INTRODUCTION
PROBLEMSTATEMENT
Smallm e d i u m enterprises(SMEs)h a v e playedanimportantr o l e i n economicde velopment.Theyarealsoanessentialsourceofjobcreation,innovation,increasingthecompeti tivenessandt h u s t h e e n g i n e o f developedandd e v e l o p i n g c o u n t r i e s (Europeia20 05).InVietNam,w i t h m o r e t h a n 3 0 0 , 0 0 0 registeredenterprises(GeneralS t a t i s t i c s
2011),SMEsplayacrucialroleintheeconomyreform,notonlyrepresentingthem a j o r perce ntage(97,6%)ofbusinessesofthecountry,butalsosignificantlycontributingt o
Fostering operational efficiency and productivity within a firm is crucial for enhancing competitiveness in the global market Innovation plays a vital role in boosting productivity, as evidenced by Crespi and Zuñiga (2012), which highlights that technological advancements lead to a more effective use of productive resources and the transformation of new ideas into economic solutions, including new products, processes, and services As a result, innovation serves as the foundation for sustainable competitive advantages and is essential for achieving lasting increases in productivity Extensive research has been conducted in this area across various developed and developing countries, including studies by Chudnovsky and López et al.
(2006),MassoandVahter(2008),RoperandLove( 2 0 0 2 ) ;however,firmsi n developingcount ries,especiallyi n S M E sector,d o n o t alwaysp r o p e r l y c o n s i d e r t h e impacto f i n n o v a t i o n o n t h e i r performance.Iti s n o s u r p r i s e t h a t t h e r e i s a l a c k o f studieso n t h i s s u b j e c t i n Asiancountries,especially Vietnam.Twoauthorshavestudiedasimilartop ic.ThefirstoneisNguyen,QuangPhametal.
2 |Page ngVietNamS M E surveyin 2 0 0 5 ; andanotherresearchwasd o n e byLang,Linetal.(2012),studiedtheeffectsofinnovationcapabilitiesonthefirm’s performance.Therefore,themainpurposeofthispaperistocontributethefindingsoftherelatio nshipbetweeni n n o v a t i o n anda f i r m ’ s productivity,whichd o e s n o t seemt o b e invest igatedforVietnameseSMEsbefore.Thispaperexaminestheimpactofinnovationo n the firm’sproductivityusingmicrodatafromVietnamSMEsurveyfortheperiodfrom2 0 0 7 -
The Douglas production function and fixed effect model reveal a strong association between innovation and productivity in Vietnamese SMEs The study highlights that factors such as firm size, location, and manufacturing sector significantly influence this relationship Notably, the impact of innovation on productivity is lower in major cities like Hanoi and Ho Chi Minh City compared to smaller cities Interestingly, the research found no significant effects of firm size or high-tech industry on the innovation-productivity relationship, with micro firms (those with fewer than 10 employees) showing similar productivity levels to larger firms Additionally, both high-tech and low-tech industries exhibit comparable productivity when they engage in innovative practices.
R ESEARCHOBJECTIVES
Withtheaboveproblemstatement,thisthesisaimstoinvestigatetherelationshipbetw eeni n n o v a t i o n a n d a firmproductivity,andt h e influencingfactorso n t h i s relationship,t oassistwithafirm’sdecisionaboutinvestingininnovationforproductivitybenefits.Specifica lly,this thesishastwomainobjectives:
R ESEARCH Q UESTIONS
Ino r d e r t o meett h e a b o v e objectives,t h i s paperattemptst o answert h e f o l l o w i n g t w o questions:
(ii) Whataret h e roleso f f i r m s i z e , firmlocation,andmanufacturings e c t o r o n t h e im pactof innovation andfirmproductivity?
S COPEOFTHESTUDY
Toansweraboveresearchquestionsandmeettheresearchobjectives,t h e s t u d y relieson datafromtheSmall andMediumEnterprises(SMEs)inVietNamfortheyear2 0 0 7 and200 9.Inthesurvey,thereareapproximately2500enterprisesof10provincesinVietNam;mostofthe maremicroandsmallenterprises,asthemajorityo f SMEsnumbers,t h u s theycanrepresentf o r
S M E s p o p u l a t i o n T h e studyhasa l s o focusedo n manufacturings e c t o r , w h i c h is consideredtohavehighertechnologicalintensitya n d i n n o v a t i o n involvementsthanoth ersectors.
S TRUCTUREOFTHESTUDY
Followedbythisintroductorychapter,chapterTwoprovidestheliteraturerevieww i t h boththeoreticalandempiricalfindingsfrompreviousstudies.Itpresentsmeaningofk e y con cepts,t h e m e a s u r e m e n t s , t h e d e b a t e s a b o u t relationshipbetweeni n n o v a t i o n andp r o d u c t i v i t y andotherinfluencingfactorsatmicrolevelandfindingsinVietNam.Thenitw i l l specifytheconceptualframework.
ChapterThreedescribesthedataandresearchmethodology.Thefirstpartdefinest h e variablesandc o n c e p t s whichareusedi n t h e t h e s i s andt h e i r measurements.T h e sec ondpartintroducestheempiricalmodelsandresearchhypothesesthatwillbetested.Itw i l l f i n a l l y presenttheestimationstrategyorregressionmethodologyforpaneldata.
T h i s chapteralsoanalyzesh o w t h e regressionresultscananswert h e researchquestions,how theresearchhypothesesinchapterThreewillbetestedandhowitrelatest o t h e previousfindings.
Then e x t chapteri s t h e conclusion,whichsummarizesallo f abovechaptersandbasedon thefindingresultsinchapterFour,italsosuggestssomestrategiestoVietnameseenterprisesfor thelong-termdevelopmentandgrowth.
All ofappendicesandreferencesshall be providedin thefinalpart.
LITERATURE REVIEW
PRODUCTIVITY : CONCEPTSAND MEASUREMENTS
Productivityiscommonlydefinedasa ratioo f a v o l u m e measureof ou tp ut toa v o l u m e measureofinputuse(SchreyerandPilat2001)orinotherwords,how muchofo u t p u t whichisobtainedfromagivensetof inputs(Syverson2010).
Measuring productivity serves to identify changes in innovation, assess efficiency from technological advancements, and evaluate real cost savings or benchmark production processes at a micro level At a macro level, it is utilized to gauge the development of living standards Various types of productivity measures exist, tailored to specific measurement purposes and data availability Generally, these measures are classified into single-factor productivity, which relates output to a single input, and multifactor productivity (MFP), which connects output to multiple inputs Single-factor measures can focus on labor or capital, while MFP can involve combinations such as capital-labor or capital-labor-materials Both types can be assessed using either the gross outputs concept, which reflects total sales or production outputs without deducting intermediate inputs, or the value-added concept, which represents gross outputs minus the cost of intermediate inputs.
Therearesomeadvantagesanddisadvantagesinusingthesetwoconcepts.OECDP r o d u c t i v i t y Manualmentionedthatiftechnicalprogressaffectsallinputsproportionately,t h e n g r o s s o u t p u t p r o d u c t i v i t y measuresgivee s t i m a t e s o f u n d e r l y i n g technicalprogre ss
The value-added measure is influenced not only by technology but also by the time paths of outputs, inputs, and prices, making it a reflection of an industry's ability to convert technical changes into income and final demand While it could accurately measure technical change if such change exclusively impacted primary inputs like capital and labor, this is not always the case Consequently, these two measures serve different purposes based on the objectives of measurement and the availability of data.
Aswehaveknown,productivityisatechnicalconceptwhichmeasuresthee f f i c i e n c y fromtheusedfactorsofproductionofaSME.Higherproductivityislikelytoi m p r o v e profitabilityandenhanceafirm’scompetitivenessrelativetoitsrivals.However,w h y dofir msdiffersomuchintheirabilitytoconverttheinputstooutputs?
According to the Cobb-Douglas theory of production (1928), productivity is fundamentally influenced by labor, capital, and total factor productivity (TFP) An increase in labor or capital inputs, along with TFP, results in higher output levels While labor and capital are tangible inputs, TFP encompasses intangible elements such as technology and worker knowledge, also referred to as human capital Since Solow's work in 1957, TFP has been recognized as a critical indicator of how efficiently firms convert inputs into outputs, playing a significant role in economic growth, alongside labor and capital investment Consequently, variations in a firm's technological innovation affect its capacity to transform inputs into outputs, as illustrated in Figure 1, which demonstrates productivity improvements driven by technical change as a component of TFP Within the production frontier, firms can achieve greater outputs with the same inputs, highlighting the importance of innovation in enhancing productivity.
Atmicrol e v e l , TFPi s a m e a s u r e o f elementssuchasmanagerialcapabilitiesrese archandd e v e l o p m e n t , technicali n n o v a t i o n Ina TFPs u r v e y a c r o s s t h e developingc ountriesd u r i n g t h e p e r i o d 2 0 0 6 - 2 0 0 9(EnterpriseN o t e N o 2 3 –
WorldB a n k G r o u p , 2011),t h e r ea r e f i v e A s i a n countriesIndonesia,Mongolia,Nepal, PhilippinesandVietNam.TheaverageTFPvalueofthesecountries is0.03.Nepalhasthehig hestaggregatep r o d u c t i v i t y l e v e l (0.38)whichi s followedbyIndonesia( 0 2 7 ) T h e lowestaggregatep r o d u c t i v i t y i s observedi n VietN a m ( -
0 0 0 4 ) T h u s , evenm o s t o f S M E s i n VietN a m believethattechnologyinnovationist hemainfactorinfluencingtheircompetitivenessint h e market,butfromabovefigures,weca neasilyseemostofVietnameseSMEscurrentlyhavel o w p r o d u c t i v i t y a n d competitivene ssb e c a u s e o f l o w investmentintechnologicali n n o v a t i o n T h e negativeaggregate p r o d u c t i v i t y s h o w e d t h a t t h e p r o d u c t i v i t y andi n n o v a t i v e investmentsare decreasingovertime.
Onanotherhand,Hansen(2006)hasf o u n d t h e p o s i t i v e andsignificanteffecto f i n n o v a t i o n o n t h e survivalo f S M E s i n h i s s t u d y u s i n g datao f VietnameseS M E s from 1990-
2000.Thus,itisvitaltostudytherelationshipbetweentechnologyinnovationandfirm productivity,whichh o p e f u l l y helpSMEs toenhancet he ir productivity throught h e t e c h n o l o g y i n n o v a t i o n , increaset h e competitivenessande f f i c i e n c y f o r t h e b e t t e r performanceanddevelopment.
INNOVATION : CONCEPTS ANDMEASUREMENTS
Innovation, as defined by Greenhalgh and Rogers (2010), involves applying new ideas to a firm's products, processes, or activities, resulting in increased value for both the firm and its consumers There are two main types of innovation: product innovation and process innovation Product innovation refers to the introduction of new products, designs, or significant qualitative changes to existing products In contrast, process innovation focuses on developing new techniques for producing or delivering goods and services Although distinct, these two types of innovation are closely linked; new processes often enable the creation of new product designs, while new products may necessitate changes in production processes to enhance efficiency and reduce energy consumption.
(product or process)Market Diffusion
(new idea, research, design) andmostimportantlyimprovingproductivity.However,creatinganewfirmormakingane winvestmentinaplantorfactoryisalsoconsideredasinnovativeactivity(Audretsch,Santar ellietal.1999).
Product and process innovation primarily leads to cost reduction in production, thereby improving a firm's competitiveness in the global market This innovation process typically begins with research and development activities, including market surveys, demand analysis, idea generation, and product testing and design Research and development (R&D) is a crucial component of the innovation process and plays a significant role in driving economic growth (Crepon, Duguet et al., 1998).
(BaldwinandBranch2000).Thisactivityhelpsformingthemarketneedsandthefirm’sres ponsest o t h e marketanalysis.Aftert h a t , t h e investmentf o r i n n o v a t i o n s h o u l d b e implementedf o r p r o d u c t , e x i s t i n g product,andt e c h n o l o g y orfo rt he w h o l e process.
A t t h i s stage,wewillprobablyknowhowmuchinnovationimpactsontheproductivity,ho wefficientofthefirm’sperformanceandhowmuchcostswillbepotentiallysaved.Lastbutn o t least,i t i s t h e t i m e f o r marketpenetrationandadaptation.T o s o m e circumstances,adjustme ntsorimprovementswill berequiredinthisstage.
Followingt o i n n o v a t i o n process,manysurveysandresearchesh a v e s t u d i e d t h e impactofR&D,innovationandproductity.Forexample,CommunityInnovationSurveys(CI
S)isthemostwell- knownsurveyexecutedbytheEuropeanUnionformeasuringtheproductandprocessinnova tion,i n n o v a t i o n a c t i v i t y andexpenditure,impactsofi n n o v a t i o n , othersourcesandfindingsofinnov ationofEuropeanenterprises.MostofthefirmsinEuropeandothercountrieshaveimplemente dthesurveysofinnovationactivitiesbasedo n R & D expendituresandpatentcountsa s indicator so f t h e i n p u t ando u t p u t o f innovation.MiguelBenavente(2006)useddataofChileanplant stostudytherelationship ofresearchinvestment(measuredbyR&Dperworker)andinnovation (withinnovation sale susedasaproxy)onlaborproductivity(measuredbyvalueaddedperworker).Inthisresearch,i n n o v a t i o n s a l e s i s measuredasa shareo f salesandusedi n thet o b i t m o d e l Otherinstrum entvariableswere alsoinvestigatedsuchasmarketshare,diversification.
1 9 9 6 Importantly,afirmiscalledaninnovatordependingonitsoutputoftheinnovation processb u t n o t o n whetheri t hasinvolvedi n i n n o v a t i v e activitieso r i nn ov at io n i n p u t s Innovatorsinthispaperwerealsoclassifiedintothreegroups:productinnovation,processi n n o v a t i o n a n d b o t h product&processinnovation.
LửửfandHeshmati( 2 0 0 6 ) definedi n n o v a t i v e firmi s wheni t s i n n o v a t i o n inve stmentandinnovativesalesarepositive.Theirmeasureofinnovationinputswasmorecomprehens ivethanotherresearches,asitnotonlyincludedR&Dspendingbutalsonon-
R&Dactivities,the outsourceservices ormachinery forinnovation activiti es, allre latedexpensesi n education,marketing,designf o r n e w products…Mohnen,Mairesseetal. (2006)consideredinnovationastheresidualofinnovationproductionfunctionandasparto f inn ovationintensitydueto theimprovementandinvestmentinnewproducts.
Insummary,manyindicatorsandproxieshavebeenusedasameasureofi n n o v a t i o n inputs andoutputs,dependingontheavailabilityofdata,thesurveyqualityandt h e purposesoftheinves tigation.ButthemostpopularindicatorsareR&Dexpenditures,patentcountsorinnovationsal es.Despiteofwhichindicatorswe useorhowwedefinetheinnovationprocess,innovatorscouldbeexpectedtohaveabetterperform anceorp r o d u c t i v i t y than non-innovators.
R ELATIONSHIPOFINNOVATION AND PRODUCTIVITY
BasingonCobbandDouglas(1928),theproductivityisthefunctionoflabor(L),capit al(K),materials(M)andtotalfactorproductivity(A)
Takingthenaturallogarithm of thefunction,wehave: log(Y)=log(A)+βlog(L) +αLβlog(K)+γlog(M)
Total Factor Productivity (TFP) is a crucial determinant of a firm's output, influenced by observable inputs such as capital (K), labor (L), and materials (M) The output elasticities of these inputs—denoted as α, β, and γ—play a significant role in this relationship While capital, labor, and materials are tangible inputs, TFP often embodies more intangible aspects of productivity Variations in TFP among firms highlight their differing capabilities to transform inputs into outputs Furthermore, innovation and technical change are integral components of TFP, which is widely recognized as a driving force behind economic growth, contributing up to 87.5% to overall productivity advancements.
%increaseoftotal- factorproductivityhascontributedtothed o u b l e d grossoutputpermanhour,andthere maining12.5%wasfromtheincreaseduseo f capital(Solow1957).Atthefirmlevel,howtheinn ovationortechnicalchangecontributest o t h e firmproductivity?
BesidesCobb-Douglasp r o d u c t i o n function,therei s a well- knownm o d e l , whichwasverypopularly used-
C D M Model,d e s c r i b i n g t h e r e l a t i o n s h i p betweeni n n o v a t i o n andproductivity.Itwas developedbyCrepon,DuguetandMairesseinCrepon,Duguetetal.
Inbelowmodel(Figure2),innovationisconsideredasaprocess,whichiscarriedo u t fr omtheengagementinR&Dactivities,i n v e s t m e n t i n technologyo r knowledgecapitalandalsoaffectedbyotherfactorssuchas:marketdemands,fi rmsizeorindustries.T h e processinnovationcanimprovetheproductionperformanceand makeitmoreefficient,thereofenhancetheproductivityofthefirm.Inthismodel,researchan ddevelopmentwass t r o n g l y emphasizedb e c a u s e o f i t s impacto n t o t h e restcomponents T h e squareboxesdenotethemeasurablequantityconcepts,whiletheovalboxesrepresentt h e immeasurablefactorsandwenormallyneed tousethe proxiesforthesefactors.
CDMmodelhasbeenappliedinmanystudiesduetoitspracticalitysuchas:LửửfandHes hmati(2002),MiguelBenavente( 2 0 0 6 ) ,MassoandVahter( 2 0 0 8 ) …IfC o b b -
Diversification Market share Size/ Industry
Douglasallowsustoconsidertherelationshipbetweeninnovationandproductivityonly,t h e n CDMmodelwouldenableustomodelthisrelationshipinthebiggerframeworkandtake into accounttheimpact ofotherinfluencing factors.Severallinksinthis structurew i l l b e capturedanddiscussed inanalysissectionofthispaper.
Atmicrolevel,innovationinfluencest h e firm’sp r o d u c t i v i t y w i t h a directandin directimpact.
(2006)strongly suggestedthatinnovatorsattain higherp r o d u c t i v i t y l e v e l s t h a n non- innovatorsi n t h e s t u d y o f Argentinem a n u f a c t u r i n g firms’behaviorsd u r i n g 1 9 9 2 – 2001.Specifically,t h e e s t i m a t i o n resultshadsuggestedt h a t t h e l a b o r productivityof innovatorsis14.1%higherthannon- innovators,whichwasasignificantdirectimpacttothefirm’sproductivity.Theformerperfor medbetterthanthel a t t e r groupi n termso f l a b o r productivity.T h i s paperhasemployeddi fferentempiricalmethodologiesf o r a n a l y z i n g t h e relationshipo f i n n o v a t i o n andp
10|Page r o d u c t i v i t y basedonC D M approach.Paneldataandfixedeffecte s t i m a t o r s hadbee nu s e d t o controlf o r
11 |Page unobservableh e t e r o g e n e i t y att h e firmlevel.Additionally,t h e a u t h o r includedt h e t i m e d u m m y tocontrolt h e specifict i m e varyingunobservableeffectso f t h e firmsovert i m e an dc l a s s i f y t h e s u r v e y e d firmsi n t o f o u r groups:l a b o r intensive,s c a l e i n t e n s i v e , R
&Dintensiveandnaturalresourcesi n t e n s i v e f o r controllingt h e changingofsectorialtechnol ogicalopportunitiesovertime.Theyareconsideredasthestrengthsofthisresearch.
(2006)alsousedCDMmodelandfoundthatproducti n n o v a t i o n wasassociatedwithhigh erproductivityinFrance,Spain,andtheUK,butnoti n Germany.Similarly,MassoandVahter(
2 0 0 8 ) suggestedt h a t firms,w h o havem o r e resourcestoinvestininnovativeactivitiesand ahigherabilitytoundertakeR&Dwillgett h e improvementinproductivity.Theyhavealsofou ndtheeffectofinnovationonp r o d u c t i v i t y notonlyontheproductivityinthelastyearoft heinnovationsurvey,butalsoo n e andt w o yearsaftert h e survey.T h i s w e n o r m a l l y calle dt h e l a g o f t h e i m p a c t o r spilled- overeffect.Onemoreinterestingfindinginthisresearchwasthedifferentresultsw i t h diff erentuseddataset.With
The Community Innovation Survey 4 (CIS4) data indicates that only process innovation significantly enhances labor productivity, unlike product innovation In contrast, the CIS3 data reveals that product innovation has a more substantial impact on productivity than process innovation Additionally, organizational innovation is found to positively influence productivity The CDM model is frequently utilized by economists and researchers to analyze the relationship between innovation and productivity at the firm level due to its comprehensive coverage and practical application.
Withanothera p p r o a c h , u s i n g linearr e g r e s s i o n m o d e l andq u a l i t a t i v e questionnaires,Barlet,Duguetetal.
1 9 9 0 T h e findingswerebiasedinproductimprovementasitachievedahighcommercial return(4%),evenw i t h moderatetechnologicalopportunities.Interestingly,productsthataren ewforthefirmb u t notforthemarketneverachieveagreatgain.Thehighestcontributiontot hemanufacturingsalescomesfromproductsthatarenewforthemarket.WhileHuergoandJau mandreu(2004)statedthatprocessinnovationsatsomepointleadtoextraproductivitygrowth.
Anothers t r o n g r e l a t i o n s h i p betweeni n n o v a t i o n ( b o t h productandproc essi n n o v a t i o n ) hasbeenassertedbyH a l l , Lottietal.
2 0 0 3 B y u s i n g t h e combinationo f CDMm o d e l andC o b b Douglasproductionfunction, t h i s researchhadd e v e l o p e d differentm o d e l s f o r e x a m i n i n g t h e relationshipbetweenR&Dandinnovation,innovationandproductiv ity.Laborproductivitywasmeasuredbyrealsalesperemployee,whileproductandprocessin novationwasusedasap r o x y forinnovation input.The resultshave shownthat produ ctinnovation haspositive impacto n l a b o r productivity,w h i l e processi n n o v a t i o n h aslargereffectv i a associatedcapitalinvestment.
However,fora less developed country likeChile,MiguelBenavente(2006)wasn o t beabletofindanysignificantimpactfrominnovationonthesalesandproductivityint h e sh ort-runin1995-
1998.Thiscouldbeexplainedthattheinnovationwillneeds o m e t i m e s t o waitf o r mar ket’sresponseso r r e a l l y impacto n t h e f i r m ’ s productivity,e s p e c i a l l y inthelong- runperiod.Thestudyalsofoundthesignificanteffectoflaborskillso n t h e estimationof productivityinstead.
Whenwetakealookatthein- directeffect,innovationislikelyleadtothesustainablecompetitiveadvantageorfirm’sperforma nce.Lengnick-
Hall (1992) stated that innovation and competitive advantage are interconnected, with successful innovations enabling firms to expand their market appeal through cost-saving systems Additionally, research by Roper and Love (2002) highlighted that innovation significantly impacts a firm's export performance In Germany, a higher intensity of innovation correlates with a lower percentage of sales from new products Furthermore, spillover effects were discussed, revealing that innovative UK plants effectively capitalize on spillovers from the innovation activities of companies within the same sector, unlike their non-innovative counterparts.
(2010)Foundthatproductinnovation- notprocessinnovationofSpanishmanufacturingfirms,affectedproductivityandhelpeds mallnon- exportingfirmstoentert h e exportmarket.Innovatingfirmshadhigherproductivity leve lsandgrownfasterthannon-innovatingfirms.
Ateconomicsl e v e l , i n n o v a t i o n o r totalfactorproductivity,w h i c h i s k n o w n i n S o l o w (1957)i s t h e corefactorandd r i v i n g f o r c e o f economicgrowth(Greenhalghand
Rogers2010).Iftheeconomybasesmerelyoncapitalaccumulationwithouttechnologicalprogre ss,thediminishing returnsoncapitalaccumulationwilleventually depresseseconomicg rowthtozero On theother hand,LeVan(2008)hasfound th at thericherac o u n t r y is,t h e m o r e mo ne y willb e investedi n newt e c h n o l o g y , t r a i n i n g andeducation.Fọre,Grossko pfetal.
(1994)usedMalmquistindexoftotalproductivitygrowthtoestimatet h e impacto f t e c h n o l o g y c h a n g e o n p r o d u c t i v i t y growtho f 1 7 OECDindustrializedcountriesoverthepe riodfrom1979–
Therearemanydifferentpointsofviewsandstudiesontheimpactofinnovationo n fir m’sproductivity,tosomecertainextent,wecaneasilyseemostofthefindingshavet h e similarc o n c l u s i o n s w i t h a p o s i t i v e r e l a t i o n s h i p andveryfewhavenegativer e s u l t s Innovat ionhasimpactedn o t o n l y atfirmlevelindifferentchannels,b u t alsoatmacrolevel.However,wi thint h e s c a l e o ft he t h e s i s , w e m o s t l y focuso n t h e f i r m levelt o s e e whatpreviousstudi eshavef o u n d t h e impacto f i n n o v a t i o n o n productivityando t h e r relateddeterminan tsof thiscorrelation.
Inanalyzingthedeterminantsoftherelationshipbetweeninnovationonproductivity,m anypapers haveu s u a l l y focusedi n C D M m o d e l (Crepon,Duguetetal.1 9 9 8 )t o evaluatet h e i m p a c t s o f i n f l u e n c i n g f a c t o r s o n t h e r e l a t i o n s h i p : firms i z e , firmlocationand manufacturing sector.Inaddition,since researchanddevelopment playsanimportantr o l e ast h e pre- innovations t e p , t h u s t h e f o l l o w i n g sectionw i l l p r o v i d e t h e empiricalreviews onthesedeterminantsontherelationshipbetweeninnovationandp r o d u c t i v i t y andth erole ofR&Don this impact. a Firmsize
Firmsizeisclassifiedbasedonnumber ofemployeesorinvestedcapitalamount.Iti s oneoftheimportantfactors,whichdirectlyaffectst hefirm’sproductivity.Withi n n o v a t i v e activities,MassoandVahter(2008)foundthatth elargerfirmsaremorelikelyt o engageininnovationthansmallfirms.Firmsizehasaninsignifi cantimpactonproduct innovationbutpositiveimpactontheprocessinnovation.Morespecifically,Chudnovsky,Lóp ezetal.
(2006)suggestedt h a t largefirmshavea h i g h e r p r o b a b i l i t y o f engagingi n i n n o v a t i o n activitiesandb e c o m i n g innovators.Similarly,a study(Dhawan2001 )ofUS industrialse ctordisplayedt h a t e v e n s m a l l e r f i r m s geta higherprofitr a t e b u t theyw i l l havelowersu rvivalprobability anddifficulty inaccessingthecapitalmarket.Thestudyuseda largepan eldatao f U S firmsf o r t h e 1 9 7 0 –
1989periods.T h e empiricalresultsindicatedthatsmallfirmsaresignificantlymoreproductiv ebutalsomoreriskythantheirlargecounterparts.S m a l l firmsfacemarketuncertainties,capit alconstraintsando t h e r challengesw h i c h m a k e t h e m m o r e efficientt h a n largefirms b u t m i g h t increaset h e i r r i s k i n e s s However,t h e largestfirmsh a v e a s i g n i f i c a n t l y h i g h e r p r o b a b i l i t y o f b e i n g i n n o v a t i v e (68%)thansmallormedium - sizedones(30%),whichwasfoundin(BaldwinandBranch2000).Andnosignificantdifferenc ewasfoundbetweensmallandmedium-s i z e d firmsintermsof theirlikelihoodofbeinginnovative. b Firm location
Research indicates that industry localization significantly enhances innovative activity, as demonstrated by Audretsch and Feldman (1996) and supported by studies such as Baptist and Swann (1998) Their findings reveal that firms in strong employment sectors within their home regions are more likely to innovate, suggesting that businesses located in robust clusters tend to attract new entrants and grow at a faster pace Additionally, CIEM (2010) provides compelling evidence that firms situated in urban areas or larger cities exhibit higher labor productivity and innovation rates compared to those in rural regions or smaller cities However, it is noteworthy that many of these innovation activities are primarily driven by customer requests rather than direct responses to market demands.
Manufacturings e c t o r i s o n e o f t h e keydeterminantso f i n n o v a t i o n b e c a u s e i t i s m u c h relatedtothetechnologyandproductionprocess.Therearemanyinvestigationsfor differentindustries.Inchemicalandtextileindustries,productandprocessinnovationare closelycorrelated(Baily,Chakrabartietal.1985).Inchemicalindustry,ifthereisanewphysicalo rstructuralproductthenthatisconsideredasthenewproducts.Newprocessiswhenthereisth enewequipmentorinstrumentinnovationtoproducethenewmaterials,helpchangingenviro nment,savingenergy…
Intextileindustry,theproductinnovationisnewyarnsorfabrics,w h i l e processi n n o v a t i o n i s t h e newequipmento r t e c h n o l o g y f o r i m p r o v i n g t h e s p e e d i n g ofoperationi n d y e i n g o r finishing,reducingt h e i n p u t requirements.Inp a p e r industry,GhosalandNai r-
Reichert(2009)f o u n d t h a t t h e greaterinvestmentinmodernizationorinnovation,thehigherpr oductivityt h a t thefirmcanachieve,especiallytheinvestmentininformationtechnologyan ddigitalmonitoringdevices.InanotherresearchofGermanindustry,FritschandMeschede(2 001)presentedt h a t low- technologym a n u f a c t u r i n g firmsl a g b e h i n d t h e i r m e d i u m andh i g h - t e c h n o l o g y counterpartsi n producti n n o v a t i o n performance,b u t t h e y appearst o per formwellandevenbetterinprocessinnovation.ForItalymanufacturingfirms,Hall,Lottietal
(2009)f o u n d t h a t t h e impactofproducti n n o v a t i o n o n p r o d u c t i v i t y wasp o s i t i v e ands l i g h t l y strongerf o r firmsi n h i g h - t e c h i n d u s t r y t h a n low- techs e c t o r However,t h e largerando l d e r firmsseemedtobelessproductivethan smallerfirms,ceterisparibus. d Research&Development
Int h e i n n o v a t i o n process,researchandd e v e l o p m e n t i s consideredast h e p r e - i n n o v a t i o n step,whichattributestodevelopingnewproductsorprocessesforfuturegrowth.
M a n y p r e v i o u s s t u d i e s havem e a s u r e d i n n o v a t i o n byp a t e n t s andR & D expendituresofthefirm.UsingdataforFrenchmanufacturingfirmsandintervaldatafori n n o v a t i v e sales,Crepon,Duguetetal.
A 1998 study estimated that a 10% increase in R&D intensity could lead to nearly a 5% rise in innovative sales Additionally, research from the 1993 Canadian Survey of Innovation indicated that firms engaged in R&D activities had a 24% higher probability of introducing product innovations and a 15% higher probability for process innovations R&D activity, along with firm size, plays a crucial role in driving innovation Specifically, firms not conducting R&D have only an 11% chance of innovating, whereas those involved in R&D enjoy a significantly higher 41% probability of innovation.
Asmentionedintheintroductionpart,thereareveryfewresearchersstudiedaboutS M E i n VietN a m , e s p e c i a l l y i n t h e topicsrelatedt o t h e i n n o v a t i o n andproductivity.Re cently,therearetwopapersdiscussingaboutinnovation anditsimpactontheSME’sper formanceinVietNam.ThefirstresearchwasconductedbyNguyen,QuangPhametal.
(2008),u s i n g VietN a m S M E s u r v e y in 2 0 0 5 W i t h t h e logitandp r o b i t m o d e l s , t h e a uthorshavefoundthatthreemeasuresofinnovation,whichareunderstoodastheinvestmentin newproductsornewproductionprocessorimprovementofexistingproducts,are significantdeterminantsofexporting.
(2012)usedsurveyquestionnairesandcollecteddatafromV i e t n a m e s e high- techm a n u f a c t u r i n g firms.T h e researchhasf o u n d t h a t investmentcapabilityhadpositi veeffectsontheirtechnologyinnovationcapabilitiesandpositivelycorrelatedtotheirfirmc ompetitiveperformance.Thus,investmentint e c h n o l o g y innovation capabilitieswashe lpfultobusinesscompetitivenessi n VietNamenterprises.
Infact,fromthestatisticsinSMEsurvey,mostofSMEsinVietnamstillusetheout- of- datet e c h n o l o g y whichmightb e 3 t o 4 generationo l d e r t h a n t h e world’stechnologies.Be sidesthat,theyevendonotputadequateinvestmentinhumancapitalfori m p r o v i n g the technologymanagementskills.
Intheshortwords,itisveryimportantandusefultoreviewthekeystudiesintherelatio nshipbetweeni n n o v a t i o n andp r o d u c t i v i t y f o r m o d e l i n g t h e researchhypotheses.P o s i t i v e relationshipshavebeenf o u n d i n manypreviouss t u d i e s ( LửửfandH e s h m a t i 2006),(Chudnovsky,Lópezetal.2006 ),( CassimanandMartinez-
Ros 2007).Thedeterminantso f t h i s relationshiparealsoinvestigated:firms i z e , l o c a t i o n a ndindustrialsector.
RESEARCH METHODOLOGYAND DATA
A NOVERVIEWOFSMESIN V IET N AM
Definitions of SMEs vary significantly across different regions and countries, primarily based on factors such as the number of employees, invested capital, total assets, sales volume, or production capability The most commonly used criteria are the number of employees and total invested capital For instance, the European Commission provides specific definitions, which can be contrasted with the definitions used in developing countries like Cambodia and Vietnam While the employee size classification appears similar, there are notable differences in capital and asset sizes due to varying economic scales This comparison indicates that Vietnam has larger firms than Cambodia but smaller firms than those in Europe regarding invested capital.
Table 1:Summaryof maindefinitionsof SME inselectedeconomies
Companycategory Employees Capital/Assets(US$)
2006,whentheenterpriselawhasbeeneffectiveandVietNamhasbecomethememberof Wo rldTradeOrganization(WTO),thenumbersofenterprisesinVietNamhaskeptincreasingo v e r years.FromF i g u r e 4 below,w e canseet h e t o t a l numberso f enterprisesin2 0 1 1 hadgrown by2.6 t i m e s comparedw i t h t h a t i n 2 0 0 6 Ita v e r a g e l y increased21percentperyear,esp eciallytheriseofnon-statedfirms.Accordingt o t h e classificationi n Decree56/2009/ND- CP,t h e n u m b e r o f S M E s hasaccountedf o r 9 7 % -
98%oftotalnumbersofenterprises.Amongthat,microandsmallfirmskeepthesignificantnu mberswhilemedium andlargeenterprisesonlyhaveasmallpercentage.
According to the SME survey report (CIEM 2010), 30% of manufacturing SMEs are concentrated in ten major provinces, including Hanoi, Phu Tho, Ha Tay, Hai Phong, Nghe An, Quang Nam, Khanh Hoa, Lam Dong, Ho Chi Minh City, and Long An Manufacturing activities are categorized into approximately 18 different industries In the overall economy, trade and service businesses represent nearly 68% of enterprises, while industry and construction contribute 31.2%, with agriculture, forestry, and fishing comprising a minimal 1% The GSO report indicates that manufacturing and construction enterprises significantly contribute to the national budget, showcasing high investment returns and being technology-intensive, which attracts a substantial labor force.
Table3showsthataveragerealrevenueperfull- timeemployeewas73.0millionVNDin2011,whilerealvalueaddedperfull- timeemployeewas20million VND.Themedium- sizedfirmshavethehighestlaborproductivity.Inaddition,thelaborproductivityi n t h e urbana reaands o u t h e r n Vietnami s h i g h e r t h a n t h e restarea.I n CIEM(2010)report,urbanarea includes HaNoiandHCMC.
Realvalueaddedper full-timeemployee full-timeemployee
(4-digit) Industry Realrevenueperf ull-timeemployee
FromtheSMEssurveydata,theinnovationratescouldbecalculatedforthemanufact uringSMEs(Table5).Innovationratesin2005areprettyhighforallcategories,comparedw i t h 2
Between 2005 and 2009, Vietnamese enterprises experienced a significant decline in innovation rates, dropping from 40.8% in 2005 to 4.9% in 2007 and further to 2.7% in 2009 This trend can be attributed to the effectiveness of the Enterprise Law in 2005, which encouraged many businesses to establish themselves and invest in innovation and market penetration However, instead of introducing new products, companies opted to improve existing ones due to lower costs and resource requirements The decline in innovation was exacerbated by a lack of capital, limited market outlets, and a reduction in the adoption of new technologies within the business environment.
(CIEM2010).Diversificationisabitdifferent,afirmisconsideredasdiversifyingenterprisei f i t producesm o r e t h a n o n e 4 - d i g i t ISICproduct.T h i s a c t i v i t y s e e m s t o b e similaramongtheyearsfrom2005–
2009withaslightdecreasein2007andrecoveredin2 0 0 9 Buti n general,firmsh a v e p u t m o r e investmentsi n newt e c h n o l o g y a n d improvementofexistingproducts.
Source:(CIEM2008),(CIEM2010)andauthor’scalculation
Ifwetakealookintothefirmsize,locationandfirmage,theSMEsreportin2009has s h o w ns o m e interestingresults.Largerfirmsweremore li ke ly tointroduceth e n e w product lineandimprovetheexistingproducts,comparedwithmicrofirms.Urbanenterprises( i n H a N o i andH C M C ) appearedt o b e m o r e i n n o v a t i v e t h a n ruralareas.Furthermore,olderfirmsimprovetheirexistingproduct moreregularin 2009 than 2007.
Intermsofsector,Table7looksatthediversificationandinnovation ratesinsomesele ctedindustries.Firmsi n f o o d s andbeveragesh a v e l o w e r diversificationandi n n o v a t i o n ratethanfirmsinothersectors.Mostofthesectorstendtoimprovingexistingproduct sthanintroducingnewproduct
Improved (4-digit) Industry product existing product
New product New technology Improvement of existing products
- - - Firm size Firm location Firm sector
CONCEPTUALFRAMEWORKANDMODELSPECIFICATION
Fromthetheoreticalandempiricalreviewinprevioussection,theinterrelationshipbetwe eninnovationandproductivityc o n c e p t s andother controlledfactorsshouldbedescribedi n t h e f o l l o w i n g f r a m e w o r k Inw h i c h , i n n o v a t i o n andp r o d u c t i v i t y a r e highlightedasthemainfocusinthisthesis.Labor,capitalandi nnovationareindependentvariableandproductivityis thedependentvariable(CobbandDouglas1928).
Ino r d e r totestthe c o r r e l a t i o n betweeni n n o v a t i o n andf i r m ’sproductivity and other influencingfactors,IamgoingtorunvariousmodelswhichapplyCobbDouglasprod uctionfunction:
WhereY i stotalo u t p u t s in a givenyear;L i sl a b o r i n p u t o r t he totaln u m b e r o fpers onworkedf o r t h e firma givenyear;K i scapitali n p u t o r t h e realvalueo f total β
L L L L physicalassets;Mistotalmaterialinputs;Aistotalfactorproductivity.Inthisequationα,β andγar ethe output elasticityofcapital,laborandmaterialsrespectively.
Productivitycanbecalculatedast h e ratioo f ou tp ut t o a specificfactoro r to allrelevantf actorsofproduction.Inthis paper,theauthorappliesthenon-parametric measureo f productivity– laborproductivityasitgivesasimpleandfullmeaningoffirmp r o d u c t i v i t y performance. Byd i v i d i n g b o t h s i d e s byL,w e havet h e n e w p r o d u c t i v i t y function: α β γ α β γ
Wew i l l havet h e m a i n m o d e l ( 1 ) below,whichi s goingt o t e s t t h e correlationb etweeninnovation andfirm productivityandhelps toanswertheresearchquestion(1):
Model1–Innovation ln(Y it /L it )= ln A +αLβln(K it /L it )+γln(M it /L it )+(αLβ+β+γ–1)ln(L it )+δI it +ε it (1)
L it )ist h e l o g ofpr od uc ti vi ty (totalo u t p u t s peremployee)o f firmi i n year2 0 0 7 , 2 0 0 9
The article analyzes firm innovation between 2007 and 2009, using a dummy variable to indicate investments in new products or technologies, as well as improvements to existing products It employs logarithmic transformations to represent physical capital per employee (ln(K it /L it)), total inputs per employee (ln(M i /L i)), and the number of employees (ln(L i )) The error term (ε it) captures unobserved variables in the models for the specified years The parameters α, β, and γ represent the productivity elasticities of capital, labor, and materials, respectively, while the coefficient (α + β + γ - 1) for ln(L it) assesses deviations from constant returns to scale Additionally, δ signifies the productivity elasticity of innovation.
Furthermore,fortestingtheimpactsoffirmsize,firmlocationandmanufacturingsec torontherelationshipofinnovationandproductivity,wearegoingtohavefollowingm o d e l s :
The model for firm size is expressed as ln(Y it /L it ) = ln A + αLβln(K it /L it ) + γln(M it /L it ) + (αLβ + β + γ - 1)ln(L it ) + δI it + δ 1 I it * dLsize + ε it, where dLsize is a dummy variable indicating micro and small-sized firms with 10 or fewer employees In this context, δ 1 represents the productivity elasticity of innovation for these smaller firms, while δ + δ 1 reflects the overall productivity elasticity of innovation.
The model examines the relationship between firm productivity and various factors, represented by the equation ln(Y_it /L_it )= ln A +αLβln(K_it /L_it )+γln(M_it /L_it )+(αLβ+β+γ–1)ln(L_it )+ δI_it + δ_1 I_it *d_location+ε_it Here, d_location is a dummy variable indicating whether a firm is situated in major cities like Hanoi or Ho Chi Minh City, where d_location equals 1 The term δ_1 reflects the productivity elasticity of innovation for firms located in these urban centers, while δ + δ_1 represents the overall productivity elasticity of innovation.
Model4–Manufacturingsector ln(Y it /L it )= ln A +αLβln(K it /L it )+γln(M it /L it )+(αLβ+β+γ–1)ln(L it )+ δI it + δ 1 I it *In_High+ε it (4) where
In_Highis dummy variableforHigh-techindustry.In_High=1ifthefirmishigh-techindustry. δ 1is the productivityelasticityofinnovationoffirms inhigh-techindustry δ+δ 1is the productivityelasticityofinnovation
R ESEARCH HYPOTHESES
Followinghypothesis i s determinedb a s i n g ont h e theoreticalandempiricalreviews whicharediscussed inabove.
(ii) Hypothesis 2: thehigherfirmsize, innovation hasmoreimpacton theproductivity
(iii) Hypothesis3:impactofinnovationonproductivityinthebigcitiesishigherthano th e r area.
(iv) Hypothesis4:innovationinthehigh- techsectorhashigher impacton thep r o d u c t i v i t y than theothersector
Innovationcanp r o m o t e t h e productivity,t h i s argumentwassupportedbymanyrese archesinthis field,suchasChudnovsky,Lópezetal.(2006),Griffith, Huergoetal.
The relationship between innovation and a firm's productivity is positively correlated, as supported by various studies (Masso and Vahter, 2008; Huergo and Jaumandreu, 2004) Larger firms tend to be more productive and efficient than smaller ones, indicating that the impact of innovation on productivity increases with firm size (Dhawan, 2001) Additionally, the location of enterprises in Vietnam plays a significant role, with urban areas like Hanoi and Ho Chi Minh City demonstrating higher innovation rates and productivity compared to rural regions (CIEM, 2010) Consequently, technology innovation significantly influences productivity, particularly in urban settings.
NoiandHC M C i s highert ha n o t h e r regions.Lastb u t n o t least,asalwaysb e i n g understoo dalonghistory,thehigh- techindustryseemstobemoreintensivetechnologicalinnovation.T h i s iswhyweassumethe impactofinnovationishigherinthissectorlikeitstatedinthefinalhypothesis.
D EFINITIONSOFVARIABLESAND CONCEPTS
Variables Definitions Measurement Unit Expectedsi g n ofcoeffi cients
Totalproductionoutputsperemployeedeflatedbytheindustrialprodu ctionpriceindex (logarithmform) thousand VND lnOAperL
Physical assetscontainvalueofland,buildings,equipment/machinery,transpo rtequipment thousand VND positive(+) lnKAperL
Labor(L) Numberoflaborsinafiscalyear Totalnumberoffull- timeemployeesattheendofafiscalyear(logarithmform) persons positive(+/-) lnL
Material (M) Valueofrawmaterialusedi n productionprocessperemplo yeein afiscalyear
Totalvalueofrawmaterialusedperemployeeattheendof afiscalyear(logarithmform) thousand VND positive(+) lnMAperL
Dummy=1ifthefirmhasaninvestmentinintroducingnewproductorn ewproductiontechnologyo r improvet h e existingproductsi n thelas ttwoyear positive(+) innovation
Dummy=0if thefirmh a s n o investmenti n introducingn e w producta n d n e w productiont echnologya n d improvet h e existingproductsi n thelasttwoyear
Firmlocation Dummyvariableforbigcities Dummy=1if thefirmis locatedinthebigcities(HaNoi, HoChiMinhCity) Dummy=0if thefirmislocatedinsmallercitiesandotherwise positive(+) dlocation
Innovation*dlocation Dummy variable forInnovationofbigcities
Dummy=Innovation*dlocation positive(+) InLocation
Dummyvariableforsmall- sizedfirmsbyemployee Dummy variable for
Dummy=1i f thefirmsizeh a s lesst h a n o r equalst o employees Dummy=0otherwiseDummy
InLsize Innovationofsmall- sizedfirms HightechIndustry Dummyvariablefor technologyindustry high Dummy=1if thefirmishigh-techindustry
Innovation*Hightech Dummy variable for Dummy=Innovation*hightech positive(+) In_High
Innovation of high-tech industry
D ATACOLLECTION
ThisstudyuseddatafromtheSMESurveyinVietNamfor2007and2009withre peatedenterprisesinmanufacturingsector.ThesesurveyswereconductedbytheVietnameseI nstituteforLaborSciencesandSocial Affairs(ILSSA),Ministry ofLabor,InvalidsandSocial Affairs(MOLISA)w i t h t h e partnershipf r o m Universityo f Copenhagen,Denmark.Data hasbeendownloadedfromthedatabaseoftheUniversityofCopenhagen:http:// www.econ.ku.dk/derg/links/
A recent survey in Vietnam examined approximately 2,500 enterprises, including nearly 2,100 repeat firms, across 10 provinces: Hanoi, Hai Phong, Ha Tay, Phu Tho, Quang Nam, Nghe An, Khanh Hoa, Lam Dong, Ho Chi Minh City, and Long An The research categorized manufacturing firms into two groups based on their location—big cities and small cities—and assessed their technological capabilities using the OECD classification of high and low technology However, the SME data faced certain limitations and challenges due to the smaller scale of the survey, especially when compared to the Vietnam Enterprise Survey and other studies Additionally, the research noted a significant drop-off of enterprises during Vietnam's recession period starting in 2007.
2 0 0 9 , thedataislackingoftheconsistencyduetothechangesinsurveyedquestionsandm i s s i n g dataoveryears.However,besidesthoselimitations, thispaneldataisastronglyb alanceddatatherefore,i t helpst o m a k e datam o r e c o n s i s t e n t , b e a b l e t o controlt h e unobservable variablesorthedifferencesin businesspracticeacross theenterprises.
METHODOLOGY
Thissectionprovidest h e estimationm e t h o d s whichrelatedt o t h e m o d e l specific ationsandcollecteddataasstatedinprevioussections.Itwillconsideralternativeregressionmod elsassociatedwithcollectedpaneldata:therandomeffectregressionmodel( R E m o d e l ) andt h e f i x e d effectregressionm o d e l (FEmodel).Byr u n n i n g t h e Haumantest,theresultswoul dallowtheauthortoselectthebetterandmoreappropriateregressionmodel for thisresearch.
3.6.1 R ANDOM E FFECT R EGRESSION M ODEL (RE)
The model is defined as ln(Y_it / L_it) = ln A + α_L β ln(K_it / L_it) + γ ln(M_it / L_it) + (α_L β + β + γ - 1) ln(L_it) + δI_it + ε_it, where y_it = βx_it + ε_it In this framework, ε_it represents the individual effect α_i, which is time-invariant and varies across individuals, while v_it is time-variant and varies unsystematically across individuals and over time The time-invariant individual effect α_i is assumed to be uncorrelated with x_it, allowing time-invariant variables to serve as explanatory variables in the model.
Similarto theREmodel,theFEmodelisstructuredasfollow: ln(Y it /L it )= ln A +αLβln(K it /L it )+γln(M it /L it )+(αLβ+β+γ –1)ln(L it )+δ I it + ε it y it =βx it +ε it whereε it =α i +v it
InFEmodel,time- invariantindividualeffectα ii s assumedtobecorrelatedwithx ito r cov(α,x it )≠0,whichisopposi te tothe REmodel T h e FEm o d e l explorest h e relationshipbetweenpredictorandoutcom evariableswithinanentity.Eachentityhasitsowni n d i v i d u a l characteristicsthat mayormaynotinfluencethepredictorvariables.
3.6.3 S ELECTIONBETWEEN RE AND FE MODELBY H AUSMAN T EST
InHausmantest,nullhypothesisH0:cov(α i ,x it )≠0orREapproachiscorrectandH1:cov(α i ,x it )=0or
(β FE – β RE )[var(βvar(β FE )–var(β RE )] -1 (β-β)̴χ 2 (k) whereβ FE ,β REare t h e vectorso f coefficientsobtainedfromr u n n i n g FEandR E models.T h i s s t a t i s t i c i s chi-squareddistributedwithkdegreeoffreedom.
TheH0i srejectedifthecalculatedstatisticislargerthanthecriticalvalueofdistributionf o r t h e levelofsignificance.Asaresult,FEmodelispreferred.Incontrast,ifH0cannotberejected,thepre fermodel is REmodel.
M EASUREMENTSOF V ARIABLES
Productivityi st o t a l v a l u e o f production/ manufacturedo u t p u t peremployee.M a n y otherstudiesusedtotalrevenueortotalvalueadde d, but thetotaloutputsseemto bem o r e efficientasitexpressestheactualvalueofthefirm’sproductionasthestatedintheor iginalproductionfunction.Theproblemisnotalloutputsmanufacturedinaneconomicaccou ntingperiodaresoldandproperlyrecordedinthefinancialstatements.Hence,ifweu s e totalrev enueorvalueadded,thepartofoutputsremainedintheinventorieswouldbeignored. b Capital
Totalphysicalassetsofthefirmareusedasaproxyfortotalcapital.Itcontainsthevalueofla nd,buildings,factory,equipment/machinery,transport equipment c Materials
Totalmaterialstakethe value ofrawmaterialsused in theproduction.
InS M E survey,totalo u t p u t s , physicalcapitala n d materialsaren o m i n a l marketval ue,thusallvalueneedtobedeflatedandconvertedtotherealvaluebyusingProductionPriceIndex (PPI)of2007and2009. d Labor
Laborasaninputofproductionfunction,itisthetotalnumber offull-timeworkerso f t h e firmat theendofthesurveyedyear. e Innovation
Innovationisthedummyvariable.Ifthefirmhasaninvestmentinintroducingnewproduct( at4 digitISIClevel)o r n e w p r o d u c t i o n t e c h n o l o g y ori m p r o v e s t h e e x i s t i n g pr oductsinthelasttwoyear,thenitissaidto beinnovative.Ifthereisnoinvestmentinallo f thesementionedabove, theybecomenon- innovative.The mainreasons ofinnovation activitiesareformeeting therequestsfro mpurchasing customersandf o r increasing t h e competitiveadvantagetootherproducersor competitors
Ha Noi Hai Phong Khanh Hoa
Phu Tho Nghe An Lam Dong Ha Tay Quang Nam HCMC f Firmsize
CP.However,duetothedataa v a i l a b i l i t y andm a k i n g i t s i m p l e r i n t h e r e s e a r c h , I h a v e dividedi n t o 2 groups.Microfirmshavenom o r e t h a n 1 0 employeesands m a l l andm e d i u m grouph a s m o r e th an 1 0 employees. g Firm Location
Firmlocationisdividedintotwogroups:bigcitiesconcludesHaNoiandHCMC,s m a l l e r citiesconsistofHaiPhong,HaTay,PhuTho,QuangNam,NgheAn,KhanhHoa,LamDonga ndLongAn.Intotal,thereare1384observationsintheHaNoiandHCMCarea,2 8 3 8 observa tionsf o r t h e o t h e r area.Figure6 s h o w s t h e shareso f enterprisesbyprovinces.
Appendix1).Accordingtothat,chemicals,machinery,transportationequipmentarehigh-tech
Textiles, Apparel; 11,0% Paper products ; 2,8% Wood products; 12,0% industries,w h i l e f o o d s & beverages,tobacco,publishing,textiles,furniture,fabricatedmeta lproductsandothersarebelongtothelow-techgroup.Inthisthesis,the authorisu s i n g t h e sameapproachtoconsiderthedifferenceimpactofinnovation on productivityoft w o groups.Belowc h a r t describest h e shareso f t h e s e industriesw i t h i n thes caleo f t h e survey.
DATA ANALYSIS
E MPIRICAL R ESULTS
Beforepresentingthedataanalysis,itisusefultodescribethesummaryofstatisticvariabl es,whichisprovidedinTable9below.Asmentionedinpreviouspart,dataofthisresearchis stronglybalancedwith statisticdescriptionbelow.
Year Variable Obs Mean Std.
The first research hypothesis is supported by the estimation results of the initial model, as detailed in Table 10, where all regression coefficients are significant at a p-value of less than 5% The estimated coefficient for innovation stands at 2.9%, indicating that if a firm has invested in new products, technologies, or improvements within the last two years, its productivity will be 2.9% higher than that of non-innovative firms, all else being equal Additionally, the coefficients for capital (K) and materials (M) are 4% and 61.2%, respectively, meaning that a 1% increase in capital and materials will lead to a productivity increase of 4% and 61.2% However, the lnL coefficient presents a different interpretation, with α + β + γ - 1 equating to -0.09.
Thesumofthreeinputs’coefficients:capital,laborandmaterialsarea p p r o x i m a t e l y closedto1(0.04+0.61+0.26),whichmeanstheconstantreturnstoscaleisobservedinprod uctionfunctioninthemanufacturingindustries.Allvaluesaresignificantasp- valuesarelessthan5%andR- squaredequals81%,whichisquitehighandgoodfore x p l a i n i n g of themodel.
VariableFixed-effect modelRandom-effect model lnOAperL
Table 10:Regressionresults lnKAperL 0.0404 (0.0075) *** 0.0753 (0.0047) *** lnMAperL 0.6120 (0.0068) *** 0.6412 (0.0046) *** lnL -0.0935 (0.0149) *** 0.0713 (0.0055) *** innovation 0.0290 (0.0136) ** 0.0186 (0.0108) *
(*)denotesignificantlevelof1%,5%,10%,respectivelyS o u r c e : Auth or'scalculation
Asw e cans e e , b o t h resultso f FEandR E supportt h e firstresearchh y p o t h e s i s , whi chstatesthepositiverelationshipbetweeninnovationandproductivity.Thecorrespondingfindi ngshavebeenfound byasignificant n u m b e r ofstudies,forinstanceGriffith,Huergoetal. (2006),MassoandVahter(2008),RoperandLove(2002)… arguedt h a t i n n o v a t i o n playsanimportantr o l e i n p r o d u c t i v i t y i m p r o v e m e n t Interestingly,t h e negativelaborcoefficient(α+β+γ-1 F = 0.0000 lnOAperL Coef Std.Err t P>|t| [95%Conf.Interval] lnKAperL 0403773 0074558 5.42 0.000 0257556 05499 lnMAperL 6119739 006756 90.58 0.000 5987246 625223 lnL -.0935365 0148831 -6.28 0.000 -.1227239 -.064349 innovation 0289898 0136353 2.13 0.034 0022495 055730
7 sigma_u 36445967 sigma_e 26717584 rho 65044999 (fraction ofvariancedue to u_i)
R-sq:w i t h i n = 0.8131 Obs pergroup:min = 1 between=0.8829 avg= 2.0 overall=0.8699 max= 2
Randomeffectsu_i~Gaussian Waldchi2(4) =2 4 9 0 8 7 2 corr(u_i,X) =0(assumed) Prob>chi2 = 0.0000 lnOAperL Coef Std.Err z P>|z| [95%Conf.Interval] lnKAperL 0753398 0047197 15.96 0.000 0660894 084590 lnMAperL 6411617 0046414 138.14 0.000 6320647 650258 lnL 071272 0054832 13.00 0.000 0605251 082018 innovation 0186494 0107798 1.73 0.084 -.0024786 039777
R-sq:w i t h i n = 0.8128 Obspergroup:min= 1 between=0.8832 avg = 2.0 overall=0.8700 max= 2
Randomeffectsu_i~Gaussian Waldchi2(5) =2 4 9 3 6 8 1 corr(u_i,X) =0(assumed) Prob>chi2 = 0.0000 lnOAperL Coef Std.Err z P>|z| [95%Conf.Interval] lnKAperL 075251 004719 15.95 0.000 066002 084500 lnMAperL 6411797 0046407 138.16 0.000 6320841 650275 lnL 0754226 0064351 11.72 0.000 06281 088035 innovation 0025202 0172401 0.15 0.884 -.0312698 036310
6 sigma_u 20876594 sigma_e 26716442 rho 37911621 (fraction ofvariancedue to u_i)
Fixed-effects(within)regression Numberofobs = 4186
R-sq:w i t h i n = 0.8235 Obspergroup:min = 1 between=0.8167 avg = 2.0 overall=0.8194 max = 2
F(5,2071) = 1933.06 corr(u_i,Xb)= 0.2050 Prob>F = 0.0000 lnOAperL Coef Std.Err t P>|t| [95%Conf.Interval] lnKAperL 0402848 007456 5.40 0.000 0256628 054906 lnMAperL 6120385 006756 90.59 0.000 5987893 625287 lnL -.0973427 0152904 -6.37 0.000 -.1273287 -.067356 innovation 0474205 0217827 2.18 0.030 0047023 090138
9 sigma_u 36428572 sigma_e 26716442 rho 65025231 (fraction ofvariancedue to u_i)
R-sq:w i t h i n = 0.8123 Obs pergroup:min = 1 between=0.8841 avg= 2.0 overall=0.8706 max= 2
Randomeffectsu_i~Gaussian Waldchi2(5) =2 5 0 1 7 8 7 corr(u_i,X) =0(assumed) Prob>chi2 = 0.0000 lnOAperL Coef Std.Err z P>|z| [95%Conf.Interval] lnKAperL 0737981 0047343 15.59 0.000 064519 083077 lnMAperL 6406835 0046376 138.15 0.000 631594 64977 lnL 0688563 0055204 12.47 0.000 0580364 079676 innovation -.0016768 0123062 -0.14 0.892 -.0257965 022442 InLocation 058822 0172885 3.40 0.001 0249372 092706 _cons 3.486165 0602105 57.90 0.000 3.368154 3.60417
5 sigma_u 20797586 sigma_e 26682879 rho 37792371 (fraction ofvariancedue to u_i)
Fixed-effects(within)regression Numberofobs =
R-sq:w i t h i n = 0.8240 Obs pergroup:min = 1 between=0.8100 avg= 2.0 overall=0.8143 max= 2
F(5,2071) = 1938.96 corr(u_i,Xb)= 0.1993 Prob>F = 0.0000 lnOAperL Coef Std.Err t P>|t| [95%Conf.Interval] lnKAperL 0400784 0074471 5.38 0.000 0254739 054682 lnMAperL 612038 0067473 90.71 0.000 5988058 625270 lnL -.0942923 0148668 -6.34 0.000 -.1234477 -.065136 innovation 0556297 0172175 3.23 0.001 0218644 089395 InLocation -.0706182 0279287 -2.53 0.012 -.1253894 -.01584 _cons 4.484393 1141296 39.29 0.000 4.260573 4.70821
4 sigma_u 37039794 sigma_e 26682879 rho 65834811 (fraction ofvariancedue to u_i)
R-sq:w i t h i n = 0.8132 Obs pergroup:min = 1 between=0.8829 avg= 2.0 overall=0.8699 max= 2
Randomeffectsu_i~Gaussian Waldchi2(5) =2 4 9 0 3 1 4 corr(u_i,X) =0(assumed) Prob>chi2 = 0.0000 lnOAperL Coef Std.Err z P>|z| [95%Conf.Interval] lnKAperL 0753286 0047202 15.96 0.000 0660772 0845 lnMAperL 6411869 0046422 138.12 0.000 6320884 650285 lnL 0712497 0054841 12.99 0.000 060501 081998 innovation 0268699 0193808 1.39 0.166 -.0111158 064855 In_High -.0100902 0197762 -0.51 0.610 -.0488508 028670 _cons 3.45932 0597695 57.88 0.000 3.342174 3.57646
6 sigma_u 21008212 sigma_e 26718228 rho 38204784 (fraction ofvariancedue to u_i)
Fixed-effects(within)regression Numberofobs = 4186
R-sq:w i t h i n = 0.8235 Obspergroup:min = 1 between=0.8161 avg = 2.0 overall=0.8190 max = 2
F(5,2071) = 1932.74 corr(u_i,Xb)= 0.2048 Prob>F = 0.0000 lnOAperL Coef Std.Err t P>|t| [95%Conf.Interval] lnKAperL 0400982 0074618 5.37 0.000 0254648 054731 lnMAperL 6118667 0067571 90.55 0.000 5986153 625118 lnL -.0939256 0148891 -6.31 0.000 -.1231249 -.064726 innovation 0507251 0266599 1.90 0.057 -.001558 103008 In_High -.0261341 0275453 -0.95 0.343 -.0801534 027885 _cons 4.485119 1143814 39.21 0.000 4.260804 4.70943
3 sigma_u 36486488 sigma_e 26718228 rho 65094416 (fraction ofvariancedue to u_i)
V_B))S E lnKAperL 0403773 0753398 -.0349625 0057718 lnMAperL 6119739 6411617 -.0291878 0049093 lnL -.0935365 071272 -.1648085 0138362 innovation 0289898 0186494 0103404 0083497 b=consistentunderHoandHa;obtainedfromxtregB =inconsi stentunderHa,efficientunderHo;obtainedfromxtreg
InLsize -.0268129 022743 -.0495559 01577 b=consistentunderHoandHa;obtainedfromxtregB =inconsist entunderHa,efficientunderHo;obtainedfromxtreg
V_B))S E lnKAperL 0400784 0737981 -.0337197 0057485 lnMAperL 612038 6406835 -.0286455 0049009 lnL -.0942923 0688563 -.1631486 0138038 innovation 0556297 -.0016768 0573065 0120415 InLocation -.0706182 058822 -.1294402 0219344 b=consistentunderHoandHa;obtainedfromxtregB =inco nsistentunderHa,efficientunderHo;obtainedfromxtreg
In_High -.0261341 -.0100902 -.0160439 0191741 b=consistentunderHoandHa;obtainedfromxtregB =inconsist entunderHa,efficientunderHo;obtainedfromxtreg
AdjR-squared = 0.8698 Total 3716.73486 4185 888108687 RootMSE = 33999 lnOAperL Coef Std.Err t P>|t| [95%Conf.Interval] lnKAperL 0764296 0045721 16.72 0.000 0674659 085393 lnMAperL 6475528 0045694 141.71 0.000 6385943 656511 lnL 0815 0049266 16.54 0.000 0718411 091158 innovation 012375 011145 1.11 0.267 -.0094752 034225
corlnOAperLlnKAperLlnMAperLlnLinnovationInLsizeInLocationIn_High(obsA86) lnOAperLlnKAperLlnMAperL lnLinnova~nInLsizeInLoca~nIn_High lnOAperL 1.0000 lnKAperL 0.4505 1.0000 lnMAperL 0.9216 0.3839 1.0000 lnL 0.2724 0.1661 0.1791 1.0000 innovation 0.1584 0.1173 0.1261 0.3140 1.0000
I S I C Revision2 (Note:in t h e s tu dy , t h e r e a r e o nl y t w o s u b - g r o u p s : low- techand h i g h - t e c h sector.The me di um high-techa n dm e d i u m low- tech a r e combinedto form an aggregatedhigh-techsector)
Source:http://epp.eurostat.ec.europa.eu/cache/ITY_SDDS/Annexes/hrst_st_esms_an9.pdf
# Authors&Title Data Variables&Concepts Model&Methodology Results
"NewTechnology,HumanC apital,TotalFactorProduc tivityandGrowthProcessfo rDevelopingCountries!"
Developingcountriesan ddevelopedcountriesi n t he20 th century.
Pooledtime- seriesdatao f 7 1 non- oilexportingcountries(B arro&Lee)
TFP,growth,technologycha nge,humancapital
“Productivitygrowth,tec hnicalprogress,andeffici encychangeinindustrializ edcountries."
Malmquist index of total factorproductivitygrowth
U.S.productivitygrowthisslightlyhigh erthanaverage,allo f whichi s d u e tote chnicalchange.Japan'sproductivitygr owthisthehighestinthesample,withal mosthalfduet o efficiencychange
"Technologicalinnovationa n d productivityinlate- transitionEstonia:econom etricevidencef r o m innova tionsurveys."
CDMModel– developedb y Crepon(1998).Thismodel explainedtheproductivityoffirmsi ncor relationwithinnovationoutput,knowled gea n d investmentinresearchdevelopm ent.
Firmsizehasaninsignificantimpacton productandapositiveimpactonthe pr ocessinnovation onlyprocessinnovationhasapositi vesignificanteffectonTFP
59 |Page effectofinnovationonproductivity – notonlyontheproductivityinthelastyearoftheinnovationsurvey,butalsooneand twoyearsafterthesurvey
R&D,innovation,firmsize Cohen and Klepper (1996a, b)approach R&Dexpenditure=asize b ln(R&Dexpenditure)= lna + b l n size standardlinearregressionmethods
R&Dexpenditureriseslessthanproport ionalwithsize smallenterprisesthatperformR&Dt e n dt o b e moreinnovativethanlargeente rprises. smallenterprisesspendamuchhigherpr oportionoftheirR&Dbudgetonnewpr oductsthanonnewprocesses
"Innovationandproductivit yindevelopingcountries:A studyofArgentinemanufac turingfirms’behavior(199
Fourgroupso f innovationoutcomes:no n- innovators,productinnovators,processi nnovatorsa n d combinedinnovators
Laborproductivityis, onaverage,1 4 1 % higherininnovat orsthaninnon-innovators
Neitherresearchexpenditurenorinnov ationhasasignificant impacto n innovationsalesandproduc tivity
RoperandLove(2002)in novationandexportperf ormance:evidencefr o mukandgerman manufacturingplants
Innovationispositivelyrelatedtoexpo rtprobabilityinbothcountries.U K: th escaleofplants’innovationis relatedp ositivelytoexportpropensity.Beeffec tiveintheir abilitytoexploitspill- oversfromtheinnovation
Germany:negativerelationshipbetwe enthescaleofinnovationactivityandex portperformance.non- innovatorsaretoabsorbregionalandsu pply-chainspill-overeffects.
Firms’age,processi nnovationand productivitygrowth
Productivitygrowth ProcessinnovationF irmage firmsenterthemarketexperiencinghig hproductivitygrowthandthatabove- averagegrowthratestendtolastforma nyyears,butalsothatproductivitygro wthofsurvivingfirmsconverges.Proc essinnovationsatsome pointthenleadtoextraproductivitygro wth,whichalsotendstopersistsomew hatattenuatedforanumberof years.
Usinga largepanelofpubl iclytradedUSfirms,para metersoftheproductiont echnologyforlargeands mallfirmsareestimatedfo rt h e 1970–1989period
CobbDouglasproductionfunctionF i x e d andRandomeffectmodel smallerfirmsexhibitahigherprofitrate, lowersurvivalprobabilitya n d difficul tyi n accessingthecapitalmarket small firmsaresignificantlymoreproductive butalsomoreriskythan theirlargecou nterparts.Smallfirmsfacingmarketunc ertainties,capitalconstraintsandotherc hallengesundertakeactionsthatmaket hemmoreefficientthanlargefirmsb u t i s achieveda t thecosto f increasingthe irriskiness
Investmentsinmodernizati on,innovationa n d gainsinp roductivity:Evidencefromf irmsintheglobalpaperindus try
Controlvariables:mergersa n d acquisitions,R&Dexpe nditure,capitalintensity
Performance=f(modernizationin vestment,innovationactivity,con trolvariables)
Innovation,exportsandp roductivity panelofSpanishmanufac turingfirms-Spanish ESEE survey
Kolmogorov–Smirnovequality-of- distributionstest,Wecomparethepro ductivitydistributions ofdifferentsamplesoffirms productinnovation– notprocessinnovation– affectsproductivityandinducessmallno n- exportingfirmstoentertheexportmarket.
Ont h e Relationshipbetwee nInnovationandPerforman ce:ASensitiveAnalysis
Mediateeffectoftechno logyinnovationcapabili tiesinvestment capabilityandfirmperform anceinVietNam
Innovation Technological innovationcapabilities Competitiveperformance surveyquestionnairewasdesigned collecteddatafromenterprisesi nt he hig h- techmanufacturingindustry&S M E s in VietNam
Investmenta n d OwnershipStructurei n ProductivityPerformanceoft h e Manu facturingSectori n VietNam–
BSPSCIEMproject” thestudyidentifiedthatVietNamenter prisearerequiredtoinvesttoenhancese vendimensionsofTICs- learning,R&D,resourceallocation,ma nufacturing,market,organizational,a ndstrategicplanningcapabilities.Inve stmentcapabilitycanleaddirectlycom petitiveperformancethroughTICs.