Dua e, Ma celo Pe ei a; Ca alho, Fe nando Manuel Pe ei a de Oli ei a
A icle
How digi al ans o ma ion shapes Eu opean
union coun ies' na ional sys ems o inno a ion: A
con igu a ional mode a ion app oach
Jou nal o Inno a ion & Knowledge (JIK)
P o ided in Coope a ion wi h:
Else ie
Sugges ed Ci a ion: Dua e, Ma celo Pe ei a; Ca alho, Fe nando Manuel Pe ei a de Oli ei a (2024) :
How digi al ans o ma ion shapes Eu opean union coun ies' na ional sys ems o inno a ion: A
con igu a ional mode a ion app oach, Jou nal o Inno a ion & Knowledge (JIK), ISSN 2444-569X,
Else ie , Ams e dam, Vol. 9, Iss. 4, pp. 1-10,
h ps://doi.o g/10.1016/j.jik.2024.100578
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/327480
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-nd/4.0/
How digi al ans o ma ion shapes Eu opean union coun ies’na ional
sys ems o inno a ion: A configu a ional mode a ion app oach
Ma celo Pe ei a Dua e*, Fe nando Manuel Pe ei a de Oli ei a Ca alho
Uni e si y o Coimb a, CeBER, Facul y o Economics, A . D . Dias da Sil a 165, Coimb a 3004-512, Po ugal
ARTICLE INFO
A icle His o y:
Recei ed 10 July 2024
Accep ed 16 Sep embe 2024
A ailable online 26 Sep embe 2024
ABSTRACT
Digi al ans o ma ion is changing he inno a ion landscape, no only o fi ms bu also a he mac o le el.
This s udy aims o analyse how configu a ions o Eu opean Union coun ies’inno a ion inpu s ha lead o
inno a ion ou pu s a e a ec ed by he le els o na ional digi al ans o ma ion. A no el configu a ional
mode a ion app oach is adop ed o assess he impac o digi al ans o ma ion on he 27 EU coun ies’
na ional sys ems o inno a ion (NSI). Resul s showed ha he p esence o all inno a ion inpu s is a su ficien
configu a ion o he p esence o high inno a ion ou pu s, while h ee di e en configu a ions we e ound o
be su ficien o i s absence. The configu a ional mode a ion analysis e ealed ha digi al ans o ma ion
e ec i ely mode a es he a angemen o inno a ion inpu s leading o he p esence o high inno a ion ou -
pu s, whe e he human capi al and esea ch pilla loses i s impo ance. The esul s o his s udy can be unde -
s ood as benchma ks, agains which less inno a i e coun ies may compa e and de elop hei NSI. They also
shed ligh on he specific configu a ions ollowed by each EU coun y in achie ing inno a ion ou pu s, which
could be an inpu o Union-based policies di ec ed a he de elopmen o conc e e inno a ion pilla s in spe-
cific Membe S a es. Fu he mo e, findings enligh en in e na ional business manage s abou he s eng hs
and weaknesses o EU coun ies’NSI, which may aid hem in hei in e na ionalisa ion decisions.
© 2024 The Au ho s. Published by Else ie España, S.L.U. on behal o Jou nal o Inno a ion & Knowledge. This
is an open access a icle unde he CC BY-NC-ND license
(h p://c ea i ecommons.o g/licenses/by-nc-nd/4.0/)
Keywo ds:
Digi al ans o ma ion
Na ional sys ems o inno a ion
Inno a ion inpu s
Inno a ion ou pu s
Configu a ional mode a ion
JEL classifica ion:
O32
O33
O52
C19
In oduc ion
Digi al ans o ma ion is shaping he inno a ion landscape and
he configu a ions o coun ies’inno a ion sys ems (Pano i e al.,
2021). The phenomenon has ga he ed subs an ial a en ion in he
fields o in o ma ion sys ems (Vial, 2019), managemen (Chen e al.,
2024;Hanel e al., 2021), inno a ion s udies (Appio e al., 2024;
Nambisan e al., 2019), and mul idisciplina y esea ch (Ve hoe e al.,
2021), o i has a significan po en ial o in oduce inno a ion and
enhance he compe i i eness o fi ms (Fleu y e al., 2024;Nambisan
e al., 2019;Zhou e al., 2023). Recen ly, howe e , schola s began
paying a en ion o he impac s o digi al ans o ma ion on le els
o he han he fi m, p oposing ha ans o ma ions a e being el a
he ins i u ional and socie al le els (D˛ab owska e al., 2022;Hinings
e al., 2018). Fo ins ance, e-go e nmen is inc easingly a eali y o
mul iple se ices p o ided by cen al o local adminis a ions (Khan
& Pa k, 2013;Rey-Mo eno e al., 2018), ans o ming he way ci izens
engage in public ini ia i es.
The li e a u e on na ional sys ems o inno a ion (NSI) a gues ha
inno a ion in fi ms does no happen in isola ion, bu a he om he
in e ac ions wi h o he ac o s (e.g., supplie s, cus ome s, compe i o s,
indus y egula o s, uni e si ies, go e nmen ) and wi hin he bound-
a ies o he na ional ins i u ional con ex (Edquis , 2006;Lund all,
2007). This a gumen implies ha inno a ion is dependen no only
on indi idual e o s by fi ms and o he ac o s, bu also on he na u e
o in e ac ions ha a e es ablished be ween hem, as well as wi h he
exis ing ins i u ional se ing. Thus, changes in he way esea ch and
de elopmen (R&D) is conduc ed and in e ac ions wi hin he sys em
a e es ablished a e likely o cause he a e and di ec ion o inno a ion
o shi as well.
Digi al ans o ma ion in his con ex en ails a es uc u ing o
inno a ion p ocesses, bo h wi hin and be ween ac o s in he sys em
(Cia li e al., 2021). Empi ically, i has been ecognised ha he in e-
g a ion o digi al echnologies a ec s he inno a ion p ocesses o
fi ms (Escoz Ba agan e al., 2024;Yu e al., 2024;Zhou e al., 2023),
ye mac o-le el analyses o he influence o digi al ans o ma ion on
NSI a e s ill sca ce. Fo ins ance, Galindo-Ma ín e al. (2023) ound
ha combina ions o digi al ans o ma ion le els, in es men in al-
en , and en u e capi al a ailabili y we e linked o coun ies’en e-
p eneu ship le els, highligh ing he impo ance o a a ou able
en i onmen in which ac o s may benefi om digi al echnologies.
Assuming a di e en pe spec i e, Ionescu e al. (2022) disco e ed
ha ins i u ional and business sophis ica ion d i e s posi i ely
* Co esponding au ho .
E-mail add ess: [email p o ec ed] (M.P. Dua e).
h ps://doi.o g/10.1016/j.jik.2024.100578
2444-569X/© 2024 The Au ho s. Published by Else ie España, S.L.U. on behal o Jou nal o Inno a ion & Knowledge. This is an open access a icle unde he CC BY-NC-ND license
(h p://c ea i ecommons.o g/licenses/by-nc-nd/4.0/)
Jou nal o Inno a ion & Knowledge 9 (2024) 100578
Jou nal o Inno a ion
&Knowledge
h ps://www.jou nals.else ie .com/jou nal-o -inno a ion-and-knowledge
influence he le els o digi al ans o ma ion o Eu opean coun ies.
Con e sely, R^
ındașu e al. (2023) ound a posi i e impac o digi al
adop ion le els on na ional inno a ion ou pu s. Taken oge he , hese
s udies sugges he exis ence o in e ac ions be ween digi al ans o -
ma ion and NSI, bu he way such in e ac ions a e es ablished s ill
eludes ou g asp. This leads us o he ollowing esea ch ques ion:
Conside ing ha coun ies’NSI in ol e an in e play among he a i-
ous inno a ion inpu s ha lead o inno a ion ou pu s, how does digi-
al ans o ma ion shape such a ela ionship?
Consequen ly, he p esen s udy aims o assess how configu a-
ions o coun ies’inno a ion inpu s ha lead o inno a ion ou pu s
a e a ec ed by he le els o na ional digi al ans o ma ion. D awing
on a configu a ional pe spec i e o NSI and digi al ans o ma ion li -
e a u e, we a gue ha he abili y o coun ies o ansla e inno a ion
inpu s in o inno a ion ou pu s is con ingen on he le el o digi al
ans o ma ion. We adop a no el, wo-s age uzzy-se Quali a i e
Compa a i e Analysis ( sQCA) app oach (Ma e al., 2024) o assess
how Eu opean Union (EU) coun ies’digi al ans o ma ion mode -
a es he configu a ional ela ionship be ween inno a ion inpu s and
inno a ion ou pu s.
The emainde o his pape is s uc u ed as ollows. The nex sec-
ion (Sec ion 2) e iews he li e a u e and ad ances he p oposi ions
o his s udy. Then, he me hodology is desc ibed in Sec ion 3 along
wi h a de ailed explana ion o he wo-s age sQCA app oach. In Sec-
ion 4 esul s a e p esen ed and discussed in Sec ion 5. Las ly, Sec ion
6 concludes and highligh s he s udy’s con ibu ions, implica ions,
limi a ions and a enues o u u e esea ch.
Theo e ical backg ound
Digi al ans o ma ion can be defined as “a socioeconomic change
ac oss indi iduals, o ganiza ions, ecosys ems, and socie ies ha is
shaped by he adop ion and u iliza ion o digi al echnologies”
(D˛ab owska e al., 2022, p. 932). Acco ding o Ve hoe e al. (2021),
digi al ans o ma ion has undamen ally changed he beha iou and
expec a ions o consume s, exe ed p essu e on adi ional busi-
nesses, and dis up ed mul iple ma ke s. F om he poin o iew o he
fi m, in eg a ion o digi al echnologies has a significan impac on
inno a ion p ocesses. Fo ins ance, Escoz Ba agan e al. (2024) ound
ha an ad anced s age o digi al ans o ma ion inc eased he likeli-
hood o adical inno a ions in SMEs. Likewise, Chen e al. (2024)
epo ed a significan impac o digi al ans o ma ion on he inno a-
ion capabili ies o fi ms. These esul s imply ha fi ms’in eg a ion
o digi al echnologies no only acili a es he de elopmen o inno a-
ion bu also enhances fi ms’capabili ies o p oduce such inno a-
ions. E en hough i has been ecognised ha he adop ion o digi al
echnologies a ec s fi ms’inno a ion p ocesses (Escoz Ba agan e
al., 2024;Nambisan e al., 2019;Zhou e al., 2023), conside ably less
a en ion has been gi en o i s impac s on he ans o ma ion o inno-
a ion sys ems (D˛ab owska e al., 2022;Reischaue , 2018).
The ela ionship be ween inno a ion inpu s and ou pu s
D awing om he NSI app oach, inno a ion sys ems a e defined
as “all impo an economic, social, poli ical, o ganiza ional, ins i u-
ional, and o he ac o s ha influence he de elopmen , di usion,
and use o inno a ions”(Edquis , 2006, p.182). This defini ion ecog-
nises ha coun ies need o possess pa icula combina ions o ac-
o s o inno a ion o occu , which migh ha e di e en deg ees o
impo ance o di e en coun ies’inno a ion ou pu s. We ex end
Edquis ’s (2006) defini ion by assuming ha he di e en ac o s can
be conside ed inno a ion inpu s, ep esen ing “in es men s”in NSI;
whe eas inno a ion ou pu s can be unde s ood as he immedia e
esul s o such “in es men s”, in e ms o he c ea ion, di usion, and
use o inno a ions. Wi h his ex ension, we in end o d aw he defini-
ion o inno a ion sys ems close o he inno a ion inpu -ou pu
amewo k (Godin, 2007), so ha meaning ul ela ionships can be
es ablished be ween inno a ion inpu s and he co esponding
esul s. We do, howe e , e ain om es ablishing he link be ween
NSI and socie al-le el ou comes, such as economic g ow h, income
inequali ies, o o he s since he ocus o his s udy lies on na ional
inno a ion p ocesses.
The e is a ich empi ical li e a u e dealing wi h he de e minan s
o inno a ion ou pu s, whe e a pa icula emphasis is gi en o he
configu a ional na u e o such ela ionships (C espo & C espo, 2016;
Ding, 2022;Dua e & Ca alho, 2024a;Khedhaou ia & Thu ik, 2017;
Tekic & Tekic, 2021;Wang e al., 2021;Yu & Hua ng, 2023). This
s and o esea ch mainly concludes ha no inno a ion inpu by i sel
is su ficien o achie ing high le els o inno a ion. Ra he , i s e ec s
on inno a ion ou pu s happen when hey combine in pa hs, o con-
figu a ions, ha become su ficien o high inno a ion ou pu s o
occu , which is exemplified by he Venn diag am in Fig. 1 and he
espec i e a ow conducing o inno a ion ou pu s.
Howe e , he e is a lack o consensus abou he specific configu a-
ions o inno a ion inpu s ha a e mo e a ou able o inno a ion o
occu (C espo & C espo, 2016;Dua e & Ca alho, 2024a;Wang e al.,
2021), which highligh s he di e se ways in which coun ies can suc-
cess ully de elop hei NSI (Can well, 2006). Hence, we ad ance wi h
he ollowing p oposi ion.
P oposi ion 1: Se e al equifinal combina ions o inno a ion inpu s
exis ha a e su ficien o high inno a ion ou pu s.
The mode a ing ole o digi al ans o ma ion
Empi ical esea ch dealing wi h he in e play be ween digi al
ans o ma ion and NSI is a he spa se, ye e idence o his in e play
exis s. Fo example, R^
ındașu e al. (2023) examined he impac o EU
coun ies’le els o digi alisa ion on inno a ion ou pu s. The au ho s
ound ha he digi al ans o ma ion o EU coun ies posi i ely influ-
ences hei inno a ion ou pu s, albei wi h a mode a e e ec size and
pa ially media ed by skills adequacy. This finding links wi h he EU
Commission’s expec a ion o digi al and inno a ion de elopmen o
i s membe -s a es (Decision (EU) 2022/2481), emphasising he
impo ance o digi al ans o ma ion o inno a ion. Ano he exam-
ple is ha o Ionescu e al. (2022), who s udied how digi alisa ion le -
els a e a ec ed by a ious inno a ion inpu s. Pa icula ly, he au ho s
ound ha he ins i u ional amewo k and he sophis ica ion o busi-
nesses a e he bes p edic o s o digi al ans o ma ion, sugges ing
ha co po a e adop ion o digi al echnologies migh benefi om
well-in eg a ed ins i u ions ac oss se e al social domains.
Taken oge he , hese s udies sugges ha digi al ans o ma ion
can conduce o highe inno a ion ou pu s and ha inno a ion inpu s
can influence digi al ans o ma ion, sugges ing an in e ac ion
be ween digi al ans o ma ion and inno a ion inpu s leading o
inno a ion ou pu s. Ye , li le is known abou how digi al ans o ma-
ion may shape coun ies’NSI, as p oposed in Fig. 1 wi h he mode -
a ing e ec o digi al ans o ma ion on he ela ionship be ween
configu a ions o inno a ion inpu s and inno a ion ou pu s.
Gi en ha di e en coun ies ely on di e en pa hs, o combina-
ions o inno a ion inpu s o achie e inno a ion ou pu s (C espo &
C espo, 2016;Hua ng & Yu, 2022;Khedhaou ia & Thu ik, 2017), digi-
al ans o ma ion is likely o wo k as a con ingency ac o a ec ing
how coun ies de elop hei pa hs o highe inno a ion ou pu s,
leading us o he ollowing p oposi ion.
P oposi ion 2: Combina ions o inno a ion inpu s leading o high
inno a ion ou pu s su e a mode a ing e ec o digi al ans o ma-
ion.
Me hods
The sQCA me hod is based on se heo y (Ragin, 2008) since many
a gumen s in he social sciences can be ou lined in e ms o
M.P. Dua e and F.M.P.d.O. Ca alho Jou nal o Inno a ion & Knowledge 9 (2024) 100578
2
ela ionships be ween se s (Schneide & Wagemann, 2012). Fo
ins ance, i can be a gued ha digi al ans o ma ion le els a e
closely ela ed o coun ies’inno a ion; his can be amed as almos
all coun ies wi h high le els o digi al ans o ma ion a e inno a i e.
In e ms o se ela ions, we could a gue ha he se o highly digi-
ally- ans o med coun ies is mos ly con ained in he se o inno a-
i e coun ies, and he ac ha he e a e coun ies wi h low le els o
digi al ans o ma ion ha a e also inno a i e does no unde mine
he claim, o he a gumen claims ha digi al ans o ma ion is su fi-
cien o inno a ion.
Rela ionships be ween se s a e analysed in e ms o necessi y and
su ficiency (Schneide & Wagemann, 2012). A condi ion is conside ed
necessa y o he ou come se i i is ully con ained in he condi ion
se (Dul, 2016). Rega ding su ficiency, a condi ion is conside ed su fi-
cien o he ou come i i s se is con ained in he ou come se (Ragin,
2008). Boolean logic is used o analyse i combina ions o condi ions
a e su ficien o an ou come.
FsQCA in ol es se e al s eps. A e da a collec ion and cleaning,
condi ions a e calib a ed, which allows o es ablish he deg ee o
membe ship o each case in each condi ion: ully in, ully ou , and
nei he ully in no ully ou (Ragin, 2008). Fuzzy-se calib a ion
en ails he assignmen o a alue be ween 0 and 1 o ep esen he
deg ee o membe ship in each condi ion, which is ypically based on
h ee ancho poin s: ull membe ship (1), ull non-membe ship (0),
and he poin o maximum ambigui y (0.5).
A e calib a ion, a u h able is cons uc ed depic ing all possible
combina ions o condi ions ha can be su ficien o he ou come
(Schneide & Wagemann, 2012). The esul ing ma ix lays down he
se o condi ions (in columns) and e e y possible combina ion o con-
di ions in ows. Condi ions can be ei he p esen o absen , esul ing
in a o al o 2
k
ows, whe e k is he numbe o causal condi ions in
he model. Due o a phenomenon called limi ed di e si y, no all
ows ha e empi ical ep esen a ion (logical emainde s) (Schneide
& Wagemann, 2012).
Two decisions mus be made once he u h able is cons uc ed:
wha is he equency h eshold, and which combina ions a e consid-
e ed su ficien o he ou come. The fi s decision is ela ed o he el-
e ance o causal combina ions. By se ing a h eshold o 1, only
empi ically obse ed combina ions a e conside ed in he analysis,
which is a easonable h eshold o small-n samples (Ragin, 2008).
Fo he second decision, consis ency sco es and p opo ional educ-
ion in inconsis ency (PRI) alues allow he esea che o dis inguish
combina ions o condi ions ha a e subse s o he ou come om
hose ha a e no (Schneide & Wagemann, 2012). A consis ency
below 0.75 is conside ed o indica e subs an ial inconsis ency (Ragin,
2008), and ecommenda ions indica e highe alues (G eckhame e
al., 2018;Pappas & Woodside, 2021). Simul aneously, PRI alues
below 0.5 indica e subs an ial inconsis ency, o he same configu a-
ion could be su ficien o bo h he p esence and he absence o he
ou come (G eckhame e al., 2018). Hence, ecommenda ions indi-
ca e ha he PRI alues should be high and close o he consis ency
h eshold (Pappas & Woodside, 2021).
The u h ables a e analysed h ough Boolean algeb a, which log-
ically minimises he su ficien configu a ions (see Schneide and
Wagemann (2012), chap e 4 o a de ailed explana ion). The analysis
e u ns h ee solu ions (Ragin, 2008): a complex solu ion ha makes
no simpli ying assump ions, a pa simonious solu ion ha makes all
simpli ying assump ions, and an in e media e solu ion ha makes
only some simpli ying assump ions. Resul s usually ollow om he
in e media e solu ion, since he complex solu ion ends o be oo
complex o be in e p e ed in a meaning ul way. Howe e , pa simoni-
ous and in e media e solu ions a e used simul aneously o iden i y
co e and pe iphe al condi ions (Fiss, 2011).
The sQCA me hod desc ibed abo e allows us o explici ly model
complex causali y (Misangyi e al., 2017) by conside ing ha : he
e ec o a single condi ion only exis s in combina ion wi h o he con-
di ions (conjunc u al causa ion), di e en combina ions o condi ions
can lead o he same ou come (equifinali y), and he combina ions o
condi ions explaining he p esence o an ou come a e no necessa ily
he exac opposi e o hose explaining i s absence (causal asymme y)
(Schneide & Wagemann, 2012).
We p esen nex he sample and da a used in his s udy (Sec ion
3.1) and he conside a ions in ol ing he calib a ion o condi ions
(Sec ion 3.2). Since his s udy in ends o unco e he configu a ional
mode a ing ole o digi al ans o ma ion on he ela ionship
be ween combina ions o inno a ion inpu s and inno a ion ou pu s,
he wo-s age sQCA app oach (Ma e al., 2024) and i s equi emen s
a e desc ibed in Sec ion 3.3.
Sample and da a
Ou sample included all 27 EU Membe S a es. Al hough EU Mem-
be S a es’le els o digi alisa ion and inno a ion a e somewha
homogenous compa ed wi h coun ies in o he egions o he wo ld,
he e a e wide dispa i ies be ween hem, as Sou he n and Eas e n
coun ies s ill lag behind mos Cen al and No dic EU coun ies. Ne -
e heless, inno a ion and digi al ans o ma ion ha e ecen ly
become cen al o he Eu opean Commission and Membe S a es
(Decision (EU) 2022/2481).
Seconda y da a used in his s udy was ga he ed om he Eu o-
pean Commission’s Digi al Decade DESI da abase and WIPO’s Global
Inno a ion Index (GII). The e a e h ee easons why we used GII da a
Fig. 1. Concep ual model.
M.P. Dua e and F.M.P.d.O. Ca alho Jou nal o Inno a ion & Knowledge 9 (2024) 100578
3
o p oxy o inno a ion: 1) I is conside ed a alid indica o o
na ional inno a ion ac i i ies (Gaul & Soe e, 2022), 2) I explici ly
dis inguishes inno a ion inpu s om inno a ion ou pu s, and 3) I s
s uc u e app oxima es he elemen s ha define NSI (Alco a & Pe es,
1998). Digi al Decade DESI da a e e s o he digi al ans o ma ion o
businesses. Da a ga he ed om bo h da abases epo o 2023, he
la es yea a ailable.
Calib a ion
Calib a ion in se - heo e ic me hods in ol es es ablishing he
deg ee o membe ship each case has in each condi ion by defining
h ee ancho poin s (Ragin, 2008): ull membe ship (1), c oss-o e
poin (0.5), and ull non-membe ship (0). Good p ac ices in sQCA
(Fiss, 2011;Misangyi e al., 2017) ecommend ha h esholds o
ancho poin s be se on heo e ical g ounds whe e a ailable. How-
e e , when heo e ical h eshold alues a e no es ablished, calib a-
ion based on he sample’s pe cen iles is he p e e ed me hod (Fiss,
2011;Khedhaou ia & Thu ik, 2017). In his s udy, we used he la e
calib a ion me hod (10 h, 50 h, and 90 h pe cen iles o ull non-
membe ship, c oss-o e poin , and ull membe ship, espec i ely)
due o a lack o heo e ical knowledge ega ding he condi ions ana-
lysed. Fu he mo e, gi en he ac ha membe ship sco es o exac ly
0.5 a e d opped in he cons uc ion o u h ables, we added a con-
s an o 0.001 o all calib a e condi ions wi h membe ship sco es
below one (Fiss, 2011).
Ou come: inno a ion ou pu s
GII’s knowledge and echnology ou pu s pilla is he ou come o
in e es in his s udy. This pilla ’s sco es anged be ween 0 and 100,
wi h highe sco es indica ing be e pe o mance. This pilla encom-
passes h ee sub-pilla s (knowledge c ea ion, impac , and di usion)
ha a e based on indica o s such as pa en s, scien ific a icles, so -
wa e spending, high- ech manu ac u ing, high- ech expo s, o in el-
lec ual p ope y eceip s (WIPO, 2023). P e ious coun y-le el
s udies used he GII’s inno a ion ou pu subindex (Ding, 2022;
R^
ındașu e al., 2023), which includes he knowledge and echnology
ou pu s and c ea i e ou pu s pilla s. In his s udy, he inno a ion ou -
pu measu e used eflec s he defini ion ad anced ea lie in Sec ion
2.1 as he immedia e esul s o “in es men s”in NSI, in e ms o he
c ea ion, di usion, and use o inno a ions.
Causal condi ions: inno a ion inpu s
The fi e inno a ion inpu pilla s o GII (ins i u ions, human capi al
and esea ch, in as uc u e, ma ke sophis ica ion, and business
sophis ica ion) a e used in his s udy as causal condi ions leading o
inno a ion ou pu s, which a e commonly used in he li e a u e (Ion-
escu e al., 2022;Khedhaou ia & Thu ik, 2017;Wang e al., 2021).
Each inpu pilla is composed o h ee sub-pilla s ha agg ega e se -
e al indica o s, such as go e nmen e ec i eness and ule o law
(ins i u ions pilla ), expendi u e on educa ion and g oss expendi u e
on R&D (human capi al and esea ch pilla ), ICT access and en i on-
men al pe o mance (in as uc u e pilla ), domes ic c edi o p i a e
sec o and domes ic ma ke scale (ma ke sophis ica ion pilla ), and
o mal aining in fi m and uni e si y-indus y collabo a ion (busi-
ness sophis ica ion pilla ) (WIPO, 2023). Sco es o each pilla ange
om 0 o 100 and highe sco es indica e be e pe o mance.
Mode a o condi ion: digi al in ensi y
The Eu opean Commission’s Digi al Decade DESI da abase is a eli-
able sou ce o digi al ans o ma ion indica o s in EU coun ies and
a ious s udies ha e used hem o assess i s ela ionship wi h sus-
ainabili y (Dabbous e al., 2024;Zoppelle o & Bullini O landi, 2022)
and inno a ion (Ionescu e al., 2022;Ma i & Pue as, 2023;R^
ındașu
e al., 2023). To measu e digi al ans o ma ion in EU coun ies, we
used he DESI indica o o digi al in ensi y, which assesses he
pe cen age o small and medium-sized en e p ises (SMEs) wi h a
leas a basic le el o digi al in ensi y. Gi en a lis o 12 selec ed ech-
nologies ha a e used by SMEs, o achie e a basic le el o digi al
in ensi y a fi m mus use a leas ou (Eu opean Commission, 2023).
This indica o is used by he EU o assess membe s a es’adhe ence
o digi al a ge s, pa icula ly “mo e han 90 % o Union SMEs each a
leas a basic le el o digi al in ensi y”(Decision (EU) 2022/2481).
Since he e was only a single poin o e e ence, we we e compelled
o calib a e his condi ion ollowing he pe cen ile me hod.
Desc ip i e s a is ics and calib a ion h esholds o all condi ions
a e shown in Table 1.
Two-s age sQCA
Resea ch wi h mode a ion analyses almos exclusi ely elies on
eg ession-based me hods (Aiken & Wes , 1991;Hai e al., 2019).
One o he main limi a ions o hese me hods is hei inabili y o es
mode a ion when he s udied ela ionship is o a configu a ional
na u e (Pappas & Woodside, 2021). This s udy uses a wo-s age con-
figu a ional mode a ion app oach de eloped by Ma e al. (2024) in a
sQCA se ing, a me hod p e iously used in he li e a u e (Dua e &
Ca alho, 2024b). As i s name sugges s, i alida es a mode a o in a
sQCA analysis h ough wo s ages. Fi s , a su ficien condi ions analy-
sis o he in ended ou come is pe o med using he causal condi ions
alone, which in ou case in ol ed he cons uc ion o a u h able
wi h he fi e inno a ion inpu s. In he second s age, he p oposed
mode a o is added o he u h able and esul s can be compa ed.
To iden i y a mode a o in he second s age o analysis, he solu ion
ob ained in he fi s s age should no change d ama ically. Hence, Ma
e al. (2024), pp., 772−773 es ablished h ee equi emen s o a mod-
e a o o be iden ified: fi s , “ he e should be no changes in he causal
ac o s wi hin he causal ecipe om he fi s o he second s age”;
second, " he e should be a leas one mode a ed configu a ion ha
has change(s) o co e/pe iphe al condi ions (i.e., om co e o pe iph-
e al o om pe iphe al o co e) be ween he wo s ages”; and hi d,
“in a leas one o he mode a ed configu a ions, he mode a o
should be a co e p esence condi ion”.
E en hough he ocus o his s udy is on he configu a ions lead-
ing o he p esence o inno a ion ou pu s, we also conduc ed he
analysis o he absence o inno a ion ou pu s o a oid common pi -
alls in QCA (Schneide & Wagemann, 2012). We ollow he con en-
ion o co e/pe iphe al condi ions p oposed by Fiss (2011), which
s a es ha a co e condi ion is one ha is bo h pa o he pa simoni-
ous and he in e media e solu ion, and a pe iphe al condi ion is one
ha exis s only in he in e media e solu ion. The no a ion used in
ables is as ollows: black ci cles (“”) indica e he p esence o a con-
di ion, c ossed ou ci cles (“⨂”) indica e i s absence, blank spaces
indica e a si ua ion in which he causal condi ion may be ei he p es-
en o absen (also e med “don’ ca e”condi ion), la ge ci cles indi-
ca e co e condi ions and small ci cles indica e pe iphe al condi ions.
Table 1
Desc ip i e s a is ics and calib a ion h esholds.
N Mean S.D. Min Max Calib a ion
h esholds
Inno a ion ou pu s 27 41.64 9.59 28.03 63.45 31.52/39.38/59.39
Ins i u ions 27 65.64 11.97 47.13 85.44 47.89/64.26/82.58
Human capi al and
esea ch
27 45.79 9.32 29.07 62.68 33.38/44.58/60.19
In as uc u e 27 57.36 4.93 48.45 69.18 51.48/56.84/66.04
Ma ke
sophis ica ion
27 44.47 9.88 30.41 67.64 33.38/44.34/60.05
Business
sophis ica ion
27 48.07 12.01 28.68 75.81 31.77/47.19/64.21
Digi al in ensi y 27 67.64 12.34 41.20 89.50 50.80/67.30/87.28
M.P. Dua e and F.M.P.d.O. Ca alho Jou nal o Inno a ion & Knowledge 9 (2024) 100578
4
Resul s
P elimina y analyses
Be o e conduc ing sQCA analyses, good p ac ices ecommend
pe o ming a con a ian case analysis o assess he exis ence o cases
ha con adic he main e ec s, which would be excluded om he
ou come o a adi ional eg ession-based analysis (G eckhame e
al., 2018;Pappas & Woodside, 2021). Table 2 shows he summa y s a-
is ics o he associa ion be ween each causal condi ion and he ou -
come, as well as he numbe o cases ha a e con a y o he
expec ed di ec ion o he ela ionships. Acco ding o his analysis, a
non- i ial numbe o cases con adic he main e ec s, suppo ing
he need o pe o m a configu a ional analysis. In keeping wi h good
p ac ices (Pappas & Woodside, 2021), a necessa y condi ions analysis
was also pe o med (Table A.1), e ealing ha no single condi ion is
necessa y o he p esence o absence o high inno a ion ou pu s.
Configu a ional mode a ion
The fi s s age o he configu a ional mode a ion analysis in ol ed
he cons uc ion o u h ables o bo h he p esence and he absence
o he ou come, whe e he fi e inpu pilla s (ins i u ions, human capi-
al and esea ch, in as uc u e, ma ke sophis ica ion, and business
sophis ica ion) en e ed as causal condi ions. The esul ing u h
ables included 2
5
= 32 ows, co esponding o all possible combina-
ions o he p esence o absence o he fi e causal condi ions. Once
comple ed, logical emainde s (i.e., combina ions o condi ions wi h-
ou empi ical obse a ions) we e emo ed (Pappas & Woodside,
2021;Ragin, 2008), which we managed by se ing a equency
h eshold o 1. We es ablished he p esence o he ou come (p esence
o absence o high inno a ion ou pu s) in combina ions ( ows) whe e
aw consis ency was abo e 0.85, which is an accep ed h eshold in
he li e a u e (G eckhame e al., 2018). Addi ionally, we conside ed
a PRI (p opo ional educ ion in inconsis ency) sco e abo e 0.75,
since i is ecommended o be high and close o he consis ency
h eshold (G eckhame e al., 2018;Pappas & Woodside, 2021). The
esul ing u h ables can be seen in Tables A.2 and A.3.
Table 3 shows he solu ions ob ained in he fi s s age, e ealing
one configu a ion ha consis en ly leads o he p esence o high
inno a ion ou pu s (consis ency=0.935, co e age=0.547), and h ee
configu a ions o he absence o high inno a ion ou pu s (consis-
ency=0.950, co e age=0.678). These esul s do no suppo P oposi-
ion 1, which p oposed se e al equifinal combina ions o inno a ion
inpu s leading o high inno a ion ou pu s. Since ou sample ocused
on EU coun ies, i is likely ha hey ha ness he ull capabili ies o
hei NSI, which migh no be he case in o he coun ies whe e spe-
cific combina ions o inno a ion inpu s a e associa ed wi h high
inno a ion ou pu s (Khedhaou ia & Thu ik, 2017).
Configu a ion C1 indica es ha he conjunc ion o all fi e inno a-
ion inpu s is su ficien o he ou come, whe e he ins i u ions pilla is
he only condi ion conside ed o be pe iphe al. Rega ding he absence
o high inno a ion ou pu s, configu a ions C2 and C3 a e neu al
pe mu a ions a ound he absence o human capi al and esea ch and
in as uc u e. C2 has he highes unique co e age (0.28), meaning ha
28% o cases a e uniquely explained by his pa h, and includes he
absence o ins i u ions and business sophis ica ion as pe iphe al condi-
ions. Configu a ion C3, on he o he hand, includes he p esence o
ins i u ions and ma ke sophis ica ion as pe iphe al condi ions. Config-
u a ion C4 shows ha he absence o ma ke sophis ica ion and busi-
ness sophis ica ion a e co e condi ions o he absence o high
inno a ion ou pu s, whe e he p esence o human capi al and esea ch
and he p esence o in as uc u e a e pe iphe al condi ions.
In he second s age, wo new u h ables we e cons uc ed wi h
he fi e inno a ion inpu pilla s plus digi al in ensi y, esul ing in
ables wi h 2
6
= 64 ows (Tables A.4 and A.5). The same equency
and consis ency/PRI h esholds we e used. Table 4 p o ides he solu-
ions o he second s age. The equi emen s o iden i ying digi al
in ensi y as a mode a o condi ion a e me o he p esence o high
inno a ion ou pu s: fi s , he single configu a ion ob ained in he fi s
s age emained unchanged (i.e., all fi e inno a ion inpu s a e p es-
en ), second, human capi al and esea ch pilla changed om co e o
pe iphe al condi ion, and hi d, digi al in ensi y appea ed as a co e
condi ion in he solu ion. This esul suppo s P oposi ion 2. Ne e -
heless, he equi emen s we e no me o he absence o high inno-
a ion ou pu s, indica ing ha digi al in ensi y only mode a es he
configu a ional ela ionship be ween inno a ion inpu s and he p es-
ence o high inno a ion ou pu s.
Configu a ion C5 shows ha when digi al in ensi y is high, he
human capi al and esea ch pilla , albei s ill p esen , becomes a
pe iphe al condi ion, hence losing i s impo ance in he configu a ion
ha explains high inno a ion ou pu s in EU coun ies. Addi ionally,
in as uc u e, ma ke sophis ica ion, business sophis ica ion, and digi-
al in ensi y we e iden ified as co e condi ions, and ins i u ions
emained a pe iphe al condi ion. An o e all consis ency o 0.937
e eals ha he empi ical e idence suppo s a s ong ela ionship
be ween he solu ion and he ou come. Mo eo e , o e all co e age
(0.528) shows ha a subs an ial p opo ion o cases a e explained by
he solu ion. EU coun ies ha a e conside ed ypical cases co e ed by
his configu a ion include Ge many, Aus ia, Ne he lands, and No dic
EU coun ies, o hey p esen ed membe ship sco es in configu a ion
C5 and he ou come abo e 0.5 (Schneide & Wagemann, 2012).
Rega ding he absence o high inno a ion ou pu s, configu a ion
C6 included he absence o in as uc u e, ma ke sophis ica ion,
business sophis ica ion, and digi al in ensi y as co e condi ions,
whe e he absence o ins i u ions is a pe iphe al condi ion. Typical
cases wi h his configu a ion include G eece, La ia, and mos Eas e n
EU coun ies. Configu a ion C7 is ep esen a i e o Luxembou g and
Li huania, whe e he absence o human capi al and esea ch, in a-
s uc u e, and digi al in ensi y, and he p esence o ins i u ions and
ma ke sophis ica ion a e all pe iphe al condi ions leading o he
Table 2
Summa y o con a ian case analysis.
Condi ion Kendall’s au-c (p- alue) Numbe o con a ian
cases (%)
Ins i u ions 0.518 (<0.001) 3 (11.1 %)
Human capi al and
esea ch
0.645 (<0.001) 1 (3.7 %)
In as uc u e 0.453 (<0.001) 2 (7.4 %)
Ma ke sophis ica ion 0.528 (<0.001) 2 (7.4 %)
Business sophis ica ion 0.641 (<0.001) 3 (7.4 %)
Digi al in ensi y 0.610 (<0.001) 3 (11.1 %)
Kendall’s au-c es assesses he deg ee o associa ion be ween each condi ion and
inno a ion ou pu s.
Table 3
Su ficien condi ions analysis.
Ou come: Inno a ion
ou pu s
»(Inno a ion ou pu s)
C1 C2 C3 C4
Ins i u ions
Human capi al and esea ch
In as uc u e
Ma ke sophis ica ion
Business sophis ica ion
Consis ency 0.935 0.949 0.984 0.936
Raw co e age 0.547 0.579 0.258 0.300
Unique co e age 0.547 0.282 0.062 0.033
Solu ion consis ency 0.935 0.950
Solu ion co e age 0.547 0.678
»deno es he absence o a condi ion.
M.P. Dua e and F.M.P.d.O. Ca alho Jou nal o Inno a ion & Knowledge 9 (2024) 100578
5
absence o high inno a ion ou pu s. Las ly, C8 shows ha he absence
o ma ke sophis ica ion and business sophis ica ion a e co e condi-
ions o he absence o high inno a ion ou pu s, whe eas he
absence o in as uc u e and he p esence o ins i u ions, human
capi al and esea ch, and digi al in ensi y a e pe iphe al condi ions.
Po ugal is he only ypical case ep esen ed by configu a ion C8.
Robus ness o findings
To a es he alidi y o he esul s and hei sensi i i y o h eshold
changes, we conduc ed se e al analyses (G eckhame e al., 2018;Pap-
pas & Woodside, 2021). Fi s , we es ed he p edic i e alidi y ( o
de ailed p ocedu es see Pappas and Woodside (2021)) o ou model o
see how well i p edic ed inno a ion ou pu s in di e en samples.
Thus, wo andom subsamples we e c ea ed (modelling and holdou )
om he ull sample, con aining a simila numbe o cases. A su ficien
condi ions analysis was hen pe o med o each subsample holding
he equency and consis ency/PRI cu -o s om he main analysis.
Table 5 e eals ha he holdou subsample consis en ly p edic s he
solu ion ob ained om he modelling subsample, and bo h o e all
consis ency and co e age alues a e close o hose o he main analysis
(Table A.6 shows he p edic i e alidi y es o he modelling subsam-
ple wi h he solu ion ob ained wi h he holdou subsample). These
esul s highligh a good p edic i e alidi y o ou model.
We also assessed he obus ness o ou esul s when calib a ion
ancho poin s and consis ency h esholds changed (G eckhame e
al., 2018). These analyses we e pe o med only o he p esence o
high inno a ion ou pu s. F equency h eshold changes we e no con-
duc ed because using a equency o 2 o highe would imply he
emo al o mo e han 20 % o cases (Ragin, 2008). Thus, we pe o med
wo analyses, conside ing consis ency/PRI h esholds o 0.9/0.8 in one
and 0.8/0.7 in ano he . Rega ding changes in calib a ion ancho
poin s, wo o he analyses we e conduc ed: one wi h a da ase cali-
b a ed wi h pe cen iles 5/50/95, and he o he wi h pe cen iles 15/
50/85. Table 6 p o ides a summa y o obus ness analyses. Conce n-
ing he main findings, no changes in he solu ion we e obse ed. Con-
sis ency and co e age alues ob ained we e simila o hose o he
main esul s. E idence con e s obus ness o ou ini ial esul s, sup-
po ing a mode a ing ole o digi al in ensi y on he ela ionship
be ween inno a ion inpu s and ou pu s.
Discussion o esul s
The pu pose o his s udy was o analyse how configu a ions o
coun ies’inno a ion inpu s leading o inno a ion ou pu s a e
a ec ed by digi al ans o ma ion le els. The s udy was mo i a ed by
a gap in he digi al ans o ma ion li e a u e. As D˛ab owska e al.
(2022) no ed, he majo i y o s udies end o ocus on digi al ans o -
ma ion a he o ganisa ional le el, while i s na ional impac is o en
o e looked. We add essed his gap by ocusing on how digi al ans-
o ma ion shapes coun ies’NSI.
Ou fi s p oposi ion s a ed ha he e a e se e al equifinal combi-
na ions o inno a ion inpu s ha a e su ficien o high inno a ion
ou pu s. The esul s did no suppo his p oposi ion, o only one
configu a ion was ound o be su ficien o high inno a ion ou pu s.
This finding is in line wi h ha o Yu and Hua ng (2023), who ound
ha a single combina ion o all fi e inpu pilla s led o high
Table 4
Mode a ing impac s o digi al in ensi y.
Ou come: Inno a ion
ou pu s
»(Inno a ion ou pu s)
C5 C6 C7 C8
Ins i u ions
Human capi al and esea ch
In as uc u e
Ma ke sophis ica ion
Business sophis ica ion
Digi al in ensi y
Consis ency 0.937 0.975 0.983 0.955
Raw co e age 0.528 0.577 0.244 0.223
Unique co e age 0.528 0.351 0.072 0.023
Solu ion consis ency 0.937 0.972
Solu ion co e age 0.528 0.672
Typical cases Aus ia, Denma k, Finland,
Ge many, Ne he lands, Sweden
Bulga ia, C oa ia, G eece, La ia, Poland, Romania, Slo akia Li huania, Luxembou g Po ugal
»deno es he absence o a condi ion.
Table 5
P edic i e alidi y es ing.
Modelling subsample (n = 14)
Ou come: Inno a ion ou pu s
M1 M2
Ins i u ions
Human capi al and esea ch
In as uc u e
Ma ke sophis ica ion
Business sophis ica ion
Digi al in ensi y
Consis ency 0.964 0.899
Raw co e age 0.277 0.493
Unique co e age 0.095 0.311
Solu ion consis ency 0.912
Solu ion co e age 0.588
Holdou subsample (n = 13)
Consis ency 0.887 0.969
Co e age 0.143 0.558
O e all consis ency 0.969
O e all co e age 0.558
Black ci cles indica e he p esence o a condi ion, and
ci cles wi h “”indica e i s absence. Blank spaces indi-
ca e “don’ ca e”condi ions.
Table 6
Summa y o obus ness analyses.
Change in h esholds Changes in solu ion Consis ency Co e age
Consis ency/PRI (0.90/0.80) No changes 0.937 0.528
Consis ency/PRI (0.80/0.70) No changes 0.937 0.528
Calib a ion (P5, P50, P95) No changes 0.935 0.556
Calib a ion (P15, P50, P85) No changes 0.944 0.490
M.P. Dua e and F.M.P.d.O. Ca alho Jou nal o Inno a ion & Knowledge 9 (2024) 100578
6
inno a ion ou pu s in ecen yea s. Ne e heless, ou esul s chal-
lenge o he s udies ha ound se e al equifinal configu a ions o
inno a ion inpu s ha a e su ficien o inno a ion ou pu s in high-
income coun ies (C espo & C espo, 2016;Wang e al., 2021). Khed-
haou ia and Thu ik (2017), on he o he hand, despi e finding wo
equifinal configu a ions leading o high inno a ion ou pu s, disco -
e ed ha only one combina ion o he fi e inno a ion inpu pilla s
was su ficien o e y high inno a ion ou pu s. This sugges s ha
only coun ies wi h well-de eloped, highly di e sified NSI a ain
e y high le els o inno a ion ou pu s.
The second p oposi ion s a ed ha combina ions o inno a ion
inpu s leading o high inno a ion ou pu s su e a mode a ing e ec
o digi al ans o ma ion. The esul s o ou configu a ional mode a-
ion analysis e ealed ha configu a ions o EU coun ies’inno a ion
inpu s ha a e conduci e o high inno a ion ou pu s change in he
p esence o high digi al in ensi y, and such change is eflec ed in he
loss o impo ance o he human capi al and esea ch pilla . This is
pe haps he mos su p ising finding, which should no be unde s ood
as a loss o jobs due o in es men s in digi al echnologies (Balsmeie
& Woe e , 2019) because i s p esence is s ill pa o a su ficien com-
bina ion o condi ions leading o high inno a ion ou pu s. Fu he -
mo e, he human capi al and esea ch pilla p ima ily eflec s he
le el o educa ion o he wo k o ce and in es men s made in R&D,
which, acco ding o ou esul s, becomes a con ibu ing condi ion
suppo ing he cen al oles o digi al in ensi y, in as uc u e, and
ma ke and business sophis ica ion. This finding could be explained
by he ac ha digi al skills a e ailing o ca ch up wi h digi al ech-
nological ad ancemen s (G eenan & Napoli ano, 2024), leading o a
sho age o skilled labou (Ska e e al., 2023). A en ion should be
gi en o his issue since li e a u e has shown ha inno a ion in Eu o-
pean SMEs is inc easingly ancho ed on employees’digi al skills
(Scuo o e al., 2021).
E en hough digi al in ensi y was no iden ified as a mode a-
o when combina ions o inno a ion inpu s led o he absence o
high inno a ion ou pu s, some findings a e wo h discussing.
Fi s , esul s e eal he causal asymme y ene o sQCA by iden-
i ying configu a ions leading o he absence o high inno a ion
ou pu s ha a eno heexac opposi eo heconfigu a ion lead-
ing o he p esence o he ou come. Second, h ee configu a ions
we e ob ained o he absence o high inno a ion ou pu s in EU
coun ies. An impo an poin o cla ifica ion is ha he absence
o high inno a ion ou pu s does no necessa ily co espond o
low inno a ion ou pu s, i a he iden ifies he g oup o coun ies
wi h membe ship sco es below 0.5 in he se o high inno a ion
ou pu s (Ragin, 2008;Schneide & Wagemann, 2012). Tha being
said, ou esul s o e a benchma k agains which he EU and
na ional go e nmen s could guide Membe S a es’policies
owa ds inc easing hei inno a ion le els. Mo eo e , since di e -
en coun ies possess di e en configu a ions, ailo ed solu ions
could be de eloped o specific membe s a es. In he case o Po -
ugal, o ins ance, in as uc u e and ma ke and business sophis-
ica ion pilla s should be he p incipal ocus o imp o emen .
Conclusion
O e all, his pape helped us o unde s and he ole o digi al
ans o ma ion in EU coun ies’NSI, which we ha e ound o be o
cen al impo ance o achie ing high inno a ion ou pu s. Resul s
e ealed ha highly di e se NSI, whe e high le els o inno a ion
inpu s a e p esen , is su ficien o a aining inno a ion ou pu s in EU
coun ies. Fu he mo e, digi al ans o ma ion was shown o influ-
ence how combina ions o inno a ion inpu s ansla e in o inno a-
ion ou pu s. Hence, policy decision-make s in en on aising hei
coun ies’inno a ion ou pu le els should acknowledge he need o
a well-de eloped NSI and imp o e he digi al ans o ma ion o hei
businesses.
Con ibu ions and implica ions
This s udy p o ides he ollowing con ibu ions. Fi s , we espond
o schola s’conce ns abou he absence o esea ch on he impac s o
digi al ans o ma ion a he coun y le el (D˛ab owska e al., 2022).
E en hough i is ecognised ha digi al echnologies influence se -
e al socie al phenomena (Ganju e al., 2016;Majch zak e al., 2016),
hei influence on coun ies’NSI is s ill unde explo ed. Thus, by ana-
lysing how digi al ans o ma ion impac s NSI, his s udy p o ides
e idence o he ans o ma i e ole o digi al ans o ma ion in coun-
ies’abili y o con e inno a ion inpu s in o inno a ion ou pu s.
Second, while p e ious s udies unco e ed he na ional inno a ion
inpu s ha ma e mos o achie ing high le els o digi al ans o -
ma ion (Ionescu e al., 2022), and desc ibed how digi al ans o ma-
ion impac s inno a ion ou pu s di ec ly and indi ec ly h ough
digi al skills (Cia li e al., 2021;R^
ındașu e al., 2023), hei adop ed
pe spec i es p e en ed a sys emic, configu a ional analysis o how
inno a ion inpu s combine wi h each o he o p oduce inno a ion
ou pu s. Hence, his s udy no only adop s a configu a ional pe spec-
i e, bu also posi ions digi al ans o ma ion as a bounda y condi ion
ha shapes how EU coun ies a ange hei inno a ion inpu s o
a ain inno a ion ou pu s. By adop ing bo h a configu a ional and
con ingency app oach o analysing he ole o digi al ans o ma ion
in coun ies’NSI, his s udy es ablished a iche and “ hick”unde -
s anding o his phenomenon (Jackson & Deeg, 2019).
Policy implica ions o his s udy a e s aigh o wa d since he
solu ion ob ained ha leads o he p esence o high inno a ion ou -
pu s can be conside ed a benchma k, agains which o he coun ies
can compa e and de elop hei NSIs acco dingly. Fu he mo e, by
knowing he specific configu a ion ollowed by each coun y, ailo ed
policies can be de eloped owa ds specific inno a ion pilla s in need
o imp o emen . Mo eo e , owing o he ac ha human capi al and
esea ch could be lagging behind he in oduc ion o new digi al
echnologies, such as big da a o gene a i e a ificial in elligence,
Union-based ecommenda ions should be made o p omo e he nec-
essa y de elopmen o digi al skills in e e y EU Membe S a e. Impli-
ca ions o businesses may also be de i ed, pa icula ly hose
de o ed o in e na ionalisa ion wi hin he EU. Ou esul s educa e
business manage s abou he specific configu a ion o EU coun ies’
NSI, e ealing he s eng hs and weaknesses o hei inno a ion
inpu s.
Limi a ions and u u e esea ch di ec ions
A limi a ion o his s udy is ha da a was ga he ed o a single
poin in ime, which p e en s us om acqui ing deepe knowledge o
he e olu ion o he configu a ional mode a ion ound in ou esul s.
Fu u e s udies migh find i ui ul o assess whe he he configu a-
ions ob ained he e emain s able ac oss ime, pa icula ly when
exogenous shocks, such as he COVID-19 pandemic, occu . Fu he -
mo e, a single inno a ion measu e− he Global Inno a ion Index
−was used in his s udy. O he measu es, such as he Eu opean Com-
mission’s Eu opean Inno a ion Sco eboa d, could be be e adjus ed
o EU Membe S a es, ye i lacks he inpu -ou pu p ope y, which is
a ele an cha ac e is ic o analysing he inno a ion pe o mance o
coun ies (Edquis e al., 2018).
Funding
This wo k was suppo ed by FCT −Funda¸c~
ao pa a a Ci^
encia e a
Tecnologia, I.P., P ojec UIDB/05037/2020 wi h DOI 10.54499/UIDB/
05037/2020. Ma celo Dua e also acknowledges unding om FCT −
Funda¸c~
ao pa a a Ci^
encia e a Tecnologia, P ojec UI/BD/150977/2021
wi h DOI 10.54499/UI/BD/150977/2021. The unding sou ce was no
in ol ed in he s udy design, da a collec ion, analysis and
M.P. Dua e and F.M.P.d.O. Ca alho Jou nal o Inno a ion & Knowledge 9 (2024) 100578
7
in e p e a ion, in he w i ing o he epo , no in he decision o sub-
mi he a icle o publica ion.
Decla a ion o compe ing in e es
The au ho s decla e ha hey ha e no known compe ing financial
in e es s o pe sonal ela ionships ha could ha e appea ed o influ-
ence he wo k epo ed in his pape .
CRediT au ho ship con ibu ion s a emen
Ma celo Pe ei a Dua e: W i ing − e iew & edi ing, W i ing −
o iginal d a , Visualiza ion, Valida ion, So wa e, Me hodology,
In es iga ion, Funding acquisi ion, Fo mal analysis, Da a cu a ion,
Concep ualiza ion. Fe nando Manuel Pe ei a de Oli ei a Ca alho:
W i ing − e iew & edi ing, Valida ion, Supe ision, P ojec adminis-
a ion, In es iga ion, Funding acquisi ion, Concep ualiza ion.
Appendix
Table A.1
Necessa y condi ions analysis.
Ou come: Inno a ion ou pu s »(Inno a ion ou pu s)
Consis ency Co e age Consis ency Co e age
Ins i u ions 0.855 0.747 0.492 0.505
»(Ins i u ions) 0.433 0.420 0.753 0.859
Human capi al and esea ch 0.863 0.796 0.447 0.485
»(Human capi al and esea ch) 0.441 0.404 0.812 0.874
In as uc u e 0.786 0.770 0.443 0.510
»(In as uc u e) 0.499 0.432 0.800 0.815
Ma ke sophis ica ion 0.790 0.811 0.384 0.464
»(Ma ke sophis ica ion) 0.478 0.398 0.844 0.825
Business sophis ica ion 0.893 0.817 0.442 0.476
»(Business sophis ica ion) 0.427 0.394 0.830 0.901
Digi al in ensi y 0.863 0.797 0.475 0.516
»(Digi al in ensi y) 0.475 0.435 0.813 0.875
»deno es he absence o a condi ion.
Table A.2
Fi s -s age u h able (p esence o inno a ion ou pu s).
Condi ions Ou come Consis ency
(1) (2) (3) (4) (5) N (6) Raw PRI
1 1 1 1 1 7 1 0,94 0,90
11011 100,85 0,64 No es:
11101 100,85 0,55 (1) Ins i u ions
01110 100,90 0,41 (2) Human capi al and esea ch
00110 100,88 0,34 (3) In as uc u e
10111 100,81 0,34 (4) Ma ke sophis ica ion
10101 100,79 0,33 (5) Business sophis ica ion
01101 200,87 0,31 (6) Inno a ion ou pu s
00010 100,85 0,22
01000 100,73 0,19
11000 100,82 0,19
10011 100,76 0,07
00000 700,46 0,06
10010 100,75 0,02
Table A.3
Fi s -s age u h able (absence o inno a ion ou pu s).
Condi ions Ou come Consis ency
(1) (2) (3) (4) (5) N »(6) Raw PRI
1111171 0,94 0,90
1101110 0,85 0,64 No es:
1110110 0,85 0,55 (1) Ins i u ions
0111010 0,90 0,41 (2) Human capi al and esea ch
0011010 0,88 0,34 (3) In as uc u e
1011110 0,81 0,34 (4) Ma ke sophis ica ion
1010110 0,79 0,33 (5) Business sophis ica ion
0110120 0,87 0,31 (6) Inno a ion ou pu s
0001010 0,85 0,22
0100010 0,73 0,19
1100010 0,82 0,19
1001110 0,76 0,07
0000070 0,46 0,06
1001010 0,75 0,02
»deno es he absence o a condi ion.
M.P. Dua e and F.M.P.d.O. Ca alho Jou nal o Inno a ion & Knowledge 9 (2024) 100578
8
