The cooperation between enterprises: significant part of the innovation process – a case study of the Czech machinery industry

110 2016, XIX, 3 Ekonomika a managemen 110 DOI: 10.15240/ ul/001/2016-3-008 In oduc ion In ecen yea s we ha e been able o see he g owing impo ance o knowledge in he con ex o he eme ging knowledge-based global economy (Mio i & Sachwald, 2003; Ca meli, Gelba d, & Rei e -Palmon, 2013; P okop & S ejskal, 2015). Töd ling and T ippl (2005) sugges ha he e is widesp ead ag eemen in academic li e a u e ha knowledge, lea ning, and inno a ion a e key o economic de elopmen and he compe i i eness o en e p ises o egions (and na ions). Gene a ing inno a ion has come o be seen as a p oduc i e opic in economics and has also has been becoming inc easingly ela ed o en e p ise’s abili y o abso b ex e nal in o ma ion, knowledge, and echnologies (Negassi, 2004; Lich en hale , 2011). I is commonly accep ed ha (i) inno a ions a e b ough o wa d h ough an in e ac i e p ocess o knowledge gene a ion, di usion, and applica ion; (ii) inno a ions a e inc easingly seen as undamen al o he compe i i eness o en e p ises and economies; and (iii) knowledge is c i ical o he p ocess o inno a ion (Te he & Taja , 2008; Töd ling, Lehne , & Kau mann, 2009). Tomlinson (2010) s a es ha knowledge ans e de i ni i ely leads o g ea e le els o bo h p oduc and p ocess inno a ion. Knowledge can be gene a ed in e nally o acqui ed ex e nally (wha an o ganiza ion knows de e mines wha i can do) and R&D coope a ion is a undamen al ing edien o he di ision o inno a i e labo (F i sch & Lukas, 2001; S ejskal & Hájek, 2015). We can s a e ha en ep eneu ship and inno a ion a e undamen al d i e s o economic g ow h, he c ea ion o weal h in an economy, and en e p ise’s long- e m su i al, p o i abili y, and sus ainable g ow h (Myško á, 2010; Classen e al., 2012; Acemoglu, Gancia, & Zilibo i, 2012; Kim e al., 2012). The e o e, en e p ises ha wan o succeed in he ma ke a e inc easingly looking o pa ne s wi h whom hey can collabo a e e ec i ely and gain a compe i i e ad an age (S ejskal & Hájek, 2012). I is clea ha he ole o R&D coope a ion wi hin he inno a ion p ocess has ecen ly inc eased (Xia, Danping, & Yue, 2012). This ac is suppo ed by se e al s udies ha con i m he impo ance o coope a ion – e.g. De Fa ia, Lima and San os (2010); Tomlinson (2010) and Bee s and Zand (2014). Collabo a ion wi h o he en e p ises and ins i u ions in R&D is a decisi e way o make ex e nal esou ces usable (Becke & Die z, 2004). En e p ises and coun ies use he expe ience acqui ed (knowledge de eloped) by o he en e p ises (o coun ies) o build hei own knowledge capi al (Negassi, 2004). Becke and Die z (2004) also sugges ha he impo ance o R&D coope a ion has isen s eadily as a consequence o he g owing complexi y, isks, and cos s o inno a ion and ha en e p ises engaged in he inno a ion p ocess a e awa e o he necessi y o es ablishing R&D coope a ion o ob ain expe ise ha canno be gene a ed in-house. Bu he ques ion is which pa ne is he mos sui able o inno a i e indus ial en e p ises. En e p ise coope a es wi h a ious pa ne s o a ious easons. The e o e, i is no possible o speci y a uni e sal pa ne ha would be sui able o all en e p ises in all sec o s. De Fa ia, Lima, and San os (2010) s a e ha he ac o s in l uencing he impo ance THE COOPERATION BETWEEN ENTERPRISES: SIGNIFICANT PART OF THE INNOVATION PROCESS – A CASE STUDY OF THE CZECH MACHINERY INDUSTRY Jan S ejskal, Beá a Mikušo á Me ičko á, Vik o P okop EM_3_2016.indd 110EM_3_2016.indd 110 8.9.2016 14:11:218.9.2016 14:11:21 111 3, XIX, 2016 Business Adminis a ion and Managemen 111 a ibu ed o coope a ion ac i i ies a e di e en om he ones behind he decision o coope a e. En e p ises a e able o coope a e wi h ou di e en ypes o R&D pa ne s: (i) compe i o s, (ii) supplie s, (iii) cus ome s, and (i ) uni e si ies and esea ch ins i u es (Belde bos, Ca ee, & Lokshin, 2004). Finding a sui able pa ne o coope a ion is a complica ed p ocess, because hese pa ne s a e di e en o each sec o . A numbe o s udies ha e add essed he issue o collabo a ion in he manu ac u ing indus y – e.g. F i sch and Lukas (2001); Becke and Die z (2004); Veugele s and Cassiman (2005); Tomlinson (2010). Howe e , none o hese s udies examine he e ec i eness o coope a i e chains o compa e he esul s o he indus y. Such in o ma ion could help manage s app op ia ely a ge e o s o i nd sui able pa ne s o coope a ion. The esea ch p esen ed in his a icle examines he impac o impo an pa ne s in one o he sec o s o manu ac u ing indus y – he machine y indus y (enginee ing). The pape in ends o analyze he issues o (i) he impo ance o collabo a ion be ween en e p ises and o he pa ne s h ough en e p ises’ inno a i e beha io and he o e all g ow h o hei pe o mance and (ii) he e i ciency o public subsidies om na ional and Eu opean unds. The pape will analyze bo h o hese o he Czech Republic’s machine y indus y by using o iginal mul iple linea eg ession models based on he da a om he Communi y Inno a ion Su ey ca ied ou in he Czech Republic o he pe iod o 2010-2012. The emainde o he pape is s uc u ed as ollows. The second sec ion discusses he heo e ical backg ound and hypo heses. The da a me hodology, esul s, and hei analysis a e p esen ed in he hi d sec ion. The i nal sec ion consis s o conclusions and ecommenda ions. 1. Theo y and Hypo heses Since he la e 1980s, i has become inc easingly acknowledged ha spillo e s o knowledge om ex e nal sou ces ha e an impo an impac on inno a ion p ocesses and economic de elopmen and ha he ole o inno a ion as a ac o o compe i i eness and echnological p og ess ha e combined o make en e p ises in ensi y and expand hei inno a i e capabili ies (Mio i & Sachwald, 2003; F i sch & F anke, 2004; Acs, Aud e sch, & Lehmann, 2013; Al a o & Chen, 2013). Negassi (2004) sugges s ha he ise o in e -en e p ise coope a ion did no begin un il he beginning o he 1980s. A numbe o s udies and esea che s sugges ha human capi al ac o s play an impo an ole in spu ing egional g ow h and ha egions (o en e p ises) unc ion as incuba o s o c ea i i y and inno a ion (Lee, Flo ida, & Acs, 2004; Ba elsman, Dobbelae e, & Pe e s, 2014, U amo á & Kožiak, 2008). The g owing impo ance o knowledge p ocesses is suppo ed by Flo ida, who de i ned he new concep o lea ning egions, which a e becoming ocal poin s o knowledge c ea ion and lea ning in he new age o global, knowledge-in ensi e capi alism (Flo ida, 1995). Ins i u ional economics is ano he heo e ical concep ha suppo ed he ole o knowledge and inno a ion. Richa d Nelson – one o he main p oponen s o his heo y – desc ibed he h ee main issues o ins i u ional economics: (i) echnology and echnological inno a ion; (ii) concep s o en e p ises, hei p inciples o ope a ion, and hei ela ion o compe i o s o subcon ac o s; and (iii) ins i u ions (Blažek & Uhlíř, 2011). The endogenous g ow h heo y, in which knowledge ep esen s an inpu in o he p ocess o gene a ing inno a i e ac i i y in he economy, gi es new insigh s in o he ole o knowledge in economic g ow h (Acs, de G oo , & Nijkamp, 2002). As we can see, knowledge has been playing an impo an ole in heo e ical concep s in ecen yea s. Howe e , in-house p oduc ion o inno a ion is no longe su i cien . The e o e, en e p ises use R&D pa ne ships o access knowledge and build global R&D ne wo ks (Mio i & Sachwald, 2003; Bee s & Zand, 2014; Hájko a & Hájek, 2014), and adop open inno a ion app oach o use o pu posi e in l ows and ou l ows o knowledge o accele a e in e nal inno a ion, and expand he ma ke s o ex e nal use o inno a ion (Chesb ough, 2006; Chesb ough & C ow he , 2006). Te he (2002) s a es ha he ela ionship be ween inno a ion and coope a ion is no s aigh o wa d and ha mos en e p ises s ill de elop hei new p oduc s, p ocesses, and se ices wi hou o ming coope a i e a angemen s o inno a ion wi h o he o ganiza ions. Howe e , en e p ises ha engage in R&D and a e a emp ing o in oduce inno a ions a a highe le el (i.e. inno a ions ha a e new o he ma ke a he han new o he en e p ise) a e much mo e likely o engage in coope a i e a angemen s o inno a ion and EM_3_2016.indd 111EM_3_2016.indd 111 8.9.2016 14:11:218.9.2016 14:11:21 112 2016, XIX, 3 Ekonomika a managemen 112 a e able o access and le e age s a egically c i ical esou ces o suppo he en e p ise´s inno a ion p ocess (Te he , 2002; Classen e al., 2012). In ac , en e p ises looking o imp o e hei compe i i eness need o de elop wo kinds o compe ences: (i) echnological, which allows en e p ises o add alue o p oduc s and p ocesses, and (ii) ne wo k, by which en e p ises a e able o link hei o ganiza ion o o he playe s in he ma ke o allow in e ac ions beyond o ganiza ional bounda ies (Ri e & Gemünden, 2004). The ela i e impo ance o coope a ion on inno a i e ac i i y a ies and depends upon a numbe o ac o s, i.e. he numbe o coope a i e ies and he con ex in which ne wo k ela ions exis (Tomlinson, 2010). Belde bos e al. (2004) explo ed he e ogenei ies in he de e minan s o inno a ing en e p ises’ decisions o engage in R&D coope a ion, di e en ia ing be ween ou ypes o coope a ion pa ne s (compe i o s, supplie s, cus ome s, and uni e si ies and esea ch ins i u es – ep esen ing ins i u ional coope a ion). The de e minan s o R&D coope a ion di e depending on he ypes o coope a ion: (i) he posi i e impac o en e p ise size, R&D in ensi y, and incoming sou ce- speci i c spillo e s is weake o compe i o coope a ion, e l ec ing g ea e conce ns abou app op iabili y and (ii) ins i u ional spillo e s a e mo e gene ic in na u e and posi i ely impac all ypes o coope a ion (Belde bos e al., 2004). Negassi (2004) analyzed he si ua ion among F ench en e p ises and sugges ed ha he c ucial de e minan s o he comme cial success o inno a ions a e en e p ise size, ma ke sha e, R&D in ensi y, and human capi al. This s udy also pinpoin ed ha spillo e measu emen s ( he acquisi ion o machine ools, o eign pa en s, licenses) and echnological oppo uni ies ha e a posi i e impac on inno a ion. Sega a-Blasco & A auzo-Ca od (2008) di ided a p opensi y o R&D coope a ion wi h o he en e p ises and ins i u ions in o h ee app oaches: (i) he ansac ion cos app oach (coope a i e R&D p ojec s enable he cos s and isks o R&D ac i i ies o be sha ed and he dissemina ion o he esul s o be p o ec ed); (ii) he s a egic managemen app oach (coope a i e beha io is a way o accessing addi ional esou ces and his leads o compe i i e ad an ages, see mo e Ga u o a (2012); and (iii) he indus ial o ganiza ion app oach (which ocuses on knowledge spillo e s be ween pa ne s). Uni e si ies cons i u e one o he coope a ion pa ne s; howe e , en e p ises do no coope a e wi h hem e y o en. Veugele s and Cassiman (2005) used Communi y Inno a ion Su ey da a o analyze Belgian manu ac u ing en e p ises and showed ha (i) only a small ac ion o inno a i e en e p ises use science (i.e. uni e si ies and public esea ch labo a o ies) as an impo an in o ma ion sou ce in hei inno a ion p ocess and (ii) coope a ing wi h uni e si ies canno be analyzed in isola ion om he o e all inno a ion s a egy o he en e p ise, because coope a ing wi h uni e si ies is complemen a y o o he inno a ion ac i i ies such as p o iding hei own R&D, sou cing public in o ma ion, and coope a ing wi h o he pa ne s. Decision-making conce ning sui able coope a i e pa ne s ha will p oduce comme cial esul s is missing om his s udy. Fu he s udies (Belde bos, Ca ee, & Lokshin, 2004) a gued ha coope a ion wi h compe i o s and uni e si ies impac ed inno a ion ou pu le els posi i ely bu coope a ion wi h cus ome s nega i ely – hese i ndings a e in con as wi h lead-use s concep (Von Hippel, 1986; 2005). This concep de i nes lead use s as use s whose p esen s ong needs will become gene al in a ma ke place mon hs o yea s in he u u e and s a es ha use -cen e ed inno a ion is a e y powe ul and gene al phenomenon suppo ing inno a i e ac i i ies, he e o e he e is a need o u he in es iga ion o he ela ionship be ween en e p ises and cus ome s. Coope a ion on inno a ion wi h domes ic uni e si ies causes he g ow h o knowledge capi al. In ac , uni e si ies a e able o signi i can ly in l uence he egional economy, play an impo an ole in inno a ion pe o mance, and a e also linked o social e ec s (Mohammadi & Ka ami, 2014). The ob ious implica ions o applying an inc ease in social capi al ( ypically on he ma ke ) is also missing om his s udy. Uni e si ies a e ega ded as a key ac o in na ional inno a ion plans and s a egies by many au ho s (e.g. Godda d, Robe son, & Vallance, 2012; Wa kins e al., 2015). Diez-Vial & Mon o o- Sanchez (2014) show a posi i e ela ionship be ween he echnological knowledge ob ained om uni e si ies and he inno a ion ca ied ou by en e p ises. We can conclude ha uni e si ies play an impo an ole in he global knowledge economy and ep esen key poin s EM_3_2016.indd 112EM_3_2016.indd 112 8.9.2016 14:11:218.9.2016 14:11:21 113 3, XIX, 2016 Business Adminis a ion and Managemen 113 o in e na ional con ac , which is complica ed o many easons – such as mobili y issues, he use o echnology, o collabo a ion (Deiaco, Hughes, & McKel ey, 2012; Al bach, 2013). The e o e, we p oceed om he a gumen s men ioned abo e and hypo hesize ha : H1: En e p ises ha coope a e on inno a i e ac i i ies in he machine y indus y in he Czech Republic demons a e a g ea e g ow h o hei pe o mance han he ones ha do no coope a e. In mos cases, hese en e p ises pa icipa e in g oups o en e p ises ha posi i ely in l uence inno a i e c ea ion wi hin he g oup. This assump ion is based on a se ies o analyses. De Fa ia, Lima and San os (2010) analyzed he e ec s o coope a ion on he o e all pe o mance o en e p ises and on inno a ion and R&D pe o mance. They s a e ha (i) en e p ise’s decision o coope a e on inno a ion is d i en by he ac ha coope a ion is an e i cien way o imp o e he p obabili y o he success o inno a ion p ojec s and (ii) coope a i e en e p ises ha e, on a e age, highe o e all pe o mance le els and highe R&D in ensi y han non-coope a i e en e p ises. Belde bos, Ca ee and Lokshin (2004) also sugges ed ha coope a ing en e p ises a e gene ally engaged in inno a i e ac i i ies o a g ea e deg ee and demons a ed he impac o coope a ion on p oduc i i y g ow h. Coope a ion wi h di e en pa ne s leads o lea ning oppo uni ies wi h ega d o bo h coope a ion and inno a ion skills and hence is expec ed o enhance en e p ise’s inno a ion pe o mance (Bee s & Zand, 2014). Chesb ough & Appleya d (2007) also sugges ed ha o make s a egic sense o inno a ion ne wo ks, collabo a ions, ecosys ems, communi ies, and hei implica ions o compe i i e ad an age, open inno a ion concep and s a egy a e needed. Pa icipa ion in a g oup o en e p ises p esen s ano he oppo uni y o coope a ion, and i also suppo s he con en ion ha inno a ions do no a ise independen ly. This issue was also analyzed by a numbe o au ho s (e.g. Lode alk, 2010; Hashi & S ojčić, 2013; Dachs & Pe e s, 2014). As we demons a ed p e iously, inno a ion c ea ion can be suppo ed by coope a ion wi h uni e si ies, because hey posi i ely impac inno a ion ou pu le els. Uni e si y-indus y collabo a ion can be ega ded as a de e minan o inno a ion, because mo e inno a i e en e p ises end o be mo e in e es ed in collabo a ing wi h uni e si ies (Oyela an- Oyeyinka & Abiola Adebowale, 2012; López e al., 2014). This assump ion leads us o he nex hypo hesis: H2: En e p ises in he machine y indus y in he Czech Republic ha coope a e wi h uni e si ies and public esea ch cen e s demons a e a g ea e g ow h o hei o e all pe o mance. The p e ious hypo heses ocus only on he impo ance o coope a ion. Howe e , many collabo a ions a e o en suppo ed by public unds (Ma , Robin, & Wol , 2012; Ho en o & Lopes-Ben o, 2014). In ecen yea s, inc easingly ine i cien spending o public unds and subsidies has been esul ing in addi ional spending on inno a ion by en e p ises bu has no lead o addi ional inno a ion ou pu (Hashi & S ojčić, 2013). The e o e, le us de i ne u he hypo hesis: H3: In he Czech Republic, public unding in he machine y indus y is being p o ided in an ine i cien way. 2. Da a Me hodology and Analysis o Resea ch Resul s A ha monized ques ionnai e o EU Membe S a es om he Communi y Inno a ion Su ey (CIS) ca ied ou in he Czech Republic o he pe iod o 2010-2012 by combining sample (s a i i ed andom sampling) and exhaus i e su eys was used o he da a collec ion. The CIS is a su ey o inno a ion ac i i y in en e p ises ha is designed o p o ide in o ma ion on he inno a i eness o sec o s by en e p ise ype, he di e en ypes o inno a ion, and a ious aspec s o inno a ion de elopmen , e.g. objec i es, sou ces o in o ma ion, public unding, and inno a ion expendi u es (Eu os a , 2015). The CIS ques ionnai e was used o examine he impac s o coope a ion by many o he au ho s (e.g. Mio i and Sachwald (2003); Belde os, Ca ee and Lokshin (2004); Veugele s and Cassiman (2005); Te he and Taja (2008); De Fa ia, Lima and San os (2010). In o al, da a on 5,151 Czech en e p ises wi h a leas 10 employees we e ob ained (wi h a esponse a e g ea e han 60%). Fo he pu pose o his s udy, we i l e ed 284 en e p ises, i.e. only en e p ises om he machine y indus y, in o ou da a g oup – speci i cally, en e p ises co e ing NACE ca ego ies 29-30. EM_3_2016.indd 113EM_3_2016.indd 113 8.9.2016 14:11:218.9.2016 14:11:21 114 2016, XIX, 3 Ekonomika a managemen 114 Fo analyzing he ela ionship be ween a iables, a mul iple linea eg ession model was used ( eg ession models a e commonly used o his kind o esea ch, e.g. Schneide and Spie h (2013); Lae en, Le ine and Michalopoulos (2015)). This model was adap ed o in es iga e he ela ionship be ween one dependen a iable, ep esen ed by he g ow h o o al u no e be ween he yea s 2010-2012, and independen a iables (see Tab. 1). Eu os a does no p o ide numbe o a iables sui able o analyzing, inpu da a om o he sou ce will no ensu e he causali y, he e o e, o u he esea ch we in es iga e o analyze he g ow h o o al u no e . Mul iple linea eg ession models ake he gene al o m shown below (Cha e jee & Hadi, 2013; Wu e al., 2013): y = β0 + β1x1 + β2x2 + … + βnxn + ε (1) whe e y is he dependen a iable; x1, x2 … xn a e independen a iables; ε is an e o e m accoun ing o he a iabili y in y ha canno be explained by he linea e ec o he n independen a iables; β1, β2 … βn, called he eg ession pa ame e s o coe i cien s, a e unknown cons an s o be de e mined (es ima ed) om he da a. Fi s ly, be o e we composed he models, e i i ca ion whe he he da a om CIS we e co ela ed was conduc ed by using Spea man’s es . Spea man’s coe i cien ( s) measu es he s eng h o he linea ela ionship be ween each wo a iables when he alues o each a iable a e ank-o de ed om 1 o N, whe e N ep esen s he numbe o pai s o alues ( he N cases o each a iable a e assigned he in ege alues om 1 o N inclusi e and no wo cases sha e he same alue). The di e ence be ween anks o each case is ep esen ed by di. The gene al o mula o Spea man’s ank co ela ion coe i cien akes he gene al o m as ollows (Weinbe g & Ab amowi z, 2002; Bo adaile, 2013): s = 1 – 6 Σd2 i N3 – N (2) All calcula ions we e made in s a is ical so wa e STATISTICA (S a So Inc., 2011). Values o Spea man´s es ejec ed he hypo hesis ha he da a a e co ela ed a he le el o signi i cance p < 0.05. A e ul i lling he i s p e equisi e (unco ela ed da a) and e usal o mul icollinea i y in he model, he analysis i sel was conduc ed. Subsequen ly, se e al mul iple linea eg ession models we e c ea ed o analyze he hypo heses. In o al, he i nal model (Model 1) desc ibing he si ua ion in he Czech Republic’s machine y indus y used 17 a iables (1 dependen , 10 independen ca ego ical, and 6 independen con inuous). All a iables a e lis ed in Tab. 1. The co ela ion coe i cien R o Model 1 had a alue o 0.979 and he coe i cien o de e mina ion R2 was 0.958. I means ha eg ession line nea ly pe ec ly i s he da a. The P- alue o his model was calcula ed a 0.000 ( he alue is ounded o h ee decimal places). The P- alue showed ha his model was signi i can a p < 0.01; he e o e, he null hypo hesis, ha he model is no signi i can , was ejec ed. The e o e, Model 1 can be ega ded as signi i can . Be o e we es he hypo hesis, i is necessa y o desc ibe he esul s p o ided by Model 1 (see Tab. 2). The esul s in Tab. 2 show ha se en a iables posi i ely in l uence he dependen a iable. We can see ha he in oduc ion o p oduc inno a ion a ec s he g ow h o pe o mance in all cases (independen ly, wi h o he en e p ises, and wi h uni e si ies and esea ch ins i u es). When he en e p ise in oduced p oduc inno a ions sepa a ely wi hou coope a ion, his a iable is signi i can a p < 0.05 (0.047). Con e sely, i en e p ise collabo a es in he de elopmen o p oduc inno a ions, i is possible o obse e ha his coope a ion has a g ea e impac on he c ea ion o inno a ions. When en e p ises use o he en e p ises o coope a e, he alue was measu ed a 0.002 (0.004 o coope a ion wi h uni e si ies). Bo h alues a e signi i can a he highe le el o signi i cance (p < 0.01). When p ocess inno a ion was implemen ed, we can see ha he en e p ise’s pe o mance is mos a ec ed when en e p ises in oduce inno a ions alone o wi h uni e si ies and esea ch ins i u ions. In bo h cases, he p- alue is measu ed a 0.000, which is signi i can a a signi i cance le el o p < 0.01. On he o he hand, he in oduc ion o p ocess inno a ions in collabo a ion wi h o he en e p ises p o ed o be insigni i can . Pa icipa ion in g oup o en e p ises (GP) and coope a ion on echnical inno a ion ac i i ies (CO) a e among he o he signi i can a iables ha a ec he g ow h o o al u no e . EM_3_2016.indd 114EM_3_2016.indd 114 8.9.2016 14:11:218.9.2016 14:11:21 115 3, XIX, 2016 Business Adminis a ion and Managemen 115 Dependen a iable Independen a iables (ca ego ical) Independen a iables (con inuous) G ow h o o al u no e Pa o a g oup o en e p ises (GP) To al inno a ion expendi u e/ o al u no e (RTOT/TOT) In oduc ion o p oduc inno a ion: a) inno a ion was de eloped independen ly by he en e p ise (INPRI); b) inno a ion was de eloped in collabo a ion wi h o he en e p ises (INPRE); c) inno a ion was de eloped in collabo a ion wi h uni e si ies and esea ch ins i u es (INPRU) In-house R&D expendi u e/ o al u no e (RRDIN/RTOT) In oduc ion o p ocess inno a ions: a) inno a ion was de eloped independen ly by he en e p ise (INPSI); b) inno a ion was de eloped in collabo a ion wi h o he en e p ises (INPSE); c) inno a ion was de eloped in collabo a ion wi h uni e si ies and esea ch ins i u es (INPSU) Ex e nal R&D expendi u e/ o al u no e (RRDEX/RTOT) Public i nancial suppo om he EU (FUNEU) Acquisi ion o equipmen / o al u no e (RMAC/RTOT) Regional and/o na ional public i nancial suppo (FUNGMT) Acquisi ion o ex e nal knowledge/ o al u no e (ROEK/RTOT) Coope a ion on echnical inno a ion ac i i ies (CO) Acquisi ion o o he inno a ion ac i i ies/ o al u no e (ROTH/ RTOT) Sou ce: own Legend: in oduc ion o p oduc inno a ion = in oduc ion o a new o imp o ed p oduc /se ice; in oduc ion o p ocess inno a ions = new o imp o ed me hod o p oduc ion o p o ision o se ices, also including dis ibu ion and s o age, p o iding business suppo ac i i ies, o signi i can changes in echnology, equipmen , o so wa e) Va iables p sd RTOT/TOT 0.689 1.739 IN/RTOT 0.694 0.011 RMAC/RTOT 0.610 0.009 GP 0.037** 1.342 INPRI 0.047** 1.479 INPRE 0.002*** 1.511 INPRU 0.004*** 1.312 INPSI 0.000*** 1.082 INPSE 0.359 0.684 INPSU 0.000*** 1.007 FUNEU 0.755 0.768 CO 0.000*** 0.706 Sou ce: own Legend: p = p- alue; sd = s anda d de ia ion; *** signi i can a p < 0.01; **signi i can a p < 0.05 Tab. 1: Inpu a iables Tab. 2: Va iables used in Model 1 EM_3_2016.indd 115EM_3_2016.indd 115 8.9.2016 14:11:228.9.2016 14:11:22 116 2016, XIX, 3 Ekonomika a managemen 116 The esul s in Tab. 2 allow us o con i m hypo hesis H1 and show ha inno a ion does no a ise in isola ion, espec i ely o cou se inno a ions can occu in isola ion, howe e coope a ion could be posi i ely a ec ed by inno a ion c ea ion. The a iable CO was analyzed as one o he mos signi i can a iables wi h g ea e in l uence on he en e p ises´ pe o mance (0.000***). Mo eo e , pa icipa ion in a g oup o en e p ises ha showed s ong chains was measu ed as signi i can . We can say ha ex e nal sou ces o knowledge gained om inno a i e collabo a ion cu en ly ep esen an impo an compe i i e elemen ; he e o e, en e p ises o en op o ex e nal coope a i e pa ne s ha ha e a signi i can impac on hei o e all pe o mance in c ea ing inno a ion (e.g. INPRE, INPRU, INPSU – see Tab. 2). A numbe o o eign s udies con i m his ac : Clausen (2013); Gallego, Rubalcaba, and Suá ez (2013); and S ock, To zaue , and Zacha ias (2014). The esul s in Tab. 2 also allow us o es hypo hesis H2. Collabo a ion wi h uni e si ies and esea ch ins i u es plays an impo an ole in he machine y indus y in he Czech Republic. In bo h cases (p oduc and p ocess inno a ion), he e a e signi i can in e ac ions be ween a iables. Fo p oduc inno a ion, he impo ance o coope a ion wi h hese ins i u ions and wi h o he en e p ises ou weighed he c ea ion o inno a ion independen ly. Fo p ocess inno a ion, c ea ing inno a ion wi h uni e si ies and esea ch ins i u ions and c ea ing hese inno a ions independen ly is gaining in impo ance (c ea ing inno a ions in collabo a ion wi h o he en e p ises is insigni i can in his case). Robin and Schube (2013) also analyzed he impac o coope a ion wi h public esea ch on en e p ises’ p oduc and p ocess inno a ions by using he CIS. They examined ha coope a ing wi h public esea ch inc eases p oduc inno a ion bu has no e ec on p ocess inno a ion, which depends mo e on he en e p ises’ openness. In ou analysis, coope a ion wi h public esea ch inc eases en e p ises’ pe o mance o bo h p oduc and p ocess inno a ion; he e o e, we can con i m hypo hesis H2. Model 1 also p o ided a numbe o addi ional highe o de in e ac ions ha suppo ou hypo hesis H2. Tab. 3 shows ha coope a ion wi h he public sec o can con ibu e o signi i can ies in some cases (mo e han when his ype o coope a ion is no used). The esul s in Tab. 2 show ha i he en e p ise pe o ms p oduc inno a ion independen ly (INPRI), he in l uence on he measu ed alue o he en e p ises’ pe o mance was 0.047 (signi i can a p < 0.05). I he e is a subsequen in e ac ion wi h INPRU o INPSU, he esul ing alue was 0.001 (signi i can a p < 0.01). Coope a ion in he c ea ion o p oduc inno a ions has p o en o be mo e signi i can in bo h cases – coope a ion on p oduc and p ocess inno a ion. Fo p ocess inno a ion, i was e ealed ha only INPRU coope a ion leads o posi i e e ec s (leading o he same alue o 0.000). The combina ion o INPSI*INPSU was analyzed o be insigni i can in his case. E en hough INPSU p o ed o be insigni i can , u he analysis e ealed a combina ion o a iables wi h INPSU ha led o impo an esul s (e.g. he combina ion wi h he INPSI a iable). The combina ion o a iables INPSU*INPRU o INPSU*INPRE showed a posi i e in e ac ion ha in l uenced he en e p ises’ o e all pe o mance. The alue o hese impac s we e measu ed iden ically a 0.001 (signi i can a p < 0.01). To es hypo hesis H3, we analyzed he impac o public subsidies on en e p ises’ pe o mance. Tab. 2 shows ha he p o ision o public subsidies om EU unds (FUNEU) was insigni i can (0.755). The e o e, we included he a iable FUNGMT and c ea ed Model 2 INPRU INPSU INPRI 0.001*** (1.323) 0.001*** (1.303) INPSI 0.000*** (0.691) 0.807 (1.175) Sou ce: own Legend: ***signi i can a p < 0.01; he alues o s anda d de ia ions a e in b acke s Tab. 3: The In l uence o Public Coope a ion on En e p ises’ Inno a i e Ac i i y EM_3_2016.indd 116EM_3_2016.indd 116 8.9.2016 14:11:228.9.2016 14:11:22 117 3, XIX, 2016 Business Adminis a ion and Managemen 117 o analyze whe he public unds inc ease he en e p ises’ pe o mance. The co ela ion coe i cien R om he second eg ession model showed a alue o 0.979. The coe i cien o de e mina ion R2 was 0.959. The p- alue o his model was measu ed a 0.000. The p- alue showed ha his model was signi i can a p < 0.01. The e o e, he null hypo hesis was ejec ed. Model 2 is ega ded as signi i can . Howe e , he esul s showed ha he e a e no signi i can in e ac ions in l uencing he en e p ises´ pe o mance. This esul allows us o con i m hypo hesis H3, in which we a gued ha he e is ine i ciency in p o iding public subsidies. The esul s in Tab. 2 also con i m his claim (FUNEU is insigni i can in Model 1). Howe e , we also ound posi i e combina ions o a iables ha con i m he signi i can in l uence o Eu opean public unds in Model 1. The esul s in Tab. 4 no only show ha he impo ance o coope a ion and pa icipa ion in a g oup o en e p ises posi i ely in l uences he spending o public unds bu also con i m ha i is possible o well a ge ed public spending o lead o e ec i e esul s. When public unds a e p o ided o en e p ises indi idually, no signi i can in e ac ions we e ound and he p alue was measu ed a 0.755 (see Tab. 2). Con e sely, s ong ies in l uencing en e p ises’ pe o mance eme ge when public suppo is a ge ed co ec ly and he combina ion o FUNEU*CO o FUNEU*GP exis s. These alues a e 0.000 (signi i can a p < 0.01). O he s ong ies occu ed a he highe o de in e ac ions. The combina ion o GP*INPSI*FUNEU o INPSI*CO*FUNEU showed signi i can in e ac ions wi h a alue o 0.000, which a e also signi i can a p < 0.01. E en hough an ine i cien use o public esou ces occu s in mos cases, i is possible o i nd cases whe e hey a e used e ec i ely. Cza ni zki and Lopes-Ben o (2014) came o a simila conclusion. They sugges ed ha public unding igge s socially bene i cial esea ch p ojec s (bu hey don’ alk abou economic e i ciency, which is o en missing in hese p ojec s) and ha he co-exis ence o na ional and Eu opean policies does no lead o c owding-ou e ec s when compa ed o a hypo he ical wo ld wi h a closed economy and no supplemen al Eu opean policies. Discussion and Conclusions The in en o his pape was o analyze whe he en e p ises´ collabo a ion on inno a i e ac i i ies in he Czech Republic’s machine y indus y has a posi i e impac on hei o e all pe o mance by examining h ee esea ch hypo heses. The esul s o he eg ession models ha we e composed allowed us o con i m indi idual hypo heses. Hypo hesis H1 indica es he need o sys ema ically p omo e coope a ion be ween en e p ises and o he pa ne s and c ea e g oups o en e p ises ha ha e an impac on he g ow h o en e p ises´ o e all pe o mance. This i nding con i m he concep o open inno a ion ha is based on di e en esea ch s eams and sugges ha aluable ideas can come om inside o ou side he en e p ise and can go o ma ke om inside o ou side he en e p ise as well (Chesb ough, 2006; Chesb ough & Appleya d, 2007), he e o e, coope a ion is seen as a c ucial way o inc ease en e p ises´ g ow h o u no e . The esul s o hypo hesis H2 allow us o s a e ha en e p ises in he Czech Republic’s machine y indus y ha coope a e wi h uni e si ies and public esea ch cen e s demons a e FUNEU CO 0.000*** (0.392) GP 0.000*** (0.965) GP*INPSI 0.000*** (0.464) CO*INPSI 0.000*** (0.403) Sou ce: own Legend: ***signi i can a p < 0.01; he alues o s anda d de ia ions a e in b acke s Tab. 4: Va iables wi h a posi i e impac on public subsidies EM_3_2016.indd 117EM_3_2016.indd 117 8.9.2016 14:11:228.9.2016 14:11:22 118 2016, XIX, 3 Ekonomika a managemen 118 a g ea e posi i e in l uence on hei o e all pe o mance. Uni e si ies ep esen impo an collabo a i e pa ne s in he Czech Republic’s machine y indus y. These collabo a ions be ween en e p ises and uni e si ies posi i ely a ec en e p ises´ c ea ion o inno a ion – bo h p oduc and p ocess. The e o e, we ecommend con inuing o suppo hese ypes o coope a ion in he machine y indus y in he Czech Republic and c ea ing links be ween en e p ises and uni e si ies. This emains ue in he case ha en e p ises do no ecei e suppo om public budge s. We a gue ha he e is ine i cien p o ision o public aid in he machine y indus y in he Czech Republic, which is possibly caused by excessi ely g an ing public unding wi hou di ec ly moni o ing i s impac ; his was subsequen ly con i med by hypo hesis H3. A numbe o p ojec s i nanced by he Eu opean Union a ise only in o de o exploi hese unds, wi hou anyone o o e see he e ec i eness o hese p ojec s. Excessi e bu eauc acy is ano he eason ha leads o ine i ciency when p o iding public aid. On he o he hand, ad anced analysis showed ha , i public unding is p ope ly a ge ed, i may be e ec i e and powe ully in l uence en e p ises´ collabo a ion ac i i ies and o e all pe o mance. The e o e, i is necessa y o adop an app op ia e way o p o ide public suppo o he machine y indus y in he Czech Republic. I should be no ed ha inno a i e ac i i ies a e essen ial pa ame e s a egic (i.e. long- e m) pe o mance. Rele an da a in an app op ia e ime ange a e c i ical o he explana o y powe o e e y simila esea ch. The limi a ion o his esea ch is a smalle ange o he analyzed sample. Fo his eason, he gene alizabili y o esul s is limi ed. We ecommend ha go e nmen s pu mo e emphasis on adap ing subsidies o speci i c en ep eneu ial ac i i ies and p omo ing he eme gence o common linkages be ween en e p ises and uni e si ies. In u u e esea ch, we plan o analyze he issue o collabo a ion pa ne s and p o iding public subsidies o o he indus ies in he Czech Republic and subsequen ly compa e he esul s wi h hose o o he Eu opean coun ies. This pape was c ea ed as a pa o he solu ion o he esea ch ask No. 14-02836S en i led “Modeling o knowledge spill-o e e ec s in he con ex o egional and local de elopmen ” and No. 16-13119S en i led “Pe o mance managemen in public adminis a ion – heo y s. p ac ices in he Czech Republic and o he CEE coun ies”, i nancially suppo ed by he G an Agency o he Czech Republic. Re e ences Acemoglu, D., Gancia, G., & Zilibo i, F. (2012). Compe ing engines o g ow h: Inno a ion and s anda diza ion. Jou nal o Economic Theo y, 147(2), 570-601. doi:10.1016/j.je .2010.09.001. Acs, Z. J., Aud e sch, D. B., & Lehmann, E. E. (2013). The knowledge spillo e heo y o en ep eneu ship. Small Business Economics, 32(15), 757-774. doi:10.1007/s11187-008-9157-3. Acs, Z. J., de G oo , H. L., & Nijkamp, P. (Eds.). (2013). The eme gence o he knowledge economy: A egional pe spec i e. Sp inge Science & Business Media. doi:10.1007/978-3-540-24823-1. Al a o, L., & Chen, M. X. (2013). Ma ke ealloca ion and knowledge spillo e : The gains om mul ina ional p oduc ion (BGIE Uni Wo king Pape ). p. 12-111. Ha a d Business School. Al bach, P. G. (2013). Ad ancing he na ional and global knowledge economy: he ole o esea ch uni e si ies in de eloping coun ies. S udies in highe educa ion, 38(3), 316-330. doi:10.1080/03075079.2013.773222. Ba elsman, E., Dobbelae e, S., & Pe e s, B. (2015). Alloca ion o human capi al and inno a ion a he on ie : i m-le el e idence on Ge many and he Ne he lands. Indus ial and Co po a e Change, 24(5), 875-949. doi:10.1093/icc/d u038. Becke , W., & Die z, J. (2004). R&D coope a ion and inno a ion ac i i ies o i ms – e idence o he Ge man manu ac u ing indus y. Resea ch policy, 33(2), 209-223. doi:10.1016/j. espol.2003.07.003. Bee s, C., & Zand, F. (2014). R&D coope a ion, pa ne di e si y, and inno a ion pe o mance: an empi ical analysis. Jou nal o P oduc Inno a ion Managemen , 31(2), 292- 312. doi:10.1111/jpim.12096. Belde bos, R., Ca ee, M., & Lokshin, B. (2004). Coope a i e R&D and i m pe o mance. Resea ch policy, 33(10), 1477- 1492. doi:10.1016/j. espol.2004.07.003. Belde bos, R., Ca ee, M., Diede en, B., Lokshin, B., & Veugele s, R. (2004). EM_3_2016.indd 118EM_3_2016.indd 118 8.9.2016 14:11:228.9.2016 14:11:22