106
ACC JOURNAL 2022, Volume 28, Issue 2 DOI: 10.15240/ ul/004/2022-2-009
HOW UNIVERSITIES DETERMINE ECONOMIC DEVELOPMENT IN 27 EU
MEMBER STATES
Olesya Pe enko
Uni e si y o Wes Bohemia, Facul y o Economics,
Depa men o Economics and Quan i a i e Me hods;
Uni e zi ní 22, 301 00 Pilsen, Czech Republic
e-mail: [email p o ec ed]
Abs ac
Many Eu opean coun ies nowadays a e collabo a i ely ocused on b inging oge he he mos
up- o-da e echnologies and he b igh es minds o deal wi h social, economic, and ecological
ma e s and ind a sus ainable equipoise. Uni e si ies may help o p oduce app op ia e
knowledge o such challenges and os e economic and social inno a ion. This pape e iews
e idence o exis ing bonds be ween e ia y educa ion and economic g ow h measu ed in
GDP pe capi a p o iding a quan i a i e e alua ion o such dependencies using he now
widely c i icized Cobb-Douglas p oduc ion unc ion o building O dina y Leas Squa es
(OLS) and Bayesian A e aging o Classical Es ima es (BACE) econome ic models. The
esul s ob ained in his esea ch showed ha e ia y educa ion expendi u es and he numbe s
o mainly male BA and MA g adua es (in echnologies, sciences and medicine obus ly and
pa ially co ela e wi h economic g ow h. Well-dis ibu ed in es men in he de elopmen o
e ia y educa ion STEM majo s can po en ially s eng hen uni e si ies’ posi i e impac on
sus ainable economic g ow h.
Keywo ds
Economic g ow h; Te ia y educa ion; Sus ainable de elopmen ; STEM majo s.
In oduc ion
Rising in e es in he ole o e ia y ins i u ions in economic g ow h is c ea ing space o
scien i ic discussions and esea ch in educa ion, he economics o educa ion and human
esou ces managemen [17], [2], [26]. I migh he e o e be impo an o de e mine wha
impac e ia y educa ion o ganiza ions ha e on economic g ow h. One majo heo e ical issue
ha has domina ed he ield o many yea s is whe he GDP and i s de i a i es a e an
adequa e measu e o economic g ow h and de elopmen . This concep has been challenged by
a numbe o empi ical s udies [7], [16], [36], [37]. Mo eo e , schola s a e now wo king ha d
o es ablish he de ini ion o ‘e ec i e’ uni e si ies and hei impac on inno a ion
implemen a ion and he g ow h o egional economies. Huggins and Johns on desc ibed
uni e si ies as d i e s o he egional inno a ion sys em in 2009. Acco ding o some
esea che s [1], [25], [15], uni e si ies a e iewed by socie y as a whole as social and
educa ional enues ha help indi iduals acqui e speci ic skills o be able o mee economic
needs, be capable o making an e ec i e con ibu ion o economic de elopmen and mee
cu en ma ke demands. Acco ding o con empo a y esea ch and ela ed policies, economic
g ow h is closely connec ed o he human capi al endowmen . Fo ins ance, Ba o and Lee [5]
poin a causal ela ion be ween yea s o educa ion and economic g ow h. Ne e heless, he
ela ion seems less s aigh o wa d han de ined in [16]. A simple causa ion ela ionship
be ween educa ion and g ow h may no always sui all he economies and is being ques ioned
107
especially du ing economic c ises like he Global Financial C isis (GFC) o he cu en Co id-
19 pandemic. Wha migh be o g ea e impo ance han he numbe o school yea s o he
a io o a well-educa ed popula ion ela ed o he whole popula ion in gene al, is s a egic
planning o e ia y educa ion. I was demons a ed in [26] ha imposing s a egical hinking
(aligning p io i ies, alues, and incen i es o he uni e si y o hose de ined by local, egional,
and o na ional au ho i ies) o uni e si y managemen is one o he key elemen s needed o
enhance a uni e si y’s pe o mance and e iciency. Se e al s udies dealing wi h his issue
ha e been published and se e as he ou look on pe o mance measu emen in highe
educa ion [8], [23], al hough e y ew models o pe o mance measu emen a e o his day
ans e ed om he o -p o i sec o and la e adjus ed o sui such public o ganiza ions.
1 Resea ch Objec i es
The p ocess o building a anspa en sys em ha could help o de e mine he e iciency o a
highe educa ion o ganiza ion would g ea ly need o be explained clea ly o manage s and
policy make s so ha syne gy among policy make s, managemen , human esou ces,
p o essionals, and s uden s can be eached and he o e all sys em can wo k mo e e ec i ely,
p o iding he wo ld wi h be e oppo uni ies o sus ainable g ow h. S a emen o he
p oblem: The cause-e ec ela ionship be ween e ia y educa ion and economic g ow h and
de elopmen has been a ocus o mul iple esea ch p ojec s. Howe e , so a he e has been
li le discussion abou he economic e ec s o e ia y educa ion in Eu opean coun ies. This
s udy a emp s o de e mine he impac o highe educa ion sys ems on economic g ow h
wi hin 27 EU Membe s a es o e he pas ew decades, as well as he e ec s o Mas e ’s and
Ph.D. le el g adua es o e all and di ided by gende in he ields o science, medicine, and
echnology. The aim o he esea ch: his s udy a emp s o ca y ou eliable empi ical
es ima es o he ela ionship be ween e ia y educa ion and economic de elopmen in he
selec ed Eu opean Union coun ies du ing he pe iod o 2013-2019 in o de o de e mine he
impac o e ia y educa ion on he economic de elopmen o he EU membe s a es.
Limi a ion o he esea ch: In he amewo k o his s udy, se e al indica o s o economic
g ow h and de elopmen we e used – he g ow h a e o eal GDP pe capi a, GDP pe capi a
in PPP p ice a es, and e ia y educa ion ep esen ed by a sha e o he popula ion holding a
e ia y deg ee. Rela ionships be ween speci ic e ia y deg ees in social sciences, echnical
sciences, chemis y, biology, e c., and economic g ow h we e no conside ed in his pape ,
which is ano he limi a ion o he cu en esea ch. Al e na i e measu emen s o economic
g ow h we e no used o c ea e a da a se . Despi e hese limi a ions, i is expec ed ha he
esul s o his esea ch will be able o be used o conside u he s eps owa ds building
sus ainable educa ion in he EU Membe S a es.
Resea ch Ques ions
1. Wha is he impac o he cha ac e is ics o highe educa ion sys ems on economic g ow h
in Eu opean coun ies?
2. Wha is he impac o human capi al educa ed in science, echnology, and medicine
(Bachelo ’s and Mas e ’ le el) on he economic g ow h?
3. Wha is he impac o male and emale g adua es (Bachelo ’s and Mas e ’ le el) o all
majo s on he economic g ow h?
4. Wha is he impac o male and emale g adua es (Bachelo ’s and Mas e ’ le el) majo ing
in science, echnologies and medicine on g ow h?
2 Theo e ical Backg ound
In oday’s economy, e ia y educa ional se ices ha e become an impo an commodi y as
HR specialis s ecognize he employee’s quali ica ions, knowledge, and skills o ensu e
108
p oduc i i y and inno a ion wi hin a numbe o businesses ( ans-na ional co po a ions,
medium-sized companies ope a ing globally). Th ee key di ec ions o educa ion-g ow h
ela ionship a e de ined in [14]: i is c ucial o enhance he le els o knowledge, expe ise, and
skills o he popula ion, ans e new knowledge and ideas esponsibly and p o ide incen i es
o inno a ion wi hin o ganiza ions and he economy as a whole.
In gene al, educa ion can be conside ed one o he main ac o s o economic g ow h and
echnological p og ess [5], [21]. I has been claimed ha economic g ow h leads o highe
pa icipa ion in e ia y educa ion as well-educa ed human capi al p omo es g ea e economic
pe o mance. A lo o esea ch has been ocused on he cause-e ec ela ionship be ween
educa ion in gene al (and e ia y educa ion in pa icula ) and economic g ow h [17], [5], [2],
[9], [13], [14] wi h da a ga he ed om de eloping and de eloped coun ies om 1960 o
2019. Bo h a iables o educa ion ( e ia y educa ion) and economic g ow h we e measu ed
in nume ous ways. The e o e, he causal ela ionship be ween e ia y educa ion and
economic g ow h can be o mally desc ibed by h ee main app oaches: 1) educa ion causes
economic g ow h; 2) economic g ow h causes mo e people o seek e ia y educa ion 3) he
cause-e ec ela ionship may unc ion bo h ways. Acco ding o Hanushek and Woessman
[19] who s udied he ela ionship o cogni i e skills in educa ion and g ow h in mul iple
coun ies, adding mo e yea s o educa ion on a e age does no ensu e economic g ow h. The
esea che s a gued ha he di e ences in cogni i e skills – in o he wo ds, wha is known in
ce ain coun ies, can be conside ed an explana o y a iable o some o he di e ences in he
economic de elopmen in a ious coun ies.
In acco dance wi h UNESCO’s Sus ainable De elopmen Goal 4 (SDG 4) o educa ion [27]
sus ainabili y in e ia y educa ion will be c ucial o p epa ing cu en s uden s o sol e he
mos challenging asks ha he wo ld’s economy encoun e s in he nea u u e. Eu opean
Commission ep esen a i es s ess in hei epo s ha mo e STEM educa ion in e ia y
disciplines is needed by 2050 [38, p.13-14]. Ano he epo clea ly s a es ha “we mus spend
mo e on esea ch and educa ion”, he e o e, i migh be implied ha mo e s uden s will be
suppo ed o g adua e om e ia y educa ion p og ams [39, p. 56]. Sus ainable e ia y
educa ion goals lea e no choice o uni e si ies bu o supply hei po en ial g adua es wi h
global educa ion and aining so ha ewe nega i e ou comes a ise o be se led by u u e
gene a ions. The e o e, in o de o comple e SDG 4 i may be impo an o ensu e ha e ia y
educa ion gi es s uden s p ope sou ces o knowledge and mo i a ion o assess p oblems om
a ious pe spec i es so ha hey a e able o make in o med decisions elying on mul iple
epu able sou ces o da a a ailable.
Te ia y educa ion o ganiza ions need o de ine wha e iciency in e ia y educa ion is
because esea che s and educa o s wo ldwide o en eel conce ned abou echniques how o
pu sue so-called “e iciency”. A ew es ima ion me hods we e de eloped in he la e 20 h
cen u y, such as non-pa ame ic: e.g. da a en elopmen analysis (DEA) desc ibed in [12] and
[4] and pa ame ic: e.g. s ochas ic on ie analysis (SFA) ope a ed by [3] and [6]. These
me hods g ea ly helped o de ine e iciency in he con ex o e ia y educa ion. These
conce ns mainly s em om he idea ha i educa o s and uni e si y s a in gene al s a
ocusing mo e on being e icien , he e y concep o highe educa ion migh su e as mo e
a en ion is gi en o mee ing ce ain equi emen s, i.e. [20] and [24].
The e m “e iciency” is ypically explained as he oppo uni y o p o ide he bes educa ional
p oduc o a gi en budge . I can gene ally wo k in wo main di ec ions: uni e si ies may use
he same amoun o esou ces o po en ially acqui e be e esul s o lowe he amoun o
esou ces o ecei e he same esul s han in p e ious pe iods. O e he las ew decades i is
becoming inc easingly impo an o calcula e and ecognize he economic impac o
uni e si ies and o he o ganiza ions p o iding e ia y educa ion, see Figu e 1.
109
Sou ce: Au ho ’s own calcula ions based on da a om Cla i a e Da abase 2022
Fig. 1: F equency change in he numbe o scien i ic esea ch pape s in economics pe yea ,
in coun s, published ia Web o Science 2001-2021 ha ea u e he wo ds “economic
impac o uni e si ies” in hei headlines and abs ac s
Huggins and Johns on [20] desc ibed uni e si ies as d i e s o egional inno a ion sys em in
2009. I is c ucial, howe e , o look a he s uc u e o he human capi al and build up a
common sys em o manage i ( ec ui , ain and sus ain) in o de o de e mine possibili ies o
sus ainable g ow h in esul s in highe educa ion.
3 Me hods
The da a was ga he ed om he Eu os a (2021), OECD (2021) and Wo ld Bank (2021)
o icial da abases. A lis o a iables and co esponding da a sou ces can be ound in Table 7.
The c i e ia o choosing he a iables we e as ollows: 1) A ailabili y – he da a a e a ailable
o he majo i y o he cu en EU Membe s S a es om 1990 o p esen ; 2) Consis ency –
cong uen e ia y educa ion da a o he 27 EU Membe S a es ha was calcula ed and
ecei ed by ca ying ou exac ly he same p ocedu e o yea s 2013-2019. Despi e he ac ha
he pe iod o six yea s could be conside ed o be an imi a ion o he cu en esea ch, i is
expec ed ha a common ela ionship be ween e ia y educa ion and economic g ow h will be
ound as he lis o coun ies chosen o his s udy seems o be homogeneous as i consis s o
coun ies wi h de eloped economies acco ding o he Wo ld Bank (2019).
The ad an age o analyzing such a da a se is da a cohe ence – an impo an da a quali y
componen ha ensu es uni o mi y as well as exis ing logical connec ions and comple eness
o he da ase . Cohe ence could also enable he making o a logical dis inc ion be ween
concep s and a ge popula ions, which means ha mos majo p oblems could be easily
de ec ed du ing he da a p epa a ion s age. These es ic ions esul ed in he selec ion o a ew
a iables, hei means and s anda d de ia ions a e shown in Table 7.
To analyze he da a, i was decided o use he Cobb-Douglas p oduc ion unc ion (CDPF) (1)
using capi al s ock, capi al and labo se ices, despi e he ac ha CDPF o en e u ned a
nega i e sign o capi al [30], [31] and has been conside ed as “ ans o ma ion o income
iden i y” [29] and [10]. The main ad an age o his unc ion used in i s gene al o m is i s
abili y o explain he agg ega e ou pu c ea ion and economic g ow h isually. The unc ion
illus a es cons an e u ns o scale (α + β = 1) when elas ici ies o p oduc ion on p oduc ion
ac o s equal ac o sha es, wi h bo h coe icien s being posi i e numbe s anging [0, 1].
-60
-40
-20
0
20
40
60
80
100
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
110
𝑌
𝑡= 𝐴𝑡∙(𝐾𝑡)𝛼∙(𝐿𝑡)𝛽, (1)
whe e A is o al ac o p oduc i i y, K is capi al, and L is labo .
Due o he ac ha mos o he unce ain y o he economic g ow h models’ hinde s
ag eemen on speci ic ac o s ha cause economic g ow h, his pape uses panel da a models
wi h coun y-speci ic ixed e ec s. Woold igde [28] poin ed ou ha i migh be mo e
e icien o use Fixed E ec s – Random E ec s models ins ead o he egula OLS eg essions
while wo king wi h panel da ase s in which he e is he e ogenei y. This app oach o assess
long- e m and sho - e m economic g ow h has been commonly used in he economic
li e a u e due o i s gene al simplici y. The main d awback migh be a limi ed numbe o
a iables ha could be explained empi ically. Thei quan i y o en depends on “wha e e lis
he i s esea che happened o selec ” [33].
To deal wi h he issue o biases, qui e a ew esea che s also conside using Bayesian model
a e aging echniques [32] which may help o de e mine model unce ain y so ha he
ela ionship be ween model-speci ic es ima es is assessed, e ealed and explained. Following
his idea, economis s Sala-I-Ma in, Doppelho e and Mille [35] c ea ed a Bayesian Model
A e aging o Classical Es ima es (a so-called SDM’s BACE app oach) in o de o unde s and
which eg esso s should be included in o c oss-coun y linea eg essions. Such models build
es ima es o e e y possible combina ion o a iables by applying he weigh ed a e aging
OLS me hod in o de o ind ou which a iables do ela e o g ow h obus ly and how s ong
hese ela ionships migh be. The e o e, when designing he cu en esea ch, i was decided
o build an al e na i e model o compa e and e alua e he esul s ecei ed om he Fixed
E ec s model.
Mo e ecen ly, li e a u e on applied econome ics o e s “ex eme bounds analysis” which
was designed o e eal obus empi ical ela ionships o he de e minan s o economic
g ow h. This es consis s o wo s eps:
1. i is necessa y o iden i y (by p io analysis) which a iables could be ela ed o economic
g ow h;
2. o check i a a iable z is obus , equa ion (2) o eg essions needs o be sol ed:
𝛾 = 𝛼𝑗+ 𝛽𝑦𝑗 ∙ 𝑦 + 𝛽𝑧𝑗 ∙ 𝑧 + 𝛽𝑥𝑗 ∙ 𝑥𝑗+ 𝜀, (2)
whe e
y is a ec o ha ep esen s ixed posi ions o he eg esso s (ce ain a iables which a e
always included in o eg essions – e.g. income, in es men a e, seconda y school en ollmen
a e, a e o popula ion g ow h [34]),
z is he a iable o be examined, and
x is a “ ool” ec o which ypically consis s o a combina ion o h ee a iables selec ed o
he analysis.
The i s es s pe o med a he beginning o he 1990s we e widely c i icized in he economic
li e a u e as hey disca ded mos o he a iables as no obus due o he ac ha hese
eg esso s did no sys ema ically co ela e wi h economic g ow h. Consequen ly, Sala-i-
Ma in [35] sugges ed making a ansi ion om “ex eme bounds” o a iables ha would
ha e a ce ain deg ee o con idence. Bo h heo y and s a is ical calcula ions based on he
BACE app oach a e explici ly explained in [35] and o ha eason u he heo e ical
desc ip ion is omi ed in his pape .
The e has been much deba e be ween economis s on he subjec o whe he o no he e is a
ixed se o a iables which can be obus ly co ela ed wi h economic g ow h. In o de o

111
answe he esea ch ques ions lis ed in he in oduc ion o his a icle, i is conside ed use ul o
ind ou es ima es o g ow h om a much la ge se o models wi h he help o he BACE
app oach.
4 Resul s
The null hypo hesis o answe he i s esea ch ques ion (“Wha is he impac o he
cha ac e is ics o highe educa ion sys ems on economic g ow h in Eu opean coun ies”) was
s a ed as: The e is no impac o e ia y educa ion cha ac e is ics on economic g ow h. We
cons uc ed an OLS model wi h ixed e ec s o he panel da ase ha ing lagged a iables o
4 yea s o bachelo s uden s and 2 yea s o mas e s uden s. The dependen a iable was
cu en GDP pe capi a in pu chasing powe pa i y (Table 1) wi h he R-squa ed o 0.79.
The eg esso s used o his model included he ones ha a e gene ally used by a ious
esea che s: g oss capi al o ma ion, Thousand hou s wo ked(as K and L a iables o he
Cobb-Douglas unc ion), gene al economic a iables (unemploymen a e, popula ion and
popula ion g ow h sha e, gene al go e nmen consump ion expendi u es, openness o he
economy) and he a iables ha e e ed o e ia y educa ion ( o al numbe o g adua ed
bachelo and mas e s uden s, go e nmen expendi u es on e ia y educa ion and sha e o he
popula ion aged 30-35 wi h a e ia y diploma). I could be assumed ha g adua ed bachelo
and mas e s uden s ha e a posi i e impac on economic g ow h. The same da ase was used
o o m mul iple BACE models, he esul s ollow in Table 2.
Tab. 1: Model 1: Fixed-e ec s, Robus (HAC) s anda d e o s, dependen a iable: LOG
GDP in cu en PPP pe capi a
Coe icien
S d. e o
- a io
p- alue
Cons an
4.891050000
0.301363000
16.2300
<0.0001
***
Thousand hou s wo ked
−0.000162264
0.000206555
−0.7856
0.4392
G oss capi al o ma ion
0.001290160
0.000750301
1.7200
0.0974
*
Sha e o age g oup 30-35 wi h a e ia y
diploma
−0.000644451
0.001286540
−0.5009
0.6206
Unemploymen a e
−0.006935170
0.002318630
−2.9910
0.0060
***
G ow h o popula ion sha e
0.000426793
0.000172614
2.4730
0.0203
**
G adua ed bachelo s uden s_4
3.11526e-06
1.53880e-06
−2.0240
0.0533
*
G adua ed mas e s uden s_2
2.02081e-06
8.50523e-07
2.3760
0.0252
**
G adua ed Ph.D. s uden s
−2.13229e-06
2.18339e-06
−0.9766
0.3378
Go e nmen expendi u es on e ia y
educa ion
0.009721160
0.038023200
0.2557
0.8002
Go e nmen consump ion expendi u es
0.006115850
0.001735630
3.5240
0.0016
***
Openness o he economy
−0.001259470
0.001395340
−0.9026
0.3750
Sou ce: Own
The BACE modelling analysis me hod applying pos e io momen s wi h uncondi ional and
condi ional inclusion e u ned he ollowing esul s: g oss capi al o ma ion, unemploymen
a e obus ly and ma ginally co ela es wi h economic g ow h. The numbe s o g adua ed
bachelo s uden s obus ly mode a ely co ela e wi h g ow h. The hypo hesis o no impac o
e ia y educa ion on economic g ow h is ejec ed because o al numbe o bachelo and
mas e s uden s obus ly and posi i ely co ela e wi h GDP g ow h pe capi a bo h in he ixed
e ec s model and BACE models.
112
Tab. 2: BACE Models (61 models accep ed ou o 1024). Dependen a iable: LOG GDP in
cu en PPP pe capi a
PIP
Mean
S d. de .
Cond.
mean
Cond.
s d. de
Cons an
1.000000
4.296766
0.171460
4.296766
0.171460
Go e nmen
expendi u es on e ia y
0.999996
−0.120231
0.022694
−0.120232
0.022693
Thousand hou s wo ked
0.914534
−0.000166
0.000075
−0.000182
0.000058
G oss capi al o ma ion
0.596062
0.002394
0.002323
0.004017
0.001592
Unemploymen a e
0.588710
−0.002529
0.002504
−0.004295
0.001750
G adua ed bachelo
s uden s_4
0.307443
0.000000
0.000000
0.000000
0.000000
G adua ed mas e
s uden s_2
0.203128
0.000000
0.000000
0.000000
0.000000
G ow h o popula ion
sha e
0.067852
−0.000004
0.000392
−0.000058
0.001503
Sou ce: Own
The hypo hesis o answe he second esea ch ques ion was s a ed as: The e is no signi ican
impac human capi al educa ed in science and echnology (Bachelo ’s, Mas e ’s and Ph.D.
le el) on economic g ow h. We cons uc ed an OLS model wi h ixed e ec s o he panel
da ase ha ing lagged a iables o 4 yea s o bachelo s uden s and 2 yea s o mas e
s uden s. The dependen a iable was cu en GDP pe capi a in pu chasing powe pa i y
(Table 3) wi h he R-squa ed o 0.788. The esul s o g adua ed Ph.D. s uden s did no e u n
any s a is ically signi ican esul s which migh mean ha i is a he cos ly o educa e
p ospec i e scien is s and i akes ime o he economy o posi i ely eac o high quali y
human capi al. I could be assumed ha g adua ed bachelo and mas e s uden s ha e
a posi i e impac on economic g ow h. The same da ase was used o o m mul iple BACE
models, he esul s ollow in Table 4.
Tab. 3: Model 2: Fixed-e ec s, Robus (HAC) s anda d e o s, dependen a iable: LOG
GDP in cu en PPP pe capi a
Coe icien
S d. e o
- a io
p- alue
Cons an
4.674090000
0.341751000
13.6800
<0.0001
***
G oss capi al o ma ion
0.001252570
0.000755304
1.6580
0.1093
Sha e o age g oup 30-35 wi h a e ia y
diploma
−0.000698896
0.001363940
−0.5124
0.6127
Unemploymen a e
−0.009768940
0.003306760
−2.9540
0.0066
***
G ow h o popula ion sha e
0.000437482
0.000160227
2.7300
0.0112
**
Go e nmen expendi u es on e ia y
educa ion
−0.033137300
0.049195400
−0.6736
0.5065
Go e nmen o al educa ion expendi u es
0.038256000
0.025660700
1.4910
0.1480
Go e nmen consump ion expendi u es
0.004575030
0.002762040
1.6560
0.1097
Openness a io
−0.000925810
0.001187970
−0.7793
0.4428
G adua ed bachelo s uden s in sciences,
echnology and medicine_4
4.35001e-06
1.88375e-06
2.3090
0.0291
**
G adua ed mas e s uden s in sciences,
echnology and medicine_2
2.11347e-06
1.32463e-06
1.5960
0.0227
**
Sou ce: Own
113
The BACE me hod e u ned he ollowing esul s: Go e nmen consump ion expendi u es, he
sha e o he popula ion wi h a e ia y educa ion diploma and go e nmen spending on e ia y
educa ion obus ly and ma ginally co ela e wi h economic g ow h. The g oss capi al
o ma ion, unemploymen a e and numbe s o g adua ed bachelo s uden s obus ly
mode a ely co ela e wi h g ow h. Wha changed om he i s se o BACE modelling esul s
is ha he numbe o bachelo s uden s g adua ing om majo s linked o science, echnology
and medicine migh co ela e obus ly and ma ginally wi h economic g ow h. The hypo hesis
o no impac o Mas e ’s and Bachelo ’s science, echnologies and medicine s uden s on
economic g ow h is ejec ed as he numbe o s uden s in such majo s obus ly and posi i ely
co ela e wi h GDP g ow h pe capi a bo h in he ixed e ec s model and BACE models.
Tab. 4: BACE Models (70 models accep ed ou o 1024), dependen a iable: LOG GDP in
cu en PPP pe capi a
PIP
Mean
S d. de .
Cond.
mean
Cond. s d.
de
Cons an
1.000000
4.262602
0.179509
4.262602
0.179509
Go e nmen expendi u es on HE
0.999910
−0.120103
0.022934
−0.120113
0.022907
Employmen by indus y
b eakdowns
0.883652
−0.000154
0.000078
−0.000175
0.000058
G oss capi al o ma ion
0.625332
0.002577
0.002358
0.004121
0.001591
Unemploymen a e
0.570417
−0.002449
0.002511
−0.004294
0.001770
To al g adua ed bachelo s uden s
in sciences and echnology
0.532771
0.000000
0.000000
0.000001
0.000000
To al g adua ed mas e s uden s
in sciences and echnology
0.180553
0.000000
0.000000
0.000000
0.000000
G ow h o popula ion sha e
0.066751
−0.000000
0.000387
−0.000007
0.001498
Sou ce: Own
The hypo hesis o answe he hi d esea ch ques ion was s a ed as: The e is no signi ican
impac he impac o men and women g adua es (Bachelo ’s and Mas e ’s le el) on he
economic g ow h. We cons uc ed an OLS model and unde wen he same p ocedu e
(Table 5) wi h he R-squa ed o 0.834 and Du bin-Wa son s a is ic o 1.72. I is isible ha
bo h emale and male mas e and bachelo s uden s end o posi i ely in luence economic
g ow h, howe e , he coe icien s o men a e la ge in his model. These indings may poin
ou ha he e migh be a gende pay gap. I could be de i ed ha male and emale g adua ed
bachelo and mas e s uden s ha e a posi i e impac on economic g ow h. The esul s
ecei ed wi h he applica ion o BACE modelling o check he same hypo hesis a e
demons a ed in Table 6 and show ha he numbe o g adua ed bachelo male s uden s migh
ha e obus ma ginal co ela ion wi h economic g ow h. Howe e , he numbe o emale
g adua e s uden s, bo h educa ed a Bachelo ’s and Mas e ’s le els, demons a e insigni ican
co ela ion wi h economic g ow h.
114
Tab. 5: Model 3: Fixed-e ec s, Robus (HAC) s anda d e o s, dependen a iable: LOG
GDP in cu en PPP pe capi a
Coe icien
S d. e o
- a io
p- alue
Cons an
4.746010000
0.301638000
15.730
<0.0001
***
G oss capi al o ma ion
0.001393160
0.000674290
2.066
0.0489
**
Unemploymen a e
−0.006738990
0.002574910
−2.617
0.0146
**
G ow h o popula ion
0.000551393
0.000192559
2.864
0.0082
***
Go e nmen expendi u es on HE
−0.062667600
0.051709000
−1.212
0.2364
Openness a io
−0.001786610
0.001423370
−1.255
0.2206
Male G adua ed bachelo s uden s
1.19535e-06
5.54429e-07
−2.156
0.0405
**
Female G adua ed bachelo s uden s
1.21790e-06
5.84251e-07
−2.085
0.0471
**
Male G adua ed mas e s uden s
3.09333e-06
8.97448e-07
3.447
0.0019
***
Female G adua ed mas e s uden s
2.51806e-06
7.73175e-07
3.257
0.0031
***
Sou ce: Own
Tab. 6: BACE Models (237 models accep ed ou o 1094), dependen a iable: LOG GDP in
cu en PPP pe capi a
PIP
Mean
S d. de .
Cond.
mean
Cond.
s d. de
Cons an
1.000000
4.247664
0.185112
4.247664
0.185112
Go e nmen
expendi u es on HE
0.999934
−0.118010
0.023213
−0.118020
0.023194
Employmen by indus y
0.848613
−0.000150
0.000082
−0.000170
0.000059
G oss capi al o ma ion
0.602108
0.002442
0.002344
0.004055
0.001607
Unemploymen a e
0.600080
−0.002640
0.002556
−0.004400
0.001771
Male g adua ed bachelo
s uden s
0.412162
0
0
0
0
Male g adua ed mas e
s uden s
0.238469
0
0
0.000001
0
Female g adua ed mas e
s uden s
0.158910
0
0
0
0.000001
Female g adua ed
bachelo s uden s
0.133351
0
0
0
0
G ow h o popula ion
sha e
0.064747
−1e-06
0.000382
−1.3e-05
0.001499
Sou ce: Own
121
JAK UNIVERZITY URČUJÍ EKONOMICKÝ ROZVOJ VE 27 ČLENSKÝCH STÁTECH EU
Mnoho e opských zemí se současnos i společně zaměřuje na spojení nejmode nějších
echnologií a nejchy řejších myslí, aby se ypořádaly se sociálními, ekonomickými
a ekologickými záleži os mi a nalezly ud ži elnou o no áhu. Uni e zi y mohou pomoci
y áře hodné znalos i p o ako é ýz y a podpo o a hospodářské a sociální ino ace.
Ten o článek sh nuje důkazy exis ujících azeb mezi e ciá ním zdělá áním a ekonomickým
ůs em měřeným HDP na hla u a posky uje k an i a i ní hodnocení ěch o zá islos í
pomocí nyní ši oce k i izo ané Cobb-Douglaso y p odukční unkce p o y áření obyčejných
nejmenších č e ců (OLS) a Bayesiánského p ůmě o ání klasických odhadů (BACE)
ekonome ických modelů. Výsledky získané om o ýzkumu ukázaly, že ýdaje na e ciá ní
zdělá ání a poč y pře ážně absol en ů - mužů BA a MA ( echnologiích, ědách
a medicíně) obus ně a čás ečně ko elují s ekonomickým ůs em. Dobře ozložené in es ice
do oz oje e ciá ního zdělá ání STEM mohou po enciálně posíli pozi i ní dopad uni e zi
na ud ži elný hospodářský ůs .
WIE DIE UNIVERSITÄTEN DIE ÖKONOMISCHE ENTWICKLUNG
IN DEN 27 MITGLIEDSSTAATEN DER EU BESTIMMEN
Viele eu opäische S aa en konzen ie en sich de zei au die Ve bindung de mode ns en Technologien
und de klügs en Geis e , dami diese sich mi sozialen, ökonomischen und ökologischen
Angelegenhei en auseinande se zen und ein nachhal iges Gleichgewich inden. Die Uni e si ä en
können bei de Scha ung geeigne e Kenn nisse ü solche He aus o de ungen hel en und
wi scha liche und soziale Inno a ionen un e s ü zen. Diese A ikel ass die Beweise exis ie ende
Bindungen zwischen de e iä en Bildung und dem im HDP p o Kop gemessenen ökonomischen
Wachs um zusammen und lie e eine quan i a i e Bewe ung diese Abhängigkei en mi Hil e de
zu zei k i isie en Cobb-Douglas’schen P oduk ions unk ion ü die Bildung gewöhnliche kleins e
Quad a e (OLS) und de Bayes’schen Be echnung des Du chschni s klassische Schä zungen (BACE)
ökonome ische Modelle. Die in diese Un e suchung gewonnenen E gebnisse haben gezeig , dass die
Ausgaben ü die e iä e Bildung und die Anzahl de Absol en en (Männe mi BA- und MA-
Abschluss in Technologien, Wissenscha en Medizin) obus und eilweise mi dem ökonomischen
Wachs um ko elie en. Gu e eil e In es i ionen in die En wicklung de e iä en Bildung STEM
können die posi i e Auswi kung de Uni e si ä en au das nachhal ige ökonomische Wachs um
s ä ken.
W JAKI SPOSÓB UNIWERSYTETY DETERMINUJĄ ROZWÓJ GOSPODARCZY
W 27 PAŃSTWACH CZŁONKOWSKICH UE
Wiele k ajów eu opejskich skupia się obecnie wspólnie na łączeniu najnowocześniejszych echnologii
i najzdolniejszych umysłów w celu ozwiązania p oblemów społecznych, gospoda czych
i ekologicznych o az znalezienia wałej ównowagi. Uniwe sy e y mogą pomóc w gene owaniu
odpowiedniej wiedzy dla akich wyzwań i wspie ać innowacje gospoda cze i społeczne. Niniejszy
a ykuł podsumowuje dowody na is niejące powiązania między szkolnic wem wyższym a wz os em
gospoda czym mie zonym w PKB pe capi a i p zeds awia ilościową ocenę ych zależności,
wyko zys ując obecnie sze oko k y ykowaną unkcję Cobba-Douglasa do wo zenia zwykłych
najmniejszych kwad a ów (OLS) i Bayesowskiego uś edniania klasycznych oszacowań (BACE)
modeli ekonome ycznych. Wyniki uzyskane w amach p zep owadzonych badań pokazały, że
wyda ki na szkolnic wo wyższe o az liczba absolwen ów, p zeważnie płci męskiej, s udiów
licencjackich i magis e skich (w zak esie echnologii, nauk i medycyny) są mocno i częściowo
sko elowane ze wz os em gospoda czym. Dob ze ozłożone inwes ycje w ozwój ksz ałcenia
wyższego na kie unkach STEM mogą po encjalnie zwiększyć pozy ywny wpływ uniwe sy e ów na
z ównoważony wz os gospoda czy.