Bach ögle , Julia; Gabelbe ge , Fabian; Ma ques San os, Anabela; Doussineau,
Ma hieu
Wo king Pape
When cohesion mee s excellence: Analysing he d i e s
o syne gies be ween EU R&I unding ins umen s in EU
egions
JRC Wo king Pape s on Te i o ial Modelling and Analysis, No. 05/2025
P o ided in Coope a ion wi h:
Join Resea ch Cen e (JRC), Eu opean Commission
Sugges ed Ci a ion: Bach ögle , Julia; Gabelbe ge , Fabian; Ma ques San os, Anabela; Doussineau,
Ma hieu (2025) : When cohesion mee s excellence: Analysing he d i e s o syne gies be ween EU
R&I unding ins umen s in EU egions, JRC Wo king Pape s on Te i o ial Modelling and Analysis,
No. 05/2025, Eu opean Commission, Join Resea ch Cen e (JRC), Se ille
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/322073
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/4.0/
When cohesion mee s excellence:
analysing he d i e s o syne gies
be ween EU R&I unding
ins umen s in EU egions
No 05/2025
2025
Au ho s:
Bach ögle
-Unge , J.
Gabelbe ge , F.
Ma ques San os, A.
Doussineau
, M.
This publica ion is a wo king pape by he Join Resea ch Cen e (JRC), he Eu opean Commission’s science and knowledge
se ice. I aims o p o ide e idence-based scien i ic suppo o he Eu opean policymaking p ocess. The con en s o his
publica ion do no necessa ily e lec he posi ion o opinion o he Eu opean Commission. Nei he he Eu opean
Commission no any pe son ac ing on behal o he Commission is esponsible o he use ha migh be made o his
publica ion. Fo in o ma ion on he me hodology and quali y unde lying he da a used in his publica ion o which he
sou ce is nei he Eu os a no o he Commission se ices, use s should con ac he e e enced sou ce. The designa ions
employed and he p esen a ion o ma e ial on he maps do no imply he exp ession o any opinion wha soe e on he
pa o he Eu opean Union conce ning he legal s a us o any coun y, e i o y, ci y o a ea o o i s au ho i ies, o
conce ning he delimi a ion o i s on ie s o bounda ies.
The JRC Wo king Pape s on Te i o ial Modelling and Analysis a e published unde he supe ision o Simone
Salo i, And ea Con e, and Anabela M. San os o JRC Se ille, Eu opean Commission. This se ies mainly add esses he
economic analysis ela ed o he egional and e i o ial policies ca ied ou in he Eu opean Union. The Wo king Pape s o
he se ies a e mainly a ge ed o policy analys s and o he academic communi y and a e o be conside ed as ea ly-s age
scien i ic pape s con aining ele an policy implica ions. They a e mean o communica e o a b oad audience p elimina y
esea ch indings and o gene a e a deba e and a ac eedback o u he imp o emen s.
Con ac in o ma ion
Name: Anabela M. San os
Add ess: Edi icio Expo, C/Inca Ga cilaso 3, 41092 Se illa (Spain)
Email: [email protected] opa.eu
Tel.: +34 95 448 71 61
EU Science Hub
h ps://join - esea ch-cen e.ec.eu opa.eu
JRC141964
Se ille: Eu opean Commission, 2025
© Eu opean Union, 2025
The euse policy o he Eu opean Commission documen s is implemen ed by he Commission Decision 2011/833/EU o 12
Decembe 2011 on he euse o Commission documen s (OJ L 330, 14.12.2011, p. 39). Unless o he wise no ed, he euse
o his documen is au ho ised unde he C ea i e Commons A ibu ion 4.0 In e na ional (CC BY 4.0) licence
(h ps://c ea i ecommons.o g/licenses/by/4.0/). This means ha euse is allowed p o ided app op ia e c edi is gi en and
any changes a e indica ed.
How o ci e his epo : Bach ögle -Unge , J.; Gabelbe ge , F.; Ma ques San os, A.; Doussineau, M., When cohesion mee s
excellence: analysing he d i e s o syne gies be ween EU R&I unding ins umen s in EU egions, Eu opean Commission,
Se ille, 2025, JRC141964.
1
Con en s
Abs ac ...................................................................................................................................................................................................... 2
Acknowledgemen s .............................................................................................................................................................................. 3
Execu i e summa y .............................................................................................................................................................................. 4
1. In oduc ion ....................................................................................................................................................................................... 6
2. Li e a u e e iew ............................................................................................................................................................................ 8
2.1. Theo ical ounda ions and policy cohe ence ........................................................................................................ 8
2.2. The impo ance o syne gies and hei po en ial e ec s ............................................................................. 8
3. Da a and Me hods ...................................................................................................................................................................... 11
3.1. P ojec da a as a ool o analyze syne gies ...................................................................................................... 11
3.2. Measu ing syne gies using he cosine simila i y index ............................................................................... 16
3.3. Reg ession analysis conside ing NUTS-3 socio-economic cha ac e is ics ....................................... 18
4. Resul s ............................................................................................................................................................................................... 20
4.1. S ylized ac s on he dis ibu ion o ERDF R&I and Ho izon 2020 unding ac oss SGCs ........ 20
4.2. The (dis)simila i y o he dis ibu ion o EU R&I unding among SGCs ............................................. 22
4.3. Socio-economic cha ac e is ics as de e minan s o po en ial unding syne gies ....................... 25
5. Conclusions ..................................................................................................................................................................................... 32
6. Discussion........................................................................................................................................................................................ 33
Re e ences ............................................................................................................................................................................................. 34
Annexes ................................................................................................................................................................................................... 38
Annex A: Dis ibu ion o Ho izon 2020 and ERDF R&I unding among NUTS-3 egions .................... 38
Annex B. Desc ip i e s a is ics and obus ness checks ......................................................................................... 42
Lis o igu es ....................................................................................................................................................................................... 53
Lis o ables ......................................................................................................................................................................................... 54
2
Abs ac
This pape in oduces a no el app oach o measu ing syne gies be ween Ho izon 2020 and cohesion
policy unding in he ield o R&I du ing he 2014-2020 p og amming pe iod. Le e aging p ojec -
le el da a, we calcula e egional cosine simila i y indices based on socie al g and challenges (SGCs)
add essed o assess alignmen be ween he wo EU unding ins umen s in EU NUTS-3 egions.
Resul s indica e ha syne gies a e less likely in u al a eas, eme ging inno a o s, and – hough no
s a is ically signi ican – less de eloped egions, highligh ing he ole o business en i onmen s and
inno a ion ecosys ems. Reg ession analysis e eals ha unding alignmen is posi i ely linked o
he p esence o uni e si ies and specializa ion in knowledge-in ensi e se ices, hough he la e
exhibi s a nonlinea e ec unde ce ain ci cums ances. Howe e , a highe numbe o SGCs
add essed in sma specializa ion (S3) policy objec i es is nega i ely associa ed wi h hema ic
unding simila i y, likely due o agmen a ion and dilu ion o ocus. Regions ha p io i ize oo many
SGCs may educe hei abili y o de elop s ong specializa ion and align di e en unding sou ces
e ec i ely.
3
Acknowledgemen s
We a e hank ul o help ul commen s by Pe e Hube , an anonymous e iewe a JRC Se ille,
Dimi i Co pakis, A naul Mo ison and F ancesco Cappellano.
Au ho s
Julia Bach ögle -Unge * (co esponding au ho )
Aus ian Ins i u e o Economic Resea ch (WIFO), Vienna, Aus ia
julia.bach oegle -[email p o ec ed]
Fabian Gabelbe ge *
Aus ian Ins i u e o Economic Resea ch (WIFO), Vienna, Aus ia
Fabian.gabelbe g[email p o ec ed]
Anabela M. San os
Eu opean Commission, Join Resea ch Cen e, Se ille, Spain
anabela.ma ques-san [email protected] opa.eu
Ma hieu Doussineau
Eu opean Fu u e Inno a ion Sys em Cen e, B ussels, Belgium
[email p o ec ed]
* The esea che s a WIFO (Julia Bach ögle -Unge and Fabian Gabelbe ge ) we e suppo ed by unds o he Aus ian
Na ional Bank (OeNB Jubiläums onds p ojec no. 18463).
Keywo ds: Syne gy; R&I unding; Cohesion policy; Ho izon 2020; EU egions
JEL Codes: O31; O52; R58.
4
Execu i e summa y
Syne gies in unding e e o he s a egic alignmen and coo dina ed use o inancial esou ces o
achie e a g ea e impac han indi idual p og ams ope a ing sepa a ely. In he Eu opean Union (EU),
unding syne gies ha e gained p ominence as a way o enhance he e ec i eness o esea ch and
inno a ion (R&I) in es men s, pa icula ly h ough cohesion policy and Ho izon 2020 p og ams. By
le e aging syne gies, he EU aims o maximize he bene i s o unding ins umen s and accele a e
p og ess in add essing socie al challenges.
The s udy examines he ex en o which cohesion policy (unde he Eu opean Regional De elopmen
Fund – ERDF) and Ho izon 2020 unding we e used syne gis ically in he 2014–2020 p og amming
pe iod o add ess socie al g and challenges (SGCs). The analysis explo es:
— The simila i y in unding dis ibu ion ac oss EU egions.
— Regional di e ences based on ac o s such as agglome a ion densi y, de elopmen le el, and
inno a ion pe o mance.
— Key egional socio-economic cha ac e is ics in luencing unding alignmen .
Using a no el da a-d i en app oach, he s udy employs ex mining and la en seman ic analysis o
classi y p ojec s based on hei hema ic ocus. A cosine simila i y index measu es he deg ee o
alignmen be ween ERDF R&I and Ho izon 2020 unding a he NUTS-3 egional le el. Reg ession
analysis hen iden i ies he ac o s in luencing he in ensi y o unding syne gies.
Key Findings
— Syne gies be ween Ho izon 2020 and ERDF R&I unding a e s onge in egions wi h well-
de eloped inno a ion ecosys ems, pa icula ly hose wi h a high concen a ion o uni e si ies
and knowledge-in ensi e se ices.
— Ru al a eas, eme ging inno a o s, and less de eloped egions show weake unding alignmen ,
indica ing po en ial ba ie s o syne gy, such as agmen ed business en i onmen s and weake
ins i u ional collabo a ion.
— Regions ha include a high numbe o SGCs in hei Sma Specialisa ion S a egy (S3) end o
exhibi lowe unding alignmen , sugges ing ha an o e ly b oad policy ocus may dilu e
specializa ion and hinde e ec i e syne gy.
— Popula ion densi y plays a signi ican ole, wi h highly u banized egions demons a ing s onge
unding complemen a i ies.
Policy Implica ions
To enhance syne gies be ween EU unding ins umen s, policymake s should:
— S eamline legal amewo ks and ins i u ional coo dina ion o acili a e collabo a ion be ween
unding p og ams.
5
— S eng hen egional inno a ion ecosys ems by in es ing in uni e si ies and knowledge-in ensi e
indus ies.
— Re ine S3 p io i ies o encou age mo e a ge ed specializa ions, p e en ing he dilu ion o
unding ac oss oo many policy a eas.
— Suppo ailo ed unding s a egies o u al and less de eloped egions o b idge he gap in
unding syne gies.
The indings con ibu e o ongoing policy discussions on op imizing he alignmen o EU R&I unding.
Fu u e esea ch should u he in es iga e how e ol ing EU policy amewo ks beyond 2020 can
build upon hese insigh s o ein o ce syne gies and enhance he EU’s global compe i i eness.
6
1. In oduc ion
The Camb idge dic iona y de ines syne gy as “ he combined powe o a g oup (…) when hey a e
wo king oge he ha is g ea e han he o al powe achie ed by each wo king sepa a ely”
(Combley, 2011:837). The concep is he e o e also linked o he bene i s and sa ings o wo king
oge he o achie e a common goal. In he ield o policy e alua ion, he academic li e a u e (see,
e.g., Van den Be gh e al., 2021) conside s ha posi i e syne gy is achie ed when he combined
e ec o mo e han one policy ins umen is g ea e han he sum o hei indi idual e ec s.
Syne gies can inc ease he e ec i eness o a policy, bu also po en ially educe nega i e e ec s
and accele a e he achie emen o esul s (Chou, 2018). The Eu opean Commission (2022) de ines
unding syne gy as he s a egic alignmen and combined use o inancial esou ces o achie e
mo e signi ican impac and e iciency han would be possible i each p og am ope a ed
independen ly. This in ol es coo dina ing e o s ac oss di e en unding sou ces, such as he
Ho izon F amewo k P og amme and he Eu opean Regional De elopmen Fund (ERDF), o mo e
b oadly unde he cohesion policy unds, o suppo common esea ch, inno a ion and egional
de elopmen objec i es and he eby maximize hei collec i e bene i s.
The policy mix, in e ms o consis ency, cohe ence, in eg a ion and coo dina ion be ween he di e -
en inancing ins umen s, is a key elemen o achie e syne gies (Ma ques San os, 2021). In his e-
spec , he OECD (2003) dis inguishes policy cohe ence om policy coo dina ion (ge ing sys ems o
wo k oge he ) and policy consis ency (a oiding con adic o y policies). F om a policy cohe ence pe -
spec i e, he concep o syne gies be ween EU unding equi es mo e han poli ical in en . Policy co-
he ence is abou ensu ing ha policies a e mu ually ein o cing o c ea e syne gies owa ds a de-
ined goal (Reid e al., 2007).
In he Eu opean Union (EU) con ex , he concep o syne gies gained a new dimension wi h he
eme gence o he Sma Specializa ion S a egy (S3) concep in 2009. S3 is a place-based inno a-
ion policy ha helps egions iden i y and de elop compe i i e ad an ages by ocusing on hei
s eng hs and po en ial (Fo ay e al., 2009). The aim is o ake he he e ogenei y o egions se i-
ously and s eng hen economic de elopmen by in es ing in he ad ancemen o echnologies ha
a e ela ed o he skills and capabili ies p e alen in each egion (e.g., Ba be o e al., 2024). Since
2014, S3 is a key equi emen o cohesion policy unding a ge ed o R&I h ough he ERDF in iew
o ensu ing e icien esou ce alloca ion and syne gies be ween policies (Eu opean Commission,
2011). The Eu opean Commission (EC) explici ly encou ages he implemen a ion o syne gies be-
ween cohesion policy and Ho izon 2020 (Eu opean Commission, 2014) and Ho izon Eu ope (Eu o-
pean Commission, 2022).
Ne e heless, acco ding o se e al s udies (see, e.g., ECA, 2022; RIMA, 2025; Segal e al., 2025) he
ull po en ial o syne gies be ween unding ins umen s has no been achie ed ye . Among he main
ba ie s o achie e syne gies a e he legal amewo k and he lack o coo dina ion be ween he di -
e en ins i u ions implemen ing he p og ams (Segal e al., 2025).
The cu en poli ical con ex unde sco es he u gency o add essing hese challenges, as he EU
s i es o enhance i s s a egic au onomy and global compe i i eness. In his ega d, he D aghi
(2024) epo highligh s he c i ical need o s eng hen syne gies be ween EU unding ins umen s,
ad oca ing o a mo e cohe en and coo dina ed app oach o policy implemen a ion. By s eamlin-
ing legal amewo ks and imp o ing ins i u ional collabo a ion, he epo highligh s ha he EU can
op imize esou ce alloca ion, os e inno a ion, and ein o ce i s economic esilience in an inc eas-
ingly compe i i e global landscape.
13
cons uc ed based on he RISIS KNOWMAK on ology (Mayna d e al., 2019)8 and supplemen ed wi h
keywo ds ex ac ed om Ho izon 2020 p ojec s, which a e al eady classi ied acco ding o SGCs. The
la e s ep ensu es he alignmen o he ca ego ies applied o p ojec s om bo h unding sou ces. In
a nex s ep, he lis o keywo ds was e ined in a ial-and-e o app oach by checking and cleaning
o keywo ds ound o be oo gene al.9 The numbe o keywo ds pe SGC anges om a ound 370
o Secu i y, o e 500 o Clima e and Socie y o o e 700 o T anspo and Heal h.
I an ERDF p ojec is ound o ela e o mo e han one SGC, i s ull inancial amoun is assigned o
each o hose ca ego ies. ERDF p ojec s ha canno be assigned o a leas one socie al g and
challenge (SGC) using he keywo d s ing sea ch app oach a e excluded om he ollowing analysis.
This yields a sample o 148,000 ERDF R&I p ojec s wi h o al eligible expendi u es o 58 billion EUR,
ep esen ing wo hi ds o all ERDF R&I p ojec s o , equi alen ly, 76% o he sum o o al eligible
expendi u es.
Table 1 shows he dis ibu ion o unding ac oss SGCs o he sample o be analyzed. The sha e o
Ho izon 2020 unding alloca ed o T anspo , Ene gy and Heal h is no ably la ge han he
espec i e sha e o ERDF R&I unding. By con as , he sha e o ERDF R&I unding alloca ed o
Clima e p ojec s is la ge han he one among Ho izon 2020 unding. Wha is s iking is he
p ominen ocus o ERDF R&I p ojec s on he SGC Socie y, which migh pa ly e lec a ela i ely
b oad de ini ion o his SGC as compa ed o o he SGCs. Howe e , he numbe o keywo ds
ep esen ing Socie y is among he lowes and he e a e subs an ial c oss-coun y di e ences: The
ERDF R&I unding sha e o Socie y anges om 9.3% in Aus ia o 45.4% in Sweden; in Li huania
he maximum sha e o 13.7% o Ho izon 2020 unding add esses his SGC (see Table A.1 in he
annex).
8 See h p://www.knowmak.eu and h ps://ga e.ac.uk/p ojec s/knowmak/ - Final e sion o he KNOWMAK on ology
e sion 1 o he lis o keywo ds [downloaded 28 July 2021].
9 A limi a ion o his me hod lies in he dependence on he quali y and deg ee o de ail o ERDF p ojec names and
desc ip ions as well as po en ially linguis ic aspec s. On he one hand, p ojec names and desc ip ions in he Kohesio
da abase a e ansla ed o English om o iginal lis s o ope a ions o en p o ided in na ional languages, which
migh igge he usage o ce ain wo ds in he ex s. On he o he hand, ce ain (key)wo ds migh be mo e o en
used in ce ain languages han in o he s. Synonyms as well as wildca ds in he lis o keywo ds applied should in
pa s ackle ha issue.
14
Table 1. Sha e o ERDF R&I (assigned o a leas one SGC) and Ho izon 2020 unding by SGC
Ca ego y
ERDF R&I Ho izon 2020*
Bio echnology
12% 13%
Ene gy
15% 19%
Secu i y
5% 6%
T anspo
15% 24%
Clima e
15% 12%
Heal h
16%
22%
Socie y
22% 4%
Basis: To al eligible expendi u e o ERDF p ojec s
assigned o a leas one SGC (coun ed mo e han once
i assigned o mo e han one SGC)
111.2 bn EUR
(wi hou mul iple
coun ing: 58 bn EUR)
Basis: To al H2020 g an
21.4 bn EUR
Sou ce: CORDIS da abase, Kohesio da abase, own elabo a ions.
No e: ERDF R&I ep esen s he o al eligible expendi u e assigned o R&I p ojec s co- unded by he ERDF and assigned o
a leas one SGC. Amoun s a e, e.g., double-coun ed i he p ojec is assigned wo SGCs. 47% o he 148,000 ERDF R&I
p ojec s a e assigned only one SGC, 37% o hem a e assigned wo SGCs. 10% o ERDF R&I p ojec s in he sample
co espond o 3 SGCs, 4% o 4 SGCs, 2% o 5 and mo e. Ho izon 2020 con ains he espec i e unding amoun . *Sha es
do no add up o 100% because pe cen age sha es a e ounded.
Dis ibu ion o Ho izon 2020 and ERDF R&I unding ac oss EU-27 NUTS-3 egions
Figu e 1 shows he dis ibu ion o Ho izon 2020 and ERDF R&I unding ac oss NUTS-3 egions in he
EU-27. Regions colou ed in g ey a e excluded om he subsequen analysis: Fi s , ERDF p ojec da a
o G eece, I eland and Mal a is only a ailable o NUTS-2 (no NUTS-3) egions. Second, egions a e
excluded om he analysis o syne gies i hey do no ecei e unding o R&I p ojec s om bo h
ins umen s. Fo example, ERDF R&I unding is epo ed o 72 o he 73 Polish NUTS-3 egions,
whe eas 25 Polish egions did no a ac any Ho izon 2020 unding. Thus, he subse o 47 Polish
NUTS-3 egions ha ha e ecei ed bo h ERDF R&I and Ho izon 2020 unding is conside ed in his
analysis. In sum, his esul s in 806 ou o 1,166 NUTS-3 egions being co e ed in he analysis. See
Table A.2 in he annex o an o e iew o he numbe o NUTS-3 egions ha did no ecei e
Ho izon 2020 and/o ERDF R&I unding by Membe S a e.
While he igh -hand side igu e (1b) mi o s he ocus o ERDF unding on Eas e n and Sou he n
Eu opean Membe S a es also when ocusing on cohesion policy- unded R&I ac i i ies, Figu e (1a)
shows ha compa a i ely la ge amoun s o Ho izon 2020 unding is a ac ed o Scandina ian and
o he No he n Eu opean egions (such as in Denma k o he Ne he lands) as well as capi al egions
(such as Mad id o Île de F ance). Fu he mo e, Figu e 1 poin s o he ac ha he egional unding
amoun pe capi a in he 806 NUTS-3 egions a ac ing bo h unds is ema kably la ge o he
ERDF (mean 167 EUR pe capi a, median 66 EUR pe capi a) han o Ho izon 2020 (mean 34 EUR
pe capi a, median 10 EUR pe capi a). 10
10 No e ha we conside unding in he 2014-2020 p og amming pe iod as a whole and no yea by yea in o de o
cap u e he o e all simila i y be ween he wo unding p og ams. The p og ams un a di e en speed (o
15
Figu e 1. Dis ibu ion o Ho izon 2020 and ERDF R&I amoun s pe capi a among NUTS-3 egions
(1a) Ho izon 2020 unding pe capi a in € (1b) To al eligible expendi u e o ERDF R&I
p ojec s (2014-2020) pe capi a in €
Sou ce: CORDIS da abase, Kohesio da abase, own elabo a ions.
No e: ERDF p ojec da a o G eece, I eland and Mal a is no p o ided a he NUTS-3 egional le el, he e o e hese egions
a e excluded om he analysis. The 2021 e sion o he Nomencla u e o Te i o ial Uni s o S a is ics (NUTS) is used.
An analysis o he syne gies be ween he wo unding ins umen s is expec ed o be mo e
meaning ul i ex eme cases a e excluded. E.g., he NUTS-3 egion o Pa is is among he op
ecei e s o Ho izon 2020 unding pe capi a and a he same ime pa o he bo om decile when
i comes o ERDF R&I unding pe capi a ecei ed. O e all, 20% o he NUTS-3 egions in he sample
(158 o 806) ecei e mo e Ho izon 2020 han ERDF R&I unding. The e a e signi ican di e ences in
he Ho izon 2020/ERDF R&I a io ac oss – in declining o de - p edominan ly u ban, in e media e
and u al a eas (see Table A.3 in he annex). Di e ences a e e en mo e p onounced when
compa ing NUTS-3 egions pa o mo e de eloped and less de eloped egions (Figu e 2) as well as
ac oss NUTS-2 egions classi ied acco ding o he egional inno a ion sco eboa d (Figu e A.1).
abso p ion) a EU, na ional and egional le el o ha e a ha monized indica o ha is no biased by p og am
cha ac e is ics and hei implemen a ion (San os e al., 2025).
16
Figu e 2. Ho izon 2020/ERDF R&I a io in di e en g oups o NUTS-3 egions based on hei economic
de elopmen
Sou ce: CORDIS da abase, Kohesio da abase, own elabo a ions.
No e: The igu e shows he ela ionship be ween Ho izon 2020 and ERDF R&I unding in 806 NUTS-3 egions ha ecei e
unding om bo h ins umen s. Due o da a es ic ions, G eece, I eland and Mal a a e no conside ed in he analysis.
3.2. Measu ing syne gies using he cosine simila i y index
To quan i y he deg ee o alignmen be ween he hema ic p io i ies o ERDF R&I in es men s and
Ho izon 2020 p ojec ac i i ies ac oss NUTS-3 egions, we employ he cosine simila i y. Cosine
simila i y measu es he simila i y be ween wo ec o s in e ms o he di e ence be ween hei
angles and is de ined as:
1 bn
2 bn
3 bn
4 bn
5 bn
17
cos(𝜃𝜃)=
𝐀𝐀∙𝐁𝐁
‖𝐀𝐀‖‖ 𝐁𝐁‖=
∑𝐴𝐴
𝑖𝑖
𝐵𝐵
𝑖𝑖
𝑛𝑛
𝑖𝑖=1
�∑𝐴𝐴𝑖𝑖
2
𝑛𝑛
𝑖𝑖=1 ∑𝐵𝐵𝑖𝑖2
𝑛𝑛
𝑖𝑖=1
(1)
whe e A and B a e he ec o s ep esen ing he ERDF R&I and Ho izon 2020 unding amoun s pe
SGC o a single egion and 𝑖𝑖 i e a es o e he se en SGCs. In con as o he Euclidean dis ance, he
cosine simila i y is always no malized be ween -1 and 1 since he absolu e magni ude o he
ec o s does no play a ole in hei angle. No e ha in ou case, he e a e no nega i e alues due o
he s ic ly posi i e unding amoun s alloca ed o NUTS-3 egions (ne alues). This means he
simila i y measu e will always be be ween 0 and 1.
This me ic is ad an ageous in si ua ions whe e he p ima y in e es lies in he p opo ional
alignmen a he han in absolu e olumes. Compa ed o co ela ion coe icien s (such as
Pea son’s), which a e p ima ily designed o de ec linea ela ionships in magni ude, cosine
simila i y compa es he di ec ion o wo ec o s in a mul idimensional hema ic space, i espec i e
o scale. Meaning ha co ela ion coe icien s a e sensi i e o co- a ia ion in magni ude and low i
he e is no linea ela ionship p esen ac oss SGCs, while cosine simila i y is high i he SGCs a e
p io i ized simila ly, e en i he ERDF and Ho izon 2020 ec o s di e in magni ude o
skewness.11Simila i y in his con ex means he simila i y o he dis ibu ion o ERDF and Ho izon
2020 unding a ac ed o alloca ed o a NUTS-3 egion in e ms o he sha es a ibu ed o
di e en SGCs. As inancial a iable o in e es , he o al eligible expendi u e alloca ed o ERDF R&I
p ojec s, including na ional co- unding, is used. In he case o Ho izon 2020 p ojec s, which do no
equi e na ional co- unding, he espec i e ne EU con ibu ion is conside ed. The cosine simila i y
index calcula ed o a NUTS-3 egion has a alue o 1 i espec i e ERDF R&I and Ho izon 2020
unding amoun s a e dis ibu ed exac ly he same ac oss SGCs. A alue o 0 indica es no o e lap in
he di ec ionali y o unding a all, i.e., o hogonal ec o s.
The cosine simila i y index canno only be used o con as unding pa e ns in NUTS-3 egions, bu
also allows o compa e he unding dis ibu ion o ERDF and Ho izon 2020 in g oups o NUTS-3
egions. The e o e, i is calcula ed o i) all (p edominan ly) u ban, u al o in e media e egions, ii)
NUTS-3 egions pa o NUTS-2 egions o di e en economic de elopmen le el (less de eloped,
ansi ion and mo e de eloped egions in he 2014-2020 p og amming pe iod), and iii) NUTS-3
egions loca ed in NUTS-2 egions wi h a di e en inno a ion pe o mance acco ding o he egional
inno a ion sco eboa d (Eu opean Commission, 2021).
As u he me ic o in e es , we calcula e he Gini index as a measu e o concen a ion o EU R&I
unding on speci ic SGCs using he ollowing equa ion:
𝐺𝐺𝑖𝑖𝐺𝐺𝑖𝑖 = ∑ ∑ �𝑥𝑥𝑖𝑖−𝑥𝑥𝑗𝑗�
𝑛𝑛
𝑗𝑗=1
𝑛𝑛
𝑖𝑖=1 2𝑛𝑛2𝑥𝑥 (2)
11 Compa ed o using loca ion quo ien s, o he concep o e ealed compa a i e ad an age, o assess syne gies (such
as in Doussineau and Bach ögle , 2021), he ad an age o he cosine simila i y index lies no only in he
independence o i s in e p e a ion om absolu e amoun s o unding, bu also in no need o choosing a benchma k
alue and g oup o egions acco ding o which a specialisa ion o unding is iden i ied. Also, loca ion quo ien s would
be calcula ed o each SGC sepa a ely and addi ional assump ions would be necessa y o build an agg ega e
measu e o syne gies a he NUTS-3 egional le el.
18
whe e x is he ec o o he o al unding amoun s (ERDF R&I o Ho izon 2020) pe SGC o a single
NUTS-3 egion and i and j i e a e o e he se en SGCs. The esul ing Gini coe icien is hen
no malized o alues be ween 0 and 1.
3.3. Reg ession analysis conside ing NUTS-3 socio-economic
cha ac e is ics
To explo e po en ial de e minan s o he syne ge ic use o EU R&I unding in NUTS-3 egions, a
s anda d OLS eg ession analysis is employed. Equa ion (3) depic s he model o be es ima ed,
𝑦𝑦𝑖𝑖,𝑟𝑟=𝛽𝛽0+𝑋𝑋𝑖𝑖,𝑟𝑟𝛽𝛽+ 𝛾𝛾𝑟𝑟+𝜀𝜀𝑖𝑖,𝑟𝑟 (3)
whe e yi, is he cosine simila i y index o NUTS-3 egion i in NUTS-2 egion , Xi, is a 1 × k ec o
o explana o y a iables and β he co esponding ec o o coe icien s. β0 deno es he in e cep
and εi, is he e o e m. As we a e pa icula ly in e es ed in he ela ionship be ween he alignmen
o ERDF R&I and Ho izon 2020 unding and s uc u al cha ac e is ics a wi hin NUTS-2 egions,
eg ession esul s a e p o ided (wi hou and) conside ing NUTS-2 egional ixed e ec s (γ ) in o de
o accoun o unobse ed di e ences ac oss NUTS-2 egions.
Gi en ha he dependen a iable is bounded be ween 0 and 1, a (quasi-likelihood) ac ional
logis ic eg ession app oach is applied as an al e na i e speci ica ion o check o he obus ness o
he esul s. Fu he mo e, eg ession esul s a e p o ided o a ‘ immed da ase ’ in which we
exclude NUTS-3 egions wi h he la ges disc epancies be ween Ho izon 2020 and ERDF R&I
unding ecei ed o ensu e he obus ness o indings.
The se o explana o y a iables consis s o a se o socio-economic cha ac e is ics a he NUTS-3
egional le el ha a e expec ed o ma e o he usage o EU R&I unding (see Table 2), ollowing
he de e minan s o he geog aphy o inno a ion (see e.g. Po e , 1998; Cice one e al., 2023;
Ma ques San os e al., 2025) and he agglome a ion economy (K ugman, 1991; O a iano and
Puga, 1998; Ba be o e al., 2025). In gene al, p e-p og amming pe iod alues o 2013 a e
conside ed o a oid e e se causali y bias o unds’ e ec s on ou explana o y a iables.
The numbe o SGCs mapped agains policy objec i es se as S3 p io i ies o he 2014-2020
p og amming pe iod is used as an indica o ha is expec ed o shape he hema ic o ien a ion o
R&I unding. As S3 p io i ies a e mos ly de ined a he NUTS-2, pa ly a he NUTS-1, and na ional
le el (only Finland and Sweden se p io i ies o NUTS-3 egions; he Czech S3 is na ional wi h
egional annexes o pa icula NUTS-3 egions), hey we e conside ed as ele an o all
subo dina ed NUTS-3 egions. Fo a mo e di e en ia ed analysis, baseline esul s conside he
numbe o SGCs add essed excluding na ional S3 p io i ies.12 No e ha his da a analysis e ealed
ha S3 policy objec i es a e de ined e y b oadly: 454 (o 56%) o NUTS-3 egions in ou sample
had all SGCs add essed as p io i ies.
12 No e ha S3 p io i ies o EU Membe S a es consis ing o one NUTS-2 egion (Cyp us, Es onia, Luxembou g, La ia,
Mal a) we e conside ed as se a he egional (no na ional) le el. Da a on S3 p io i ies was downloaded om he
Eye@RIS3 da abase (h ps://s3pla o m.j c.ec.eu opa.eu/map), adap ed o i he NUTS 2021 classi ica ion and
assigned o SGCs (which is i ial in mos cases).
19
Table 2. Explana o y a iables a he NUTS-3 egional le el: a iable desc ip ion and sou ce
Name
Desc ip ion
Sou ce
Capi al egion
Dummy a iable equal o 1 i he
espec i e coun y’s capi al is
loca ed in he NUTS-3 egion; 0
o he wise
Own elabo a ion based on ci y
loca ion wi hin NUTS-3 egion.
Uni e si y13
Dummy a iable equal o 1 i
he e is an uni e si y in he
egion; 0 o he wise
Own elabo a ion based on da a
om The Eu opean Highe
Educa ion Sec o Obse a o y
(2022)
LQ Knowledge-in ensi e se ices
Concen a ion o knowledge-
in ensi e se ices es ima ed
using he loca ion quo ien (LQ),
i.e. he sha e o g oss alue
added (GVA) in NACE sec o s J-N
in egion 𝑖𝑖 o e he sha e o GVA
in sec o s J-N in he EU-27
Own elabo a ion based on
ARDECO (SOVGZ)
GDP pe capi a (ln)
Real GDP pe capi a, EUR2015, by
inhabi an , in logs (ln)
Own es ima ion based on
ARDECO (GDP - SOVGD;
Popula ion - SNPTD)
In es men a e
Ra io o G oss Fixed Capi al
Fo ma ion (GFCF) o e G oss
Domes ic P oduc (GDP)
Own es ima ion based on
ARDECO (GFCF - ROIGT; GDP -
SOVGD)
Compe i ion
Compe i ion le el = 1 –
He indahl-Hi schman-Index (10
NACE sec o s)
Own es ima ion based on
ARDECO (SOVGZ)
Popula ion densi y (ln)
Inhabi an pe km
2
, in logs (ln)
Own es ima ion based on
ARDECO (Popula ion - SNPTD)
and EUROSTAT (a ea - eg_a ea3)
No. o SGCs in S3 p io i ies (excl.
na ional p io i ies)
Numbe o SGCs mapped agains
policy objec i es se as pa o S3
p io i ies o he 2014-2020
p og amming pe iod.
No e ha his a iable is
a ailable a NUTS-3 le el only o
Sweden and Finland as well as
ce ain Czech NUTS-3 egions. Fo
he o he Membe S a es, S3
p io i ies a e se a he NUTS-2,
NUTS-1 and/o na ional le el and
conside ed as such o
subo dina ed NUTS-3 egions.
Eye@RIS3 da abase, own
elabo a ion
Sou ce: own elabo a ion. Summa y s a is ics a e p o ided in Table B.1 in he annex.
13 Fo a obus ness check, we conside he numbe o uni e si ies and he numbe o s uden s (ISCED le els 5-7),
espec i ely, pe NUTS-3 egion in 2013 (see Tables B.5 and B.6 in he annex).
20
4. Resul s
4.1. S ylized ac s on he dis ibu ion o ERDF R&I and Ho izon 2020
unding ac oss SGCs
Figu e 3 shows he dis ibu ion o he EU’s wo main R&I unding ins umen s in di e en g oups o
NUTS-3 – and co esponding NUTS-2 – egions. As al eady discussed in he Da a sec ion, ERDF
p ojec s a e mo e s ongly ocused on he SGC Socie y, whe eas Ho izon 2020 p ojec s a e in
gene al mo e a ge ed a T anspo and Ene gy. Compa ing especially Ho izon 2020 unding sha es
ac oss g oups o egions, howe e , e eals some pa e ns: The unding sha es o Ho izon 2020 o
T anspo and Heal h decline wi h he le el o de elopmen as well as wi h inno a ion pe o mance.
By con as , he sha e alloca ed o Bio echnology is highes o NUTS-3 egions loca ed in ansi ion
and eme ging inno a o egions, espec i ely. Fo ERDF unding (in he ield o R&I), sys ema ic
di e ences ac oss g oups o egions appea o be less p onounced. All in all, hese s ylized ac s
mo i a e an analysis o socio-economic indica o s ha con ibu e o he ex en o alignmen in he
hema ic o ien a ion o he wo unding schemes a NUTS-3 le el.
Figu e 3. Funding sha es alloca ed o di e en SGCs in NUTS-3 egions pa o …
Con inued on he nex page …
21
Sou ce: CORDIS da abase, Kohesio da abase, own elabo a ions.
No e: Fi s classi ica ion based on he economic de elopmen le el o co esponding NUTS-2 egions in he 2014-2020
p og amming pe iod. Second classi ica ion o co esponding NUTS-2 egions based on egional inno a ion sco eboa d
(Eu opean Commission 2021).
The S3 concep mo i a es egions o ocus hei unding on speci ic s a egic a eas, such as policy
objec i es. The e o e, he Gini index, anging om one (only one SGC add essed) o ze o (equal
dis ibu ion o unding ac oss he se en SGCs), is calcula ed o assess concen a ion o ERDF R&I
and Ho izon 2020 unding on a numbe o SGCs in NUTS-3 egions.
The Gini index indica es high a ia ion in he concen a ion o egional EU R&I unding among NUTS-
3 egions and as well ac oss unding ins umen s (see Table A.4). Ho izon 2020 unding is, on
a e age, much mo e ocused on ce ain SGCs han ERDF R&I unding ( he mean o he Gini o ERDF
p ojec s is 0.54, o Ho izon 2020 p ojec s 0.77). While his is pa ly d i en by design (ERDF p ojec s
can be a ibu ed o mo e han one SGC, see no es o Table 1), he highe concen a ion o Ho izon
2020 unding is no su p ising gi en i s excellence-based alloca ion mechanism.
Ano he con ibu ing ac o may be ha al eady S3 p io i ies in sma specializa ion s a egies ha
go e n ERDF alloca ions in he ield o R&I, i.e., policy objec i es, a e chosen e y b oadly: Mo e han
hal o he NUTS-3 egions in he sample chose S3 policy objec i es ha co espond o all he se en
SGCs. In his espec , ecen li e a u e inds o echnological domains ha ac ual R&I unding is
mo e selec i e (in e ms o he ela edness o egional capabili ies) han S3 p io i ies se in sma
specializa ion s a egies (Hellinga e al., 2025). This mo i a es o check whe he he numbe o
SGCs p io i ized h ough S3 p io i ies con ibu ed o he (dis)simila i y in he dis ibu ion o EU R&I
unding in (di e en g oups o ) NUTS-3 egions.
In e es ingly, p edominan ly u al NUTS-3 egions pe o m be e in channeling bo h hei ERDF R&I
and Ho izon 2020 unding o speci ic SGCs han in e media e and e en mo e han u ban a eas (see
22
Table A.4), which will pa ially be due o a b oade a ie y o ypes o o ganiza ions (companies,
esea ch ins i u es, uni e si ies, e c.) and expe ise o ac o s loca ed in mo e densely popula ed
egions. A simila pa e n is ound o Ho izon 2020 unding in g oups o NUTS-3 egions by
de elopmen le el and inno a ion pe o mance o co esponding NUTS-2 egions: concen a ion in
less de eloped and ansi ion egions is highe han in mo e de eloped ones, and concen a ion in
egions classi ied as eme ging inno a o s is signi ican ly highe han in inno a ion leade s and
s ong inno a o s. Howe e , he di e ences in mean alues ac oss g oups o egions a e ela i ely
small (Table A.4).
By con as , concen a ion o ERDF R&I unding on ce ain SGCs is signi ican ly highe in mo e
de eloped and ansi ion egions as compa ed o less de eloped ones. Also, he deg ee o
concen a ion o inc eases ema kably (also in e ms o mean alues) wi h imp o ing inno a ion
pe o mance.
4.2. The (dis)simila i y o he dis ibu ion o EU R&I unding among SGCs
To explo e he syne ge ic use o he wo main EU R&I unding ins umen s in e ms o add essing
he same socie al challenges, he cosine simila i y index measu ing he alignmen o Ho izon 2020
and ERDF R&I unding is calcula ed o he le el o NUTS-3 egions. This index amoun s o one i
he wo unding ins umen s a e alloca ed o ha e been a ac ed - o add ess he same SGCs in he
same p opo ions. A alue o ze o indica es no o e lap a all.
The cosine simila i y index a ies ema kably ac oss he 806 Eu opean NUTS-3 egions ha
ecei ed EU R&I unding om bo h ERDF and Ho izon 2020 (see Figu e 4). The a e age cosine
simila i y index amoun s o 0.51 ( he median is 0.55, see Table 3). 46 o he 806 NUTS-3 egions
do no o ien hei EU R&I unding owa ds he same SGCs a all (mos o hem a e pa o mo e
de eloped NUTS-2 egions), while he dis ibu ion o EU R&I unding ac oss SGCs is ully aligned in
wo u al NUTS-3 egions in Ge many.14
Table 3 p o ides summa y s a is ics on he di e ences o he simila i y o he dis ibu ion o ERDF
R&I and Ho izon 2020 unding ac oss g oups o NUTS-3 egions. Fi s , he alignmen o he wo
unding ins umen s in e ms o SGCs add essed is ound o be signi ican ly highe in p edominan ly
u ban egions han in in e media e and, e en mo e, u al egions. In he p e ious sec ion we
discussed he s ylized ac ha concen a ion o bo h unding ins umen s on ce ain SGCs is
ela i ely low in u ban a eas, which is likely o be explained by he ac ha he numbe and a ie y
o bene icia ies and hei expe ise is highe in mo e densely popula ed egions. The e o e, he
inding ha bo h unding s eams a e ne e heless used in a mo e simila ashion in u ban a eas,
indeed poin s o syne gies in he a ac ion o R&I p ojec g an s. This could be suppo ed by a good
inno a ion ecosys em and ne wo k, po en ially cen e ed a ound a uni e si y, esea ch ins i u es o
inno a i e companies.
14 I NUTS-3 egions ha ecei e mos (10 h decile) and lowes (1s decile) o ERDF and Ho izon 2020 unding,
espec i ely, a e excluded om he sample (518 NUTS-3 egions le ), he a e age cosine simila i y amoun s o 0.53
(median 0.56). In his case, he cosine simila i y index amoun s o 0 only in 20 NUTS-3 egions; he wo egions wi h
he o ally simila unding dis ibu ion emain in he sample.
29
Table 6. Co espondence be ween he cosine simila i y index and socio-economic cha ac e is ics in NUTS-3 egions by inno a ion g oup
OLS eg ession by
inno a ion g oup
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
(12)
VARIABLES
Eme ging inno a o
Mode a e inno a o
S ong inno a o
Inno a ion leade
Capi al egion
-0.212
-0.196
-0.311
0.159
0.120
0.085
-0.220*
-0.197
-0.140
-0.008
0.763***
0.475**
(0.315)
(0.389)
(0.364)
(0.255)
(0.228)
(0.235)
(0.122)
(0.223)
(0.205)
(0.134)
(0.200)
(0.178)
Uni e si y
0.148**
0.187*
0.192**
0.115**
0.103**
0.106**
0.075*
0.074
0.096**
0.045
0.147*
0.160*
(0.064)
(0.098)
(0.090)
(0.046)
(0.050)
(0.048)
(0.044)
(0.049)
(0.044)
(0.071)
(0.080)
(0.081)
LQ knowledge-in ensi e
-0.069
-0.312
-0.132
0.781
0.983*
0.705
0.554
0.742
0.628
0.765***
0.937**
0.661
se ices (KIS)
(0.421)
(0.584)
(0.499)
(0.519)
(0.579)
(0.576)
(0.399)
(0.535)
(0.508)
(0.237)
(0.414)
(0.420)
LQ KIS squa ed
0.375
0.357
0.309
-0.399
-0.526
-0.333
-0.282
-0.414
-0.310
-0.202***
-0.256*
-0.116
(0.354)
(0.437)
(0.395)
(0.314)
(0.343)
(0.330)
(0.172)
(0.272)
(0.250)
(0.064)
(0.142)
(0.138)
ln_GDP pe capi a
0.045
0.199
0.163
0.187
0.172
0.217
0.540***
0.454**
(0.200)
(0.258)
(0.116)
(0.133)
(0.134)
(0.154)
(0.136)
(0.221)
In es men a e
1.186
4.079
5.088***
5.384***
1.016
1.860
1.367
2.996
(1.398)
(2.872)
(1.941)
(1.989)
(2.838)
(3.077)
(2.270)
(2.979)
Compe i ion le el
-0.571
-0.063
-1.201
0.460
0.590
0.546
-0.475
-0.212
-0.747
0.332
-0.807
-1.665*
(1.169)
(1.377)
(1.043)
(0.880)
(0.944)
(0.827)
(1.142)
(1.058)
(1.002)
(0.776)
(1.092)
(0.871)
ln_Popula ion densi y
-0.002
0.048
0.046
0.024
0.013
0.037
0.060*
0.060*
0.074***
-0.118***
-0.202***
-0.123**
(0.044)
(0.055)
(0.054)
(0.030)
(0.033)
(0.031)
(0.031)
(0.031)
(0.028)
(0.042)
(0.054)
(0.051)
No. o SGC in S3 p io i ies
-0.037
-0.102*
-0.056
0.040
-0.016
-0.011
-0.048
-0.056
-0.053
-0.079***
-0.062**
-0.068**
(excl. na ional S3 p io i ies)
(0.038)
(0.057)
(0.047)
(0.046)
(0.015)
(0.015)
(0.046)
(0.049)
(0.047)
(0.026)
(0.028)
(0.027)
Obse a ions
169
114
114
236
214
214
274
242
242
127
75
75
NUTS-2 FE
YES
YES
YES
YES
YES
YES
YES
YES
YES
YES
YES
YES
TRIMMED da ase
NO
YES
YES
NO
YES
YES
NO
YES
YES
NO
YES
YES
R-squa ed
0.592
0.710
0.692
0.492
0.495
0.470
0.378
0.407
0.398
0.482
0.575
0.516
Adjus ed R-squa ed
0.328
0.414
0.399
0.263
0.237
0.211
0.160
0.169
0.166
0.258
0.301
0.239
VIF
7.50
11.91
8.60
19.37
21.20
19.95
6.47
6.83
5.29
9.61
8.54
7.49
VIF > 10 (excl. squa ed
e ms)
S3 p ios,
ln_GDP_pc,
In a e,
LQ_KIS
S3 p ios,
ln_GDP_pc,
In a e,
LQ_KIS,
ln_PopDen
s
S3 p ios,
ln_PopDen
s (10.6)
S3 p ios,
In Ra e,
ln_GDP_pc
S3 p ios,
In Ra e,
ln_GDP_pc
S3 p ios
S3 p ios
In Ra e
S3 p ios
In Ra e
S3 p ios
-
-
-
No es: The dependen a iable is he cosine simila i y index. Signi icance le el: *** p < 0.01, ** p<0.05, * p<0.1. He e oskedas ici y- obus s anda d e o s (clus e ed a he NUTS-3 egional
le el) a e in pa en heses. The cons an as well as he speci ica ion excl. he no. o SGC in S3 p io i ies is no epo ed he e. Fo he la e , as in baseline esul s, coe icien s o he o he
a iables do only change ma ginally.
As discussed in p e ious sec ions, di e ences in he cosine simila i y index ac oss g oups o egions
seem o be linked o he concen a ion o unding. As i would induce endogenei y issues, i is no
possible o include he Gini coe icien as a measu e o concen a ion in he eg essions, hough.
The e o e, we un a seemingly un ela ed eg ession (SUR) model o all obse a ions as well as
subse s o he da a, wi h he cosine simila i y index and he Gini indices measu ing he
concen a ion o ERDF R&I unding and Ho izon 2020 unding, espec i ely, as dependen a iables
ha a e join ly de e mined by egional cha ac e is ics.
Table 7 p esen s he esul s o he SUR model. Wha s ikes ou is ha he s a is ically signi ican
posi i e ela ionship be ween he simila i y o EU R&I unds usage and he p esence o a uni e si y
in he NUTS-3 egion goes along wi h, a he same ime, a signi ican nega i e ela ionship be ween
uni e si y loca ion and he concen a ion o bo h unding ins umen s on ce ain SGCs. A i s sigh ,
mo e alignmen h ough a hema ically b oade unds usage appea s no o mi o he idea o S3.
Howe e , gi en he di e en objec i es and alloca ion p inciples o Ho izon 2020 and ERDF unding,
i seems mo e in ui i e ha NUTS-3 egions able o a ac unding in mo e policy a eas each ace
mo e po en ial o syne gies.
In he same ashion, when conside ing a iables ha we e de-meaned a he NUTS-2 egional le el
( o conside NUTS-2 ixed e ec s in his se ing), a highe specializa ion in knowledge-in ensi e
se ices is ound o be associa ed wi h less concen a ion o unding and, a he same ime, a be e
alignmen in he hema ic o ien a ion o he wo unds.
Resul s o he SUR model un o di e en g oups o egions and obus ac oss all speci ica ions
(Tables B.3 and B.4) indica e in line wi h OLS esul s in Table 5 ha he specializa ion in knowledge-
in ensi e se ices plays a ole in NUTS-3 pa o less de eloped egions, while he p esence o a
uni e si y ma e s in ansi ion and mo e de eloped egions. Fu he mo e, a highe numbe o SGCs
add essed ia S3 p io i ies (policy objec i es) is s a is ically signi ican ly associa ed wi h less
alignmen o he unding ins umen s in mo e de eloped egions, when conside ing a iables de-
meaned by NUTS-2 a e ages (in he speci ica ion wi hou de-meaned a iables, i is s a is ically
signi ican ly nega i ely linked wi h alignmen in less de eloped egions). By inno a ion g oup, he
esul ha he p esence o a uni e si y co esponds o less concen a ion, bu be e hema ic
alignmen holds ac oss all g oups (using he immed da ase ). In he case o inno a ion leade s,
he same esul is ound o a highe specializa ion in knowledge-in ensi e indus ies, e en i he e
is a maximum specializa ion le el up om which no be e alignmen is achie ed.
31
Table 7. Seemingly un ela ed eg ession (SUR) esul s
All obse a ions
T immed da ase
VARIABLES
Cosine_sim
Gini_Ho izon
Gini_ERDF
Cosine_sim
Gini_Ho izon
Gini_ERDF
Capi al egion
0.124**
-0.282***
-0.159***
0.166**
-0.286***
-0.151***
(0.059)
(0.036)
(0.045)
(0.077)
(0.047)
(0.057)
Uni e si y
0.134***
-0.154***
-0.122***
0.144***
-0.160***
-0.138***
(0.020)
(0.012)
(0.015)
(0.022)
(0.013)
(0.016)
LQ knowledge-in ensi e
-0.024
-0.090
-0.162**
0.069
-0.077
-0.324***
se ices (KIS)
(0.104)
(0.063)
(0.079)
(0.145)
(0.089)
(0.108)
LQ KIS squa ed
-0.009
0.021
0.069**
-0.058
0.012
0.125**
(0.043)
(0.026)
(0.033)
(0.069)
(0.042)
(0.051)
ln_GDP pe capi a
0.025
-0.013
0.111***
0.017
0.007
0.089***
(0.020)
(0.012)
(0.015)
(0.024)
(0.015)
(0.018)
In es men a e
-0.432
0.063
1.033***
-0.397
0.153
1.058***
(0.267)
(0.161)
(0.203)
(0.294)
(0.181)
(0.219)
Compe i ion le el
0.536*
-0.696***
0.030
0.744**
-0.745***
-0.128
(0.275)
(0.166)
(0.209)
(0.337)
(0.207)
(0.251)
ln_Popula ion densi y
0.019*
-0.015**
0.009
0.015
-0.016**
0.016*
(0.010)
(0.006)
(0.007)
(0.011)
(0.007)
(0.008)
No. o SGC in S3 p io i ies
-0.008
0.005
0.005
-0.011*
0.004
0.009**
(0.005)
(0.003)
(0.004)
(0.006)
(0.003)
(0.004)
Obse a ions
806
806
806
645
645
645
R-squa ed
0.104
0.376
0.193
0.111
0.338
0.216
De-meaned a iables
VARIABLES
Cosine_sim
Gini_Ho izon
Gini_ERDF
Cosine_sim
Gini_Ho izon
Gini_ERDF
Capi al egion
-0.023
0.023
0.041
0.048
0.021
0.154**
(0.087)
(0.048)
(0.053)
(0.115)
(0.064)
(0.068)
Uni e si y
0.094***
-0.079***
-0.087***
0.111***
-0.085***
-0.095***
(0.020)
(0.011)
(0.012)
(0.022)
(0.012)
(0.013)
LQ knowledge-in ensi e
0.324***
-0.495***
-0.373***
0.335**
-0.469***
-0.601***
se ices (KIS)
(0.123)
(0.068)
(0.074)
(0.160)
(0.089)
(0.094)
LQ KIS squa ed
-0.087*
0.143***
0.095***
-0.114
0.131***
0.186***
(0.045)
(0.025)
(0.027)
(0.069)
(0.039)
(0.041)
ln_GDP pe capi a
0.212***
-0.118***
-0.089***
0.188***
-0.125***
-0.085**
(0.054)
(0.030)
(0.032)
(0.065)
(0.036)
(0.038)
In es men a e
2.357***
-0.857**
-1.104**
3.184***
-1.172**
-1.286**
(0.779)
(0.429)
(0.469)
(0.942)
(0.526)
(0.552)
Compe i ion le el
0.111
-0.398**
0.233
0.178
-0.420*
0.439*
(0.328)
(0.181)
(0.198)
(0.387)
(0.216)
(0.227)
ln_Popula ion densi y
0.019
-0.048***
0.007
0.025
-0.049***
0.013
(0.014)
(0.008)
(0.008)
(0.016)
(0.009)
(0.009)
No. o SGC in S3 p io i ies
-0.043
-0.031**
0.002
-0.049
-0.029*
-0.010
(0.028)
(0.016)
(0.017)
(0.031)
(0.017)
(0.018)
Obse a ions
806
806
806
645
645
645
R-squa ed
0.176
0.459
0.212
0.178
0.458
0.261
No es: Signi icance le el: *** p<0.01, ** p<0.05, * p<0.1. S anda d e o s in pa en heses. The cons an is no epo ed o sa e space.
32
5. Conclusions
This s udy p o ides a no el quan i a i e app oach o assessing syne gies be ween Ho izon 2020
and cohesion policy (ERDF) unding in he ield o esea ch and inno a ion (R&I) du ing he 2014-
2020 p og amming pe iod. By le e aging p ojec -le el da a, we de elop a cosine simila i y index
based on socie al g and challenges (SGCs) o e alua e he hema ic alignmen be ween hese wo
majo EU unding ins umen s in EU NUTS-3 egions. Ou indings highligh signi ican dispa i ies in
syne gies ac oss di e en egional ypologies, wi h u al a eas, eme ging inno a o s, and less
de eloped egions exhibi ing lowe alignmen le els. The analysis unde sco es he c ucial ole o
egional inno a ion ecosys ems, pa icula ly he p esence o uni e si ies and specializa ion in
knowledge-in ensi e se ices, in os e ing s onge unding complemen a i ies.
The esul s also indica e ha a highe numbe o SGCs add essed in Sma Specializa ion S a egy
(S3) policy objec i es is nega i ely linked o hema ic unding simila i y. This sugges s ha an o e ly
b oad policy ocus may hinde e ec i e alignmen be ween Ho izon 2020 and ERDF R&I
in es men s. Mo eo e , he obse ed co-exis ence o highe simila i y and lowe concen a ion o
unding schemes implies ha Ho izon 2020 and ERDF R&I p ojec s o en add ess di e en SGCs
wi hin a egion, likely e lec ing he di e si y o ac o s in ol ed in he local inno a ion landscape.
C ucially, he s udy iden i ies obus pa e ns ac oss egional inno a ion capaci ies. The p esence o
a uni e si y ma e s o hema ic alignmen o EU R&I unding in mo e de eloped and ansi ion
egions, as well as e e y g oup o NUTS-3 egions when dis inguished by inno a ion pe o mance o
co esponding NUTS-2 egions. By con as , specializa ion in knowledge-in ensi e se ices is ound
o play a ole only o mode a e inno a o s and inno a ion leade s, and in less (and mo e)
de eloped egions, whe eby excessi e specializa ion appea s o ha e he opposi e e ec .
Fu he mo e, in less de eloped and ansi ion egions, as well as in s ong inno a o egions, a
highe popula ion densi y is linked o a mo e syne ge ic use o he wo unding ins umen s.
These insigh s con ibu e o ongoing policy discussions on imp o ing he s a egic alignmen o EU
R&I unding ins umen s. The e idence sugges s ha e ining S3 p io i ies o os e mo e a ge ed
specializa ions and s eng hening egional inno a ion ecosys ems, pa icula ly h ough uni e si ies
and knowledge-in ensi e indus ies, can enhance syne ge ic unding use. Fu u e esea ch should
u he explo e he mechanisms unde pinning hese ela ionships, pa icula ly in ligh o e ol ing EU
policy amewo ks beyond 2020.
33
6. Discussion
Se e al limi a ions o he analysis should be aken in o accoun . The esul s a e only alid o he
sample conside ed, meaning ha i should be conside ed ha one hi d o ERDF R&I p ojec s in he
en iched Kohesio da ase could no be assigned o a SGC based on p ojec names, desc ip ions, and
he lis o keywo ds used. Fu he mo e, limi a ions a ising om he ex mining app oach and
linguis ic aspec s (e.g., ce ain keywo ds used mo e o en in some na ional languages o ex
ansla ed o English) ha e o be kep in mind.
This esea ch does no e alua e he success ul implemen a ion o sma specializa ion policies in
egions. In o de o con ibu e o such an e alua ion, a ollow-up pape should analyze he indi idual
SGCs add essed by ERDF R&I and Ho izon 2020 p ojec s be o e he backg ound o S3 p io i ies
de ined in sma specializa ion s a egies. In addi ion, i would also be necessa y o s udy he
ou comes o he p ojec s, which equi es a se o ou come indica o s mi o ing p og ess in
add essing p e alen SGCs a he egional le el.
34
Re e ences
Almudi, I.; Fa as-Villa anca, F.; Fe nández-Má quez, C.M; Po s, J.; Vazquez, F.J. (2020). “Abso p i e
capaci y in a wo-sec o neo-Schumpe e ian model: a new ole o inno a ion policy”, Indus ial and
Co po a e Change, 29(2):507–531. h ps://doi.o g/10.1093/icc/d z052.
Bach ögle -Unge , J., A nold, E., Doussineau, M. and Reschenho e , P. (2021a), UPDATE: Da ase o
p ojec s co- unded by he ERDF du ing he mul i-annual inancial amewo k 2014-2020,
Publica ions O ice o he Eu opean Union, Luxembou g, ISBN 978-92-76-43735-2,
doi:10.2760/440159, JRC125008.
Bach ögle -Unge , J., Ma ques San os, A. and Con e, A. (2021b), ERDF BENEFICIARIES DATASET
2014-2020: AN OVERVIEW FOR POLICY-MAKERS, Eu opean Commission, JRC127403.
Ba be o, J., Diukano a, O., Gianelle, C., Salo i, S., & San oalha, A. (2024). “Technologically ela ed
di e si ica ion: One size does no i all Eu opean egions”, Resea ch Policy, 53(3), 104973.
Bo ás, S. and Schwaag Se ge , S. (2022). “The design o ans o ma i e esea ch and inno a ion
policy ins umen s o g and challenges: The policy-nes ing pe spec i e”, Science and Public Policy,
49(5): 659–672. h ps://doi.o g/10.1093/scipol/scac017.
Cappellano, F.; Molica, F. and Makkonen, T. (2024). “Missions and Cohesion Policy: is he e a
ma ch?”, Science and Public Policy, 51(3):360–374. h ps://doi.o g/10.1093/scipol/scad076.
Chou, T.C (2018). “The combina ion index (CI < 1) as he de ini ion o syne gism and o syne gy
claims”, Syne gy 7: 49-50. h ps://doi.o g/10.1016/j.syn es.2018.04.001.
Cice one, G.; Faggian, A.; Mon eso , S. and Ren occhini, F. (2023). “Regional a i icial in elligence and
he geog aphy o en i onmen al echnologies: does local AI knowledge help egional g een- ech spe-
cializa ion?”, Regional S udies 57, no. 2 (2023): 330-343.
Cince a, M. and San os, A. M. (2015). Inno a ion and Access o Finance – A Re iew o he Li e a u e,
Wo king Pape s TIMES² 2015-16, ULB - Uni e si e Lib e de B uxelles.
Combley, R. (2011). Camb idge Business English Dic iona y. Camb idge Uni e si y P ess.
C escenzi, R., de Blasio, G., & Giua, M. (2020). “Cohesion Policy incen i es o collabo a i e indus ial
esea ch: e alua ion o a Sma Specialisa ion o e unne p og amme”, Regional S udies, 54(10):
1341–1353. h ps://doi.o g/10.1080/00343404.2018.1502422.
D’Adda, D., Iacobucci, D., & Pe ugini, F. (2021). “Sma Specialisa ion S a egy in p ac ice: ha e egions
changed he alloca ion o S uc u al Funds?”, Regional S udies, 56(1), 155–170.
h ps://doi.o g/10.1080/00343404.2021.1890326.
Deegan, J., B oekel, T., & Fi ja , R. D. (2021). “Sea ching h ough he Hays ack: The Rela edness and
Complexi y o P io i ies in Sma Specializa ion S a egies”, Economic Geog aphy, 97(5): 497– 520.
h ps://doi.o g/10.1080/00130095.2021.1967739.
Deleidi, M. and Mazzuca o, M. (2021). "Di ec ed inno a ion policies and he supe mul iplie : An empi -
ical assessmen o mission-o ien ed policies in he US economy", Resea ch Policy, 50(2), 104151.
h ps://doi.o g/10.1016/j. espol.2020.104151
35
Di Ca aldo, M., Monas i io is, V., & Rod íguez‐Pose, A. (2021). “How ‘Sma ’ a e sma specializa ion
s a egies?”, Jou nal o Common Ma ke S udies, 60(5): 1272–1298.
h ps://doi.o g/10.1111/jcms.13156.
Doussineau, M. and Bach ögle -Unge , J. (2021). Explo ing Syne gies be ween EU Cohesion Policy
and Ho izon 2020 Funding ac oss Eu opean Regions: An analysis o egional unding concen a ion
on key enabling echnologies and socie al g and challenges, Publica ions O ice o he Eu opean Un-
ion. h ps://da a.eu opa.eu/doi/10.2760/218779.
D aghi, M. (2024). “The D aghi epo : A compe i i eness s a egy o Eu ope”. A ailable he e:
h ps://commission.eu opa.eu/ opics/s eng hening-eu opean-compe i i eness/eu-compe i i eness-
looking-ahead_en.
ECA (2022). Special Repo 23/2022: Syne gies be ween Ho izon 2020 and Eu opean S uc u al and
In es men Funds - No ye used o ull po en ial, Eu opean Cou o Audi o s. A ailable a :
h ps://www.eca.eu opa.eu/en/publica ions/SR22_23
Eu opean Commission (2011). Ex-an e Condi ionali y in Cohesion Policy, Thi d Mee ing o he
Condi ionali y Task Fo ce, 8 Ap il 2011.
Eu opean Commission (2014). Enabling syne gies be ween Eu opean S uc u al and In es men
Funds, Ho izon 2020 and o he esea ch, inno a ion and compe i i eness- ela ed Union
p og ammes. A ailable he e: h p://
ec.eu opa.eu/ egional_policy/sou ces/docgene /guides/syne gy/syne gies_en.pd .
Eu opean Commission (2016). EU unds wo king oge he o jobs & g ow h – Examples o
syne gies be ween he amewo k p og ammes o esea ch and inno a ion (Ho izon 2020) and he
Eu opean S uc u al and In es men Funds (ESIF), Di ec o a e-Gene al o Resea ch and Inno a ion,
Publica ions O ice, 2016, h ps://da a.eu opa.eu/doi/10.2777/678944.
Eu opean Commission (2021), Regional inno a ion sco eboa d 2021, Publica ions O ice.
h ps://da a.eu opa.eu/doi/10.2873/674111.
Eu opean Commission (2022). Communica ion o he Commission, App o al o he con en o a
d a Commission No ice on he syne gies be ween ERDF p og ammes and Ho izon Eu ope. B ussels,
5.7.2022, C(2022) 4747 inal.
Fo ay, D., Da id, P., and Hall, B. H. (2009). Sma Specialisa ion – The Concep . Knowledge
Economis s, Policy B ie Numbe 9, June. B ussels: Eu opean Commission, DG esea ch.
F a esi U., Gianelle, C. & Guzzo, F. (2021). Assessing Sma Specialisa ion: Policy Implemen a ion
Measu es, EUR 30758 EN, Publica ions O ice o he Eu opean Union, Luxembou g, 2021, ISBN 978-
92-76-40034-9, h ps://doi.o g/10.2760/71769, JRC123821.
Gianelle, C., D. Ky iakou, C. Cohen and M. P zeo (2016). “Implemen ing Sma Specialisa ion: A
Handbook”, B ussels: Eu opean Commission, EUR 28053 EN, doi:10.2791/53569.
Gianelle, C., Guzzo, F., & Mieszkowski, K. (2020). “Sma Specialisa ion: wha ge s los in ansla ion
om concep o p ac ice?”, Regional S udies, 54(10): 1377–1388.
h ps://doi.o g/10.1080/00343404.2019.1607970.
Hellinga, Z., Bach ögle -Unge , J., Balland, P., & Boschma, R. (2025). Beyond he Bluep in . F om Sma
Specializa ion S a egies o R&I Funding, Pape s in E olu iona y Economic Geog aphy (PEEG) 2502,
36
U ech Uni e si y, Depa men o Human Geog aphy and Spa ial Planning, G oup Economic Geog a-
phy, h p://econ.geo.uu.nl/peeg/peeg2502.pd .
Huggins, R. and Thompson, P. (2015). “En ep eneu ship, inno a ion and egional g ow h: a ne wo k
heo y”, Small Business Economics, 45:103–128 (2015). h ps://doi.o g/10.1007/s11187-015-9643-
3.
K ame , J. P. e al. (2021). S udy on p io i isa ion in Sma Specialisa ion S a egies in he EU.
Repo , Eu opean Commission, Luxembou g: Publica ions O ice o he Eu opean Union.
h ps://doi.o g/10.2776/60867.
K ugman, P. (1991). “Inc easing e u ns and economic geog aphy”, Jou nal o Poli ical Economy, 99,
no. 3: 483-499.
Ma ques San os, A. (2021). Linking he ‘Reco e y and Resilience Plan’ and Sma Specialisa ion. The
Po uguese case, JRC Wo king Pape s on Te i o ial Modelling and Analysis, No. 05/2021, a ailable
a : h ps://publica ions.j c.ec.eu opa.eu/ eposi o y/handle/JRC126178.
Ma ques San os, A.; Molica, F. and To ecilla-Salinas, C. (2025). “EU- unded in es men in A i icial
In elligence and egional specializa ion3, Regional Science Policy & P ac ice, 17(7).
h ps://doi.o g/10.1016/j. spp.2025.100190.
Ma ocu, E., Paci, R., Rigby, D. L., & Usai, S. (2022). “E alua ing he implemen a ion o Sma
Specialisa ion policy”, Regional S udies, 57(1), 112–128.
h ps://doi.o g/10.1080/00343404.2022.2047915.
Mayna d, D., Pe ak, J., Song, X. & Funk, A. (2019), KNOWMAK Repo on On ologies and Tagging,
a ailable a h ps://ga e.ac.uk/p ojec s/knowmak/D2.4- inal.pd .
Mazzuca o, M. (2018). “Mission-o ien ed inno a ion policies: challenges and oppo uni ies”, Indus ial
and Co po a e Change, 27(5):803–815. h ps://doi.o g/10.1093/icc/d y034
McCann, P. and Soe e, L. (2020). “Place-based inno a ion o sus ainabili y”, Publica ions O ice o he
Eu opean Union, doi:10.2760/250023, JRC121271
Molica, F. and Ma ques San os, A. (2024). “In sea ch o he bes ma ch. Complemen a i ies be ween
R&I unds ac oss EU egions”, Te i o ial De elopmen Insigh s Se ies, JRC136780.
OECD (2003). Policy cohe ence: Public go e nance and e i o ial de elopmen di ec o a e.
OECD (2023). D i ing Policy Cohe ence o Sus ainable De elopmen : Accele a ing P og ess on he
SDGs. OECD Publishing.
O a iano, G.IP., and Puga, D. (1998). “Agglome a ion in he global economy: a su ey o he ‘new
economic geog aphy’”, The Wo ld Economy, 21, no. 6 (1998): 707-731.
Po e , M. E. (1998). Clus e s and he new economics o compe i ion. Vol. 76, no. 6. Bos on: Ha a d
Business Re iew.
Po e s, L. (2009). “R&D in Low-Tech sec o s”, IPTS Wo king Pape on Co po a e R&D and Inno a ion,
08/2009, Eu opean Commission.
Reid, A., Miedzinski, M., B uno, N., Le Ga s, G., (2007) Syne gies be ween he EU 7 h Resea ch F ame-
wo k P og amme, he Compe i i eness and Inno a ion F amewo k P og amme and he S uc u al
37
Funds, Technical epo o he Eu opean Pa liamen 's commi ee on Indus y, Resea ch and Ene gy
(ITRE) IP/A/ITRE/ST/2006-16, DOI: 10.13140/RG.2.2.25650.20169
RIMA, Resea ch & Inno a ion and Cohesion Managing Au ho i ies Ne wo k (2025). Repo on
Eu opean Syne gies o Funds, a ailable a : h ps://ho izoneu opencppo al.eu/ eposi o y/10d06752-
371b-47 -89ab-54265c 2e 13.
San os, A. M.; Cince a, M. and Ce ulli, G. (2024). “Sou ces o inancing: Which ones a e mo e
e ec i e in inno a ion–g ow h linkage?”, Economic Sys ems, 48(2), 101177,
h ps://doi.o g/10.1016/j.ecosys.2023.101177
San os, A., Edwa ds, J., and Ne o, P. (2022). “Does Sma Specialisa ion imp o e any inno a ion
subsidy e ec on egional p oduc i i y? The Po uguese case”, Eu opean Planning S udies, 31(4),
758–779. h ps://doi.o g/10.1080/09654313.2022.2073787
San os, A.M. and Con e, A. (2024). "Regional pa icipa ion o Resea ch and Inno a ion p og ammes
unde Nex Gene a ion EU: The Po uguese case", Pape s in Regional Science, 103(1), 100006.
h ps://doi.o g/10.1016/j.pi s.2024.100006.
San os, A. M., Con e, A., & Molica, F. (2025). “Financial abso p ion o cohesion policy unds: how do
p og ammes and e i o ial cha ac e is ics in luence he pace o spending?”, JCMS: Jou nal o Com-
mon Ma ke S udies, 63(1), 227-245.
Segal L. J., G iepink M., Bosche i A., Ma chioni M., Vilajosana Guillén X., Rie a Du an M., and Lupiáñez-
Villanue a F. (2025). “Explo ing syne gies be ween Ho izon Eu ope and he EU Cohesion Policy”, Pub-
lica ion o he Commi ee on Indus y, Resea ch and Ene gy (ITRE), Policy Depa men o T ans o -
ma ion, Inno a ion and Heal h, Eu opean Pa liamen , Luxembou g.
Sö ik, J. & Kleib ink, A. (2015). ‘Mapping Inno a ion P io i ies and Specialisa ion Pa e ns in Eu ope’,
JRC Technical Repo s JRC95227.
Van den Be gh, J.; Cas o, J.; D ews, S.; Exadak ylos, F.; Fo ami i, J.; Klein, F.; Konc, T. and Sa in, I.
(2021). “Designing an e ec i e clima e-policy mix: accoun ing o ins umen syne gy”, Clima e
Policy, 21(6):745-764, DOI: 10.1080/14693062.2021.1907276
Annexes
Annex A: Dis ibu ion o Ho izon 2020 and ERDF R&I unding among NUTS-3
egions
Table A.1. Minimum and maximum sha es o ERDF R&I (assigned o a leas one SGC) and Ho izon 2020
unding by SGC in 806 NUTS-3 egions ha ecei e unding om bo h ins umen s
Ca ego y ERDF R&I Ho izon 2020
Minimum sha e Maximum sha e Minimum sha e Maximum sha e
Clima e 5.7% (Bulga ia) 20.1% (Cyp us) 6.2% (Luxembou g) 18.7% (Finland)
Ene gy 7.4% (Sweden) 42.9% (Luxembou g) 13.8% (Luxembou g) 35.9% (Bulga ia)
Food 0.0% (Luxembou g) 21.8% (Slo enia) 6.2% (Luxembou g) 41.8% (Slo akia)
Heal h 3.1% (Aus ia) 31.4% (Bulga ia) 6.4% (Bulga ia) 36.7% (Ne he lands)
Secu i y 1.2% (Belgium) 20.0% (Bulga ia) 1.9% (Denma k) 20.1% (Cyp us)
Socie y 9.3% (Aus ia) 45.4% (Sweden) 2.0% (F ance) 13.7% (Li huania)
T anspo 5.9% (Luxembou g) 40.8% (Aus ia) 5.1% (Es onia) 39.8% (Czechia)
Sou ce: CORDIS da abase, Kohesio da abase, own elabo a ions.
No e: ERDF R&I ep esen s he o al eligible expendi u e assigned o R&I p ojec s co- unded by he ERDF and assigned o
a leas one SGC. Ho izon 2020 con ains he espec i e unding amoun . This able p o es a conside able a ia ion in he
dis ibu ion o unding ac oss SGC in he Membe S a es.
Table A.2. Numbe o NUTS-3 egions (NUTS e sion 2021) ha ecei e bo h Ho izon 2020 and ERDF R&I
unding
Membe
S a e
No. o
NUTS
-3
egions
…
wi hou
Ho izon
2020
… wi hou
ERDF R&I
unding
… conside ed
in his
analysis
Membe
S a e
No. o
NUTS
-3
egions
…
wi hou
Ho izon
2020
… wi hou
ERDF R&I
unding
… conside ed
in his
analysis
AT
35 3 9 26
IE
8 8 n.a. 0
BE
44 7 20 22
IT
107 5 7 96
BG
28 10 0 18
LT
10 4 0 6
CY
1 0 0 1
LU
1 0 0 1
CZ
14 1 0 13
LV
6 0 0 6
DE
401
73
128
231
MT
2
2
n.a.
0
DK
11 0 1 10
NL
40 0 4 36
EE
5 0 0 5
PL
73 25 1 47
EL
52 46 n.a. 0
PT
25 1 1 23
ES
59 5 5 54
RO
42 4 10 30
FI
19 0 1 18
SE
21 0 0 21
FR
101 5 7 90
SI
12 0 0 12
HR
21 7 0 14
SK
8 0 0 8
HU
20 2 0 18
To al
1,166 208 194 806
Sou ce: CORDIS da abase, Kohesio da abase, own elabo a ions.
No e: P ojec unding da a o I ish, G eek and Mal ese egions is no p o ided a he NUTS-3 (bu only NUTS-2) le el and
he e o e is no conside ed in his analysis. No e ha , e.g., in Romania, e e y NUTS-3 egion ecei es ERDF g an s
acco ding o he Kohesio da abase. The same is ue, e.g., o Belgium, howe e , only ew ERDF p ojec s a e epo ed o
Eas Flande s (NUTS-2 egion BE23) which explains why 20 NUTS-3 Belgian NUTS-3 egions ha e no ecei ed any ERDF
R&I unding.
T ansi ion egions (capi al egion and no. o SGC in S3 p io i ies omi ed)
All obse a ions T immed da ase
DE-MEANED VARIABLES
(ins ead o NUTS-2 FE)
Cosine_sim Gini_Ho izon Gini_ERDF Cosine_sim Gini_Ho izon Gini_ERDF
Uni e si y 0.207*** -0.128*** -0.172*** 0.223*** -0.130*** -0.182***
(0.053) (0.025) (0.032) (0.054) (0.025) (0.032)
LQ knowledge-in ensi e -0.284 -0.684 -1.050* -0.192 -0.898* -1.143*
se ices (KIS) (1.057) (0.493) (0.638) (1.061) (0.495) (0.637)
LQ KIS squa ed 0.005 0.427 0.642 -0.051 0.516 0.675
(0.713) (0.332) (0.430) (0.707) (0.330) (0.424)
ln_GDP pe capi a -0.433** -0.180* 0.195 -0.508** -0.203** 0.194
(0.218) (0.102) (0.132) (0.222) (0.104) (0.133)
In es men a e 1.356 2.243 0.800 -0.279 2.598 1.462
(3.571) (1.665) (2.156) (3.583) (1.671) (2.151)
Compe i ion le el 2.181 -0.131 -0.001 1.778 0.978 0.869
(1.584) (0.739) (0.956) (1.769) (0.825) (1.062)
ln_Popula ion densi y 0.137*** -0.031 -0.050* 0.155*** -0.019 -0.043
(0.045) (0.021) (0.027) (0.048) (0.023) (0.029)
Cons an 0.000 0.000 -0.000 0.004 -0.001 0.005
(0.020) (0.009) (0.012) (0.020) (0.009) (0.012)
Obse a ions 115 115 115 108 108 108
R-squa ed 0.234 0.430 0.345 0.246 0.460 0.364
Con inued on he nex page …
Mo e de eloped egions
All obse a ions
T immed da ase
DE-MEANED VARIABLES
(ins ead o NUTS-2 FE)
Cosine_sim
Gini_Ho izon
Gini_ERDF
Cosine_sim
Gini_Ho izon
Gini_ERDF
Capi al egion
0.013
-0.064
0.046
0.167
-0.146*
0.174*
(0.104)
(0.056)
(0.061)
(0.155)
(0.084)
(0.089)
Uni e si y
0.074***
-0.077***
-0.066***
0.093***
-0.080***
-0.073***
(0.027)
(0.014)
(0.016)
(0.029)
(0.016)
(0.017)
LQ knowledge-in ensi e
0.325**
-0.442***
-0.450***
0.271
-0.433***
-0.746***
se ices (KIS)
(0.160)
(0.086)
(0.094)
(0.213)
(0.115)
(0.122)
LQ KIS squa ed
-0.092*
0.123***
0.123***
-0.091
0.114**
0.233***
(0.055)
(0.030)
(0.032)
(0.086)
(0.047)
(0.049)
ln_GDP pe capi a
0.314***
-0.086**
-0.146***
0.281***
-0.099**
-0.125***
(0.067)
(0.036)
(0.039)
(0.082)
(0.044)
(0.047)
In es men a e
2.814**
-1.520***
-0.087
3.521***
-2.018***
-0.433
(1.095)
(0.587)
(0.645)
(1.233)
(0.667)
(0.705)
Compe i ion le el
0.165
-0.214
0.058
0.189
-0.191
0.375
(0.427)
(0.229)
(0.251)
(0.494)
(0.267)
(0.283)
ln_Popula ion densi y
-0.000
-0.055***
0.014
0.003
-0.054***
0.024**
(0.018)
(0.009)
(0.010)
(0.019)
(0.010)
(0.011)
No. o SGC in S3 p io i ies
-0.057*
-0.025
0.008
-0.056*
-0.024
-0.001
(excl. na ional p io i ies)
(0.031)
(0.017)
(0.019)
(0.034)
(0.018)
(0.019)
Cons an
-0.000
0.000
0.000
-0.001
-0.004
-0.004
(0.009)
(0.005)
(0.006)
(0.011)
(0.006)
(0.006)
Obse a ions
480
480
480
393
393
393
R-squa ed
0.184
0.448
0.221
0.166
0.450
0.261
No es: The immed da ase is excluding he op and bo om decile o NUTS-3 egions acco ding o he a io be ween Ho izon 2020
and ERDF R&I unding. Signi icance le el: *** p < 0.01, ** p<0.05, * p<0.1. S anda d e o s a e in pa en heses.
SUR esul s by inno a ion g oup
Table B.4. Seemingly un ela ed eg ession esul s by inno a ion g oup
DE-MEANED VARIABLES
All obse a ions
T immed da ase
Eme ging inno a o s
(no. o SGCs in S3 p io i ies omi ed)
Cosine_sim
Gini_Ho izon
Gini_ERDF
Cosine_sim
Gini_Ho izon
Gini_ERDF
Capi al egion
-0.212
0.575***
-0.139
-0.136
0.456**
-0.022
(0.275)
(0.176)
(0.199)
(0.303)
(0.203)
(0.216)
Uni e si y
0.148***
-0.085***
-0.137***
0.175***
-0.124***
-0.166***
(0.045)
(0.029)
(0.033)
(0.052)
(0.035)
(0.037)
LQ knowledge-in ensi e
-0.069
-0.109
-0.123
-0.106
-0.145
-0.032
se ices (KIS)
(0.405)
(0.259)
(0.292)
(0.428)
(0.287)
(0.305)
LQ KIS squa ed
0.375
-0.366*
0.029
0.288
-0.278
0.005
(0.322)
(0.206)
(0.233)
(0.343)
(0.230)
(0.244)
ln_GDP pe capi a
0.045
0.023
-0.136
0.119
0.061
-0.211**
(0.130)
(0.083)
(0.094)
(0.146)
(0.098)
(0.104)
In es men a e
1.186
-0.368
-2.301***
2.674*
-0.599
-2.805**
(1.226)
(0.786)
(0.886)
(1.567)
(1.052)
(1.118)
Compe i ion le el
-0.571
0.200
-0.431
-0.351
0.465
-0.540
(0.704)
(0.451)
(0.509)
(0.865)
(0.581)
(0.617)
ln_Popula ion densi y
-0.002
-0.066***
0.034
0.019
-0.075***
0.010
(0.032)
(0.021)
(0.023)
(0.040)
(0.027)
(0.028)
Cons an
0.000
0.000
-0.000
0.025
-0.022**
0.005
(0.013)
(0.008)
(0.009)
(0.016)
(0.011)
(0.011)
Obse a ions
169
169
169
114
114
114
R-squa ed
0.261
0.533
0.233
0.319
0.541
0.352
Con inued on he nex page …
Mode a e inno a o s
Cosine_sim
Gini_Ho izon
Gini_ERDF
Cosine_sim
Gini_Ho izon
Gini_ERDF
Capi al egion
0.159
-0.229
0.093
0.123
-0.232
0.049
(0.262)
(0.146)
(0.154)
(0.271)
(0.153)
(0.155)
Uni e si y
0.115***
-0.102***
-0.080***
0.104***
-0.096***
-0.083***
(0.034)
(0.019)
(0.020)
(0.035)
(0.020)
(0.020)
LQ knowledge-in ensi e
0.781
-0.687***
-0.756***
0.886*
-0.739**
-1.032***
se ices (KIS)
(0.475)
(0.265)
(0.279)
(0.528)
(0.298)
(0.302)
LQ KIS squa ed
-0.399
0.299*
0.319*
-0.463
0.340*
0.485**
(0.313)
(0.174)
(0.184)
(0.343)
(0.194)
(0.197)
ln_GDP pe capi a
0.163
-0.100*
-0.007
0.179
-0.095
-0.041
(0.107)
(0.059)
(0.063)
(0.116)
(0.065)
(0.066)
In es men a e
5.088***
0.036
-2.527**
5.143***
-0.014
-2.845***
(1.680)
(0.936)
(0.985)
(1.751)
(0.988)
(1.002)
Compe i ion le el
0.460
-0.498
0.581
0.614
-0.511
0.912**
(0.683)
(0.380)
(0.401)
(0.732)
(0.413)
(0.419)
ln_Popula ion densi y
0.024
-0.061***
-0.004
0.017
-0.064***
0.004
(0.026)
(0.014)
(0.015)
(0.028)
(0.016)
(0.016)
No. o SGC in S3 p io i ies
0.040
-0.086**
-0.007
0.017
-0.074
-0.042
(excl. na ional p io i ies)
(0.071)
(0.039)
(0.041)
(0.087)
(0.049)
(0.050)
Cons an
0.000
0.000
-0.000
0.002
-0.003
0.003
(0.012)
(0.007)
(0.007)
(0.013)
(0.007)
(0.008)
Obse a ions
236
236
236
214
214
214
R-squa ed
0.214
0.518
0.251
0.199
0.499
0.289
Con inued on he nex page …
S ong inno a o s
Cosine_sim
Gini_Ho izon
Gini_ERDF
Cosine_sim
Gini_Ho izon
Gini_ERDF
Capi al egion
-0.220
0.096
0.257**
-0.209
0.074
0.302**
(0.200)
(0.105)
(0.109)
(0.255)
(0.133)
(0.138)
Uni e si y
0.075**
-0.041**
-0.055***
0.075*
-0.047**
-0.063***
(0.037)
(0.019)
(0.020)
(0.039)
(0.020)
(0.021)
LQ knowledge-in ensi e
0.554
-0.638***
-0.948***
0.686
-0.865***
-0.913***
se ices (KIS)
(0.389)
(0.203)
(0.213)
(0.468)
(0.245)
(0.253)
LQ KIS squa ed
-0.282
0.222**
0.341***
-0.378
0.374***
0.315**
(0.184)
(0.096)
(0.101)
(0.241)
(0.126)
(0.131)
ln_GDP pe capi a
0.172
-0.216***
-0.143**
0.210*
-0.246***
-0.145**
(0.108)
(0.056)
(0.059)
(0.116)
(0.061)
(0.063)
In es men a e
1.016
-2.860**
-0.154
1.308
-3.080**
-0.006
(2.231)
(1.166)
(1.221)
(2.376)
(1.242)
(1.285)
Compe i ion le el
-0.475
-0.441
0.974**
-0.369
-0.469
0.780*
(0.789)
(0.412)
(0.431)
(0.832)
(0.435)
(0.450)
ln_Popula ion densi y
0.060**
-0.039***
0.016
0.062**
-0.041***
0.018
(0.024)
(0.013)
(0.013)
(0.025)
(0.013)
(0.013)
No. o SGC in S3 p io i ies
-0.048
-0.014
-0.020
-0.054
-0.007
-0.017
(excl. na ional p io i ies)
(0.054)
(0.028)
(0.029)
(0.055)
(0.029)
(0.030)
Cons an
-0.000
0.000
0.000
-0.004
-0.002
0.000
(0.013)
(0.007)
(0.007)
(0.014)
(0.007)
(0.008)
Obse a ions
274
274
274
242
242
242
R-squa ed
0.152
0.430
0.283
0.166
0.445
0.309
Con inued on he nex page …
Inno a ion leade s
Cosine_sim
Gini_Ho izon
Gini_ERDF
Cosine_sim
Gini_Ho izon
Gini_ERDF
Capi al egion
-0.008
-0.039
-0.162*
0.753
-0.208
0.108
(0.153)
(0.081)
(0.093)
(0.472)
(0.266)
(0.303)
Uni e si y
0.045
-0.090***
-0.080**
0.141**
-0.089***
-0.118***
(0.053)
(0.028)
(0.032)
(0.060)
(0.034)
(0.039)
LQ knowledge-in ensi e
0.765***
-0.379**
0.007
0.914*
-0.515*
-0.672**
se ices (KIS)
(0.286)
(0.151)
(0.174)
(0.478)
(0.269)
(0.306)
LQ KIS squa ed
-0.202**
0.103**
-0.001
-0.263*
0.121
0.194*
(0.083)
(0.044)
(0.050)
(0.155)
(0.087)
(0.099)
ln_GDP pe capi a
0.540***
-0.110*
-0.205***
0.497***
-0.155
-0.114
(0.116)
(0.061)
(0.071)
(0.171)
(0.097)
(0.110)
In es men a e
1.367
-0.534
1.744
3.690
-1.054
0.475
(1.912)
(1.011)
(1.166)
(2.421)
(1.363)
(1.553)
Compe i ion le el
0.332
-0.397
-0.857**
-0.281
-0.160
0.361
(0.702)
(0.371)
(0.428)
(0.909)
(0.512)
(0.583)
ln_Popula ion densi y
-0.118***
-0.030
0.032
-0.169***
-0.001
0.069**
(0.037)
(0.019)
(0.022)
(0.050)
(0.028)
(0.032)
No. o SGC in S3 p io i ies
-0.079*
-0.030
0.015
-0.061
-0.033
0.008
(excl. na ional p io i ies)
(0.041)
(0.022)
(0.025)
(0.044)
(0.025)
(0.028)
Cons an
0.000
-0.000
-0.000
-0.002
-0.007
-0.009
(0.018)
(0.010)
(0.011)
(0.025)
(0.014)
(0.016)
Obse a ions
127
127
127
75
75
75
R-squa ed
0.313
0.455
0.266
0.335
0.450
0.234
No es: The immed da ase is excluding he op and bo om decile o NUTS-3 egions acco ding o he a io be ween
Ho izon 2020 and ERDF R&I unding. Signi icance le el: *** p < 0.01, ** p<0.05, * p<0.1. S anda d e o s a e in
pa en heses.
Robus ness checks: OLS eg ession conside ing he numbe o uni e si ies and he numbe o
s uden s ins ead o he Uni e si y indica o .
Table B.5. Conside ing he numbe o uni e si ies pe NUTS-3 egion in 2013 - Co espondence be ween he
cosine simila i y index and socio-economic cha ac e is ics (OLS eg ession)
OLS eg ession
(1)
(2)
(3)
(4)
(5)
(6)
VARIABLES
All
Obse a ions
T immed
All
Obse a ions
T immed
T immed
T immed
Capi al egion
0.014
0.039
-0.069
0.014
-0.085
-0.079
(0.042)
(0.052)
(0.089)
(0.098)
(0.082)
(0.081)
ln_Numbe o uni e si ies 0.117*** 0.128*** 0.095*** 0.121*** 0.124*** 0.126***
(0.017)
(0.018)
(0.024)
(0.027)
(0.025)
(0.025)
Loca ion Quo ien (LQ)
-0.032
0.087
0.309**
0.355*
0.310*
0.292
knowledge-in ensi e se ices (KIS)
(0.113)
(0.157)
(0.136)
(0.185)
(0.182)
(0.182)
LQ KIS squa ed
-0.009
-0.070
-0.080
-0.135
-0.090
-0.086
(0.048)
(0.078)
(0.049)
(0.088)
(0.086)
(0.086)
ln_GDP pe capi a
0.028
0.023
0.172**
0.126
(0.021)
(0.026)
(0.070)
(0.087)
In es men a e
-0.511**
-0.495*
2.388***
3.753***
(0.255)
(0.280)
(0.855)
(1.186)
Compe i ion le el
0.491*
0.690**
-0.081
-0.099
-0.380
-0.380
(0.298)
(0.342)
(0.403)
(0.503)
(0.429)
(0.429)
ln_Popula ion densi y 0.010 0.005 0.013 0.015 0.024 0.023
(0.010)
(0.011)
(0.019)
(0.021)
(0.020)
(0.020)
No. o SGC in S3 p io i ies
-0.007
-0.011**
-0.047**
-0.061**
-0.062**
(excl. na ional S3 p io i ies)
(0.005)
(0.005)
(0.023)
(0.026)
(0.026)
Cons an
-0.129
-0.238
-1.552
-1.350
1.024**
0.725
(0.337)
(0.419)
(0.999)
(1.208)
(0.472)
(0.459)
Obse a ions
800
642
800
642
642
642
NUTS-2 FE
NO
NO
YES
YES
YES
YES
R-squa ed
0.104
0.113
0.454
0.498
0.485
0.482
Adjus ed R-squa ed
0.094
0.100
0.240
0.262
0.247
0.244
Wald- es o join signi icance
(p- alue)
0.000
0.000
0.000
0.000
VIF
4.22
4.59
3.55
3.66
3.12
2.48
VIF > 10 (excl. squa ed e ms)
-
-
S3 p ios,
ln_GDP_pc,
In a e
S3 p ios,
ln_GDP_pc,
In a e
S3 p ios
-
No es: The dependen a iable is he cosine simila i y index. Ins ead o he Uni e si y dummy in he baseline eg ession,
he (na u al log o he) numbe o uni e si ies pe NUTS-3 egion in 2013 is used. Signi icance le el: *** p < 0.01, **
p<0.05, * p<0.1. He e oskedas ici y- obus s anda d e o s a e in pa en heses.
Table B.6. Conside ing he numbe o s uden s (ISCED 5-7) pe NUTS-3 egion in 2013 - Co espondence
be ween he cosine simila i y index and socio-economic cha ac e is ics
OLS eg ession
(1)
(2)
(3)
(4)
(5)
(6)
VARIABLES
All Obse a ions
T immed
All
Obse a ions
T immed
T immed
T immed
Capi al egion
0.124***
0.168***
0.023
0.121
0.012
0.017
(0.037)
(0.048)
(0.097)
(0.127)
(0.097)
(0.096)
ln_Numbe o s uden s
0.017***
0.018***
0.012***
0.015***
0.015***
0.015***
(0.002)
(0.002)
(0.003)
(0.003)
(0.003)
(0.003)
LQ knowledge-in ensi e
-0.087
-0.029
0.248*
0.294
0.259
0.246
se ices (KIS)
(0.101)
(0.152)
(0.141)
(0.199)
(0.193)
(0.192)
LQ KIS squa ed
0.012
-0.019
-0.065
-0.116
-0.072
-0.068
(0.038)
(0.070)
(0.047)
(0.090)
(0.087)
(0.087)
ln_GDP pe capi a
0.025
0.022
0.200***
0.154*
(0.021)
(0.027)
(0.069)
(0.087)
In es men a e
-0.480*
-0.469
2.403***
3.701***
(0.260)
(0.288)
(0.897)
(1.301)
Compe i ion le el
0.569*
0.784**
0.160
0.151
-0.217
-0.226
(0.300)
(0.333)
(0.396)
(0.500)
(0.423)
(0.421)
ln_Popula ion densi y
0.019*
0.014
0.017
0.023
0.036*
0.035*
(0.010)
(0.012)
(0.019)
(0.021)
(0.019)
(0.019)
No. o SGC in S3 p io i ies
-0.008*
-0.012**
-0.028
-0.048
-0.044
(excl. na ional p io i ies)
(0.005)
(0.005)
(0.028)
(0.033)
(0.032)
Cons an
-0.204
-0.327
-2.124**
-1.903
0.776*
0.571
(0.343)
(0.427)
(0.971)
(1.197)
(0.463)
(0.447)
Obse a ions
779
622
779
622
622
622
NUTS-2 FE
NO
NO
YES
YES
YES
YES
R-squa ed
0.124
0.134
0.461
0.502
0.489
0.488
Adjus ed R-squa ed
0.113
0.121
0.244
0.260
0.244
0.244
Wald- es o join signi icance
(p- alue)
0.000
0.000
0.000
0.000
Mean VIF
4.18
4.56
3.78
4.06
3.51
2.47
VIF > 10 (excl. squa ed e ms)
-
-
S3 p ios,
ln_GDP_pc,
In a e
S3 p ios,
ln_GDP_pc,
In a e
S3 p ios
-
No es: The dependen a iable is he cosine simila i y index. Ins ead o he Uni e si y dummy in he baseline eg ession,
he (na u al log o he) numbe o s uden s (ISCED le els 5-7) pe NUTS-3 egion in 2013 is used. Signi icance le el: ***
p < 0.01, ** p<0.05, * p<0.1. He e oskedas ici y- obus s anda d e o s a e in pa en heses.
Lis o igu es
Figu e 1. Dis ibu ion o Ho izon 2020 and ERDF R&I amoun s pe capi a among NUTS-3 egions 15
Figu e 2. Ho izon 2020/ERDF R&I a io in di e en g oups o NUTS-3 egions based on hei
economic de elopmen .................................................................................................................................................................... 16
Figu e 3. Funding sha es alloca ed o di e en SGCs in NUTS-3 egions pa o … .................................. 20
Figu e 4. O e all simila i y o he ERDF R&I and Ho izon 2020 unding dis ibu ion ac oss SGCs by
NUTS-3 egion (cosine simila i y index) ................................................................................................................................. 23
Lis o ables
Table 1. Sha e o ERDF R&I (assigned o a leas one SGC) and Ho izon 2020 unding by SGC ........ 14
Table 2. Explana o y a iables a he NUTS-3 egional le el: a iable desc ip ion and sou ce ........... 19
Table 3. Cosine simila i y index by NUTS-3 egions ha ecei e bo h ERDF R&I and Ho izon 2020
unding ...................................................................................................................................................................................................... 24
Table 4. Co espondence be ween he cosine simila i y index and socio-economic cha ac e is ics in
EU NUTS-3 egions (OLS eg ession) ....................................................................................................................................... 26
Table 5. Co espondence be ween he cosine simila i y index and socio-economic cha ac e is ics in
NUTS-3 egions by de elopmen g oup ................................................................................................................................. 28
Table 6. Co espondence be ween he cosine simila i y index and socio-economic cha ac e is ics in
NUTS-3 egions by inno a ion g oup ....................................................................................................................................... 29
Table 7. Seemingly un ela ed eg ession (SUR) esul s .............................................................................................. 31
