de las Me cedes Menéndez, Ma ía
Wo king Pape
Twin ansi ion ade based on mul i-dimensional
economic complexi y
UNU-MERIT Wo king Pape s, No. 2024-032
P o ided in Coope a ion wi h:
Maas ich Economic and Social Resea ch Ins i u e on Inno a ion and Technology (UNU-MERIT),
Uni ed Na ions Uni e si y (UNU)
Sugges ed Ci a ion: de las Me cedes Menéndez, Ma ía (2024) : Twin ansi ion ade based on mul i-
dimensional economic complexi y, UNU-MERIT Wo king Pape s, No. 2024-032, Uni ed Na ions
Uni e si y (UNU), Maas ich Economic and Social Resea ch Ins i u e on Inno a ion and Technology
(UNU-MERIT), Maas ich
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/326928
S anda d-Nu zungsbedingungen:
Die Dokumen e au EconS o dü en zu eigenen wissenscha lichen
Zwecken und zum P i a geb auch gespeiche und kopie we den.
Sie dü en die Dokumen e nich ü ö en liche ode komme zielle
Zwecke e iel äl igen, ö en lich auss ellen, ö en lich zugänglich
machen, e eiben ode ande wei ig nu zen.
So e n die Ve asse die Dokumen e un e Open-Con en -Lizenzen
(insbesonde e CC-Lizenzen) zu Ve ügung ges ell haben soll en,
gel en abweichend on diesen Nu zungsbedingungen die in de do
genann en Lizenz gewäh en Nu zungs ech e.
Te ms o use:
Documen s in EconS o may be sa ed and copied o you pe sonal
and schola ly pu poses.
You a e no o copy documen s o public o comme cial pu poses, o
exhibi he documen s publicly, o make hem publicly a ailable on he
in e ne , o o dis ibu e o o he wise use he documen s in public.
I he documen s ha e been made a ailable unde an Open Con en
Licence (especially C ea i e Commons Licences), you may exe cise
u he usage igh s as speci ied in he indica ed licence.
h ps://c ea i ecommons.o g/licenses/by-nc-sa/4.0/
#2024-032
Twin ansi ion ade based on mul i-dimensional economic
complexi y
Ma ía de las Me cedes Menéndez
Published 21 No embe 2024
Maas ich Economic and social Resea ch ins i u e on Inno a ion and Technology (UNU-MERIT)
email: in o@me i .unu.edu | websi e: h p://www.me i .unu.edu
Boschs aa 24, 6211 AX Maas ich , The Ne he lands
Tel: (31) (43) 388 44 00
UNU-MERIT Wo king Pape s
ISSN 1871-9872
Maas ich Economic and social Resea ch Ins i u e on Inno a ion and Technology
UNU-MERIT
UNU-MERIT Wo king Pape s in end o dissemina e p elimina y esul s o esea ch ca ied
ou a UNU-MERIT o s imula e discussion on he issues aised.
1
Twin ansi ion ade based on
mul i-dimensional economic complexi y*
Ma ía de las Me cedes Menéndez†
21 No embe 2024
Abs ac
T ade wo ldwide is being eshaped by wo majo mega ends: ad anced digi aliza ion o p oduc ion and he
ansi ion owa ds en i onmen ally sus ainable goods. This chap e examines o he i s ime he pe spec i e o
win ansi ion expo and impo di e si ica ion wi hin a mul i-dimensional economic complexi y app oach
(Nomale & Ve spagen, 2024b, 2024d) and in es iga e whe he his ype o p oduc i e ans o ma ion pe pe ua es
pa h-dependency p ocesses in 80 coun ies o e 2000-2018. The esul s sugges ha an expo /impo p oduc i e
s uc u e based on win ansi ion p oduc s exhibi di e en economic pe o mance, sus ainabili y, and inequali y
implica ions. P oduc i e specializa ion in hese p oduc s has been e y pa h-dependen and wi h a low engagemen
o de eloping coun ies and hence, ein o cing he co e-pe iphe y ade di ision. Fu he mo e, esul s sugges ha
de elopmen s in digi al and g een echnological pa adigms mainly ake place in a selec ed numbe o coun ies ha
a e al eady highly de eloped.
JEL codes: F14, O10, Q01
Keywo ds: mul i-dimensional economic complexi y, canonical co espondence analysis, win ansi ion,
sus ainable de elopmen .
* The esul s o his pape we e p esen ed a he UNU-MERIT In e nal Con e ence, 26-27 Sep embe in Maas ich , The Ne he lands and a
he VIII Con e ence o he La in Ame ican and he Ca ibbean Inno a ion and En ep eneu ship Economics Ne wo k, 8-9 No embe 2024 in
Mon e ideo, U uguay. I am speci ically g a e ul o Ba Ve spagen and Önde Nomale o hei help on he me hodological aspec s o his
wo k and hei ca e ul eading o he manusc ip and commen s o imp o e he pape . The opinions exp essed and any e o s emain my sole
esponsibili y.
† PhD Fellow a UNU-MERIT, menendez@me i .unu.edu
2
1. In oduc ion
One o he mos p essing challenges cu en ly aced by coun ies wo ldwide is he impe a i e o ans o m
hei p oduc i e s uc u e owa ds g eene and digi al p oduc s. Such ans o ma ion mus be in ha mony
wi h he dimensions o sus ainable de elopmen , which emb ace economic g ow h, social inclusion, and
en i onmen al p o ec ion. Coun ies’ p oduc i e ans o ma ion s a egies ha encompass p oduc s linked
o in o ma ion and communica ion (IC) and Fou h Indus ial Re olu ion (4IR) echnologies, as well as
di e en ypes o goods and echnologies ha make up a g een p oduc i e sys em, play a undamen al ole
in his con ex . This a icle p o ides an o e iew o he ela ionship be ween coun ies' expo /impo
specializa ion on g een and digi al ( win) ansi ion p oduc s and di e se dimensions o sus ainable
de elopmen . Addi ionally, i examines whe he his ype o p oduc i e ans o ma ion pe pe ua es pa h-
dependence p ocesses in coun ies’ p oduc i e specializa ion.
Economic complexi y cons i u es a b oad academic ield o he comp ehensi e analysis o socie al
challenges (Balland e al., 2022; Calda ola e al., 2024), including he p ocess o p oduc di e si ica ion
owa ds c ucial p oduc s o he g een and digi al ansi ion. I is a well-es ablished academic ield3 ha
ound i s oo s in Schumpe e ian heo y, which pos ula es ha he economy is a complex sys em in
cons an e olu ion, whe e economic change is go e ned by dis-equilib ium esul ing om he eme gence
o inno a ions (Nomale & Ve spagen, 2024a; Robe & Yoguel, 2016). Mo eo e , i d aws on Albe
Hi schman’s g ow h heo y and s uc u alis heo y, whe e de elopmen is concei ed as a p ocess ha
gene a es sec o al and geog aphical imbalances and social and income inequali ies (Hidalgo, 2023; Ne ke
e al., 2024). Inc easing compu a ional capaci y o da a p ocessing has p omp ed a shi in he ield o
economic complexi y away om a heo y o unde s and economic de elopmen “big” p oblems o
emb ace a da a-d i en esea ch p og am. This esea ch p og am, known as “economic complexi y heo y”
o he “new pa adigm o economic complexi y” has been cha ac e ized by employing ne wo k heo y and
machine lea ning echniques on big da ase s, such as ade da e, pa en da a, among o he s (Balland e al.,
2022; Hidalgo & Hausmann, 2009; Tacchella e al., 2012, among o he s).
3 Mo e abou he e olu ion o Economic Complexi y as a ield, see Balland e al (2022); Hidalgo (2023); Nomale & Ve spagen
(2024a).
3
Wi hin he economic complexi y amewo k, he p oduc i e ans o ma ion o a coun y owa ds win
ansi ion p oduc s can be concei ed as an e ol ing and dynamic p ocess dependen on he cu en
p oduc i e specializa ion. Imbalances end o occu when coun ies inco po a e hese p oduc s due o
changes in compa a i e ad an age (Coniglio e al., 2021). The ac ha he pa e n o p oduc i e
specializa ion ends o pe sis o e ime, o in o he wo ds, is pa h-dependen , di ec ly challenges he
possibili y o some coun ies success ully in eg a ing in o a g een and digi al p oduc i e pa h. The
ela ionship be ween he pa h dependence concep , b oadly unde s ood as a dynamic p ocess whose
e olu ion depends on pas decisions o economic agen s and he con ex (An onelli, 1997; Da id, 2007)
and he ans o ma ion o he p oduc i e s uc u e has been ex ensi ely s udied by an e olu iona y-neo-
Schumpe e ian pe spec i e. This pe spec i e explains ha he pe sis ence o coun ies in compa a i e
ad an age e lec s he in e dependence be ween hei p oduc i e and inno a ion sys ems which a e i mly
oo ed in he social and ins i u ional s uc u e (Dosi e al., 1990; Reins alle & Reschenho e , 2019).
The ela ionship be ween pa h dependence and p oduc i e s uc u e has been ope a ionalized by he
economic complexi y app oach h ough he concep o p oduc ela edness. F om an expo pe spec i e,
his concep pos ula es ha a coun y will be mo e likely o de elop compa a i e ad an age in expo ing
(p oducing) win ansi ion p oduc s i hey a e ela ed o i s cu en expo s, a he han expo s ha a e
un ela ed. The e o e, he po en ial di e si ica ion o economies owa ds c ucial p oduc s o digi al and
g een ansi ion depends on he p e-exis ence expo baske o coun ies, which is de e mined by he e-
combina ion o cu en coun ies’ a ailable capabili ies (e.g., physical capi al, human capi al,
echnologies, ins i u ions) (Coniglio e al., 2021). Howe e , he me hod does no add ess how hese
capabili ies accumula e. In o he wo ds, he concep o p oduc ela edness ein o ces he no ion o pa h
dependence and he asse ion ha coun ies a e “condemned” o p oduce p oduc s only ela ed o hei
de elopmen oppo uni ies. This de e minis ic pe spec i e o coun ies’ possibili ies o s uc u al
ans o ma ion has been ecen ly empi ically e isi ed by au ho s who a gue ha coun ies can di e si y
ega dless o hei p oduc ela edness o cha ac e is ic p oduc s (Coniglio e al., 2018, 2021; Nomale &
Ve spagen, 2024c). In his ega d, Talebzadehhosseini e al (2020) ound o a se o coun ies ha hei
abili y o di e si y hei g een p oduc baske is no only pa h-dependen on hei cu en capabili ies bu
also iden i y non-pa h-dependen p ocesses o p oduce p oduc s un ela ed o hei cu en p oduc baske .
4
The concep o p oduc ela edness om he pe spec i e o impo s has been much less add essed by he
economic complexi y li e a u e. A ecen pape by Jun e al (2020) s udies ela edness om he pe spec i e
o expo e /impo e coun ies. B ie ly, he au ho s in oduce a measu e named “impo ela edness” o
es ima e “ he ac ion o he geog aphic neighbo s o a coun y ha impo a p oduc om he same
o igin” (p.250-251). They a gue ha i e lec s he knowledge on how o dis ibu e wi hin a egion wi h
ce ain cul u e o knowledge on logis ic in o ma ion abou shipping ou es. Adap ing he expo p oduc
ela edness measu e o impo s, his pape assumes ha a coun y will be mo e likely o de elop a
compa a i e ad an age in impo ing (using) win ansi ion p oduc s i hey a e ela ed o i s cu en impo
baske , a he han un ela ed impo s. The po en ial di e si ica ion o economies owa ds p oduc s c ucial
o he digi al and g een ansi ion h ough impo s is de e mined by he ecombina ion o impo
capabili ies linked o he usage o hese p oduc s. Unlike o expo s, no s udy has e isi ed empi ically
he de e minis ic iew ha coun ies a e des ined o use only p oduc s ela ed o hei knowledge impo
oppo uni ies. Impo s o capi al play a key ole o coun ies, especially o de eloping coun ies, o
achie e highe a es o in es men and g ow h (Can e al., 2023). This is pa icula ly impo an in he
con ex o he win ansi ion as hey cons i u e a key aspec o acing challenges o digi aliza ion and
clima e change. In essence, ollowing he au ho s, impo di e si ica ion migh a oid pa h dependence as
i allows coun ies o access new capi al goods and he e o e o he p oduc ion o new p oduc s, new
echnology and knowledge.
Pa h dependence on ade specializa ion based on p oduc simila i y can be isualized h ough a ne wo k
ep esen ing he “p oduc space” (PS). PS de i es om he applica ion o he dimensionali y- educ ion
echniques on ade specializa ion ma ices (Hidalgo, 2021), whe e he nodes a e expo (impo ) p oduc s
and links ep esen p oduc s ha a e co-expo ed (co-impo ed) by a coun y (based on Hidalgo e al.,
2018). In o he wo ds, he PS e lec s he pai -wise p oximi y be ween p oduc s as condi ional
p obabili ies o co-occu ence (Nomale & Ve spagen, 2022a). The sugges ion is ha coun ies mo e
h ough he PS by de eloping goods close o hose ha hey cu en ly p oduce. Less sophis ica ed
coun ies a e in he pe iphe y, no so well connec ed, while sophis ica ed p oduc s a e connec ed in he
co e (Hidalgo e al., 2007). So a , he PS ool only allows o isualize p oduc s and coun ies wi hou
including how hey may be associa ed wi h he di e en a iables o sus ainable de elopmen .
Fu he mo e, economic complexi y es ima es he a ailabili y o capabili ies embedded in he PS hough
indexes such as he Economic Complexi y Index (ECI), which e lec s he cu en s a e o a coun y's
5
p oduc i e knowledge and he P oduc Complexi y Index (PCI) ha measu es he p oduc i e knowledge
equi ed o p oduce a p oduc . O e all, economic complexi y has un a eled he ela ionship be ween
economic s uc u es and he dimensions o sus ainable de elopmen 4, simply h ough only he pa allel
assessmen o me ics such as he Economic Complexi y Index (ECI) on economic g ow h, inequali y,
and emissions in ensi y5.
On his las poin , he mul i-dimensional economic complexi y app oach in oduced by Nomale &
Ve spagen (2024b, 2024d) ep esen s a mo e comp ehensi e ope a ionalisa ion o he PS. This app oach
pe mi s examining he changes o pe sis ence in he coun ies’ ade specialisa ion and includes coun ies’
sus ainable de elopmen cha ac e is ics in he analysis. This app oach con as wi h he PS om Hidalgo
e al (2007), which solely includes p oduc s and coun ies. B ie ly, he mul i-dimensional economic
complexi y app oach is based on an o dina ion echnique, canonical co espondence analysis (CCA),
commonly employed in ecological s udies (Legend e & Legend e, 2012; Te B aak, 1986). The s a ing
poin o he CCA is co espondence analysis (CA), as, o es ima ing he PS, i is a dimensionali y
educ ion me hod ha e eals he ela i e ela ionship be ween and wi hin wo g oups o a iables based
on da a gi en in a con ingency able. Fo ins ance, his ype o analysis can summa ize he numbe o
species (p oduc s) in di e en loca ions (coun ies) whe e he species ha end o occu in he same
loca ion a e plo ed oge he . The "o dina ion diag ams" o " iplo s" display he esul o CCA whe e
poin s commonly ep esen species (p oduc s) and loca ions (coun ies), and a ows ep esen
en i onmen al (sus ainable de elopmen ) a iables (Te B aak, 1986). As a PS o con en ional economic
complexi y, he o dina ion diag am allows o in e he specializa ion pa e ns o coun ies, bu also
includes how his specializa ion is associa ed wi h sus ainable de elopmen a iables. Like Hidalgo e al.
(2007) o he PS, he p oximi y o he p oduc s in he o dina ion diag am can be in e p e ed as he
p oduc s elying on simila capabili ies o hei p oduc ion. Simila ly, in he case o coun ies, he
closeness o he poin s can be unde s ood as indica ing a simila capabili y endowmen . Unlike PS, he
4 See Fe az e al (2021) and Mon iel-He nández e al (2024) o an ex ensi e li e a u e e iew on economic complexi y and
sus ainable de elopmen dimensions.
5 Fo economic pe o mance (Hausmann e al., 2014; Hidalgo & Hausmann, 2009), inequali y (Bandei a Mo ais e al., 2018;
Ha mann e al., 2017) and sus ainabili y5 (Can & Gozgo , 2017; Rome o & G amkow, 2021)). So a , he way ha economic
complexi y has explo ed he mul idimensionali y app oach o sus ainable de elopmen is by es ima ing he ECI using di e en
sou ces o da a (pa en s, ade, and publica ions da a) and analyzing how hei in e ac ions impac on economic pe o mance,
inequali y, and sus ainabili y indi idually (S ojkoski e al., 2023), which limi s he scope o a b oade in e p e a ion o he
comp ehensi e impac o hese indica o s on sus ainable de elopmen .
6
dis ance among coun ies and among p oduc s e lec s chi-squa e dis ances and inc eases acco ding
u he apa he coun ies/p oduc s a e in he diag am (Te B aak & Ve donscho , 1995).
The aim o his chap e is o un a el he ela ionship be ween coun ies' ade specializa ion owa ds
p oduc s c ucial o he g een and digi al ( win) ansi ion and a ious dimensions o sus ainable
de elopmen om a mul idimensional economic complexi y app oach. In his ega d, i con ibu es
h ough a no el me hodology o a) examine o he i s ime he pe spec i e o win ansi ion expo and
impo specializa ion wi hin a mul i-dimensional economic complexi y app oach and b) in es iga e
whe he his ype o p oduc i e ans o ma ion pe pe ua es pa h-dependency p ocesses in coun ies
p oduc i e specializa ion. Fu he mo e, i p o ides a baske o c ucial p oduc s connec ed wi h he digi al
and g een ansi ion. The p ima y da a sou ce is ade da a ex ac ed om BACI-CEPII o 2000-2018.
This wo k elies on Ha monized Sys em classi ica ion co espondence ables o iden i y p oduc s linked
o win ansi ion o e ime. Mo eo e , his pape collec s in o ma ion om se e al sou ces on o b oadly
con empla e he sus ainable de elopmen dimensions.
The esul s sugges ha an expo /impo p oduc i e s uc u e based on win ansi ion p oduc s has
di e en economic pe o mance, sus ainabili y, and inequali y implica ions. High income, pollu an and
less unequal coun ies end o specialize in expo ing g een, digi al and g een-digi al p oduc s while mo e
unequal and high g ow h coun ies specialize in expo ing c i ical mine als o g een echnologies. On he
o he hand, digi al and digi al-g een p oduc s impo specializa ion associa es wi h highe GDP g ow h
a es while g een impo p oduc specializa ion associa es wi h highe le els o inequali y. C i ical mine al
o g een echnologies is he only se o p oduc s om he win ansi ion baske ha has seen a mo e
p onounced inc ease in ubiqui y and shi in he diag am, con i ming hei key ole o he Eu opean ne
ze o s a egy.
Addi ionally, he esul s con i m ha win ansi ion expo s specializa ion is pa h dependen and
ein o ces global ade di ision. In his ega d, g een expo specializa ion engages only a limi ed g oup
o mo e di e si ied de eloped coun ies, which pu s a isk he possibili y o a b oade ange o de eloping
coun ies pa icipa ing in a g een p oduc i e ans o ma ion p ocess alongside a educ ion in inequali ies.
Fu he mo e, he posi ion o hese ad anced economies in he g een p oduc i e ace, coupled wi h high
pollu an emissions, se es o ein o ce hei undamen al ole in he igh agains clima e change. The
13
Figu e 2.3 T iplo wi h CCA dimensions 1 and 2
14
2.3 Twin ansi ion p oduc s
Al hough iden i ying win ansi ion p oduc s is di icul because he e is no s anda dized classi ica ion in
he academic li e a u e o de ec ing hese p oduc s, se e al e o s ha e been made o iden i y and analyze
he e olu ion o p oduc s in ol ed in g een and digi al ansi ion using ade da a. The ollowing
subsec ions e iew he app oaches used o iden i y hese p oduc s.
2.2.1 G een p oduc s
To begin, he e is no an in e na ionally accep ed classi ica ion o g een p oduc s (Mealy & Tey elboym,
2022) no a single de ini ion o he classi ica ion o g een p oduc s (o g een expo s). In his ega d, he
Ha monized Sys em (HS) does no iden i y g een/non-g een p oduc s. Despi e all hese limi a ions, e o s
o iden i y g een/sus ainable/clean p oduc s o "p oduc s wi h en i onmen al bene i s" a e se e al and
eme ge because o in e na ional nego ia ions o ade libe aliza ion on a g oup o p oduc s ha a e ele an
o deca boniza ion and sus ainable de elopmen (Bon adini & Vona, 2023). As i is poin ed ou by he
au ho s, some o hese a emp s we e guided mo e owa ds poli ical aspec s a he han he eal g een
con en and no ha ing as a esul and uni ica ion o a lis o p oduc s (Mao e al., 2023; Sau age, 2014).
In a b oad sense, a emp s o iden i y g een en i onmen al goods can be di ided in o h ee main g oups:
adi ional en i onmen al goods, en i onmen ally p e e able p oduc s (EPPs), and en i onmen al
echnologies goods. The i s a e hose whose main unc ion is o add ess o con ibu e o sol ing an
en i onmen al issue o challenge. EPPs a e p oduc s p oduced in ways ha a e less de imen al o he
en i onmen , such as o ganic co on (B en on & Chemu ai, 2021). Finally, en i onmen al echnologies
goods a e b oadly unde s ood as hose used solely o sol ing en i onmen al p oblems.
A pionee ing a emp o iden i y he en i onmen al goods and se ices indus y is a epo by he OECD
(1999) ha de ines en i onmen al goods:
“The en i onmen al goods and se ices indus y consis s o ac i i ies which p oduce goods and se ices
o measu e, p e en , limi , minimise o co ec en i onmen al damage o wa e , ai and soil, as well as
15
p oblems ela ed o was e, noise and eco-sys ems. This includes cleane echnologies, p oduc s and
se ices ha educe en i onmen al isk and minimise pollu ion and esou ce use.”
The en i onmen al goods a e classi ied in o ele en ca ego ies: 1) ai pollu ion con ol, 2) Cleane o mo e
esou ce e icien echnologies and p oduc s, 3) en i onmen ally p e e able p oduc s based on end use o
disposal cha ac e is ics, 4) hea and ene gy managemen , 5) en i onmen al moni o ing, analysis and
assessmen equipmen , 6) na u al esou ces p o ec ion, 7) noise and ib a ion aba emen , 8) enewable
ene gy plan , 9) managemen o solid and haza dous was e and ecycling sys ems, 10) clean up o
emedia ion o soil and wa e and 11) was e wa e managemen and po able wa e ea men .
Fo ins ance, he Asian Paci ic Economic Coope a ion (APEC) lis o En i onmen al Goods10 in oduced
in 2012 include 54 en i onmen al goods. App oxima ely h ee-qua e s o he lis a e goods ela ed o
enewable ene gy p oduc ion, en i onmen al moni o ing analysis, and managemen o solid and haza dous
was e and ecycling sys ems. Addi ionally, he lis comp ises a educed numbe o p oduc s o ai
pollu ion con ol, was ewa e managemen , and po able wa e ea men . Finally, he lis includes one
en i onmen ally p e e able p oduc and one associa ed wi h na u al isk managemen . Fu he mo e, he
In e na ional Mone a y Fund (IMF) de ines en i onmen al goods as hose ha
"include bo h goods connec ed o en i onmen al p o ec ion (connec ed goods) and goods ha ha e been
adap ed o be mo e en i onmen ally iendly o "cleane " (adap ed goods). Connec ed goods a e goods
whose use di ec ly se es en i onmen al p o ec ion pu poses…Adap ed goods a e goods whose use is
bene icial o en i onmen al p o ec ion"11.
IMF iden i ies he p oduc s based on OECD and Eu os a , The En i onmen al Goods & Se ices Indus y:
Manual o Da a Collec ion and Analysis, 1999, and IMF esea ch.
The mos ecen a emp o iden i y g een p oduc s was ca ied ou by (Mealy & Tey elboym, 2022),
whose main aim was o b ing oge he a ious lis s o en i onmen al goods classi ica ions om he WTO,
he O ganiza ion o Economic Co-ope a ion and De elopmen (OECD), and he Asia-Paci ic Economic
10 h ps://www.apec.o g/mee ing-pape s/leade s-decla a ions/2012/2012_aelm/2012_aelm_annexc
11 h ps://clima eda a.im .o g/documen s/ad5179b954ed4a8389b 6400324a901e/explo e
16
Coope a ion (APEC) in o a single and comp ehensi e da a se o g een p oduc lis s. This s udy de ines
g een p oduc s as hose p oduc s wi h en i onmen al bene i s. The lis includes 293 p oduc s classi ied as
HS1992.12
Wi hin en i onmen al echnologies goods, i is possible o iden i y se e al p oduc classi ica ions. Low
Ca bon echnology p oduc s a e hose " ha p oduce less pollu ion han hei adi ional ene gy
coun e pa s and will play a i al ole in he ansi ion o a low ca bon economy. LCTs include mechanics
like wind u bines, sola panels, biomass sys ems and ca bon cap u e equipmen . The iden i ica ion is
based on (Piga o e al., 2020). One hund ed wen y- ou p oduc s a e iden i ied acco ding o his
me hodology. Ano he a emp is p o ided by Rosenow & Mealy (2024). The au ho s iden i y
deca boniza ion echnologies ha include adable p oduc s associa ed wi h he alue chains o sola PV,
wind u bines, and elec ic ehicles. The iden i ica ion o hese p oduc s elies on h ee s eps: e iew o
academic li e a u e, des ock esea ch o classi y each p oduc in o ou alue chain segmen s: aw ma e ial,
p ocessed ma e ials, subcomponen s, and end p oduc s. The p oduc s we e hen alida ed by indus y
expe s.
2.2.2 Digi al p oduc s
As is o en he case when iden i ying digi al echnologies, when i comes o p oduc s, he e does no seem
o be ag eemen on a se o p oduc s associa ed wi h he digi al ansi ion. In many o hese classi ica ions,
digi isa ion is mo e closely associa ed wi h In o ma ion and Communica ion Technologies (ICT), while
in o he s, i is mo e closely associa ed wi h echnologies ela ed o he Fou h Indus ial Re olu ion. To a
g ea e o lesse ex en , se e al in e na ional o ganiza ions ha e ocused on he ele ance and challenges
o measu ing digi al ade. Fo he la e , he p oduc -le el classi ica ions a e mo e complex. This is due
o he in angible capi al componen o he digi al ansi ion, which is o en no cap u ed in expo p oduc s.
And eoni e al. (2023) buil a no el Digi al P oduc ion Technology Classi ica ion based on Ha monized
Sys em (HS2017- 6 Digi s) wi h p oduc s connec ed o digi al p oduc ion echnologies (DPTs). The
p oduc s linked o DPTs a e clus e ed in o di e en ca ego ies: inal digi al echnologies, componen s o
digi al echnologies, and ins umen a ion o digi al echnologies. Au ho s iden i y p oduc s based on
12 Mo e in o ma ion abou g een p oduc s a ailable in h ps://g een- ansi ion-na iga o .o g/.
17
selec ing HS ele an chap e s o digi al p oduc ion (1) Chap e 84, "Nuclea eac o s, boile s,
machine y, and mechanical appliances; pa s he eo "; 2) Chap e 85, "Elec ical machine y and
equipmen and pa s he eo ; sound eco de s and ep oduce s; ele ision image and sound eco de s
and ep oduce s, pa s and accesso ies o such a icles"; and 3) Chap e 90 "Op ical, pho og aphic,
cinema og aphic, measu ing, checking, medical o su gical ins umen s and appa a us; pa s and
accesso ies") and BECs (Chap e 41, Chap e 42, Chap e 22). Then, o na ow he selec ion a sys ema ic
iden i ica ion o DPTs was ca ied ou using keywo ds selec ion and p oduc analysis p oduc in
consul a ion wi h expe s o assess i s connec ion o digi aliza ion.
UNCTAD de eloped a lis o ICT goods b oadening he de ini ion p oposed by (OECD, 2011) de ining
ICT goods as hose ha “p ima ily ul il o enable he unc ion o in o ma ion p ocessing and/o
communica ion by elec onic means, such as ansmission o display"13. The p oduc s iden i ied a e
g ouped in o he ollowing ca ego ies: 1) compu e s and ICT pe iphe al equipmen , 2) communica ion
equipmen , 3) consume elec onic equipmen , 4) elec onic componen s and 5) miscellaneous.
Fos e -McG ego e al. (2022) iden i y a lis o Fou h Indus ial Re olu ion p oduc s linked o CAD-
CAM, obo s, au oma e welding, 3D p in ing, egula ing ins umen s and ICT based on a Fos e ‐McG ego
e al. (2019) and Acemoglu & Res epo (2022). The o me iden i y p oduc s linked o Fou h Indus ial
Re olu ion (4IR) associa ed wi h ou speci ic echnologies: (i) indus ial obo s; (ii) addi i e
manu ac u ing (o 3D p in ing); (iii) compu e -aided design and compu e aided manu ac u ing
(CAD/CAM) echniques; (i ) big da a and cloud compu ing; and ( ) a i icial in elligence and machine
lea ning. Iden i ying his ype o p oduc s p esen s some limi a ions: mos pa o his echnology is
so wa e, and i is complica ed o iden i y using ade da a, hese echnologies depend on ha dwa e, and i
is usually a mul i-pu pose ha dwa e, echnologies a e embodied in manu ac u ing capi al goods and hence
he e is no a dis inc ion be ween sma and no mal e sion o hese p oduc s. Au ho s selec ed p oduc s
linked o obo s, 3D p in ing and CAD CAM. The p oduc s belong o he HS Chap e 84 and he au ho s
wa n he eade s ha he e a e possibili ies ha hese classi ica ions e ie e in o ma ion o p oduc s om
olde echnological e olu ions.
13 h ps://unc ad.o g/sys em/ iles/o icial-documen / n_unc ad_ic 4d20_en.pd
18
Finally, ano he a emp is p o ided by Zhang e al. (2023) iden i y digi al p oduc s using S a is ical
Classi ica ion o he Digi al Economy and I s Co e Indus ies (2021)" eleased by he Na ional Bu eau o
S a is ics o China (NBSC) and he lis o in o ma ion and communica ions echnology (ICT) p oduc s
published by he Uni ed Na ions Con e ence on T ade and De elopmen (UNCTAD) o cons uc a digi al
p oduc ca alogue manually. 209 p oduc s a e iden i ied.14 The p oduc s (HS2017) iden i ied by he
au ho s belong o Chap e 84, Chap e 85, Chap e 95, Chap e 90, Chap e , Chap e 88, Chap e 87
Chap e 86, and Chap e 98.
2.2.3 G een-Digi al p oduc s
Since no uni e sally accep ed de ini ion exis s o p oduc s in ol ed in he win ansi ion, his documen
c ea es a baske by combining lis s o g een-digi al p oduc s and iden i ying po en ial o e laps. To he
bes o my knowledge, no o he s udy ocused on iden i ying hese p oduc s. I de ine g een-digi al
p oduc s as hose included in he G een P oduc s & Deca bonisa ion Technologies and Digi al lis s. These
p oduc s a e a small sample (less han 50 p oduc s). B oadly, his ca ego y includes p oduc s belonging o
Chap e s 84, 85 and 90 om COMTRADE classi ica ion and include machine ools and mechanical
appliances, machine y o semiconduc o manu ac u ing, elecommunica ion equipmen , elec onic
componen s, op ical and scien i ic ins umen s and egula ing and con ol ins umen s.
2.2.3 C i ical mine als o g een echnologies
I decided o include c i ical mine als in o he win ansi ion baske o p oduc s as ha a e a c ucial inpu
o low ca bon and digi al echnologies. Val e de-Ca bonell e al. (2024) poin ed ou ha he e is no
unique lis o c i ical mine als since he e is no s anda dized de ini ion o mine al c i icali y o a s anda d
me hodology o es ima e i . While he concep o c i ical mine als has long been pa o na ional secu i y
agendas, e o s o concep ualize and iden i y hese ypes o mine als ha e in ensi ied because o he high
demand o hese p oduc s in he g een and digi al ansi ion.
Recen ly, he inno a ion s udies agenda has been a ac ed o s udy he associa ion be ween coun ies'
na u al esou ces and he supply o ma e ials in he e olu ion o echnological pa adigms. These s udies
14 See Appendix in he pape .
19
ha e iden i ied c i ical ma e ials based on hei in ensi y in echnological inno a ions ela ed o he
in o ma ion and communica ion echnologies (ICT) pa adigm and ha e ound an associa ion be ween
echnological dynamics and he demand o c i ical ma e ials o e ime (Dieme e al., 2022; Li e al.,
2024). Fo ins ance, Dieme e al. (2022) concen a e on six mine als: in, ungs en, an alum, gold, cobal
and li hium while Li e al. (2024) iden i y 13 di e en a e ea h me als. The in ensi e use o c i ical aw
ma e ials is also ound in g een echnologies, especially he mo e ma u e ones, such as me al p ocessing
inno a ions (de Cunzo e al., 2023). These au ho s iden i y a lis o 39 c i ical aw ma e ials in which is
possible o iden i y o e laps wi h mo e ICT mine als-o ien ed s udies. Val e de-Ca bonell e al. (2024)
buil a lis o c i ical mine als mos in ensi ely used by ene gy ansi ion echnologies based on an
ex ensi e li e a u e e iew including ecen s udies by In e na ional Ene gy Agency and he Wo ld Bank.
The inal sample consis ed in 42 mine al p oduc s dis ibu ed as ollows: i on p oduc s (6);
aluminum/bauxi e p oduc s (4); coppe p oduc s (4); nickel p oduc s (4); molybdenum p oduc s (3);
li hium p oduc s (3); zinc p oduc s (3); g aphi e p oduc s (2); silicon p oduc s (2); ch omium p oduc s (1);
sil e p oduc s (1); a e ea hs p oduc (1); cobal p oduc (1); niobium, an alum & anadium p oduc s
(1); manganese p oduc (1); lead p oduc (1); u henium, osmium and i idium p oduc (1); hodium
p oduc (1); palladium p oduc (1); and pla inum p oduc (1)15.
2.2.4 Towa d he iden i ica ion o win ansi ion p oduc baske
As men ioned a he beginning o his documen , one o he con ibu ions o his pape is he iden i ica ion
o a baske o p oduc s connec ed wi h win ansi ion based on an ex ensi e li e a u e e iew (See Sec ion
E o ! Re e ence sou ce no ound.). Fo he inal decision on which classi ica ion o digi al and g een
p oduc s o adop , I op ed o hose ha ha e he g ea es possible o e lap be ween bo h ypes o p oduc s.
In his ega d, digi al p oduc s comp ise digi al p oduc ion echnologies p oduc s classi ica ion p oposed
by And eoni e al. (2023). G een p oduc s comp ise g een p oduc s (Mealy & Tey elboym, 2022) and
deca boniza ion echnologies p oduc s (Rosenow & Mealy, 2024). G een-Digi al p oduc s a e hose ha
all a he in e sec ion o he las wo lis s men ioned. C i ical Mine als o g een echnologies comp ise
hose p oduc s collec ed by (Val e de-Ca bonell e al., 2024) (See Annex). E o ! Re e ence sou ce no
ound. displays he numbe o p oduc s o 2000 (HS96) and 2018 (HS17).
15 See Annex A in Val e de-Ca bonell e al (2024) o mo e in o ma ion abou he p oduc s.
20
Table 2: To al Numbe o p oduc s
2000
(HS96)
2018
(HS17)
Digi al
111
127
G een
542
563
Digi al∩G een
47
52
No Digi al
5000
5254
No G een
4569
4818
No Digi al∩No G een
4505
4743
To al p oduc s
5111
5381
Spin-o
C i ical Mine als o g een ech
42
42
Sou ce: own elabo a ion based on BACI-CEPII and li e a u e e iew
To illus a e he p oduc s ha cons i u e he win ansi ion ade baske , Table 3 exhibi s some examples.
Table 3: Desc ip ion he win ansi ion p oduc s
Type
HS
Desc ip ion
G een
850.231
Elec ic gene a ing se s: wind-powe ed, (excluding hose wi h spa k-igni ion o
comp ession-igni ion in e nal combus ion pis on engines).
Digi al
847.180
Da a p ocessing machines: au oma ic, uni s he eo n.e.s. in heading no. 8471.
G een-
Digi al
901.042
Pho og aphic labo a o y appa a us and equipmen (including cinema og aphic):
o he p ojec ion o d awing o ci cui pa e ns on sensi ised semiconduc o
ma e ials, s ep and epea aligne s.
C i ical
Mine als
o g een
ech
281.820
Aluminum oxide: o he han a i icial co undum.
Sou ce: own elabo a ion
21
2.3 Da a
I use ade da a ex ac ed om BACI-CEPII16 o he pe iod 2000-2018. This da ase con ains in o ma ion
abou ade lows by coun y expo e and impo e and by p oduc s classi ied by Ha monized Sys em
(HS) 6-digi nomencla u e. The alue o ade is exp essed in housand USD. This s udy ex ac s
in o ma ion o expo e and impo e coun ies o wo yea s, 2000 and 2018. Fo each ime obse a ion,
he HS e sion na i e o ha yea is used. Consequen ly, he numbe o p oduc codes a ies yea ly (See
Table A 2). I use he ade da a o mo e han 200 coun ies and calcula e he RCA ma ices based on he
comple e da ase . Finally, ou analysis limi s o a se o 80 coun ies (A ica (12), Ame icas (17), Asia
(20), Eu ope (29), and Oceania (2))17 based on he a ailabili y o sus ainable de elopmen indica o s, use
RCA as calcula ed o he mo e ex ensi e se o coun ies.
The HS p oduc s classi ica ion equen ly unde goes e isions o ake in o conside a ion echnological
changes e lec ed in he decline o old and he ise o new p oduc s. This aspec is especially ele an o
win ansi ion ela ed p oduc s. Consequen ly, he p oduc s in expo /impo da a a e no compa able
be ween yea s, and hence, his pape adop s a pooling scheme, as in Nomale & Ve spagen (2024), ha
assumes no such compa abili y, lea ing space o changes in p oduc s and coun ies o dina ion o be
e lec ed in he esul s. In his ega d, each p oduc -yea combina ion and each coun y-yea combina ion
can appea as a single obse a ion in he biplo s. Based on his poin , he au ho s claim ha i is possible
o connec obse a ions o a single coun y o p oduc (o g oups o p oduc s o coun ies) h ough ime,
and hence, i is possible o c ea e a ajec o y o he p oduc (g oup) o coun y (g oup). The pooling
p ocedu e consis s o s acking he wo yea ly coun y a iable ma ices on op o each o he o expo s
and impo s. In his ega d, he RCA ma ices a e s acked, ensu ing ha he o de o he yea s is he same
as he o de used in cons uc ing he Y ma ix. This will o m a la ge ma ix (25,957 by 180).
P oduc s a e also classi ied in e ms o he echnological classi ica ion o expo s in oduced by Lall
(2000) based on he na u e o echnology wi hin indus ies (See Table A 4). Relying on his classi ica ion
o he in e es o his pape is o majo ele ance as allows ela ing g oups o p oduc s wi h echnological
e olu ions (Fage be g & Ve spagen (2021). Fo example, high ech class includes p oduc s linked o he
16 Ve sion 202401b
17 See he de ailed lis in he Annex.
22
las echnological e olu ion pushed by ICT echnologies and expec ed o be close o he digi al p oduc s
g oup iden i ied in Sec ion 2.2.4.
I use 10 di e en sus ainable de elopmen a iables o 80 coun ies. The economic dimension includes
he log o GDP pe capi a in cons an in e na ional $ and he a e age annual compound g ow h a e o
GDP pe capi a (o e 8 yea s wi h yea o he ade da a in he cen e ). The sus ainabili y dimension
includes CO2 emissions pe capi a, consump ion emissions pe capi a, ma e ial oo p in pe capi a,
domes ic ma e ial consump ion pe capi a, and non- enewable ene gy %. The social dimension includes
Gini index, ICT in as uc u e o con empla e digi al di ide and go e nance quali ies (see de ails o he
sou ces in he Annex). All he a iables a e z-sco ed each yea o emo e he in luence o di e ences in
scales and uni s.
3 Resul s
3.2 Twin ansi ion expo iplo s
Figu es 3.1 and 3.2 show he esul s o he CCA o expo s, o 2000 and 2018, espec i ely. These iplo s
display he 11 Lall p oduc ca ego ies and p oduc s linked o win ansi ion (g een, digi al, g een-digi al,
and c i ical mine als demanded by g een echnology). Each ca ego y ep esen s a cen oid, which is a
weigh ed a e age o all he p oduc s in ha ca ego y. The a ows ep esen he a iables linked o
sus ainable de elopmen . The size o he bubbles exp esses, o p oduc classes, he a e age expo
p oduc s ubiqui y, and o coun ies, hei expo di e si y.
The sus ainable de elopmen a iables ep esen ed by he a ows con i m ha he selec ed a iables a e
s ongly and posi i ely co ela ed wi h CCA1, excep o GDP pe capi a g ow h and he Gini index. This
sugges s a ade-o be ween imp o ing economic pe o mance (measu ed by GDP pe capi a), decoupling
he en i onmen , and achie ing social equi y.18 Addi ionally, he diag am con i ms a close o -180 deg ees
18 CCA ope a es unde he assump ion ha he ela ionship be ween he sus ainable de elopmen a iables is linea .
Ne e heless, I acknowledged ha a subs an ial body o li e a u e exis s which es s o non-linea ela ionships among he
a iables. The exis ing li e a u e was de eloped a ound es ing he in e ed U-cu e p oposed by Simon Kuzne s and la e he
U-shaped En i onmen al Kuzne s Cu e (EKC).
29
3.3 Twin ansi ion impo iplo s
Figu e 3.3 and Figu e 3.4 display he co ela ion o sus ainable de elopmen and specializa ion in
impo ing win ansi ion p oduc s. The size o he bubbles exp esses, o p oduc s, he a e age impo
p oduc s ubiqui y, and o coun ies, hei impo di e si y. Compa ed o expo s, a di e en loca ion o
he p oduc s is iden i ied. Fo bo h 2000 and 2018, coun ies' impo baske s a e mo e di e si ied, as can
be seen om he size o he blue bubbles. In 2000, g een, digi al, and g een-digi al p oduc s emain close
o he cen e o he g aph, which means ha hese p oduc s a e impo ed wi h high in ensi y ac oss mos
coun ies. Ano he no ewo hy aspec is ha om 2000 o 2018, impo specializa ion becomes mo e
p onounced, wi h highe in ensi y especially in he e ical (CCA2) dimension. Con a ily, he g oup o
mo e di e si ied coun ies emains in he same posi ion wi h simila di e si ica ion in he ho izon al
(CCA1) dimension. This g oup o coun ies consis s o he ad anced Eu opean economies, he Uni ed
S a es and Japan. Fo example, o bo h yea s, CCA2 e ains g een, digi al and g een-digi al p oduc s
associa ed wi h highe le els o inequali y and high g ow h. Con a ily, c i ical mine als o g een
echnologies emain in he ho izon al dimension (CCA1) o e he yea s close o mo e di e si ied
coun ies.
As was isualized in Figu es in Sec ion 3.1 o expo s, Figu e 3.3 and Figu e 3.4 con i m a clea
specializa ion o impo s ollowing he co e-pe iphe y logic. Fo ins ance, he igh side o he diag am in
which coun ies such as Eu opean ad anced economies, he USA and Japan, a e close o p oduc s PPm,
PPo and c i ical mine al p oduc s associa ed wi h clean echnology. The le side o he diag am con ains
a b oade a ie y o coun ies and is associa ed wi h he es o he coun ies. In u n, i iden i ies ha
be ween 2000 and 2018 all he ca ego ies o p oduc s mo e ou , which sugges s ha he capabili ies o
knowledge equi ed o hei usage became no so simila . The change appea s o be o g ea e magni ude
o p oduc s in he ca ego ies han o hose associa ed wi h he win ansi ion (excep o p oduc s linked
o c i ical mine als). Fu he mo e, bo h igu es con i med ha he ubiqui y o he mos echnologically
ad anced p oduc s is g ea e han ha o he less ad anced ones.
The pa h dependence s o y seems o be epea ed when analyzing he impo specializa ion on win
ansi ion p oduc s. The p ojec ion o g een p oduc s cen oid on sus ainable de elopmen a iables
con i ms ha impo specializa ion on hese p oduc s ocuses on unequal, and high g ow h coun ies.
30
G een impo specializa ion is close o a wide a ie y o coun ies, and by 2000, his ype o p oduc s is
close o medium ech and low- ech echnologies. By 2018, no only he ubiqui y o g een p oduc s
educes, bu also, i emains sepa a ed om he o he p oduc s ca ego ies which can e lec a p ocess o
g een di e si ica ion. In o he wo ds, he capabili ies equi ed o g een impo specializa ion become
mo e speci ic, and hence in ol ing e y educed numbe o coun ies ha a e impo ing “g een
capabili ies” in expo s. I is expec ed ha he inc ease in he di e si y o impo ed g een p oduc s will
po en ially con ibu e o he inc ease in he coun ies’ g een p oduc di e si y. In o he wo ds, he
po en iali ies inc easing he di e si y o impo ed g een p oduc s seem o be e y limi ed acco ding o he
iplo s. A he same ime, he compa able posi ioning o g een p oduc s be ween 2000 and 2018 in he
diag am migh indica e he p esence o pa h dependence in he g een impo specializa ion. Howe e , i is
possible o iden i y a sligh mo emen o he le o he g een do . G een impo specializa ion seems o
be becoming mo e associa ed wi h highe le els o inequali y.
Rega ding digi al p oduc s impo s, o 2000 and 2018, he o dina ion diag am e i ies ha he impo
p oduc i e specializa ion o digi al and digi al-g een p oduc s is associa ed wi h highe GDP g ow h a es.
By 2000, his specializa ion was close o coun ies such as Ko ea and Singapo e. In e es ingly, by 2018,
i is possible o iden i y a shi in coun ies, such as Malaysia, India, and Thailand, ha 2000 we e mo e
associa ed wi h highe le els o inequali y owa ds a highe associa ion wi h highe g ow h GDP. This
p ocess can be unde s ood as a p ocess o assimila ing echnological change and p onouncing he digi al
echnological pa adigm. Also, his diag am con i ms he low engagemen o he leas de eloped coun ies
in he ade o digi al and digi al g een goods.
By 2000, he o dina ion diag am e i ies ha he impo p oduc i e specializa ion o hese c i ical mine als
p oduc s o g een ech is associa ed wi h ele a ed le els o pollu ion, speci ically o non enewable
ene gy. Simul aneously, i is possible o iden i y ha he size o he impo s o hese p oduc s is e y low
wha con i ms ha only ew coun ies specialize in his ype o p oduc s. By 2018, hese p oduc s change
comple ely hei posi ion in he diag am. These p oduc s become being impo ed by mo e coun ies wha
is e lec ed in he size o he pu ple bubble bu also, hey change comple ely hei posi ion in he diag am
and now a e mo e associa ed wi h a iables such as oice and accoun abili y and close o EU coun ies.
In his sense, his could be shedding some ligh on he deepening o he g een ansi ion and he ole o
31
c i ical mine als in sus aining he Eu opean ne -ze o s a egy, he digi al indus y, ae ospace, and de ence
sec o s.
Figu e 3.3 Impo s: Biplo wi h CCA dimensions 1 and 2 (2000, pooling)
Sou ce: idem Figu e 3.1
32
Figu e 3.4 Impo s: Biplo wi h CCA dimensions 1 and 2 (2018, pooling)
Sou ce: idem Figu e 3.1
33
4 Concluding ema ks and limi a ions
T ade wo ldwide is being eshaped by wo majo mega ends: ad anced digi aliza ion o p oduc ion and
he ansi ion owa ds en i onmen ally sus ainable goods. This chap e examines o he i s ime he
pe spec i e o win ansi ion expo and impo di e si ica ion wi hin a mul i-dimensional economic
complexi y app oach and in es iga es whe he his ype o p oduc i e ans o ma ion pe pe ua es pa h-
dependency p ocesses in 80 coun ies p oduc i e specializa ion o e 2000-2018. Fu he mo e, i p o ides
a baske o c ucial p oduc s connec ed wi h he digi al and g een ansi ion.
In his chap e , I a gue ha he mul idimensional economic complexi y app oach con i ms a mo e
comp ehensi e pic u e o unde s anding he esponse o he di e en win ansi ion expo /impo
specialisa ion o sus ainable de elopmen a iables. The implica ions o hese a iables in coun ies ha
begin a p ocess o p oduc i e specializa ion owa ds g een, digi al, g een-digi al p oduc s o c i ical
mine als o he g een ech seem o be e y di e en . On he one hand, expo o g een, g een digi al and
digi al p oduc s is associa ed wi h highe income le els, less inequali y and highe pollu ion le els. The
p oduc i e specializa ion o c i ical mine als o g een ech mine als con i ms being associa ed wi h highe
le els o inequali y and high g ow h. On he o he hand, he impo p oduc i e specializa ion o digi al
and digi al-g een p oduc s is associa ed wi h highe GDP g ow h a es while he specializa ion o g een
impo s is associa ed wi h highe le els o inequali y. Fo he c i ical mine als, his diag am con i ms he
inc easing o he ubiqui y o hese c i ical also, i is con i med a shi om non enewable o oice and
accoun abili y.
The key inding o his esea ch is he con i ma ion o he s ong sou ce o pa h dependence o he win
ansi ion pa e n o specializa ion ha ein o ces he coun y's cen e -pe iphe y ade di ision. G een
expo specializa ion engages only a limi ed g oup o mo e di e si ied de eloped coun ies, which pu s a
isk he possibili y o a b oade ange o de eloping coun ies pa icipa ing in a g een p oduc i e
ans o ma ion p ocess alongside a educ ion in inequali ies. The esul s also sugges ha he possibili ies
o de eloping coun ies o acqui e g een capabili ies h ough impo s a e limi ed oo and associa ed wi h
inequali y o e ime. Con a ily, digi al, and g een-digi al impo specializa ion allow coun ies such as
Ko ea and Singapo e o acqui e echnological capabili ies ha allow hem o consolida e hei posi ion in
he digi al echnological pa adigm. Rega ding c i ical mine als o g een echnologies expo
specializa ion, he indings con i med ha hese p oduc s p esen li le p oximi y wi h o he mo e dynamic
34
p oduc s and pe sis in associa ed wi h poo e coun ies; on he con a y, he impo specializa ion o hese
p oduc s emains associa ed wi h high-income coun ies.
Beyond he win ansi ion p oduc s, he esul s allow iden i ying he exis ence o i uous s uc u al
ans o ma ion p ocess based on acqui ing compa a i e ad an age in adically di e en p oduc s in
opposi ion o he hypo hesis o pa h dependence based on he cu en capabili ies.
Some o he esul s ob ained in his pape can in o m policymake s in he de elopmen o s a egies o
s uc u al ans o ma ion unde win ansi ion. F om he poin o iew o he sou ces o pa h dependence
and b oaden he possibili ies o o he coun ies, specially de eloping, o ansi a digi al and g een pa h o
s uc u al ans o ma ion, he esul s o his pape a e mo e aligned wi h he necessi y o mo e gene al
di e si ica ion policies (Nomale & Ve spagen, 2024c). This mo e gene al di e si ica ion policies mus
ha e as a co e o gene a e he mechanisms o acili a e knowledge among coun ies and p oduc s.
This s udy is limi ed in se e al ways. Fi s , his s udy only conside s in e na ional ade o goods, no
se ices, as he la e is a c ucial componen o digi al echnologies p oduc s. Secondly, due o he opic's
no el y, he magni ude o he numbe o p oduc s linked o win ansi ion p oduc s is ew and o cou se,
only co e a educed sample o coun ies. Thi dly, CCA has he disad an age o assuming ha he
esponse o coun ies and p oduc s o sus ainable dimension is lineal. S ill, I belie e he e is space o s udy
changes in he loca ion o coun ies/p oduc s wi hin a he le el o he indus ies (e.g., All ca ego ies) o
iden i y i he posi ion o coun ies/p oduc s a ies depending on he echnological le el o he p oduc s.
Also, hese ypes o analysis can be used o analyse se s o speci ic coun ies, which I did no con empla e
because hey a e beyond he aim o his chap e . Also, i is possible o include mo e a iables connec ed
wi h sus ainable de elopmen .
35
Re e ences
Acemoglu, D., & Res epo, P. (2022). Demog aphics and au oma ion. The Re iew o Economic S udies,
89(1), 1–44.
And eoni, A., Anzolin, G., Lab unie, M., & Spinola, D. S. (2023). Un eiling s uc u e and dynamics o
global digi al p oduc ion echnology (044; UNU-MERIT Wo king Pape s).
h ps://www.me i .unu.edu/publica ions/wppd /2023/wp2023-044.pd
An onelli, C. (1997). The economics o pa h-dependence in indus ial o ganiza ion. In e na ional
Jou nal o Indus ial O ganiza ion, 15(6), 643–675. h ps://doi.o g/10.1016/S0167-7187(97)00006-
4
Balland, P.-A., B oekel, T., Dioda o, D., Giuliani, E., Hausmann, R., O’Cle y, N., & Rigby, D. (2022).
The new pa adigm o economic complexi y. Resea ch Policy, 51(3), 104450.
h ps://doi.o g/h ps://doi.o g/10.1016/j. espol.2021.104450
Bandei a Mo ais, M., Swa , J., & Jo daan, J. (2018). Economic Complexi y and Inequali y: Does
P oduc i e S uc u e A ec Regional Wage Di e en ials in B azil? U ech School o Economics.,
(No. 18-11).
Bon adini, F., & Vona, F. (2023). Ana omy o G een Specialisa ion: E idence om EU P oduc ion
Da a, 1995–2015. En i onmen al and Resou ce Economics, 85(3–4), 707–740.
h ps://doi.o g/10.1007/s10640-023-00781-7
B en on, P., & Chemu ai, V. (2021). The T ade and Clima e Change Nexus: The U gency and
Oppo uni ies o De eloping Coun ies. The Wo ld Bank. h ps://doi.o g/doi:10.1596/978-1-4648-
1770-0
Calda ola, B., Mazzilli, D., Napoli ano, L., Pa elli, A., & Sba della, A. (2024). Economic complexi y
and he sus ainabili y ansi ion: a e iew o da a, me hods, and li e a u e. Jou nal o Physics:
Complexi y, 5(2), 022001. h ps://doi.o g/10.1088/2632-072X/ad4 3d
Can, M., Ahmed, Z., Ahmad, M., & Oluc, I. (2023). Economic p og ess in eme ging coun ies: he oles
o di e si ica ion o impo and expo p oduc s and ene gy consump ion. En i onmen ,
De elopmen and Sus ainabili y, 26(7), 18207–18229. h ps://doi.o g/10.1007/s10668-023-03386-8
Can, M., & Gozgo , G. (2017). The impac o economic complexi y on ca bon emissions: e idence om
F ance. En i onmen al Science and Pollu ion Resea ch, 24(19), 16364–16370.
h ps://doi.o g/10.1007/s11356-017-9219-7
Coniglio, N. D., Lag a inese, R., Vu chio, D., & A menise, M. (2018). The pa e n o s uc u al change:
es ing he p oduc space amewo k. Indus ial and Co po a e Change, 27(4), 763–785.
h ps://doi.o g/10.1093/icc/d y009
Coniglio, N. D., Vu chio, D., Can o e, N., & Cla a, M. (2021). On he e olu ion o compa a i e
ad an age: Pa h-dependen e sus pa h-de ying changes. Jou nal o In e na ional Economics, 133,
103522. h ps://doi.o g/10.1016/j.jin eco.2021.103522
Da id, P. A. (2007). Pa h dependence: a ounda ional concep o his o ical social science. Cliome ica,
1(2), 91–114. h ps://doi.o g/10.1007/s11698-006-0005-x
de Cunzo, F., Consoli, D., Pe uchas, F., & Sba della, A. (2023). Mapping C i ical Raw Ma e ials in
G een Technologies (2322; Pape s in E olu iona y Economic Geog aphy (PEEG)).
Dieme , A., Iamma ino, S., Pe kins, R., & G os, A. (2022). Technology, esou ces and geog aphy in a
pa adigm shi : he case o c i ical and con lic ma e ials in ICTs. Regional S udies, 1–13.
h ps://doi.o g/10.1080/00343404.2022.2077326
Dosi, G., Pa i , K., & Soe e, L. (1990). Technology and ade: An o e iew o he li e a u e. In The
Economics o Technical Change and In e na ional T ade (Issue 2, pp. 15–39). Labo a o y o
36
Economics and Managemen (LEM), San ’Anna School o Ad anced S udies, Pisa, I aly.
h ps://EconPape s. epec.o g/RePEc:ssa:lemchs:dosie al-1990-2
Fage be g, J., & Ve spagen, B. (2021). Technological Re olu ions, S uc u al Change, and Ca ching up.
In New Pe spec i es on S uc u al Change (pp. 131–155). Ox o d Uni e si y P ess.
h ps://doi.o g/10.1093/oso/9780198850113.003.0007
Fe az, D., Falgue a, F. P. S., Ma iano, E. B., & Ha mann, D. (2021). Linking Economic Complexi y,
Di e si ica ion, and Indus ial Policy wi h Sus ainable De elopmen : A S uc u ed Li e a u e
Re iew. Sus ainabili y, 13(3), 1265. h ps://doi.o g/10.3390/su13031265
Fos e ‐McG ego , N., Nomale , Ö., & Ve spagen, B. (2019). Measu ing he C ea ion and Adop ion o
New Technologies using T ade and Pa en Da a. In UNU-MERIT Wo king Pape Se ies (Vol. 053).
h ps://doi.o g/10.20955/ .85.67
Fos e -McG ego , N., Nomale , Ö., & Ve spagen, B. (2022). The ADB-ADBI Inno a ion and S uc u al
T ans o ma ion Da abase: A Guide. ADB-ADBI Inno a ion and S uc u al T ans o ma ion
Da abase.
F ohbe g, K., & Ha mann, M. (1997). Compa ing measu es o compe i i eness (Issue 2). Ins i u e o
Ag icul u al De elopmen in Cen al and Eas e n Eu ope (IAMO).
h ps://hdl.handle.ne /10419/28566
G eenac e, M. (2010). Canonical co espondence analysis in social science esea ch. Classi ica ion as a
Tool o Resea ch: P oceedings o he 11 h IFCS Biennial Con e ence and 33 d Annual
Con e ence o he Gesellscha Fü Klassi ika ion EV, D esden, Ma ch 13-18, 2009, 279–286.
G eenac e, M. (2017). Co espondence analysis in p ac ice. chapman and hall/c c.
Ha mann, D., Gue a a, M. R., Ja a-Figue oa, C., A is a án, M., & Hidalgo, C. A. (2017). Linking
Economic Complexi y, Ins i u ions, and Income Inequali y. Wo ld De elopmen , 93, 75–93.
h ps://doi.o g/10.1016/j.wo ldde .2016.12.020
Hausmann, R., Hidalgo, C. A., Bus os, S., Coscia, M., & Simoes, A. (2014). The a las o economic
complexi y: Mapping pa hs o p ospe i y. MIT P ess.
Hidalgo, C. A. (2023). The policy implica ions o economic complexi y. Resea ch Policy, 52(9),
104863. h ps://doi.o g/10.1016/j. espol.2023.104863
Hidalgo, C. A., Balland, P.-A., Boschma, R., Delgado, M., Feldman, M., F enken, K., Glaese , E., He,
C., Kogle , D. F., Mo ison, A., Ne ke, F., Rigby, D., S e n, S., Zheng, S., & Zhu, S. (2018). The
P inciple o Rela edness (pp. 451–457). h ps://doi.o g/10.1007/978-3-319-96661-8_46
Hidalgo, C. A., & Hausmann, R. (2009). The building blocks o economic complexi y. P oceedings o
he Na ional Academy o Sciences, 106(26), 10570–10575.
h ps://doi.o g/10.1073/pnas.0900943106
Hidalgo, C. A., Klinge , B., Ba abási, A.-L., & Hausmann, R. (2007). The P oduc Space Condi ions he
De elopmen o Na ions. Science, 317(5837), 482–487. h ps://doi.o g/10.1126/science.1144581
Hill, M. O. (1973). Recip ocal A e aging: An Eigen ec o Me hod o O dina ion. The Jou nal o
Ecology, 61(1), 237. h ps://doi.o g/10.2307/2258931
Jun, B., Alshamsi, A., Gao, J., & Hidalgo, C. A. (2020). Bila e al ela edness: knowledge di usion and
he e olu ion o bila e al ade. Jou nal o E olu iona y Economics, 30(2), 247–277.
h ps://doi.o g/10.1007/s00191-019-00638-7
Lall, S. (2000). The Technological S uc u e and Pe o mance o De eloping Coun y Manu ac u ed
Expo s, 1985‐98. Ox o d De elopmen S udies, 28(3), 337–369.
h ps://doi.o g/10.1080/713688318
Legend e, P., & Legend e, L. (2012). Nume ical ecology. Else ie .
37
Li, G. Y., Ascani, A., & Iamma ino, S. (2024). The ma e ial basis o mode n echnologies. A case s udy
on a e me als. Resea ch Policy, 53(1), 104914. h ps://doi.o g/10.1016/j. espol.2023.104914
Mao, X., Liu, H., Gui, J., & Wang, P. (2023). Towa d inclusi e lis -making o ade libe aliza ion in
en i onmen al goods o educe ca bon emissions. Geog aphy and Sus ainabili y, 4(3), 200–212.
h ps://doi.o g/10.1016/j.geosus.2023.04.002
Mealy, P., Fa me , J. D., & Tey elboym, A. (2018). A new in e p e a ion o he economic complexi y
index. Alexande , A New In e p e a ion o he Economic Complexi y Index (Feb ua y 4, 2018).
Mealy, P., Fa me , J. D., & Tey elboym, A. (2024). In e p e ing economic complexi y. Science
Ad ances, 5(1), eaau1705. h ps://doi.o g/10.1126/sciad .aau1705
Mealy, P., & Tey elboym, A. (2022). Economic complexi y and he g een economy. Resea ch Policy,
51(8), 103948.
Mon iel-He nández, M. G., Pé ez-He nández, C. C., & Salaza -He nández, B. C. (2024). The In insic
Links o Economic Complexi y wi h Sus ainabili y Dimensions: A Sys ema ic Re iew and Agenda
o Fu u e Resea ch. Sus ainabili y, 16(1), 391. h ps://doi.o g/10.3390/su16010391
Ne ke, F., Sba della, A., Sche e , U., & Tacchella, A. (2024). Economic Complexi y Analysis (2430).
Nomale , Ö., & Ve spagen, B. (2022a). Some New Views on P oduc Space and Rela ed Di e si ica ion.
Nomale , Ö., & Ve spagen, B. (2022b). The canonical co ela ion complexi y me hod (015; UNU-
MERIT Wo king Pape s). Maas ich Economic and Social Resea ch Ins i u e on Inno a ion
and …. h ps://unu-me i .nl/publica ions/wppd /2022/wp2022-015.pd
Nomale , Ö., & Ve spagen, B. (2024a). Complexi y Resea ch in Economics: Pas , P esen , and Fu u e.
In Technological Re olu ion and New D i ing Fo ces o Global Sus ainable De elopmen (pp.
109–124). Sp inge Na u e Singapo e. h ps://doi.o g/10.1007/978-981-97-7332-9_15
Nomale , Ö., & Ve spagen, B. (2024b). Rein e p e ing economic complexi y in mul iple dimensions.
A Xi P ep in .
Nomale , Ö., & Ve spagen, B. (2024c). Rela ed o un ela ed di e si ica ion: Wha is sma
specializa ion? S uc u al Change and Economic Dynamics, 70, 503–515.
h ps://doi.o g/10.1016/j.s ueco.2024.05.019
Nomale , Ö., & Ve spagen, B. (2024d). The e olu ion o de elopmen wi h ade in global alue chains
(013; UNU-MERIT Wo king Pape s).
h ps://c is.maas ich uni e si y.nl/ws/ iles/206011394/wp2024-013.pd
Nomale , Ö., & Ve spagen, B. (2024e). The e olu ion o de elopmen wi h ade in global alue chains.
In Fo hcoming (013; UNU-MERIT Wo king Pape s).
h ps://c is.maas ich uni e si y.nl/ws/po al iles/po al/206011394/wp2024-013.pd
OECD. (2011). ICT P oduc s. In OECD Guide o Measu ing he In o ma ion Socie y 2011 (pp. 19–47).
OECD. h ps://doi.o g/10.1787/9789264113541-4-en
Piga o, M., Black, S., Dussaux, D., Mao, Z., Ra a y, R., & Touboul, S. (2020). Technology ans e and
inno a ion o low-ca bon de elopmen . Wo ld Bank Publica ions.
Reins alle , A., & Reschenho e , P. (2019). The impac o he scope o echnological sea ch on pa h-
dependence in expo specializa ion: e idence o Eu opean coun ies. Indus ial and Co po a e
Change, 28(6), 1611–1635. h ps://doi.o g/10.1093/icc/d z026
Robe , V., & Yoguel, G. (2016). Complexi y pa hs in neo-Schumpe e ian e olu iona y economics,
s uc u al change and de elopmen policies. S uc u al Change and Economic Dynamics, 38, 3–14.
h ps://doi.o g/10.1016/j.s ueco.2015.11.004
Rome o, J. P., & G amkow, C. (2021). Economic complexi y and g eenhouse gas emissions. Wo ld
De elopmen , 139, 105317. h ps://doi.o g/h ps://doi.o g/10.1016/j.wo ldde .2020.105317
38
Rosenow, S. K., & Mealy, P. (2024). Tu ning Risks in o Rewa d: Di e si ying he Global Value Chains
o Deca boniza ion Technologies. In Policy Resea ch Wo king Pape s. The Wo ld Bank.
h ps://doi.o g/doi:10.1596/1813-9450-10696
Sau age, J. (2014). The s ingency o en i onmen al egula ions and ade in en i onmen al goods.
Sol , F. (2020). Measu ing Income Inequali y Ac oss Coun ies and O e Time: The S anda dized Wo ld
Income Inequali y Da abase. Social Science Qua e ly, 101(3), 1183–1199.
h ps://doi.o g/h ps://doi.o g/10.1111/ssqu.12795
S ellian, R., & Danna-Bui ago, J. P. (2022). Re ealed Compa a i e Ad an age and Con ibu ion- o- he-
T ade-Balance indexes. In e na ional Economics, 170, 129–155.
h ps://doi.o g/10.1016/j.in eco.2022.02.007
S ojkoski, V., Koch, P., & Hidalgo, C. A. (2023). Mul idimensional economic complexi y and inclusi e
g een g ow h. Communica ions Ea h & En i onmen , 4(1), 130. h ps://doi.o g/10.1038/s43247-
023-00770-0
Tacchella, A., C is elli, M., Calda elli, G., Gab ielli, A., & Pie one o, L. (2012). A New Me ics o
Coun ies’ Fi ness and P oduc s’ Complexi y. Scien i ic Repo s, 2(1), 723.
h ps://doi.o g/10.1038/s ep00723
Talebzadehhosseini, S., Scheine , S. R., & Ga ibay, I. (2020). G owing G een: he Role o Pa h
Dependency and S uc u al Jumps in G een Economy Expansion. Jou nal on Policy and Complex
Sys ems, 6(1). h ps://doi.o g/10.18278/jpcs.6.1.2
Te B aak, C. J. F. (1986). Canonical Co espondence Analysis: A New Eigen ec o Technique o
Mul i a ia e Di ec G adien Analysis. Ecology, 67(5), 1167–1179.
h ps://doi.o g/10.2307/1938672
Te B aak, C. J. F. (1987). The analysis o ege a ion-en i onmen ela ionships by canonical
co espondence analysis. Vege a io, 69(1–3), 69–77. h ps://doi.o g/10.1007/BF00038688
Te B aak, C. J. F. (2014). His o y o canonical co espondence analysis. Visualiza ion and
Ve baliza ion o Da a, 61–75.
Te B aak, C. J. F., & Ve donscho , P. F. M. (1995). Canonical co espondence analysis and ela ed
mul i a ia e me hods in aqua ic ecology. Aqua ic Sciences, 57, 255–289.
Val e de-Ca bonell, J., Pie obelli, C., & Menéndez, M. de las M. (2024). Mine als’ c i icali y and
coun ies’ mining compe i i eness: Two aces o he same coin. Resou ces Policy, 98, 105359.
h ps://doi.o g/10.1016/j. esou pol.2024.105359
an Dam, A., Dekke , M., Mo ales-Cas illa, I., Rod íguez, M. Á., Wichmann, D., & Baudena, M.
(2021). Co espondence analysis, spec al clus e ing and g aph embedding: applica ions o ecology
and economic complexi y. Scien i ic Repo s, 11(1), 8926. h ps://doi.o g/10.1038/s41598-021-
87971-9
Zhang, H., Liu, Q., & Wei, Y. (2023). Digi al p oduc impo s and expo p oduc quali y: Fi m-le el
e idence om China. China Economic Re iew, 79, 101981.
h ps://doi.o g/h ps://doi.o g/10.1016/j.chieco.2023.101981
