Lau e bach, Rodol o
A icle
Di e se knowledge o di e se inno a ion: E idence om
Chilean i ms
Es udios de Economía
P o ided in Coope a ion wi h:
Depa men o Economics, Uni e si y o Chile
Sugges ed Ci a ion: Lau e bach, Rodol o (2024) : Di e se knowledge o di e se inno a ion: E idence
om Chilean i ms, Es udios de Economía, ISSN 0718-5286, Uni e sidad de Chile, Depa amen o de
Economía, San iago de Chile, Vol. 51, Iss. 1, pp. 85-115
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/312791
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-nc-sa/4.0/
85
Es udios de Economía, Vol.51 - Nº 1, Junio 2024. Págs 85-115
Di e se knowledge o di e se inno a ion; e idence om Chilean
i ms
Conocimien os di e sos pa a inno aciones di e sas, e idencia de i mas
chilenas.
RODOLFO LAUTERBACH*
Abs ac
Using he Chilean Inno a ion Su ey o 2019-2020, his wo k s udies he e -
ec s o di e en knowledge sou ces on a ange o inno a ion ou pu s. Findings
e eal dis inc impac s o sou cing in o ma ion om compe i o s, cus ome s,
and go e nmen agencies on p oduc , p ocess, ma ke ing, o ganiza ional, and
social inno a ion ou pu s. In o ma ion om cus ome s has a posi i e e ec
on o e all inno a ion. Social inno a ion is posi i ely in luenced by in o ma-
ion sou ced om go e nmen agencies. These indings con ibu e o he un-
de s anding o how di e en knowledge sou ces shape inno a ion ou pu s on
de eloping coun ies. They p o ide aluable insigh s o i ms, policymake s,
and esea che s seeking o enhance inno a ion capabili ies and in o m e i-
dence-based policies.
Key wo ds: Inno a ion ou pu , Di e se knowledge sou ces, Chilean Inno a-
ion su ey, Bina y ins umen al a iable model.
JEL Classi ica ion: O31, O32, D22.
*Uni e sidad de Las Amé icas Chile, Facul ad de Ingenie ía y Negocios, Sede
P o idencia, Manuel Mon 948, San iago, Chile. [email p o ec ed]
Recei ed: Ma ch, 2023 Accep ed: Ap il, 2024
86 Es udios de Economía, Vol.51 - Nº 1
Resumen
U ilizando la Encues a de Inno ación de Chile pa a 2019-2020, es e abajo
es udia los e ec os de di e en es uen es de conocimien o en una a iedad de
esul ados de inno ación. Los hallazgos e elan dis in os impac os de la ob-
ención de in o mación de compe ido es, clien es y agencias gube namen ales
en los esul ados de inno ación social, o ganizacional, de ma ke ing, de p o-
cesos y de p oduc os. La in o mación de clien es iene un e ec o posi i o en la
inno ación gene al. La inno ación social se e in luenciada posi i amen e po
la in o mación p oceden e de agencias gube namen ales. Es os hallazgos con-
ibuyen a comp ende cómo las di e en es uen es de conocimien o dan o ma
a los esul ados de la inno ación en los países en desa ollo. P opo cionan
in o mación aliosa pa a emp esas, o mulado es de polí icas e in es igado-
es que buscan mejo a las capacidades de inno ación e in o ma polí icas
basadas en e idencia.
Palab as cla e: Resul ados de inno ación, Di e sas uen es de in o mación,
Encues a de inno ación chilena, Modelo de a iable ins umen al bina ia.
Clasi icación JEL: O31, O32, D22
1. INTRODUCTION
Inno a ion has long been ecognized as a c ucial d i e o economic g ow h
and compe i i eness. As socie ies and economies become inc easingly com-
plex and in e connec ed, he abili y o i ms o adap and inno a e becomes
e e mo e essen ial. Unde s anding he ac o s ha con ibu e o success ul in-
no a ion is he e o e o pa amoun impo ance o policymake s, esea che s,
and business leade s alike. I has been widely acknowledged ha i ms need o
look beyond hei in e nal esou ces and ap in o ex e nal knowledge o os e
inno a ion. Howe e , he speci ic mechanisms h ough which di e se knowl-
edge sou ces in luence inno a ion ou comes equi e u he in es iga ion.
O e he yea s, schola s ha e made signi ican p og ess in de eloping mod-
els o comp ehend he dynamics o inno a ion. Da a collec ed h ough inno-
a ion su eys has played a pi o al ole in un a eling he causali y behind in-
no a ion success. By examining a ious i m-le el a iables, such as esea ch
and de elopmen (R&D) expendi u e, human capi al, sales, and o al employ-
ees, esea che s ha e sough o iden i y he de e minan s o inno a ion ou pu .
Howe e , a undamen al ques ion emains: A e he e speci ic a iables ha
ha e dis inc impac s on speci ic ypes o inno a ion ou pu s? To shed ligh on
87
Di e se knowledge o di e se inno a ion; e idence om Chilean... / Rodol o Lau e bach
his ma e and disen angle he in ica e causal ela ionships behind inno a-
ion success, i is c ucial o de elop obus models ha conside bidi ec ion-
al e ec s. The mo e we unde s and which ac o s con ibu e o speci ic ypes
o inno a ion, he be e equipped we a e o o mula e e ec i e go e nmen
policies ha p omo e desi able ou comes o local economies, pa icula ly in
de eloping coun ies.
While p e ious s udies ha e shed ligh on he e ec s o knowledge sou c-
es on inno a ion pe o mance in a ious con ex s, he e is a need o explo e
hese issues wi hin he unique con ex o Chilean i ms. Chile is a dynamic
and eme ging economy ha has made signi ican e o s o os e inno a ion
and en ep eneu ship. The e o e, examining he ole o di e se knowledge in
Chilean i ms’ inno a ion ou pu s can p o ide a b oade unde s anding o in-
no a ion dynamics in eme ging economies.
Building upon he seminal wo k o Cohen and Le in hal (1990), his pape
ocuses on he i m’s capaci y o acqui e and u ilize in o ma ion om di e se
sou ces as a c i ical de e minan o inno a ion. Recognizing he impo ance
o add essing endogenei y conce ns, ou app oach d aws inspi a ion om he
wo k by C epon, Dugue , and Mai esse (1998).
The p ima y objec i e o his pape is o es ima e an empi ical model ha
e eals he causal ela ionships be ween di e en ypes o in o ma ion sou ces
and a ious o ms o inno a ion ou pu . To accomplish his, we le e age eli-
able inno a ion su ey da a collec ed in Chile du ing he pe iod o 2019-2020.
Ou model conside s he e olu ion o empi ical esea ch on he de e minan s
o inno a ion ou pu and employs ins umen al a iables o es ima e a bina y
ea men model wi h idiosync a ic a e age e ec s.
Ou indings demons a e ha he u iliza ion o di e se inno a ion in o -
ma ion sou ces has a ying impac s on di e en ypes o inno a ion ou pu s,
each wi h i s unique magni ude. No ably, while in o ma ion sou ced om cus-
ome s posi i ely in luences mos ypes o inno a ion, we ound no disce nible
e ec om in o ma ion ob ained om compe i o s. Fu he mo e, go e nmen
in o ma ion eme ges as a pa icula ly aluable esou ce, bene i ing social in-
no a ion signi ican ly while also exhibi ing posi i e e ec s on p ocess and o -
ganiza ional inno a ions.
By shedding ligh on he in ica e ela ionships be ween in o ma ion sou c-
es and inno a ion ou pu s, his s udy p o ides aluable insigh s o policymak-
e s, esea che s, and i ms seeking o enhance hei inno a ion capabili ies. The
empi ical e idence p esen ed he ein se es as a ounda ion o e idence-based
policy ecommenda ions aimed a os e ing speci ic ypes o inno a ion ha
can d i e he local economies o de eloping coun ies o wa d.
O e all, his esea ch con ibu es o he exis ing li e a u e by o e ing a
comp ehensi e analysis o he ole o di e se knowledge in d i ing di e se
88 Es udios de Economía, Vol.51 - Nº 1
inno a ion ou comes. By highligh ing he nuanced ela ionships be ween in-
o ma ion sou ces and inno a ion ou pu s, we aim o s imula e u he esea ch
and in o m s a egic decision-making p ocesses in bo h he public and p i a e
sec o s.
The es o his pape is s uc u ed as ollows. Sec ion 2 p esen s a li e a u e
e iew wi h p e ious indings in he opic o his wo k. Sec ion 3 p oposes a
heo e ical model by which he a iables a e ela ed. Sec ion 4 p esen s he
da abase while sec ion 5 p esen s he empi ical s a egy. Resul s a e discussed
on sec ion 6 and sec ion 7 concludes wi h a discussion abou he alue o ou
indings.
2. PREVIOUS LITERATURE
RESEARCH ON INNOVATION PERFORMANCE
This li e a u e e iew i s ly highligh s he impo ance o in es iga ing in-
no a ion pe o mance and i s de e minan s in a ious con ex s. Unde s and-
ing he ac o s ha con ibu e o success ul inno a ion ou comes is c ucial o
i ms and policymake s alike. By e iewing p e ious s udies on inno a ion
pe o mance, his pape aims o con ibu e o he exis ing body o knowledge
by examining he ole o di e se knowledge in os e ing inno a ion, speci ical-
ly ocusing on e idence om Chilean i ms.
By examining he li e a u e on inno a ion pe o mance, his pape aims o
discuss he impo ance o ex e nal knowledge, owne ship s uc u e, o gani-
za ional p ac ices, sec o al di e ences, cus ome pa icipa ion, and he e ec-
i eness o di e en knowledge sou ces in d i ing inno a ion. Unde s anding
hese ac o s can help i ms and policymake s de elop s a egies and policies
ha p omo e inno a ion and enhance o e all economic pe o mance.
Nume ous s udies ha e ocused on in es iga ing inno a ion pe o mance
and i s de e minan s. C epon, Dugue , and Mai esse (1998) de eloped a model
ha es ablished a amewo k o explo ing he causa ion o inno a ion ou pu
and p oduc i i y g ow h by linking inno a ion su ey a iables. Building on
his model, subsequen esea ch has u he examined he ela ionship be ween
inno a ion su ey a iables and inno a ion ou pu .
The impo ance o ex e nal knowledge o inno a ion has been emphasized
in a ious s udies. So ka and G impe (2010) a gued ha i ms should de elop
s a egies o le e age ex e nal in o ma ion, and he success o his s a egy sig-
ni ican ly in luences inno a ion ou comes. They demons a ed ha combining
in-house R&D in es men s wi h a ma ke -o ien ed sea ch s a egy enhances
he e ec i eness o inno a ion e o s.
89
Di e se knowledge o di e se inno a ion; e idence om Chilean... / Rodol o Lau e bach
Owne ship s uc u e has also been iden i ied as a ac o in luencing inno-
a ion pe o mance. Choi, Lee, and Williams (2011) ound ha i ms wi h o -
eign owne ship ha e a highe p obabili y o success ul inno a ion. Thei s udy,
conduc ed on Chinese i ms, e ealed ha o eign owne ship and a ilia ion
wi h a business g oup s ongly in luence he olume o pa en egis a ions.
This sugges s ha owne ship s uc u e plays a i al ole in de e mining inno-
a ion ou comes.
O ganiza ional p ac ices ha e been ecognized as c ucial ac o s o inno-
a ion success. Mol and Bi kinshaw (2014) highligh ed he signi icance o ce -
ain o ganiza ional p ac ices in os e ing inno a ion. They emphasized he ole
o ex e nal in ol emen in he inno a ion managemen p ocess, which no only
p o ides di ec inpu om ex e nal change agen s bu also b ings p io ex e nal
expe ience as an in e nal agen o change.
Analyzing sec o al di e ences in inno a ion ou comes is also impo an .
Cas ellacci (2008) p esen ed a sec o al axonomy ha in eg a ed manu ac u -
ing and se ice indus ies wi hin a comp ehensi e amewo k. This app oach
unde sco ed he inc easing impo ance o e ical linkages and in e -sec o al
knowledge exchanges be ween hese in e connec ed b anches o he economy.
Božić and Mohnen (2016) conduc ed a quan i a i e analysis using C oa ian
Communi y Inno a ion Su ey da a and ound ha while he e a e some di -
e ences, se ice and manu ac u ing SMEs sha e simila de e minan s o in-
no a ion ac i i ies. Howe e , se ice SMEs ely mo e on acqui ed knowledge
compa ed o hei manu ac u ing coun e pa s.
The ela ionship be ween cus ome pa icipa ion and inno a ion pe o -
mance has been explo ed in se e al s udies. Chang and Taylo (2016) con-
duc ed a me a-analysis ha examined he e ec s o con ex ual ac o s on he
ela ionship be ween cus ome pa icipa ion and new p oduc de elopmen
pe o mance. Thei analysis e ealed ha in ol ing cus ome s in he idea ion
and launch s ages o new p oduc de elopmen imp o es new p oduc inancial
pe o mance di ec ly, as well as indi ec ly h ough accele a ed ime o ma ke .
Howe e , cus ome pa icipa ion in he de elopmen phase slows down ime o
ma ke , leading o a de e io a ion in new p oduc inancial pe o mance.
The s udy by Anzola-Román, Bayona-Sáez, and Ga cía-Ma co (2018) in-
es iga ed he in luence o in e nal and ex e nally sou ced inno a ion p ac ices
on he likelihood o achie ing p oduc and p ocess inno a ions. Thei indings
indica ed posi i e e ec s o in e nal R&D and ex e nally sou ced inno a ion
p ac ices, as well as a posi i e in luence o o ganiza ional inno a ion on he
ealiza ion o echnological inno a ions.
Unde s anding he mos e ec i e sou ces o inno a i e ideas emains a
signi ican challenge in echnological inno a ion managemen . C iscuolo e
al. (2018) examined he e ec i eness o di e en combina ions o knowledge
90 Es udios de Economía, Vol.51 - Nº 1
sou ces o achie ing inno a i e pe o mance. Thei s udy, based on a la ge-
scale sample o UK i ms, e ealed impo an di e ences be ween p oduc and
p ocess inno a ion, wi h b oade knowledge sea ches associa ed wi h he o -
me .
THE MANAGEMENT OF INNOVATION AND FIRM PERFORMANCE
Inno a ion is widely ecognized as a c ucial d i e o i m success, con ib-
u ing o compe i i e ad an age, ma ke g ow h, and long- e m sus ainabili y.
As he business landscape becomes inc easingly dynamic and complex, o ga-
niza ions mus con inuously adap and inno a e o s ay ahead. Consequen ly,
unde s anding i m managemen ac o s ha in luence inno a ion pe o mance
has become a opic o g ea in e es o esea che s and p ac i ione s alike.
Cohen and Le in hal (1990) highligh ed he concep o abso p i e capac-
i y, which e e s o a i m’s abili y o acqui e, assimila e, and u ilize ex e nal
knowledge o os e inno a ion. They emphasized ha p io knowledge and
expe iences signi ican ly in luence a i m’s abso p i e capaci y. This pe spec-
i e unde sco es he impo ance o le e aging di e se knowledge sou ces and
lea ning om ex e nal in o ma ion o enhance inno a ion capabili ies. By ex-
plo ing he ela ionship be ween di e se knowledge and inno a ion ou comes,
aluable insigh s can be gained in o how i ms can e ec i ely ap in o a ange
o knowledge domains.
While much o he exis ing li e a u e has p ima ily ocused on p oduc
and p ocess inno a ion, he e is a g owing ecogni ion o o he dimensions
o inno a ion ha ex end beyond angible ou pu s. These dimensions include
managemen , o ganiza ional, and social inno a ions, which encompass no -
el p ac ices, s uc u es, and echniques ha ad ance o ganiza ional goals.
OECD/Eu os a (2018) p oposed a comp ehensi e amewo k encompassing
hese a ious inno a ion ypes. Acknowledging and explo ing hese di e se
dimensions o inno a ion con ibu e o a mo e comp ehensi e unde s anding
o he inno a ion p ocess and i s impac on i m pe o mance. Chen, Wang,
and Huang (2019) in es iga ed he ela ionship be ween o ganiza ional inno a-
ion and echnological inno a ion capabili ies, explo ing hei impac on i m
pe o mance. Th ough s uc u al equa ion modeling, hei s udy e ealed ha
inno a ion capabili ies pa ially media e he link be ween o ganiza ional inno-
a ion and i m pe o mance.
Fu he mo e, e ec i e inno a ion managemen p ac ices play a i al ole in
ealizing he ull po en ial o inno a ion. Bi kinshaw and Mol (2008) iden i ied
ou key p ocesses—mo i a ion, in en ion, implemen a ion, and heo iza ion
and labeling— ha collec i ely shape managemen inno a ion. By examining
he oles o change agen s wi hin and ou side he o ganiza ion, aluable in-
91
Di e se knowledge o di e se inno a ion; e idence om Chilean... / Rodol o Lau e bach
sigh s can be gained in o how inno a ion managemen p ac ices can be op i-
mized o maximize he bene i s de i ed om inno a ion e o s.
Howe e , despi e he ecognized impo ance o inno a ion and i s mul idi-
mensional na u e, challenges pe sis in ealizing signi ican economic e u ns
om inno a ion. Teece (1986) highligh ed ha p o i s o en acc ue o com-
plemen a y asse owne s, cus ome s, and imi a o s a he han o he o iginal
de elope s o in ellec ual p ope y. This aises impo an ques ions ega ding
he alignmen o inno a ion s a egies wi h app op ia e managemen p ac ices
o ensu e ha i ms cap u e and capi alize on he economic bene i s o hei
inno a i e endea o s.
Gi en he mul i ace ed and ongoing na u e o inno a ion, i is essen ial
o del e in o he li e a u e o gain a comp ehensi e unde s anding o he e-
la ionship be ween di e se knowledge and di e se inno a ion ou comes. By
explo ing he in e play be ween abso p i e capaci y and di e en dimensions
o inno a ion, in he con ex o e ec i e inno a ion managemen p ac ices, his
s udy aims o p o ide e idence on he ela ionship be ween di e se knowledge
and di e se inno a ion ou comes among Chilean i ms. Th ough his in es i-
ga ion, aluable insigh s can be ob ained o in o m i ms’ inno a ion s a egies
and enhance hei abili y o d i e success ul inno a ion ou comes while ealiz-
ing economic e u ns.
INFORMATION SOURCES AND INNOVATION
The s udy o in o ma ion sou ces and hei impac on i m-le el inno a ion
pe o mance is highly mo i a ed by he ecogni ion o inno a ion as a c i i-
cal d i e o i m success. In oday’s compe i i e business en i onmen , i ms
a e cons an ly seeking ways o imp o e hei inno a ion capabili ies and ou -
comes. Unde s anding he ole o in o ma ion sou ces in his p ocess is essen-
ial o i ms aiming o le e age knowledge e ec i ely and achie e sus ainable
inno a ion pe o mance.
P e ious esea ch has shed ligh on he in luence o di e en ypes o in o -
ma ion sou ces on inno a ion. A ani is, Lokshin, Mohnen, and Wö e (2013)
conduc ed a s udy based on panels o Du ch and Swiss inno a ing i ms, ind-
ing ha bo h “buying” and “coope a ing” ha e a posi i e e ec on inno a ion.
Howe e , simul aneous u iliza ion o hese in o ma ion sou ces does no nec-
essa ily lead o highe inno a ion pe o mance. Pejić Bach e al. (2015) em-
phasized he ca aly ic ole o in o ma ion sou ces in inno a ion imp o emen ,
u ilizing CIS da a om C oa ia, F ance, and he Ne he lands. Thei indings in-
dica ed ha in e nal sou ces, cus ome s, supplie s, and uni e si ies a e impo -
an in o ma ion sou ces o bo h in e nal and ex e nal R&D ac i i ies ac oss
he h ee coun ies. In e es ingly, i ms om he Ne he lands exhibi di e en
92 Es udios de Economía, Vol.51 - Nº 1
pa e ns in u ilizing in o ma ion sou ces, elying mo e on compe i o s com-
pa ed o i ms om C oa ia and F ance. Addi ionally, go e nmen in o ma ion
sou ces had a ela i ely smalle impac on i ms’ inno a ion pe o mance.
The dis inc ion be ween in e nal and ex e nal sou ces o in o ma ion has
been explo ed in ela ion o he gene a ion o p oduc and p ocess inno a ion.
Gómez, Salaza , and Va gas (2016) examined he usage o in e nal and ex e -
nal sou ces o in o ma ion by Spanish i ms, including cus ome s, supplie s,
compe i o s, consul an s, and uni e si ies. They ound ha he impo ance o
ex e nal sou ces o in o ma ion a ies depending on he ype o inno a ion
conside ed. Fo p ocess inno a ion, i ms mainly ely on supplie s, while o
p oduc inno a ion, he main con ibu ion comes om cus ome s. Damanpou ,
Sanchez-Hen iquez, and Chiu (2018) in es iga ed he dual ole o in e nal and
ex e nal sou ces o knowledge and in o ma ion in he adop ion o manage ial
inno a ions. Thei indings indica ed ha in e nal implemen a ion ac ions ha e
a s onge e ec han ex e nal implemen a ion ac ions in in luencing inno a-
ion adop ion. Do zel and Faggian (2019) analyzed he ela ionship be ween
ex e nal knowledge sou cing and a ious inno a ion ou comes in u al and u -
ban es ablishmen s in he U.S. Thei esul s sugges ed ha ex e nal knowledge
sou cing speci ically p omo es p oduc , p ocess, and g een inno a ion in U.S.
i ms. They also highligh ed he po en ial impo ance o knowledge sou cing
om non-local o ganiza ions, pa icula ly in suppo ing inno a ion in u al
ma ke s compa ed o u ban ma ke s.
Fu he mo e, he li e a u e has explo ed he e ec s o di e en combina-
ions o knowledge sou ces on inno a ion ou pu . Basi and Medase (2019a)
highligh ed he posi i e link be ween knowledge di e si y and i m-le el in-
no a ion pe o mance, emphasizing he impo ance o knowledge om cus-
ome s in he p i a e and public sec o s, as well as knowledge om compe i-
o s. Basi (2021) ex ended his esea ch by examining he impac o ex e nal
knowledge sou ces on he willingness o small and medium-sized en e p ises
(SMEs) o in oduce o ganiza ional inno a ion, e ealing he g ea e impo -
ance o ex e nal knowledge o small i ms and hei p opensi y o u ilize di-
e se se s o ex e nal knowledge.
By del ing in o he li e a u e on in o ma ion sou ces and inno a ion, i
becomes e iden ha di e se knowledge u iliza ion plays a i al ole in d i ing
i m-le el inno a ion pe o mance. The in e play be ween di e en ypes o
in o ma ion sou ces, whe he o igina ing om paid deals o coope a ion ag ee-
men s, and whe he de i ed om in e nal o ex e nal agen s, o e s aluable
insigh s o i ms aiming o enhance hei inno a ion capabili ies and achie e
supe io inno a ion ou comes. The e o e, his s udy seeks o con ibu e o he
exis ing body o knowledge by examining he ela ionship be ween di e se
knowledge sou ces and di e se inno a ion ou comes wi hin he con ex o
99
Di e se knowledge o di e se inno a ion; e idence om Chilean... / Rodol o Lau e bach
TABLE 2
DESCRIPTION OF VARIABLES
Va iable Name Type Desc ip ion
Any inno a ion Dummy 1 i he i m in oduced any inno a ion in
2019-2020 and 0 o he wise
P oduc inno a ion Dummy
1 i he i m in oduced new o signi i-
can ly imp o ed p oduc o se ice in
2019-2020 and 0 o he wise
P ocess inno a ion Dummy
1 i he i m in oduced new o signi i-
can ly imp o ed ope a ional p ocesses in
2019-2020 and 0 o he wise
Ma ke ing inno a ion Dummy
1 i he i m in oduced ma ke ing inno-
a ion (i.e. signi ican modi ica ion in
design o packaging o goods o se ices)
in 2019-2020 and 0 o he wise
O ganiza ional inno a ion Dummy
1 i he i m in oduced o ganiza ional
inno a ion (i.e. new business p ac ices
o o ganizing p ocedu es) in 2019-2020
and 0 o he wise
Social inno a ion Dummy
1 i he i m in oduced social inno a ion
in 2019-2020 (i.e. sus ainable inno a-
ion) and 0 o he wise
In e nal R&D Dummy 1 i he i m ca ied ou in e nal R&D
ac i i ies
O he Inno a i e In es men s Dummy 1 i he i m ca ied ou in es men s
Sou ce o knowledge om
compe i o s Dummy
1 i he i m ge in o ma ion sou ce o
new ideas in cu en inno a ion p ojec s
om he compe i o s in 2019–2020 and
0 o he wise
Sou ce o knowledge om he
cus ome s Dummy
1 i he i m ge in o ma ion sou ce
o new ideas in cu en inno a ion
p ojec s h ough he cus ome s sec o in
2019–2020 and 0 o he wise
Sou ce o knowledge om he
go e nmen Dummy
1 i he i m ge in o ma ion sou ce
o new ideas om in e ac ion wi h
go e nmen agencies in 2019-2020 and
0 o he wise
G adua e employees Con inuous
s anda dized o 0-1
Numbe o g adua e employees (p o-
essional, mas e o PhD) o he o al
numbe o employees in 2020.
Employmen log The log o he numbe o employees as a
measu e o i m size
High Expo s Dummy
1 i expo s we e highe han
USD$500,000 o e 2019-2020 and 0
o he wise.
Public Funding Dummy
1 i i m ecei ed inno a ion unding
om he public sec o in 2019-2020 and
0 o he wise
Ac
A
c
ii
11
3
,,…Dummy 1 i i m belongs o speci ic sec o and 0
o he wise
Sou ce: Va iables de ined based on da a om he Chilean Inno a ion Su ey 2019-2020.
100 Es udios de Economía, Vol.51 - Nº 1
TABLE 3
CORRELATION MATRIX
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
1. Any Inno a ion 1,00
2. P oduc Inno a ion 0,66 1,00
3. P ocess Inno a ion 0,93 0,51 1,00
4. Ma ke ing Inno a ion 0,50 0,37 0,53 1,00
5. O ganiza ional Inno a ion 0,62 0,36 0,66 0,52 1,00
6. Social Inno a ion 0,32 0,31 0,28 0,22 0,24 1,00
7. R&D Ac i i y 0,52 0,48 0,46 0,30 0,32 0,31 1,00
8. O he Inno a i e In es men s 0,76 0,53 0,72 0,40 0,47 0,29 0,46 1,00
9. Sou ce Compe i o s 0,12 0,12 0,13 0,11 0,10 0,09 0,17 0,14 1,00
10. Sou ce Clien s 0,54 0,54 0,49 0,41 0,37 0,27 0,49 0,59 0,20 1,00
11. Sou ce Go e nmen 0,25 0,23 0,22 0,16 0,18 0,24 0,31 0,26 0,31 0,29 1,00
12. Highly Educa ed 0,12 0,14 0,11 0,07 0,09 0,08 0,12 0,09 0,04 0,10 0,08 1,00
13. Log To al Employees 0,19 0,13 0,17 0,08 0,14 0,09 0,22 0,18 0,07 0,13 0,08 -0,20 1,00
14. Expo s USD$500.000+ 0,12 0,10 0,08 0,06 0,07 0,06 0,19 0,10 0,05 0,10 0,07 0,00 0,33 1,00
15. Public Inn. Funding 0,22 0,22 0,19 0,10 0,15 0,17 0,32 0,23 0,07 0,19 0,24 0,09 0,04 0,05 1,00
Sou ce: Own calcula ions based on he Chilean Inno a ion Su ey 2019-2020.
101
Di e se knowledge o di e se inno a ion; e idence om Chilean... / Rodol o Lau e bach
5. EMPIRICAL STRATEGY
To empi ically es ima e he heo e ical model, ou i s s ep is o examine
he p esence o endogenei y ela ed o a selec i i y p oblem. The li e a u e p o-
ides se e al compelling easons why inno a ion could also in luence R&D,
which ha e been well-documen ed. Mans ield (1969) p esen ed one o he ea -
lies wo ks on his ela ionship, a guing ha success ul inno a ion inc eases
a i m’s echnological oppo uni ies, making u he inno a ion e o s mo e
likely.
Ano he a gumen o he impac o inno a ion on R&D is he di icul-
y i ms may ace in ob aining unding o inno a ion p ojec s om ex e nal
sou ces due o hei inhe en iskiness (Pe e s, 2009). I success ul inno a ions
lead o inc eased p o i abili y and access o ex e nal unding, i ms a e mo e
likely o engage in u he R&D.
Fu he mo e, he ela ionship be ween inno a ion, expo ing, and R&D has
been discussed as a bidi ec ional o ce by Ha is and Mo a (2011). Some
s udies ha e emphasized he pe sis ence o inno a ion and i s posi i e impac
on subsequen R&D in es men . Ge oski e al. (1997) and Male ba and O s-
enigo (1999) ha e also explained he mechanism h ough which inno a ion
in luences R&D.
F om he li e a u e, i can be concluded ha he de e minan s o R&D ex-
pendi u e o an indi idual i m a e no comple ely independen o he i m’s
p obabili y o inno a ing. Inno a ing i ms alloca e esou ces o R&D o
achie e inno a ions, while non-inno a ing i ms may in es in R&D o en-
hance hei abso p i e capaci ies. Addi ionally, he a iables ha explain R&D
may di e depending on whe he he i m is inno a ing o no . Hence, he e is
a selec i i y p oblem.
To add ess he selec i i y p oblem, one pe spec i e is o conside inno a-
ion as an au o-selec ion p ocess. Expec ed R&D in es men depends on he
i m’s inno a ion s a us, making he selec i i y p oblem mo e complex han
a simple sample selec i i y bias. K iaa and Ka ay (2010) sugges ha one
app oach o sol ing his p oblem is o limi obse ed he e ogenei y be ween
i ms while also con olling o unobse ed he e ogenei y. O he esea che s
ha e used an app oach based on Heckman (1979) o add ess selec i i y p ob-
lems in his model. Following Basi and Medase (2019b), his wo k adop s an
ins umen al a iable (IV) bina y ea men model wi h a selec ion equa ion
based on a se o ins umen s as he empi ical me hodology.
The econome ic model aims o s udy he ela ionship be ween i m-le el
inno a ion, human capi al, in e nal R&D ac i i ies, and sou ces o knowledge
lows. Gi en he bina y na u e o he endogenous and ins umen al a iables,
his s udy employs an IV bina y ea men model. The es ima ion me hod is a
102 Es udios de Economía, Vol.51 - Nº 1
wo-s age Heckman bina y ea men model. This empi ical se up allows us
o add ess po en ial endogenei y p oblems. The bina y ea men model used
in his esea ch has been ho oughly explained by Woold idge (2010) and has
been employed by au ho s such as Basi (2021) and Ce ulli (2012). The wo-
s age Heckman bina y ea men model wi h he e ogeneous ea men esponse
helps o add ess he endogenei y issues ha a ise in his con ex , whe e he
ela ionship be ween inno a ion ou pu and pe o mance di e s be ween i ms
in es ing in R&D and hose ha do no .
The speci ica ion o he ins umen al a iable model is as ollows:
(2) ywxwxewe e
x
00
01
0
Whe e we assume ha obse able and unobse able he e ogenei y a e no
he same, so
ee
10
. Following he p inciple o he wo-s age sample se-
lec ion es ima ion o Heckman (1979), we assume ha on a bina y ea men
model we can s ill obse e no mali y o he e o e m. This way we use a
gene al model i s ly speci ying a undamen al eg ession.
(3) yx
ii i
1
Whe e selec ion implies ha he dependen a iable is known unde he
condi ion ha
z
ii
20
Whe e
12 12
001~,,~,, ,NNand Co
. And i we
could assume ha
0
, we could igno e he selec ion p oblem.
The s a egy hen implies he es ima ion o wo equa ions, he main equa-
ion wi h inno a ion ou pu as dependen a iable, and a selec ion equa ion
wi h R&D dummy as a dependen a iable. Inno a ion ou pu is a se o dum-
my a iables ha can desc ibe each ype o inno a ion sepa a ely wi h:
(4) INNxRD
ii
ii
12
&
whe e: xInnAc In In In HCEmp Ac
A
c
iiii ii ii i
n
,,,,,,,,
12
31
, and:
(5) RD xz
ii
ii
&
112
whe e: zHExpPublicFunding
ii i
,,
Whe e InnAc i is a dummy wi h alue 1 i he i m has spen on any o he
o he non-R&D ac i i ies: machine y, knowledge acquisi ion, aining. In i
1
103
Di e se knowledge o di e se inno a ion; e idence om Chilean... / Rodol o Lau e bach
is a dummy indica ing whe he he sou ce o ideas o inno a ion de eloped
wi h in o ma ion om compe i o s, In i
2 is a dummy indica ing whe he he
sou ce o ideas o inno a ion de eloped wi h in o ma ion om cus ome s,
In i
3 is a dummy o whe he he ideas o inno a ions came om go e nmen
sou ces. HCi is human capi al in ensi y measu ed as highly educa ed employ-
ees di ided by o al employees, Empi is he log o he numbe o employees
as a measu e o i m size. Ac
A
c
ii
n1,,… a e economic sec o dummies. Fo
he i s s age equa ion he ins umen ed a iable
RD
i
& is a dummy equal
o 1 i he i m has done R&D in es men . We used wo ins umen s ha a e
s a is ically alid wi h signi ican ly highe co ela ion o he ins umen ed
a iable compa ed o he endogenous a iable1. The ins umen s a e HExp
i
ha is a dummy indica ing whe he expo s we e highe han USD$500,000,
and PublicFundingi, ha is a dummy equal o 1 i he i m ecei ed any kind
o public unding o inno a ion du ing he pe iod 2019-2020, and 0 o he -
wise. The ins umen s we e chosen conside ing ha bo h, access o expo ing
ma ke s and access inno a ion public unding, a e expec ed o ha e a g ea e
impac o e R&D e o s compa ed o inno a ion ou pu s because he la e
esul om a mo e complex knowledge gene a ion p ocesses ha is a ec ed by
inno a ion e o s, in o ma ion sou ces and i ms’ human capi al.
6. MAIN RESULTS
The indings o his pape a e p esen ed in his sec ion. The es ima ion
me hod begins wi h a se o p elimina y bina y p obi eg essions using 5,519
obse a ions. This s ep is aken be o e conside ing any endogenei y p oblems.
Table 4 displays plausible esul s ha align wi h he heo e ical model.
All ypes o inno a ion ou pu s conside ed in he model a e posi i ely and
signi ican ly in luenced by bo h R&D in es men and o he inno a i e in es -
men s. The p opo ion o employees wi h a p o essional i le o highe le el
o educa ion also has a posi i e and signi ican impac on inno a ion ou pu
in all eg essions. Fi m size, measu ed as he loga i hm o o al employees,
consis en ly shows a posi i e pa ame e in all eg essions, al hough i s impac
appea s o be lowe compa ed o he o he a iables. Fu he mo e, i m size
has a signi ican impac on p ocess, o ganiza ional, and social inno a ion, bu
i s signi icance is no obse ed in he case o p oduc and ma ke ing inno a-
ion. This p elimina y esul sugges s ha smalle i ms may ha e he abili y
o achie e hese ypes o inno a ion ou pu wi hou acing clea disad an ages
due o hei size.
1 See he co ela ion de ails on able 3
104 Es udios de Economía, Vol.51 - Nº 1
Based on hese ini ial esul s, he e is e idence ha in o ma ion om com-
pe i o s may ha e e y li le o no impac on all ypes o inno a ion ou pu .
This esul could be due o biases caused by endogenei y p oblems. I could
also be a ibu ed o he ac ha a ela i ely small pe cen age o i ms u i-
lize in o ma ion om compe i o s. Rega ding sou cing inno a ion in o ma ion
om cus ome s, he esul s indica e ha i is an impo an and signi ican a i-
able ha posi i ely a ec s all ypes o inno a ion ou pu s. This inding sug-
ges s ha i ms a ach g ea e impo ance o cus ome eedback, indica ing ha
cus ome -o ien ed i ms a e mo e likely o succeed in hei inno a i e endea -
o s. This obse a ion aligns wi h ecen managemen li e a u e ha emphasiz-
es he impo ance o ocusing business models on cus ome s. The eg essions
also e eal ha sou cing in o ma ion om go e nmen agencies is associa ed
wi h speci ic ypes o inno a ion ou pu s. The esul s p opose ha go e nmen
in o ma ion has a signi ican impac only in he case o social inno a ions.
Howe e , ollowing ou empi ical s a egy and in line wi h p e ious li e a-
u e2 on he es ima ion o inno a ion de e minan s, Table 5 examines he same
ques ion using a Heckman wo-s age bina y ins umen al a iable ea men
model. This es ima ion me hod has been employed in o he pape s, including
Basi and Medase (2019a, 2019b). The bina y selec ion a iable is R&D ac-
i i y, and we use dummy a iables as ins umen s o indica e whe he expo s
exceed USD$500,000 and whe he any public unding o inno a ion was e-
cei ed. Bo h ins umen s exhibi conside ably highe co ela ion wi h he R&D
ac i i y dummy compa ed o inno a ion ou pu a iables. Like on he p e ious
eg essions, con ol a iables o economic sec o a e included bu no epo ed
in he able.
Fi s s age esul s a e epo ed on he i s column. No e ha he i s -s age
equa ion is he same o all six inno a ion equa ions. We ind ha bo h ins u-
men s, expo s and public unding ha e a posi i e and signi ican impac on
R&D e o s a he i m le el.
The esul s o he ollowing columns sugges ha inno a i e in es men s
o he han R&D is also a signi ican de e minan o all ypes o inno a ion. The
p e ious inding indica ing a low and insigni ican impac o in o ma ion om
compe i o s on inno a ion ou pu is also suppo ed by hese esul s. In o ma-
ion om clien s as a sou ce o inno a ion ideas has a posi i e and signi ican
e ec in all cases excep o social inno a ion, whe e go e nmen agencies
eme ge as he only impo an and signi ican in o ma ion sou ce. Go e nmen
in o ma ion also has a posi i e and signi ican impac on p ocess and o gani-
za ional inno a ion, albei wi h smalle pa ame e sizes. The posi i e e ec o
he p opo ion o highly educa ed employees on di e en ypes o inno a ion
2 The a gumen o why R&D should be conside ed endogenous on an Inno a ion equa-
ion is pa icula ly well explaiden in he wo k by C epon, Dugue and Mai esse (1998).
105
Di e se knowledge o di e se inno a ion; e idence om Chilean... / Rodol o Lau e bach
ou pu is obse ed, al hough he e ec is smalle han wha was obse ed in
he p e ious able and is no signi ican in he case o social inno a ion. The
esul s also demons a e ha he loga i hm o he o al numbe o employees
has a posi i e and signi ican impac on inno a ion ou pu s, excep in he case
o p oduc and ma ke ing inno a ion, which is consis en wi h he indings
om he p e ious able.
Fo each o he second s age equa ions, he wo-s age Heckman model es-
ima es ho (ac ually, he in e se hype bolic angen o ho) ha ep esen s
he co ela ion o he esiduals in he wo equa ions. Addi ionally, i p esen s
he es ima ion o sigma (ac ually, he log o sigma) which ep esen s he s an-
da d e o o he esiduals o he second s age equa ion. Lambda is ho*sigma,
which is ound o be signi ican on all bu one o he equa ions which sugges s
ha he es ima ion o R&D in he i s equa ion is ele an o he es ima ion
o he second s age equa ions o all kinds o inno a ion ou pu excep o gani-
za ional inno a ion.
These esul s a e ele an because, based on a p e iously alida ed empi -
ical s a egy3 ha akes endogenei y in o accoun , hey explain he impo ance
o di e en in o ma ion sou ces o he se e al dis inc ypes o inno a ion
ou pu s. The esul s show some simila i y wi h p e ious wo ks4 ega ding he
impo ance o cus ome in o ma ion o inno a ion ou pu bu also di e ind-
ing ha o he case Chilean i ms he impo ance o in o ma ion om compe -
i o s is no an impo an de e minan o inno a ion ou pu . I could be he case
ha his esul is obse ed because Chilean i ms ha e a e y low p obabili y
o sou cing inno a ion in o ma ion om compe i o s, and hence he e is no
enough a ia ion o ind a signi ican pa ame e . In ac , only 0.4% o i ms
decla ed o ha e used in o ma ion om compe i o s as a sou ce o inno a ion
ideas. Addi ionally, i could also be he case ha he low use o compe i o
in o ma ion is he esul o low us o highe le els o sec ecy among indus-
y-le el compe i o s. In any case, his esul calls o u he esea ch ha can
dig in o indus y-le el in o ma ion lows o explain his low equency and low
impac i m compe i o ela ion.
Bu e en hough in o ma ion om compe i o s has no p o en o be el-
e an o inno a ion ou pu , we ound ha inno a ion ou pu among Chilean
i ms is d i en o a la ge ex en by ma ke o ien a ion, pa icula ly by sou cing
in o ma ion om cus ome s. In he line wi h he indings by Anzola-Román
(2018), his esul is impo an om a manage ial poin o iew because i
shows ha conside ing cus ome da a is an impo an d i e o inno a ion suc-
cess. Addi ionally, om he public policy pe spec i e, his esul implies he
oppo uni y o de eloping public ins umen s o p omo e cus ome - i m in e -
3 Heckman (1979)
4 Basi and Medase (2019a, 2019b).
106 Es udios de Economía, Vol.51 - Nº 1
ac ions such as expe imen al ai s o a ge ed consume su eys.
Bu he mos ele an and no el esul ound on his wo k is ela ed o he
es ima ion o social inno a ion de e minan s. The wo k by To ia e al (2020)
discussed how social inno a ion in e plays wi h en ep eneu ship in public and
p i a e ins i u ions. Social inno a ions imply he achie emen o esul s ha
bene i socially ulne able g oups o he en i omen , i should be inancially
sus ainable, and i unc ions based on he use o new app oaches and ideas o
sol e a pa icula social p oblem.
We ind ha in e ms o in o ma ion sou ces, social inno a ion is mainly
d i en by in o ma ion om go e nmen ins i u ions, while cus ome and com-
pe i o sou ces a e no ele an when he ull model is es ima ed. This esul ,
i con i med by u he esea ch, could ha e impo an public policy impli-
ca ions. The wo k by Mulgan (2007) discussed ha social inno a ion o en
in ol es uni e si ies, go e nmen agencies and p i a e companies wo king
oge he . He also showed ha social inno a ion is mo e ela ed o he combi-
na ion o knowledge om di e en ac o s a he han he ad ancemen o new
echnologies a he indi idual o ganiza ional le el. Pa icula ly, conside ing
ha social inno a ion has a la ge posi i e ex e nali y componen , ou esul s
sugges ha public unding ins umen s o p omo e collabo a ion wi h go e n-
men ins i u ions could help p omo e inno a ions ha ha e he highes social
alue.
107
Di e se knowledge o di e se inno a ion; e idence om Chilean... / Rodol o Lau e bach
TABLE 4
PRELIMINARY BINARY PROBIT REGRESSION
Any
Inno a ion
P oduc
Inno a ion
P ocess
Inno a ion
Ma ke ing
Inno a ion
O ganiza ional
Inno a ion
Social
Inno a ion
R&D Ac i i y 1.313 0.863 0.798 0.412 0.406 0.666
(0.106)** (0.081)** (0.086)** (0.089)** (0.080)** (0.100)**
Inno a ion Ac i i y 2.397 1.183 2.125 1.062 1.240 0.806
(0.078)** (0.071)** (0.069)** (0.080)** (0.069)** (0.097)**
Sou ce Compe i o s -0.403 -0.381 0.216 0.089 -0.111 -0.349
(0.495) (0.283) (0.405) (0.282) (0.268) (0.301)
Sou ce Clien s 0.586 0.868 0.296 0.676 0.378 0.296
(0.120)** (0.084)** (0.094)** (0.090)** (0.083)** (0.105)**
Sou ce Go e nmen 0.152 0.048 -0.024 0.105 0.151 0.573
(0.214) (0.143) (0.163) (0.149) (0.137) (0.146)**
Highly Educa ed 0.412 0.439 0.390 0.307 0.383 0.349
(0.095)** (0.109)** (0.092)** (0.118)** (0.101)** (0.147)*
To al Employees 0.062 0.029 0.059 0.010 0.077 0.066
(0.017)** (0.019) (0.016)** (0.021) (0.018)** (0.025)**
Cons an -1.760 -2.097 -1.768 -1.818 -2.244 -2.093
(0.171)** (0.197)** (0.168)** (0.183)** (0.195)** (0.203)**
N 5519 5519 5519 5519 5519 5519
* p<0.05; ** p<0.01
Sou ce: Own calcula ions based on he Chilean Inno a ion Su ey 2019-2020.
108 Es udios de Economía, Vol.51 - Nº 1
TABLE 5
BINARY TWO-STAGE HECKMAN INSTRUMENTAL VARIABLE REGRESSION
1s S age
R&D Ac i i y
Any
Inno a ion
P oduc
Inno a ion
P ocess
Inno a ion
Ma ke ing
Inno a ion
O ganiza ional
Inno a ion
Social
Inno a ion
Inno a ion Ac i i y 0.941 0.758 0.212 0.731 0.165 0.279 0.048
(0.074)** (0.014)** (0.013)** (0.015)** (0.012)** (0.014)** (0.009)**
Sou ce Compe i o s 0.117 -0.054 -0.084 0.068 0.060 0.016 -0.037
(0.323) (0.059) (0.051) (0.061) (0.046) (0.054) (0.035)
Sou ce Clien s 0.997 0.250 0.271 0.218 0.230 0.129 0.015
(0.083)** (0.021)** (0.019)** (0.022)** (0.017)** (0.020)** (0.013)
Sou ce Go e nmen 0.575 0.143 0.012 0.113 0.044 0.058 0.138
(0.145)** (0.031)** (0.027) (0.032)** (0.024) (0.029)* (0.019)**
Highly Educa ed 0.792 0.093 0.040 0.091 0.037 0.043 0.008
(0.123)** (0.014)** (0.012)** (0.014)** (0.011)** (0.013)** (0.008)
To al Employees 0.203 0.019 -0.000 0.019 0.003 0.010 0.000
(0.021)** (0.003)** (0.002) (0.003)** (0.002) (0.002)** (0.002)
Expo s USD$500.000+ 0.348
(0.098)**
Public Inn. Funding 1.080
(0.116)**
R&D Ac i i y -0.177 0.295 -0.272 -0.041 0.051 0.198
(0.041)** (0.039)** (0.043)** (0.035) (0.042) (0.027)**
Cons an -3.162 0.002 0.019 -0.003 0.044 -0.017 0.052
(0.239)** (0.024) (0.021) (0.026) (0.019)* (0.023) (0.015)**
5519 5519 5519 5519 5519 5519 5519
lambda 0.265 0.022 0.277 0.063 0.026 -0.055
(0.022)** (0.021)** (0.023)** (0.019)** (0.023) (0.015)**
Rho 0.962 -0.096 0.962 0.281 0.054 -0.395
Sigma 0.275 0.236 0.287 0.215 0.252 0.165
* p<0.05; ** p<0.01
Sou ce: Own calcula ions based on he Chilean Inno a ion Su ey 2019-2020.
115
Di e se knowledge o di e se inno a ion; e idence om Chilean... / Rodol o Lau e bach
Teece, D. J. (1986). “P o i ing om echnological inno a ion: Implica ions o
in eg a ion, collabo a ion, licensing and public policy.” Resea ch Policy,
15(6), 285-305.
To ia, E manno C., Flo ence Dega e and Simone Poled ini (2020) Why a e
social en e p ises good candida es o social inno a ion? Looking o
pe sonal and ins i u ional d i e s o inno a ion. © 2020 The Au ho s
Annals o Public and Coope a i e Economics © 2020 EMF
Tsai, K.H. (2009), “Collabo a i e ne wo ks and p oduc inno a ion
pe o mance: owa d a con ingency pe spec i e”, Resea ch Policy, Vol.
38 No. 5, pp. 765-778.
Vega-Ju ado, J., Gu ié ez-G acia, A. and Fe nández-de-Lucio, I. (2009),
“Does ex e nal knowledge sou cing ma e o inno a ion? E idence
om he Spanish manu ac u ing indus y”, Indus ial and Co po a e
Change, Vol. 18 No. 4, pp. 637-670.
