Gomes, O lando
A icle
Economic g ow h in he age o ubiqui ous h ea s: How
global isks a e eshaping g ow h heo y
Economics: The Open-Access, Open-Assessmen Jou nal
P o ided in Coope a ion wi h:
De G uy e B ill
Sugges ed Ci a ion: Gomes, O lando (2024) : Economic g ow h in he age o ubiqui ous h ea s: How
global isks a e eshaping g ow h heo y, Economics: The Open-Access, Open-Assessmen Jou nal,
ISSN 1864-6042, De G uy e , Be lin, Vol. 18, Iss. 1, pp. 1-15,
h ps://doi.o g/10.1515/econ-2022-0059
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/306075
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/
Re iew A icle
O lando Gomes*
Economic G ow h in he Age o Ubiqui ous
Th ea s: How Global Risks a e Reshaping
G ow h Theo y
h ps://doi.o g/10.1515/econ-2022-0059
ecei ed July 13, 2023; accep ed No embe 24, 2023
Abs ac : One o he mos ou s anding accomplishmen s o
he economic science o e he las decades is he de elop-
men o a sound and cohe en heo y o economic g ow h.
Resea ch in g ow h heo y has demons a ed ha signifi-
can and sys ema ic inc eases in well-being a e a ainable
whene e he igh o mula is implemen ed. When com-
bined wi h efficiency, he ing edien s o his o mula –
inno a ion, he diffusion o ideas, and human capi al accu-
mula ion –can d i e he economy owa d a i uous pa h
o sus ained g ow h. No wi hs anding, his is an o e ly
op imis ic iew o g ow h ha does no accoun o he
many obs acles ha he c ea ion o weal h may encoun e .
The cu en essay su eys cu ing-edge esea ch on g ow h
heo y o conclude in a o o a pa adigm shi : he main
conce n is no longe jus wi h how o co ec ly combine
p oduc ion inpu s, bu wi h how hei efficien use is e en-
ually hampe ed by a la ge collec ion o wo ldwide isks
and h ea s. Global isks come in many shapes ( hey can be
classified as economic, en i onmen al, geopoli ical, soci-
e al, and echnological) bu , in any case, hey call o a
eexamina ion o g ow h heo y.
Keywo ds: g ow h heo y, g ow h models, global isks, eco-
nomic disas e s, a e e en s
JEL classifica ion: O41, O33, O43, O44
1 In oduc ion
The Wo ld Economic Fo um, an independen in e na ional
o ganiza ion whose main pu pose is o os e public-p i-
a e coope a ion a he highes le els o decision-making,
publishes e e y yea , since 2006, he Global Risks Repo .
The aims and scope o his publica ion consis o a ho -
ough sys ema iza ion and assessmen o he main and mos
p essing h ea s ha humani y cu en ly aces. The epo
defines global isk as “ he possibili y o he occu ence o
an e en o condi ion which, i i occu s, would nega i ely
impac a significan p opo ion o global GDP, popula ion
o na u al esou ces.”(2023 Repo , p. 5). As cha ac e ized,
he no ion o isk should be in e p e ed loosely, o include
e e y dange , menace, and po en ial disas e ha h ea-
ens ou agile collec i e exis ence.
The men ioned epo compa men alizes isks in o
fi e b oad ca ego ies: economic, en i onmen al, geopoli-
ical, socie al, and echnological. The con en s o each
ca ego y a e sel -explana o y. On he economic on ,
mac oeconomic isks a e highligh ed; hese include he p o-
spec o economic s agna ion and ecessions, ising infla ion
and unemploymen , asse bubbles, and deb c ises, especially
in la ge economies. Also ele an , ega ding he h ea s posed
o he wo ld economy, a e he possibili y o commodi y shocks,
he collapse o supply chains, and he p oli e a ion o illici
ac i i ies, such as o ganized c ime, ade in coun e ei goods,
and ax e asion and aud.
In wha conce ns he second ca ego y, he en i onmen ,
a long lis o h ea s can also be enuncia ed, including human-
made en i onmen al damages, o e exploi a ion and misman-
agemen o c i ical na u al esou ces, clima e change, he loss
o biodi e si y, ex eme wea he e en s, and geophysical dis-
as e s. Geopoli ical isks encompass e o ism, he h ea
posed by weapons o mass des uc ion, geoeconomic and
geopoli ical con on a ions, ci il wa s, and he dismembe -
men o mul ila e al o ganiza ions and a angemen s.
The socie al ca ego y co e s a wide a ay o isks, om
hose associa ed wi h he sp ead o in ec ious diseases
* Co esponding au ho : O lando Gomes, Lisbon Accoun ing and
Business School –Lisbon Poly echnic Ins i u e (ISCAL-IPL) and CEFAGE
(Uni . É o a –ISCAL) Resea ch Cen e , ISCAL, A . Miguel Bomba da, 20,
1069-035 Lisboa, Po ugal, e-mail: [email p o ec ed],
el: +351 217 984 500
ORCID: O lando Gomes 0000-0002-7251-8736
Economics 2024; 18: 20220059
Open Access. © 2024 he au ho (s), published by De G uy e . This wo k is licensed unde he C ea i e Commons A ibu ion 4.0 In e na ional License.
(epidemics and pandemics) o many o he issues aised by
ou coexis ence in socie y (e.g., he e osion o public ins i-
u ions and social cohesion, he de e io a ion o wo king
condi ions and job oppo uni ies, he disillusionmen o he
you h, o he eme gence o la ge-scale in olun a y mig a-
ions). Finally, echnological isks a e, cu en ly, associa ed
wi h cybe c ime and cybe espionage, digi al inequali y
and digi al powe concen a ion, and, among o he s, he
e en ual inad e en o malicious b eakdown o c i ical
in o ma ion in as uc u es.
Mos o he a o emen ioned isks do no mani es
hemsel es in isola ion. Al hough a global c isis may e up
om a single seed o dys opia, his can sp ead as , o o he
a eas o he economy and he socie y, c ea ing wha one
migh designa e as a pe ec s o m. The COVID-19 pandemic
and he ecen escala ion o geopoli ical ensions a e wo
p o o ypical examples o seeds o dys opia ha ueled he
up ise o many o he meaning ul h ea s (e.g., soa ing
infla ion, inc easingly wo isome cybe -secu i y b eaches,
deepe social agmen a ion, massi e e ugee c ises, o he
ise o inequali ies wi hin and among coun ies). The sub-
s an ia ion o some o he enuncia ed h ea s, and e en he
me e pe cep ion ha hey migh somehow ma e ialize,
may se iously hampe economic g ow h in a a ie y o
ways. The challenge ha g ow h heo is s nowadays ace
is p ecisely o inco po a e hese h ea s in o hei models
and o effec i ely explain how hey migh influence he
pace o ma e ial p og ess as we know i oday.
This essay unde akes a selec i e su ey o g ow h
heo y (o con ibu ions published om 2020 onwa d) o
cla i y ha mos ecen addi ions o he heo y acknowl-
edge and a e awa e o he main obs acles ha wo ldwide
economic g ow h aces. This con as s wi h ea lie con i-
bu ions, which we e much mo e ocused on efficiency
issues and on how coun ies should success ully combine
he a ailable physical, human, and echnological inpu s
wi h he objec i e o maximizing in e empo al u ili y.
Al hough he aim con inues o be he same, i.e., he p omo-
ion o ma e ial well-being, he ocal poin is ha schola s
ha e, oday, a much clea e pe cep ion ha he exis ing
isks migh h ea en he efficacy o he con en ional o -
mulas leading o sus ained economic g ow h.
The emainde o he a icle is o ganized as ollows.
Sec ion 2 highligh s he pieces o li e a u e ha di ec ly
and gene ically app oach he impac o p obable signifi-
can disas e s and a e e en s on he economy’s g ow h
a e. Special ocus is placed on an analy ical amewo k
capable o quan i ying he g ow h effec s o a disas e and,
also, he g ow h impac o he isk i sel . In Sec ion 3, ech-
nological isks a e add essed. The ele an li e a u e is
su eyed, and he ypical endogenous g ow h model is
ein e p e ed in ligh o he p esence o an addi ional
inpu : obo ic capi al and/o a ificial in elligence.
Sec ion 4 p oceeds wi h a eflec ion on he in e play
be ween he sp ead o in ec ious diseases and economic
g ow h. To sha e ideas and knowledge, human con ac is
equi ed; howe e , wi h inc eased human con ac comes
he possibili y o as e dissemina ion o diseases. The
wo ldwide as dissemina ion o he COVID-19 pandemic
was he di ec consequence o he globalized and in e con-
nec ed wo ld we li e in oday, which leads o an undeni-
able piece o e idence: he close he globaliza ion p ocess
b ings us oge he , he s onge i becomes he isk o ca -
as ophic public heal h e en s. In Sec ion 5, en i onmen al
isks a e highligh ed. En i onmen al conce ns a e p og es-
si ely becoming an insepa able pa o g ow h analyses.
This is illus a ed by explo ing an adap ed e sion o a
ecen ly p oposed model o clima e change and g ow h.
In Sec ion 6, geopoli ical isks a e b iefly deba ed. These
may ake many o ms and hey can be associa ed wi h g ow h
models in many ways. A ypical neoclassical g ow h model
allowing o poli ical ins abili y is cha ac e ized o illumina e
he impac ulness o his ype o isk.
In Sec ion 7, a ew addi ional no es on economic isks
a e added o he su ey, and Sec ion 8 concludes.
2 The Accommoda ion o Risks and
Disas e s in G ow h Models
Global economic g ow h is subjec o a wide a ie y o
isks. Al hough hese may be somehow in e connec ed,
hey a e diffe en in na u e, and he e o e, as expec ed,
diffe en s ands o li e a u e deal wi h he impac o dis-
simila h ea s in dis inc ways, as he sec ions ha ollow
will highligh . Despi e his di e si y, he e a e a ew ecen
s udies ha add ess, in a gene ic and abs ac way, he po en-
ial impac o menaces and ac ual disas e s on g ow h. These
include Ba o and Jin (2021), Douenne (2020), Hao e al. (2020),
Jo ano ic and Ma (2022), and K ishna e al. (2023). The
common poin in he men ioned s udies is he p esence o
unce ain y associa ed wi h some aspec s o he g ow h p o-
cess: he ou come o he adop ion o new echnologies migh
be unce ain, in es men decisions migh be unp edic able,
o s ochas ic a e e en s may cause un o eseeable changes in
consump ion.
The model by Douenne (2020) is pa icula ly well-
sui ed o app oach he impac o isks on g ow h. I is a
ela i ely s anda d op imal g ow h model whe e he com-
bina ion o ecu si e u ili y wi h a s ochas ic capi al
2O lando Gomes
accumula ion p ocess allows o he quan ifica ion o he
effec o disas e isks and o he consequences o ac ual
disas e s o e he g ow h a e de i ed unde an endo-
genous g ow h se up. In his sec ion, Douenne’s amewo k
is eco e ed, and i s discussion is u he ex ended.
Le
K
() ep esen he s ock o physical capi al a da e
. In his se ing, he isk is defined as he p obabili y o
occu ence o a nega i e shock affec ing he s ock o capi al. I
he shock ma e ializes,
K
() alls o
∼=∈
K
ωK ω,0,1() () (
)
;
he lowe he alue o pa ame e
ω
, he s onge a e he
damages caused by he disas e . Al hough he impac o he
disas e on g ow h is unequi ocally nega i e, he isk ha i
poses may accele a e o decele a e g ow h, depending on he
effec on consump ion and sa ings. When aced wi h a isk,
he ep esen a i eagen mayei he ans e consump ion
om he p esen o he u u e (p ecau iona y sa ings) o
he o he way a ound (p ecau iona y consump ion). The
key elemen in his ega d is he in e empo al elas ici y o
subs i u ion: a low elas ici y o subs i u ion (lowe han 1)
s imula es an inc ease in sa ings; a high elas ici y o subs i u-
ion (highe han 1) leads o an an icipa ion o consump ion in
he aceo he isk.E iden ly, hep ecau iona ysa ingssce-
na io is he one leading o as e long- e m g ow h.
Douenne’s model is pa icula ly appealing because i
allows o a clea dis inc ion be ween he no ions o in e -
empo al elas ici y o subs i u ion and he coefficien o
ela i e isk a e sion (CRRA). This sepa a ion is easible
i he agen ’s p e e ences a e ep esen ed h ough a ecu -
si e u ili y unc ion o he Eps ein-Zin ype. In his model,
he u ili y unc ion akes he o m:
=−⎧
⎨
⎩+−)+
⎫
⎬
⎭
−− −
−
−
−
U
γC e γ U
1
1d1 d
.
θρ θ
γ
γ
θ
1d 1
1
1
1
() () [( ( )] (1)
In exp ession (1), C (
)
and U ()s and o consump ion
and u ili y, espec i ely. The ope a o
designa es he expec-
a ion abou u u e u ili y. The pa ame e s a e he ollowing:
≥ρ
0
is he a e o ime p e e ence; ∈+∞θ0, 1[){
}
is he
in e se o he in e empo al elas ici y o subs i u ion; and
∈+∞γ0, 1[){
}
is he CRRA; he highe he alue o
γ
, he
s onge is he a e sion o isk.
The maximiza ion o u ili y is subjec o a cons ain on
he accumula ion o capi al. This is a s ochas ic diffe en ial
equa ion, which is ep esen ed unde he ollowing o m:
∑
=− − −
∼
=
K Y C K K q
d
dd
.
i
n
ii
1
() [ () ()] [ () ()] () (2)
In equa ion (2), Y () ep esen s ou pu . In o de o
gua an ee he ac abili y o he model and also an endo-
genous g ow h ou come, cons an ma ginal e u ns a e
aken, such ha =>Y AK A,0() () . In he exp ession,
he accumula ion o capi al can be hi by ndiffe en shocks
o ampli ude −∼=−
K
K ωK 1
ii
() () ( )(); he equency
o he shocks is de e mined by a Poisson p ocess, such
ha =q λ
dd
ii
() ep esen s he p obabili y o he occu -
ence o a disas e o ype i;
λ
i
is a posi i e pa ame e . No
o he e en ual fluc ua ions besides hose igge ed by he
a e ca as ophic e en s a e conside ed in his se up, as
p esen ed.
Fo he cha ac e ized dynamics, he a e age g ow h o
he s ock o capi al is w i en as ollows:
∑
⎡
⎣
⎢⎤
⎦
⎥=− − −
=
K
K AC
K ωλ
d/d 1
.
i
n
ii
1
()
() ()
() () (3)
The maximiza ion o u ili y in (1) subjec o capi al
accumula ion cons ain (2) equi es employing op imal
con ol echniques o s ochas ic p oblems. Following he
same p ocedu e as in Douenne (2020), he compu a ion
o he Hamil on–Jacobi–Bellman equa ion conduc s o an
op imal solu ion in which he consump ion–capi al a io is
cons an . Unde he p oposed o mula ion, on he op imal
pa h, he consump ion–capi al a io is as ollows:
∑
⎜⎟
=+
−⎛
⎝−−−
⎞
⎠
=
−
C
K ρ
θθθAωγλ
11
1
.
i
niγ
i
1
1
()
() (4)
The impac o isk and isk a e sion o e he consump-
ion-capi al a io is con ingen on he alue o he elas ici y
o in e empo al subs i u ion, −
θ
1
.I <
−
θ
1
1, hen he con-
sump ion-capi al a io dec eases wi h isk (wi h a highe
p obabili y o disas e s –highe
λ
i
–and wi h a highe
in ensi y o disas e s –lowe ωi), and wi h isk a e sion
(highe
γ
). The opposi e esul s a e ob ained o >
−
θ
1
1.I
=θ
1
, hen he consump ion–capi al a io is equal o he
a e o ime p e e ence. When <
−
θ
1
1, people inc ease sa -
ings in he ace o a gi en isk; his is a scena io o p ecau-
iona y sa ings. When >
−
θ
1
1, people p e e o inc ease
consump ion when con on ed wi h highe unce ain y, a
phenomenon ha can be designa ed as p ecau iona y
consump ion.
Subs i u ing he op imal consump ion–capi al a io in
(4) in o (3), one ob ains he expec ed g ow h a e o he
a e age long- un g ow h a e (o capi al, consump ion, and
income),
∑∑
≡⎡
⎣
⎢⎤
⎦
⎥=⎡
⎣
⎢⎤
⎦
⎥=⎡
⎣
⎢⎤
⎦
⎥
=−+
−−
−−−
=
−
=
g
K
K C
C Y
Y
θAρ θθωγλωλ
d/d d/d d/d
111
11
.
i
niγ
ii
n
ii
1
1
1
()
() ()
() ()
()
() ()
(5)
The g ow h a e in exp ession (5) in ol es h ee e ms
wi h diffe en meanings. The fi s e m co esponds o he
Economic G ow h in he Age o Ubiqui ous Th ea s 3
no- isk ou come, he well-known Eule equa ion esul ,
acco ding o which he pace o g ow h is essen ially de e -
mined by he diffe ence be ween he ma ginal e u n on
capi al and he a e o ime p e e ence. The second e m
ep esen s he impac o isks on g ow h; he isk may
inc ease he pace o g ow h i <
−
θ
1
1(p ecau iona y sa -
ings igge highe long- e m g ow h). The hi d e m is he
impac o he ac ual occu ence o he disas e , which is
necessa ily nega i e. Hence, he p oposed amewo k has
he me i o sepa a ing he consequences o he h ea om
hose o he disas e i sel ; hey will bo h nega i ely influence
wel a e, bu hei join effec ong ow hmigh no benega i e
i he isk induces, o a la ge ex en , p ecau iona y sa ings.
The abo e easoning conside s mul iple isks (n isks,
o be p ecise) bu no associa ion be ween hem. As men-
ioned in he in oduc ion, he h ea o a la ge-scale ne a -
ious e en (e.g., a pandemic o a wa ) is jus a p obable
seed o dys opia ha easily sp eads o many o he a eas o
socie y o he economy. Hence, one may concei e a scena io
in which an ini ial high-p obabili y–high-in ensi y isk is
jus he fi s s ep in a chain o o eseeable e en s wi h p o-
g essi ely lowe in ensi y and p obabili y o occu ing. A
s ylized o m o ep esen ing he abo e easoning consis s
o aking a fi s isko p obabili y
=
λλ
1
and in ensi y
=ω
ω
1
, and a se ies o subsequen isks obeying condi ions
=∈
+
λ
ϕλ ϕ,0,1
ii1(
)
and =∈
+
ωδωδ,0,1
ii1()
.
Taking →∞
n
(i.e., an infini e se ies o p og essi ely
lowe p obabili y –lowe in ensi y po en ial disas e s), he g ow h
a e in exp ession (5)isp esen ableasa unc iono heeigh
ele an pa ame e s o he model (Aρθγλωϕ
δ
,,,,,,,):
=−+
−−
−−−
−−
−−−−
−−
−−
−
g
θAρ θθλγϕδ ϕ ω
ϕϕδ
λϕδ ϕ ω
ϕϕδ
11
111
11
11
11 .
γγ
γ
11
1
() ()
()( )
()
()()
(6)
I isks ollowing he ini ial h ea a e o some signifi-
cance, meaning ha he alues o
ϕ
and
δ
a e ela i ely
high, hen he ini ial effec o he seed o dys opia is p o-
longed in ime. This effec is clea ly nega i e in espec o
he impac o he disas e . Howe e , as ema ked ea lie , i
can be ei he posi i e o nega i e ega ding he isk i sel ,
gi en he alue o pa ame e
θ
.
Douenne (2020) in oduces an addi ional ele an opic,
namely he possibili y o delibe a e isk mi iga ion. In wha
espec s global isks, he effo o lowe hem equi es
an in e na ional coo dina ion o effo s, because he la ge
majo i y o he al eady highligh ed isks a e associa ed wi h
global commons (e.g., he p ese a ion o he en i onmen ,
peacekeeping, o he p e en ion o in ec ious diseases).
Because ee iding is una oidable, he in e na ional com-
muni y should a leas gua an ee a coali ion o he willing.
Analy ically, in he con ex o he model, isk mi iga-
ion consis s o di e ing a sha e o income, ∈
τ
0,1(
)
, o
educe he p obabili y o he disas e . Unde isk alle ia-
ion, he p obabili y o a disas e alls om
λ
o
−
λ
τ1α
(
)
,
<<α0
1
. Sol ing he model in his scena io yields an
op imal esul o sha e
τ
,
=⎡
⎣
⎢−−⎤
⎦
⎥
−−
τ
ωλα
Aγ
11
.
γα
11
1
()
()
(7)
One conside s ha he isk educ ion effo is exe ed
only upon he fi s isk ( he seed o dys opia). Because all
o he isks depend on he fi s , he isk educ ion sp eads
o e all po en ial subsequen disas e s. In his case, he
op imal consump ion–capi al a io is
=+
−⎡
⎣
⎢−
−− −
−−−
−− ⎤
⎦
⎥
−−
−
C
K ρ
θθθτA
τλγϕδ ϕ ω
ϕϕδ
11
1111
11 ,
αγγ
γ
11
1
()
() ()
() ()
()( )(8)
and he expec ed g ow h a e comes,
=−−+
−
−−
−−−
−−
−− −−−
−−
−−
−
g
θτA ρ θθ
τλγϕδ ϕ ω
ϕϕδ
λτ ϕδ ϕ ω
ϕϕδ
1111
111
11
111
11 .
αγγ
γ
α
11
1
[( ) ] (
)()
()( )
() ()
()()
(9)
The p e en ion o disas e s has a nega i e di ec impac
on g ow h because i di e s esou ces om capi al accumu-
la ion, bu i has a posi i e effec ia disas e a oidance. The
g ow h effec ia isk is, again, dependen on he in e em-
po al elas ici y o subs i u ion.
The cha ac e ized model is gene al enough o be
applicable o any kind o global isk. Howe e , diffe en
ypes o isks ha e specifici ies, conce ning g ow h, ha
a e wo h explo ing. This explo a ion begins in he ol-
lowing sec ion, wi h a discussion abou h ea s o a ech-
nological na u e.
3 The Wonde s o Au oma ion and
A ificial In elligence: Wha Can
Go W ong?
The p og ess associa ed wi h compu a ional capabili ies
and a ificial in elligence opens new significan p omising
p ospec s ega ding long- e m g ow h. In his espec , a
pe inen ques ion is aised by No dhaus (2021): a e we
4O lando Gomes
heading owa d a singula i y poin , i.e., owa d a momen
in his o y in which, wi hou much human in e en ion,
g ow h could accele a e u he and u he ? This idyllic
scena io is apidly disca ded by he au ho , based on
empi ical es ima es and he use o a ew logical a gumen s.
The s ong idea is ha echnological wonde s a e necessa-
ily accompanied by ele an echnological isks ha mus
be accoun ed o in o de o p e en majo echnological
disas e s.
Technological isks a e an una oidable side effec o
he p og essi e sophis ica ion o digi al ools and o he
echnical no el ies. Such ools ely on inc easingly high
le els o connec i i y and in eg a ion, which is necessa ily
accompanied by ising ulne abili ies. One mus no o ge
ha he echnologies ha os e g ow h a e he same ech-
nologies ha can be used o c iminal ac i i ies, espionage,
and o he audulen and des uc i e ac i i ies. Mo eo e , he
e en ual pa h owa d he c ea ion o supe -in elligen machines
can be a h ea on i s own, because wi h in elligence comes he
abili y o eason and o c ea e and ame mo al no ms. Fo
hese easons, and o he s (namely, he sca ci y and non- enew-
able na u e o mos physical esou ces), i is sa e o asse ha
we a e no heading owa d a singula i y.
Mos o he endogenous g ow h li e a u e ha equa es
he ole o au oma ion and a ificial in elligence is a li le
bi mo e down o ea h han wha he abo e pa ag aphs
migh sugges . The main conce n ha anspi es om such
li e a u e espec s o he sho - and medium- e m impac
o he new echnologies on employmen and income dis-
ibu ion. These new echnologies suppo a new o m o
capi al ha , unlike physical capi al, is a subs i u e and no
a complemen o labo . Recen s udies add essing au oma ion
and g ow h include Acemoglu and Res epo (2022), Abe-
liansky and P e ne (2023), Gas eige and P e ne (2022),
Hémous and Olsen (2022), Kla l (2022), Lu (2021, 2022), Moll
e al. (2022), I men (2021), P e ne and S ulik (2020), Ray and
Mookhe jee (2022), and Sasaki (2023).
The abo e-men ioned esea ch p oposes a wide a ie y
o models and amewo ks ha a e dis inc in hei s uc u e
and app oach, bu ha sha e some common g ound: in any o
he cases, au oma ion eplaces labo (a leas low-skilled
labo ), and i allows o enhanced p oduc i i y. A he end
o he day, he new p oduc ion capabili ies a e likely o os e
g ow h, bu one should no jump immedia ely o his conclu-
sion. Wi h au oma ion comes he pola iza ion o jobs and
wages and he concomi an inc ease in income inequali y
(low-skilled wo ke s lose o high-skilled wo ke s and capi al
owne s). As a significan pa o he popula ion loses income,
wo po en ially damaging consequences eme ge: a all in
agg ega e demand and an inc ease in social discon en men .
These colla e al effec s migh o e come he p oduc i i y gains
om au oma ion, in wha espec s g ow h and, mos e i-
den ly, in wha conce ns social wel a e.
Accoun ing o au oma ion in s anda d g ow h ana-
lysis equi es adding a new inpu o he sho lis o p o-
duc ion ac o s ha a e ypically assumed. This new inpu
is obo ic capi al (au oma ed machines and p ocesses, and a i-
ficial in elligence algo i hms). As highligh ed by Abeliansky and
P e ne (2023), obo ic capi al mixes ea u es o bo h adi ional
inpu s: i is like labo , becausei occupies hesame oleas
human labo in he p oduc ion p ocess, and i is like capi al,
because i can be accumula ed and i ep esen s he non-human
con ibu ion o p oduc ion. In he s udy by Lu (2021), he au o-
ma ion inpu is di ec ly in e p e ed as a ificial in elligence.
This is a special o m o capi al, wi h singula and non- i ial
p ope ies. I has simila i ies wi h human capi al, because i can
lea n and accumula e knowledge by i sel ; i has simila i ies
wi h ideas, because hey a e bo h non i al.
In he s udy by Bloom e al. (2023), he dis inc ion
be ween indus ial obo s and a ificial in elligence is made
analy ically explici . These a e in e p e ed as wo sepa a e
inpu s in p oduc ion. Roughly speaking, while obo s a e a
eplacemen o low-skilled wo ke s in he de elopmen o
ou ine asks, a ificial in elligence subs i u es o high-skilled
wo ke s, who pe o m non- ou ine c ea i e asks. This dis-
inc ion is impo an and has consequences o he o ganiza-
ion o wo k, income dis ibu ion and, ul ima ely, g ow h.
While con en ional au oma ion p ocesses place downwa d
p essu e on he wages o low-skilled wo ke s, hus
inc easing he skill p emium (i.e., he alue o he a io
be ween he wages o high-skilledwo ke sandlow-
skilled wo ke s), Cha GPT and ela ed echnologies ha epli-
ca e human hinking and c ea i i y p edominan ly influence
he wages o high-skilled labo (al hough hese cu ing-edge
echnologies may pu p essu e as well on he subsis ence o
low-skilled jobs).
Consequen ly, a ificial in elligence is likely o con i-
bu e o a educ ion in he skill p emium, as he pe o -
mance o high-skilled asks becomes no longe exclusi e
o imagina i e human minds. In ei he case (i.e., whe he
machines and algo i hms eplace humans in he comple-
ion o ou ine ac i i ies o cogni i e demanding asks),
he phenomenal inc ease in he efficiency o echnologies
ha eplace human pa icipa ion and effo in p oduc ion
h ea ens jobs, wel a e, and also g ow h, as i concen a es
he means o p oduc ion in he hands o a ew capi al
owne s in de imen o he la ge a my o wo ke s ha
popula e socie y.
Besides indus ial obo s and Cha GPT-like echnolo-
gies, ye ano he no el inpu migh be conside ed o com-
pose he agg ega e p oduc ion unc ion ha unde lies
g ow h analyses. This inpu is big da a (Cong e al., 2022),
Economic G ow h in he Age o Ubiqui ous Th ea s 5
and i diffe s om obo ic capi al/a ificial in elligence in
he sense ha i is no a subs i u e o labo . Howe e ,
hese ac o s also sha e some p ope ies: hey a e non i al
and, unlike human capi al, hey can be de ached om
people and concen a ed in he hands o a ew, hus con-
ibu ing o s ong le els o income and weal h inequali y.
Besides his, he use o da a aises ano he c i ical isk o
people, namely he isk associa ed wi h hei p i acy.
Based on he men ioned li e a u e, a syn hesis model
can be compiled. S a by assuming a Cobb–Douglas p o-
duc ion unc ion, wi h obo ic capi al deno ed by
R
(
)
,
=+>∈
−
Y AK L R A α,0, 0,1
.
αα1
() ()[() ()] ( ) (10)
In equa ion (1), s anda d no a ion is adop ed: Y K ,
,
() ()
and L( ) ep esen , espec i ely, ou pu , physical capi al, and
labo ( o he sake o he exposi ion le his las a iable be
cons an o e ime, L( )=L); Ais he p oduc i i y index and α
he ou pu –capi al elas ici y. The subs i u abili y be ween labo
and obo ic capi al is e iden om he exp ession; in he limi ,
i all labo is eplaced by machines, p oduc ion is s ill possible.
Define ≡ R
Lϱ
/() () and assume he commonly used
no a ions o pe capi a income and capi al. Equa ion (10)
is equi alen o i s in ensi e o m coun e pa ,
=+
−
y
Ak 1ϱ
.
αα1
() ()[ ()]
(11)
In a compe i i e economy, ac o e u ns co espond
wi h hei espec i e ma ginal p oduc s. In he de ised
scena io, he wage a e is iden ical o he a e o e u n
on obo ic capi al,
==−
⎡
⎣
⎢+⎤
⎦
⎥
w αA k
11ϱ
.
R
α
() () ( ) ()
() (12)
The a e o e u n on physical capi al is:
=⎡
⎣
⎢+⎤
⎦
⎥
−
αA
k
1ϱ
.
α1
() ()
()
(13)
Unde his simple o mula ion, i is s aigh o wa d o
obse e ha he labo income sha e alls wi h an inc ease
in he employmen o obo ic capi al:
=−
+
w
y α
1
1ϱ
.
()
() () (14)
In con as , i one defines capi alis s as he agen s who
hold any o m o capi al (physical and obo ic), hei
income sha e is:
+=+
+
k
y α
ϱϱ
1ϱ
.
R
()() ()()
() ()
()
(15)
F om exp ession (15), one concludes ha as he pa i-
cipa ion o obo ic capi al in p oduc ion inc eases, he
income sha e o capi alis s inc eases as well.
The abo e logical a gumen a ion can be ex ended o
include he sepa a ion be ween indus ial obo s and a i-
ficial in elligence (Bloom e al., 2023). In his case, p oduc-
ion unc ion (10) can be augmen ed by spli ing he human
wo k o ce in o unskilled labo (
L
u
) and skilled labo (
Ls
),
and by dis inguishing be ween obo ic capi al,
R
(
)
, and
a ificial in elligence,
R
AI(). As ema ked ea lie ,
R
(
)
is
a di ec eplacemen o
L
u
, while
R
AI()is a di ec eplace-
men o
Ls
. A concei able enla ged p oduc ion unc ion
is, hen,
=+++
>∈
−
Y AK L R L R
Aαβ
,
0, , 0,1 .
αuβsβαβ
AI 1/
() (){[ ()] [ ()]}
()
()
(16)
In in ensi e o m,
=+++
−
y
Ak l lς ϱ
.
αuβsβαβ1/
() (){[ ()] [ ()]}
()
(17)
In equa ion (17),
lu
and
l
s
ep esen , espec i ely, he
sha es o unskilled and skilled labo ( hei sum is equal o
1);
ϱ
()
is defined as be o e (pe capi a obo ic capi al) and
≡
ς
R
L
AI
() ()(pe capi a a ificial in elligence).
In he cu en case, wages and e u ns om diffe en
o ms o capi al a e, espec i ely,
== −
+++ +
−
w αy
l l lς
1
ϱϱ ,
uRuu
βsβ1
() () ()()
() [ ()] [ ()]
(18)
== −
+++ +
−
w αy
lς l lς
1ϱ,
ssu
βsβ
RAI 1
() () ()()
() [ ()][ ()]
(19)
= αy
k
.
() ()
()
(20)
The skill p emium is s aigh o wa d o display, gi en
exp essions (18) and (19),
=+++ +
+++ +
−
−
w
w l l lς
lς l lς
ϱϱ
ϱ
.
s
u
uu
βsβ
su
βsβ
1
1
()
() () [ ()] [ ()]
() [ ()][ ()] (21)
The compu a ion o de i a i e ∂⎛
⎝⎞
⎠
∂ς
ws
wu
()
()
() allows us o
e i y wha is he impac o an inc ease in a ificial in elli-
gence echnologies o e he diffe ence in income be ween
non-skilled and high-skilled wo ke s. Fo he p oposed
echnology o p oduc ion, he de i a i e has a nega i e
sign unde condi ion
+>+
l
ς l ϱ
su
() (
)
, i.e., whene e he
numbe o skilled asks su passes he olume o unskilled
asks de eloped in he economy, he inc ease in he use o
a ificial in elligence educes income inequali y (lowe s he
skill p emium).
O he meaning ul esul s a e a ached o he weigh
o wages and capi al e u ns on income. Equa ions (18)
and (19) di ec ly indica e ha he highe he alues o
ϱ( ) and ς( ), he lowe will be he wage income sha e;
6O lando Gomes
i.e., addi ional au oma ion, ia obo ic capi al and a ificial
in elligence, con ibu es o a decline o he ela i e income
o wo ke s. Consequen ly, as he alues o ϱ( ) and ς( ) ise,
eflec ing he s onge ela i e p esence o au oma ed p o-
cesses in p oduc ion, he mo e income will be concen a ed
in he hands o capi al owne s, which poses a eal and
conc e e dange o social and poli ical s abili y.
To associa e all he abo e easoning o a g ow h model,
one would need o conside a s anda d physical capi al
accumula ion equa ion and an in e empo al elici y unc-
ion. Then, i would be necessa y o add one (o mo e)
obo ic capi al sec o (s) o he analysis: he sel - eplica ing
ea u es o obo s and, essen ially, o a ificial in elligence
make i easonable o conside ha no o he inpu is
equi ed o i s gene a ion and ha , p obably, in he cu -
en s age o de elopmen , his inpu would escape he
p e alence o diminishing ma ginal e u ns. As a esul , in
his amewo k, obo ic capi al and Cha GPT-like echnologies
become he d i e s o endogenous g ow h. Howe e , his is
adiffe en ype o g ow h; i is a g ow h p ocess ha la gely
amplifies inequali ies and ha changes he s uc u e o
demand in he economy.
Hence, he analysis o g ow h in he au oma ed economy
clea ly equi es a modeling amewo k wi h he e ogeneous
agen s: by sepa a ing wo ke s om capi al owne s, one will
be able o disce n how he ongoing uncons ained e olu ion
o echnology ep esen s a isk, no only o hose who
di ec ly suffe wi h he loss o jobs, bu o all people ha
may end up li ing in a dys opian wo ld popula ed by an
e e -inc easing a my o excluded.
4 Lessons om he Pandemic
The a aging global pandemic o he ea ly 2020s aised
disquie ing in e oga ions abou he eali y ha we had
aken o g an ed conce ning wo ld p ospe i y and g ow h.
I e ealed how a low- isk, huge-impac e en may suddenly
affec he li es o e e yone on his plane . I also showed ha
accoun ing o g ow h is no jus an exe cise o measu ing he
quan i y and quali y o inpu s and he efficiency in hei use;
he e a e ele an socie al issues, in his case abou public
heal h, ha mus be accoun ed o . As i is e iden , he COVID
pandemic led o a e hinking o g ow h heo y in he p e-
sence o heal h eme gencies and disas e s. Meaning ul ecen
wo k on he mac oeconomic consequences o he sp ead o
in ec ious diseases comp ehends he wo ks o Ca mona and
León (2023), Fogli and Veldkamp (2021), Hao e al. (2023), Lu
(2023), and Shi (2023).
The mos common s a egy in assessing he g ow h
implica ions o he p opaga ion o in ec ious diseases,
ollowed by mos o he abo e-men ioned li e a u e, con-
sis s o me ging benchma k op imal g ow h models wi h
s anda d epidemiological analy ical amewo ks o he
SIR (suscep ible–in ec ious– eco e ed) ype. As indi iduals
pass om each epidemiological s a e o he nex , he
economy also e ol es om one g ow h s age o ano he .
E iden ly, pe iods in which a significan pe cen age o
he popula ion is in an in ec ious s a e a e pe iods o
slowe g ow h. The channels om disease o g ow h a e
essen ially h ee: labo p oduc i i y, human capi al accu-
mula ion, and popula ion g ow h. Combined, he a ious
nega i e effec s migh ha e de as a ing consequences o
he wo ld economy and he li ing s anda ds o people
a ound he wo ld.
Some o he wo k on he impac o in ec ious diseases
on g ow h, mos no iceably ha o Fogli and Veldkamp
(2021), es ablish a link be ween he sp ead o diseases
and he diffusion o ideas and echnology. The a gumen
is ha in e ac ion among people diffuses bo h ideas and
diseases. The e o e, gi en hei heal h condi ions and sys-
ems, coun ies mus choose an adequa e balance be ween
knowledge diffusion and he isk o he ansmission o
infi mi ies. Knowledge and in ec ious pa hogens ha e one
cha ac e is ic in common: hey a e bo h non i al; howe e ,
hey ha e an an agonis ic na u e in he sense ha he fi s is a
global good, while he second is a global bad. The assessmen
o ex e nali ies mus be ponde ed: he posi i e ex e nali ies
o igina ing in knowledge diffusion mus be weighed agains
he nega i e ex e nali ies ha he diffusion o i us and
ge ms b ings.
As an illus a ion o he g ow h implica ions o disease
p opaga ion, conside he ollowing s aigh o wa d ea-
soning. Imagine a s anda d g ow h model, wi h physical
and human capi al as p oduc ion inpu s. In his se up, he
d i e o g ow h is human capi al accumula ion; hus, le
us concen a e on he mo ion o his inpu , ep esen ed in
ime
( )
0.5
Figu e 1: Double-logis ic epidemic diffusion.
Economic G ow h in he Age o Ubiqui ous Th ea s 7
wha ollows by a iable
H
(). Human capi al is subjec o
obsolescence a a e ∈δ0,1
,
h()
and i s p oduc ion is subjec
o cons an e u ns; howe e , he e is a p oduc i i y loss in
he educa ion sec o di ec ly a ibu able o illness. Le
a iable
ϕ
(
)
ep esen he p e alence o an in ec ious dis-
ease, and le
x
(
)
be he p oduc i i y loss di ec ly a ibu-
able o he disease.
Wi h he abo e in o ma ion, one can display he g ow h
a e o human capi al (which will also be he g ow h a e o
he economy unde a i ial wo-sec o op imal g ow h se up)
in he ollowing e ms:
=− − >
H
H Bϕ x δB
1,0
.
h
()
() [()()] (22)
Assume ha
x
(
)
is ime-in a ian and ha he p e a-
lence a e e ol es, as in he s udy by Hao e al. (2023),
ollowing a double-logis ic ule, i.e.,
=⎛
⎝+−+⎞
⎠
−−
ϕ
ϕ ee
1
11
1
.
a a
012
()
(23)
All pa ame e s in equa ion (23) a e posi i e alues.
Figu e 1 illus a es he e olu ion o he in ec ion a e o
== =
ϕ
aa2.5, 1, and 0.4
012
. A e a fi s phase o as
inc ease in he sha e o in ec ed in he popula ion, his
alue g adually alls o ze o.
In his simple amewo k, gi en equa ion (22), in he
absence o he disease, he economy g ows a a cons an
a e. The effec o he epidemic is o p o oke a ansien all
in he g ow h a e. Figu e 2 illus a es his effec o he
sp eading mechanism displayed in Figu e 1 and cha ac e -
ized h ough equa ion (23).
The abo e easoning di ec ly applies o he dissemina-
ion o a disease bu , in ac , i is adap able o many o he
socie al h ea s. Any e en leading o social dis us o he
b eakdown o social ies (e.g., he g owing you h disillu-
sionmen men ioned in he in oduc ion) may cause a
nega i e impac on he accumula ion o human capi al.
In he ske ched amewo k, he impac is ansi o y, in
he sense ha i is expec ed ha he heal h issue will be
esol ed soone o la e . Some socie al p oblems migh be
mo e p o ound and e en ually igge a g ow h slowdown
o a mo e pe manen na u e.
5 The G ea es o Them All: The
En i onmen al Ex e nali y
As ema ked in he la es edi ions o he Global Risks
Repo , en i onmen al h ea s (clima e change, ex eme
wea he episodes, biodi e si y losses, deple ion o na u al
esou ces, and man-made disas e s) occupy he fi s place
in he anking o global isks, in e ms o bo h likelihood
and expec ed damaging impac . Due o hei ca as ophic
na u e, en i onmen al isks a e ha d o econcile wi h
economic heo y and, in pa icula , wi h g ow h heo y,
which p i ileges “business as usual.”None heless, he e
is a oluminous new li e a u e sea ching o a cohe en
in eg a ion be ween he wo. I is sa e o say ha en i on-
men al conce ns ha e become an inc easingly ele an
pa o he heo y o economic g ow h.
Con ibu ions a e dispe sed and app oach di e se
aspec s o he en i onmen al menace. One o he mos
p ominen opics conce ns he impac o pollu ion o ,
mo e p ecisely, ca bon emissions (Oli ei a & Lima, 2022;
Olijslage s e al., 2023; an den B eme & an de Ploeg,
2021). Measu ing he social cos o ca bon is a complica ed
ask, gi en he inhe en long- e m unce ain y ha makes
i un easible o compu e undispu able discoun a es o
quan i y he cu en alue o u u e damages. In g ow h
models, he en i onmen is equen ly added o he ana-
lysis h ough he explo a ion o he pollu ion-g ow h
ade-off: pollu ion is a by-p oduc o p oduc ion, while
en i onmen al quali y is an a gumen o he u ili y unc-
ion. The solu ion o he unde lying conund um consis s
o p omo ing he ansi ion o clean p oduc ion echnol-
ogies (Ha , 2020).
Casey (2024) and Hassle e al. (2021) de elop g ow h
models in which echnical change endogenously e ol es o
inc ease ene gy efficiency and o adap o en i onmen al
changes. Ene gy dependence will hen de e mine he s uc-
u e o p oduc ion and he pace o g ow h. Fabozzi e al.
(2022) look a he economy om he pe spec i e o g een
g ow h. G een g ow h is associa ed wi h he no ion o pu -
ing science and echnology a he se ice o en i on-
men al p ese a ion, a he same ime hey acili a e
ime
()
( )
Figu e 2: T ansi ional pa h implied by he sp ead o an in ec ious
disease.
8O lando Gomes
Sasaki, H. (2023). G ow h wi h au oma ion capi al and declining popu-
la ion. Economics Le e s,222(Janua y), 110958.
Shi, S. (2023). Knowledge, ge ms, and ou pu . Re iew o Economic
Dynamics,48(Ap il), 297–319.
S ike , H., & Suen, R. M. (2022). Sou ces o economic g ow h in models wi h
non- enewable esou ces. Jou nal o Mac oeconomics,72(June), 103416.
Thies, C. F., & Baum, C. F. (2020). The effec o wa on economic g ow h.
Ca o Jou nal,40(1), 199–212.
Tohmé, F., Ca aballo, M. A., & Dabús, C. (2022). Ins abili y, poli ical
egimes and economic g ow h. A heo e ical amewo k.
Me oeconomica,73(1), 291–317.
an den B eme , T. S., & an de Ploeg, F. (2021). The isk-adjus ed ca bon
p ice. Ame ican Economic Re iew,111(9), 2782–2810.
Va a igos, D. (2023). Cul u al pe sis ence in co up ion, economic
g ow h, and he en i onmen . Jou nal o Economic Dynamics and
Con ol,147(Feb ua y), 104590.
Wei, S., & Aadland, D. (2022). Physical capi al, human capi al, and he
heal h effec s o pollu ion in an OLG model. Mac oeconomic
Dynamics,26(6), 1522–1563.
Wo ld Economic Fo um. (2023). Global Risks Repo , 2023 edi ion. Gene a,
Swi ze land.
Xa ie , I. (2023). Bubbles and s agna ion. Jou nal o he Eu opean Economic
Associa ion,21(6), 2460–2484.
Zhou, S., & Smulde s, S. (2021). Closing he loop in a ci cula economy:
Sa ing esou ces o suffoca ing inno a ions? Eu opean Economic
Re iew,139(Oc obe ), 103857.
Economic G ow h in he Age o Ubiqui ous Th ea s 15
