Foun a, Kons an ina; Kollias, Ch ēs os; Tze emes, Panayio is
A icle
"You don' pay you bills you ge no p o ec ion": A T ump
e ec on NATO membe s' mili a y expendi u es?
Peace Economics, Peace Science and Public Policy (PEPS)
P o ided in Coope a ion wi h:
De G uy e B ill
Sugges ed Ci a ion: Foun a, Kons an ina; Kollias, Ch ēs os; Tze emes, Panayio is (2025) : "You don'
pay you bills you ge no p o ec ion": A T ump e ec on NATO membe s' mili a y expendi u es?,
Peace Economics, Peace Science and Public Policy (PEPS), ISSN 1554-8597, De G uy e , Be lin, Vol.
31, Iss. 2, pp. 145-160,
h ps://doi.o g/10.1515/peps-2025-0024
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/333341
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/
Kons an ina Foun a, Ch is os Kollias* and Panayio is Tze emes
“You don’ Pay You Bills You Ge No
P o ec ion”: A T ump Effec on NATO
Membe s’Mili a y Expendi u es?
h ps://doi.o g/10.1515/peps-2025-0024
Recei ed Ma ch 18, 2025; accep ed Ma ch 28, 2025; published online Ap il 16, 2025
Abs ac : Dispensing wi h he cus oma y cou eous diploma ic language, much o
he 47 h P esiden ’s he o ic conce ning he o he NATO Allies echoes he alid US
c i icism ha many o hem ee ide on he US p o ec i e de ence umb ella wi hou
con ibu ing hei ai sha e o he cos s o NATO’s collec i e de ence. “You don’ pay
you bills, you ge no p o ec ion…”, he is quo ed saying in a p e-elec ion speech while
is his Da os 2025 add ess, he men ioned, e e ing o he 2 % NATO guideline and he
effec o his fi s p esiden ial e m, ha “…mos na ions didn’ pay un il I came
along”. Using such s a emen s as an impe us, he pape se s ou o examine whe he a
T ump effec o his fi s p esidency on NATO membe s’mili a y spending is
empi ically aceable. Resul s epo ed he ein seem o offe e idence in a o o a
en a i e T ump effec in e ence.
Keywo ds: NATO; mili a y spending; T ump; MDS; K-Means; s uc u al b eaks
1 In oduc ion
Janua y 20, 2025, ma ked he beginning o Donald T ump’s second e m in he Whi e
House’s O al Office. In an in e iew a ew weeks be o e his inaugu a ion as he 47 h
P esiden o he USA he s a ed ha he would “absolu ely”conside aking he
Uni ed S a es ou o he Alliance unless membe s a e “ ea ing us ai ly”, ha is i
allies “pay hei bills”.
1
He was e e ing o he lowe de ence bu dens o mos NATO
*Co esponding au ho : Ch is os Kollias, Labo a o y o Economic Policy & S a egic Planning,
Depa men o Economics, Uni e si y o Thessaly, Volos, G eece, E-mail: [email p o ec ed]. h ps://o cid.o g/
0000-0003-2876-4304
Kons an ina Foun a, Labo a o y o Economic Policy & S a egic Planning, Depa men o Economics,
Uni e si y o Thessaly, Volos, G eece
Panayio is Tze emes, Depa men o Accoun ing and Finance, Uni e si y o Thessaly, Volos, G eece
1h ps://www.ba ons.com/news/ ump-says-will-absolu ely-conside -lea ing-na o-i -us-no -
ea ed- ai ly-9d3163c (accessed 4 Janua y 2025).
Peace Econ. Peace Sci. Pub. Pol. 2025; 31(2): 145–160
Open Access. © 2025 he au ho (s), published by De G uy e . This wo k is licensed unde he
C ea i e Commons A ibu ion 4.0 In e na ional License.
membe s compa ed o he US de ence budge . Indeed, in his fi s speech o an
in e na ional audience a he Da os 2025 Wo ld Economic Fo um Annual Mee ing, he
s essed ha he in ended o ask allies o inc ease hei de ence spending o 5 % o
GDP.
2
A ound a yea ea lie , in Feb ua y 2024, be o e his official p esiden ial
nomina ion a he Republican Na ional Con en ion in Milwaukee in July 2024, he is
quo ed saying du ing his speech a a campaign e en in Sou h Ca olina in a a he
bla an and diploma ically unbecoming manne ha i “You don’ pay you bills, you
ge no p o ec ion. I ’s e y simple”.
3
In an in e iew, once again dispensing wi h he
usual diploma ic ac ulness owa ds he o he NATO allies, he was mo e explici
saying ha in he cases o NATO membe s ha did no con ibu e enough unds
owa ds he Alliance’s collec i e de ence he would “encou age”Russia “ o do
wha e e he hell hey wan ”, indica ing ha hese Allies should no ely on US
suppo ing hem mili a ily.
4
E en hough such s a emen s a e plainly de oid o he expec ed ypical cou -
eous diploma ic language especially when add essed o allies, hey none heless echo
he long-s anding, and o all in en s and pu poses alid US c i icism ha many o
hei allies ee ide on he US p o ec i e de ence umb ella wi hou con ibu ing hei
ai sha e o he cos s o NATO’s collec i e de ence. The issue o bu den sha ing
be ween NATO’s membe s is by no means a new one. Hence he a en ion i has
ecei ed in he ele an li e a u e ha has been expanding s eadily o e he yea s.
Recen pape s on he heme o bu den sha ing and in e dependency among Wes e n
Allies include Geo ge and Sandle (2024, 2022), Ki imäki (2019), Pede sen, Oh , and
S endsen (2023), Kim and Sandle (2020), Mombelli (2024), Ca uso and Di Domizio
(2016), Sandle and Shimizu (2014), Zyla (2016), Becke (2017). Random examples o
mo e in age pape s on his heme a e hose by Sandle and Ha ley (2001), Ha ley
and Sandle (1999), Oneal (1990), Solomon (2004), Mu doch and Sandle (1984),
Sandle and Fo bes (1980). A ecen sys ema ic and comp ehensi e e iew o his
li e a u e can be ound in Boge s, Bee es, and Bollen (2022) and in Geo ge and
Sandle (2024, 2022). Consequen ly, o easons o b e i y, we e ain om engaging
in a undamen ally simila discussion he e. Mo eo e , he aim o he p esen pape is
no o engage wi h he bu den sha ing issue pe se.
Du ing his fi s P esiden ial e m, T ump did no spa e his s e n c i icism o-
wa ds o he NATO membe s, especially owa ds majo Eu opean coun ies, o no
shoulde ing he cos s o hei de ence. Fo example, du ing he 2018 B ussels NATO
2h ps://www.eu onews.com/business/2025/01/23/ ump-a -da os-na o-5-push- a iff-wa nings- o -
eu ope (accessed 23 Janua y 2025).
3h ps://abcnews.go.com/Poli ics/ ump- ussia-na o-de ense- unds/s o y?id=107136736 (accessed 23
Janua y 2025).
4ibid.
146 K. Foun a e al.
Summi i was epo ed ha he ga e some hing o an ul ima um ega ding allies’
de ence spending: “He said hey mus aise spending by Janua y 2019, o he
Uni ed S a es would go i alone”.
5
A s a emen e y much con ex ually simila o he
ecen one ci ed abo e conce ning a possible US wi hd awal om NATO. In a simila
ein, a he ou se o ha Summi he is quo ed calling o he NATO membe s delin-
quen in hei de ence expendi u es and demanded ha hey should inc ease hei
de ence budge s immedia ely,
6
while he has also asked Eu opean coun ies o buy
Ame ican weapons o keep NATO s ong.
7
Fo eign sales a e a e y impo an sou ce
o e enues o he US de ence indus y, a subs an ial sha e o which a e expo s o
o he NATO membe s. Fo ins ance, du ing 2000–2024 US manu ac u ed weapon
sys ems impo ed by Eu opean NATO Allies accoun ed o almos a qua e o he US
indus y’s o al expo s
8
while, he e enues o he US de ence indus ies a e also
affec ed by he US poli ical cycles (Bales a and Ca uso 2025) as well as by o eign
in e en ions (Ba onchelli and Ca uso 2024).
In his Da os 2025 speech when calling o a 5 % o GDP alloca ion o de ence,
P esiden T ump also e e ed o his fi s e m in office and he effec i had on allies’
mili a y spending: “I’m also going o ask all NATO na ions o inc ease de ence
spending o 5 % o GDP, which is wha i should ha e been yea s ago. I was only a
2 %, and mos na ions didn’ pay un il I came along”.
9
In wha ollows we examine he
impac exe ed by he fi s T ump P esidency on NATO Allies’mili a y spending,
ying o de e mine whe he he b agged T ump effec ci ed abo e is in ac
empi ically aceable. The sec ion ha ollows con ains a b ie o e iew o he
empi ical me hodologies used o p obe in o he issue a hand. The findings a e
epo ed and discussed in sec ions h ee and ou while sec ion fi e concludes
he pape .
2 A Bi d’s Eye View o he Da a and Me hodologies
Two me ics a e used o examine empi ically whe he a T ump effec is iden ifiable
on NATO membe s’alloca ions o de ence: mili a y spending as a sha e o GDP and
5h ps://www. eu e s.com/a icle/wo ld/ ump-wa ned-na o-allies-us-would-go-i -alone-i - hey-
did-no -spend-sou ces-idUSKBN1K21KY/(accessed 26 Janua y 2025).
6h ps://edi ion.cnn.com/2018/07/10/poli ics/donald- ump-na o-summi -2018/index.h ml (accessed
26 Janua y 2025).
7h ps://www.bloombe g.com/news/a icles/2025-02-13/ ump- ells-eu ope- o-buy-ame ican-
weapons- o-keep-na o-s ong.
8Da a calcula ed om SIPRI’s A ms T ans e s Da abase.
9h ps://www.we o um.o g/s o ies/2025/01/da os-2025-special-add ess-donald- ump-p esiden -
uni ed-s a es/(accessed 31 Janua y 2025).
You don’ Pay You Bills You Ge No P o ec ion 147
equipmen expendi u e as a sha e in o al de ence spending. Da a o he o me a e
d awn om SIPRI’s da abase and o he la e om NATO. Since he 2014 NATO
Wales Summi , bo h a e he me ics used o de e mine he minimum h esholds ha
Alliance membe s should adhe e o in o de o mee hei obliga ion o he quan i y
o esou ces alloca ed owa ds hei mili a y capabili ies (in e alia: Kim, Sandle ,
and Shimizu 2024; Haeseb ouck 2022; Blum and Po a ke 2020; A ani idis, Kollias,
and Messis 2017). To p obe in o he T ump effec hypo hesis, hese wo me ics a e
compa ed and analyzed du ing wo P esiden ial pe iods: he Obama wo e m
p esidency (2009–2016) and he fi s T ump p esidency. To allow o a possible
hys e esis in a T ump effec , he fi s yea o he Biden adminis a ion is included in
he la e pe iod. Hence he second pe iod in he analyses and empi ical es s ha
ollow co e s he yea s 2017–2021. Albei da a is a ailable o mo e yea s, we op o
ha e 2021 as he cu -offpoin in he se ies gi en he p o ound effec he ull-scale
Russian in asion o Uk aine in Feb ua y 2022 had on NATO de ence planning and
budge s (in e alia: Bu s, Ch is ie, and Du Bois 2024; Magula, Rouland, and Zwack
2022). All cu en NATO membe s
10
a e included in he empi ical analyses conduc ed
he ein. Howe e , due o da a a ailabili y cons ain s in he case o he second me ic,
i.e. he sha e o equipmen expendi u e in o al mili a y spending, Mon eneg o,
No h Macedonia, Finland and Sweden, a e no included in he sample and he
es ima ions. Using boxplo s, Figu es 1 and 2 offe a bi d’s eye iew o he da a du ing
he wo pe iods men ioned abo e. Tha is, he wo Obama p esiden ial e ms (2009–
2016) and he fi s T ump p esidency whe ein he fi s yea o he Biden adminis-
a ion ha ollowed is included (2017–2021) as p e iously poin ed ou .
Figu e 1: Mili a y spending as a sha e o GDP du ing he Obama and ump P esidencies. The mili a y
spending as a sha e o GDP da a is d awn om SIPRI’s mili a y expendi u e da abaseen.
10 Al hough Finland and Sweden joined NATO in 2023 and 2024 espec i ely, we op ed o include
hem in he analysis.
148 K. Foun a e al.
Fo he fi s o he wo me ics, he median alue o he en i e fi s pe iod was
1.33 % and o he second pe iod i was 1.43 % while he co esponding mean alues
o de ence spending as a sha e o GDP we e 1.49 % and 1.59 % espec i ely. Du ing
bo h pe iods he a e age de ence bu den o he USA was he highes among all he
NATO membe s s anding a 4.2 % du ing 2009–2016 and 3.4 % du ing 2017–2021. The
second and hi d la ges de ence spending o GDP a ios we e hose o G eece (2.6 %)
and he UK (2.3 %) du ing he Obama adminis a ion pe iod. The a e age a ios o
majo Eu opean powe s such as Ge many, F ance, I aly and Spain we e 1.2 %, 1.9 %,
1.4 %, 1.3 % espec i ely du ing he fi s pe iod depic ed in Figu e 1. Du ing he
second pe iod (2017–2021), he second and hi d la ges de ence spending o GDP
a ios we e hose o G eece (3 %) and Tü kiye (2.3 %) while he co esponding a -
e ages o he la ges Eu opean economies we e: Ge many 1.3 %, UK 2 %, F ance
1.9 %, I aly 1.5 % and Spain 1.3 %.
Tu ning o he second me ic, i.e. equipmen expendi u e as a sha e o o al
mili a y spending
11
depic ed in he boxplo s o Figu e 2, he median alue o he
en i e fi s pe iod was 13.48 % and o he second i was 19.8 % while he co e-
sponding mean alues we e 14.45 % and 21.52 % espec i ely. Du ing he wo Obama
p esidencies pe iod, he highes a e age sha e o equipmen expendi u e was
F ance’s 27.3 % ollowed by he USA (25.5 %) and Tü kiye (25.3 %). The co esponding
a e ages o he majo Eu opean powe s we e as ollows: Ge many 14.7 %, UK 22 %,
I aly 11.9 % and Spain 13.3 %. Du ing he second pe iod depic ed in Figu e 2, Tü kiye
(32.5 %), Li huania (32.4 %), Hunga y (29.9 %), Slo akia (28.9 %) and he USA (28 %)
eme ge as he fi e NATO membe s wi h he highes sha es in equipmen expendi-
u e, well abo e he 2014 Wales Summi a ge . The co esponding a e ages du ing
Figu e 2: De ence equipmen expendi u e as a sha e o o al mili a y spending du ing he Obama and
ump P esidencies. The de ence equipmen expendi u e as a sha e o o al mili a y spending da a is
d awn om NATO’s da abaseen.
11 As p e iously men ioned, Finland, Sweden, Mon eneg o and No h Macedonia a e no included
due o da a a ailabili y cons ain s.
You don’ Pay You Bills You Ge No P o ec ion 149
2017–2021 o he majo Eu opean powe s we e: F ance 25.4 %, UK 24.1 %, Spain 21 %,
I aly 19.7 % and Ge many 14.6 %.
To p obe in o he issue a hand, a ba e y o me hodologies is applied. The
empi ical analysis ha ollows in he nex sec ion s a s wi h he applica ion o a
mul idimensional scaling (MDS) app oach o map he ela i e posi ion o NATO
membe s o each o he wo me ics used he ein. B iefly, MDS is a s a is ical ech-
nique used o isualize he simila i y o dissimila i y be ween da a poin s in a lowe -
dimensional space. In ou case he e, in e ms o mili a y spending as a sha e o GDP
and equipmen expendi u e’s sha e in he NATO membe s’annual de ence budge s.
The undamen al idea behind MDS is o ep esen objec s as poin s in a Euclidean
space such ha he pai wise dis ances be ween hem app oxima e hei o iginal
dissimila i ies. In his case s udy, we use “Classical MDS”, a ype o MDS whe e he
inpu is a ma ix o Euclidean dis ances and eigen alue decomposi ion is pe o med
on a ans o med dis ance ma ix (Pawliczek and Dzwinel 2007).
The p ocess o MDS begins wi h a dis ance o dissimila i y ma ix, whe e each
elemen ep esen s he pe cei ed diffe ence be ween a pai o objec s. In classical
MDS, he dissimila i y ma ix is con e ed in o a squa ed dis ance ma ix, ollowed
by double-cen e ing o compu e a ma ix sui able o eigen decomposi ion. The
la ges eigen alues and hei co esponding eigen ec o s a e hen ex ac ed o
de e mine he coo dina es in he lowe -dimensional space (Pawliczek and Dzwinel
2007). In mul idimensional scaling he subjec o in es iga ion is da a space RNda a
ec o s Xmapped o he des ina ion space, usually Y∈R2. The analyzed dis ances
a e R
ij
=D(Xi,Xj) be ween Xiand Xjin RNand he dis ances
ij
=d(Yi,Yj)inR2. MDS
aims o find a map X→Y=M(X), minimizing global measu es o he opog aphic fi ,
i.e. he diffe ence be ween R
ij
and
ij
(Naud 2006). Acco ding o K uskal (1964), a
c i ical componen o MDS is he s ess alue, which quan ifies he disc epancy
be ween he obse ed dis ances in he o iginal space and he dis ances in he MDS-
gene a ed space. A lowe s ess alue indica es a be e ep esen a ion o he o iginal
da a. The mos commonly used measu e o s ess is K uskal’s s ess o mula, which is
defined as he squa e oo o he sum o squa ed diffe ences be ween he o iginal
dissimila i ies and he fi ed dis ances, no malized by he sum o squa ed o iginal
dissimila i ies. Ma hema ically, i is exp essed as:
s ess =S=
∑i<jdij −δij
()
2
∑i<jd2
ij
√
√
whe e d
ij
ep esen s he dis ances in he lowe -dimensional space, and δ
ij
ep esen s
he o iginal dissimila i ies. A lowe s ess alue indica es a be e fi be ween he
150 K. Foun a e al.
ans o med and o iginal da a. The dis ances a e calcula ed by Euclidian dis ance
o mula:
dij =xi1−xj1
()
2+xi2−xj2
()
2+…+xin −xjn
()
2
{}
1
/
2
In e p e ing s ess alues in ol es gene al guidelines, wi h lowe alues signi ying
be e ep esen a ions. A s ess alue below 0.05 is conside ed an excellen fi , be-
ween 0.05 and 0.1 is an accep able fi , be ween 0.1 and 0.2 sugges s a ai ep e-
sen a ion, and alues abo e 0.2 indica e significan dis o ion, meaning he MDS
solu ion may no be eliable (Naud 2006).
In p ac ical applica ions, minimizing s ess is a c ucial s ep in MDS analysis. This
is ypically achie ed h ough i e a i e op imiza ion echniques. Despi e i s impo -
ance, s ess alue should no be he sole c i e ion o e alua ing MDS solu ions.
Addi ional diagnos ic ools, such as Shepa d diag ams, which plo he ela ionship
be ween o iginal dissimila i ies and dis ances in he MDS space, can p o ide deepe
insigh s in o he quali y o he ep esen a ion (Naud 2006). In summa y, s ess alue
se es as a key me ic in MDS o assess he accu acy o lowe -dimensional embed-
dings. A lowe s ess alue is desi able as i indica es a mo e ai h ul ep esen a ion
o he o iginal da a ela ionships.
Then, as he second s ep in he empi ical analysis we op o employ he K-means
clus e ing echnique. B iefly, K-Means analysis is a widely used clus e ing echnique
ha o ganizes da a in o dis inc g oups based on simila i y. K-Means clus e ing is
based on an algo i hm ha is used o pa i ion da a in o k clus e s, whe e each da a
poin belongs o he clus e wi h he nea es mean. The algo i hm aims o minimize
he wi hin-clus e a iance, ensu ing ha poin s wi hin he same clus e a e mo e
simila o each o he han o poin s in o he clus e s (Pa el e al. 2022). The K-Means
p ocess s a s by selec ing k ini ial clus e cen oids, which can be andomly assigned
o chosen using specific ini ializa ion me hods, i.e. K-Means++. Each da a poin is
assigned o he nea es cen oid based on Euclidean dis ance, o ming k clus e s.
Once all poin s a e assigned, he cen oids a e ecalcula ed as he mean o he poin s
in each clus e . The p ocess i e a es, eassigning poin s o he closes cen oid and
upda ing cen oids un il con e gence is eached. Con e gence occu s when he
clus e assignmen s no longe change, o a p edefined s opping c i e ion is me . To
imp o e s abili y, k-means is o en un mul iple imes wi h diffe en ini ializa ions,
and he bes esul is chosen based on he lowes o al wi hin-clus e a iance
(Leono e al. 2019).
Bo h MDS and K-Means se e c i ical oles in explo a o y da a analysis. MDS
p o ides a meaning ul low-dimensional ep esen a ion o complex da a ela ion-
ships, aiding in pa e n ecogni ion and isualiza ion. K-Means, on he o he hand, is
effec i e in segmen ing da a in o meaning ul g oups, making i a popula choice o
You don’ Pay You Bills You Ge No P o ec ion 151
applica ions such as pa e n ecogni ion and anomaly de ec ion. The wo echniques
can be used oge he , whe e MDS educes dimensionali y be o e applying k-Means
clus e ing o enhance in e p e abili y and efficiency.
As a final s ep in he empi ical analysis, we op o examine o he possible
p esence o s uc u al b eaks in he ime se ies o he wo a iables o each NATO
membe . S uc u al b eaks in economic a iables a e in a iably he ou come o
significan changes, domes ic and/o ex e nal, ha exe a s a is ically aceable
effec on he a iables conce ned. Clea ly, in he case o mili a y expendi u es hese
include geopoli ical de elopmen s ha impose apid adjus men s o such public
ou lays, an unexpec ed majo change in he global o egional s a egic en i onmen ,
a majo na ional secu i y ela ed e en , a shi in de ence policy due o a go e n-
men al change, a fiscal c isis such as he ecen so e eign deb c isis ha esul s in
significan educ ions and cu backs in public expendi u es. E iden ly, he Russian
in asion o Uk aine as well as he shockwa es o he new T ump p esidency a e he
ype o geopoli ical de elopmen s ha lea e an indelible ma k and cause apid
adjus men s o de ence expendi u es. Building on he me hodology o Bai and Pe on
(1998, 2003), we employ a mul iple s uc u al b eak model ha does no equi e any
p edefined b eakpoin s. Ins ead, po en ial b eak da es a e es ima ed di ec ly om
he da a. Each se ies can be modeled sepa a ely wi h mb eaks (leading o m+1
dis inc segmen s) using an equa ion o he o m:
Φτ=Ψj+ωτ,τ=Tj −1+1,Tj −1+2,…,Tj,
whe e Φτ is he mili a y expendi u e as a sha e o GDP (o equipmen sha e in o al
de ense spending) a ime τ,j= 1,2,…,m+1,Ψjis he mean le el o Φτ in he j- h
segmen , and ωτ cap u es any addi ional dis u bances, including se ial co ela ion
o he e oscedas ici y. The unknown b eakpoin s (T1,T2,…,Tm) a e ound by mini-
mizing a chosen c i e ion (e.g., he Bayesian In o ma ion C i e ion, BIC, o he e-
sidual sum o squa es, RSS).
3 The Findings om MDS Mapping and K-Means
Clus e ing
We begin he p esen a ion and discussion o he findings yielded om he empi ical
analysis wi h he esul s om applying he MDS and K-Means me hodologies o he
wo me ics used he ein o p obe in o he ques ion a hand. These a e shown in
Figu e 3 in he case o he mili a y expendi u e o GDP a io a iable and in Figu e 4
o he equipmen pe cen age sha e in o al mili a y spending me ic. In each figu e
he co esponding maps and clus e s o each o he wo pe iods a e included.
152 K. Foun a e al.
Boge s, M., R. Bee es, and M. Bollen. 2022. “NATO Bu den Sha ing Resea ch along Th ee Pa adigms.”
De ence and Peace Economics 33 (5): 534–47.
Bu s, C., E. H. Ch is ie, and C. Du Bois. 2024. “Explo ing he Speed o Change: Eu opean De ence Budge s in
an Age o Dange .”De ense and Secu i y Analysis 40 (3): 319–40.
Ca uso, R., and M. Di Domizio. 2016. “In e dependence be ween US and Eu opean Mili a y Spending: A
Panel Coin eg a ion Analysis (1988–2013).”Applied Economics Le e s 23 (4): 302–5.
Geo ge, J., and T. Sandle . 2022. “NATO De ense Demand, F ee Riding, and he Russo-Uk ainian Wa in
2022.”Jou nal o Indus ial and Business Economics 49: 783–806.
Geo ge, J., and T. Sandle . 2024. “A Spa ial Analysis o NATO Bu den Sha ing a he Ope a ional Le els.”
Kyklos 77 (4): 1026–47.
Haeseb ouck, T. 2022. “NATO Bu den Sha ing a e he Wales Summi : A Gene alized Se Quali a i e
Analysis.”De ence and Peace Economics 33 (6): 637–54.
Ha ley, K., and T. Sandle . 1999. “NATO Bu den-Sha ing: Pas and Fu u e.”Jou nal o Peace Resea ch 36 (6):
665–80.
Kim, W., T. Sandle , and H. Shimizu. 2024. “An Expanded In es iga ion o Alliance Secu i y F ee Riding.”
Global Policy 15 (4): 570–82.
Kim, W., and T. Sandle . 2020. “NATO a 70: Pledges, F ee Riding, and Benefi -Bu den Conco dance.”
De ence and Peace Economics 31 (4): 400–13.
Ki imäki, T. 2019. “Powe , Con ibu ion and Dependence in NATO Bu den Sha ing.”Eu opean Secu i y 28
(1): 66–84.
Leono , P., V. Suy s, O. Ko elyane s, and N. I ano . 2019. “K-means Me hod as a Tool o Big Da a Analysis in
Risk-O ien ed Audi .”In Big Da a Inno a ions And Applica ions, edi ed by M. Younas, I. Awan, and
S. Benbe nou, 206–16. Cham, Swi ze land: Sp inge .
Magula, J., M. Rouland, and P. Zwack. 2022. “NATO and Russia: De ense and De e ence in a Time o
Conflic .”De ence S udies 22 (3): 502–9.
Mombelli, S. 2024. “De ense Bu den Sha ing and Mili a y Coope a ion in he EU27: A Desc ip i e Analysis
(2002–2023).”Peace Economics, Peace Science and Public Policy 30 (4): 423–53.
Mu doch, J., and T. Sandle . 1984. “Complemen a i y, F ee Riding, and he Mili a y Expendi u es o NATO
Allies.”Jou nal o Public Economics 25 (1–2): 83–101.
Naud, A. 2006. “An Accu a e MDS-Based Algo i hm o he Visualiza ion o La ge Mul idimensional
Da ase s.”In A ificial In elligence and So Compu ing –ICAISC 2006, 643–52. Be lin, Heidelbe g:
Sp inge .
Oneal, J. R. 1990. “The Theo y o Collec i e Ac ion and Bu den-Sha ing in NATO.”In e na ional O ganiza ion
44 (3): 379–402.
Pa el, S. S., S. S. Pa el, N. Kuma , N. Kuma , J. Aswa hy, J. Aswa hy, S. K. Vaddadi, e al. 2022. “K-Means
Algo i hm: An Unsupe ised Clus e ing App oach Using Va ious Simila i y/Dissimila i y Measu es.”
In In elligen Sus ainable Sys ems, edi ed by J. Raj, R. Palanisamy, I. Pe ikos, and Y. Shi, 805–13.
Singapo e: Sp inge .
Pawliczek, P., and W. Dzwinel. 2007. “Visual Analysis o Mul idimensional Da a Using Fas MDS Algo i hm.”
Pho onics Applica ions in As onomy, Communica ions, Indus y, and High-Ene gy Physics Expe imen s
6937 (2M): 755–65.
Pede sen, R., A. Oh , and G. S endsen. 2023. “F ee Riding in NATO a e he Rise o Russia: Cos Sha ing,
F ee Riding and Selec i e Incen i es in NATO om 2009 o 2019.”Jou nal o T ansa lan ic S udies 21 (1):
54–72.
Sandle , T., and J. F. Fo bes. 1980. “Bu den Sha ing, S a egy, and he Design o NATO.”Economic Inqui y 18
(3): 425–44.
You don’ Pay You Bills You Ge No P o ec ion 159
Sandle , T., and K. Ha ley. 2001. “Economics o Alliances: The Lessons o Collec i e Ac ion.”Jou nal o
Economic Li e a u e 39 (3): 869–96.
Sandle , T., and H. Shimizu. 2014. “NATO Bu den-Sha ing 1999–2010: An Al e ed Alliance.”Fo eign Policy
Analysis 10 (1): 43–60.
Solomon, B. 2004. “NATO Bu den Sha ing Re isi ed.”De ence and Peace Economics 15 (3): 251–8.
Zyla, B. 2016. “Who Is Keeping he Peace and Who Is F ee-Riding? NATO Middle Powe s and Bu den
Sha ing, 1995–2001.”In e na ional Poli ics 53 (3): 303–23.
160 K. Foun a e al.
