Soysal, Yasemin Nuhoḡlu; Bal a u, Roxana D.; Cebolla-Boado, Héc o
A icle — Published Ve sion
Me i oc acy o epu a ion? The ole o ankings in he
so ing o in e na ional s uden s ac oss uni e si ies
Globalisa ion, Socie ies and Educa ion
P o ided in Coope a ion wi h:
WZB Be lin Social Science Cen e
Sugges ed Ci a ion: Soysal, Yasemin Nuhoḡlu; Bal a u, Roxana D.; Cebolla-Boado, Héc o (2024) :
Me i oc acy o epu a ion? The ole o ankings in he so ing o in e na ional s uden s ac oss
uni e si ies, Globalisa ion, Socie ies and Educa ion, ISSN 1476-7732, Taylo & F ancis, London, Vol.
22, Iss. 2, pp. 252-263,
h ps://doi.o g/10.1080/14767724.2022.2070131
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/265088.2
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 p://c ea i ecommons.o g/licenses/by/4.0/
Me i oc acy o epu a ion? The ole o ankings in he so ing o
in e na ional s uden s ac oss uni e si ies
Yasemin Nuhog
lu Soysal
a,b
, Roxana D. Bal a u
c
and Héc o Cebolla-Boado
d
a
Depa men o Sociology, Uni e si y o Essex, Colches e , UK;
b
WZB, Be lin Social Science Cen e , Be lin, Ge many;
c
Uni e si y o Wa wick, Co en y, UK;
d
UNED, Mad id, Spain
ABSTRACT
Uni e si y ankings ha e gained p ominence in andem wi h he global ace
owa ds excellence and as pa o he g owing expec a ion o a ional,
scien ific e alua ion o pe o mance ac oss a ange o ins i u ional sec o s
and human ac i i y. While hei omnip esence is acknowledged,
empi ically we know less abou whe he and how ankings ma e in
highe educa ion ou comes. Do uni e si y ankings, p edica ed on
uni e salis ic s anda ds and sha ed me ics o quali y, unc ion
me i oc a ically o le el he impac o long-es ablished epu a ions? We
add ess his ques ion by analysing he ex en o which changes in he
posi ion o UK uni e si ies in anking ables, beyond exis ing epu a ions,
impac on hei s a egic goal o in e na ional s uden ec ui men . We
d aw upon an ad hoc da ase me ging agg ega e (uni e si y) le el
indica o s o anking pe o mance and epu a ion wi h indica o s o o he
ins i u ional cha ac e is ics and in e na ional s uden numbe s. Ou
findings show ha ec ui men o in e na ional s uden s is p ima ily
de e mined by uni e si y epu a ion, socially media ed and sedimen ed
o e he long e m, a he han uni e si ies’yea ly upda ed anking
posi ions. We conclude ha while he e is insufficien e idence ha
imp o ing ankings changes uni e si ies’in e na ional ec ui men
ou comes, hey a e ne e heless consequen ial o uni e si ies and
s uden s as s a egic ac o s in es ing in ankings as pu pose and iden i y.
ARTICLE HISTORY
Recei ed 24 Ap il 2021
Accep ed 20 Ap il 2022
KEYWORDS
Uni e si y ankings;
epu a ional p es ige;
quan ifica ion and me ics;
UK highe educa ion;
in e na ional s uden
ec ui men ; ins i u ional
le el analysis
Uni e si y ankings a e now ubiqui ous in he inc easingly expanding highe educa ion (HE) field.
1
They a e pa o he b oade mo e owa d quan i a i e benchma king and e alua ion o pe o m-
ance ac oss a ange o ins i u ional sec o s and human ac i i y (Be man and Hi schman 2018).
Accoun ing o excellence is now a widely adop ed p ac ice in e e y o ganisa ional field, public
o p i a e. A g owing li e a u e, o igina ing in No h Ame ican and Eu opean schola ship, has
paid bo h heo e ical and empi ical a en ion o ankings (See Lamon 2012 o a e iew). Uni e -
si y ankings as a esea ch opic has also a ac ed a en ion in ecen yea s, mos o he empi ical
wo k ha ing been done in he US and ocusing on how ankings change uni e si y o ganisa ional
beha iou , adminis a ion, college admission policies and s uden choices (Espeland and Saude
2016; Monks and Eh enbe g 1999). In Eu ope, exis ing empi ical s udies a e mos ly desc ip i e,
ei he ocusing on he sho comings o he exis ing ankings and ways o imp o e hem o aking
a c i ical iew o highligh mo e gene ally hei biases and conflic wi h he undamen al missions o
HE (Al bach 2012; Amsle and Bolsmann 2012; Dill and Soo 2005; Pusse and Ma ginson 2013;
Shin, Tou koushian, and Teichle 2011; Taylo and B addock 2007). While hei pe asi eness in
© 2022 The Au ho (s). Published by In o ma UK Limi ed, ading as Taylo & F ancis G oup
This is an Open Access a icledis ibu ed unde he e ms o he C ea i e Commons A ibu ion 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 he o iginal wo k is p ope ly ci ed.
CONTACT Yasemin Nuhog
lu Soysal [email p o ec ed] WZB, Be lin Social Science Cen e , Be lin, Ge many
GLOBALISATION, SOCIETIES AND EDUCATION
2024, VOL. 22, NO. 2, 252–263
h ps://doi.o g/10.1080/14767724.2022.2070131
he Eu opean con ex is now widely commen ed upon, sys ema ic s udies abou whe he , and how,
uni e si y ankings ma e in specific ins i u ional ou comes a e missing.
In his pape , we unde ake an empi ical analysis o he ela ionship be ween ankings and he
in e na ionalisa ion o UK uni e si ies, mo e specifically he in e na ionalisa ion o hei s uden
body. Apa om i s po en ial o gene a ing e enue, he in e na ionalisa ion o he s uden
body is highly in eg a ed in o he idea o a wo ld-class uni e si y. A ac ing s uden s om ab oad,
o sending hei own ab oad, is widely adop ed as a p io i y and explici ly s a egized by go e n-
men s, e ia y educa ion sec o s and ins i u ions alike (Buckne 2019; IAU 2019).
Wha is he ole o ankings in shaping he so ing o in e na ional s uden s ac oss uni e si ies?
Unlike long eng ained, socially media ed epu a ions, uni e si y ankings a e designed o measu e
‘objec i e’changes in quali y o e ime based on sha ed and s anda dised me ics. Thus, shi s in
ankings should impac on s uden ec ui men o e and beyond long-exis ing epu a ions. To
in es iga e, we use agg ega e le el ins i u ional da a om 88 UK uni e si ies obse ed longi udin-
ally and analyse he changes o e ime in he s ock o in e na ional s uden s in ela ion o changes in
ankings. UK is he second mos popula des ina ion, a e he US, in he wo ld o in e na ional
s uden s (Hubble and Bol on 2021). In 2019/20 in e na ional s uden s (excluding he ones om
he Eu opean Union) cons i u ed 22 pe cen o o al s uden popula ion s udying a he UK uni e -
si ies, hei numbe s ebled since he u n o he wen y-fi s cen u y. Using o ganisa ional le el
and longi udinal da a, and con olling o se e al uni e si y cha ac e is ics, we analyse he pe o m-
ance o change in ankings e sus long-held epu a ion in p edic ing he a ia ion o e ime in
in e na ional s uden en olmen ac oss uni e si ies.
How do uni e si y ankings wo k empi ically?
The po ency o ankings as a ool o e alua ion has been highligh ed ex ensi ely in he inc easingly
impo an field o sociology o e alua ion (Espeland and S e ens 2008; Fou cade and Healy 2017;
Lamon 2012). The b oade li e a u e emphasises he ise o ‘new public managemen ’in a neolib-
e al en i onmen , esul ing in amewo ks and ins umen s o e alua ion apidly diffusing om p i-
a e o public sec o s. In dealing wi h de egula ion and e e g ea e budge cu s, go e nmen s and
public ins i u ions u n o he ools o new public managemen o ensu e g ea e efficacy, hus blu -
ing he bounda ies be ween ma ke and non-ma ke e alua i e schemes.
All e alua i e schemes a e based on classifica ions, and quan ifiable indica o s ha e become he
‘g amma ’unde lying such classifica ion p ac ices (Be man and Hi schman 2018; Espeland and S e-
ens 2008;Fou cade2016;Mau2019). Quan ified ep esen a ions o m pa o assessmen s o pe -
o mance, impac and isk in fields as wide- anging as heal h, he economy, finance, educa ion,
human igh s, co up ion, he en i onmen , ood and leisu e (Ke in e al. 2012). In he ealm o
HE go e nance, he cul u e o nume ical e alua ion mani es s i sel as uni e si y ankings bu is
also p esen in quali y assessmen o a ious kinds, such as he Resea ch E alua ion F amewo k in
he UK and academic annual e iews and p omo ions in he US and o he pa s o he wo ld (Rami ez
2013). Uni e si y ankings a e based on a numbe o a ionalised, quan ifiable pe o mance me ics
(based on algo i hms and weigh s) which ac as ‘p oxies o quali y’. These include unding and ci a-
ions ( esea ch excellence), en y le els (s uden selec i i y), good honou s and employabili y (s uden
ou comes), s uden /s aff a io ( eaching en i onmen ), and budge spen on academic se ices and
acili ies (quali y o in as uc u e). Al hough diffe en weigh s a e gi en o diffe en indica o s ac oss
he exis ing anking schemes, he main ca ego ies do no change significan ly – esea ch indica o s o
example ha e a significan place in all o hem –affi ming a con e gen model o uni e si y excellence
(Hazelko n 2007, 3; Hazelko n 2014;Ushe andSa ino2006). Bu mo e ema kably, like o he edu-
ca ional indica o s ha unde line global e alua ion schemes (p omulga ed by he UN, UNESCO and
OECD), hese me ics a e pe cei ed as ep esen ing uni e salis ic s anda ds. They a e expec ed o be
applicable independen o local and na ional con ex s; ac oss HE sys ems, old o new; and HE ins i-
u ions la ge, small, p i a e and public.
GLOBALISATION, SOCIETIES AND EDUCATION 253
The uni e salis ic p ojec ion o s anda ds diffe en ia es uni e si y ankings om o he e alua-
i e schemes ha ha e been ex ensi ely s udied in he sociological li e a u e, such as he ag o-
ood sec o , pa icula ly he wine indus y. In he wine indus y, quali y con en ions (measu es
mobilised o e alua e quali y) ake in o conside a ion bo h p oduc and p oduce ca ego ies,
hus he e is plu ali y in he logics o e alua ion. Diffe en classifica ion con en ions apply o
la ge indus ial es a es and c a sman wine ies (Diaz-Bone 2013). Fo he p oduc , quali y is linked
wi h e i o ial and empo al ancho age: a ce ain egion, a ce ain g ape, a ce ain ime pe iod c e-
a es he dis inc ion in quali y (Chau in 2009). T adi ion plays a ole.
HE ins i u ions long had epu a ions based on adi ion, age, and e en e i o y, which landed
hem a op o he p es ige hie a chy. I y League, Oxb idge, Russel G oup a e sho cu e minology
o such epu a ions. Uni e si y ankings a e supposed o go beyond such epu a ion-based ca eg-
o iza ions and hie a chies by edis ibu ing p es ige based on uni e salis ic s anda ds and sha ed
me ics o quali y. They ope a ionalise p es ige as a me ic sys em in ela ion o ‘objec i e’pe o m-
ance, no as long-held, o en socially media ed, pe cep ions abou he quali y o he ins i u ion. They
a e no o bes ow dis inc ion based on adi ion, ounda ion yea o any o he pa icula is ic
c i e ia.
2
The diffe en ial alloca ion o ins i u ional s a us ia ankings is hus pe cei ed as a me i oc a ic
p ocess based on demons able ou comes, which can be enhanced by s a egic o ganisa ional
decisions and in es men s (Bal a u 2018; Kei h 2001). Rankings a e expec ed o signal imp o e-
men s in quali y and in o m and influence a a ie y o uni e si ies’s akeholde s (e.g. applican s,
s uden s, employe s, unde s, o he uni e si ies, minis ies), who should be esponsi e o s a us
changes as eflec ed in ankings.
To wha ex en do uni e si y ankings deli e on hei p omise o me i oc a ic mapping o qual-
i y and ela ed s a us change? We in es iga e his ques ion by examining whe he changes in ank-
ing pe o mance impac on uni e si ies’ ec ui men o in e na ional s uden s. In e na ional
s uden flows, mo e so han local s uden s, a e expec ed o be awa e o and sensi i e o ankings
as hey a e he p ima y sou ce o compa a i e in o ma ion on highe educa ion ins i u ions beyond
he na ional con ex (Mazza ol and Sou a 2002; Hazelko n 2014). I indeed ankings dis ibu e
p es ige me i oc a ically, we should find ha shi s in ankings, as opposed o long-ancho ed ank-
ing epu a ions, a e consequen ial o in e na ional ec ui men .
Uni e si y ankings, epu a ion and so ing o in e na ional s uden flows
The ela ionship be ween ankings and epu a ion has no been explici ly heo ised in he li e a u e.
Bou dieu has a p ominen place in s udies on uni e si y ankings, and ankings in gene al (Heffe -
nan 2022). F om a Bou dieusian pe spec i e, ankings a e means o econfigu ing epu a ion in
uni e si ies’s uggle o gaining symbolic capi al, and ela edly o he o ms o capi al. Uni e si ies
ha a e success ul a he ‘ anking game’a e expec ed o a ac bes academics and s uden s, win
mos esea ch unding and cha ge highe ui ion. Ye , uni e si ies do no en e he anking com-
pe i ion om equal places; hey ca y he ‘cumula ed sco es’o he pas , as eflec ed in long-
ancho ed epu a ions (B osnan 2017). S i ing o dis inc ion and ela i e ad an age, po en ial s u-
den s and hei amilies pay a en ion o ins i u ional epu a ion o benefi om uni e si y’s sym-
bolic capi al and o be e hei social posi ion down he oad. Ma ginson (2014) a gues ha , as a
p oxy o epu a ion, ankings ha e become in eg al pa o ‘s a us cul u e’.
Rankings also figu e hea ily in o he s udies o highe educa ion om a neo-ins i u ionalis pe -
spec i e. A pa icula a gumen has been he cul u al ole o ankings in cons i u ing s a egic and
compe i i e ac o s and iden i ies a bo h o ganisa ional and indi idual le els (Saude and Espeland
2009; Hasse and K ucken 2013; Rami ez 2013). A s anda dised concep ion o highe educa ion s u-
den s ollows: hose who pu sue he bes educa ional e u ns and who can easily mo e and lea n
ac oss HE ins i u ions and sys ems. The p ojec ed ac o hood o he uni e si y and i s s uden s
boos s he global p ominence o ankings and hei wide-sp ead cu ency e en in highly
254 Y. N. SOYSAL ET AL.
esou ce-limi ed highe educa ion con ex s. The neo-ins i u ionalis schola s, howe e , also s ess
ins i u ional pa h dependency and acknowledge he epu a ional s uc u e o HE (Rami ez and Tip-
lic 2014). Once achie ed, uni e si y p es ige may exe cise long e m influence on esou ces and
legi ima ion. Thus, while s a ing om diffe en an age poin s, bo h Bou desian and neo-ins i u-
ionalis schola ship sugges ha ins i u ional epu a ion can ha e independen and ino dina e
influence on uni e si y ou comes such as a ac ing in e na ional s uden s –an indica o o p es-
ige i sel .
Empi ically, sys ema ic s udies in o he ela ionship be ween ankings, epu a ion and s uden
ec ui men a e sca ce. Resea ch conduc ed in he US shows ha shi s in an ins i u ion’s pos-
i ion in ankings can cause changes in he numbe and cha ac e is ics o domes ic applican s
(Bowman and Bas edo 2009; Me edi h 2004; Monks and Eh enbe g 1999). In he UK, esea ch
ound epu a ional signals o be impo an in decisions o s udy ab oad (Findlay e al. 2010) and
anking eme ge as a significan d i e o in e na ional s uden s’decision-making in choosing
uni e si ies based on analysis o he Highe Expec a ions Su ey (Cebolla-Boado, Hu, and
Nuhog
lu Soysal 2017). Sou o-O e o and Ende s (2017), using mic o-le el da a om he In e -
na ional S uden Ba ome e su ey, conclude ha epu a ion (along wi h o he ac o s such as
ee le els and quali y o eaching) is mo e impo an han ankings in in o ming and influencing
s uden s, he po en ial use s o ankings.
Apa om he la e , hese s udies do no explici ly diffe en ia e be ween anking posi ion and
epu a ion, which is ou analy ical in e es he e. Fu he mo e, hey a e mainly conduc ed om he
lens o indi idual s uden s, as he ‘consume s’o ankings, ocusing on how ankings shape s uden
decision-making, along wi h o he ac o s such as he pe sonal en i onmen o s uden s. We sup-
plemen his li e a u e mo ing o an o ganisa ional le el analysis, inqui ing whe he agg ega e le el
in e na ional s uden flows con inue o be so ed and s a ified ac oss uni e si ies by his o ically
consolida ed ins i u ional epu a ions a he han by changes in ankings as sugges ed by he me -
i oc a ic a ionale.
Analysis
Da a and a iables
Ou da a a e a he o ganisa ional le el, and we analyse he ex en o which uni e si y anking pos-
i ion in con as o epu a ion p edic s he a ia ion o e ime in in e na ional s uden ec ui men .
We se ou dependen a iable as he numbe o in e na ional, non-EU s uden s en olled in B i ish
uni e si ies, using he da a p o ided yea ly by he Highe Educa ion S a is ics Agency (HESA). Fo
ou analysis, we cons uc ed an ad-hoc longi udinal da ase con aining uni e si y le el da a o ou
da a poin s: 2010/2011, 2014/2015, 2016/2017, and 2018/19. These ou da a poin s ep esen he
a ailable HESA ables p o iding longi udinal and compa able da a ac oss all ou a iables o
in e es .
Ou analy ical s a egy equi es ope a ionalisa ion o ankings posi ion and epu a ion, ou key
p edic o s, as dis inc a iables om each o he . While ankings a e expec ed o be esponsi e o
yea ly changes in pe o mance, epu a ional p es ige, once o med, has an ancho ing effec , whe he
i is s ill dese ed o no , and hus has long- e m pe sis ence (Bal a u 2018; Bowman and Bas edo
2011; Kei h and Babchuk 1998; Schul z, Mou i sen, and Gab ielsen 2001; Taylo and B addock
2007; see also Fine 2012). To cap u e hese diffe ences, we ely on da a om wo sou ces: da a
on uni e si y ankings om he Comple e Uni e si y Guide (CUG) and da a on ins i u ional epu-
a ion om his o ical anking ables published by The Times and a ailable a he Times Digi al
A chi e. We chose CUG da a, published since 2007, o e o he a ailable ankings (such as The
Gua dian,Times Highe Educa ion) o h ee easons: i s longi udinal ankings da a a e publicly
accessible; i has consis en ly a ailable da a on all UK uni e si ies; and impo an ly i does no
include subjec i e assessmen s o uni e si y epu a ion as one o i s anking c i e ia.
3
The Times
GLOBALISATION, SOCIETIES AND EDUCATION 255
es ablished he fi s anking ables, as ea ly as 1993; we use hem o cap u e he ini ial o ma ion o
epu a ion abou he UK’s‘ op’uni e si ies.
4
We a e awa e ha some ankings, such as Times
Highe Educa ion o example, include in e na ional ou look (in e ms o s uden and s affnum-
be s) as a c i e ion; his is no he case o CUG o he ini ial pe iod Times ankings, hus we do
no isk mul icollinea i y wi h ou independen a iables.
We cons uc ed he a iables o anking and epu a ion as ollows. The pe cen age change in
uni e si y ank, om 2010/2011 ( he ea lies publicly a ailable CUG da a) o 2018/2019 ( he
mos ecen c oss-sec ion in ou da a ac oss all a iables o in e es ), ep esen s how much he uni-
e si y ‘objec i ely’imp o ed i s posi ion in ankings be ween hese wo ime poin s. We ha e used
he wides a ailable measu e o change in ankings, om 2010 o 2018, as opposed o yea ly changes
o allow as much a ia ion as possible in he changes. In con as , epu a ional p es ige is ime in a -
ian and ope a ionalised as a bina y a iable on he basis o (a) hose uni e si ies which consis en ly
anked in he Top 20 in he ea ly yea s (1993, 1994, 1995 and 1996) o The Times ables, and (b) he
uni e si ies ha a e among he ounde s o he Russel G oup in he 1990s o joined sho ly a e .
5
These uni e si ies we e coded ‘1’as opposed o all o he uni e si ies, which we e coded ‘0’. The pe -
cen age change in uni e si y ank on he one hand, and epu a ion on he o he , allow us o diffe -
en ia e be ween he effec s o uni e si ies’inc emen al successes in he anking ables and b oade
pe cep ions o uni e si ies’p e iously buil epu a ions.
As con ol a iables, we in oduce se e al uni e si y le el cha ac e is ics (da a also om HESA)
ha migh bols e he numbe o in e na ional s uden s. Whe e applicable, we ope a ionalise hese
indica o s as p opo ions (in o al s uden s o in o al income) o ne ou he effec o uni e si y size.
Fi s ly, ha ing an in e na ional ou look may inc ease uni e si ies’ isibili y and appeal o p o-
spec i e in e na ional s uden s. We ha e wo indica o s o ope a ionalise: he in e na ional esea ch
unding (Eu opean Union and non-Eu opean Union) as a p opo ion o o al unding om esea ch
g an s and con ac s (measu ed in housands o pounds) and he Eu opean Union (EU) s uden s as a
p opo ion o o al s uden s.
Secondly, uni e si ies inc easingly in es in s uden -o ien ed se ices o a ac and ca e o
in e na ional s uden s. Uni e si ies’se ice o ien a ion has been ope a ionalised as he expendi-
u e spen on s uden and s aff acili ies as a p opo ion o o al expendi u e (measu ed in hou-
sands o pounds). These acili ies include ca ee s ad iso y se ices, g an s o s uden socie ies,
accommoda ion offices, a hle ic and spo ing acili ies, anspo , chaplaincy, s uden counsel-
ling and heal h se ices. Whils expendi u e on s uden acili ies is included as a c i e ion in
he CUG ankings, we ha e no iden ified mul icollinea i y be ween he wo independen
a iables.
Thi dly, we include con ols o income om ui ion ees and educa ion con ac s as a p opo ion
o o al income. Rec ui men o in e na ional s uden s is o en seen as a su i al s a egy in he con-
ex o an inc easingly de egula ed and unde unded HE sec o , especially since hey pay highe ees
compa ed o home s uden s. I a uni e si y’s budge hea ily depends on ui ion ees, he uni e si y
is likely o be mo e eage o accep in e na ional s uden s. Some uni e si ies may e en se hei en y
s anda ds low o inc ease hei in e na ional in ake. We do no include en y s anda ds in he con-
ols, as hey a e among he c i e ia ha CUG uses in calcula ing he uni e si y ank and he wo
a iables co ela e highly (ρ> .80). The a iable is measu ed in housands pounds and includes all
income om s uden ees om indi idual s uden s as well as o he sou ces such as he S uden
Loans Company, Local Educa ion Au ho i ies, he S uden Awa ds Agency o Sco land, and he
Depa men o Educa ion o No he n I eland.
And las ly, we include he o al numbe o s uden s o con ol o uni e si y size, as well as he
numbe o in e na ional s uden s he p e ious yea . La ge bodies o in e na ional s uden s may
help aise he p ofile and bols e he epu a ion o he uni e si ies in e na ionally.
All independen and con ol a iables (excep epu a ion, a ime in a ian a iable) we e
measu ed longi udinally, in line wi h he dependen a iables. Table A1 in he Appendix p o ides
he desc ip i e s a is ics.
256 Y. N. SOYSAL ET AL.
Me hod
We an wo models, he second one including he lagged dependen a iable (T−1) as a p edic o
o model ine ia as p oxied by he p e ious le els o in e na ional s uden en olmen s. The da a is
based on 88 uni e si ies (app oxima ely 70% o he UK uni e si y sec o ), obse ed a ou ou
poin s in ime. These a e all uni e si ies wi h a ailable longi udinal da a ac oss all ou a iables
o in e es .
6
The sample includes a ious uni e si ies, la ge and small, mo e and less in e na ional
in e ms o he s uden body, and o diffe en anks and financial esou ces, as illus a ed in Table
A1 in he Appendix.
We op ed o modelling o ime se ies a e confi ming ha he clus e ing o obse a ions o e
ime is s a is ically ele an .
7
Reg ession models wi h andom cons an effec s decompose he o e -
all a ia ion wi hin (c oss- ime effec ) as well as be ween yea s o obse a ion.
Ou model specifica ion includes bo h ime a ian p edic o s ( he pe cen age change in anking
posi ion, numbe o s uden s, expendi u e) and ou ime in a ian p edic o : epu a ion. The
inclusion o lagged independen a iables helps us p edic whe he changes in uni e si y le el
cha ac e is ics nu u e o inhibi he numbe o in e na ional s uden s, o e and abo e uni e si y
epu a ion. We addi ionally con ol o ime (by including dummy a iables o each yea ) o
accoun o he o e all g ow h in in e na ional s uden s o e he pe iod. The specifica ion o ou
model is:
Yi =
a
i+ui( −1) +
b
1Xi( −1) +
e
i( −1)
whe e α
i
is he numbe o in e na ional s uden s in each uni e si y, which is adjus ed yea ly by a
esidual called u
i( -1).
The slope modelling he effec o ou p edic o s and he uni e si y le el
esidual ɛ
i( -1)
is in e p e ed as in a s anda d eg ession model.
Las bu no leas , ou VIF analysis confi ms ha he e is no conce n o mul icollinea i y (VIF <
2.5).
Findings
The esul s om he model p edic ing he numbe o in e na ional s uden s a e shown in Figu e 1
( he able epo ing exac es ima es is p esen ed in he Appendix A2). The figu e shows he ma ginal
Figu e 1. A e age ma ginal effec s p edic ing s anda dised in e na ional s uden s.
GLOBALISATION, SOCIETIES AND EDUCATION 257
effec s a he 95 pe cen confidence in e al in each case. We find ha he epu a ion o he uni-
e si y is he main ac o boos ing he numbe s o in e na ional s uden s. By con as , imp o e-
men s in anking ables as measu ed in pe cen changes in able posi ion ha e li le o no
impac ( he size o he co esponding es ima e is ai ly small in bo h models and s a is ically
insignifican a he 5% le el). This confi ms ha he d i e o uni e si y in e na ionalisa ion as
measu ed by he numbe o in e na ional s uden s is long- e m consolida ed epu a ional p es ige
and no ou al e na i e sho - e m sensi i e p oxy (shi in anking posi ion).
Ou con ols appea o ha e a limi ed ole in p edic ing he s ock o in e na ional s uden s ac oss
ime. In Model 1, we see ha in e na ional s uden s a e especially concen a ed in la ge uni e si ies,
wi h a la ge sha e o EU s uden s and who a e mo e elian on ui ion ees. Howe e , uni e si y size
and ui ion ees a e no longe s a is ically significan in Model 2, when con olling o he p e ious
le el o in e na ional s uden s. This is expec ed, as he s ock o in e na ional s uden s has di ec
implica ions in e ms o he size o he s uden body and he magni ude o ui ion ee income.
The sha e o EU s uden s emains a posi i e and s a is ically significan p edic o in bo h models.
I is possible ha he p esence o EU s uden s plays a ole in defining a uni e si y’s in e na ional
ou look, hus making i an a ac i e des ina ion o he non-EU, in e na ional s uden s oo.
Expendi u e on s uden acili ies, which uni e si ies in es in s a egically, as hey a e hough o
as d aws o in e na ional s uden s and s uden s in gene al, does no ha e any consis en effec .
Nei he do in e na ional esea ch unds.
Finally, a e in oducing in ou models he numbe o in e na ional s uden s in he p e ious
yea , i becomes clea ha a ce ain so o Ma hew-effec applies in a ac ing in e na ional s u-
den s. Those uni e si ies which al eady a ac ed mo e in e na ional s uden s in he pas , cu en ly
also a ac mo e o hem: uni e si ies b and hemsel es ‘in e na ionalized’.
Fu he checks
We u he explo ed al e na i e ope a ionalisa ions o anking posi ion o iden i y whe he ank-
ings become a significan p edic o unde a diffe en model specifica ion. Fi s ly, we ha e ope -
a ionalised he changes in ankings based on he in e na ional QS Rankings. This
ope a ionalisa ion accoun s o he possibili y ha in e na ional s uden s consul in e na ional
ankings mo e han na ional ankings. We ha e no used QS Rankings in he p ima y model
because he sample size is limi ed; in e na ional ankings a e a ailable only o abou hal o
ou cu en sample. We aim o explain a ia ion in in e na ional s uden numbe s ac oss a
ep esen a i e sample o UK uni e si ies, no only hose who a e included in he in e na ional
ankings. Ne e heless, ou esul s we e eplica ed by e- unning he model on he limi ed
sample size affo ded by in e na ional ankings (N= 46) as ollows: uni e si y p es ige was posi-
i e and s a is ically significan , while changes in QS Rankings we e no a s a is ically significan
p edic o o in e na ional s uden s (p> .05).
Secondly, we added a bina y indica o o cap u e uni e si ies whose anking change pushed
hem in o ‘Top 10’ca ego y. Being labelled as ‘Top Tie ’migh ca y mo e weigh han gene ally
mo ing up in he ankings (Me edi h 2004). Fo hese uni e si ies he anking change was a guably
s a is ically significan , i.e. a he 10% le el (p= .051), along epu a ion, as opposed o al e na i e
ope a ionalisa ions (e.g. Top 25) which we e insignifican . The pe sis en effec o uni e si y epu-
a ion when con olling o he mos subs an ial change in uni e si y ankings (en e ing he Top
10), confi ms he impo an quali a i e diffe ence be ween uni e si y epu a ion and ankings.
We find ha epu a ion emains he s onges p edic o o in e na ional s uden numbe s, com-
pa ed wi h a ied ope a ionalisa ion o change in ankings.
Fu he mo e, we also un he models wi h an indica o cap u ing a mo e di ec , s a egic a emp
o signalling in e na ionalisa ion: ha ing a P o-Vice Chancello (PVC) In e na ional (coded ‘1’), as
opposed o no ha ing one (coded ‘0’). We ha e no included his indica o in he main analysis due
258 Y. N. SOYSAL ET AL.
o he limi ed a ailabili y; he indica o is a ailable only o 2017. The esul s showed no s a is ically
significan effec o ha ing a PVC In e na ional on he numbe o in e na ional s uden s.
Finally, o disca d he possibili y ha he age o uni e si y is he main ca alys o epu a ion, we
es ima ed ea men effec s models se ing epu a ion as he ea men a iable. Repu a ion in hese
models is endogenized using ounda ion yea . Ou esul s show ha epu a ion emains a signifi-
can p edic o o ou dependen a iable (numbe o in e na ional s uden s) beyond he ime since
ounda ion.
Discussion
Empi ically ocusing on in e na ional s uden ec ui men , ou findings poin o a a he diffuse and
symbolic unc ioning o uni e si y ankings. We show ha in e na ional s uden s a e so ed ac oss
B i ish uni e si ies based on long-ancho ed ins i u ional epu a ions, mo e so han on he basis o
objec i e, p ecise indica o s o quali y con eyed by changing ins i u ional posi ion in anking
ables. Uni e si ies’ epu a ional p es ige, as opposed o hei o ganisa ional pe o mance as
measu ed in ankings, is he main de e minan o hei ec ui men success. This con as s sha ply
wi h he pe cep ion ha effo s o boos yea ly changes in ankings a e consequen ial, a iew ha
unde pins significan changes in he go e nance and o ganisa ion o uni e si ies. Rankings ail o
deli e hei me i oc a ic p omise; i any hing, hey help o u he legi ima e epu a ion as a sym-
bolic good. Ins i u ions long epu ed o be a he op o he ‘p es ige hie a chy’benefi bo h om
socially media ed pe cep ions o epu a ion and om ankings ha ein o ce he symbolic alue o
epu a ion.
I his is so, why do HE ins i u ions and he sec o con inue o engage wi h ankings? In ela ion
o in e na ional s uden mobili y, ankings’pe sis en p esence in HE is bu essed by widely
adhe ed human capi al heo ies, which p opaga e a ision o educa ion as an in es men wi h
na ional and indi idual le el e u ns. This is he p edominan amewo k ha uni e si ies and pol-
icy make s use o make sense o in e na ional s uden mobili y. Sha e o in e na ional s uden s is
iewed as a ba ome e o uni e si ies’abili y o effec i ely gene a e human capi al and global alen
by pu suing ‘wo ld class educa ion’. Concomi an ly, in e na ional s uden s a e ega ded as a ional
e alua o s o uni e si ies’pe o mance, elying on ankings in o de o compa e he ‘uni e si y
ma ke ’and disce n he wo hies uni e si ies a a pa icula poin in ime. Rankings echnology,
and i s communica ion, makes possible compa ison o ins i u ions ha a e o he wise un-compa -
able on na ional and wo ld scales, while also p o iding a common defini ion o he mos aluable
and wo hies educa ion o in es in (Espeland and Saude 2012; Rami ez 2013).
Mo e p o oundly, ankings help pe pe ua e cogni i e sel -cons uc ions o uni e si ies and s u-
den s as a ional, goal-o ien ed agen s (Hasse and K ucken 2013; Ma celo and Powell 2020). Uni-
e si y ankings hold sway no only because hey p ojec uni e sal quan ifiable and compa able
s anda ds o wo h, bu also because in es men in hem con eys pu pose ul and capable ac o hood
wi h aspi a ions o mobili y. Fo p ospec i e s uden s and g adua es, pu suing ankings ac s as a
p oxy o signal hei p oac i e exe cise o agency. F om uni e si ies poin o iew, imp o ing ank-
ings is a s a egic ac ion, and in pa seeking legi imacy in he inc easingly s anda dised and ans-
na ionalized sphe e o highe educa ion. In o he wo ds, ankings become HE pu pose and iden i y
in hemsel es, decoupled om he e y goal o enhancing educa ional excellence and in e na iona-
lisa ion (Bal a u and Soysal Nuhog
lu 2018; B omley and Powell 2012; Espeland and Saude 2012).
Uni e si ies explici ly in eg a e imp o ing s andings in ankings in o hei mission s a emen s, as
an indica o o hei ambi ion o excellence. The wide na a i es and publica ion o ankings
in he media makes i difficul o uni e si ies o igno e hese lis ings. Small diffe ences a e mag-
nified, and ela i e posi ioning is celeb a ed (B anko ic, Ringel, and We on 2018; Fowles, F ede -
ickson, and Koppell 2016), despi e he e being li le empi ical e idence on hei impac on
o ganisa ional ou comes. An annual cycle o key pe o mance indica o s (KPI) akes hold (e.g.
inc eased en olmen s, inc eased g adua ion a es, inc eased esea ch unding, inc eased ci a ions,
GLOBALISATION, SOCIETIES AND EDUCATION 259
