o s
in he pas h ee decades (i.e. “globaliza ion”). Empi ically, he ade iew
is consis en wi h employmen mo ing om skill-unin ensi e owa ds skill-
in ensi e sec o s ( i ms o plan s).
The b oad consensus eme ging om he ea ly empi ical li e a u e, gen-
e ally based on s udies o indus y da a, is ha , while in e na ional ade
accoun s o no mo e han 15-20% o he ise in wage di e en ials, he es
can be explained by skill-biased echnical p og ess (see e.g. Bound and John-
son (1992) and Be man e al. (1994) o he Uni ed S a es, bu also Be man
e al. (1998) and Machin and Van Reenen (1998) o an in e na ional pe -
spec i e).8This conclusion is suppo ed by wo main indings. Fi s , mos
o he agg ega e skill upg ading is due o changes wi hin indus ies, whe eas
he ealloca ion o employmen be ween indus ies plays a smalle ole. Sec-
ond, wi hin-indus y skill upg ading is signi ican ly ela ed o a numbe o
indica o s o echnological change.
This explana ion has been ecen ly challenged, bo h empi ically and he-
o e ically. A he empi ical le el, a numbe o s udies based on i m- o
plan -le el da a each conclusions signi ican ly di e en om hose ob ained
on he basis o indus y da a.9Be na d and Jensen (1997), o example,
ind ha wi hin-indus y inc eases in he demand o skilled labo can be
la gely a ibu ed o shi s in employmen be ween plan s o he same indus-
y (see also Be na d and Jensen (1995)), wi h expo ing plan s playing a
majo ole.10 Ea lie s udies, i is a gued, ha e igno ed impo an dynam-
7See Richa dson (1995), Wood (1995) and Slaugh e (1998) o ecen su eys on he
e ec s o ade on wage dynamics.
8A simila conclusion has been eached using bo h p ice (e.g. Leame (1996), Feens a
and Hanson (1996)) and olume (e.g., Bo jas, F eeman and Ka z (1997)) da a o cap u e
he e ec o ade on he labo ma ke .
9Mos plan - and i m-le el analyses aim a assessing he links be ween expo ing ac-
i i y and p oduc i i y (see e.g. Be na d and Jensen (1999) and Be na d e al. (2000))
o he exis ence o lea ning e ec s associa ed wi h he expo s s a us o i ms (see e.g.
Cla ides, Lauch and Tybou (1998)).
10Fo a heo e ical explana ion o his e idence see Manasse and Tu ini (2001).
5
ics occu ing a he le el o indi idual i ms and es ablishmen s, and hus
ha e la gely unde es ima ed he ole o demand and ade. A he heo e -
ical le el, ade heo is s ha e a gued ha wha ma e s o ac o p ices
(in a wo-sec o wo- ac o Hecksche -Ohlin economy) is he sec o in which
echnical p og ess occu s, a he han i s ac o bias (see e.g. Leame (1994,
1998)).11
The e a e ela i ely ew s udies on he I alian case. Mos o he exis -
ing e idence o I aly is based on indus y-le el da a. Bella and Quin ie i
(2000) analyze a panel o manu ac u ing indus ies, and a gue ha ade
compe i ion has had a small impac on employmen changes, whe eas ech-
nological p og ess has played a majo ole. Faini e al. (1999) each simila
conclusions on he limi ed ole o ade o labo ma ke dynamics, using
a panel o ou een manu ac u ing sec o s be ween 1985 and 1995. Among
i m-le el s udies, Dell’A inga and Luci o a (1994) look a a c oss sec ion o
me al-mechanical i ms o discuss he ole o ade unions in a ec ing wage
di e en ials.12 Casa ola e al. (1996) conside a la ge panel o i ms be-
ween 1986 and 1990, inding ha echnological change explains mos o he
inc ease in ela i e employmen . Mo e ecen ly, Manasse e al. (2001) an-
alyze a panel o me al-mechanical i ms and ind ha skill-biased echnical
change is he main de e minan o skill upg ading, aising wage inequali y
wi hin skilled wo ke s (i.e. be ween manage s and cle ks) mo e han be ween
manual and non-manual wo ke s.13
Agains his backg ound, ou pape con ibu es o he li e a u e in se e al
espec s: da a, me hodology and, we hink, esul s. As o he i s aspec ,
we exploi a new and much mo e comp ehensi e da a se o I aly, illing
an impo an gap o assessing he ole o echnology and ade o his
coun y; as o me hodology, we p o ide a gene al and consis en app oach o
11K ugman (1995), howe e , shows ha his c i icism es s on he assump ion o local
echnical change a ec ing a small open economy. See Haskel (2000) o an in e p e a ion
o his deba e, and Haskel and Slaugh e (2001a) o empi ical e idence on he ole o
sec o bias o he dynamics o wage di e en ials.
12E ickson and Ichino (1995) and Dell’A inga and Luci o a (2000) discuss he ole o
labo ma ke ins i u ions in explaining a comp essed wage s uc u e in I aly. Fe agina
and Quin ie i (1998) examine he ela ionship be ween expo ac i i y, p oduc i i y and
pe o mance. See also Quin ie i and Rosa i (1995) o an in es iga ion o in e -indus y
wage di e en ials.
13This s udy also inds ha ade has dampened he e ec s o echnology on he labo
ma ke , as employmen has shi ed owa ds unskilled-in ensi e i ms (see also Faini e al.
(1999)).
6
i m-le el be ween/wi hin decomposi ions; mo eo e , we show how p e ious
es ima es o he ole o echnical p og ess may con ain a “bias o he bias”, by
a ibu ing changes in ela i e hou s o ac o p ices a he han quan i ies.
3 The da a
Ou analysis is based on i m-le el da a o he I alian manu ac u ing sec o .
The da a se is d awn om he S a is ical In o ma ion Sys em on En e -
p ises (SISSI), de eloped by he I alian S a is ical Ins i u e (ISTAT, Cen al
Di ec o a e o S a is ics on Ins i u ions and En e p ises), and i combines in-
o ma ion om ou sou ces: he Sys em o Accoun s o Fi ms (SCI) and he
Su ey on Technological Inno a ion o Indus ial En e p ises (INN), bo h
collec ed on a yea ly basis; he mon hly s a is ics on Fo eign T ade Flows
(COE), and he A chi es o Ac i e Fi ms (ASIA, SIRIO, NAI).14
The da a se p o ides in o ma ion on he p o i and loss accoun (sales,
ou pu , cos s and ou lays, alue added, labo cos s, capi al dep ecia ion and
allowances, in e es s on deb s, axes, p o i s, e c.), he asse and liabili ies
accoun ( eal asse s, inancial asse s and liabili ies, inancial and comme -
cial c edi s and debi s, e c.), i ms’ employmen and wages, ixed capi al
o ma ion, R&D and expo s. Ou sample consis s o a balanced panel o
8441 manu ac u ing i ms, co e ing abou 22 pe cen o o al manu ac u ing
employmen , wi h annual obse a ions om 1989 o 1995. Da a on employ-
men and wages a e a ailable sepa a ely o manual wo ke s ( ainees and
p oduc ion wo ke s) and non-manual wo ke s (cle ks and execu i es).15 The
majo i y o i ms in he sample (63%) alls in o he ca ego y o “medium”
i ms (be ween 25 and 100 employees), while 23% a e “la ge” (mo e han 100
employees) and he emaining 14% a e “small” (below 25 employees). As o
he geog aphic dis ibu ion, 80% o he i ms in he sample a e loca ed in
No he n I aly, 15.5% in Cen al I aly and he emaining 4.5% in he Sou h.16
Table 1 p o ides a p elimina y desc ip ion o he da a, epo ing sample
14See So ce and Fazio (1999) and Co sini, Di F ancescan onio and Monducci (1998) o
a mo e de ailed desc ip ion o he cons uc ion o he da a se .
15Wages include sala ies, social con ibu ions paid by he i m, and con ibu ions paid
by he i m o he se e ance-paymen und (TFR).
16The h ee geog aphic a eas a e de ined as ollows. No h: Piemon e, Valle D’Aos a,
Lomba dia, Al o Adige, Vene o, F iuli Venezia Giulia, Ligu ia, Emilia Romagna. Cen e :
Toscana, Umb ia, Lazio, Ma che, Ab uzzo, Molise. Sou h: Campania, Basilica a, Puglia,
Calab ia, Sicilia, Sa degna.
7
and (app op ia ely de ined) sub-sample a e ages o a numbe o wage and
employmen indica o s. Column 1 shows he sha e o non-manual wo ke s in
he wage bill (W Bn
W B ), while columns 2 and 3 display i s componen s: he a io
o he wage a e o non-manual wo ke s o e he a e age wage (Wn
W, hence o h
“skill p emium”), and he sha e o non-manual wo ke s in employmen (En
E,
hence o h “skill in ensi y”). In he pe iod 1989-1995, on a e age, he sha e
o non-manual wo ke s in he wage bill was 43.3 pe cen , he skill p emium
135.9 pe cen , and skill in ensi y 31.8 pe cen . Table 1 also epo s, in
columns 4 and 5, he a e age annual wage a e o non-manual and a e age
wo ke s (Wn= 68.2 and W= 50.2 millions I alian li a, espec i ely); and,
in columns 6 and 7, he numbe o non-manual and o al employees in he
sample (En= 43.3 and E= 136 housands, espec i ely).
Be ween 1989 and 1995 he sha e o non-manual wo ke s in he wage
bill ose by 3.5 pe cen age poin s (0.58 pe cen a yea , on a e age). This
e lec ed a signi ican inc ease in skill in ensi y (2.4 pe cen ), wi h a ela i ely
modes 0.8 pe cen ise in he skill p emium. Hence, ela i e wages in ou
sample con o m o he “s icky” pa e n ound in o he s udies o ea lie
pe iods (e.g. E ickson, Ichino (1995)). The ise in skill-in ensi y, in u n,
e lec ed an absolu e inc ease o a e age non-manual employmen ( om 41.6
o 43.4 housands) despi e he con ac ion, om 137.8 o 133.2 housands,
o o al employmen (no e ha his implies ha manual employmen ell by
6.4 housand uni s in ou sample o i ms).
The ollowing blocks in Table 1 documen he signi ican he e ogenei y o
i ms in he sample as a as wages and employmen a e conce ned. G ouping
i ms acco ding o hei size, la ge i ms pay subs an ially highe nominal
wages han small and medium i ms. Skill p emia a e highes in medium-
size i ms (134.7 pe cen ) and lowes in small i ms (128.5 pe cen ), while
he wage bill sha e and skill in ensi y a e inc easing in size. Conside ing a
classi ica ion based on he geog aphic dis ibu ion, i ms loca ed in he Sou h
a e on a e age smalle (119.5 employees) and pay subs an ially lowe wages
han hose in he es o he coun y. Also, hey appea o pay highe skill
p emia (140.2%) han hose in he es o he coun y, al hough hey a e
cha ac e ized by lowe wage bill sha es (36%) and skill in ensi y (25.7%).
Nex , we conside wo u he classi ica ions, acco ding o hei expo
ac i i y and compu e in ensi y. “High-expo ” (“low-expo ”) i ms a e de-
ined as hose whose sha e o expo s in o al sales is abo e (below) he
8
subs i u ed manual wi h non-manual wo ke s no only in e ms o employ-
men le els (on he ex ensi e ma gin),a he annual a e o Ewi = 0.63,
bu also in e ms o hou s (on he in ensi e ma gin), a he annual a e o
Hwi = 0.19. The la e phenomenon is simply obscu ed when he s an-
da d de ini ion o annual wages is used. Rela i e o al non-manual hou s
ha e he e o e isen app oxima ely a he annual a e o 0.63+ 0.19 = 0.82
which is abou one hi d abo e he es ima e in Table 3. We now u n o he
in e p e a ion o hese decomposi ions.
6 In e p e ing he decomposi ions
So a we ha e in e p e ed he wi hin and be ween componen s as e lec ing
echnology and demand shocks, espec i ely. This in e p e a ion, howe e ,
is no wa an ed: wi hin- i m changes may also be due o demand shocks.
Suppose, o example, ha he domes ic ela i e p ice o unskilled-in ensi e
(“ adi ional”) goods ises, due o a change in p e e ences o o ade libe al-
iza ion.23 As new i ms en e he “ adi ional” sec o , he sha e o unskilled
wo ke s in employmen ises (be ween e ec ).The esul ing excess demand
o unskilled wo ke s lowe s he wage p emium, and induces i ms o subs i-
u e manual wi h non-manual wo ke s (a posi i e employmen wi hin e ec ).
In his case, a demand shock (be ween i ms) indi ec ly causes a (wi hin i m)
change in ac o p opo ions. A ibu ing he la e o echnology would be
inco ec , and i would esul in o e es ima ing he ole o echnology (and
unde es ima ing ha o demand o ade).
In his sec ion we he e o e examine whe he i is co ec o in e p e
wi hin and be ween componen s as e lec ing echnology and demand, e-
spec i ely. We eg ess he be ween and wi hin changes o wages (bo h annual
and hou ly), employmen and hou s, on a iables ha p oxy o i m-le el
demand and echnology shocks. I he s anda d in e p e a ion is co ec ,
wi hin- i m changes should be signi ican ly ela ed o echnology bu no o
demand a iables, while he con e se should be ue o be ween changes.
We use he a e o g ow h o o al sales as an indica o o he change
in demand o a i m’s ou pu , and conside wo al e na i e indica o s o
echnological change a i m-le el: he a io o in es men in compu e s o e
o al in es men , and he a io o esea ch and de elopmen expendi u es o e
23We hank Paolo Epi ani o aising his poin .
15

o al sales24. All he eg essions include size, egion, and indus y dummies o
allow o di e en i m and indus y cha ac e is ics. The gene al speci ica ion
is he e o e:
∆Ci
d=α+β1∆lSi+β2ICIi+β3RDSi+X
j
γjDUMj(5)
whe e ∆Ci
dindica es i m i0s con ibu ion o he o e all change in he el-
a i e wage bill, employmen and (annual and hou ly) wage (∆C=W B,
E, W, HW, H),and he subsc ip d=be , wi deno es be ween and wi hin
componen s, espec i ely; ∆lS is he g ow h a e o o al sales, ICI is he
a io o he i m’s in es men in compu e s o e o al in es men , RDS is he
a io o Resea ch and De elopmen expendi u es o e o al sales, and DUM
ep esen s a se o indus y, size and geog aphic dummies.
The esul s o OLS es ima ion o equa ion (5), p esen ed in able 6,
a e qui e e ealing 25. The g ow h a e o sales has a posi i e and highly
signi ican coe icien in all be ween eg essions (wi h he excep ion o he
hou ly wage equa ion): demand shocks a e posi i ely ela ed o be ween- i m
changes in bo h employmen and annual wages, bu no o wi hin changes
(wi h he excep ion o hou s, Hwi , and hou ly wages, HW wi ). Looking
a he echnology indica o s, he compu e sha e o in es men ICI is posi-
i e and signi ican in he wage bill and employmen wi hin equa ions, while
nega i e bu ne e signi ican in he be ween equa ions. The esea ch and
de elopmen indica o RDS is posi i e bu only ma ginally signi ican in he
wage bill and employmen wi hin equa ions. In e es ingly, i is posi i e and
s ongly signi ican in he equa ion o he wi hin i m ela i e numbe o
hou s, Hwi . The esul s o hou ly wages a e less clea -cu : he wi hin com-
ponen is signi ican ly ela ed o bo h he g ow h o sales (posi i ely) and
he R&D indica o (nega i ely); he be ween componen is no signi ican ly
a ec ed by ei he demand o echnology indica o s.
O e all, he e idence sugges s ha be ween- i m changes o all he indi-
ca o s examined a e posi i ely and signi ican ly ela ed o changes in demand.
In addi ion, he e is a posi i e and signi ican ela ionship be ween echnical
change, as measu ed by in es men in compu e s and R&D in ensi y, and
24The R&D a iable also con ains expendi u es o pa en s, concessions, and copy igh s.
25The lowe numbe o obse a ions ( om 8203 in he decomposi ions o 7377 in he
eg essions) is la gely due o da a limi a ions on he echnology indica o s: only 8005 and
7830 obse a ions, espec i ely, a e a ailable o he compu e in ensi y and esea ch and
de elopmen indica o s.
16
wi hin- i m skill upg ading bo h on he ex ensi e ma gin (numbe o em-
ployees) and he in ensi e ma gin (numbe o hou s wo ked pe employee).
7 The biased bias o echnical change
In he p e ious sec ions we ound ha main de e minan o he ise o
he non-manual employmen and wage bill sha es is i ms subs i u ing non-
manual o manual wo ke s. In his sec ion, we use a cos unc ion ame-
wo k o measu e e ec o echnical change on he ela i e p oduc i i y o
non-manual wo ke s ( he so called skill-bias o echnical change). We ind a
posi i e and signi ican skill-bias. Also, we show ha he common p ac ice
o de ining ela i e wages in e ms o annual, a he han hou ly, sala ies,
p oduces a downwa d bias in he es ima es o he skill-bias as well as o he
elas ici y o ac o subs i u ion.
In o de o isola e he e ec o echnical p og ess on ac o sha es, one
needs o con ol o changes in ac o p ices and capi al in ensi y: he ise
in he sha e o skilled wo ke s wi hin i ms may be simply due o a all in
hei ela i e ac o p ices o o capi al deepening when skills and capi al
a e complemen . Following he li e a u e (see Binswange , 1974) we he e-
o e de ine echnical p og ess as a educ ion in uni cos (an inwa d shi o
he uni -isoquan ) a cons an ac o p ices and capi al in ensi y. Technical
p og ess is neu al i , despi e lowe uni cos s, i ms on a e age do no change
ac o p opo ions, a gi en ac o p ices and capi al in ensi y. Howe e , i
hey inc ease on a e age he p opo ion o skilled wo ke s in employmen
(when hey pick a new angency poin on an lowe isocos line o he same
slope), hen echnical p og ess aises he ela i e p oduc i i y o non-manual
wo ke s and is de ined skill biased.
Empi ically, we implemen his app oach ollowing Be man e al. (1994),
and B own and Ch is ensen (1981). An equa ion o he wage bill sha e can
be de i ed om a anslog cos unc ion wi h quasi- ixed ac o s o p oduc-
ion. Assume ha i ms choose a iable ac o s, manual and non manual
labo , in o de o minimize cos s, subjec o an ou pu cons ain . P oduc-
ion equi es (manual and non-manual) labo and capi al, which is ixed in
he sho un. The cos unc ion has he anslog unc ional o m, and e-
u ns o scale a e cons an . Unde hese assump ions he change in he sha e
17
o non-manuals in he wage bill can be w i en as ollows:
∆(W Bi
n
WBi) = α+β∆ ln( wi
n
wi
m
) + γ∆ ln(Ki
Yi) + εi(6)
whe e Kiand Yi ep esen capi al and alue added, espec i ely ( he ac ual
speci ica ion also includes a se o indus y, size and geog aphic dummies, as
in (5)). No e ha he in e cep αmeasu es o he a e age bias in echnical
change, and he esidual εip o ides an es ima e o he i m-speci ic bias. I
he slope coe icien βis posi i e (nega i e) a change in ela i e p ice o a ac-
o aises (lowe s) i s cos sha e, implying ha he elas ici y o subs i u ion
be ween inpu s is below (abo e) uni y (σ=−β+sn(1−sn)
sn(1−sn),whe e sn=W Bn
W B ).
A posi i e (nega i e) es ima e o γimplies ha capi al is complemen (sub-
s i u e) o non-manual labo , since i aises (lowe s) i s wage bill sha e a
cons an ac o p ices. In he ollowing we p esen esul s ob ained es ima -
ing he abo e equa ion using ei he annual o hou ly wages (w=W, ω) as
explana o y a iable.
Table 7 epo s OLS es ima ion esul s using annual wages. We es ima e
equa ion (6) in i s basic e sion , and subsequen ly add, ei he indi idually
o join ly, he wo indica o s o echnological change desc ibed abo e (com-
pu e s as a sha e o o al in es men and R&D o e sales). S a ing om he
basic speci ica ion, we see ha he cons an is posi i e and signi ican : he
inc ease in he ela i e p oduc i i y o non manual wo ke s ( he bias o ech-
nical p og ess) occu s a an annual a e o 0.48 and hus aises he wage bill
sha e o skilled wo ke s by almos hal o a pe cen age poin pe yea . The
change in ela i e wages ha e a posi i e and s a is ically signi ican coe i-
cien , implying an elas ici y o subs i u ion be ween labo inpu s o σ= 0.49.
The coe icien o he capi al- alue added a io is also posi i e and signi i-
can , indica ing complemen a i y be ween capi al and skilled labo . Capi al
deepening has hus con ibu ed o skill upg ading. When we add o he basic
model echnology indica o s indi idually (equa ions 2-3) o join ly (equa ion
4), bo h he compu e sha e o o al in es men and R&D expendi u es as a
ac ion o sales ha e posi i e and highly s a is ically signi ican coe icien s.
The es ima e o he skill bias alls sligh ly ( o 0.44) when explici p oxies o
echnical p og ess a e included in he equa ion, while he es ima ed elas ici y
o subs i u ion is ema kably s able ac oss di e en speci ica ions.26
26The esul s a e also obus s o he use o beginning-o -pe iod le els o he echnology
indica o s.
18
Nex we e-es ima e he p e ious equa ion wi h hou ly wages on he igh
hand side, and ob ain he esul s shown o Table 8. Compa ed wi h hose
in Table 7, he pa ame e s o capi al deepening and he compu e sha e in
in es men a e i ually unchanged, while he es ima es o he a io o R&D
expendi u es o e sales a e mo e p ecisely es ima ed and almos double in
size. Mo ing o hou ly wages has wo mo e impo an consequences: he
es ima ed skill-bias ises consis en ly in all speci ica ions, espec i ely om
αin he ange (0.48, 0.44) o α0in he ange (0.52-0.48). Simila ly, he
es ima ed elas ici y o subs i u ion ises om σ= 0.49 o σ0= 0.67. The
eason o he la ge es ima ed skill bias is he ollowing: echnical p og ess
aises, as we saw, he ela i e numbe o non-manual hou s, bu when wages
a e inco ec ly measu ed (on an annual, a he han hou ly basis), his e ec
is a ibu ed o highe ela i e ac o p ices, a he han o he bias. As
o he la ge elas ici y o subs i u ion, no e ha in he second speci ica ion
his elas ici y e ec i ely measu es he change in o al hou s (employmen
plus a e age hou s) induced by a change in ela i e ac o p ices, so ha he
es ima ed elas ici y mus also be la ge As long as echnical inno a ion and
skilled hou s a e complemen , p e ious es ima es o he skill-bias (and o he
elas ici y o subs i u ion), e.g. Be man e al., 1994, Be man e al., 1998, a e
he e o e likely o be biased downwa ds.
Summing up, ou es ima e sugges s ha skill-biased echnological change
has aised he ela i e p oduc i i y o non manual wo ke s an annual a e
o oughly hal o a pe cen age poin , and hus was he key de e minan o
he inc ease in he demand o non-manual wo ke s in I alian manu ac u ing
du ing he 1990s. We also ound ha in o de o assess he ole o echnical
p og ess on wage inequali y and skill upg ading, i is impo an o disagg e-
ga e annual wage a es in o he numbe o hou s wo ked and hei hou ly
p ice. The cu en p ac ice in he li e a u e ails o do so, and he e o e
e oneously a ibu es changes in hou s o ac o p ices a he han quan i-
ies. This p oduces a downwa d bias in he es ima ed skill-bias o echnical
p og ess.
8 Discussion and conclusions
This pape has p esen ed i m-le el e idence on he dynamics o wage p emia
and ela i e employmen and hou s in I alian manu ac u ing in he nine ies.
We ha e exploi ed a b and new da a se , p e iously una ailable o esea ch,
19
ha co e s a balanced panel o 8441 manu ac u ing i ms be ween 1989 and
1995. The analysis has eached a numbe o in e es ing esul s on he e ec s
o echnology and ade on employmen and wages in I alian manu ac u ing
i ms.
Fi s , I alian i ms ha e subs i u ed unskilled o skilled wo ke s a a a e
compa able o hose expe ienced in o he indus ialized coun ies, wi h high-
ech i ms playing a leading ole in his p ocess (wi hin- i m skill upg ading is
he main de e minan o he shi in ela i e labo demand in he nine ies).
This esul is a new, and somewha unexpec ed, esul , gi en ha mos
s udies on Eu opean economies ind signi ican e ec o echnical p og ess a
sec o le el only a e 1995 (e.g. Da e i, 2000).
By con as , demand changes associa ed o ade ha e mo ed employ-
men away om skill-in ensi e i ms, con ibu ing o mode a e he change
in ela i e ac o p ices (be ween- i m employmen shi s has educed he el-
a i e demand o skills). This inding is consis en wi h esul s in Manasse
e al (2002), and is p obably due o he anomalous specializa ion pa e n o
I alian ade. Du ing he nine ies i ms has become inc easingly specialized
in unskilled-in ensi e “ adi ional” goods (such as shoes, ex iles, u ni u e
e c., see Chia lone, 2001).
Second, he ela i e s abili y o wage di e en ials wi hin i ms hides an
impo an composi ion e ec : he ela i e numbe o hou s wo ked by skilled
wo ke s has isen whe eas ela i e hou ly wages ha e allen. The na owing
o hou ly skill p emia in he ace o echnical p og ess may come unexpec ed,
pa icula ly o eade s un amilia wi h he ea u es o he I alian labo ma -
ke . Ye i is well known he I alian cen alized sys em o wage ba gaining
sys ema ically ails o ailo wages o i ms and wo ke s p oduc i i y, wi h
unions ac ing as a powe ul ins umen o wage equaliza ion. Fo example,
sala ies in he Sou h a e equalized o sala ies in he No h, despi e la ge p o-
duc i i y gaps, and his is gene ally ega ded as an explana ion o a a e o
unemploymen which is ou imes la ge in he Sou h han in he No h. In
addi ion, possibly as a esul o his comp ession in ela i e hou ly wages,I aly
has been expo ing college g adua es and skilled wo ke s (“ he b ain d ain”)
a a a e ha has no compa ison in Eu ope (see Becke , Ichino and Pe i
(2002))
Thi d, wi hin- i m skill upg ading, measu ed by changes in bo h ela i e
employmen and numbe o hou s, is s ongly and signi ican ly ela ed o
in es men in compu e s and R&D).
Fou h, echnical p og ess has aised he ela i e p oduc i i y o non
20

manual-wo ke a an annual a e o hal a pe cen age poin , and he e o e he
skill bias has been a key de e minan o he inc ease in he ela i e demand
o non-manual wo ke s in I alian manu ac u ing in he las decade.
Finally, he pape makes an impo an me hodological poin : in o de o
assess he ole o echnical p og ess on wage inequali y and skill upg ading,
i is essen ial o disagg ega e hou s wo ked om hei p ice. Failing o do so,
and a ibu ing hou s o ac o p ices a he han quan i ies, biases downwa d
he es ima es o he skill bias whene e echnical p og ess and hou s a e
complemen .
Whe he hese esul s ex end beyond he manu ac u ing sec o is one o
he ques ions o be in es iga ed in u he esea ch.
21
9 Appendix
This appendix p o ides some de ails on he de i a ion o he con ibu ions
o employmen , hou s wo ked and hou ly wages o he wage bill p esen ed in
sec ion 5:
∆WBn
WB = ∆
I
X
iEi
n
E
Wi
n
W
=
I
X
i"∆Ei
n
EWi
n
W+ ∆ Wi
n
WEi
n
E#=
=
I
X
i



∆Ei
n
EWi
n
W+ ∆ hi
n
hωi
n
ωEi
n
E+
∆ωi
n
ωhi
n
hEi
n
E

=
=
I
X
i









∆Ei
n
EiWi
n
W
Ewi
+ ∆ Ei
EWi
n
W
Ebe
+
∆hi
n
hiωi
n
ωEi
n
E
Hwi
+ ∆ hi
hωi
n
ωEi
n
E
Hbe
+
∆ωi
n
ωihi
n
hEi
n
E
HW wi
+ ∆ ωi
ωhi
n
hEi
n
E
HW be









The i s , second and hi d line abo e co espond o he i s , second and
hi d e m o equa ion (4) in he ex
22
Re e ences
Becke , Sascha O., And ea Ichino, and Gio anni Pe i (2003) ‘How la ge is
he ’b ain d ain’ om i aly?’ Gio nale degli Economis i e Annali di
Economia
Bella, Ma iano, and Beniamino Quin ie i (2000) ‘The e ec o ade on em-
ploymen and wages in i alian indus y.’ Labou 14, 291–309
Be man, Eli, John Bound, and S ephen Machin (1998) ‘Implica ions o skill-
biased echnological change: In e na ional e idence.’ Qua e ly Jou nal
o Economics pp. 1245–1280
Be man, Eli, John Bound, and Z i G iliches (1994) ‘Changes in he demand
o skilled labo wi hin u.s. manu ac u ing: E idence om he annual
su ey o manu ac u es.’ Qua e ly Jou nal o Economics pp. 367–397
Be na d, And ew B., and B ad o d J. Jensen (1995) ‘Expo e s, jobs and
wages in u.s. manu ac u ing: 1976-1987.’ B ookings Pape s on Economic
Ac i i y: Mic oeconomics pp. 67–119
(1997) ‘Expo e s, skill upg ading and he wage gap.’ Jou nal o In e -
na ional Economics 42, 3–31
(1999) ‘Excep ional expo e pe o mance: Cause, e ec , o bo h?’
Jou nal o In e na ional Economics 47, 1–25
Binswange , Hans P. (1974) ‘The measu emen o echnical change biases
wi h many ac o s o p oduc ion.’ Ame ican Economic Re iew pp. 964–
976
Black, Sand a E., and Lisa M. Lynch (2000) ‘Wha ’s d i ing he new econ-
omy: The bene i s o wo kplace inno a ion.’ Technical Repo 7479,
NBER Wo king Pape
Bo jas, G. J. (1980) ‘The ela ionship be ween wages and weekly hou s o
wo k: The ole o di ision bias.’ Jou nal o Human Resou ces 15, 409–
423
23
Bo jas, Geo ge J., Richa d F eeman, and Law ence Ka z (1992) ‘On he
labo -ma ke e ec s o immig a ion and ade.’ In ‘Immig a ion and he
Wo k o ce’ (Chicago: Uni e si y o Chicago P ess) chap e aa, pp. 213–
44
Bound, John, and Geo ge Johnson (1992) ‘Changes in he s uc u e o waes
du ing he 1980s: An e alua ion o al e na i e explana ions.’ Ame ican
Economic Re iew 82, 371–92
B own, Randall S., and Lau i s R. Ch is ensen (1981) ‘Es ima ing elas ici ies
o subs i u ion in a model o pa ial s a ic equilib ium: An applica ion
o u.s. ag icul u e, 1947 o 1974.’ In Modeling and Measu ing Na u al
Resou ce Subs i u ion, ed. E ns R. Be nd and Ba y C. Field (Cam-
b idge, MA: MIT P ess) pp. 209–29
Ca d, D., F. K ama z, and T. Lemieux (1998) ‘Changes in he ela i e s uc-
u e o wages and employmen : A compa ison o canada, ance and he
uni ed s a es.’ Technical Repo 2008, CEPR Wo king Pape
Casa ola, Paola, And ea Ga os o, and Paolo Ses i o (1996) ‘Technical
p og ess and wage dispe sion in i aly: E idence om i ms’ da a.’ An-
nales d’Economie e S a is ique 0, 387–412
Cle ides, S., S. Lauch, and J. Tybou (1998) ‘Is lea ning by expo ing im-
po an ?’ Qua e ly Jou nal o Economics 454, 903–47
Co sini, V., T. Di F ancescan onio, and R. Monducci (1998) ‘Cos uzione
di base da i pe l’analisi della s u u a e della dinamica del sis ema di
imp ese.’ Technical Repo , ISTAT, Di ezione Cen ale delle S a is iche
su Is i uzioni e Imp ese
Da e i, F ancesco, and Guido Tabellini (1997) ‘Unemploymen , g ow h and
axa ion in indus ial coun ies.’ Technical Repo 122, IGIER Wo king
Pape
Dell’A inga, Ca lo, and Claudio Luci o a (1994) ‘Collec i e ba gaining
and ela i e ea nings in i aly.’ Eu opean Jou nal o Poli ical Economy
10, 727–47
(2000) ‘Inside he black box: Labou ma ke ins i u ions, wage o ma-
ion and unemploymen in i aly.’ Ri is a di Poli ica Economica 90, 13–55
24
Table 4: Hou s and hou ly wages: o e all and sub-sample a e ages
Sample ωn
ω
hn
hωnω hnhN.Obs.
O e all 131.5 103.3 39.6 30.1 1720.6 1665.2 59083
1989 131.2 102.9 32.1 24.4 1718.7 1669.8 8441
1990 130.4 103.3 34.8 26.7 1708.7 1653.4 8440
1991 130.1 104.0 38.1 29.3 1714.6 1648.6 8441
1992 131.9 103.3 40.6 30.8 1723.6 1668.2 8441
1993 130.1 104.1 41.6 32.0 1724.1 1655.6 8439
1994 131.4 102.8 43.8 33.4 1718.8 1671.6 8440
1995 132.4 102.7 46.0 34.7 1736.1 1690.8 8441
High exp. 131.0 103.7 39.4 30.1 1723.7 1661.7 29538
Low exp. 132.3 102.8 39.9 30.2 1716.5 1670.1 29545
High ech. 130.5 103.2 39.9 30.6 1709.2 1656.8 29523
Low ech. 132.5 103.8 39.2 29.5 1740.5 1676.5 29560
No e:ωn= non-manual hou ly wage pe wo ke ; hn= non-manual
a e age numbe o hou s wo ked pe employee (see sec ion 5).
Table 5: Hou s and hou ly wages in wage bill decomposi ions
Sample WB o Ewi Ebe Hwi Hbe HWwi HWbe N.Obs.
1989-95 0.58 0.63 -0.12 0.19 -0.00 -0.18 0.06 8203
89-92 1.00 0.74 0.11 0.41 0.04 -0.34 0.04 8204
92-95 0.17 0.57 -0.38 -0.03 -0.09 -0.06 0.16 8267
High exp. 0.28 0.33 -0.11 0.12 -0.01 -0.09 0.05 4168
Low exp. 0.30 0.30 -0.01 0.07 0.01 -0.09 0.01 4035
High ech. 0.29 0.39 -0.15 0.14 -0.00 -0.15 0.06 4154
Low ech. 0.29 0.24 0.03 0.05 0.00 -0.03 0.00 4049
No e:W B o = non-manual wage bill sha e, Ewi = Employmen wi hin,
Ebe = Employmen be ween, Hwi = Hou s wi hin, Hbe = Hou s be ween,
HW wi = Hou ly wage wi hin, HW be = Hou ly wage be ween.
31

Table 6: De e minan s o wage bill componen s
Dep. Va . ∆lS ICI RDS R2N.Obs.
WBwi -0.02 0.07 0.09 0.04 7377
( -1.24) ( 2.09) ( 1.29)
WBbe 0.72 -0.20 0.12 0.04 7377
( 9.91) ( -1.37) ( 0.46)
Ewi -0.04 0.08 0.12 0.03 7377
( -1.69) ( 2.04) ( 1.34)
Ebe 0.62 -0.19 0.23 0.04 7377
( 10.07) ( -1.50) ( 0.92)
Wwi 0.01 -0.01 -0.03 0.01 7377
( 0.94) ( -0.79) ( -0.71)
Wbe 0.10 -0.01 -0.11 0.02 7377
( 3.98) ( -0.35) ( -0.69)
Hwi -0.02 0.02 0.34 0.00 7377
( -1.87) ( 0.64) ( 3.78)
Hbe 0.11 -0.07 -0.03 0.01 7377
( 3.52) ( -1.23) ( -0.17)
HW wi 0.03 -0.03 -0.37 0.01 7377
( 2.22) ( -1.02) ( -3.87)
HW be -0.01 0.06 -0.07 0.00 7377
( -0.50) ( 1.41) ( -1.43)
No e: -s a is ics in pa en heses. All speci ica ions include size,
geog aphy and indus y sec o dummies, as de ined in sec ion 3.
Legend: ∆lS = g ow h a e o sales; ICI = Compu e sha e
o o al in es men ; RDS =R&D/ sales.
32
Table 7: De e minan s o wi hin i m skill upg ading
Equa ion α∆lWnm ∆lKY ICI RDS R2N.Obs.
(1) 0.48 0.11 0.34 0.23 8136
( 31.21) ( 32.51) ( 2.10)
(2) 0.43 0.11 0.36 0.56 0.24 7742
( 24.84) ( 32.32) ( 2.22) ( 5.91)
(3) 0.49 0.11 0.40 0.52 0.24 7684
( 30.87) ( 31.89) ( 2.39) ( 3.36)
(4) 0.44 0.11 0.43 0.58 0.51 0.25 7321
( 24.53) ( 31.82) ( 2.52) ( 5.94) ( 3.62)
No e: -s a is ics in pa en heses. Dependen a iable: ∆wbsh = change in log ela i e
wage bill; ∆lW nm = change in log ela i e annual wage; ∆lKY = change in log capi al
ou pu a io; ICI = Compu e sha e o o al in es men ; RDS =R&D/ sales.
Table 8: De e minan s o wi hin i m skill upg ading (hou ly wages)
Equa ion α∆lHW nm ∆lKY ICI RDS R2N.Obs.
(1) 0.52 0.07 0.46 0.14 8135
( 32.14) ( 20.48) ( 2.71)
(2) 0.47 0.07 0.46 0.59 0.14 7741
( 25.67) ( 19.86) ( 2.66) ( 5.76)
(3) 0.53 0.07 0.50 0.97 0.14 7684
( 31.53) ( 19.98) ( 2.82) ( 5.46)
(4) 0.48 0.07 0.50 0.63 0.97 0.15 7321
( 25.03) ( 19.50) ( 2.77) ( 5.92) ( 5.57)
No e: -s a is ics in pa en heses. Dependen a iable: ∆wbsh = change in log ela i e
wage bill; ∆lHW nm = change in log ela i e hou ly wage; ∆lKY = change in log capi al
ou pu a io; ICI = Compu e sha e o o al in es men ; RDS =R&D/ sales.
33