Baek, Yaein; Lee, Jiyun
A icle
Analysis o A i icial In elligence Exposu e Ac oss
Indus ies in Sou h Ko ea and he Uni ed S a es
Eas Asian Economic Re iew (EAER)
P o ided in Coope a ion wi h:
Ko ea Ins i u e o In e na ional Economic Policy (KIEP), Sejong-si
Sugges ed Ci a ion: Baek, Yaein; Lee, Jiyun (2025) : Analysis o A i icial In elligence Exposu e
Ac oss Indus ies in Sou h Ko ea and he Uni ed S a es, Eas Asian Economic Re iew (EAER), ISSN
2508-1667, Ko ea Ins i u e o In e na ional Economic Policy (KIEP), Sejong-si, Vol. 29, Iss. 1, pp.
3-40,
h ps://doi.o g/10.11644/KIEP.EAER.2025.29.1.443
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/316639
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/
PISSN 2508-1640 EISSN 2508-1667 an open access jou nal
Eas Asian Economic Re iew ol. 29, no. 1 (Ma ch 2025) 3-40
h ps://dx.doi.o g/10.11644/KIEP.EAER.2025.29.1.443
ⓒ Ko ea Ins i u e o In e na ional Economic Policy
Analysis o A i icial In elligence Exposu e Ac oss Indus ies
in Sou h Ko ea and he Uni ed S a es*
Yaein Baek†
Sogang Uni e si y
[email p o ec ed]
Jiyun Lee††
Ko ea Ins i u e o In e na ional Economic Policy
[email p o ec ed]
1
This s udy examines he impac o AI exposu e on indus ies in Sou h Ko ea and he
Uni ed S a es om 2019 o 2022, using he AI Indus y Exposu e (AIIE) index
de eloped by Fel en e al. (2021). In Sou h Ko ea, AI exposu e is posi i ely associa ed
wi h employmen bu nega i ely associa ed wi h eal labo income pe capi a, wi h
labo p oduc i i y po en ially d i ing he employmen gains. Addi ionally, an analysis
o occupa ional employmen in Sou h Ko ea con i ms a posi i e co ela ion wi h AI
exposu e. Fo he U.S., AI exposu e shows mo e p onounced labo ma ke in luence
han in Sou h Ko ea.
Keywo ds: A i icial In elligence, Indus ies, Labo Ma ke
JEL Classi ica ion: O33, J24, J23
I. In oduc ion
Cha GPT gained signi ican global a en ion a e a ac ing o e 1 million use s
wi hin 5 days o i s ini ial elease in No embe 2022. The global a i icial in elligence
ma ke is expec ed o g ow om $515.31 billion in 2023 o $621.19 billion in 2024,
* This s udy is an ex ension o Chap e 5 o Yoon e al. (2024).
† Co esponding au ho , Depa men o Economics, Sogang Uni e si y, 35 Baekbeom- o, Mapo-gu,
Seoul 04107, Republic o Ko ea. This wo k was suppo ed by he Sogang Uni e si y Resea ch G an
o 2024 (Numbe : 202410012.01).
†† In e na ional Mac oeconomics & Finance Depa men , Ko ea Ins i u e o In e na ional Economic
Policy.
ID
ID
4 Yaein Baek and Jiyun Lee
ⓒ Ko ea Ins i u e o In e na ional Economic Policy
eaching $2.74 illion in 2032, wi h a p ojec ed compound annual g ow h a e o 20.4%
o e he o ecas pe iod (Kim, 2024). While he impac o gene a i e AI on he global
economy can be examined om a ious pe spec i es, i s implica ions o he labo
ma ke s and ou .
Recen s udies ha e inc easingly ocused on he e ec s o AI on labo ma ke s,
indus ial s uc u es, and economic g ow h. Acco ding o esea ch by Cazzaniga e al.
(2024), 40% o global employmen is exposed o AI. AI could lead o a es uc u ing o
labo ma ke s by eplacing high-skilled and high-wage wo ke s, he eby enhancing
p oduc i i y, c ea ing new jobs, and d i ing economic g ow h. Ha zius e al. (2023)
sugges s ha while mos jobs and indus ies a e pa ially exposed o AI, hey a e mo e
likely o be augmen ed a he han eplaced. His o ically, jobs displaced by au oma ion
ha e been o se by he c ea ion o new oles, wi h he eme gence o new occupa ions
due o echnological inno a ion accoun ing o mos long- e m employmen g ow h.
They p edic ha widesp ead adop ion o gene a i e AI could inc ease he annual U.S.
labo p oduc i i y g ow h a e by 1.5 pe cen age poin s o e he nex decade and
ul ima ely boos global GDP by 7% annually. This e lec s bo h conce ns abou labo
displacemen and op imism o u u e economic g ow h d i en by AI.
The de elopmen and adop ion o gene a i e AI is p og essing apidly, leading o
signi ican economic impac , bu he e is insu icien da a o p o e i s e ec s. Recen
esea ch has u ilized case s udies and ex mining o indi ec ly explo e he impac o
gene a i e AI. Chui e al. (2023) inds ha gene a i e AI is expec ed o gene a e be ween
$2.6 illion and $4.4 illion in annual economic alue. Gene a i e AI will impac all
indus ies, wi h banking, high- ech, and li e sciences expec ed o be mos impac ed as a
pe cen age o e enue. Addi ionally, hey sugges ed ha while gene a i e AI can
signi ican ly imp o e labo p oduc i i y ac oss he economy, in es men s in wo ke
eskilling o edeploymen will be necessa y.
Howe e , Sou h Ko ea’s in es men in and u iliza ion o AI is ela i ely low
compa ed o o he coun ies, aising conce ns abou i s u u e economic g ow h.
Acco ding o he 2024 AI Index Repo (Maslej e al., 2024), p i a e in es men in AI in
he Uni ed S a es in 2023 was $67.2 billion, app oxima ely 48.4 imes he in es men
amoun o Ko ea, which anked nin h ($1.4 billion). Since 2013, he U.S. has consis en ly
anked i s in p i a e AI in es men , and he gap be ween he U.S. and o he coun ies
has widened o e ime. This dispa i y is pa icula ly e iden in in es men s ela ed o
gene a i e AI, such as Cha GPT. In 2022, he U.S. exceeded he combined gene a i e
AI in es men o he EU and he UK by app oxima ely $1.9 billion, and his igu e
Analysis o A i icial In elligence Exposu e Ac oss Indus ies in Sou h Ko ea and he Uni ed S a es 5
ⓒ 2025 Eas Asian Economic Re iew
expanded o $21.1 billion in 2023. Acco ding o PwC (2024), Sou h Ko ea’s indus ies
ha e been slow o adop and u ilize AI. This is a ibu ed o a sho age o skilled
pe sonnel, sha ed da a, and pla o m in as uc u e, as well as insu icien policy suppo
and in es men o AI de elopmen . The gap is pa icula ly e iden in sec o s such as
inance, logis ics, media, manu ac u ing, and law, compa ed o global s anda ds.
The launch o gene a i e AI and i s apid adop ion by businesses ac oss a ious
indus ies sugges ha he echnology’s impac a ies by indus y. While p io esea ch
has p ima ily ocused on he e ec s o AI a he occupa ional o i m le el, his s udy
examines key indica o s o indus y-le el exposu e o AI in Sou h Ko ea and compa es
hese indings wi h U.S. indus ies leading in p i a e AI in es men . Using he AI
Indus ial Exposu e (AIIE) index de eloped by Fel en e al. (2021), we examine he
e ec o AI on p oduc i i y, employmen , and wages in indus ies in Sou h Ko ea and
he U.S. since 2019. We es ima e he ela ionship be ween he AIIE and indus y
indica o s using OLS wi h wo-way ixed e ec s.
Ou indings indica e ha Sou h Ko ean indus ies wi h g ea e exposu e o AI
expe ienced inc eases in employmen and sales, bu a decline in eal labo income pe
capi a. The esul s u he sugges ha he posi i e associa ion be ween AI exposu e and
employmen is d i en by inc eases in labo p oduc i i y. In he U.S., indus ies wi h
highe AI exposu e expe ienced no only employmen g ow h bu also inc eases in bo h
hou ly compensa ion and o al labo compensa ion. Addi ionally, we in es iga e he
ela ionship be ween AI exposu e and employmen in Sou h Ko ea a he occupa ional
le el, inding ha highe occupa ional exposu e o AI is linked o inc eases in sho - e m
employmen sha es.
The es o he pape p oceeds as ollows. Sec ion 2 e iews he li e a u e ela ed o
ou s udy. Sec ion 3 discusses he da a and he cons uc ion o he AIIE index. Sec ion 4
p esen s empi ical analysis esul s on he ela ionship be ween indus y-speci ic
indica o s and he AIIE index o Sou h Ko ea and he Uni ed S a es, and Sec ion 5
concludes.
II. Li e a u e Re iew
S udies on he adop ion o AI and obo ics a e g ounded in he li e a u e on inno a ion
and echnological de elopmen . F ey and Osbo ne (2017) p edic ed ha compu e iza ion
would impac jobs in ol ing non- ou ine asks, es ima ing ha 47% o U.S. jobs we e a
6 Yaein Baek and Jiyun Lee
ⓒ Ko ea Ins i u e o In e na ional Economic Policy
high isk o compu e iza ion. S udies such as B ynjol sson e al. (2018) and Fel en e al.
(2018) ha e examined he impac o AI and au oma ion on he U.S. economy.
B ynjol sson e al. (2018) measu ed he sui abili y o asks o machine lea ning and
a gued ha he eshaping o jobs due o machine lea ning would a ec he labo ma ke
di e en ly om pas au oma ion d i en by obo s.
The e ha e also been s udies ha iden i y “winne s” and “lose s” due o AI adop ion
and analyze he dis ibu ional e ec s o new echnologies. Acemoglu e al. (2022)
demons a ed ha he e was a signi ican inc ease in AI- ela ed employmen be ween
2010 and 2018, p ima ily a ibu able o i ms wi h a high deg ee o AI exposu e. These
i ms dec eased hi ing o non-AI ela ed oles as hey adop ed AI. Howe e , he o e all
impac o labo displacemen due o AI on employmen and wage g ow h in occupa ions
and indus ies was ound o be minimal a p esen . Au o (2022) e iews he e olu ion
o economic hinking on he ela ionship be ween digi al echnology and inequali y
ac oss ou decades, and p esen s ou pa adigms (educa ional ace, ask pola iza ion,
au oma ion- eins a emen ace, and he e a o AI unce ain y). While echnological
change c ea es winne s and lose s, he au ho concludes ha complemen a y ins i u ional
in es men s a e necessa y o gene a e sha ed bene i s.
Webb (2019) and Fel en e al. (2021) cons uc s a measu e o an occupa ion’s
exposu e o AI. Webb (2019) shows ha high-skilled jobs a e ela i ely mo e exposed
o AI echnology, compa ed o obo ics and so wa e. In a simila s udy, Tolan e al.
(2021) linked a ious asks and cogni i e abili ies o a lis o AI benchma ks used o
e alua e p og ess in AI echniques. An applica ion o occupa ional da abases p o ides
esul s ha some jobs ha we e p e iously hough o be immune o au oma ion could
be mo e exposed o AI. Han and Oh (2023) employed he Webb (2019) occupa ional AI
exposu e index, adjus ed o he Ko ean S anda d Classi ica ion o Occupa ions, o
iden i y Ko ean occupa ions suscep ible o AI and examine he consequen e ec s on
employmen and wages. Thei indings indica e ha occupa ions wi h highe AI
exposu e we e mo e p one o job losses and slowe wage g ow h. Geo gie and Hyee
(2022) adap he measu e de eloped by Fel en e al. (2021) o analyze he link be ween
AI and employmen in a c oss-coun y con ex . Za i hona a (2024) uses ex mining
on he In e na ional S anda d Classi ica ion o Occupa ions (ISCO) da abase and inds
ha 32.8% o all occupa ions will be signi ican ly impac ed by gene a i e AI, and 36.5%
will be pa ially a ec ed. Bon iglioli e al. (2023) classi y AI- ela ed occupa ions as
hose whose job pos ings mos equen ly equi e specialized so wa e used o machine
lea ning and da a analysis. The au ho s use a shi -sha e ins umen ha combines
Analysis o A i icial In elligence Exposu e Ac oss Indus ies in Sou h Ko ea and he Uni ed S a es 7
ⓒ 2025 Eas Asian Economic Re iew
indus y-le el AI adop ion wi h local indus y employmen , and ind nega i e e ec s o
AI exposu e on employmen ac oss commu ing zones and ime.
Song e al. (2021) empi ically analyzed he ela ionship be ween AI adop ion and
p oduc i i y in domes ic i ms. While hei s udy ound no s a is ically signi ican
co ela ion be ween AI adop ion and p oduc i i y in he domes ic manu ac u ing sec o
o e all, hey disco e ed a posi i e impac o AI on p oduc i i y when i ms we e
ca ego ized by hei owne ship s uc u e, speci ically hose wi h mul iple business uni s.
Cza ni zki e al. (2023) conduc ed an empi ical analysis o he impac o AI adop ion on
p oduc i i y in Ge man i ms, using e enue as a p oxy o p oduc i i y. Thei indings
sugges ha he e is a s a is ically signi ican posi i e ela ionship be ween AI adop ion
and p oduc i i y.
III. Da a
We cons uc ed he AI Indus ial Exposu e (AIIE) o he Uni ed S a es and Sou h
Ko ea by calcula ing a weigh ed a e age o he AI Occupa ional Exposu e (AIOE) o
each occupa ion wi hin an indus y, using he employmen sha e o each occupa ion as
he weigh . The AIOE alues we e ob ained om Fel en e al. (2021). Fo Sou h Ko ea,
he AIOE alues we e mapped o he Ko ean S anda d Classi ica ion o Occupa ions
(KSCO, 7 h edi ion) h ough he In e na ional S anda d Classi ica ion o Occupa ions
(ISCO-08). Simila ly, Han and Oh (2023) calcula ed Sou h Ko ean occupa ional and
indus ial AI exposu e index based on he measu e by Webb (2019). Fel en e al. (2021)
cons uc ed he AIOE using da a om he 2019 O*NET da abase. They selec ed en
ca ego ies om he Elec onic F on ie Founda ion (EFF)’s AI applica ion de ini ions
and linked hese ca ego ies o wo kplace abili ies. By de eloping a da ase based on su ey
esponses collec ed om Amazon Mechanical Tu k (mTu k) gig wo ke s, hey
es ablished he connec ions be ween AI applica ions and wo kplace abili ies, enabling
he calcula ion o abili y-le el exposu e. Fo de ails on he calcula ion p ocess, see Fel en
e al. (2021). They de i ed he AIOE by weigh ing he abili y-le el exposu e sco es and
he p e alence and impo ance indica o s o each abili y. The o al occupa ional exposu e
o AI was calcula ed by summing all he weigh ed abili y-le el AI exposu es. They hen
cons uc ed an AIIE by calcula ing a weigh ed a e age o he AIOE sco es, whe e
weigh s a e de e mined by indus y employmen igu es based on he ou -digi NAICS
classi ica ion sys em.
8 Yaein Baek and Jiyun Lee
ⓒ Ko ea Ins i u e o In e na ional Economic Policy
In his s udy, we cons uc ed an upda ed AIIE o he U.S. using employmen da a
om 2020 o 2022. Simila ly, we cons uc ed he AIIE o Sou h Ko ea using
employmen da a om 2019 o 2022, wi h de ails p o ided in he Appendix A.3.
S uc u ally, a high AIIE o a gi en indus y indica es ha a signi ican p opo ion o
wo ke s in ha indus y a e employed in occupa ions wi h high AI exposu e. Since
indus ies ha ac i ely u ilize AI echnologies may no necessa ily align wi h indus ies
ha employ a la ge numbe o occupa ions wi h high AI exposu e, we c oss- alida ed
whe he he AIIE ep esen s AI exposu e in Ko ean indus ies using da a om he
Minis y o Science and ICT’s “Su ey on Co po a e In o ma ioniza ion S a is ics.”1 By
calcula ing he p opo ion o companies using AI echnologies and se ices in he 1-digi
indus y classi ica ions in Sou h Ko ea and conduc ing a sign es wi h AIIE, we could
no conclude ha he AIIE ankings de i ed om each index we e di e en .2
Fo he U.S., we calcula ed AIIE a he NAICS 2-digi (18 ca ego ies), NAICS 3-digi
(59 ca ego ies), and NAICS 4-digi (195 ca ego ies) indus y le els, ollowing he No h
Ame ican Indus y Classi ica ion Sys em (NAICS). Fo he eg ession analysis, we use
U.S. indus y da a a he NAICS 3-digi le el (38 ca ego ies) and NAICS 4-digi le el (96
ca ego ies) o he yea 2022. Fo Sou h Ko ea, we calcula ed AIIE a he KSIC 1-digi
(21 ca ego ies), KSIC 2-digi (75 ca ego ies), and KSIC 3-digi (213 ca ego ies) indus y
le els, using he Ko ean S anda d Indus ial Classi ica ion (KSIC). To compa e AIIE in
bo h coun ies, we used he In e na ional S anda d Indus ial Classi ica ion o All
Economic Ac i i ies (ISIC Re .4). Fo u he de ails on he da a and he anking o AI
exposu e o indus ies, see Appendix A.3.
IV. Empi ical Analysis
This sec ion p esen s he eg ession speci ica ion and esul s. In sec ion 4.1, we
p o ide he esul s on he ela ionship be ween AI exposu e and indus y-le el measu es.
1 The Su ey on Co po a e In o ma ioniza ion S a is ics in es iga ed whe he p i a e sec o companies
u ilize a i icial in elligence (AI) echnologies and se ices. Fo he 19 1-digi indus y classi ica ions in
Sou h Ko ea (KSIC 10 h e ision), we calcula ed he p opo ion o companies using AI by using he
numbe o companies ha use o do no use AI.
2 The sign es , a s a is ical me hodology o compa ing wo popula ions o o dinal da a, was employed.
When compa ing he AIIE-de i ed ankings wi h hose based on he p opo ion o companies using
AI echnologies, he p- alue o he null hypo hesis ha he wo ankings a e iden ical was 0.607,
ailing o ejec he null hypo hesis.
Analysis o A i icial In elligence Exposu e Ac oss Indus ies in Sou h Ko ea and he Uni ed S a es 9
ⓒ 2025 Eas Asian Economic Re iew
All indus y- le el eg essions a e pe o med sepa a ely o Sou h Ko ea and he Uni ed
S a es. In sec ion 4.2, we u he analyze he impac o AI on he labo ma ke in Sou h
Ko ea wi h esul s on he occupa ion-le el AI exposu e and employmen .
1. Main Reg ession Resul s
Using he da a om 2019 o 2023 desc ibed in sec ion 3 and Appendix 5 we es ima e
he ollowing eg ession model:
yi, = αi + δ + βAIIEi, −1 + γRobo i, −1 + εi, (1)
whe e i is he indus y and is yea . The dependen a iable yi, is he ou come o in e es ,
which a e he indus y-le el indica o s in Tables 1 and 2 o Sou h Ko ea, and Table 6
o he U.S. He e αi a e indus y ixed e ec s and δ a e ime ixed e ec s. We include
he lagged AIIE index AIIEi, −1 in o de o mi iga e he endogenei y o con empo aneous
a iables. I is also likely ha AI impac s an indus y wi h a lag. The es ima es o β a e
p esen ed in Tables 3 o 4 o Sou h Ko ea. Fo he U.S., indica o s o NAICS 4-digi
indus ies a e used and he indus y ixed e ec s in (1) a e based on 3-digi indus ies:
yi, = αs + δ + βAIIEi, −1 + γRobo i + εi, (2)
whe e s indica es he NAICS 3-digi indus y and i is he 4-digi indus y. The es ima es
a e p o ided in Table 7.
In addi ion, we conside he usage o obo ics echnologies in indus ies. Bo h AI and
obo ics a e capable o au oma ion bu he e ec s may di e ac oss he wo echnologies.
I is likely ha echnologies ha inco po a e AI will be able o au oma e a mo e asks
han pu ely obo -based echnologies (Raj and Seamans, 2019). We include he lagged
pe cen age o i ms wi hin an indus y ha use obo ics, Robo i, −1 as a con ol in (1).3
Fo he U.S. eg ession, we use he pe cen age o i ms wi hin an indus y ha use
3 The a e o i ms ha use obo ics wi hin an indus y is a ailable o KSIC 1-digi indus ies. The
es ima ion esul s in Table 4 uses KSIC 2-digi indus y measu es, hence he no a ion Robo s, −1 is
app op ia e.
10 Yaein Baek and Jiyun Lee
ⓒ Ko ea Ins i u e o In e na ional Economic Policy
obo ics in yea 2020, Robo i as a con ol a iable in (2).4
We i s examine he esul s o indus ies in Sou h Ko ea. Table 3 p esen s es ima es
o KSIC 1-digi indus ies’ alue added c ea ion & dis ibu ion indica o s, and g ow h
& p o i abili y indica o s. Based on he esul s in column (1), indica o s ha show
s a is ically signi ican posi i e ela ionship wi h he AIIE a e alue added g ow h pe
capi a, sales g ow h pe capi a, and sales g ow h a e. Indica o s ha ha e a s a is ically
signi ican nega i e ela ionship wi h he AIIE a e sales pe capi a, capi al in ensi y, labo
sha e, eal labo income pe capi a, o al capi al g ow h a e, equi y capi al g ow h a e,
and ne p o i as a sha e o o al capi al. Indus ies ha a e hea ily exposed o AI
echnology expe ience a educ ion in eal labo income pe capi a. This esul is simila
o Han and Oh (2023) which inds ha in Sou h Ko ea, occupa ions exposed o AI
expe ience a decline in wages. Addi ionally, AI-exposed indus ies ace a educ ion in
labo sha e, which is he p opo ion o labo income ela i e o he o al o ope a ing
income and labo income. This sugges s ha wo ke s in AI-exposed indus ies a e being
nega i ely impac ed, wi h a decline in labo sha es and eal labo income.
The esul s in column (2), which inco po a e obo ic usage, a e e y simila o he
es ima es in column (1). The pe cen age o i ms ha use obo ics echnologies and he
AIIE index has co ela ion o 0.32. The indus ies wi h high AI exposu e a e inancial
se ice, insu ance, and in o ma ion whe eas indus ies wi h high usage o obo ics a e
manu ac u ing.
Table 4 p esen s es ima ion esul s o KSIC 2-digi indus ies. F om column (1),
indica o s ha show s a is ically signi ican posi i e ela ionship wi h he AIIE a e he
numbe o wo ke s, emale wo ke s, egula wo ke s, and sales. Highe AI exposu e is
associa ed wi h an inc ease in employmen in he indus y, and in pa icula , emale
wo ke s and egula wo ke s. These esul s di e om he occupa ion-le el employmen
analysis in Han and Oh (2023), which shows ha occupa ions wi h highe AI exposu e
expe ienced a decline in employmen sha es. Taking hei indings and ou s oge he , we
can specula e ha he ela ionship be ween AI exposu e and occupa ional employmen
a ies by indus y. Fo example, he ole o a bookkeepe is highly suscep ible o AI,
po en ially leading o a decline in o e all employmen ac oss indus ies. Howe e , as AI-
in ensi e indus ies expand, he demand o bookkeeping wi hin hese indus ies may
4 The Na ional Cen e o Science and Enginee ing S a is ics (NSF) da a in he Annual Business Su ey
a ies sligh ly om yea o yea , making i di icul o cons uc he a iable by yea .
Analysis o A i icial In elligence Exposu e Ac oss Indus ies in Sou h Ko ea and he Uni ed S a es 17
ⓒ 2025 Eas Asian Economic Re iew
Acco ding o Acemoglu e al. (2022), human-complemen a y AI can inc ease labo
demand by enhancing human p oduc i i y in asks whe e AI is adop ed. I labo and AI
algo i hms a e highly subs i u able in a i m’s p oduc ion p ocess, imp o emen s in AI
ask pe o mance ini ially lead o labo subs i u ion, pa icula ly a ec ing he hi ing o
AI-specialized wo ke s. Howe e , i ms ha lowe p oduc ion cos s h ough AI adop ion
can maximize p o i s by educing p oduc p ices, which subsequen ly inc eases demand.
This p ocess c ea es a labo demand expansion e ec (p oduc ion e ec ). I he
p oduc i i y gains a e subs an ial, i m-le el employmen will ise alongside sales
g ow h (Han, 2023).
Ou inding ha AI exposu e is posi i ely co ela ed wi h labo compensa ion can be
explained by he s udy o Cazzaniga e al. (2024). I AI signi ican ly enhances human
labo in speci ic occupa ions and gene a es subs an ial p oduc i i y gains, he esul ing
economic g ow h and inc eased labo demand could ou weigh he pa ial eplacemen o
labo asks, leading o labo income g ow h ac oss much o he income dis ibu ion.
Cazzaniga e al. (2024) highligh s h ee channels h ough which AI a ec s he economy:
labo displacemen , complemen a i y, and p oduc i i y gains. When AI s ongly
complemen s labo , he posi i e complemen a i y e ec ou weighs he displacemen
e ec , esul ing in a smalle p opo ion o high-income wo ke s being ad e sely a ec ed
compa ed o he lowe -complemen a i y case. Addi ionally, when AI-d i en
p oduc i i y gains a e accoun ed o , labo income inc eases o all wo ke s, as highe
p oduc i i y boos s demand o all ac o s o p oduc ion in he economy.
In addi ion, we ake he i s di e ences o he U.S. indus y indica o s and es ima e
eg ession (2). Table A11 and Table A12 in he Appendix p esen he es ima es based
on he indica o s ha a e con e ed o yea -o e -yea pe cen age change. G ow h a es
wi h a s a is ically signi ican nega i e ela ionship wi h he AIIE a e labo p oduc i i y,
capi al p oduc i i y, capi al cos s, capi al sha e, combined inpu cs s, and combined
inpu s p ice de la o . G ow h a es wi h a s a is ically signi ican posi i e ela ionship
wi h he AIIE a e employmen , labo compensa ion, hou s wo ked, labo sha e, and
capi al inpu . Hence, highe AI exposu e is associa ed wi h an inc ease in indus y
employmen and labo compensa ion, in le els and g ow h a e. This sugges s ha U.S.
indus ies a e mo e in luenced by AI echnology compa ed o indus ies in Sou h Ko ea.7
7 Hou ly compensa ion and labo compensa ion a e nominal a iables. Fo Sou h Ko ea indus y indica o s
in Table 4, he ela ionship be ween AI and he yea -o e -yea pe cen age change a e s a is ically
insigni ican .
18 Yaein Baek and Jiyun Lee
ⓒ Ko ea Ins i u e o In e na ional Economic Policy
Howe e , i should be no ed ha he analysis used da a om 2019 o 2023, and he
impac o COVID-19 on he U.S. labo ma ke mus be aken in o accoun . In pa icula ,
indus ies wi h high exposu e o AI, such as inance and insu ance, in o ma ion and
communica ions, and p o essional, scien i ic, and echnical se ices, expe ienced highe
wage g ow h a es and elewo k adop ion a es compa ed o o he indus ies. All h ee
indus ies ansi ioned o emo e wo k easily, and wo ke s in he in o ma ion and
communica ions and echnical se ices sec o s expe ienced longe wo king hou s due o
he lexibili y o emo e wo k.8
Table 6. U.S. NAICS 4-digi Indus y Indica o s Summa y S a is ics
Obse a ions Mean S anda d De ia ion Min Max
Labo P oduc i i y 781 104.34 19.73 54.91 195.71
Capi al P oduc i i y 258 101.59 7.94 80.35 122.68
To al Fac o P oduc i i y 258 9.96 8.12 73.64 146.06
In e media e Inpu s P oduc i i y 258 101.12 11.40 75.39 177.83
ln(Employmen ) 805 5.11 1.24 1.41 9.31
Hou ly Compensa ion 804 115.38 12.65 71.11 164.89
ln(Labo Compensa ion) 804 9.34 1.30 5.28 12.66
ln(Hou s Wo ked) 805 5.73 1.20 1.88 9.60
Ou pu pe Wo ke 781 102.80 16.12 59.32 190.06
Uni labo Cos s 781 111.88 15.53 56.64 183.90
ln(Capi al Cos s) 258 8.47 1.84 2.72 11.73
Labo Sha e 258 22.05 8.50 2.80 41.00
Capi al Sha e 258 19.49 12.24 0.90 80.50
Real Sec o al Ou pu 781 100.38 15.98 43.99 224.84
ln(Sec o al Ou pu ) 781 11.00 1.46 6.69 14.12
Capi al Inpu 258 97.93 6.19 73.84 114.36
Combined Inpu s 258 95.91 8.36 61.07 126.40
In e media e Inpu s 258 95.37 11.44 48.11 150.82
In e media e Inpu s P oduc i i y 258 101.12 11.40 75.39 177.83
No es: Da a om 2019-2023, Bu eau o Labo S a is ics (BLS) (h ps://www.bls.go /p oduc i i y/, accessed on
July 9, 2024).
8 Bu eau o Labo S a is ics (h ps://www.bls.go /spo ligh /2021/impac -o - he-co ona i us-pandemic- on-
businesses-and-employees-by-indus y, accessed on July 9, 2024), Wo ld Economic Fo um (h ps://
www.we o um.o g/agenda/2021/10/mic oso -s udy-co id19-wo k-hou s/, accessed on July 9, 2024).
Analysis o A i icial In elligence Exposu e Ac oss Indus ies in Sou h Ko ea and he Uni ed S a es 19
ⓒ 2025 Eas Asian Economic Re iew
Table 7. Rela ionship be ween NAICS 4-digi Indus y Indica o s in U.S. and
AI Exposu e Index
Dependen Va iable (1)
R
obo exclude
d
(2)
R
obo include
d
Labo P oduc i i y 6.4943 6.6150
(
7.3360
)
(
7.2287
)
Capi al P oduc i i
y
4.3859 4.6588
(
11.4701
)
(
11.1006
)
To al Fac o P oduc i i
y
-14.1212 -14.2046
(
10.9279
)
(
10.8775
)
In e media e Inpu s P oduc i i
y
-22.6724 -22.5405
(
14.2032
)
(
14.2944
)
ln(Employmen ) 1.4656** 1.3219**
(
0.6349
)
(
0.5975
)
Hou ly Compensa ion 8.8521* 8.1414
(
5.1437
)
(
5.0757
)
ln(Labo Compensa ion) 1.9893*** 1.8327***
(
0.6882
)
(
0.6472
)
ln(Hou s Wo ked) 11.9028** 1.2542**
(
5.8893
)
(
0.5915
)
Ou pu pe Wo ke
6.5950 6.9131
(
7.9288
)
(
7.7649
)
Uni Labo Cos s 7.4786 6.6959
(
8.0637
)
(
7.4342
)
ln
(
Ca
p
i al Cos s
)
0.4218 0.0231
(1.3161) (1.0938)
Labo Sha e 1.3488 1.7808
(5.7993) (5.7366)
Ca
p
i al Sha e -8.3574 -8.4172
(5.1952) (5.2496)
Real Sec o al Ou
p
u 19.4436 19.1482
(11.9588) (11.9253)
ln
(
Sec o al Ou
p
u
)
1.7201** 1.5495**
(0.7461) (0.7120)
Ca
p
i al In
p
u 4.0590 3.9299
(5.0737) (5.2332)
Combined In
p
u s 18.0555* 18.0274*
(9.7637) (9.7300)
In e media e In
p
u s 22.3869* 22.3193*
(12.4003) (12.4038)
In e media e In
p
u s P oduc i i
y
-22.6724 -22.5405
(14.2032) (14.2944)
Obse a ions 124-441 124-441
No es: The uni s o he a iables ha ha e no been log- ans o med a e Index (2017=100) o pe cen age.
Robus s anda d e o s in pa en heses a e clus e ed a he indus y le el. *, **, *** deno e signi icance
a he 10%, 5%, and 1% le els, espec i ely.
20 Yaein Baek and Jiyun Lee
ⓒ Ko ea Ins i u e o In e na ional Economic Policy
Table 8. Rela ionship be ween Indus y Employmen Indica o s in U.S. and
AI Exposu e Index
ln(Employmen ) Hou ly Compensa ion ln(Labo Compensa ion)
AIIE 1.4258** 0.0759* 1.9519***
(0.6433) (0.0424) (0.6980)
Labo P oduc i i y 0.0888 0.2567*** 0.0188
(0.6194) (0.0416) (0.7025)
αs yes yes yes
δ yes yes yes
Obse a ions 424 424 424
No es: Signed log ans o ma ion is used. Robus s anda d e o s in pa en heses a e clus e ed a he indus y
le el. *, **, *** deno e signi icance a he 10%, 5%, and 1% le els, espec i ely.
F om 2020 o 2021, a e age weekly wages in he in o ma ion indus y inc eased by
12.3% and 9.8% in inance and insu ance.9 The e o e, i is challenging o analyze he
impac o AI on indus ies using da a om he o e hea ed U.S. labo ma ke caused by
COVID-19.
2. Reg ession Resul s o Occupa ions in Ko ea
This sec ion examines occupa ion-le el employmen and exposu e o AI in Sou h
Ko ea du ing he pe iod om 2018 o 2022. We es ima e he ollowing eg ession model:
∆yi,o = αi + β1AIOEo + β2AIOEo × AIIEi + γ′Xi,o + εi,o, (4)
whe e ∆yi,o is he employmen DHS di e ence10 o occupa ion o in indus y i om 2018
o 2022, αi is he indus y ixed e ec s, AIOEo is he occupa ional AI exposu e index,
and Xi,o is he ec o o con ol a iables ha ep esen occupa ion cha ac e is ics in he
KLIPS da ase . These include he a e age mon hly wage, a e age weekly hou s wo ked,
p opo ion o emale wo ke s, and he p opo ion o ull- ime ( egula ) wo ke s, all in
9 Pew Resea ch Cen e , “How has he Co id-19 pandemic eshaped how US ge paid?” Jan 6, 2022
(h ps://www.we o um.o g/s o ies/2022/01/many-u-s-wo ke s-a e-seeing-bigge -paychecks-in-pandemic-
e a-bu -gains-a en- -sp ead-e enly/, accessed on Feb ua y 17, 2025).
10 The DHS di e ence is calcula ed as 2 × (s1 − so)/(s1 + s0), whe e s1 and s0 ep esen he p opo ions
a wo di e en ime poin s.
Analysis o A i icial In elligence Exposu e Ac oss Indus ies in Sou h Ko ea and he Uni ed S a es 21
ⓒ 2025 Eas Asian Economic Re iew
hei le els in 2018. In addi ion, we include he in e ac ion e m o he AIOE and he
AIIE index in 2019 in some o he speci ica ions.11
As Table 9 shows, AI exposu e has a posi i e impac on he change in employmen .
The ela ionship is s a is ically signi ican in columns (3) and (4), which includes a e age
wage, a e age hou s wo ked, and he sha e o emale wo ke s. Fo example, mo ing
om he 25 h o he 75 h pe cen ile o exposu e o AI is associa ed wi h an inc ease in
wi hin-indus y employmen sha es o 5.2%p, based on es ima es in column (4).12 The
s a is ical signi icance disappea s when he sha e o ull- ime wo ke s o he in e ac ion
e m is included. The e o e, he impac o AI exposu e on occupa ional employmen
changes does no depend on whe he he indus y as a whole is mo e o less exposed o
AI.13
Highe exposu e o AI is associa ed wi h an inc ease in wi hin-indus y occupa ional
employmen sha es. This is consis en wi h he Sou h Ko ea’s indus y-le el esul s in
sec ion 4.1; indus ies wi h high exposu e o AI a e associa ed wi h an inc ease in
employmen . No e ha we measu e he occupa ional employmen change om 2018 o
2022, which is a sho - e m change compa ed o he analysis in Han and Oh (2023) and
Webb (2019).14 Ou es ima es o he ela ionship be ween AI exposu e and occupa ional
employmen di e om hei s mainly because we a e examining sho - e m changes.
The sho - e m inc ease in occupa ional employmen wi h high AI exposu e can be
a ibu ed o h ee key ac o s. Fi s , he ex en o compu e usage is a signi ican
de e minan . Occupa ions wi h high AI exposu e, such as hose in he inance and
insu ance sec o s, ha e been ound o expe ience highe employmen g ow h. Acco ding
o Geo gie and Hyee (2022), s udies ha e shown ha jobs in ol ing ex ensi e
compu e use exhibi a s onge posi i e co ela ion be ween AI exposu e and
11 The in e ac ion e m AIOEo × AIIEi is compu ed a e each index is scaled be ween 0 and 1.
12 The 25 h pe cen ile o AIOE is -0.8537, he 75 h pe cen ile is 0.9342, hence 2.9280∙(0.9342−(−0.8537)) ≈
5.2. The minimum AIOE is -1.6079 (KSCO 910, cons uc ion and mining labo ) and he maximum
AIOE is 1.4045 (KSCO 271, human esou ces and managemen p o essionals).
13 Following a e e ee’s sugges ion, we also es ima e he ime a ia ion in β1 using he one-yea DHS
di e ence in occupa ional employmen . The esul s a e p esen ed in Appendix Table A8. Unlike he
es ima es om model 4, he in e ac ion e m is s a is ically signi ican . Gi en he sho ime ame
o ou sample, we ocus on he ime-in a ian ela ionship be ween AI exposu e and changes in
occupa ion-le el employmen .
14 Han and Oh (2023) examines he occupa ional employmen change be ween 2000 and 2021, and
Webb (2019) be ween 1980 and 2010.
22 Yaein Baek and Jiyun Lee
ⓒ Ko ea Ins i u e o In e na ional Economic Policy
employmen g ow h. This is because pa ial au oma ion d i en by AI no only di ec ly
enhances p oduc i i y bu also shi s he job ask composi ion owa ds highe - alue
ac i i ies, u he boos ing p oduc i i y. Such inc eases in labo p oduc i i y and ou pu
o se he di ec displacemen e ec s o au oma ion o wo ke s wi h s ong digi al skills.
Wo ke s ind i easie o e ec i ely u ilize AI and ansi ion o highe - alue asks wi hin
hei jobs ha canno be au oma ed.
Secondly, AI has he po en ial o bo h displace ce ain jobs and gene a e new demand
o labo (Guliye , 2023). AI can gene a e new jobs and expand exis ing ones. E en in
occupa ions wi h signi ican AI exposu e, human o e sigh and e iew can con ibu e o
employmen g ow h. Ins ead o eplacing jobs, AI can enhance human p oduc i i y,
d i ing job c ea ion and imp o ing e iciency. Due o he cu en limi a ions o AI
echnology, human judgmen emains indispensable in he sho e m. Consequen ly, in
highly AI-exposed occupa ions, human cogni ion and expe ise a e c ucial o boos ing
p oduc i i y.
Table 9. Rela ionship be ween Occupa ional Employmen in Sou h Ko ea and
AI Exposu e Index
(1) (2) (3) (4) (5) (6) (7) (8)
AIOE 3.0453** 2.2986 2.4595* 2.9280* 1.6073 2.0664 2.4186 1.5433
(1.4680) (1.4568) (1.4214) (1.4861) (1.4692) (1.5363) (2.6146) (2.7681)
Wage 0.0132 0.0113 0.0065 0.0092 0.0050 0.0113 0.0049
(0.0096) (0.0098) (0.0098) (0.0092) (0.0092) (0.0099) (0.0093)
Hou s Wo ked 0.1928 0.1467 0.1486 0.1087 0.1929 0.1094
(0.1647) (0.1677) (0.1726) (0.1753) (0.1644) (0.1751)
Female Wo ke s -5.4800 -4.9666 -5.0001
(3.3798) (3.3365) (3.3417)
Regula Wo ke s 8.4400* 8.0972 8.1223*
(4.4701) (4.4212) (4.4507)
AIOE×AIIE 0.2930 3.7503
(16.8702) (16.6251)
αi yes yes yes yes yes yes yes yes
Obse a ions 1279 1279 1279 1279 1279 1279 1279 1279
No es: Robus s anda d e o s in pa en heses a e clus e ed a he indus y le el. *, **, *** deno e signi icance
a he 10%, 5%, and 1% le els, espec i ely.
Analysis o A i icial In elligence Exposu e Ac oss Indus ies in Sou h Ko ea and he Uni ed S a es 23
ⓒ 2025 Eas Asian Economic Re iew
Finally, he issue o cos wa an s conside a ion. As echnologies capable o eplica ing
ask comp ehension and p o essional expe ise in highly AI-exposed occupa ions
con inue o e ol e, i emains unce ain whe he he associa ed cos s can be ully o se .
Gi en he incomple e s a e o echnological ad ancemen s, i canno be conclusi ely
asse ed ha occupa ions wi h high AI exposu e will be apidly eplaced by AI in a sho
pe iod. Mo eo e , he ime equi ed o secu e adequa e unding o AI echnology and
i s implemen a ion sugges s ha employmen in highly AI-exposed occupa ions is
unlikely o decline apidly in he sho e m. Consequen ly, employmen in hese
occupa ions may expe ience g ow h in he nea e m.
V. Conclusion
This s udy empi ically analyzes he impac o AI exposu e on employmen and
p oduc i i y in he indus ies o Sou h Ko ea and he Uni ed S a es. In Sou h Ko ea, key
indica o s ha showed a s a is ically signi ican posi i e co ela ion wi h he AI exposu e
index we e he numbe o employees, he numbe o emale employees, he numbe o
egula wo ke s, and sales. Con e sely, indica o s showing a nega i e co ela ion
included sales pe capi a, he labo sha e, and eal labo income pe capi a. Howe e ,
eg ession analysis con olling o labo p oduc i i y in indus ies e ealed ha he AI
exposu e index did no exhibi a s a is ically signi ican posi i e co ela ion wi h he
numbe o employees, emale employees, o egula wo ke s. This esul may align wi h
p io s udies sugges ing ha he adop ion o AI enhances labo p oduc i i y, he eby
inc easing employmen . None heless, ou analysis is based on da a om 2019 o 2022,
p io o he widesp ead adop ion o gene a i e AI. Gi en he po en ial ime lag in
ealizing he e ec s o new echnology adop ion and p oduc i i y gains, he esul s
canno de ini i ely con i m he labo p oduc i i y-enhancing e ec s o AI.
Unlike p e ious li e a u e on occupa ional employmen and AI exposu e, ou indings
show ha indus y-le el employmen is posi i ely co ela ed wi h AI exposu e. We u he
examine occupa ional employmen in Sou h Ko ea and ind ha i is also posi i ely
co ela ed wi h AI exposu e. The sho - e m inc ease in employmen o occupa ions
wi h high AI exposu e may be d i en by enhanced p oduc i i y h ough pa ial
au oma ion, he c ea ion o new labo demand equi ing human o e sigh , and he cu en
limi a ions and cos s o AI echnology ha delay ull au oma ion.
In he Uni ed S a es, indus ies wi h highe AI exposu e exhibi a s a is ically
24 Yaein Baek and Jiyun Lee
ⓒ Ko ea Ins i u e o In e na ional Economic Policy
signi ican posi i e ela ionship wi h employmen , labo compensa ion, and hou ly
compensa ion. This sugges s ha AI-exposed indus ies in bo h Sou h Ko ea and he U.S.
expe ience employmen g ow h. Howe e , a key di e ence be ween he wo coun ies
is ha , while labo compensa ion and hou ly compensa ion a e posi i ely co ela ed wi h
AI exposu e in he U.S., eal labo income pe capi a in Sou h Ko ea show a nega i e
co ela ion wi h he AI exposu e index. Mo eo e , he AI exposu e index demons a ed
a s a is ically signi ican posi i e co ela ion wi h indus ial employmen e en a e
con olling o labo p oduc i i y o he Uni ed S a es. In pa icula , he g ow h a e o
employmen and labo compensa ion a e also signi ican ly posi i ely co ela ed wi h
he AI exposu e. This inding sugges s ha AI exposu e di ec ly impac s employmen
and wages, and U.S. indus ies a e mo e in luenced by AI compa ed o Sou h Ko ea.
Howe e , ou analysis elies on da a om 2019 onwa d and we mus conside he
p o ound e ec s o COVID-19 on he U.S. labo ma ke .
In summa y, ou indings indica e a sho - e m inc ease in employmen wi hin
indus ies highly exposed o AI. In Sou h Ko ea, we also obse ed g ow h in
occupa ional employmen . Labo ma ke adjus men s o AI will be shaped by a ious
ac o s, including labo displacemen , AI complemen a i y, and p oduc i i y gains
(Cazzaniga e al., 2024). While AI has he po en ial o eplace a subs an ial sha e o
human labo , his displacemen is likely o occu g adually. This slowe ansi ion
p o ides policymake s wi h an oppo uni y o implemen s a egies ha mi iga e job
losses h ough wo k o ce e aining, social suppo sys ems, and incen i es o
businesses o adop AI in ways ha enhance a he han eplace human labo (S anbe g
e al., 2024).
Meanwhile, al hough he e is g owing in e es in he economic e ec s o gene a i e
AI, su icien da a o speci ically e i y he e ec s beyond 2022 a e no ye a ailable.
The e o e, i is impo an o no e ha he esul s o ou s udy e lec he sho - e m
impac s o AI and do no ully accoun o he implica ions o gene a i e AI. The
empi ical in es iga ion o he impac o gene a i e AI ac oss indus ies is le o u u e
esea ch.
Analysis o A i icial In elligence Exposu e Ac oss Indus ies in Sou h Ko ea and he Uni ed S a es 25
ⓒ 2025 Eas Asian Economic Re iew
REFERENCES
Acemoglu, D., Au o , D., Hazell, J., and P. Res epo. 2022. “A i icial in elligence and jobs:
E idence om online acancies.” Jou nal o Labo Economics, 40 (S1), S293-S340.
Au o , D. 2022. “The labo ma ke impac s o echnological change: F om unb idled en- husiasm
o quali ied op imism o as unce ain y,” NBER Wo king Pape , no. 30074. Na ional
Bu eau o Economic Resea ch.
Babina, T., Fedyk, A., He, A., and J. Hodson. 2024. “A i icial in elligence, i m g ow h, and
p oduc inno a ion.” Jou nal o Financial Economics, 151, 103745.
Bon iglioli, A., C inò, R., Gancia, G., and I. Papadakis. 2023. “A i icial in elligence and jobs:
E idence om us commu ing zones.” CESi o Wo king Pape , no. 10685.
B ynjol sson, E., Mi chell, T., and D. Rock. 2018. “Wha can machines lea n and wha does i
mean o occupa ions and he economy?” AEA pape s and p oceedings, ol. 108, pp. 43-47.
B ynjol sson, E., Rock, D., and C. Sy e son. 2019. “A i icial in elligence and he mode n
p oduc i i y pa adox.” The economics o a i icial in elligence: An agenda, ol. 23, pp. 23-
57.
Cazzaniga, M., Jaumo e, M. F., Li, L., Melina, M. G., Pan on, A. J., Pizzinelli, C., Rockall, E. J.,
and M. M. Ta a es. 2024. Gen-ai: A i icial in elligence and he u u e o wo k. In e na ional
Mone a y Fund.
Chui, M., Hazan, E., Robe s, R., Singla, A., and K. Smaje. 2023. The economic po en ial o
gene a i e AI. McKinsey&Company.
Cza ni zki, D., Fe nández, G. P., and C. Ramme . 2023. “A i icial in elligence and i m-le el
p oduc i i y.” Jou nal o Economic Beha io & O ganiza ion, ol. 211, pp. 188-205.
Fel en, E., Raj, M., and R. Seamans. 2018. “A me hod o link ad ances in a i icial in elligence o
occupa ional abili ies.” AEA Pape s and P oceedings, ol. 108, pp. 54-57.
———. 2021. “Occupa ional, indus y, and geog aphic exposu e o a i icial in elligence: A
no el da ase and i s po en ial uses.” S a egic Managemen Jou nal, ol. 42, no. 12, pp.
2195-2217.
F ey, C. B. and M. A. Osbo ne. 2017. “The u u e o employmen : How suscep ible a e jobs o
compu e isa ion?” Technological o ecas ing and social change, ol. 114, pp. 254-280.
Geo gie , A. and R. Hyee. 2022. “A i icial in elligence and employmen : New c oss-coun y
e idence.” F on ie s in a i icial in elligence, ol. 5, 832736. h ps://doi.o g/10.3389/ ai.
2022.832736
Guliye , H. 2023. “A i icial in elligence and unemploymen in high- ech de eloped coun ies:
New insigh s om dynamic panel da a model.” Resea ch in Globaliza ion, ol. 7, 100140.
Han, J. and S. Oh. 2023. “AI and he labo ma ke .” BOK Issue No e, no. 2023-30.
Han, J. 2023. “The impac o A i icial In elligence on he labo ma ke and policy implica ions.”
Resea ch Monog aph, no. 2023-03. Ko ea De elopmen Ins i u e.
26 Yaein Baek and Jiyun Lee
ⓒ Ko ea Ins i u e o In e na ional Economic Policy
Ha zius, J., B iggs, J. Kodnani, D., and G. Pie domenico. 2023. “The po en ially la ge e ec s o
a i icial in elligence on economic g ow h (b iggs/kodnani).” Economics Resea ch, Ma ch,
2023. Goldman Sachs.
Kim, S. 2024. “The AI epo ( ech. ep.).” Na ional In o ma ion Socie y Agency (NIA).
Lane, M. 2024. “Algo i hm and e e: How ai will impac women a wo k.” OECD Policy B ie ,
Decembe 6, 2024. Pa is, F ance, OECD.
Maslej, N., Fa o ini, L., Pe aul , R., Pa li, V., Reuel, A., B ynjol sson, E., E chemendy, J., Lige ,
K., Lyons, T., Manyika, J., Niebles, J. C., Shoham, Y., Wald, R., and J. Cla k. 2024. A i icial
In elligence Index Repo 2024. 10.48550/a Xi .2405.19522.
PwC S. 2024. The cu en s a us o business using gene a i e AI ( ech. ep.). Samil
P icewa e houseCoope s (in Ko ean).
Raj, M. and R. Seamans. 2019. “P ime on a i icial in elligence and obo ics.” Jou nal o
O ganiza ion Design, ol. 8, no. 11. h ps://doi.o g/10.1186/s41469-019-0050-0
Song, D., Cho, J., Choi, M., and H. Kim. 2021. “A i icial in elligence and i m p oduc i i y.” i-
KIET Issues&Analysis, 2021-12. Ko ea Ins i u e o Indus ial Economics & T ade (KIET).
S anbe g, M., Li, W., Fleming, M., Goeh ing, B., and N. Thompson. 2024. “Beyond ai exposu e:
Which asks a e cos -e ec i e o au oma e wi h compu e ision?” Wo king Pape .
h p://dx.doi.o g/10.2139/ss n.4700751
Tolan, S., Pesole, A., Ma ínez-Plumed, F., Fe nández-Macías, E., He nández-O allo, J., and E.
Gómez. 2021. “Measu ing he occupa ional impac o ai: Tasks, cogni i e abili ies and ai
benchma ks.” Jou nal o A i icial In elligence Resea ch, ol. 71, pp. 191-236.
Webb, M. 2019. “The impac o a i icial in elligence on he labo ma ke .” h p://dx.doi.
o g/10.2139/ss n.3482150
Yoon, S.-H., Yoon, J. E., Cho, S., Lee, J., Baek, Y., and N. S. Son. 2024. In es men in in angible
asse s and economic g ow h: Global ends and policy implica ions. Ko ea Ins i u e o
In e na ional Economic Policy.
Za i hona a , A. 2024. “Economics o Cha GPT: A labo ma ke iew on he occupa ional
impac o a i icial in elligence.” Jou nal o Elec onic Business & Digi al Economics, ol.
3, no. 2, pp. 100-116.
Fi s e sion ecei ed on Decembe 31, 2024
Pee - e iewed e sion ecei ed on Feb ua y 28, 2025
Final e sion accep ed on Ma ch 3, 2025
© 2025 EAER a icles a e dis ibu ed unde he e ms o he C ea i e Commons A ibu ion 4.0 In e na ional
License (h p://c ea i ecommons.o g/licenses/by/4.0/), which pe mi s un es ic ed use, dis ibu ion, and
ep oduc ion in any medium, p o ided you gi e app op ia e c edi o he o iginal au ho (s) and he sou ce, and p o ide a
link o he C ea i e Commons license.
Analysis o A i icial In elligence Exposu e Ac oss Indus ies in Sou h Ko ea and he Uni ed S a es 33
ⓒ 2025 Eas Asian Economic Re iew
Table A9. Top 15 U.S. AIIE
Ran
k
NAICS 3-digi AIIE NAICS 2-digi
1 C edi In e media ion and Rela ed Ac i i ies 1.17 Finance and Insu ance (52)
2 Insu ance Ca ie s and Rela ed Ac i i ies 1.13 Finance and Insu ance (52)
3 Funds, T us s, and O he Financial Vehicle 1.12 Finance and Insu ance (52)
4 Publishin
g
Indus ies 1.07 In o ma ion
(
51
)
5 P o essional, Scien i ic, and Technical Se ices 0.73 P o essional, Scien i ic, and Technical
Se ices (54)
6 Educa ional Se ices 0.66 Educa ional Se ices
(
61
)
7 Religious, G an making, Ci ic, P o essional,
and Simila O ganiza ions
0.62 O he Se ices (excep Public
Adminis a ion) (81)
8 Compu e and Elec onic P oduc Manu ac u ing 0.53 Manu ac u ing (33)
9 Mo ion Pic u e and Sound Reco din
g
Indus ies 0.44 In o ma ion
(
51
)
10 Ambula o
y
Heal h Ca e Se ices 0.39 Heal h Ca e and Social Assis ance
(
62
)
11 T ansi and G ound Passen
g
e T ans
p
o a ion 0.32 T ans
p
o a ion and Wa ehousin
g
(
48
)
12 Social Assis ance 0.31 Heal h Ca e and Social Assis ance (62)
13 Hospi als 0.23 Heal h Ca e and Social Assis ance (62)
14 Fede al, S a e, Local go e nmen , excluding
Educa ion and Hos
p
i als
0.21
15 Me chan Wholesale s, Du able Goo
d
0.18 Wholesale T ade (42)
No es: NAICS 3-digi AIIE alues a e au ho ’s calcula ion. We conduc ed hei analysis using he AIOE da a
om Fel en e al. (2021) (accessed on Ma ch 25, 2024), along wi h he Bu eau o Labo S a is ics
(BLS) Occupa ional Employmen P ojec ions Da a and BLS Occupa ional Employmen and Wage
S a is ics (h ps://www.bls.go /oes/, bo h accessed on July 9, 2024). AIIE’s mean alue is -0.09, he
median is -0.13, he 25 h pe cen ile is -0.53 and he 75 h pe cen ile is 0.16.
Table A10. Bo om 15 U.S. AIIE
Ran
k
NAICS 3-digi AIIE NAICS 2-digi
1 Cou ie s and Messenge s -1.12 T anspo a ion and Wa ehousing (49)
2 Fo es y and Logging -0.82 Ag icul u e, Fo es y, Fishing and Hun ing
(11)
3 Wood P oduc Manu ac u ing -0.73 Manu ac u ing (32)
4 Food Se ices and D inking Places -0.71 Accommoda ion and Food Se ices (72)
5 Hea y and Ci il Enginee ing Cons uc ion -0.70 Cons uc ion (23)
6 Food Manu ac u ing -0.68 Manu ac u ing (31)
7 Special y T ade Con ac o s -0.66 Cons uc ion (23)
8 Amusemen , Gambling, and Rec ea ion
Indus ies
-0.62 A s, En e ainmen , and Rec ea ion (71)
9 Mining (excep Oil and Gas) -0.61 Mining, Qua ying, and Oil and Gas
Ex ac ion (21)
10 Was e Managemen and Remedia ion
Se ices
-0.58 Adminis a i e and Suppo and Was e
Managemen and Remedia ion Se ices (56)
34 Yaein Baek and Jiyun Lee
ⓒ Ko ea Ins i u e o In e na ional Economic Policy
Table A10. Con inued
Ran
k
N
AICS 3-digi AIIE
N
AICS 2-digi
11 Pape Manu ac u ing -0.58 Manu ac u ing (32)
12 P ima y Me al Manu ac u ing -0.57 Manu ac u ing (33)
13 Plas ics and Rubbe P oduc s Manu ac u ing -0.56 Manu ac u ing (32)
14 Repai and Main enance -0.56 O he Se ices (excep Public
Adminis a ion) (81)
15 Adminis a i e and Suppo Se ices -0.53 Adminis a i e and Suppo and Was e
Managemen and Remedia ion Se ices (56)
No es: NAICS 3-digi AIIE alues a e au ho ’s calcula ion. We conduc ed hei analysis using he AIOE da a
om Fel en e al. (2021) (accessed on Ma ch 25, 2024), along wi h he Bu eau o Labo S a is ics
(BLS) Occupa ional Employmen P ojec ions Da a and BLS Occupa ional Employmen and Wage
S a is ics (h ps://www.bls.go /oes/, bo h accessed on July 9, 2024). AIIE’s mean alue is -0.09, he
median is -0.13, he 25 h pe cen ile is -0.53 and he 75 h pe cen ile is 0.16.
Table A11. Rela ionship be ween NAICS 4-digi Indus y Indica o s ( i s di e ences) in
U.S. and AI Exposu e Index (1)
Dependen Va iable (1)
R
obo exclude
d
(2)
R
obo include
d
Labo P oduc i i
y
-3.1934* -3.1439*
(1.8469) (1.8647)
Ca
p
i al P oduc i i
y
-8.0898* -7.9177*
(4.7866) (4.6632)
To al Fac o P oduc i i
y
-3.1521 -3.3194
(3.9488) (2.2322)
In e media e In
p
u s P oduc i i
y
-0.9266 -0.8963
(4.8160) (4.7580)
Em
p
lo
y
men 3.4795** 3.379**
(1.6425) (1.6701)
Hou l
y
Com
p
ensa ion -0.2763 -0.3755
(1.1084) (1.1001)
Labo Com
p
ensa ion 4.4548** 4.2490**
(2.0279) (2.0760)
Hou s Wo ke
d
4.5388*** 4.4296***
(1.6384) (1.6723)
Ou
p
u
p
e Wo ke
-2.1525 -2.1004
(1.9585) (1.9772)
Uni Labo Cos s 2.2089 2.1059
(1.5829) (1.5832)
Ca
p
i al Cos s -67.0484** -72.0847**
(26.7975) (27.8977)
Labo Sha e 7.2122* 7.0761*
(3.6302) (3.6538)
Ca
p
i al Sha e -47.6055** -51.6484**
(
21.1652
)
(
21.9740
)
Obse a ions 124-441 124-441
No es: Robus s anda d e o s in pa en heses a e clus e ed a he indus y le el. *, **, *** deno e signi icance
a he 10%, 5%, and 1% le els, espec i ely.
Analysis o A i icial In elligence Exposu e Ac oss Indus ies in Sou h Ko ea and he Uni ed S a es 35
ⓒ 2025 Eas Asian Economic Re iew
Table A12. Rela ionship be ween NAICS 4-digi Indus y Indica o s ( i s di e ences) in
U.S. and AI Exposu e Index (2)
Dependen Va iable (1)
R
obo exclude
d
(2)
R
obo include
d
Real Sec o al Ou pu 1.6047 1.5301
(
2.5853
)
(
2.5954
)
Sec o al Ou pu -1.4244 -1.4413
(
3.3809
)
(
3.3746
)
Sec o al Ou pu P ice De la o
-3.3794 -3.3194
(
2.2095
)
(
2.2322
)
Capi al In ensi
y
2.1664 2.2204
(
3.2665
)
(
3.2256
)
Capi al Inpu 3.4652** 3.5826**
(
1.5015
)
(
1.5932
)
Combined Inpu s -0.9760 -0.7420
(
3.1567
)
(
3.1239
)
Combined Inpu s Cos s -11.8287** -11.7143**
(
5.0846
)
(
5.0925
)
Combined Inpu s P ice De la o
-10.0052*** -10.1115***
(
3.5398
)
(
3.5830
)
In e media e Inpu s -4.3809 -4.0093
(
4.7569
)
(
4.7611
)
In e media e Inpu s Cos s -6.5572 -5.9597
(
5.5691
)
(
5.5702
)
Con ibu ion o Ca
p
i al In ensi
y
o Labo P oduc i i
y
0.7528 0.7362
(0.6809) (0.6717)
Obse a ions 124-441 124-441
No es: Robus s anda d e o s in pa en heses a e clus e ed a he indus y le el. *, **, *** deno e signi icance
a he 10%, 5%, and 1% le els, espec i ely.
Table A13. U.S. NAICS 4-digi Indus y Indica o s Summa y S a is ics
Obse a ions Mean S anda d De ia ion Min Max
Sec o al Ou pu P ice De la o 781 114.29 17.80 73.81 215.62
Capi al In ensi y 258 101.59 73.94 80.35 122.68
ln(Combined Inpu Cos s) 258 10.33 1.34 6.69 13.22
Combined Inpu s P ice De la o 258 107.35 11.86 74.71 155.39
ln(In e media e Inpu s Cos s) 258 9.78 1.34 6.13 13.07
Con ibu ion o Capi al In ensi y o Labo
P oduc i i
y
258 100.27 1.83 95.13 115.25
No es: Da a om 2019-2023, Bu eau o Labo S a is ics (BLS) (h ps://www.bls.go /p oduc i i y/, accessed on
July 9, 2024).
36 Yaein Baek and Jiyun Lee
ⓒ Ko ea Ins i u e o In e na ional Economic Policy
Table A14. Rela ionship be ween NAICS 4-digi Indus y Indica o s in U.S. and
AI Exposu e Index
Dependen Va iable (1)
R
obo exclude
d
(2)
R
obo include
d
Sec o al Ou pu P ice De la o
-7.6538 -7.6419
(
6.6971
)
(
6.7905
)
Capi al In ensi
y
-3.2700 -3.4608
(
4.9602
)
(
5.0457
)
ln(Combined Inpu Cos s) 1.2891 0.9557
(
1.0105
)
(
0.8674
)
Combined Inpu s P ice De la o
-11.0004 -11.1848
(
9.1387
)
(
9.0355
)
ln(In e media e Inpu s Cos s) 1.4210 1.0831
(
1.0095
)
(
0.8765
)
Con ibu ion o Capi al In ensi y o Labo P oduc i i
y
0.2500 0.3371
(
0.9683
)
(
0.9893
)
Obse a ions 124-441 124-441
No es: The uni s o he a iables ha ha e no been log- ans o med a e Index (2017=100) o pe cen age.
Robus s anda d e o s in pa en heses a e clus e ed a he indus y le el. *, **, *** deno e signi icance
a he 10%, 5%, and 1% le els, espec i ely.
A.3. De ails abou Da a and Indexes
This sec ion desc ibes he da a and he cons uc ion o he AI Indus ial Exposu e
(AIIE) o Sou h Ko ea and he U.S. To cons uc he AIIE, he AI Occupa ional
Exposu e (AIOE) de eloped by Fel en e al. (2021) was u ilized o bo h coun ies.15
The AIOE can be used as a e e ence indica o o p edic which occupa ions a e expec ed
o be mos signi ican ly a ec ed by u u e ad ances in AI echnology. A high AIIE
indica es a g ea e employmen sha e o wo ke s in jobs wi h high AI exposu e wi hin
ha indus y. Re e o he Appendix o he op and bo om indus ies o AIIE in Sou h
Ko ea and he Uni ed S a es o 2022.
A.3.1. Sou h Ko ea
The AIOE o Fel en e al. (2021) is based on he U.S. S anda d Occupa ional
Classi ica ion (SOC 2010) so i was con e ed o he Ko ean S anda d Classi ica ion o
Occupa ions (KSCO, 7 h edi ion) in o de o measu e he Sou h Ko ean AIOE.16
15 h ps://gi hub.com/AIOE-Da a/AIOE (accessed on Ma ch 25, 2024).
16 The U.S. S anda d Occupa ional Classi ica ion (SOC 2010, 6-digi ) was con e ed o he In e na ional
S anda d Classi ica ion o Occupa ions (ISCO-08), and hen u he con e ed o he Ko ean
Analysis o A i icial In elligence Exposu e Ac oss Indus ies in Sou h Ko ea and he Uni ed S a es 37
ⓒ 2025 Eas Asian Economic Re iew
To gene a e he Sou h Ko ean AIIE om 2019 o 2022, we u ilize da a om he Sou h
Ko ean AIOE and he Ko ean Labo & Income Panel S udy (KLIPS).17 To add ess he
issue o non-one- o-one ma ching be ween occupa ions and indus ies (i.e., when a single
occupa ion belongs o mul iple indus ies o mul iple occupa ions belong o a single
indus y, esul ing in a one- o-many ela ionship), he p opo ion o each occupa ion
belonging o a speci ic indus y was calcula ed o de i e weigh s.18 These weigh s we e
hen used o calcula e a weigh ed a e age ha con e s he AIOE in o AIIE.
The Sou h Ko ean indus y da a mainly come om he Ko ea P oduc i i y Cen e
(KPC) and S a is ics Ko ea (KOSIS), whe e he pe iod s a s om 2019 un il 2022. Table
1 p esen s he Sou h Ko ea 1-digi indus ial indica o s summa y s a is ics o he alue-
added analysis o lis ed companies om he KPC.19 The KPC indica o s can be
classi ied in o alue-added c ea ion and dis ibu ion indica o s, as well as g ow h and
p o i abili y indica o s by indus y. Table 3 uses he alue-added c ea ion & dis ibu ion
indica o s and he g ow h & p o i abili y indica o s by 1-digi indus y.
Tables 2 and 5 da a a e om he KOSIS Business Ac i i y Su ey.20 Table 4 also
shows a iables by 2-digi indus y using KOSIS da a.21 The indus ial obo ic usage
da a in Sou h Ko ea is sou ced om he KOSIS Business Ac i i y Su ey, speci ically
he da a on he de elopmen and u iliza ion o Fou h Indus ial Re olu ion echnologies
S anda d Classi ica ion o Occupa ions (KSCO, 7 h edi ion, 3-digi ). The con e sion om SOC o
ISCO was pe o med using da a om he U.S. Bu eau o Labo S a is ics (BLS) (h ps://
www.bls.go /soc/isco_soc_c osswalk.xls, accessed on Ma ch 25, 2024), and he con e sion om
ISCO o KSCO u ilized he conco dance able om he Ko ean S a is ical Classi ica ion Po al
(h ps://kssc.kos a .go.k :8443/ksscNew_web/index.jsp, accessed on Ma ch 29, 2024).
17 The Occupa ional Classi ica ion Codes in he Ko ean Labo and Income Panel S udy (KLIPS) a e based
on he Ko ean S anda d Classi ica ion o Occupa ions (KSCO, 7 h Edi ion) a he 3-digi le el, while
he indus y classi ica ion codes a e based on he Ko ean S anda d Indus ial Classi ica ion (KSIC,
10 h Edi ion) a he 3-digi le el. Acco ding o he KLIPS use guide (h ps://sma klips.kli. e.
k /klips/sma klips, accessed on Ap il 22, 2024), i he p o ided Open Code in o ma ion was
insu icien o classi y down o he 3-digi le el du ing he occupa ion and indus y coding p ocess,
a ’0’ was appended o he 2-digi alue o c ea e a 3-digi code.
18 Weigh s we e de i ed sepa a ely o each le el o indus y classi ica ion: 1-3 digi .
19 h ps://s a .kpc.o .k /in eg a ion/index (accessed on June 11, 2024).
20 h ps://kosis.k /s a H ml/s a H ml.do?o gId=101& blId=DT_1KI2001_S&conn_pa h=I2 (accessed on
July 8, 2024).
21 h ps://www.k-s a .go.k /me as c/msea100/s a sdcd a-popup?s a sCon mNo=101066 (accessed on July 8,
2024).
38 Yaein Baek and Jiyun Lee
ⓒ Ko ea Ins i u e o In e na ional Economic Policy
by 1-digi indus y.22 Table 9 shows he a iables om KLIPS.23
A.3.2. U.S.
Fel en e al. (2021) used 2019 employmen da a o calcula e he AIIE in he Uni ed
S a es. In his s udy, we cons uc ed an upda ed U.S. AIIE using he Occupa ional
Employmen and Wage S a is ics (OEWS) om he U.S. Bu eau o Labo S a is ics
(BLS) da a om 2020 o 2022.24
U.S. indus ial indica o s mainly come om BLS. The indica o s in Tables 6 and 7
a e om BLS OPT (U.S. Bu eau o Labo S a is ics, he O ice o P oduc i i y and
Technology’s da a se ies), whe e he pe iod s a s om 2019 un il 2023. 25 This
p oduc i i y da a is a ailable in qua e ly o annual o ma s and is p o ided ac oss h ee
da abases. The da ase includes De ailed Indus y P oduc i i y me ics (Labo , To al
Fac o , and S a e Labo ), and he comple e da ase was downloaded o use. I
encompasses 27 a iables and is classi ied by indus y acco ding o he NAICS a he 2-
o 6-digi le el.26 The indus ial obo ic usage da a in U.S. is sou ced om he Na ional
Cen e o Science and Enginee ing S a is ics (NSF) Annual Business Su ey: 2020.27
22 h ps://kosis.k /s a H ml/s a H ml.do?o gId=101& blId=DT_1EP1237_1&conn_pa h=I2 (accessed on
July 8, 2024).
23 h ps://sma klips.kli. e.k /klips/sma klips (accessed on No 1, 2024).
24 h ps://www.bls.go /p oduc i i y/da a.h m (accessed on July 9, 2024).
25 h ps://www.bls.go /p oduc i i y/ (accessed on July 9, 2024).
26 Va iables: Capi al cos s, Capi al inpu , Capi al in ensi y, Capi al p oduc i i y, Capi al sha e, Combined
inpu , Combined inpu cos s, Combined inpu p ice de la o , Con ibu ion o capi al in ensi y o labo
p oduc i i y, Con ibu ion o in e media e inpu s in ensi y o labo p oduc i i y, Employmen , Hou ly
compensa ion, Hou s wo ked, In e media e inpu , In e media ed inpu cos s, In e media e inpu
in ensi y, In e media e inpu s p oduc i i y, In e media e inpu s sha e, Labo compensa ion, Labo
p oduc i i y, Labo sha e, Ou pu pe wo ke , Real sec o al ou pu , Sec o al ou pu , Sec o al ou pu
p ice de la o , To al ac o p oduc i i y, Uni labo cos s. No e ha he uni s o he a iables ha
ha e no been log- ans o med a e Index (2017=100) o pe cen age. All da a is p o ided by he BLS
as % Change om he p e ious yea . Please e e o Table A15.
27 h ps://ncses.ns .go /pubs/ns 22344 (accessed on July 9, 2024).
Analysis o A i icial In elligence Exposu e Ac oss Indus ies in Sou h Ko ea and he Uni ed S a es 39
ⓒ 2025 Eas Asian Economic Re iew
Table A15. U.S. BLS Da a
Measu e Uni s Digi Desc ip ion
Labo P oduc i i y Index
(2017=100)
2-6 digi The e iciency wi h which goods and se ices a e
p
oduced ia labo hou s; o en e e ed o as ou pu
p
e hou .
Capi al
P oduc i i
y
Index
(
2017=100
)
3-6 digi The e iciency a which capi al inpu is used o p oduce
ou
p
u o
g
oods and se ices.
To al Fac o
P oduc i i
y
Index
(
2017=100
)
3-6 digi The e iciency a which combined inpu s a e used o
p
oduce ou
p
u o
g
oods and se ices.
In e media e Inpu s
P oduc i i
y
Index
(2017=100)
3-6 digi The e iciency a which in e media e inpu s a e use
d
in he p oduc ion o goods and se ices.
Employmen Index
(2017=100)
Thousands o
jobs
2-6 digi The numbe o jobs in a gi en sec o . An indi idual
who wo ks mul iple jobs has each o hei jobs coun e
d
in he employmen measu e, as his measu e is a coun
o jobs, no pe sons. The ypes o wo ke s in he
employmen measu e may ei he include (a)
employees-only (also e e ed o as wage and sala
y
wo ke s) o (b) all wo ke s—which includes employees,
uninco po a ed sel -employed wo ke s, and unpai
d
amily wo ke s.
Hou ly
Com
p
ensa ion
Index
(
2017=100
)
2-6 digi The sum o wage and bene i s paid pe hou o wo k.
Labo
Compensa ion
Index
(2017=100)
Millions o
cu en dolla s
2-6 digi Paymen s o labo o
p
oduce goods and se ices,
including wages, bene i s and o he mone a y o
nonmone a y paymen s.
Hou s Wo ked Index
(2017=100)
Millions o
hou s
2-6 digi The numbe o labo hou s wo ked by all wo ke s,
including wage and sala y wo ke s, uninco po a e
d
sel -employed wo ke s, and unpaid amily wo ke s, i
n
he p oduc ion o goods and se ices.
Ou pu pe Wo ke Index
(
2017=100
)
2-6 digi The e iciency wi h which goods and se ices a e
p
oduced ia wo ke s.
Uni Labo Cos s Index
(2017=100)
2-6 digi The paymen s o labo se ices used o p oduce eac
h
uni o goods and se ices.
Capi al Cos s Millions o
cu en dolla s
3-6 digi The paymen s o capi al inpu o use in he p oduc ion o
goods and se ices.
Labo Sha e Pe cen age 3-6 digi The p opo ion o cu en -dolla ou pu a ibu ed o
he use o labo .
Capi al Sha e Pe cen age 3-6 digi The p opo ion o cu en -dolla ou pu p oduc io
n
a ibu ed o he use o capi al inpu .
Real Sec o al
Ou pu
Index
(2017=100)
2-6 digi The amoun o goods and se ices p oduced by a
n
indus y o deli e y o consume s ou side ha indus y.
Sec o al Ou pu Millions o
cu en dolla s
2-6 digi The cu en dolla alue o goods and se ices p oduce
d
b
y an indus y o deli e y o consume s ou side ha
indus y.
40 Yaein Baek and Jiyun Lee
ⓒ Ko ea Ins i u e o In e na ional Economic Policy
Table A15. Con inue
d
Measu e Uni s Di
g
i Desc i
p
ion
Capi al Inpu Index
(2017=100)
3-6 digi The con ibu ion o p oduc ion om capi al asse s.
Capi al asse s a e he
p
oduc i e ools (equipmen ,
s uc u es, in en o ies, land, in ellec ual p ope y, e c.)
ha can be e-used in u u e ime pe iods a e hey a e
p
u chased.
Combined Inpu s Index
(2017=100)
3-6 digi The agg ega e o measu ed inpu s ha a e used o
p oduce goods and se ices. These can include capi al,
labo , ene
gy
, aw ma e ials, and
p
u chased se ices.
In e media e Inpu s Index
(2017=100)
3-6 digi The goods and se ices (including ene gy, aw ma e ials,
semi inished goods, and se ices ha a e pu chase
d
om all sou ces) ha a e used in he p oduc ion o
o he goods o se ices a he han o inal consump ion.
In e media e Inpu s
P oduc i i
y
Index
(2017=100)
3-6 digi The e iciency a which in e media e inpu s a e use
d
in he p oduc ion o goods and se ices.
Sec o al Ou pu
P ice De la o
Index
(
2017=100
)
2-6 digi The ela i e change in he p ice o sec o al ou pu o e
ime.
Capi al In ensi y Index
(2017=100)
3-6 digi The a io o he amoun o capi al inpu used ela i e
o he amoun o labo hou s used o p oduce ou pu o
g
oods and se ices.
Combined Inpu s Index
(2017=100)
3-6 digi The agg ega e o measu ed inpu s ha a e used o
p oduce goods and se ices. These can include capi al,
labo , ene gy, aw ma e ials, and pu chased se ices.
Combined Inpu
Cos s
Millions o
cu en dolla s
3-6 digi The paymen s and implici compensa ion o u ilize all
in
p
u s o
p
oduce
g
oods and se ices.
Combined Inpu s
P ice De la o
Index
(2017=100)
3-6 digi The ela i e change o e ime in he alue o money
spen o use all inpu s o p oduc ion. This in o ma io
n
helps make combined inpu u iliza ion in di e en
ime pe iods compa able.
In e media e Inpu s
Cos s
Millions o
cu en dolla s
3-6 digi The paymen s o pu chase in e media e inpu s (ene gy,
ma e ials, and se ices
)
o
p
oduce
g
oods and se ices.
Con ibu ion o
Capi al In ensi y o
Labo P oduc i i y
Index
(2017=100)
3-6 digi The po ion o labo p oduc i i y change a ibu ed o
he change in he use o capi al ela i e o hou s wo ked.
No es: All measu es ha e hei con e ed alues as % Change om p e ious yea included in he da a. Sou ce
om BLS (h ps://www.bls.go /p oduc i i y/glossa y.h m#P, accessed on July 9, 2024).
