British and Spanish entrepreneurial intentions: a comparative study

A ailable in: h p://www. edalyc.o g/a iculo.oa?id=86626373004 Red de Re is as Cien í icas de Amé ica La ina, el Ca ibe, España y Po ugal Sis ema de In o mación Cien í ica Liñán, F ancisco; Nabi, Ghulam; K uege , No is B i ish and Spanish En ep eneu ial In en ions: A Compa a i e S udy Re is a de Economía Mundial, núm. 33, 2013, pp. 73-103 Sociedad de Economía Mundial Huel a, España How o ci e Comple e issue Mo e in o ma ion abou his a icle Jou nal's homepage Re is a de Economía Mundial, ISSN (P in ed Ve sion): 1576-0162 [email p o ec ed] Sociedad de Economía Mundial España www. edalyc.o g Non-P o i Academic P ojec , de eloped unde he Open Acces Ini ia i e e i S a d e ec o n o m í a mu n d i a l 33, 2013, 73-103 ISSN: 1576-0162 B i iSh a n d Sp a n i S h en e p e n e u i a l in e n i o n S : a co m p a a i e S u d y la i n E n c i ó n E m p E n d E d o a E n E i n o un i d o y Es p a ñ a : u n E s u d i o c o m p a a i o F ancisco Liñán Uni e sidad de Se illa linan@us .es Ghulam Nabi Business School, Manches e Me opoli an Uni e si y d _g_nabi@ho mail .com No is K uege En ep eneu ship No hwes (Boise, ID, USA) and Max Planck Ins i u e o Economics (Jena, Ge many) no is .k uege @gmail .com Recibido: oc ub e de 2012; acep ado: ene o de 2013 aB S a c Ex ending om Ajzen’s (1991) heo y o planned beha iou , his pape de- elops a mo e in eg a ed en ep eneu ial in en ion model . This inco po a es he ole o cul u e, along wi h mo i a ions, skills and knowledge o he en e- p eneu ial en i onmen . The c oss-cul u al applicabili y o he model is es ed ac oss wo di e en coun ies, G ea B i ain and Spain, on a o al sample o 1005 . Pa ial Leas Squa es echnique is used o y and o e come limi a ions o p e ious esea ch . The model b oadly holds o bo h coun ies . Implica ions o public decision make s and en ep eneu ship educa ion a e discussed . In pa icula , enhancing he le el o knowledge and awa eness abou en ep e- neu ship would inc ease sel -e icacy pe cep ions and, hence, en ep eneu ial in en ions . Keywo ds: En ep eneu ial In en ion ; C oss-cul u al Analysis ; En ep e- neu ial Knowledge ; En ep eneu ial Skills ; En ep eneu ial Valua ion . e S u m e n Es e abajo, pa iendo del en oque de Acción Plani icada (Ajzen), desa- olla un modelo más in eg ado de in ención emp endedo a, inco po ando el papel de la cul u a, las mo i aciones, habilidades y conocimien o emp esa ial . La aplicabilidad anscul u al del modelo se p ueba analizando dos países di- e en es (G an B e aña y España), en una mues a de 1005 indi iduos . Se u i- lizan Mínimos Cuad ados Pa ciales pa a supe a las limi aciones de an e io es in es igaciones . El modelo en gene al es álido pa a los dos países, de i án- dose implicaciones pa a la oma de decisiones y la educación emp endedo- a . En pa icula , mejo a el conocimien o y la oma de conciencia sob e el emp endimien o aumen a ía la pe cepción de au oe icacia y, po lo an o, las in enciones emp endedo as . Palab as cla e: In ención emp endedo a; Análisis anscul u al; Conoci- mien o emp endedo ; Habilidades emp endedo as; Valo ación emp endedo- a . JEL Classi ica ion: A13, L26, R11 . e i S a d e ec o n o m í a mu n d i a l 33, 2013, 73-103 1 . in o d u c i o n A signi ican body o li e a u e on en ep eneu ship highligh s he impo ance o s udying cogni i e ac o s, such as en ep eneu ial mo i a ion (e .g ., a i udes, pe cep ions) and in en ion, in o de o p o ide insigh s in o he complex p ocess o new en u e c ea ion (Au io, Keeley, Klo s en, Pa ke and Hay, 2001; Fo bes, 1999; K uege , Reilly and Ca s ud, 2000; Pe e man and Kennedy, 2003; Tkache and Kol e eid, 1999) . Such a cogni i e pe spec i e is aluable because i ep esen s an a emp o unde s and he o ma ion o new en u es and he unde lying s uc u es and p ocesses (c . Hen y, Hill and Lei ch, 2003) . While he e ha e been de elopmen s in his ield, he e is a lack o esea ch ac oss di e en cul u es (Fo bes, 1999; Liñán and Chen, 2009) . Indeed, i is ecognized ha “cul u al alues […] will posi i ely in luence he p esence o en ep eneu ial cogni ion on bo h he indi idual and socie al le els” (Fo bes, 1999: 421) . McG a h and MacMillan’s (1992) seminal s udy sugges ed ha en ep eneu s in di e en se ings we e mo e alike han was ob ious . Impo an ly, despi e hese ea lie p o oca i e s udies, he e a e only limi ed heo e ically-d i en empi ical es s o cul u al ac o s in en ep eneu ial in en ion models . I is clea ha mo e c oss-cul u al esea ch is needed o shed u he ligh on he in luence o di e en cul u es and alues on en ep eneu ial in en ion . One o he main aims o his pape is o add ess his gap by using samples om wo di e en coun ies in Eu ope (G ea B i ain –G .B .– and Spain) ha ha e a ac ed li le compa a i e esea ch . This will be achie ed by using an ex ended heo y o planned beha iou (TPB) ha includes cul u al and o he psychosocial a iables (discussed below) in he model . Tha is, in addi ion o he co e TPB mo i a ional elemen s o : pe sonal a i ude, subjec i e no m and pe cei ed beha iou al con ol . The wo chosen Eu opean Union coun ies, G .B . and Spain, do na u ally ha e a dis inc cul u e and he i age, bu hey a e no comple ely di e en in se e al ega ds . Bo h coun ies a e ‘inno a ion d i en’, e lec ing ma u e economies, shi ing owa ds he se ice sec o and ca e ing o an inc easingly a luen popula ion, combined wi h a ocus on knowledge gene a ion and de elopmen o inno a i e, oppo uni y seeking en ep eneu ial ac i i y and a he ime o he s udy had easonably compa able economic pe o mance (Bosma, Acs, Au io, Codu as and Le ie, 2008) . Bo h he UK (o which GB o ms 76 a n c i S c o li ñ á n , Gh u l a m na B i , no i S k u e G e he majo i y) and Spain ha e signi ican popula ions (abou 60 .2 million and 43 .4 million espec i ely based on Wo ld Bank 2005 Da a) . They a e classi ied as high income coun ies wi h o example, o al GDP o he UK and Spain a $2244 .1 billion and $1129 .7 billion espec i ely (Wo ld Bank 2005 da a), and GDP pe capi a (PPP-adjus ed) o UK and Spain o he same yea s a $31580 and $27270 espec i ely (Wo ld Bank 2005 da a) . Bo h coun ies ha e simila le els o nascen en ep eneu s (new businesses o less han 3 mon hs) o a ound 3 .1 and 3 .3% ( o UK and Spain espec i ely), ea ly-s age en ep eneu ial ac i i y (new businesses o 3-42 mon hs) a a ound 5 .9 o 7% (Bosma e al ., 2008) . His o ically bo h coun ies ha e also had simila ea ly- s age en ep eneu ial ac i i y o abou 5-8% (Reynolds, Byg a e, Au io and Hay, 2002) . Thus, despi e ha ing dis inc di e ences in cul u al he i age, he wo coun ies a e easonably compa able o he pu poses o his esea ch . A second aim o he pape is ela ed o ying o d aw oge he some o he s ands in TPB-based in en ion models . The TPB is inc easingly been seen as a use ul heo e ical amewo k in new en u e c ea ion because i helps o explain he complex and in ica e cogni i e p ocesses leading o i m c ea ion (Au io e al ., 2001; K uege e al ., 2000; Liñán, 2008) . Au ho s sugges he TPB-a i ude app oach p o ides g ea e p edic i e capaci y in explaining en ep eneu ial in en ion and beha iou o e some p e ious app oaches, such as, he pe sonali y ai o demog aphic app oaches (Au io e al ., 2001; Mo iano, Go gie ski and Lukes, 2008) . Fu he , esea che s sugges ha in en ion models a e mo e ui ul because unde lying an eceden s such as a i udes owa ds en ep eneu ship can be modi ied wi h educa ional in e en ions (Liñán, Rod íguez-Coha d and Rueda, 2011a; Pe e man and Kennedy, 2003) . No wi hs anding such heo e ical ad an ages, TPB-based wo k can be iden i ied as ha ing 4 di e en s ands . Tha is, (1) he e ec o co e TPB elemen s on en ep eneu ial in en ion (e .g ., Kol e eid, 1996; K uege e al ., 2000); (2) he e ec o human and social capi al (e .g ., en ep eneu ial skills, social ela ionships) on he TPB elemen s and EI (e .g ., Chen, G eene and C ick, 1998; Liñán and San os, 2007); (3) he e ec o knowledge and awa eness o he en ep eneu ial en i onmen on TPB elemen s and EI (e .g ., Liñán, 2008; Lu hje and F anke, 2003; Sche e , B odzinsky and Wiebe, 1991); and/o (4) c oss-cul u al esea ch (e .g ., Au io e al ., 2001; Mo iano, Go gie ski, Laguna, S ephan and Za a shani, 2012) . Taken collec i ely, his body o wo k has p o ided use ul insigh s in o, and sound e idence o , he TPB model . Howe e , i has a guably been agmen ed, wi h he g ea e sha e o he wo k unde he i s s and, and he leas unde he hi d and ou h s ands . O he issues wi h he in en ions model ha e also su aced (e .g ., he need o conside he dynamics and looking a deepe unde pinnings) wi h calls o mo e sophis ica ed models (K uege , 2009; K uege and Day, 2010) . The e is a lack o wo k ha has a emp ed o combine hese s ands in o a mo e in eg a ed model . A inal aim o his pape is o add ess some limi a ions o p e ious esea ch . Mos pas esea ch in TPB [no jus o en ep eneu ship] has used linea 77 e i S a d e ec o n o m í a mu n d i a l 33, 2013, 73-103 B i iSh a n d Sp a n i S h en e p e n e u i a l in e n i o n S : a co m p a a i e S u d y eg ession models (e .g ., Au io e al ., 2001; Tkache and Kol e eid, 1999), despi e he isk o missing indi ec and complex e ec s . Hence, in his esea ch, s uc u al equa ion modelling is used o allow o mo e obus , sophis ica ed and ecu si e es ing in o de o be e unde s and hese e ec s . This pape has implica ions o educa o s, policy make s and esea che s by d awing oge he p e ious published wo k . I will also ac as a compa ison in wo di e en cul u es, allowing o a obus es o a model ha may help o explain en ep eneu ial in en ion in di e en con ex s . This could be use ul o policy make s o unde s and no only he pa e n o ela ionships among in en ion an eceden s, bu also i s implica ions o in e en ions and de eloping en ep eneu ial in en ion . The pape is s uc u ed a ound he ollowing ou main sec ions: a li e a u e e iew p esen ing a concep ual amewo k and e lec ing on p e ious esea ch o unde pin he model and hypo heses; a me hodology sec ion o explain he sample and measu es used; a esul s sec ion epo ing on he s uc u al equa ion models om he c oss-coun y analysis; and a inal sec ion o discuss and conclude he pape . 2 . h e o y a n d h y p o h e S e S En ep eneu ial in en ion can be de ined as a conscious awa eness and con ic ion by an indi idual ha hey in end o se up a new business en u e and plan o do so in he u u e (see Bi d, 1988; Thompson, 2009) . The p ocess o s a ing a new i m can hus be ega ded as olun a y wi h conscious in en ionali y . Impo an ly, in en ion has been conside ed as he single mos powe ul p edic o o en ep eneu ial beha iou s (Au io e al ., 2001; K uege e al ., 2000), and also an impo an dependen a iable in i s own igh (Thompson, 2009) . Applied o en ep eneu ial in en ion, he TPB posi s h ee mo i a ional an eceden s ha ac as p ecu so s o new en u e c ea ion . These can be de ined in he ollowing way . Pe sonal a i ude (PA) owa ds becoming an en ep eneu e e s o he ex en o posi i e alua ion abou he s a -up o a new en u e . Subjec i e no m (SN) e lec s he p essu e and app o al om signi ican o he s o becoming an en ep eneu , hus aking in o accoun he indi idual’s social con ex . Pe cei ed beha iou al con ol (PBC) measu es he pe cei ed ease o di icul y o becoming an en ep eneu . Resea ch has gene ally p o ided suppo o he TPB in he con ex o en ep eneu ship (e .g ., K uege e al ., 2000), bu he applica ion and empi ical es o he TPB model o di e en cul u es has been su p isingly limi ed . Because o c oss-coun y di e ences in en ep eneu ial ac i i y, i is o heo e ical and p ac ical impo ance o examine how cul u e ela es o le els o en ep eneu ial ac i i y (Hay on, Geo ge and Zah a, 2002) . This would include en ep eneu ial in en ion as his o ms pa o he en ep eneu ial p ocess . Cul u e can be de ined as a se o sha ed belie s, alues and expec a ions 78 a n c i S c o li ñ á n , Gh u l a m na B i , no i S k u e G e (Hay on e al ., 2002) . Simila ly, Ho s ede (2003: 9) de ines i as “ he collec i e p og amming o he mind ha dis inguishes membe s o one g oup o ca ego y o people om ano he ” . Impo an ly, cul u al alues p o ide an indica o as o he ex en o which a socie y iews en ep eneu ial ac i i y as a ac i e o no . Cul u es ha alue and ein o ce such en ep eneu ial ac i i y may acili a e mo e isk- aking and adical inno a ion, whe eas hose ha alue and ein o ce con o mi y and con ol o e he u u e a e less likely o display en ep eneu ial ac i i y (Hay on e al ., 2002) . Fo example, some coun ies e .g ., USA, encou age a can-do isk- aking a i ude whe eas o he s do no (Hen y e al ., 2003; S ephan, 2008) . Cul u e should in luence en ep eneu ship in wo main ways: Bo om-up e sus op-down . The i s way ( he ‘agg ega e ai ’ explana ion) sugges s ha i a coun y has mo e indi iduals wi h en ep eneu ial alues and ai s, mo e indi iduals will become en ep eneu s, a iew ha sugges s alues a e agg ega ed . The second way ( he ‘socie al legi ima ion’ explana ion) sugges s ha a highe le el o mo al app o al o legi imiza ion o en ep eneu ship wi hin a cul u e is e lec ed in ha socie y’s p ac ices . Fo example, paying mo e a en ion o en ep eneu ship in educa ion, conside ing en ep eneu ship desi able, and allowing a mo e a ou able and suppo i e s a -up en i onmen , a iew ha highligh s cul u e is a unc ion o socie al p ac ices (S ephan, 2008) . I we a e o unde s and he p ocesses o in en ions o ma ion, i has become clea ha we need o look mo e closely a he deepe belie s ha se e as ancho ing assump ions o en ep eneu ial decision making . Cul u al and social no ms can be a key sou ce o hose deep ancho ing belie s (K uege , 2007) . The e is a pauci y o esea ch looking a en ep eneu ial mo i a ion and in en ion ac oss di e en cul u es . This is impo an as he ela i e e ec s o TPB elemen s on en ep eneu ial in en ion could be di e en in di e en coun ies . Recognizing his, Liñán and Chen (2009) called o mo e c oss- cul u al esea ch in his ield wi h wide samples om di e en coun ies o be es ed . The e is, howe e , limi ed a ailable esea ch on he applica ion o he TPB o di e en cul u es . Au io e al . (2001) examined he TPB co e cons uc s amongs uni e si y s uden s om he USA, Finland and Sweden, bu hey did no look a he in luence o cul u al a iables di ec ly . In o he wo ds, hey did no look a social alues, such as close en i onmen alua ions (e .g ., o signi ican o he s in he indi idual’s close en i onmen ), based on he social capi al li e a u e, no social alua ions (e .g ., wide cul u al alues o socie y) o en ep eneu ship in he chosen coun ies . This is impo an because such social and cul u al alues can in luence he TPB mo i a ional elemen s (Ajzen, 2001) and his is discussed la e wi h e e ence o ou wo chosen coun ies . In a mo e in eg a ed a emp , as men ioned ea lie , Liñán (2008) de eloped and es ed an en ep eneu ial in en ion model on a Spanish sample . This inco po a ed he a o emen ioned cul u al a iables o social alua ion (SV) 79 e i S a d e ec o n o m í a mu n d i a l 33, 2013, 73-103 B i iSh a n d Sp a n i S h en e p e n e u i a l in e n i o n S : a co m p a a i e S u d y and close en i onmen alua ion (CV), along wi h en ep eneu ial skill pe cep ions as hey could make indi iduals eel mo e capable o s a ing a new en u e . None heless, esea che s ha e also no ed he impo ance o wo o he conside a ions (Ajzen, 1991; Au io e al ., 2001; Bosma e al ., 2008; Liñán, 2008; Liñán, Ba is elli and Mo iano, 2008; Lu hje and F anke, 2003; Sche e e al ., 1991) . The i s is he ele ance o a g ea e knowledge o he en ep eneu ial en i onmen (e .g ., knowing abou sou ces o en ep eneu ial knowledge and suppo assis ance) . The second is he impo ance o a c oss- cul u al pe spec i e o be e unde s and he e ec o di e en cul u al en i onmen s on en ep eneu ial mo i a ion and in en ion . In e es ingly, he mos ecen analysis o he Global En ep eneu ship Moni o da a ound wo key p edic o s o en ep eneu ial ac i i y: pe sonal p epa edness o en u e and social/cul u al no ms, he wo elemen s jus desc ibed (Reynolds, 2011) . The o e all model in his esea ch, hus, d aws on Liñán (2008) bu wi h hese impo an ex ensions ha a e suppo ed by GEM da a . Based on ou model, howe e , some hypo heses a e gene ic and should hold in bo h coun ies . Mo e speci ically, he model sugges s ha PA and PBC should in luence in en ion, ega dless o coun y . In c oss-cul u al esea ch wi h o he coun ies, his has gene ally been shown o be he case (e .g ., Au io e al ., 2001; Liñán and Chen, 2009) . The case o SN is mo e in iguing and ypically he weakes p edic o o in en ion . S udies ha e ound a weak o no di ec e ec be ween SN and in en ion (Au io e al ., 2001; K uege e al ., 2000) . Ra he , ecen a gumen and e idence sugges s ha he e ec o SN may be indi ec ia PA and PBC, and his has been epo ed in Spain and Taiwan (Liñán, 2008; Liñán and Chen, 2009) . The e o e, ega dless o he coun y, he i s se o hypo heses is: H1a . Pe sonal a i ude posi i ely in luences in en ion H1b . Pe cei ed beha iou al con ol posi i ely in luences in en ion . H1c . Subjec i e no m will ha e a posi i e impac on pe sonal a i ude . H1d . Subjec i e no m will ha e a posi i e impac on pe cei ed beha iou al con ol . Gi en simila le els o (pe cei ed) knowledge and skills in B i ain and Spain o s a a new en u e (Bosma e al ., 2008, sugges 45% and 43% espec i ely), he coun ies a e no comple ely dissimila in e ms o pe cei ed human capi al . Thus, a numbe o en a i e c oss-cul u al hypo heses can be de i ed om he a ailable li e a u e . Mos esea ch on en ep eneu ship and cul u e has employed Ho s ede’s i e cul u al dimensions (see Ho s ede, 2003) . These a e: indi idualism s . collec i ism (ex en o which people in a socie y p e e o ac as indi iduals a he han in g oups), unce ain y a oidance ( he ex en o which people p e e s uc u ed o e uns uc u ed si ua ions), powe dis ance ( he ex en o which he exis ence o inequali y among people in a coun y is accep ed), masculini y- eminini y ( he ex en o which ough alues like asse i eness and compe i ion p e ail o e ende ones like quali y o li e and ca ing) and long- 80 a n c i S c o li ñ á n , Gh u l a m na B i , no i S k u e G e sho e m o ien a ion ( he ex en o which alues a e ocused owa ds he u u e -like pe sis ence- o he pas and p esen ) . These alues a e impo an o in luencing en ep eneu ial cogni ions . Fo example, a long- e m o ien ed cul u e is likely o gene a e people wi h posi i e en ep eneu ial alues, belie s, a i udes and ac ions (Buseni z and Lau, 1996) . In his esea ch, wo Eu opean bu cul u ally di e en coun ies a e used . Using Ho s ede’s dimensions, he e a e some no ewo hy simila i ies and di e ences . Long-sho e m o ien a ion is ai ly simila (25 o B i ain s . 19 o Spain) . This sugges s simila le els o long e m ocus owa ds u u e ewa ds like pe se e ance and h i . On he o he hand, he e a e subs an ial di e ences o all o he dimensions . Rega ding indi idualism, B i ain sco es much highe han Spain (89 and 51 espec i ely), and appea s much mo e indi idualis ic-ac ing, implying a mo e en ep eneu - iendly cul u e . In ela ion o unce ain y a oidance, B i ain sco es conside ably lowe han Spain (35 and 86 espec i ely) . This sugges s ha B i ish cul u e is ela i ely mo e ole an o unce ain y and eel less h ea ened by unce ain si ua ions . Thus, in his dimension, Spanish cul u e is ela i ely mo e opposed o en ep eneu ship . Wi h ega ds o powe dis ance, again B i ain sco es conside ably lowe han Spain (35 and 57 espec i ely), sugges ing ha , in Spain, inequali y in in e pe sonal ela ionships ha unde lies unc ioning in ha socie y is pe cei ed as mo e na u al . In his dimension, he e o e, Spanish cul u e is again ela i ely less conduci e o en ep eneu ship . Finally, B i ain sco es conside ably highe in masculini y (66 and 42 espec i ely) . This sugges s ha B i ish cul u e (as pa o Anglo coun ies along wi h US) is ela i ely mo e masculine-o ien a ed wi h a highe endency owa ds asse i eness, ea nings, ad ancemen and ‘li e o wo k’ cul u e compa ed o Spain (as pa o La in coun ies, Ho s ede, 2003) . This would imply ha B i ish cul u e is pe haps mo e a ou able owa ds en ep eneu ship . The economic si ua ion has gene ally been a ou able in bo h coun ies since he mid-nine ies up un il ecen ly . GB has, howe e , expe ienced be e economic g ow h and lowe le els o unemploymen . Al hough Spain’s unemploymen has been highe , i has also been dec easing since he mid- nine ies . Thus, he impac o he economic si ua ion can be conside ed o be b oadly compa able and neu al o he pu poses o his s udy . As a as he au ho s a e awa e, he e is no TPB-based esea ch con as ing he e ec s o cul u al di e ences on en ep eneu ial mo i a ion and in en ion o he wo coun ies o in e es . Howe e , he a ailable li e a u e does allow o some en a i e hypo heses o be de eloped ega ding pe sonal a ac ion and pe cei ed beha iou al con ol . Bosma e al . (2008) ound ha pe sonal a i ude owa ds an en ep eneu ial ca ee is ela i ely s onge and mo e desi able in Spain han in B i ain (68% and 52% espec i ely) . A simila inding was epo ed by Uslay, Teach and Schwa z (2002) . They ound ha mo e Spanish s uden s ag eed ha en ep eneu ship o e s job sa is ac ion han hei US coun e pa s . Since B i ain is in he same Anglo g oup (Ho s ede, 2003) as he US, his inding could also be conside ed o B i ish esponden s . This sugges s ha he salien 87 e i S a d e ec o n o m í a mu n d i a l 33, 2013, 73-103 B i iSh a n d Sp a n i S h en e p e n e u i a l in e n i o n S : a co m p a a i e S u d y a B l e 2 . u l l -S a m p l e m e a S u e m e n m o d e l ( eliaBili y indica o S) Cons uc / Indica o Loadings Composi e Reliabili y AVE EI 0 .895 0 .589 A4 0 .7106 A6 0 .7665 A9 ecod 0 .6603 A13 0 .8540 A17 0 .8077 A19 ecod 0 .7901 PA 0 .858 0 .605 A10 0 .8279 A12 ecod 0 .6624 A15 0 .8526 A18 0 .7536 PBC 0 .836 0 .562 A1 0 .7662 A7 0 .7662 A14 0 .8009 A20 0 .6580 SN 0 .877 0 .704 A3 0 .7951 A8 0 .8375 A11 0 .8817 SV 0 .824 0 .700 C2 0 .8524 C6 0 .8208 CV 0 .842 0 .640 C1 0 .7706 C4 0 .8287 C7 0 .7992 SkillSe 0 .871 0 .531 D1 0 .7842 D2 0 .7016 D3 0 .6848 D4 0 .7730 D5 0 .7416 D6 0 .6782 KnowSupp 0 .922 0 .662 F3a 0 .7642 F3b 0 .7895 F3c 0 .8206 F3d 0 .8184 F3e 0 .8662 F3 0 .8208 88 a n c i S c o li ñ á n , Gh u l a m na B i , no i S k u e G e Fo added con i ma ion, A e age Va iance Ex ac ed (AVE) assesses he amoun o a iance ha a cons uc cap u es om i s indica o s ela i e o he amoun due o measu emen e o (Chin, 1998) . I is usually conside ed ha a le el abo e 0 .5 indica es adequa e eliabili y . In his case, all cons uc s had AVEs o e ha le el . Disc iminan alidi y may also be assessed compa ing AVE and he a iance sha ed be ween his cons uc and he o he s in he model . Tha is, he squa ed co ela ion be ween each pai o cons uc s . In his sense, Table 3 p esen s AVE sco es on he main diagonal, oge he wi h squa ed co ela ions . As may be obse ed, AVE sco es a e always highe , indica ing adequa e disc iminan alidi y . a B l e 3 . diSc iminan alidi y EI PA PBC SN SV CV SkillSe K n o w Supp EI 0 .589 PA 0 .588 0 .605 PBC 0 .352 0 .269 0 .562 SN 0 .154 0 .197 0 .140 0 .704 SV 0 .108 0 .106 0 .101 0 .042 0 .700 CV 0 .145 0 .119 0 .077 0 .031 0 .151 0 .640 SkillSe 0 .140 0 .110 0 .235 0 .072 0 .052 0 .069 0 .531 KnowSupp 0 .076 0 .028 0 .120 0 .003 0 .026 0 .052 0 .048 0 .662 Resul s om he measu emen model indica e cons uc s p esen adequa e p ope ies . Rega ding he s uc u al model, he main c i e ion o assess hei adequacy is he coe icien o de e mina ion (R2) o each endogenous la en a iable (Hensele , Ringle and Sinko ics, 2009), which should exceed 0 .10 (Falk and Mille , 1992) . Figu e 2 shows he a iance explained (R2) in he endogenous cons uc s and he pa h coe icien s (β) o all signi ican ela ionships . Consis en wi h Chin (1998), boo s apping (500 esamples) has been used o gene a e s anda d e o s and -s a is ics . Boo s apping ep esen s a non- pa ame ic app oach o es ima ing he accu acy o he PLS es ima es . This allows us o assess he s a is ical signi icance o he pa h coe icien s . Figu e 2 p esen s he esul s o he join sample . To make i clea e , con ol a iables a e no shown . Ins ead, Table 4 p esen s he signi ican coe icien s o hose con ol a iables on each o he cons uc s conside ed . In his sense, he coun y dummy has u ned ou o be e y ele an , signi ican ly a ec ing almos all cons uc s . Only EI is no a ec ed by his a iable . This would mean ha in en ions a e explained by he ele an cons uc s and any di e ence in in en ions by coun y is explained by he di e ences in he o he cons uc s . TPB assumes ha exogenous ac o s ope a e on 89 e i S a d e ec o n o m í a mu n d i a l 33, 2013, 73-103 B i iSh a n d Sp a n i S h en e p e n e u i a l in e n i o n S : a co m p a a i e S u d y in en ion h ough he key an eceden s; hese indings suppo his ‘condui ’ e ec . Addi ionally, i may be said ha he Spanish sub-sample has lowe le els o all o he cons uc s (excep SN), ein o cing he idea ha GB is a mo e en ep eneu ial coun y han Spain . i G u e 2 . e S u l S o Sem o J o i n S a m p l e No e: Only signi ican (p<0 .05) pa hs shown . Dashed lines indica e non-hypo hesized ela ionships . a B l e 4 . SiGni ican c o n o l a i a B l e S o h e J o i n S a m p l e (p<0 .05) EI PA SN PBC Skills se Knowldg . suppo CV SV Age 0 .100 Gende 0 .115 0 .108 0 .073 Yea S d -0 .041 0 .091 Role Model 0 .093 0 .067 0 .129 0 .157 0 .096 LabExp 0 .074 0 .090 0 .099 Sel Empl Exp 0 .052 0 .078 0 .078 Spain -0 .126 0 .075 -0 .155 -0 .083 -0 .272 -0 .156 -0 .289 90 a n c i S c o li ñ á n , Gh u l a m na B i , no i S k u e G e I we now compa e he esul s om each sub-sample, we can see some ele an di e ences . Figu e 3 p esen s hem . Solid a ows ep esen pa h coe icien s signi ican in bo h sub-samples . Meanwhile, do ed lines a e used when he pa h was only signi ican in one sub-sample . i G u e 3 . co m p a a i e e S u l S o Sem o e a c h S u B -S a m p l e No e: Only signi ican (p<0 .05) pa hs shown . The i s igu e co espond he GB sample; he second o ha o Spain . Do ed lines indica e ela ionships signi ican only in one sub-sample . Finally, he mul ig oup analysis was ca ied ou o es whe he he e we e any s a is ically signi ican di e ences among bo h sub-samples wi h espec o pa h coe icien s (Chin, 2000) . In acco dance wi h his p ocedu e, a -s a is ic has been calcula ed (equa ion 1 in Appendix 2), which ollows a -dis ibu ion wi h m + n – 2 deg ees o eedom, Sp (equa ion 2 in Appendix 2) being he pooled es ima o o he a iance, m he numbe o cases o he B i ish sample, n he numbe o cases o he sample om Spain, and SE he s anda d e o o he pa h p o ided by PLS-G aph in he boo s ap es . Resul s a e summa ized in Table 5 . 91 e i S a d e ec o n o m í a mu n d i a l 33, 2013, 73-103 B i iSh a n d Sp a n i S h en e p e n e u i a l in e n i o n S : a co m p a a i e S u d y a B l e 5 . mu l i G o u p a n a l y S i S βGB βSpain βGB - βSpain -s uden Signi . CV -> PA 0,155 0,205 -0,050 -0,7599 ns CV -> SkillSe 0,171 0,215 -0,044 -0,5981 ns CV -> KnowSuppo 0,225 0,154 0,071 1,0204 ns SV -> SkillSe 0,125 0,131 -0,006 -0,0894 ns SV -> PA 0,166 0,085 0,081 1,3432 ns SV -> SN 0,213 0,141 0,072 1,1444 ns SkillSe -> SN 0,275 0,205 0,070 0,9696 ns SkillSe -> PBC 0,311 0,344 -0,033 -0,5212 ns KnowSupp -> PBC 0,195 0,160 0,035 0,6193 ns SN -> PA 0,483 0,302 0,181 2,8833 ** SN -> PBC 0,321 0,219 0,102 1,5247 ns PBC -> EI 0,268 0,149 0,119 2,3646 * PA -> EI 0,521 0,682 -0,161 -3,5221 *** No e: ns = non-signi ican ; * = p < 0 .05; ** = p < 0 .01; *** = p < 0 .001 I is now possible o d aw some conclusions ega ding s uc u al ela ionships among he cons uc s . O e all, as can be seen om Figu e 2, he co e in en ion model o he join sample is ully suppo ed by he model . Thus, H1a h ough o H1d a e con i med, indica ing ha he in luence o SN on in en ions is indi ec ia PA and PBC, pe haps explaining he o en- ound weakness o SN as a p edic o The model explains 65% o he a iance in en ep eneu ial in en ion . This is e y high gi en ha mos p e ious linea models end o explain less han 40% o he a iance in EI . The model also explains abou 35% and 43% o he a iance in PA and PBC, due o he signi ican con ibu ion o SN . Looking mo e closely a he esul s o each sub-sample, hey also p o ide suppo o H1a, b, c and d . Consis en wi h he join sample, he a iance explained by he model is abou he same (a ound 65%) . Thus, adding suppo o applicabili y o he basic TPB model in e ms in en ion o ma ion being simila in he wo coun ies examined . As expec ed, wi h espec o he con ol a iables (demog aphic and human capi al), he e a e ela i ely ew signi ican e ec s ei he on en ep eneu ial in en ion, o on he an eceden s o en ep eneu ial in en ion . These e ec s we e gene ally qui e small ( anging om .04 o .10) . The highes e ec s we e o ha ing a ole model bo h on en ep eneu ial en i onmen knowledge ( .13) and on a mo e posi i e alua ion in he close en i onmen , such as amily and iends ( .15) . Hypo heses 2a and b conce ned he ela i e in luence o PA and PBC on EI o each o he wo sub-samples would be di e en . PA did indeed exe a s onge e ec on EI in he Spanish sub-sample ( .682 s . .521) . This suppo s hypo hesis H2a . In con as , PBC exe ed a s onge e ec on EI in he B i ish 92 a n c i S c o li ñ á n , Gh u l a m na B i , no i S k u e G e sub-sample ( .321 s . .219) . This suppo s hypo hesis H2b . As shown in Table 5, bo h o hese e ec s we e signi ican ly di e en , adding obus ness o hese indings . Hypo heses 3a-d ocused on he in luence o social and close alua ion on co e TPB elemen s o SN, PBC and PA . The esul s sugges a signi ican posi i e pa h be ween social alua ion and SN o bo h he join sample and wo sub-samples, p o iding suppo o H3a . Rega ding hypo hesis H3b, he e was also a signi ican posi i e pa h be ween social alua ion and PBC o he join sample and in one sub-sample (Spain) only, p o iding pa ial suppo o his hypo hesis . Hypo hesis H3c sugges ed ha close alua ion would in luence PA . This was suppo ed o bo h he join sample and each sub-sample, p o iding suppo o H3c . Hypo hesis 3d sugges ed ha close alua ion would in luence SN . Howe e , his hypo hesis is no con i med o ei he he join sample o he wo sub-samples . Hypo hesis H4a-e conce ned he pa hs om en ep eneu ial skills o he TPB elemen s, and he in luence o social and close alua ion on en ep eneu ial skills . These hypo heses we e la gely con i med as en ep eneu ial skills we e signi ican p edic o s o PA, SN and PBC espec i ely . This applied o he join sample and he wo sub-samples, wi h he excep ion o skills-PA (H4a) which was no signi ican in he B i ish sub-sample . Apa om his pa ial excep ion, i is he e o e con i med ha pe cei ed en ep eneu ial skills signi ican ly in luence he h ee mo i a ional cons uc s in bo h he Spanish and B i ish sub-samples . H4d and H4e we e ully suppo ed because social alua ion and close alua ion we e signi ican p edic o s o en ep eneu ial skills . Hypo heses H5a-c we e only pa ially suppo ed . H5c was signi ican , bu H5a and H5b we e no . En ep eneu ial en i onmen knowledge only appea ed o be a signi ican p edic o o PBC . Howe e , his inding was applicable o bo h he join sample and each sub-sample . Rega ding hypo heses, H6a and b, again hese we e only pa ially suppo ed . H6a (close alua ion would posi i ely in luence en ep eneu ial en i onmen knowledge) was suppo ed, bu H6b (social alua ion would do he same) was no . The alua ion in he close en i onmen a he han socie y in gene al he e o e appea s o in luence awa eness o he en ep eneu ial en i onmen . 5 . di S c u S S i o n The main aim o his pape was ying o d aw oge he some o he s ands in TPB-in en ion based models om a c oss-cul u al pe spec i e . Social and skills pe cep ions, combined wi h en ep eneu ial en i onmen knowledge we e in es iga ed o see how hey may a ec he mo i a ional an eceden s o en ep eneu ial in en ion . To a la ge ex en , he majo i y o he hypo heses we e suppo ed and he model explained a highly sa is ac o y pe cen age o 93 e i S a d e ec o n o m í a mu n d i a l 33, 2013, 73-103 B i iSh a n d Sp a n i S h en e p e n e u i a l in e n i o n S : a co m p a a i e S u d y he a iance in en ep eneu ial in en ion and i s mo i a ional an eceden s . Mo e speci ically, he hypo hesized model linkages ecei ed qui e s ong suppo . Mos hypo hesized ela ionships (H1a-d, H2a and H2b, H3a-c, H4a-e, H5c, and H6a) we e signi ican . Impo an ly, hese hypo heses we e con i med o bo h he join sample and o each o he sub-samples, wi h only wo excep ions (H3b, H4a) . O e all hough, he model seems ai ly obus because i was la gely applicable o bo h B i ish and Spanish sub-samples, ha is, despi e na ional di e ences be ween he wo coun ies . E en when a limi ed numbe o hypo heses we e no con i med (H3d, H5a, H5b and H6b), hey we e simila ly no con i med in he join sample no o ei he o he na ional sub-samples . This a leas indica es c oss-cul u al consis ency . In en ion was explained by an essen ially consis en pa e n o (mo i a ional, social capi al, skill se and en i onmen knowledge) an eceden s, sugges ing ha he o ma ion o in en ions and in e nal cogni i e mechanisms is b oadly compa able in he B i ish and Spanish sub-samples . This p o ides solid suppo o he c oss- cul u al applicabili y o he en ep eneu ial in en ion model . The gene al pic u e ob ained om hese esul s sugges s a ce ain pa h in he con igu a ion o en ep eneu ial in en ions . As may be expec ed, close and social alua ions exe a di ec e ec on PA and SN . A mo e a ou able en i onmen owa ds en ep eneu ship will con ibu e o people eeling mo e a ac ed and mo e suppo ed o become en ep eneu s . Bu , addi ionally, pe cei ed close alua ions o en ep eneu ship (and social alua ions wi h espec o skills) con ibu e o aising awa eness, knowledge and skills which, in u n, also con ibu e o gene a ing mo e a ou able mo i a ional an eceden s and, h ough hem, highe in en ion . This sugges s ha close en i onmen alua ions o en ep eneu ship con ibu e owa ds encou aging he acquisi ion o en ep eneu ial skills, oge he wi h knowledge and awa eness o he en ep eneu ial ca ee pa h, lending indi ec suppo o he idea ha g adua es alue in o mal mo e han o mal suppo sys ems (Shape o and Sokol, 1982; Tackey and Pe yman, 1999) . The inclusion o en ep eneu ial en i onmen knowledge in o he model (H5 and H6) ex ended om p e ious esea ch (Liñán, 2008) . This cons uc hus made an impo an con ibu ion . The e ec o knowledge (and skills) is el especially on PBC, as may be expec ed, and, o a lesse ex en , on SN . Mo eo e , i is a signi ican p edic o o PBC, ega dless o coun y, sugges ing a consis en e ec o g ea e knowledge o en ep eneu ial en i onmen and suppo sys ems con ibu ing o he sense o capaci y o i m c ea ion . In o he wo ds, hey en ep eneu ial knowledge di ec ly con ibu es o eeling able o engage in en ep eneu ial beha iou and pe cei ing con ollabili y o ha beha iou . One o he s onges indings was he hypo hesized c oss-cul u al di e ences we e con i med, sugges ing he ole o cul u e in explaining in en ion is likely o be qui e impo an . In pa icula he ela i e e ec s o PA and PBC on EI we e 94 a n c i S c o li ñ á n , Gh u l a m na B i , no i S k u e G e signi ican ly di e en . PA exe ed a s onge e ec on EI in he Spanish sub- sample . This is a no able inding and suppo s Bosma e al . (2008) and simila wo k (Uslay e al ., 2002) ha en ep eneu ship is mo e s ongly associa ed wi h pe sonal sa is ac ion and enjoymen in Spain . Con e sely, he indings o H2b also suppo he Ho s ede-based idea ha PBC is a s onge p edic o o EI in B i ain han Spain . Tha is, he highe unce ain y a oidance in Spain sugges s ha en ep eneu ship will be conside ed o be a mo e unce ain and ambiguous ca ee ou e, and hence po en ially cul u ally less accep able and endo sed . We a gued his could explain why eelings o en ep eneu ial sel -e icacy do no lead o highe s a - up in en ions . This, combined wi h a highe ea o ailu e in Spain (Bosma e al ., 2008), could explain why PBC was a weake p edic o o EI in Spain . Only wo hypo heses a e held in one sub-sample alone . Social alua ion did signi ican ly con ibu e o pe cei ed beha iou al con ol, albei no in he B i ish sub-sample (H3b) . In a less en ep eneu ial cul u e ( ha o Spain), pe cei ing low social alua ions would end o dec ease PBC, possibly due o sha ed ea o ailu e and unce ain y a oidance belie s . Addi ionally, possessing en ep eneu ial skills does no inc ease PA in B i ain (H4a) . Gi en in his la e coun y he e is a highe social alua ion which di ec ly inc eases PA, he ole o skills may no be so ele an in his espec . Meanwhile in Spain (lowe social alua ion and weake e ec on PA) skills become impo an in inc easing a ac ion owa ds en ep eneu ship . Simila ly, when Spain was used as a dummy con ol a iable in he join - sample analysis, i u ned ou o be e y in luen ial, wi h signi ican pa hs o almos all cons uc s (excep EI) . The Spanish sub-sample had lowe le els on mos cons uc s, namely, PA, PBC, skills se s, en ep eneu ial knowledge and close /social alua ions . This suppo s he idea ha B i ain, a he han Spain, was he mo e en ep eneu ial coun y . Taken oge he wi h he ela i e e ec s o PA and PBC on EI in he wo sub-samples, his sugges s possible cul u al and social nuances in he o ma ion o in en ion om i s an eceden s . O e all, howe e , he indings sugges a heo y-d i en explana ion o McG a h & MacMillan’s (1992) indings: G ea e simila i ies han di e ences in he con igu a ion o mo i a ional an eceden s o en ep eneu ial in en ion ha sugges pe haps a easonable deg ee o con e gence on he p e- ounding cogni i e p ocesses o en ep eneu s in B i ain and Spain . 5 .1 . limi a ionS a n d implica ionS Despi e he s ong suppo o he en ep eneu ial in en ion model in his esea ch, he indings should be ea ed en a i ely because o he lack o p e ious c oss-cul u al wo k using B i ish and Spanish samples . Tha said, he model and indings seem qui e obus because hey: (a) explain a e y high pe cen age o he a iance in en ep eneu ial in en ion and also in PA and PBC; (b) we e based on wo coun ies which allowed o c oss-cul u al analysis; and 95 e i S a d e ec o n o m í a mu n d i a l 33, 2013, 73-103 B i iSh a n d Sp a n i S h en e p e n e u i a l in e n i o n S : a co m p a a i e S u d y (c) s a is ical analysis ( ac o and s uc u al analysis) was b oadly equi alen in B i ish and Spanish samples . Taking he abo e in o accoun , wo main implica ions may be sugges ed . The i s ela es o he indings ega ding he ela i e e ec s o PA and PBC on en ep eneu ial in en ion in he B i ish and Spanish sub-samples (H2) . This has impo an implica ions o policy make s . The e is an a gumen o en ep eneu ship o ecei e g ea e posi i e social legi ima ion . Ou esul s sugges s ha Spain is he less en ep eneu ial coun y, wi h a ela i ely weake e ec be ween he c i ical mo i a ional cons uc , pe cei ed beha iou al con ol, and en ep eneu ial in en ion, despi e highe le els o pe sonal a ac ion o i m c ea ion . This p esen s Spanish policy make s wi h an oppo uni y o ake s ock and add ess he issue o how en ep eneu ship is cul u ally endo sed and app o ed in Spain, pe haps obse ing cul u al p ac ices in B i ain . Fo example, how public policy in Spain, as well as Spanish uni e si ies, encou age he image o en ep eneu ship and impo an ly acili a e posi i e en ep eneu ial cogni ions . Tha is, a highe social p o ile, posi i e image in he media and socie y, and mo e legal/ ax e o ms ha acili a e en ep eneu ship, o ein o ce he social message ha i m c ea ion is an a ac i e, alued and easible ca ee op ion . This should help o educe unce ain y a oidance abou en ep eneu ship (pe cep ions ha his is an ambiguous and unce ain ca ee pa h) and bols e PBC (g ea e sense o capaci y), and hence aise le els o en ep eneu ial in en ion and beha iou . All hese a e es able p oposi ions ha a e wo h explo ing . The second main implica ion conce ns he e ec o en ep eneu ial en i onmen knowledge o e PBC . Gi en ha his held, ega dless o coun y, he e a e g ounds o sugges ha decision make s and en ep eneu ship educa o s conside wha hey can do o enhance he le el o knowledge and awa eness s uden s ha e abou he en ep eneu ial en i onmen . This could ake he o m, o example, o aising awa eness o s uden s (pe haps a a uni e si y-wide le el) o suppo bodies (e .g ., Business Link and simila o ganiza ions, egional/local de elopmen agencies), aining and suppo measu es (e .g ., echnical aid, incuba ion cen es) o de elop quali y businesses and access o p e e en ial loans . Again his should help in sending ou he igh signals ha becoming an en ep eneu is socially alued and enhance pe cei ed capaci y abou i m c ea ion . 6 . co n c l u S i o n S This esea ch has con ibu ed owa ds he li e a u e on en ep eneu ial in en ion by de eloping and es ing an en ep eneu ial in en ion model in wo coun ies, conside ing he ole o cul u e, along wi h mo i a ional and o he cons uc s, especially, knowledge o he en ep eneu ial en i onmen . This mo e in eg a ed and c oss-cul u al app oach has ecei ed li le a en ion in p e ious esea ch . 96 a n c i S c o li ñ á n , Gh u l a m na B i , no i S k u e G e The esul s suppo he majo i y o he hypo heses and he model b oadly holds o he wo di e en coun ies . The ole o cul u al and social nuances, do eme ge, especially in he ela i e s eng h o e ec s be ween PA and PBC on en ep eneu ial in en ion . Salien cul u al belie s associa ed wi h an en ep eneu ial ca ee appea o be a play, explaining why PA is mo e s ongly linked o in en ion amongs ou Spanish sample, and PBC is mo e s ongly linked o in en ion amongs ou B i ish sample . None heless, ou esul s gene ally appea o sugges he c oss-cul u al applicabili y o he in en ion model, a leas in e ms o he h ee mo i a ional an eceden s o en ep eneu ial in en ion . The ela i e impo ance o he an eceden s o hese mo i a ional ac o s (social and skills pe cep ions, en ep eneu ial en i onmen knowledge) also appea s o be b oadly simila . Fu he esea ch should be ca ied ou o co obo a e ou indings in he same coun ies unde in es iga ion in his esea ch and o he cul u ally di e en coun ies o es he in e na ional applicabili y o he model . We now ha e new, obus insigh s in o wha “lies benea h” he o ma ion o en ep eneu ial in en ions ha sugges use ul esea ch di ec ions and also sugges p ac ical implica ions o policy make s . As such, ano he impo an a enue o esea ch will be o examine he link be ween en ep eneu ial in en ion and beha iou in di e en coun ies, wi h a iew o unde s anding he ex en o which cul u al di e ences in luence his p ocess . BiBlioG aphic e e e n c e S Ajzen, I . (1991): “The Theo y o Planned Beha io ”, O ganiza ional Beha io and Human Decision P ocesses, 50(2), 179-211 . Ajzen, I . (2001): “Na u e and Ope a ion o A i udes”, Annual Re iew o Psychology, 52, 27-58 . Au io, E ., Keeley, R . H ., Klo s en, M ., Pa ke , G . G . C ., and Hay, M . (2001): “En ep eneu ial In en Among S uden s in Scandina ia and in he USA”, En e p ise and Inno a ion Managemen S udies, 2(2), 145-160 . Bi d, B . (1988): “Implemen ing En ep eneu ial Ideas: he Case o In en ion”, Academy o Managemen Re iew, 13(3), 442-453 . Bosma, N ., Acs, Z . J ., Au io, E ., Codu as, A ., and Le ie, J . (2008): Global En ep eneu ship Moni o . 2008 execu i e epo , Global En ep eneu ship Resea ch Conso ium (GERA), Wellesley, Massachuse s and London . Boyd, N . G ., and Vozikis, G . S . (1994): “The In luence o Sel -e icacy on he De elopmen o En ep eneu ial In en ions and Ac ions”, En ep eneu ship Theo y and P ac ice, 18, 63-77 . Buseni z, L . W ., and Lau, C . M . (1996): “A C oss-cul u al Cogni i e Model o New Ven u e C ea ion”, En ep eneu ship Theo y and P ac ice, 20(4), 25-39 . Chen, C . C ., G eene, P . G ., and C ick, A . (1998): “Does En ep eneu ial Sel - e icacy Dis inguish En ep eneu s F om Manage s?”, Jou nal o Business Ven u ing, 13(4), 295-316 . 103 e i S a d e ec o n o m í a mu n d i a l 33, 2013, 73-103 B i iSh a n d Sp a n i S h en e p e n e u i a l in e n i o n S : a co m p a a i e S u d y ap p e n d i x 2 . mu l i G o u p a n a l y S i S Equa ion 1 . T-s a is ic wi h m + n - 2 deg ees o eedom Equa ion 2 . Pooled es ima o o he a iance