He e ogenei y, Di e si y, and Complemen a i y in Alliance Po olios
Abs ac :
The di e gence ha a i m manages o achie e because o i s pa ne s is a undamen al
ques ion in an alliance po olio con igu a ion. Di e si y can e e no only o he
di e ences be ween he ocal i m and i s pa ne s o be ween he pa ne s hemsel es
bu also o he di e ences a ising om a ious esou ce endowmen s in he alliance
po olio. Unde s anding he signi icance o hese di e en sou ces, how hey
in e ela e, and how hey a ec he i m pe o mance is an un esol ed ques ion as
unclea de ini ions and opposing a gumen s a e p oposed. This pape expounds he
concep s o pa ne he e ogenei y, alliance po olio di e si y, and ne wo k esou ce
complemen a i y o gain a deepe comp ehension o he alliance po olio con igu a ion
and how i a ec s pe o mance. Ou analysis o ai line alliances a a global le el e eals
he cen al ole o esou ce complemen a i y in he ocal i m pe o mance.
Keywo ds:
He e ogenei y, Di e si y, Complemen a y, Alliance Po olio, Ne wo k Resou ces.
1. In oduc ion
S a egic alliances play a i al ole in i m su i al as long as hey con ibu e
ex e nal esou ces ha allow i ms o gain and main ain compe i i e ad an ages (Mu,
Lo e & Peng, 2008). These ex e nal esou ces a e e e ed o as ne wo k esou ces
(Gula i, 2007). Access o hem and hei e en ual use is a key mo i a ion o he
company o in ol e i sel in in e o ganiza ional ies and alliance o ma ion.
In e y compe i i e and highly luc ua ing business en i onmen s, some
o ganiza ions usually en e in o mul iple ag eemen s wi h o he i ms o de elop a ious
compe i i e ad an ages simul aneously. The se o i m alliances is e e ed o as i s
alliance po olio, and i gene a es a need o global and simul aneous managemen o
all alliances so ha he company achie es i s objec i es (Wassme , 2010). Ho mann
(2007: 834) s a ed ha alliance po olio composi ions “de e mine he quali y, quan i y,
and di e si y o in o ma ion and esou ces o which he ocal company has access.” In
o he wo ds, he global i m’s ne wo k esou ce access will be de e mined by who i s
pa ne s a e (in di e en alliances) and by hose pa ne s’ esou ces.
P e ious li e a u e abou Ne wo k Theo y and Resou ce-based iew has highligh ed
ha ne wo k esou ces o which a company has access h ough i s ela ions should be
complemen ed and/o combined wi h i s own esou ces (Chung, Singh & Lee, 2000;
Eisenha d & Schoonho en, 1996; La ie, 2006; Mi suhashi & G e e, 2009; S ua ,
2000; Wassme , 2007; Zheng, Li & Wu, 2013). The e o e, i is ele an o ind pa ne s
ha a e di e en om he ocal i m and pa ne s wi h esou ces ha di e om i s own
esou ces.
Baum, Calab ese and Sil e man (2000) p e iously indica ed he need o con igu e
alliance po olios, wi hin hei classic concep ualiza ion, o pe mi access o mo e
di e gen in o ma ion and capabili ies. Di e si y, non edundancy, syne gy, and he
b ead h o pa ne esou ce cha ac e is ics a e ele an elemen s connec ed o he
company’s pe o mance ha o m pa o he esea ch agenda on alliance po olio
con igu a ion (Wassme , 2010). In he ecen li e a u e, e en g ea e in e es has been
exp essed in e idence ha shows how di e en aspec s o he alliance po olio
con igu a ion (pa icula ly di e si y and/o simila i y be ween i ms and be ween he
esou ces ha hey own) a ec company pe o mance (Collins & Riley, 2013; Cui &
O’Conno , 2012; De Leew, Lokshin, & Duys e s, 2014; Duys e s & Lokshin, 2011;
Kim, 2014; La ie, 2007; Sa ka , Aulakh, & Madhok, 2009).
Howe e , a a concep ual le el, he e is s ill a lack o cla i y o e he use o di e en
wo ds, and he e a e le els o analysis ha desc ibe he ci cums ances in which an
alliance po olio con igu a ion imp o es pe o mance. Wo ds such as he e ogenei y,
di e si y, o complemen a i y a e used abundan ly o e e o di e en si ua ions and
le els. On occasions, he di e en e ms appea mixed up wi h he same de ini ion:
“Alliance pa ne he e ogenei y e e s o he b ead h o di e si y o he complemen a y
capabili ies held by di e en alliance pa icipan s” (Lin, 2012) and “Di e si y e e s o
he he e ogenei y o esou ces o knowledge and hus yields he po en ial o no el
combina ions o eme ge” (Vanha e beke, Gilsing, Bee kens, & Duys e s, 2009).
The e o e, he e is a need o cla i y he concep s ha a e used in he p e ious li e a u e
be o e e e ing o he alliance po olio di e si y.
Al hough he di e en sou ces can be b oade and can e e o s uc u al ques ions
and o bo h geog aphic and cul u al dis ances (Ahuja, Polido o & Mi chell, 2009; La ie
& Mille , 2008), he p e ious li e a u e has indica ed ha he basic di e gence sou ces
ocus on he di e ences be ween bo h pa ne s and esou ces (Gula i, La ie &
Madha an, 2011; Kale & Singh, 2009). The di e ences be ween he alliance po olio
pa ne s may a ise be ween hem o wi h he ocal i m. Following p e ious pape s on
ne wo k app oach s a egy, his wo k cen e s on he di e en sou ces ha , in a di ec
way, a e linked o he ne wo k esou ces and unde s and ne wo k esou ce as pa ne
esou ce endowmen (La ie, 2008).
Ea lie esea ch has no ed i m he e ogenei y as a p elimina y ac o in he exis ence
o complemen a i y, and bo h concep s a e linked o business pe o mance (Baum e al.,
2000; Bu , 1992; La ie, 2006; Mi suhashi & G e e, 2009; Wassme , 2007). O he
s udies indica e he way in which company homogenei y s imula es inno a ion and
imp o es esul s (Ahuja, 2000; Cui & O’Conno , 2012). The e o e, he e is a ce ain
con o e sy abou wha he bes ype o pa ne migh be o imp o e he business
pe o mance. P e ious s udies ha e p oposed a gumen s and e idence ha indica e ha
holding a mo e di e se alliance po olio and one con ibu ing complemen a y esou ces
yields posi i e o nega i e esul s o esul s ha change o e ime. The e a e no e iews
ha discuss he di e en o ms o di e ences ha can a ise in an alliance po olio and
how hese o ms a e in e ela ed wi h each o he (e.g., i some may be he an eceden s o
consequences o o he s). The e o e, he e is a gap in he li e a u e ha co e s he
a ious ways o conside ing di e gence o con e gence in he alliance po olio
con igu a ion, he possible connec ions be ween hem, and how di e ences a ec i m
pe o mance in di e en ways.
The aim o his wo k is o look mo e deeply a he ques ion o he alliance po olio
con igu a ion ollowing he lines ma ked ou by Ho mann (2007) and Wassme (2010).
The objec i e is o con ibu e o a be e unde s anding o he ela ions be ween he h ee
basic o ms o unde s anding he di e ences in alliance po olio con igu a ion
(he e ogenei y be ween he ocal i m and i s pa ne s, di e si y be ween he pa ne s,
and esou ce complemen a i y) and how hese (each one sepa a ely and in ela ion o
each o he ) a ec he business pe o mance. The e o e, his wo k will seek o speci y
whe he hey a e ela ed in ei he a posi i e o a nega i e way o he esul s (o in a
nonlinea way) and whe he hey do so di ec ly o indi ec ly. Mo eo e , his pape
de ines hese h ee o ms o di e ence in an alliance po olio.
Codesha e alliances sha ed be ween ai lines a an in e na ional le el ha e been
conside ed o his pu pose, employing da a om 135 alliance po olios ha include he
main companies in he sec o and ep esen o e 70% o he o al.
2. Theo e ical Re iew
The p oblem o how he ocal i m selec s mo e o less di e en pa ne s is a key in
he s udy o alliance po olios (Cas o, Casanue a & Galán, 2014). The a ie y o
pa ne s and he di e gence o he esou ces ha hey con ibu e a e decisi e o he
alliance po olio con igu a ion (Ho mann, 2007).
Howe e , he ollowing key ques ions su ound wha we conside a di e en
pa ne . How do we analyze he di e ences? Wha is he basis o hose di e ences?
Placing he ocus on ne wo k esou ces (Gula i, 2007; La ie, 2008), alliance po olio
di e gence has been conside ed in nume ous p e ious s udies. A e iew o he li e a u e
shows ha an un esol ed concep ual p oblem exis s o e he name o each concep (see
Appendix).
In his esea ch, h ee di e en ia ed concep s a e p oposed o analyze di e gence
and con e gence wi hin an alliance po olio: he e ogenei y, po olio di e si y, and
complemen a i y. We conside ed ha he e ogenei y e e s o “how di e en he
o ganiza ion is om i s pa ne s” (Gula i e al., 2011), di e si y is “ he ex en o
a iance in a ocal i m’s alliance pa ne s” (Collins & Riley, 2013), and
complemen a i y is “ he ex en o which a pa ne con ibu es non-o e lapping esou ces
o he ela ionship” (Kale & Singh, 2009).
Business ne wo ks ha e o be conside ed a wo le els. The i s is he dyadic le el
ha ela es he ocal i m wi h each one o i s pa ne s and p oposes whe he hey a e
di e en because o hei cha ac e is ics o because o hei esou ce endowmen . The
second is a he egone le el—alliance po olio. A his le el, he di e ences be ween
he ocal i m egone membe s a e conside ed: be ween hei pa ne s. A dis inc ion
should be made be ween whe he he di e ences e e o he i m cha ac e is ics o o
hei esou ce endowmen .
Table 1 se s ou a concep ual amewo k wi h which o analyze he di e ence in he
alliance po olio, which conside s h ee sou ces o di e gence: di e ences be ween he
cha ac e is ics o ei he he i ms o he esou ces and he conside a ion o wo ne wo k
le els (dyad and egone ). Each new pa ne choice in he alliance po olio con igu a ion
is a commi men and has an e ec a he dyadic le el (speci ic alliance wi h he new
pa ne ) and a he global o ne wo k le el as long as i shapes and modi ies he egone
cha ac e is ics and s uc u e (alliance po olio).
In Table 1, he e ogenei y is linked o he exis ence o di e en a ibu es be ween
wo pa ne s. I is a simila concep o he one used by Gula i e al. (2011), whe e hey
e e o di e ences be ween he i ms’ cha ac e is ics ha c ea e an alliance and he
Pa khe’s (1991) ype II di e si y. The e o e, he e ogenei y ocuses on he di e ences
be ween dyadic pai s ( ocal company and each o hei pa ne s’ cha ac e is ics, one by
one). Di e si y cen e s on he ego ne wo k: all ocal i m pa ne s wi hou he ocal i m
i sel . In his way, he di e ences be ween he alliance po olio membe s a e
conside ed, wi hou conside ing ocal o ganiza ion cha ac e is ics. This de ini ion o
di e si y is simila o hose gi en by Collins and Riley (2013) and Jiang, Tao and
San o o (2010). Finally, complemen a i y e e s o he compa ison o he esou ce
endowmen s (and he na u e o hose esou ces) be ween he ocal i m and i s pa ne s.
Complemen a i y e e s o inding he esou ces ha he ocal i m needs in ano he i m
when hey ind he ideal combina ion and hei e ec is g ea e oge he han i is
indi idually (Gula i, 1999). I bo h he pa ne s’ esou ces a e simila , hen hose
esou ces will be edundan (Huggins, 2010).
Wi h ega d o he ela ions be ween hese h ee di e gence sou ces (pa ne
he e ogenei y, alliance po olio di e si y, and ne wo k esou ce complemen a i y),
p e ious li e a u e has some imes eso ed o one o hose ela ions as a p oxy o he
o he ; a o he imes, hese ela ions ha e coincided wi h each o he and he esul s ha e
had a iable e ec s. Thus, some s udies ha e conside ed he e ogenei y as a synonym o
di e si y (Beckman & Haunschild, 2002; De Leeuw e al., 2014; Duys e s & Lokshin,
2011). In o he s udies, he e is no clea sepa a ion be ween di e si y and he e ogenei y
(Collins & Riley, 2013; Jiang e al., 2010). Va ious s udies ha e linked
complemen a i y o c ea ing alliances wi h he e ogeneous pa ne s, which con ibu e o
he minimiza ion o he o e lapping (Abuzaid, 2014; Chung e al., 2000; Duys e s &
Lokshin, 2011; Kale & Singh, 2009; Kim, 2014; Shan & Hamil on, 1991). The e o e, i
is necessa y o unde s and how each alliance po olio di e gence sou ce (sepa a ely
and combina ion wi h he o he s) a ec s i m pe o mance.
Table 1: Alliance po olio di e gence sou ces
Pa ne
He e ogenei y/Homogenei y
Alliance Po olio
Di e si y/Simila i y
Ne wo k Resou ce
Complemen a i y/Redundancy
Ne wo k
le el
Dyad
Egone
Dyad
Objec o
Fi ms
Fi ms
Fi m esou ces
he
di e ence
Objec o
compa ison
Focal i m cha ac e is ics s.
pa ne cha ac e is ics
Di e ences be ween
all pa ne
cha ac e is ics,
wi hou including he
ocal i m
Focal i m esou ces s.
pa ne s’ esou ces
2.1. Pa ne he e ogenei y and pe o mance
He e ogenei y means ha he ocal company cha ac e is ics ha e o be compa ed
wi h each one o hei alliance po olio pa ne s. The main a gumen ha links
he e ogenei y o pe o mance is di ec ly linked o esou ces. P e ious wo ks ha e
de ined he e ogenei y in opposi ion o homogenei y and ha e a gued ha he mos
aluable i ms a e companies ha ha e he e ogeneous pa ne s as hose o ganiza ions
a e mo e e icien and con ibu e di e en esou ces ha encou age he compe i i e
ad an ages de elopmen (Ahuja e al., 2009; Baum e al., 2000; Beckman &
Haunschild, 2002; Lee, 2007; Pa khe, 1991; Wuy s, Du a & S eme sch, 2004). In
o he wo ds, a company mus ha e allies ha b ing in new esou ces—di e en om
hose ha i al eady has— o imp o e i s pe o mance (Baum e al., 2000; Bu , 1992;
Gula i e al., 2011; Mi suhashi & G e e, 2009; Reagan & Zucke man, 2008).
Ano he impo an a gumen on how he e ogenei y a o s he alliance pe o mance
cen e s on he di icul y o coope a ion be ween pee s. An alliance wi h di ec
compe i o s implies ha hey a e homogenous o ocal i m, which can p o oke
nume ous con lic s as hey collide in he pu sui o hei aims ha e ode he ela ionship
(Baum e al., 2000; Cane & Tyle , 2013; De Leeuw e al., 2014; Gomes-Casse es,
1994; La ie, 2006; Pa k, S i as a a & Gnyawali, 2014) and gene a e oppo unis ic
beha io (Gula i, Noh ia, & Zahee , 2000). Hi , Hoskisson and Kim (2001) iden i ied
ha homogeneous i ms a e no able o exploi all he alliance oppo uni ies.
I seems ha he e ogeneous collabo a ions a e mo e aluable owing o hei enla ged
ange o a ailable esou ces and wha ha implies: mo e lexibili y o adap o
en i onmen changes because o di e en compe i i e ad an ages c ea ed h ough
combina ions o e y di e en esou ces. These combina ions a e also almos inimi able
because o he combina ion uniqueness, and in he las ins ance, hey could, acco ding
o Dye (1996), inc ease hei p oduc i i y.
Howe e , he e a e also heo e ical a gumen s and e idence in he li e a u e ha
show ha i m he e ogenei y can ha e a nega i e e ec on pe o mance and is linked o
an inc ease in he cos s. He e ogenei y in ol es a high cos oo because a company has
o in es mo e esou ces o o e come he di e ences, and hese cos s educe he alliance
pe o mance (Baum e al., 2000; Collins & Riley, 2013; De Leeuw e al., 2014).
Despi e he abo e, i seems ha he ype o pa ne ha p omp s be e pe o mance
will depend on he ype o esou ces ha he ocal i m wishes o ob ain. G an and
Baden-Fulle (2004) showed ha i he wish is o access physical o angible esou ces,
such as ma ke s (Kogu , 1988), he bes op ion will be o associa e wi h he e ogeneous
pa ne s as he alliance pa ne p oduces a pooling o esou ces ha will gene a e
mul iple bene i s; hus, he ocal i m expands in o a ious ma ke s. On he con a y, i
he idea is o imp o e he alue chain o o inno a e, he bes app oach is o choose
homogenous pa ne s, con ac s ha will cause cospecializa ion in he company and a
con e gence p ocess (G an & Baden-Fulle , 2004; Jiang e al., 2010; Pa k e al., 2014),
because con e gen i ms lea n in mo e e ec i e ways (Lane & Luba kin, 1998; Roge s,
1995) and a e be e a he coc ea ion p ocesses (Pa k e al., 2014).
Thus, he e a e a gumen s ha poin o bo h a posi i e and nega i e ela ion be ween
pa ne he e ogenei y and ocal i m pe o mance; e en a nonlinea ela ion be ween
he e ogenei y and pe o mance is possible. The e o e, a be e unde s anding is equi ed
abou he ype o ela ion ha appea s be ween hese a iables wi hou p oposing he
ela ion sign a p io i. Acco dingly, we may es ablish he ollowing hypo hesis:
H1: Alliance pa ne he e ogenei y is ela ed o he ocal i m pe o mance.
2.2. Alliance po olio di e si y and pe o mance
Alliance po olio di e si y compa es ocal i m pa ne s wi h each o he ( hei
simila i ies and di e ences wi hou aking he ocal i m in o accoun ). In an egone
analysis, di e si y is he a ie y o he pa ne s ha pa icipa e in he alliance po olio.
Re e sely, simila i y implies all po olio pa ne s i in he same p o ile.
The mo i es ha d i e he alliance po olio c ea ion usually ocus on ob aining
ne wo k esou ces because hey help c ea e di e en compe i i e ad an ages, eac
be e o en i onmen al changes (Dess, Rasheed, McLaughlin, & P iem, 1995; Duys e s,
De Man, & Wildeman, 1999; Lahi i & Na ayanan, 2013), o inno a e (Golonka, 2015),
and educe cos s (Be is, B adley & Hamel, 1992; Lahi i & Na ayanan, 2013).
The e o e, i would be logical o hink ha owning mul iple pa ne s ha con ibu e e y
a ied esou ces be ween each o he will inc ease he possibili ies o di e en and
o iginal compe i i e ad an ages de elopmen . These compe i i e ad an ages will be
based on combina ions ha compe i o s a e unable o imi a e (Beckman & Haunschild,
2002; Collins & Riley, 2013; Duys e s, Heime iks, Lokshin, Meije & Sabidussi, 2012;
Faems, De Visse , And ies & Van Looy, 2010; Ho man, 2007; Jiang e al., 2010; Kim,
2014; Lahi i & Na ayanan, 2013).
The le el o alliance po olio di e si y has caused ce ain esea che s conce n
because a high le el o di e si y can imply high cos s (De Leeuw e al., 2014; Duys e s
& Lokshin, 2011; Faems e al., 2010; Lahi i & Na ayanan, 2013; Pa k e al., 2014) and
g ea e o in alliance coo dina ion and in decisions abou po olio con igu a ion o
a oid con lic s and incompa ibili ies (Collins & Riley, 2013). Some s udies show ha
alliance po olio di e si y is only posi i e up o a ce ain poin (Duys e s & Lokshin,
2011; Koka & P esco , 2002). Ne e heless, i hey o e come he di icul ies, o
example, wi h a special alliance unc ion dedica ed o coo dina e and o acili a e
esou ce exchanges (Dye , Kale & Singh, 2001; Faems e al., 2010; Kale, Dye &
Singh, 2002), i ms will sha e a ious esou ces, and hose asse s will, in u n, gi e
hem s a egic lexibili y o ge a be e ma ke posi ion and will also inc ease hei
su i al, allowing hem o ob ain be e pe o mance (Collins & Riley, 2013; Hamel,
Doz & P ahalad, 1989).
Once again, as wi h he e ogenei y, p e ious s udies ad anced bo h e idence and
a gumen s ha indica e he posi i e ela ion be ween alliance po olio di e si y and
ocal i m pe o mance, whe eas o he s showed ha di e si y is only posi i e up o a
ce ain le el o ha i can be nega i e. The e o e, ou hypo hesis sugges s he ela ion
be ween bo h a iables wi hou p esupposing he ela ion sign:
H2: Alliance po olio di e si y will be ela ed o ocal i m pe o mance.
2.3. Ne wo k esou ces complemen a i y and pe o mance
Mos o he alliance po olio con igu a ion has ocused on he di e ences be ween
i ms. Howe e , on many occasions, he essen ial a gumen was he esou ce sea ch ha
hey do no own. Complemen a i y e e s o he di e ences be ween ocal i m
esou ces and hei pa ne asse s. In o he wo ds, complemen a i y e e s o pa ne s
wi h a esou ce endowmen ha ocal i m does no possess ha a e nono e lapping and
syne gic wi h hose ha i does own. Syne gy implies ha wo esou ces a e
complemen a y when hei join pe o mance, which hey gene a e oge he , is g ea e
han each one by i sel (Milg om & Robe s, 1990).
Resou ce complemen a i y is decisi e in an alliance as such a combina ion o
esou ces manages o gene a e e ec s ha a e supe io o ha gene a ed by he
esou ces sepa a ely by each pa ne . In addi ion, i his combina ion is inimi able, i
will achie e ad an ages o e he o he i als. We may add ha a i m wi h an alliance
po olio can e en solely combine esou ces om a ious pa ne s o wi h i s own,
making i mo e di icul o o he s o manage o imi a e such a combina ion.
In di e en s udies, he idea ha ne wo k esou ce complemen a i y gua an ees he
alliance success has been es ed as i inc eases he company alue and he le el o i m
pe o mance (Chung e al., 2000; De Leeuw e al., 2014; Kale & Singh, 2009; Wassme ,
2007; Zheng e al., 2013). O he s udies ha e e en a i med ha complemen a i y
should be conside ed be o e he alliance o ma ion and ha i in luences pe o mance
(Chung e al., 2000; Gula i, 1995; Gula i, 2007; Kim, 2014; La ie, 2007; Zheng e al.,
2013).
I complemen a i y is impo an a a dyadic le el, i is also impo an a he le el o
he alliance po olio. Thus, di e en esea che s emphasized ha when inco po a ing
new alliances, he new complemen a i ies ha hey will con ibu e should be ca e ully
examined o see whe he hey will cause any con lic wi h he p e ious membe s (Cane
& Tyle , 2013; Wassme , 2007) and whe he hey will c ea e less alue.
Ne e heless, he e a e companies ha p e e edundan esou ces, pa icula ly in
knowledge and in o ma ion, because hey b ing a highe le el o us and eliabili y
(Gula i, 2007; Lee, 2007). Dussauge, Ga e e and Mi chell (2000) de ailed ha he
pa ne s in a la ge-scale alliance end o coope a e wi h hose wi h simila esou ces and
skills. Ahuja e al. (2009) p oposed a nonmono onic esou ce simila i y e ec on
alliance o ma ion. Once again, he ela ion be ween complemen a i y and pe o mance
is no clea ly es ablished in he li e a u e, p oducing con ex s whe e i is be e o ob ain
pa ne edundan esou ces and whe e he ela ion migh no be linea . The e o e, we
p opose he ollowing hypo hesis:
H3: Alliance esou ce complemen a i y will be ela ed o he ocal i m
pe o mance.
Mo eo e , in ligh o he abo e, complemen a i y is usually associa ed wi h
he e ogenei y and di e si y because i is belie ed ha a ocal i m will seek di e en
esou ces o hose ha i al eady has when i c ea es an alliance. Thus, he
he e ogeneous pa ne s a e he ones who p o ide he esou ces ha di e om he ocal
company esou ces. Some s udies such as hose by Baum e al. (2000), La ie (2007),
Bu (1992), Sakakiba a (1997), Mi suhashi and G e e (2009) and Wassme (2007)
indica e his endency. Complemen a i y is he e o e de ined in opposi ion o pa ne
o e lapping and ne wo k edundancy and ine icacy (Abuzaid, 2014; Lin, 2012), caused
by company homogenei y, which leads us o hink ha he e ogenei y will gene a e
u he complemen a i y. Mo eo e , alliance po olio di e si y—o egone di e si y—
implies mul iple sou ces ha will gene a e bene i s om he c ea ed syne gies (Duys e s
& Lokshin, 2011). In o he wo ds, he e ogeneous pa ne s and alliance po olio
di e si y will mean ha he i m has di e en and complemen a y esou ces o i s own
because pa ne and esou ce edundancy is ei he elimina ed o se e ely limi ed.
Finally, complemen a i y may also be said o ha e a media o y e ec in he ela ion ha
bo h pa ne he e ogenei y and alliance po olio di e si y ha e wi h pe o mance.
The e o e, we p opose he ollowing hypo heses:
H4a: The mo e he pa ne ’s he e ogenei y, he mo e likely hey will be able o
achie e complemen a y esou ces.
H4b: The g ea e he alliance po olio di e si y, he g ea e he likelihood o hem
achie ing mo e complemen a y esou ces.
H5a: Resou ce complemen a i y in an alliance po olio media es he ela ion
be ween pa ne he e ogenei y and ocal i m pe o mance.
H5b: Resou ce complemen a i y in an alliance po olio media es he ela ion
be ween alliance po olio di e si y and ocal i m pe o mance.
3. Me hods
3.1. Sample
The ai line indus y was selec ed o ou analysis o how hese h ee aspec s (pa ne
he e ogenei y, alliance po olio di e si y, and esou ce complemen a i y) in luence
pe o mance. In he ai line business, he e is a mul i ude o alliances and o he
collabo a i e ag eemen s be ween i ms and la ge in e na ional g oups ha ope a e
a ound he wo ld and compe e a a global le el. Ai lines en e in o simul aneous alliance
ag eemen s wi h a ela i ely la ge numbe o ai line companies such ha he po olio
join managemen is also a s anda d p ac ice. In his way, he mul iple ne wo ks c ea ed
in his indus y may be examined. Al hough in e i m collabo a ion has exis ed in his
sec o e e since i s c ea ion, he p ocess has speeded up o e ecen yea s, wi h he
alliance and g oup eme gence ope a ing oge he . I is a widely used con ex in he
s a egic alliance analysis (Casanue a, Gallego & Sancho, 2013; Gimeno, 2004;
Gomes-Cassa es, 1994; Lazza ini, 2007; Shah & Swamina han, 2008; Wassme , 2007).
In addi ion, he ai line sec o pe mi s a clea iden i ica ion o he key esou ces ha
de ine compe i ion wi hin i . Mo e speci ically, i we conside only ma ke esou ces,
des ina ions o which he ai lines ly and hei ou es ep esen one o he mos ele an
esou ces in he indus y’s success (Wassme , 2007). Howe e , des ina ions ha e o he
added componen s ha in ol e physical esou ces such as land-based se ices,
main enance, and so on. The e o e, ai line sec o allows he p oposed ela ions o be
s udied, bo h because o he clea iden i ica ion o each alliance ha pa icipa es in he
alliance po olio and he possibili y o sha ing wi h he pa ne s.
In his s udy, i wo ai lines en e ed in o a codesha e ag eemen , an alliance was
conside ed o exis be ween hem (Min & Mi suhashi, 2012). A codesha e means an
ai line can sell sea s o i s clien s on he o he company’s ligh s wi h which i has an
ag eemen . This a angemen means ha an ai line company can ly o des ina ions ha
i does no i sel o e because o i s alliances wi h i s pa ne s. In a codesha e alliance
ha is sha ed, pa ne s can inco po a e he des ina ion in i s ou es, bu he esou ce
emains unde he con ol o he i m ha holds i . In his indus y, des ina ions a e a
key esou ce, whe he unde s ood as an elemen in a ne wo k s uc u e o as a ma ke
(Wassme & Dussauge, 2012).
3.2. Da a
Da a om he la ges 200 i ms in he indus y, epo ed by Ai line Business jou nal,
we e used in he s udy. Wi hin his se o i ms, he e we e some g oups, which mean
ha a o al o 214 ai line companies we e ac ually selec ed. Finally, hose ha had no
codesha e ag eemen s we e il e ed ou , lea ing a o al o 135 ai line companies wi h a
u no e o 75% o he o al indus y income.
The da a equi ed o pe o m he s udy we e compiled on he 135 i ms ha
cons i u e he 135 po olios unde analysis. In o al, o e 1700 codesha e ag eemen s
we e iden i ied among hem. The da ase s o he iden i ica ion o he alliances we e
ex ac ed om he da abases o Ai T anspo a ion In elligence, Business-Alliance
Su ey, and In e na ional Ci il A ia ion O ganiza ion (ICAO) co esponding o 2011,
acili a ed by Fligh global, and a ime se ies upda ed o 2012 was used o o he so s o
ope a i e, inancial, and a ic da a.
The ela ional da a ha pe mi ed he cons uc ion o he alliance po olio we e
sou ced om he Business-Alliance Su ey. This in o ma ion implies a da a se ha has
been used in con en ional wo ks and has been upda ed e e y yea o o e a decade
(Gimeno, 2004). As hese da a sou ces a e many and a ied, hey we e compa ed wi h
he p og ammed codesha es in GDS, published in he Ai line Rou e epo o Augus 8,
2011. The cong uency o he codesha e alliances in he s udy was almos comple e. The
da a on he des ina ions we e ob ained om he indi idual eco ds o each ai line
included in he Fligh global da abase. The ope a ional, inancial, and a ic da a we e
consul ed om hose same eco ds bu wi h a 2-yea di e ence, which had been
ob ained om ICAO and om he annual epo s o he ai lines. Comple e da a we e
a ailable, bu he inancial in o ma ion p esen ed he e is s anda dized in dolla s as hey
a e published in he local cu ency (indica ing he exchange a e a each poin in ime).
O he his o ic da a on he ai lines (employees, des ina ions, and ypes o ai line) we e
also indi idually consul ed in hose eco ds. In a ew cases, whe e he da abase
p esen ed no up- o-da e o eliable da a on he numbe o employees and on he
pa icula des ina ions o some small pa ne s o he 135 ai lines unde analysis, hose
pa ne s we e no conside ed as hey ep esen ed only a e y small pa o he 1700
alliances unde analysis.
3.3. Va iables
3.3.1. Dependen a iable
1) Resul s: Ope a ing esul s a e an acknowledged e e ence in he ai line sec o and
we e used as an indica o o s udy pe o mance. Ope a ing esul s we e measu ed in a
simila way o Casanue a e al. (2013). The companies’ o al u no e h oughou 2009–
2012 was used o calcula e esul s exp essed in mone a y and physical e ms (Assa and
Josiassen, 2012). We selec ed 2012 because ha yea e lec s he alliance pe o mance
in ea lie yea s. The used indica o s we e as ollows:
a) Mingpaskm: income pe passenge anspo ed o e kilome e s lown. This a io
is o en employed o exp ess demand o each company in he ai line indus y. I
exp esses he income gene a ed by passenge a ic.
b) Load ac o : his a iable indica es he passenge s pe kilome e lown in ela ion
o a ailable sea kilome e s. The e o e, i implies a capaci y u iliza ion measu e in he
indus y (Lazza ini, 2007).
c) Ope a ing esul (MResOpe ): yielded by he di e ence be ween income and
ope a ing cos s. These come om i s own business ac i i y.
3.3.2. Independen a iables
1) He e ogenei y: He e ogenei y is a measu e o he di e ences be ween he ocal
i m’s cha ac e is ics and each one o hei pa ne s’ cha ac e is ics. I may be measu ed
h elDes ina ions -> HETEROGENEITY
0.610
h elFlee -> HETEROGENEITY
0.528
h elSize -> HETEROGENEITY
0.729
Ha ing analyzed he measu emen model, we mo ed on o he s uc u al model.
Fi s , we analyzed cons uc mul icollinea i y. In Table 6, we see VIF s a is ic o he
cons uc s. In his case, VIF alue should be below 5 (Hai e al., 2014). All cons uc s
complied wi h his c i e ion.
Table 6: S uc u al model VIF alues
COMPLEMENTARITY
RESULTS
COMPLEMENTARITY
2.207
complemen a i y2
2.186
DIVERSITY
1.006
1.566
di e si y2
1.288
HETEROGENEITY
1.006
1.250
he e ogenei y2
1.401
RESULTS
age
1.131
lee
1.138
size
1.540
Finally, we e alua ed he pa h coe icien s ha showed he ela ions be ween he
a iables. To do so, boo s apping was applied o 5000 samples (Hai e al., 2014). The
esul s a e gi en in Table 7. The explained a iance (R2) eached a le el o 0.293 o
complemen a i y and 0.470 o he esul s. Those alues a e abo e he minimum o 0.1,
as indica ed by some au ho s (Chin, 1998; Falk & Mille , 1992), and e en 0.25 ha Hai
e al (2014) ecommended.
Wi h ega d o he con ol a iables, he only one ha showed a signi ican and
posi i e ela ion wi h he esul s (β = 0.454 and p = 0.000) was size. This esul means
ha ai lines imp o e hei esul s in ela ion o hei po olio size, in o he wo ds, when
he numbe o pa ne s inc eases.
I we analyze he linea ela ions, nei he he e ogenei y (H1) no di e si y (H2) has a
signi ican ela ion wi h he esul s. Mo eo e , he e ogenei y is no associa ed wi h
complemen a i y in a signi ican way (H4a). We do see a signi ican and posi i e
ela ion (β = 0.394 and p = 0.000) be ween di e si y and complemen a i y (H4b),
indica ing ha he g ea e he di e ences be ween he pa ne s ha cons i u e he
alliance po olio, he highe he p obabili y o ob aining complemen a y esou ces.
Finally, complemen a i y is he only a iable ha shows a signi ican linea ela ion
wi h he esul s (β = 0.512 and p = 0.000), indica ing ha he mo e complemen a y he
alliance po olio esou ces, he be e he i m pe o mance (H3).
Table 7: Pa h coe icien s
Pa h
coe .
s a is ics
P Values
COMPLEMENTARITY ->
0.512
5.003
0.000
RESULTS
compleman a i y2 -> RESULTS
0.124
2.150
0.032
DIVERSITY ->
COMPLEMENTARITY
0.394
3.849
0.000
DIVERSITY -> RESULTS
−0.026
0.270
0.787
di e si y2 -> RESULTS
0.025
0.347
0.728
HETEROGENEITY ->
COMPLEMENTARITY
−0.342
1.174
0.240
HETEROGENEITY ->
RESULTS
0.149
0.872
0.384
he e ogenei y2 -> RESULTS
−0.011
0.223
0.823
age -> RESULTS
0.157
1.494
0.135
lee -> RESULTS
−0.050
0.615
0.538
size -> RESULTS
0.454
4.949
0.000
I we analyze he nonlinea ela ions, we see ha complemen a i y is he only one
ha con inues i s signi ican ela ion wi h he esul s (β = 0.124 and p = 0.032). As he
sign is posi i e, we ha e a U-shaped ela ion. This esul shows ha complemen a i y
be ween he ocal i m esou ces and i s pa ne s’ esou ces gene a es be e esul s o
ha i m i ha complemen a i y is qui e small o qui e la ge, whe eas in e media e
le els o complemen a i y a ec he esul s o a lesse deg ee.
Mo eo e , a mo e de ailed analysis o he ela ions in he model allows us o analyze
he media ing ole o complemen a i y in he ela ions be ween pa ne he e ogenei y
and alliance po olio di e si y wi h ega d o he i m pe o mance. To do so, we used
he analysis sugges ed by Cas o & Roldán (2013) and Real, Roldán & Leal (2014). In
his way, he alues a e examined o see whe he hose ob ained in he boo s apping
p ocess o 5000 subsamples o he indi ec ela ions (yielded by mul iplying he alues
ob ained in he ela ion he e ogenei y esul s by complemen a i y esul s and di e si y
esul s by complemen a i y esul s) aking 0 as a alue, a a con idence le el o 95%, o
which pu pose pe cen iles o 2.5 and 97.5 we e applied (Chin, 2010; Hayes, 2009;
Williams & MacKinnon, 2008). In he case whe e some o hese in e als con ain a
nega i e numbe , i means ha i is inclusi e o he alue 0, so he ela ion o
in e media ion is no signi ican . The da a show ha his condi ion is only me in he
ela ion in which alliance po olio di e si y and esul s a e media ed by ne wo k
esou ce complemen a i y, ha ing an es ima ed alue o 0.202, whe eas he pe cen ile
2.5 assumed a alue o 0.089 and he pe cen ile 97.5 assumed a alue o 0.265. We
managed o demons a e hypo hesis 5b, al hough 5a did no ha e su icien suppo .
Finally, indica o 2 o he cons uc s ha we e signi ican ob ained he ollowing
alues: 0.224 o complemen a i y esul s, 0.060 o complemen a i y2 esul s, 0.218 o
di e si y complemen a i y, and 0.252 o size esul s. In line wi h Cohen (1988), alues
o be ween 0.15 and 0.35 indica e a mode a e size e ec . The e o e, only he quad a ic
ela ion o alliance po olio di e si y wi h ne wo k esou ce complemen a i y has a
small e ec . This esul shows a g ea e explana o y powe o he linea ela ions a he
han he nonlinea ela ions included in he model.
5. Conclusions and discussion
The analysis o he 135 global ai lines alliance po olios con igu ed on he basis o
hei codesha e alliances has p o ided an insigh in o he consequences o he h ee
di e gence sou ces o hese alliance po olios. Mo eo e , di e gence sou ces a e
linked o ne wo k esou ces and i m pe o mance. In pa icula , he analysis has cha ed
he way in which pa ne he e ogenei y, alliance po olio di e si y be ween, and key
ne wo k esou ce complemen a i y in codesha e alliances (des ina ions) a ec s he
ai line ope a ing esul s. These esul s ha e shown suppo o some, bu no all, o he
ela ions p esen ed in he hypo heses.
Fi s , i has been con i med ha he ne wo k esou ce complemen a i y be ween
pa ne s in luences be e pe o mance achie emen a an ope a ional le el (H3). The
in luence o complemen a i y in he esul s is g ea e i ne wo k esou ce
complemen a i y is qui e low o qui e high, as shown in he analysis o he quad a ic
e ec s. These esul s con ibu e new e idence o he deba e abou whe he a be e
pe o mance is linked o ob aining pa ne edundan esou ces ha a e supplemen a y
o hose he ocal i m al eady has o , on he con a y, whe he i is linked o ob aining
complemen a y esou ces ha p oduce syne gies. The esul s sugges ha some
po olios unc ion well when simila esou ces a e accumula ed o hose ha he ocal
i m al eady owns. This is consis en wi h he p e ious a gumen s ha pa ne s ha gi e
he i m a highe le el o con idence and eliabili y a e sough (Gula i, 2007; Lee,
2007). Ano he ac o ha can explain he cu ilinea o m o he ela ion is ha he
i ms can use link alliances (in which he complemen a y esou ces will be impo an ;
his kind o alliance i s in ou dyadic alliance be ween he e ogeneous pa ne s o ou
di e se alliance po olio) o scale alliances (in which he i ms end o ally hemsel es
wi h i ms ha ha e simila esou ces; his ype o alliance co esponds wi h ou dyadic
alliance be ween homogeneous pa ne s o ou simila alliance po olio) (Dussauge e
al., 2000). Acco ding o a p e ious a gumen , Ga e e, Dussage, Cas añe and Mulo e
(2007) sugges ha domain-speci ic know-how gene a es be e pe o mance in scale
alliances. The e o e, ou esul s sugges ha i ms should be specialized in one ype o
alliance. Tha can be explained because companies each hei own bes pe o mance
when hey a e specialized in homogeneous (simila ) pa ne s—sha ing edundan o
supplemen a y esou ce—o he e ogeneous (di e se) ones—sha ing complemen a y
esou ces. In o he wo ds, ou esul s indica e ha he bes op ion is o ha e an alliance
po olio p o ile specialized ei he in link alliances o in scale alliances. The decision
will depend on he alliance’s p e ious expe ience.
Howe e , i may also be because ma ke esou ces ha e been used and, due o hese
so s o esou ce cha ac e is ics, ha i is posi i e o gain bo h supplemen a y and
complemen a y esou ces. The e o e, ai lines will be in e es ed in gaining hei pa ne ’s
new des ina ions no only o comple e hei ope a ional ne wo k bu also o access he
des ina ions, h ough pa ne esou ces, ha hese ai lines al eady ha e bu which
inc ease he ligh equency and he possibili ies o onwa d connec ions. Mo eo e , he
esul s also showed ha a e y high pe cen ages o complemen a i y, he esul s also
imp o e, e e ing di ec ly o he syne gy and he a gumen s ad anced in a la ge pa o
he p e ious li e a u e (Chung e al., 2000; De Leeuw e al., 2014; Kale & Singh, 2009;
Wassme , 2007; Zheng e al., 2013).
Wi h ega d o he e ogenei y, i was possible o nei he demons a e ha he bes
ope a i e esul s a e p oduced when he pa ne s di e g ea ly om he ocal i m no
demons a e ha g ea e pa ne he e ogenei y will achie e, as a esul , mo e
complemen a y ne wo k esou ces. Bo h hese ela ions ha e been emphasized in a la ge
pa o he p e ious li e a u e. Fu he mo e, in line wi h many s udies in he li e a u e,
which associa e he alliance po olio di e si y wi h g ea e ne wo k esou ce
complemen a i y and s a e ha his di e si y leads o a be e pe o mance, he da a on
he alliance po olio ob ained om he codesha e alliances showed no di ec ela ion
be ween alliance po olio di e si y and pe o mance; howe e , he e ec s o media ion
do appea , which poin o he indi ec in luence o di e si y on pe o mance h ough
access o complemen a y ne wo k esou ces (H4b and H5b). These esul s highligh he
cen al ole o ob aining pa ne esou ce complemen a i y as a pi o al elemen in he
alliance po olio cons uc ion.
The main con ibu ions o his wo k a e wo- old. Fi s , he di e ences be ween a
ocal i m and i s pa ne s in cha ac e is ics—pa ne he e ogenei y—a ec company
pe o mance in a posi i e way no he key ne wo k esou ce ha a e complemen a y o
he ocal i m esou ce, whe eas he di e si y o a ie y be ween pa ne s wi hin a
po olio—alliance po olio di e si y—posi i ely in luence bo h pe o mance (in an
indi ec manne ) and he complemen a i y ne wo k esou ce achie emen . This idea
sugges s ha he e ec o di e gences o di e ences is mo e powe ul and posi i e a
he egone le el han a he dyadic ela ion le el wi h he ocal i m. The e o e, he
a gumen s linked o he egone s uc u e should p e ail in he analysis o he alliance
con igu a ion (Bu , 1992; Gula i, 1995). Second, he mo e complemen a y he alliance
esou ces, he be e he i m pe o mance. I appea s ha he ne wo k esou ce
complemen a i y has an impo an weigh age in he ela ion. In o he wo ds, a ele an
ac o o he pe o mance imp o emen in he alliance po olio con igu a ion is o
achie e access o ne wo k esou ces ha a e complemen a y o he i m esou ce
endowmen and ha may be achie ed wi h a mo e di e se and a ied alliance po olio.
This obse a ion implies si ua ing he ne wo k esou ces a he cen e o he analysis o
he po olio con igu a ion, p oposing ha he ac o s linked o he pa ne
cha ac e is ics and o hei own esou ces can ha e a mo e impo an e ec on
pe o mance compa ed o o he s ha we e analyzed mo e widely by Ne wo k Theo y,
such as ne wo k s uc u e and he pa ne ela ionship in ensi y (Gula i, 2007; La ie,
2008). P e ious s udies ha used he same de ini ion o ne wo k esou ces (La ie,
2008) ha e shown ha hei access imp o ed he company’s pe o mance (Casanue a e
al., 2013; La ie, 2007), bu hey did no include any analysis o whe he hose ne wo k
esou ces we e complemen a y o he ocal i m’s esou ces.
In line wi h he s udies by Kim (2014), Wassme (2007) o Zheng e al. (2013), hese
esul s suppo he idea ha complemen a i y in luences he economic pe o mance
achie emen in a posi i e way. In addi ion, i has an indi ec in luence on he ela ion
be ween di e si y and esul s because i is impo an o possess di e en and
complemen a y esul s, as Lin (2012) and Sakakiba a (1997) ha e a gued.
The e o e, as main ained by he p oposed hypo heses, he da a ha e shown ha
ha ing a e y di e se alliance po olio in no way implies excessi e expendi u e on
asks ela ed o coo dina ion o o e come all di icul ies (De Leeuw e al., 2014;
Duys e s & Lokshin, 2011; Faems e al., 2010; Lahi i & Na ayanan, 2013; Pa k e al.,
2014).
This wo k has shown ha he physical and ma ke esou ce complemen a i y in
alliances may, in an accep able way, combine associa ions wi h highly di e gen
pa ne s. These esul s se ou he need o esea ch o he indus ies, wi h physical
esou ces and esou ces ela ed o knowledge and inno a ion. Unlike ea lie s udies,
alliance po olio di e si y and pa ne he e ogenei y wi hin he same indus y ha e
been s udied by analyzing subsec o membe ship o segmen s a he han in e indus y
di e ences.
A a heo e ical le el, ou esea ch has con ibu ed o he be e knowledge o
a ious impo an issues aised in he s a egic li e a u e. Fi s , a concep ual cla i ica ion
has been made o he h ee p incipal sou ces o po olio di e gence ha a e linked o
ne wo k esou ces. This will mean ha we may ad ance wi h he p oposed esea ch
agenda in a mo e o de ly manne (Gula i e al., 2011; Wassme , 2010). Second, he
implica ions o he s udy ha e poin ed owa d si ua ing alliance managemen wi hin a
sui able analysis o he ne wo k esou ces ha a e sough and sha ed wi hin he alliance
(Gula i, 2007; La ie, 2008). An accep able analysis o ne wo k esou ces wi h ega d o
ypology, endowmen , accumula ion, access, and mobiliza ion by pa ne s will pe mi
he maximum exploi a ion o bo h syne gies and complemen a i y, he eby imp o ing
i m pe o mance and achie ing compe i i e ad an ages (Gula i e al., 2011). Thi d, as
his wo k has highligh ed, he conside a ion o physical and ma ke esou ces can be
e y ele an when s udying i m beha io in e ms o hei in e o ganiza ional ies, as
opposed o a la ge pa o he li e a u e ha has ocused (and i has gene a ed heo y) on
in angible esou ces.
Ou s udy also has implica ions o p ac ice. Acco dingly, on he one hand, i ms
should o m alliances wi h o he companies unlike hemsel es o ob ain be e
pe o mance, and on he o he hand, hey should sea ch o pa ne s endowed wi h
complemen a y esou ces o imp o e hei pe o mance. This ecommenda ion se s ou
he need o global alliance po olio managemen o ha e a place in he i m, whe he
h ough line manage s o speci ic depa men wo k, simila o ha in some i ms and
pa icula ly in he indus y unde s udy (De Man, Roijakke s, & De G aauw, 2010; Dye
e al., 2001; Faems e al., 2010; Kale e al., 2002).
The s udy also p esen s a se ies o limi a ions. Fi s , only one esou ce has been
analyzed o s udy complemen a i y. Despi e i being a e y ele an indus y esou ce
(Wassme & Dussauge, 2012), which is he p incipal mo i e o o ming he ype o
alliances unde conside a ion, i.e., codesha e, he in o ma ion ha ela es o he a iable
ne wo k esou ce complemen a i y has, i appea s, been excessi ely in luenced by his
ele ance. Second, he indica o s used o measu e he h ee concep s (pa ne
he e ogenei y, alliance po olio di e si y, and ne wo k esou ce complemen a i y) ha
ha e been analyzed we e he esea che ’s choice. O he i m cha ac e is ics and o he
ypes o measu es o con i m di e ences and simila i ies may be conside ed. Equally, a
u he limi a ion o ou esea ch is ha we ga e no conside a ion o o he alliances
excep codesha es so ha each ai line alliance po olio unde s udy is eally comple e.
Ano he limi a ion is ha mos o he he e ogenei y and di e si y indica o s a e ela ed
o i m size (numbe o employees, numbe o des ina ions, and lee size), and hey a e
co ela ed. Finally, he chosen indus y is complex as i b ings oge he e y di e en
i ms and hei pa icula ci cums ances abou compe ence and globaliza ion, which
may pe haps be e lec ed in he esul s.
Fu u e lines o esea ch will mainly seek o ackle hese limi a ions. Fo example,
ou s udy could be epea ed in ano he indus y—o he han ai lines—and in a pa icula
ai line company segmen o es he esul s. A second esea ch line would be he
conside a ion o o he ypes o in angible esou ces, as new a gumen s may a ise o
es ablish di e ences in he ela ions, hus paying a en ion o he ype o esou ces unde
conside a ion and con i m how he esul s ob ained om he con ex unde analysis i
in wi h he p e ious li e a u e. Mo eo e , o he indica o s could also be used o he
a iables ha we e employed. In addi ion o seeking al e na i e means o measu emen ,
i would be in e es ing o ind a way o measu ing he pa ne ’s willingness o join he
alliance po olio. Ano he limi a ion was ha we wo ked by aking seconda y da a in o
accoun . An in e es ing esea ch line would be he inclusion o o he ypes o alliances,
whe he ho izon al, such as equen lye p og ams, o e ical, and alliances wi h i ms
om o he sec o s. This app oach would pe mi new di e gence sou ces o be included
in he alliance po olio as done in p e ious esea ch (Duys e s & Lokshin, 2011; La ie
& Mille , 2008) o include he in e es ing ques ion o mul iplexi y in he alliance
po olio (Cui & O’Conno , 2012; Sa ka e al., 2009). Finally, i would be in e es ing o
include o he ways o conside ing he di e ences p oposed by ne wo k app oach, such
as s uc u al he e ogenei y (Ahuja e al., 2009), cul u al dis ances (La ie & Mille ,
2008), and mul iplexi y in s a egic alliances (Cui & O’Conno , 2012).
6. Acknowledgmen
This esea ch was suppo ed by he Minis e io de Economía y Compe i i idad
(ECO2013-45329-R) and he Minis e io de Educación, Cul u a y Depo e (P og ama de
Ayudas de Fo mación del P o eso ado Uni e si a io), Spain.
7. Re e ences
Abuzaid, A.N.A. (2014). The Impac o S a egic Alliance Pa ne Cha ac e is ics on
Fi ms' Inno a ion: E idence om Jo dan. In e na ional Jou nal o Business and
Managemen , 9(3), 77-87.
Ahuja, G. (2000). The duali y o collabo a ion: Inducemen s and oppo uni ies in he
o ma ion o in e i m linkages. S a egic Managemen Jou nal, 21(3), 317-343.
Ahuja, G., Polido o, F., & Mi chell, W. (2009). S uc u al homophily o social
asymme y? The o ma ion o alliances by poo ly embedded i ms. S a egic
Managemen Jou nal, 30(9), 941-958.
Assa , A.G., & Josiassen, A. (2012). Eu opean s. U.S. ai lines: Pe o mance
compa ison in a dynamic ma ke . Tou ism Managemen , 33(2), 17-326.
Bagozzi, R.P. (1994). S uc u al equa ion models in ma ke ing esea ch: Basic
p inciples. In R.P. Bagozzi, (Ed.), P inciples o ma ke ing esea ch (pp. 317–385).
Ox o d, UK: Blackwell.
Baum, J.A.C., Calab ese, T., & Sil e man, B.S. (2000). Don' go i alone: Alliance
ne wo k composi ion and s a ups' pe o mance in Canadian bio echnology.
S a egic Managemen Jou nal, 21(3), 267-294.
Beckman, C.M., & Haunschild, P.R. (2002). Ne wo k Lea ning: The E ec s o
Pa ne s’ He e ogenei y o Expe ience on Co po a e Acquisi ions. Adminis a i e
Science Qua e ly, 47(1), 92–124.
Be is, R.A., B adley, S.P., & Hamel, G. (1992). Ou sou cing and indus ial decline. The
Execu i e, 6(1), 7-22.
Blau, P.M. (1977). Inequali y and he e ogenei y. New Yo k, NY: F ee P ess.
Bu , R.S. (1992). S uc u al holes: The social s uc u e o compe i ion. Camb idge,
MA: Ha a d Uni e si y P ess.
Cane , T., & Tyle , B.B. (2013). Alliance po olio R&D in ensi y and new p oduc
in oduc ion. Ame ican Jou nal o Business, 28(1), 38-63.
Casanue a, C., Gallego, A., & Sancho, M. (2013). Ne wo k esou ces and social capi al
in ai line alliance po olios. Tou ism Managemen , 36, 441-453.
Cas o, I., Casanue a, C., & Galán, J.L. (2014). Dynamic e olu ion o alliance
po olios. Eu opean Managemen Jou nal, 32(3), 423-433.
Cas o, I., & Roldán, J. L. (2013). A media ion model be ween dimensions o social
capi al. In e na ional Business Re iew, 22(6), 1034–1050.
Chin, W.W. (1998). The pa ial leas squa es app oach o s uc u al equa ion modeling.
In Ma coulides, G.A. (Ed.), Mode n Me hods o Business Resea ch (pp. 295–
336). Mahwah, NJ: Law ence E lbaum.
Chin, W.W. (2010). How o w i e up and epo PLS analyses. In V.E. Vinzi, W.W.
Chin, J. Hensele and H. Wang (Eds.), Handbook o Pa ial Leas Squa es:
Concep s, Me hods and Applica ions (pp. 655–690). Be lin: Sp inge .
Chin, W.W., & News ed, P.R. (1999). S uc u al equa ion modelling analysis wi h small
samples using pa ial leas squa es. In R. Hoyle, (Ed.), S a is ical s a egies o
small sample esea ch (pp. 307–341). Thousand Oaks, CA: Sage.
Chung, S., Singh, H., & Lee, K. (2000). Complemen a i y, s a us simila i y and social
capi al as d i e s o alliance o ma ion. S a egic Managemen Jou nal, 21(1), 1-
22.
Cohen, J. (1988). S a is ical powe analysis o he beha io al sciences (2nd ed.).
Hillsdale, NJ: E lbaum
Collins, J., & Riley, J. (2013). Alliance Po olio Di e si y and Fi m Pe o mance:
Examining Mode a o s. Jou nal o Business and Managemen , 19(2), 35-50.
C osby, L.A., E ans, K.R., & Cowles, D. (1990). Rela ionship Quali y in Se ices
Selling: An In e pe sonal In luence Pe spec i e. Jou nal o Ma ke ing, 54(July),
68–81.
Cui, A.S., & O'Conno , G. (2012). Alliance Po olio Resou ce Di e si y and Fi m
Inno a ion. Jou nal o Ma ke ing, 76(4), 24-43.
De Leeuw, T., Lokshin, B., & Duys e s, G. (2014). Re u ns o alliance po olio
di e si y: The ela i e e ec s o pa ne di e si y on i m's inno a i e pe o mance
and p oduc i i y. Jou nal o Business Resea ch, 67(9), 1839-1849.
De Man, A.-P., Roijakke s, N., & De G aauw, H. (2010). Managing dynamics h ough
obus alliance go e nance s uc u es: The case o KLM and No hwes ai lines.
Eu opean Managemen Jou nal, 28(3), 171–181.
Dess, G.G., Rasheed, A.M., McLaughlin, K.J., & P iem, R.L. (1995). The new
co po a e a chi ec u e. The Academy o Managemen Execu i e, 9(3), 7-18.
Diaman opoulos, A., & Siguaw, J.A. (2006). Fo ma i e e sus e lec i e indica o s in
o ganiza ional measu e de elopmen : A compa ison and empi ical illus a ion.
B i ish Jou nal o Managemen , 17(4), 263–282.
Dussauge, P., Ga e e, B., & Mi chell, W. (2000). Lea ning om compe ing pa ne s:
Ou comes and du a ions o scale and link alliances in Eu ope, No h Ame ica and
Asia. S a egic Managemen Jou nal, 21(2), 99.
Duys e s, G., De Man, A.P., & Wildeman, L. (1999). A ne wo k app oach o alliance
managemen . Eu opean Managemen Jou nal, 17(2), 182-187.
Duys e s, G., Heime iks, K.H., Lokshin, B., Meije , E., & Sabidussi, A. (2012). Do
Fi ms Lea n o Manage Alliance Po olio Di e si y? The Di e si y-Pe o mance
Rela ionship and he Mode a ing E ec s o Expe ience and Capabili y. Eu opean
Managemen Re iew, 9(3), 139-152.
Duys e s, G., & Lokshin, B. (2011). De e minan s o Alliance Po olio Complexi y and
I s E ec on Inno a i e Pe o mance o Companies. The Jou nal o P oduc
Inno a ion Managemen , 28(4), 570-585.
Dye , J.H. (1996). Specialized supplie ne wo ks as a sou ce o compe i i e ad an age:
E idence om he au o indus y. S a egic Managemen Jou nal,17(4), 271-291.
Dye , J.H., Kale, P., & Singh, H. (2001). How o Make S a egic Alliances Wo k. MIT
Sloan Managemen Re iew, 42(4), 37-43.
Dye , J.H., & Singh, H. (1998). The ela ional iew: Coope a i e s a egy and sou ces
o in e o ganiza ional compe i i e ad an age. Academy o Managemen . The
Academy o Managemen Re iew, 23(4), 660-679.
Eisenha d , K.M., & Schoonho en, C.B. (1996). Resou ce-based iew o s a egic
alliance o ma ion: S a egic and social e ec s in en ep eneu ial i ms.
O ganiza ion Science, 7(2), 136-150.
Faems, D., De Visse , M., And ies, P., & Van Looy, B. (2010). Technology Alliance
Po olios and Financial Pe o mance: Value-Enhancing and Cos -Inc easing
E ec s o Open Inno a ion. The Jou nal o P oduc Inno a ion Managemen ,
27(6), 785-796.
Falk, R.F., & Mille , N.B. (1992). A p ime o so modeling. Ak on, OH: The
Uni e si y o Ak on.
Fo nell, C. (1982). A second gene a ion o mul i- a ia e analysis: an o e iew. In C.
Fo nell, (Ed.), A second gene a ion o mul i a ia e analysis (pp. 1–21). New Yo k,
NY: P aege .
Ga e e, B., Dussauge, P., Cas añe , X., & Mulo e, L. (2007). Do scale alliances wi h
compe i o s imp o e p oduc pe o mance? A s udy o he ai c a indus y, 1949-
2000. Re ie ed om h p://www.hec. /Facul e-Reche che/Publica ions/Cahie s-
de- eche che/(id)/2720
Ge en, D., Rigdon, E.E., & S aub, D. (2011). An Upda e and Ex ension o SEM
Guidelines o Adminis a i e and Social Science Resea ch. MIS Qua e ly, 35(2),
iii–A7.
Gimeno, J. (2004). Compe i ion wi hin and be ween Ne wo ks: The Con ingen E ec
o Compe i i e Embeddedness on Alliance Fo ma ion. Academy o Managemen
Jou nal, 47(6), 820-842.
Goe zen, A., & Beamish, P.W. (2005). The e ec o alliance ne wo k di e si y on
mul ina ional en e p ise pe o mance. S a egic Managemen Jou nal, 26(4), 333–
354.
Golonka, M. (2015). P oac i e coope a ion wi h s ange s: Enhancing complexi y o he
ICT i ms’ alliance po olio and hei inno a i eness. Eu opean Managemen
Jou nal, 33(3), 168–178.
Gomes-Casse es, B. (1994). G oup e sus g oup: How alliance ne wo ks compe e.
Ha a d Business Re iew, 72(4), 62-67.
G an , R.M., & Baden-Fulle , C. (2004). A Knowledge Accessing Theo y o S a egic
Alliances. Jou nal o Managemen S udies, 41(1), 61-84.
Gula i, R. (1995). Social s uc u e and alliance o ma ion pa e ns: A longi udinal
analysis. Adminis a i e Science Qua e ly, 40(4), 619-652.
Gula i, R. (1999). Ne wo k loca ion and lea ning: The in luence o ne wo k esou ces
and i m capabili ies on alliance o ma ion. S a egic Managemen Jou nal, 20(5),
397-420.
Gula i, R. (2007). Managing ne wo k esou ces, Ox o d: Ox o d Uni e si y P ess.
Gula i, R., La ie, D., & Madha an, R. (2011). How do ne wo ks ma e ? The
pe o mance e ec s o in e o ganiza ional ne wo ks. Resea ch in O ganiza ional
Beha io , 31, 207-224.
Gula i, R., Noh ia, N., & Zahee , A. (2000). S a egic ne wo ks. S a egic Managemen
Jou nal, 21(3), 203–215.
Hai , J.F.J , Sa s ed , M., Hopkins, L., Kuppelwiese , V.G. (2014). Pa ial leas squa es
s uc u al equa ion modeling (PLS-SEM): An eme ging ool in business esea ch,
Eu opean Business Re iew, 26(2), 106 – 121.
Hamel, G., Doz, Y., & P ahalad, C.K. (1989). Collabo a e wi h you pa ne s—and win.
Ha a d Business Re iew, 67(1), 133-139.
Hayes, A.F. (2009). Beyond Ba on and Kenny: s a is ical media ion analysis in he new
millennium, Communica ion Monog aphs, 76(4), 408–420.
Hensele , J., Fasso , G., Dijks a, T.K., & Wilson, B. (2012). Analysing quad a ic
e ec s o o ma i e cons uc s by means o a iance-based s uc u al equa ion
modelling. Eu opean Jou nal o In o ma ion Sys ems, 21(1), 99–112.
Hi , M.A., Hoskisson, R.E., & Kim, H. (1997). In e na ional di e si ica ion: E ec s on
inno a ion and i m pe o mance in p oduc -di e si ied i ms. Academy o
Managemen Jou nal, 40(4), 767-798.
Ho mann, W.H. (2007). S a egies o managing a po olio o alliances. S a egic
Managemen Jou nal, 28(8), 827-856.
Huggins, R. (2010). Fo ms o ne wo k esou ce: Knowledge access and he ole o
in e - i m ne wo ks. In e na ional Jou nal o Managemen Re iews, 12(3), 335–
352.
Ja is, C.B., MacKenzie, S.B., & Podsako , P.M. (2003). A C i ical Re iew o
Cons uc Indica o s and Measu emen Model Misspeci ica ion in Ma ke ing and
Consume Resea ch. Jou nal o Consume Resea ch, 30(2), 199–218.
Jiang, R.J., Tao, Q.T., & San o o, M.D. (2010). Alliance po olio di e si y and i m
pe o mance. S a egic Managemen Jou nal, 31(10), 1136-1144.
Kale, P., Dye , J., & Singh, H. (2002). Value c ea ion and success in s a egic alliances:
Alliancing skills and he ole o alliance s uc u e and sys ems. Eu opean
Managemen Jou nal, 19(5), 463-471.
Kale, P., & Singh, H. (2009). Managing S a egic Alliances: Wha Do We Know Now,
and Whe e Do We Go F om He e? Academy o Managemen Pe spec i es, 23(3),
45-62.
Kim, S. (2014). In luence o Fi m and Pa ne Resou ces on Fi m Pe o mance in he
Alliance Po olio. M@n@gemen , 17(2), 88-109.
Kogu , B. (1988). Join en u es: heo e ical and empi ical pe spec i es. S a egic
Managemen Jou nal (1986-1998),9(4), 319-332.
Koka, B.R., & P esco , J.E. (2002). S a egic alliances as social capi al: A
mul idimensional iew. S a egic Managemen Jou nal, 23(9), 795-816.
Lahi i, N., & Na ayanan, S. (2013). Ve ical in eg a ion, inno a ion, and alliance
po olio size: Implica ions o i m pe o mance. S a egic Managemen
Jou nal, 34(9), 1042-1064.
Lane, P.J., & Luba kin, M. (1998). Rela i e abso p i e capaci y and in e o ganiza ional
lea ning. S a egic Managemen Jou nal, 19(5), 461-477.
La ie, D. (2006). The Compe i i e Ad an age o In e connec ed Fi ms: an Ex ension o
he Resou ce-Based View. Academy o Managemen Re iew, 31(3), 638-658.
La ie, D. (2007). Alliance po olios and i m pe o mance: A s udy o alue c ea ion
and app op ia ion in he U.S. so wa e indus y. S a egic Managemen Jou nal,
28(12), 1187-1212.
