How$Internationalization$and$Culture$Influence$
Corporate$Social$Responsibility$and$Financial$
Performance$
$
Master$Thesis$
MSc$Business$Administration$
International$Management$
$
Amsterdam$Business$School,$University$of$Amsterdam$
$
$
$
Peter$Sponer$
Student$Number:$10712216$
Supervisor:$Dr.$Alan$Muller$
Aug$31,$2015$
Abstract$
This$thesis$explores$the$corporate$social$responsibility$and$financial$performance$relationship$ and$proposes$that$it$is$moderated$by$internationalization$and$culture.$The$relationship$is$tested$ on$a$sample$of$S&P$500$companies.$Even$though$the$study$does$not$find$a$direct$support$for$the$ hypotheses,$it$finds$that$social$responsibility$is$related$to$financial$performance.$$
Statement$of$Originality$
This$document$is$written$by$the$author$who$declares$to$take$full$responsibility$for$its$content.$ The$work$presented$in$this$document$is$original$and$no$sources$other$than$those$mentioned$in$ the$text$and$its$references$have$been$used$in$creating$it.$$ The$Faculty$of$Economics$and$Business$is$responsible$solely$for$the$supervision$of$completion$of$ the$work,$not$for$the$contents.$$Table$of$Contents
$ Introduction$ ... 5$ Theoretical$Framework$ ... 8$ Literature$Review$ ... 8$ Shared$Value$ ... 8$ Measuring$CSR$ ... 11$ Causality$... 12$$ CSR$and$Financial$Performance$... 13$ CSR$and$Stakeholder$Theory$ ... 14$ CSR$and$Internationalization$... 15$ CSR$and$Culture$... 16$ Hypothesis$Development$... 18$ Methods$... 25$ Dataset$ ... 25$ Independent$Variable$... 25$ Dependent$Variable$ ... 27$ Moderating$Variables$... 28$Control$Variables$... 29$ Sample$Size$ ... 30$ Analysis$ ... 31$ Check$for$Normality$ ... 32$ Results$... 34$ KLD$Strengths$as$predictor$and$ROA$as$dependent$variable$ ... 36$ KLD$Concerns$as$predictor$and$ROA$as$dependent$variable$... 38$ KLD$Strengths$as$predictor$and$ROS$as$dependent$variable$... 42$ KLD$Concerns$as$predictor$and$ROS$as$dependent$variable$ ... 43$ KLD$Strengths$as$predictor$and$ROE$as$dependent$variable$... 46$ KLD$Concerns$as$predictor$and$ROE$as$dependent$variable$ ... 47$ Discussion,$Conclusion$ ... 51$ Limitations$... 52$ Implications$and$Future$Research$... 53$ Acknowledgements$... 55$ References$ ... 56$ $
Introduction
$ The$world$has$been$changing.$We$are$facing$some$big$issues.$Global$warming$has$ become$one$of$the$biggest$challenges$of$our$generation.$Climate$change$will$have$negative$ consequences$on$the$civilization,$including$shortage$of$fresh$water$in$parts$of$the$world,$ droughts$and$floods,$food$scarcity,$decreasing$biodiversity,$and$health.$Weather$will$become$ more$extreme$as$the$amount$of$carbon$dioxide$increases$in$the$atmosphere$and$the$sea$levels$ will$rise$(Metz,$2011).$$ There$are$nine$planetary$boundaries:$climate$change,$rate$of$biodiversity$loss,$ interference$with$nitrogen$and$phosphorus$cycles,$stratospheric$ozone$depletion,$ocean$ acidification,$global$fresh$water$use,$change$in$land$use,$and$chemical$pollution.$These$are$ interconnected$and$need$to$be$dealt$with$together.$We$have$evidence$that$critical$thresholds$ have$already$been$exceeded$in$three$categories:$climate$change,$global$nitrogen$cycle,$and$rate$ of$biodiversity$loss$meaning$that$there$will$be$some$irreversible$changes.$We$know$that$ companies$have$largely$contributed$to$these$(Whiteman$et$al.,$2013).$$ The$population$has$been$growing$rapidly$over$the$past$decades$and$is$expected$to$grow$ in$the$future.$In$2011,$the$world$population$reached$7$billion$and$will$top$9$billion$in$2050$(UN,$ 2013).$This$may$pose$significant$threads$including$resource$depletion,$lack$of$food,$and$a$need$ for$new$approaches$to$urbanization.$$ Finally,$even$though$this$number$has$been$decreased$since$1990s,$there$are$still$over$ one$billion$people$in$poverty$and$the$income$inequality$gap$has$been$rising$in$many$countries$ (UN,$2013).$Multinational$companies$(MNEs)$should$identify$the$opportunities$in$developing$ markets$and$create$new$business$models$to$target$the$bottom$of$the$pyramid.$By$doing$so,$companies$would$not$only$make$profits,$but$could$help$eradicate$poverty$(Prahalad,$Hart,$2002;$ Karnani,$2007).$Firms$are$increasingly$expected$to$play$an$active$role$in$addressing$these$issues$ and$promoting$sustainable$development$(Kolk,$van$Tulder,$2010).$$ Moreover,$the$recent$developments$in$new$institutional$infrastructure,$such$as$Global$ Reporting$Initiative,$the$social$investment$movement,$and$others$increase$the$pressure$on$ businesses$to$behave$responsibly$by$emphasizing$CSR,$accountability,$transparency,$and$ sustainability.$This$new$setting$changes$the$rules$of$the$game,$particularly$for$multinational$ companies$that$are$increasingly$expected$to$be$more$transparent$about$their$activities,$adopt$ new$standards$and$practices,$work$proactively$with$various$stakeholders,$and$engage$in$new$ partnerships$(Waddock,$2008).$$ In$response$to$these$issues$both$governments$and$businesses$have$come$up$with$ specific$tools$to$address$them.$On$the$government$level,$there$has$been$the$adoption$of$the$UN$ Millennium$Development$Goals,$introduction$of$the$Kyoto$Protocol,$the$EU$Emission$Trading$ Scheme,$and$others.$These$initiatives$have$had$direct$influence$on$corporations.$$$ Companies$have$begun$to$deal$with$these$issues$by$adopting$codes$of$good$governance,$ signing$the$UN$Global$Compact,$and$promoting$initiatives$created$in$cooperation$with$various$ partners$(OECD,$2008).$Even$in$the$times$of$the$global$economic$slowdown$companies$invest$in$ CSR.$According$to$the$report$of$Ernst$&$Young$(2010),$almost$two$thirds$of$the$companies$with$ revenues$over$a$billion$dollars$expressed$a$commitment$to$increase$their$climatebrelated$ investments.$$ It$is$important$to$define$the$concept$of$CSR.$For$the$purposes$of$this$paper,$I$will$use$the$ definition$of$Aquinis$and$Glavas$(2012)$who$define$CSR$as$“contextbspecific$organizational$ actions$and$policies$that$take$into$account$stakeholders’$expectations$and$the$triple$bottom$line$
of$economic,$social,$and$environmental$performance.”$It$is$also$worth$mentioning$that$ companies$involve$in$CSR$voluntarily$and$that$these$activities$go$beyond$what$is$required$by$law.$$ The$benefits$for$companies$to$pursue$CSR$activities$include$risk$management,$improved$ reputation,$increased$efficiencies,$brand$differentiation,$innovation,$better$employee$ motivation,$etc.$$ I$will$focus$in$this$paper$on$discussing$the$benefit$of$CSR$in$terms$of$better$financial$ performance$of$a$company.$In$the$next$section,$the$most$relevant$literature$on$the$topic$will$be$ discussed$and$a$research$gap$will$be$identified.$$$
Theoretical$Framework$
In$this$section,$literature$review$on$the$topic$of$CSR$and$financial$performance$will$be$presented$ and$hypotheses$developed.$$ $Literature$Review$
The$researchers$have$long$talked$about$corporate$social$responsibility$(CSR).$The$first$ appearance$can$be$traced$to$the$beginning$of$20th$century;$however,$the$topic$started$to$ flourish$in$the$literature$in$1970s$with$the$article$of$Friedman$(1970).$He$notes$that$CSR$ activities$go$far$beyond$the$responsibilities$of$the$business$and$that$these$should$be$performed$ by$the$government$that$collects$taxes.$He$states$that$pursuing$CSR$activities$is$comparable$to$ decreasing$the$value$of$the$business$and$therefore$stealing$from$the$shareholders.$Banerjee$ (2008)$argues$that$companies$involve$themselves$in$CSR$to$only$serve$their$own$interests$and$ satisfy$the$demands$coming$from$external$stakeholders.$These$actions$give$large$multinational$ companies$power$and$legitimize$their$actions.$$ $Shared$Value$
On$the$other$hand,$Porter$and$van$der$Linde$(1995),$Porter$and$Kramer$(2006)$claim$ that$CSR$is$closely$related$to$company’s$strategy$and$being$sustainable$will$benefit$both$the$ company$and$the$society.$Instead$of$looking$at$CSR$the$old$way$as$a$cost$that$is$a$burden$to$ business,$companies$should$capitalize$on$opportunities$and$pursue$sustainability$to$eliminate$waste,$decrease$overall$cost,$and$improve$the$product$quality.$They$argue$that$pollution$is$ highly$inefficient.$Companies$should$harvest$earlybmover$advantages,$improve$resource$ productivity,$and$trigger$innovations$that$will$increase$their$competitiveness.$Companies$can$ choose$to$support$general$causes,$for$example$by$donating$money;$however,$if$they$are$to$turn$ CSR$activities$to$their$competitive$advantage,$they$should$focus$on$the$social$impacts$in$their$ value$chain$and$dimensions$of$competitive$context.$Strategic$CSR$is$achieved$through$creating$ shared$value$when$both$the$business$and$society$benefit.$They$present$a$case$of$Nestle$in$India$ when$the$company$invested$in$an$impoverished$region.$After$the$company$built$infrastructure$ to$support$the$local$farmers,$provided$education$and$technical$assistance$to$them,$the$region$ substantially$grew$and$the$company$ensured$itself$a$supply$of$highbquality$commodities$and$ created$a$competitive$advantage.$$ Shared$value$is$a$concept$that$is$perceived$to$be$a$winbwin$situation$for$both$business$ and$society,$but$there$are$many$problems$implementing$it.$Despite$large$benefits$that$have$ been$proven$by$the$number$of$companies,$creating$shared$value$is$a$strategic$task$that$requires$ an$approval$of$a$CEO$or$executive$board$and$it$is$often$beyond$the$authority$of$most$CSR$ managers.$Furthermore,$creating$shared$value$may$be$perceived$as$undertaking$a$big$risk$with$ potential$results$that$may$not$be$visible$in$the$shortbterm$(Rangan$et$al.,$2012).$Yet$the$concept$ of$shared$value$seems$to$pay$off.$$ Examples$of$other$success$stories$regarding$sustainability,$cluster$development,$and$ competitive$advantage$can$be$found$at$Unilever.$The$new$strategy$of$the$company$lies$on$ sustainability$where$the$company$reduces$its$environmental$impact,$improves$health$and$wellb being$of$people,$and$the$working$conditions$of$its$workforce.$After$implementing$the$strategy$in$
2010,$the$company$stock$increased$by$40%$and$the$employee$satisfaction$has$risen$significantly$ (Economist,$2014).$$ No$matter$if$a$company$is$involved$in$creating$shared$value$or$has$smaller$CSR$programs$ it$should$create$a$CSR$strategy$and$align$it$with$its$business$goals.$$ The$question$that$arises$is$whether$to$pursue$CSR,$because$a$multinational$company$is$ expected$to$do$so,$if$it$increases$financial$performance,$or$whether$it$only$creates$additional$ cost.$There$are$two$underlying$attitudes$towards$CSR.$Some$believe$that$the$company$should$ not$pursue$something$that$is$an$exclusive$role$of$the$government$and$that$involving$in$CSR$ activities$only$increases$costs$and$therefore$puts$the$company$in$disadvantage$compared$to$its$ competitors.$The$funds$should$be$redistributed$to$the$shareholders$or$used$to$improve$ efficiencies$in$the$business.$Others$claim$that$the$business$needs$to$pursue$CSR$to$offset$its$ negative$impact$on$the$environment$(Barnett,$2007).$Managing$stakeholders$effectively$is$ crucial$to$company’s$success$(Freeman,$1984).$Stakeholder$management$can$be$compatible$ with$maximizing$firm’s$financial$performance$(Brammer,$Millington,$2008).$It$is$no$wonder$ therefore,$that$the$link$between$CSR$and$financial$performance$has$been$discussed$in$the$ literature$extensively.$$ Ambec$and$Lanoie$(2008)$propose$that$the$company$can$increase$financial$performance$ resulting$from$sustainability$in$two$ways:$by$reducing$cost$or$increasing$revenues.$Increased$ revenues$would$be$ensured$by$better$access$to$certain$markets,$differentiating$products,$and$ selling$pollutionbcontrol$technologies.$Risk$management$and$relations$with$external$ stakeholders;$cost$of$material,$energy,$and$services;$cost$of$capital;$and$cost$of$labor$would$then$ be$determinants$for$reduced$cost$of$operations.$Green$management$should$be$pursued$ regardless$its$effect$on$financial$performance$since$the$society$expects$firms$to$use$resources$
wisely$and$responsibly,$protect$the$environment,$and$eliminate$toxins$that$could$potentially$ harm$people$and$communities.$Green$management$has$also$positive$impact$on$the$health$of$ workforce$and$recruiting$and$retaining$talent$(Marcus,$Fremeth,$2009).$$
$
Measuring$CSR$
CSR$ratings$have$become$increasingly$important$in$the$past$couple$of$years.$For$some$ companies,$it$is$important$to$score$well$on$these$rankings$so$that$they$are$included$in$social$ investment$funds.$For$others,$CSR$ratings$can$be$a$way$to$comply$with$various$regulations$that$ can$come$from$the$national$governments$or$international$bodies.$$ To$be$able$to$measure$the$output$of$CSR,$the$metrics$have$to$be$established.$The$level$ of$social$responsibility$is$measured$by$corporate$social$performance$(CSP).$Waddock$and$Graves$ (1997)$note,$though,$that$one$of$the$problems$with$measuring$CSP$is$that$many$metrics$that$ have$been$used$in$the$research$are$onebdimensional.$Firms$can$pursue$different$activities$ ranging$from$inputs,$such$as$protecting$the$environment,$labor$issues,$or$product$safety,$to$ outputs,$such$as$relationship$with$the$local$communities$or$various$philanthropic$activities.$A$lot$ of$studies$in$the$past$focused$on$surveys,$reputational$and$social$responsibility$indexes,$content$ analysis$of$documents,$and$social$audits.$The$problem$with$these$measures$is$that$they$were$ often$biased,$the$data$had$been$gathered$for$a$different$purpose,$or$they$did$not$sufficiently$ cover$a$broad$range$of$CSR$activities.$$CSP$can$significantly$differ$across$industries,$since$each$ industry$has$its$own$dynamics.$CSP$should$therefore$be$treated$as$a$multibdimensional$construct$ covering$various$areas$of$activities.$$As$the$research$on$CSR$evolved,$more$studies$started$using$KLD$database$as$a$measure$ of$CSP.$This$database$covers$more$aspects$of$social$responsibility$and$should$therefore$ represent$a$wider$range$of$CSR$activities.$$ $
Causality$$
Another$problem$that$comes$up$when$analyzing$the$CSP$and$financial$performance$ relationship$is$the$causality$of$the$relationship.$Are$companies$performing$better$financially$ because$they$are$more$responsible$or$are$they$pursuing$responsibility$because$they$have$ resources$for$it?$Waddock$and$Graves$(1997)$refer$to$this$as$the$slack$resources$theory.$Those$ companies$that$have$a$lot$of$resources$available$can$afford$to$invest$in$various$social$ responsibility$initiatives,$which$increases$their$CSP.$In$this$case,$financial$performance$is$a$ predictor$of$higher$CSP$and$not$the$other$way$around.$$ Other$groups$of$researchers$argue$that$CSP$goes$in$line$with$stakeholder$management,$ since$being$socially$responsible$can$increase$satisfaction$of$the$employees$and$therefore$ motivation$or$other$aspects,$such$as$product$quality$that$can$become$a$competitive$advantage.$ Pursuing$stakeholder$management$can$then$increase$CSP.$$ Many$studies$have$focused$on$the$CSP$and$financial$performance$relationship,$but$did$ not$answer$the$question$of$causality.$$ $CSR$and$Financial$Performance$
Margolis$et$al.$(2009)$conducted$a$metabanalysis$of$251$studies$to$find$out$whether$CSR$ has$a$positive$impact$on$companies’$financial$performance.$The$studies$reviewed$are$from$1972$ to$2007.$The$authors$tested$corporate$social$performance$(CSP)$and$the$studies$were$sorted$ into$nine$categories$according$to$their$character:$corporate$policies,$disclosure,$environmental$ performance,$philanthropic$donations,$revealed$misdeeds,$selfbreported$social$performance,$ observers’$perceptions,$third$party$audits,$and$screened$mutual$funds.$This$was$done$to$find$out$ whether$certain$types$of$CSP$have$different$effect$on$financial$performance$than$others.$The$ results$show$that$there$is$a$nonbsignificant$relationship.$Only$revealed$misdeeds$show$higher$ significance.$This$implies$that$the$consequences$of$negative$corporate$actions$have$a$bigger$ impact$on$the$income$statement$than$pursuing$different$CSR$activities.$Overall,$the$study$finds$ that$there$is$a$small,$but$positive$link$between$CSP$and$financial$performance.$This$goes$ together$with$conclusions$of$previous$studies$trying$to$test$the$same$relationship$(Orlitzky$et$al.,$ 2003;$Allouche,$Laroche,$2005;$Wu,$2006).$These$studies$also$confirm$that$CSP$shows$bigger$ influence$on$accountingbbased$measures$of$financial$performance$compared$to$the$marketb based$measures.$$$ Barnett,$Salomon$(2012)$address$the$shape$of$CSP$and$financial$performance$ relationship.$The$study$included$a$sample$of$1,214$companies$during$a$period$from$1998$to$ 2006.$They$find$that$firms$with$low$level$of$CSP$perform$financially$better$than$those$with$ moderate$level.$Those$firms$with$the$highest$CSP$show$highest$financial$performance.$The$CSP$ and$financial$performance$is$therefore$Ubshaped.$Even$though$the$socially$responsible$ investment$may$initially$be$costly$for$the$companies$to$pursue,$the$role$of$stakeholder$influence$ capacity$ensures$that$the$companies$that$have$the$strongest$ties$with$their$stakeholders$will$yield$the$highest$returns.$This$is$because$more$socially$responsible$companies$will$be$perceived$ as$more$credible$among$stakeholders.$$$ $
CSR$and$Stakeholder$Theory$
A$big$stream$of$researchers$consider$CSR$to$be$a$response$to$stakeholder$pressures$and$ therefore$a$strategic$tool$that$has$ultimately$an$influence$on$firm’s$performance.$In$today’s$era$ of$global$connectedness,$information$is$easily$available$to$consumers.$Some$companies$pursue$ CSR$as$a$way$of$insurance$for$their$reputation$and$risk$management.$Other$companies$use$CSR$ as$a$reactive$tool$to$mitigate$harm$from$their$negative$behavior.$For$example,$Groza$et$al.$ (2011)$note$that$firm’s$motives$for$CSR$influence$purchase$intentions$or$the$willingness$to$ recommend$a$product$or$service.$Consumers$can$therefore$be$considered$an$important$ stakeholder,$to$which$a$firm$can$positively$respond$by$engaging$in$CSR$programs.$Based$on$the$ stakeholder$theory,$a$company$would$identify,$evaluate,$and$manage$relationships$with$anyone$ who$is$impacted$by$the$actions$of$the$organization.$Stakeholder$management$should$be$ accounted$for$when$investigating$the$CSR$and$financial$performance$relationship$(Barnett,$ 2007).$This$phenomenon$causes$that$the$results$are$inconsistent$across$firms$and$time.$He$notes$ that$companies$with$the$same$intention$to$invest$in$CSR$would$yield$different$financial$ performance$depending$on$how$they$are$perceived$by$stakeholders.$$ Peloza,$Papania$(2008)$note$that$stakeholder$saliency,$or$the$extent$to$which$ stakeholders$can$influence$the$company$based$on$their$power,$legitimacy,$and$urgency,$also$ plays$a$role$in$the$CSR$and$financial$performance$relationship.$Berman$et$al.$(1998)$find$that$ especially$two$groups$of$stakeholders,$employees$and$consumers,$have$a$direct$impact$on$a$firm’s$financial$performance.$When$it$comes$to$managing$the$relationship$with$consumers,$ product$safety$and$quality$are$the$most$important$issues.$Furthermore,$when$looking$at$the$ investment$decisions,$many$institutional$investors$now$allocate$funds$based$on$social$screening$ and$invest$in$those$companies$that$behave$responsibly$(Pava,$Krausz,$1996).$$ CSR$also$plays$a$major$role$in$how$employees$perceive$the$organization$and$can$ significantly$influence$their$attitudes$and$behaviors$towards$the$organization$(Rupp$et$al.,$2006).$ The$research$has$shown$that$the$perception$of$fairness$and$social$justice$are$important$factors$ to$develop$organizational$commitment.$This$goes$in$line$with$one$of$the$drivers$of$CSR,$which$ are$the$personal$values$of$executives.$Some$argue$that$engagement$in$CSR,$no$matter$if$for$ strategic$reasons,$is$to$some$extent$always$driven$by$some$kind$of$selfbinterest$(Hemingway,$ MacLagan,$2004).$$ $
CSR$and$Internationalization$
We$know$that$stakeholder$management$has$an$impact$on$the$CSR$and$financial$ performance$relationship.$However,$attention$in$the$research$has$not$been$given$much$to$ internationalization$as$a$moderating$variable$influencing$the$relationship.$Muller$(2006)$ addressed$the$question$whether$to$pursue$a$global$CSR$policy$coming$from$the$headquarters$or$ customize$it$to$address$the$local$needs.$He$expressed$concerns$that$if$CSR$was$pursued$ independently$by$MNE’s$foreign$subsidiaries,$these$could$adopt$lower$standards$of$CSR.$ Surprisingly,$he$found$that$pursuing$CSR$locally$can$even$increase$CSR$performance.$There$is$no$ clear$answer$on$whether$to$pursue$global$or$local$CSR$strategy.$Some$studies$found$support$for$ standardizing$CSR$activities$across$subsidiaries,$others$point$out$the$importance$of$adapting$to$the$local$environment.$Bondy$and$Starkey$(2014)$concluded$that$selection$of$global$versus$local$ strategy$often$depends$on$the$company$and$specific$context.$Attig$et$al.$(2014)$find$on$a$sample$ of$3,040$U.S.$firms$that$internationalization$is$positively$related$to$company’s$CSR$rating.$We$do$ not$know;$however,$what$role$internationalization$plays$in$the$CSR$and$financial$performance$ relationship.$$ $
CSR$and$Culture$
Expanding$abroad$may$create$various$challenges$for$a$company.$One$of$these$is$ inevitably$a$different$culture.$The$role$of$culture$in$CSR$practices$is$not$very$well$explored$yet$ (Sachs$et$al.,$2005).$Williams$and$Zinkin$(2008)$find$that$consumers$tend$to$punish$irresponsible$ companies$in$different$ways$depending$on$their$cultural$values.$The$variance$was$analyzed$using$ Hofstede$cultural$dimensions$and$explains$that$using$different$CSR$orientation$in$different$ countries$matters.$$ Hofstede$(2011)$introduced$five$cultural$dimensions$that$reflect$society’s$values$and$ vary$across$countries:$powerbdistance,$individualism$vs.$collectivism,$masculinity$vs.$femininity,$ uncertainty$avoidance,$and$longbterm$vs.$shortbterm$orientation.$Powerbdistance$relates$to$the$ extent$by$which$people$accept$hierarchies$and$respect$authorities.$In$countries$with$lowbpower$ distance$organization$structures$are$flatter$and$employees$are$more$equal.$Individualism$vs.$ collectivism$is$about$the$degree$to$which$societies$value$achievement$and$time,$or$tend$to$be$ more$consensusbseeking$and$promote$gender$equality.$Masculinity$vs.$femininity$deals$with$the$ level$of$competitiveness,$distinction$between$male$and$female$roles$in$the$society,$or$how$ people$are$concerned$with$quality$of$life.$Uncertainty$avoidance$has$to$do$more$with$the$extent$to$which$people$are$willing$to$accept$ambiguity.$Some$cultures$need$rules$and$structure$more$ than$others.$The$same$applies$to$the$level$of$risk$people$from$different$cultures$are$willing$to$ accept.$Longbterm$vs.$shortbterm$orientation$suggests$than$some$cultures$value$traditions,$ whereas$other$cultures$cope$better$with$change.$This$has$also$to$do$with$the$role$of$the$family$ vs.$equality$and$seeking$selfbactualization.$Recently,$after$the$research$of$Minkov$(2010)$a$sixth$ dimension$has$been$added$–$indulgence$vs.$restraint.$This$cultural$dimension$has$to$do$with$the$ perception$of$people$having$control$over$their$lives.$Cultures$with$high$indulgence$allow$ themselves$to$pursue$their$desires,$whereas$cultures$with$low$indulgence$tend$to$be$restrained$ by$social$norms.$Hofstede’s$research$has$become$a$basis$for$intercultural$communication$and$ for$understanding$that$businesses$need$to$be$culturally$sensitive$when$doing$business$abroad$ and$interacting$with$different$cultures.$$ Williams$and$Zinkin$(2008)$find$that$their$research$is$consistent$with$Hofstede$measures.$ Consumers$in$high$individualism$countries,$for$example,$tend$to$punish$companies$for$bad$ actions$more$often$than$those$from$countries$that$score$on$this$ranking$lower.$Similarly,$ consumers$in$countries$with$shortbterm$orientation$tend$to$behave$in$the$same$manner$as$ opposed$to$countries$with$longbterm$orientation.$The$study$concludes$that$managers$need$to$ take$into$account$different$cultural$values$when$selecting$and$implementing$CSR$strategies.$$ Husted$(2005)$argues$that$the$current$models$by$which$we$assess$sustainability$are$ inappropriate.$We$should$not$only$look$at$the$economic$values$as$drivers$of$sustainability,$but$ we$should$also$include$national$culture.$Even$though$economic$development$inevitably$plays$an$ important$role,$he$finds$that$countries$with$low$powerbdistance,$high$individualism,$and$less$ masculinity$have$higher$social$and$institutional$capacity$to$pursue$sustainability.$The$impact$of$ these$dimensions,$however,$varies$with$the$level$of$economic$development.$$
Hypothesis$Development$
The$literature$on$the$relationship$of$CSR$and$financial$performance$has$been$quite$ extensive$over$the$past$decades$with$researchers$finding$inconsistent$results.$There$has$been$ found$a$positive$impact$of$CSR$on$financial$performance,$but$I$assume$this$is$only$valid$under$ certain$conditions.$Many$studies$focus$on$why$companies$involve$in$CSR$and$what$the$outcomes$ are;$however,$there$is$a$big$gap$in$understanding$the$mediating$variables.$The$new$studies$ should$try$to$analyze$what$underlying$mechanisms$lead$to$particular$CSR$outcomes$(Aquinis,$ Glavas,$2012).$The$previous$research$implies$that$internationalization$and$culture$might$have$an$ effect$in$the$CSR$and$financial$performance$relationship.$In$this$paper,$I$will$try$to$fill$in$the$gap$ in$the$literature$and$focus$on$these$variables.$$ $ $ $ $ As$companies$internationalize,$they$face$different$forms$of$distance.$The$distance$can$ be$economic,$financial,$political,$administrative,$cultural,$demographic,$knowledge,$global$ connectedness,$and$geographic$(Berry$et$al.,$2010).$These$forms$of$distance$initially$put$the$ CSR! Financial!! Performance! Internationalization! Culture!companies$in$disadvantage$over$their$competitors$and$their$goal$is$to$overcome$these$ differences$in$the$shortest$possible$time.$$The$concept$of$distance$plays$an$important$role$in$ many$decisions$MNEs$are$facing,$such$as$the$internationalization$process$or$the$selection$of$an$ entry$mode.$Managing$activities$in$different$countries$involves$multiple$institutional$ environments$and$stakeholder$pressures.$CSR$can$be$a$way$for$the$MNEs$to$overcome$liability$ of$foreignness$(LOF)$and$manage$these$differences$in$a$strategic$way$(Campbell$et$al.,$2012).$ Liability$of$foreignness$arises$when$a$company$enters$a$new$market,$in$which$it$is$not$ established$yet$and$needs$to$compete$with$the$local$firms.$Since$the$company$needs$to$get$ familiar$with$the$new$environment,$which$often$features$different$cultural,$political,$and$ economical$systems,$it$will$develop$a$firmbspecific$advantage$that$will$help$it$overcome$these$ differences$(Zaheer,$1995).$$ Being$socially$responsible$while$investing$abroad$increases$the$firm’s$legitimacy.$ Legitimacy$is$perceived$as$critical$for$an$MNE$to$establish$itself$and$become$successful$in$a$host$ country.$This$is$especially$important$for$those$companies$expanding$to$emerging$economies$ (Reimann$et$al.,$2012).$$ As$mentioned$earlier,$a$study$conducted$by$Attig$et$al.$(2014)$found$that$ internationalization$is$related$to$companies’$CSR$ratings.$Those$companies$that$operate$in$more$ countries$tend$to$score$better$on$CSR$strengths.$Moreover,$the$companies$that$have$foreign$ subsidiaries$in$countries$with$wellbfunctioning$legal$and$political$systems$score$better$on$overall$ CSR$ratings.$This$may$be$because$these$companies$use$strategic$CSR$as$a$tool$to$gain$legitimacy$ while$going$abroad.$$ Internationalization$can$be$seen$from$different$points$of$view.$It$can$be$expressed$by$ the$number$of$countries$a$company$operates$in,$the$distribution$of$domestic$vs.$foreign$sales,$or$
whether$the$company$is$present$in$the$triad,$ie.$it$must$have$operations$in$three$geographical$ regions$–$the$EU,$NAFTA,$and$Asia$(Rugman,$Verbeke,$2004).$In$this$study,$the$degree$of$ internationalization$will$be$expressed$by$the$number$of$countries$the$company$operates$in,$ despite$its$geographical$dispersion,$based$on$the$logic$that$every$foreign$market$features$ institutional$differences$and$creates$LOF.$$ I$hypothesize$that$the$degree$of$internationalization$expressed$by$the$number$of$ countries$an$MNE$operates$in$will$have$a$moderating$effect$on$the$CSP$and$CFP$relationship$ since$more$host$countries$will$bring$a$higher$complexity,$multiple$institutional$environments,$ and$stakeholder$pressures$the$company$has$to$deal$with.$The$higher$number$of$markets$will$ also$increase$the$distance$the$company$will$have$to$overcome.$For$those$companies$that$are$ less$socially$responsible,$the$higher$number$of$foreign$markets$will$have$a$negative$moderating$ effect$since$these$companies$will$more$likely$have$troubles$to$gain$legitimacy$among$the$local$ stakeholders.$$ H1:$The$degree$of$internationalization$will$have$a$moderating$effect$in$the$CSP$and$CFP$ relationship.$$ $ MNEs$are$believed$to$have$a$significant$potential$to$contribute$alleviate$poverty$and$ address$the$needs$of$various$communities$through$their$CSR$programs.$Researchers$suggest,$ however,$that$institutions$are$crucial$for$these$initiatives$to$have$impact$and$that$in$developing$ countries$with$nonbfunctioning$governments$the$positive$effect$of$the$MNEs$on$the$economy$is$ hindered$(Ite,$2004).$$
Whereas$in$developed$markets$companies$often$pursue$CSR$as$a$way$of$reputation$and$ risk$management,$this$does$not$necessarily$have$to$be$true$in$developing$countries.$Consumers$ in$developing$countries$are$often$unaware$and$unsupportive$towards$these$programs.$ Consumers$in$developing$countries$with$low$purchasing$power$often$buy$cheaper$and$more$ affordable$products$over$the$green$products$that$can$be$more$costly$to$purchase,$even$though$ they$may$bring$savings$in$the$longbterm.$The$social$responsibility$aspect$is$not$that$important$ yet$for$these$consumers$when$making$purchasing$decisions$(Arli,$Lasmono,$2010).$$ The$same$holds$for$the$institutions$in$the$developed$vs.$developing$countries.$ Companies$in$the$developed$countries$have$higher$social$and$institutional$capacity$to$pursue$ sustainability$(Husted,$2005).$The$level$of$development$plays$a$crucial$role$since$countries$that$ are$less$developed$have$a$lower$chance$to$pursue$research$and$development$and$invest$in$ programs$to$preserve$the$environment$and$biodiversity.$On$the$other$hand,$consumers$in$the$ developed$countries$are$generally$well$informed$and$are$able$to$put$demands$on$multinational$ companies.$$ I$hypothesize$that$the$level$of$development$will$play$a$role$in$the$CSP$and$CFP$ relationship$since$consumers$in$less$developed$countries$do$not$give$CSR$as$much$importance$ due$to$their$lower$purchasing$power$and$awareness.$Companies$will$also$be$hindered$to$pursue$ sustainability$because$of$the$lack$of$institutions.$Less$responsible$companies$will$therefore$have$ lower$financial$performance$in$more$developed$countries.$$ H2:$The$level$of$development$of$the$countries$where$an$MNE$operates$will$have$a$moderating$ effect$in$the$CSP$and$CFP$relationship.$$ $
Hofstede$and$his$cultural$framework$influenced$many$areas$of$business$and$it$has$ become$the$foundation$stone$for$intercultural$management.$Hofstede$argues$that$the$ differences$in$national$culture$set$boundaries$for$management$theories.$Culture$and$cultural$ differences$have$to$be$accounted$for$when$doing$business$abroad,$negotiating$with$people$from$ different$countries,$or$managing$employees$in$a$multinational$company.$Understanding$one’s$ cultural$background$is$crucial$to$be$able$to$predict$their$present$and$future$behavior.$Building$an$ organizational$culture$and$the$development$of$common$organizational$practices$can$help$ multinationals$overcome$the$differences$in$cultural$values$of$their$employees$across$countries$ or$functional$areas$(Hofstede,$2011).$$ Cultural$dimensions$have$also$been$of$interest$of$consumer$behavior$research.$ Strategies$of$multinational$companies$should$not$be$only$duplicated$in$the$developing$markets$ to$reach$the$enlarging$middle$class.$As$income$distribution$becomes$equal$across$countries,$ cultural$differences$become$stronger.$Multinational$companies$need$to$adapt$to$these$ differences$when$implementing$new$retailing$strategies$(Mooij,$Hofstede,$2002).$$ As$previously$noted,$research$also$suggests$that$culture$is$an$important$determinant$of$ environmental$sustainability.$Ringov$and$Zolo$(2007)$applied$the$Hofstede$framework$on$a$ sample$of$463$companies$and$conclude$that$cultural$dimensions$have$influence$on$corporate$ social$and$environmental$performance.$Personal$experience$and$the$inability$to$associate$ directly$oneself$with$the$problem$of$global$warming$together$with$societal$values$were$also$the$ main$constraints$to$why$the$UK$public$did$not$get$involved$enough$in$tackling$the$problem$of$ climate$change,$despite$the$research$pointing$at$its$effects$(Lorenzoni$et$al.,$2007).$$ CSR$activities$differ$across$regions.$In$the$United$States,$CSR$has$been$adopted$much$ earlier$than$in$the$rest$of$the$world.$Matten$and$Moon$(2008)$explain$that$this$is$due$to$differing$
historical$development$of$institutions,$business,$political,$financial,$education$and$labor$systems,$ culture,$and$other$factors.$This$is$also$the$case$of$differing$corporate$governance$systems,$where$ one$is$more$focused$on$agency$relationship,$whereas$the$other$lies$on$the$foundations$of$the$ stakeholder$theory.$One$of$the$reasons$to$explain$this$could$be$that$Americans$believe$in$much$ lower$role$of$the$government$in$their$lives$and$business,$therefore$companies$would$likely$ engage$in$CSR$activities$earlier$and$on$larger$scale$to$bring$the$benefits$that$would$have$already$ been$enacted$in$Europe,$which$has$traditionally$had$more$social$orientation.$On$the$other$hand,$ CSR$activities$are$claimed$to$be$getting$standardized$around$the$world$as$multinationals$face$ more$pressure$from$the$stakeholders.$This$could$also$be$due$to$the$fact$that$CSR$has$become$a$ part$of$management$practices$that$are$being$adopted$worldwide.$$ I$argue$that$culture$will$have$a$moderating$effect$in$the$CSP$and$CFP$relationship$since$ attitudes$towards$CSR$may$vary$across$cultures$on$the$corporate$level$and$CSR$initiatives$may$ be$perceived$differently$among$consumers$as$well.$This$might$be$because$of$different$values$ that$translate$into$different$institutions$across$countries,$role$of$the$government,$and$ stakeholder$needs.$If$the$company$operates$in$countries$that$are$more$culturally$distant,$cost$of$ not$being$responsible$will$probably$have$an$impact$on$the$financial$performance.$$ H3:$The$CSP$and$CFP$relationship$will$be$moderated$by$cultural$distance.$$$$ $ As$mentioned$previously,$countries$with$low$powerbdistance,$high$individuality,$and$low$ masculinity$have$higher$social$and$institutional$capacity$to$pursue$sustainability$(Husted,$2005).$ Park$et$al.$(2007)$find$similar$results$with$their$study$testing$the$influence$of$cultural$dimensions$ on$the$environmental$sustainability$index$(ESI).$They$concluded$that$powerbdistance$and$ masculinity$are$both$negatively$related$to$ESI.$Interestingly,$their$study$also$shows$that$
countries$with$high$individuality$have$significantly$higher$per$capita$income$than$countries$with$ low$individuality.$$ As$mentioned$earlier,$Williams$and$Zinkin$(2008)$confirm$that$cultural$differences$have,$ indeed,$influence$on$consumer$behavior.$They$found$out$that$consumers$from$low$power$ distance$and$high$individuality$countries$are$more$likely$to$punish$irresponsible$companies$for$ their$behavior.$Following$these$assumptions,$I$hypothesize$that$the$CSP$and$CFP$relationship$will$ be$moderated$by$the$fact$whether$the$company$operates$in$low$powerbdistance,$high$ individuality,$low$masculinity$countries.$These$countries$should$have$higher$social$and$ institutional$capacity$to$pursue$sustainability$and$the$consumers$in$these$countries$should$also$ be$more$concerned$about$CSR.$$ H4:$The$CSP$and$CFP$relationship$is$moderated$by$the$company’s$presence$in$low$powerH distance,$high$individuality,$and$low$masculinity$countries.$$ $
Methods$$
In$this$section$the$data$will$be$described.$Analysis$will$follow.$$$
$
Dataset$
The$sample$of$the$study$were$S&P$500$companies$with$the$panel$data$from$2008b2012.$ The$sample$was$chosen$for$a$number$of$reasons.$First$of$all,$to$avoid$any$possible$biases$it$was$ important$to$set$up$a$home$country$against$which$the$data$was$compared.$Secondly,$S&P$500$is$ a$good$sample$of$the$companies$that$drive$the$U.S.$economy$and$firms$from$multiple$sectors$ are$represented.$Third,$most$of$these$companies$do$business$internationally$and$their$data$is$ publicly$available,$therefore$there$is$enough$information$on$them.$Finally,$the$previous$research$ often$used$Fortune$500$companies$as$a$sample$and$there$is$a$need$for$an$alternative$source$of$ data$(Margolis$et$al.,$2009).$$ $Independent$Variable$
The$predictor$variable$was$CSP.$It$was$operationalized$using$KLD$database.$This$ database$includes$strengths$and$concerns$of$each$company$in$various$categories$including$ community,$corporate$governance,$diversity$of$the$board$of$directors,$employee$relations,$ environment,$human$rights,$product$safety,$involvement$in$producing$and/or$selling$alcohol,$ tobacco,$fire$arms,$military,$gambling,$and$nuclear$power.$The$database$is$the$most$commonly$used$source$of$CSP$data$in$research.$Since$every$category$includes$different$number$of$ parameters$and$each$parameter$may$have$different$weight,$the$CSP$index$was$divided$to$total$ strengths$and$total$concerns,$or$in$other$words,$responsible$and$irresponsible$companies.$This$is$ a$more$accurate$way$to$use$KLD$data$than$subtracting$concerns$from$strengths$as$has$been$ commonly$used$in$previous$studies$(Muller,$Kraussl,$2010).$Some$companies$can$be$doing$good$ and$bad$at$the$same$time$and$since$people$usually$give$higher$importance$to$negative$ information$due$to$negative$bias,$doing$bad$can$ruin$a$company’s$reputation.$Moreover,$a$ company$that$would$score$10$on$strengths$and$8$on$concerns$would$have$a$net$score$2.$The$ same$score$could$be$obtained$by$the$company$scoring$2$on$strengths$and$zero$on$concerns.$This$ method$could$be$very$biased$since$it$would$give$the$same$score$to$two$companies$that$can$be$ very$different$concerning$their$CSR$activities.$CSP$was$therefore$divided$into$strengths$and$ concerns$and$both$were$compared$separately$to$financial$performance.$The$KLD$data$was$ retrieved$from$2008b2011.$Since$I$wanted$to$show$how$CSP$predicts$CFP,$the$measure$of$social$ performance$was$lagged$one$year$behind$and$compared$to$the$financial$performance$in$the$ following$year.$The$underlying$assumption$here$was$that$companies’$behavior$in$terms$of$social$ responsibility$would$be$reflected$in$the$financial$performance$in$the$following$year$since$annual$ reports,$social$indexes,$KLD$database$scores,$and$other$sources$of$data$dealing$with$companies’$ CSR$activities$become$available$at$the$end$of$the$year.$The$final$number$of$both$KLD$strengths$ and$concerns$was$averaged$over$four$years.$$ There$were$the$minimum$of$0$(National$Oilwell$Varco$Inc,$Procter$&$Gamble$Co,$ Regeneron$Pharmaceuticals,$Scripps$Networks$Interactive,$Stericycle$Inc,$Teradata$Corp,$Urban$ Outfitters$Inc,$Vertex$Pharmaceuticals$Inc)$and$maximum$of$20$strengths$(IBM)$in$the$sample$(N$ =$250,$M$=$5.6,$SD$=$4.16).$$
The$minimum$number$of$concerns$was$0$(Pitney$Bowes$Inc)$and$the$maximum$was$16$ (WalbMart$Stores$Inc)$in$the$sample$(N$=$250,$M$=$4.19,$SD$=$2.9).$$ $
Dependent$Variable$
The$dependent$variable$in$this$research$was$CFP.$Financial$performance$was$ operationalized$by$return$on$assets$(ROA),$return$on$sales$(ROS),$and$return$on$equity$(ROE).$ This$is$in$line$with$previous$research$(Tsoutsora,$2004).$The$data$was$obtained$from$ COMPUSTAT$for$2009b2012.$The$financial$performance$figures$were$averaged$over$four$years.$$$$ Financial$performance$in$ROA$ranged$from$b0.14$to$0.47$(N$=$250,$M$=$0.13,$SD$=$0.07).$ The$company$with$the$lowest$ROA$was$Vertex$Pharmanceuticals$Inc$and$the$company$with$the$ highest$ROA$was$Coach$Inc.$ROS$was$between$b2.6$and$0.57$(N$=$250,$M$=$0.16,$SD$=$0.2).$The$ company$that$had$the$lowest$EBIT$compared$to$the$sales$was,$once$again,$Vertex$ Pharmaceuticals$Inc$and$the$company$with$the$highest$EBIT$compared$to$the$sales$was$Visa$Inc.$ Finally,$the$lowest$ROE$was$b12.84$and$the$highest$was$28.69$(N$=$250,$M$=$0.48,$SD$=$2.45).$The$ company$that$performed$the$worst$in$this$measure$was$Verisign$Inc$and$Lockheed$Martin$Corp$ had$the$best$results.$The$reason$that$some$companies$had$negative$financial$ratios$was$likely$ due$to$the$fact$that$the$sample$was$from$the$period$between$2009b2012$when$many$companies$ had$to$cope$with$the$negative$effects$of$the$global$economic$slowdown.$$ $Moderating$Variables$
The$moderating$variables$were$the$degree$of$internationalization$measured$by$the$ number$of$markets,$HDI,$cultural$distance,$and$the$fact$whether$the$company$operates$in$low$ PD,$high$IND,$low$MAS$country.$The$number$of$markets$was$obtained$by$looking$at$foreign$level$ one$subsidiaries$of$each$company.$Each$subsidiary$was$considered$for$the$sample$if$the$parent$ company$owned$at$least$50.01%$stake$in$this$subsidiary.$This$data$was$obtained$as$of$2015$from$ the$Orbis$database.$A$drawback$here$is$that$there$might$be$have$been$minor$differences$in$the$ number$of$subsidiaries$between$2008b2015.$The$minimum$number$of$markets$each$company$ operated$in$was$4$(Best$Buy$Co$Inc,$CocabCola$Enterprises$Inc,$Dr$Pepper$Snapple$Group$Inc,$ Hasbro$Inc,$Nabors$Industries$Ltd,$Regeneron$Pharmaceuticals,$Scripps$Networks$Interactive,$ Sempra$Energy,$Textron$Inc,$Time$Warner$Inc,$Under$Armour$Inc,$Visa$Inc,$Apache$Corp)$and$the$ maximum$was$91$(Ecolab$Inc)$in$the$sample$(N$=$250,$M$=$25,$SD$=$19.3).$$ The$HDI$was$computed$as$an$average$number$of$all$markets$the$company$operates$in,$ controlled$for$the$variance.$The$source$of$data$was$HDI$for$2010$retrieved$from$the$Human$ Development$Reports$of$the$UNDP.$To$control$for$variance,$the$HDI$for$each$country$was$ squared$and$divided$by$variance.$The$result$was$an$index,$where$the$minimum$value$was$3.96$ (The$Democratic$Republic$of$Congo)$and$maximum$value$was$34.41$(Norway).$The$minimum$ average$value$was$18.61$(Pioneer$Natural$Resources$Co)$and$the$maximum$average$value$was$ 31.91$(ArcherbDaniels$Midland$Company)$in$the$sample$(N$=$250,$M$=$27.26,$SD$=$2.35).$$$ Cultural$distance$was$computed$using$Kogut$and$Singh$formula$(Kogut,$Singh,$1988).$ The$U.S.$was$selected$as$a$home$country$against$which$other$markets$were$compared.$The$ formula$used$was$the$difference$on$each$Hofstede$dimension$between$home$and$host$country$result$was$an$index$with$the$lowest$number$0.03$(Australia)$and$highest$number$4.83$(Russia).$ The$drawback$of$computing$the$cultural$distance$was$that$there$were$only$complete$cultural$ values$available$for$72$countries.$$The$cultural$distance$was$computed$for$the$majority$of$foreign$ markets$of$each$company$though,$therefore$it$should$be$a$credible$measure$in$the$study.$$The$ company$that$experienced$the$lowest$cultural$distance$between$its$home$country$and$host$ countries$was$Nabors$Industries$Ltd$with$the$value$of$0.12;$Davita$Healthcare$Partners$was$then$ the$company$with$the$highest$value$of$2.75$(N$=$250,$M$=$1.84,$SD$=$0.45).$$ There$were$the$total$of$10$countries,$whose$cultural$dimensions$were$low$PD,$high$IND,$ and$low$MAS.$These$countries$were$Denmark,$Estonia,$Finland,$Iceland,$Latvia,$Lithuania,$ Luxembourg,$Netherlands,$Norway,$and$Sweden.$A$company$that$had$foreign$subsidiaries$in$all$ of$these$countries$was$given$a$score$10,$whereas$a$company$with$no$presence$in$these$markets$ was$given$a$score$zero.$$The$minimum$value$represented$in$the$sample$was$0$and$the$maximum$ value$was$9$(N$=$250,$M$=$2.66,$SD$=$2.18).$Most$companies$were$represented$only$in$one$low$ PD,$high$IND,$and$low$MAS$country$(31.2%)$and$the$least$number$of$companies$(1.2%)$were$ represented$in$9$of$these$countries.$$$ $
Control$Variables$
The$study$controlled$for$industry,$size,$risk,$R&D$expenditure,$and$advertising$ expenditure.$Margolis$et$al.$(2009)$recommended$that$any$future$study$on$CSP$and$financial$ performance$should$control$at$least$for$these$factors.$It$is$important$to$control$for$industry$since$ every$industry$is$very$different$and$companies$in$some$industries$tend$to$invest$in$CSR$activities$ more$than$the$others.$In$this$study,$companies$were$sorted$into$industries$based$on$their$fourbdigit$GICS$code.$The$second$variable$that$was$found$to$have$an$impact$on$the$CSP$and$CFP$ relationship$is$firm’s$size.$Companies$that$are$bigger$tend$to$have$more$slack$resources$that$can$ be$used$to$invest$in$CSR$activities,$as$opposed$to$the$smaller$companies$that$often$have$tighter$ budgets.$In$this$study,$firm$size$was$operationalized$by$the$company’s$total$assets.$The$research$ suggests$that$higher$CSP$results$in$lower$financial$risk.$Risk$was$defined$as$total$debt/total$ assets$(Tsoutsoura,$2004;$Lopez$et$al.,$2007).$Finally,$R&D$expenditure$and$advertising$ expenditure$were$additional$variables$that$were$controlled$for.$Both$of$these$variables$are$ expected$to$have$an$impact$on$longbterm$profitability$as$companies$that$spend$more$money$on$ innovation$and$advertising$usually$have$higher$levels$of$sales.$Since$not$all$companies$reported$ on$either$R&D$expenditure$or$advertising$expenditure$or$both,$dummy$variables$were$created.$ Companies$that$disclosed$the$measure$were$given$value$1$and$companies$that$did$not$disclose$ the$measure$were$given$value$0.$$ $
Sample$Size$
The$original$sample$included$502$S&P$companies;$however,$there$were$not$data$ available$for$all$companies.$Some$companies$also$did$not$operate$internationally.$Other$ companies$had$structures$with$foreign$subsidiaries$only$in$tax$havens,$such$as$Bahamas$or$ Luxembourg.$The$record$was$in$such$cases$deleted$from$the$sample.$Financial$companies$were$ also$removed$from$the$sample$as$these$are$highly$regulated$entities$(Attig$et$al.,$2014).$The$final$ sample$was$250$companies$from$20$different$industries.$Capital$goods$industry$was$represented$ with$the$highest$number$of$companies$(35)$and$Transportation,$Automobiles$&$Components,$ Food$&$Staples$Retailing,$and$Household$&$Personal$Products$were$industries$with$the$least$ number$of$companies$(3).$$$
Representation$of$Industries$in$the$Sample$
Industry$ 4bdigit$GICS$Code$ Number$of$Companies$
Energy$ 1010$ 19$ Materials$ 1510$ 21$ Capital$Goods$ 2010$ 35$ Commercial$&$Professional$Services$ 2020$ 5$ Transportation$$ 2030$ 3$ Automobiles$&$Components$ 2510$ 3$ Consumer$Durables$&$Apparel$ 2520$ 14$ Consumer$Services$ 2530$ 6$ Media$ 2540$ 9$ Retailing$ 2550$ 10$ Food$&$Staples$Retailing$ 3010$ 3$ Food,$Beverage,$&$Tobacco$ 3020$ 18$ Household$&$Personal$Products$ 3030$ 3$ Healthcare$Equipment$&$Services$ 3510$ 23$ Pharmaceuticals,$Biotechnology$&$Life$Sciences$ 3520$ 20$ Software$&$Services$ 4510$ 23$ Technology$Hardware$&$Equipment$ 4520$ 14$ Semiconductors$&$Semiconductor$Equipment$ 4530$ 12$ Telecommunication$Services$ 5010$ 4$ Utilities$ 5510$ 5$ $ $
Analysis$
The$initial$idea$was$to$compare$the$change$in$KLD$strengths$and$concerns$lagged$one$ year$behind$to$change$in$financial$performance.$However,$due$to$change$in$methodology$of$KLD$ in$2008,$the$data$was$inconsistent$and$therefore$this$method$could$not$be$used.$The$final$ methodology$used$for$the$analysis$was$the$comparison$of$KLD$strengths$and$concerns$lagged$ one$year$behind$to$ROA,$ROS,$ROE.$$The$data$was$analyzed$using$hierarchical$regression$analysis$to$see$whether$there$is$a$ causal$relationship$between$the$predictor$(degree$of$internationalization,$HDI,$cultural$distance,$ low$PD,$high$IND,$low$MAS)$and$the$dependent$variable$(ROA,$ROS,$ROE).$To$see$the$effect$of$ each$variable,$different$models$were$constructed$to$see$how$much$each$contributes$to$the$total$ variance$explained$by$the$regression$analysis.$The$overview$of$each$model$and$variables$is$ explained$below.$$ $ The$formula$used$was:$$ Sum$of$KLD$strengths$and$concerns/4tb1$+$degree$of$internationalization$+$HDI$+$cultural$distance$ +$low$PD,$high$IND,$low$MAS$+$industry$+$firm$size$+$risk$+$R&D$expenditure$+$advertising$ expenditure$=$Sum$of$ROA/4$and$Sum$of$ROS/4$and$Sum$of$ROE/4$$ $
Check$for$Normality$
The$frequencies$of$the$variables$were$checked$to$make$sure$there$were$no$errors.$Some$ variables$had$high$skewness$and$kurtosis,$which$would$have$led$to$bias$in$the$analysis.$ Therefore,$these$variables$needed$to$be$transformed.$Total$Assets$variable$was$adjusted$using$ log10(x).$This$variable$had$extremely$high$skewness$and$kurtosis.$This$was$due$to$the$fact$that$ there$was$large$variation$in$assets$between$the$companies.$The$transformation$resulted$in$ normal$distribution$of$the$variable.$Skewness$and$kurtosis$fell$within$the$range$of$b2$and$2,$ which$was$deemed$acceptable$for$the$analysis.$There$was$a$big$variation$in$ROS$and$ROE$as$well.$ These$variables$were$transformed$using$log10(x)+1,$since$they$also$had$negative$values.$The$transformation$also$resulted$in$normal$distribution$and$skewness$and$kurtosis$were$within$the$ acceptable$range.$$
Results$
The$correlation$matrix$was$constructed$to$see$how$the$variables$are$related$to$each$ other.$It$was$also$used$to$detect$multicollinearity,$which$occurs$if$two$variables$are$highly$ correlated$and$it$can$make$the$results$biased.$If$there$was$a$value$of$0.7$or$higher$between$two$ variables,$these$would$be$highly$correlated.$$ All$values$were$below$the$value$of$0.7,$except$in$one$case.$The$degree$of$ internationalization$variable$was$highly$positively$correlated$with$the$low$PD,$high$IND,$low$MAS$ variable$(0.84,$p<0.01).$Because$of$the$problem$of$multicollinearity,$the$low$PD,$high$IND,$low$ MAS$value$was$removed$from$the$model$and$tested$separately.$$ As$seen$in$the$correlation$matrix,$the$degree$of$internationalization$variable$is$also$ positively$correlated$with$cultural$distance$(0.59,$p<0.01)$and$negatively$correlated$with$HDI$(b 0.6,$p<0.01).$This$is$expected$since$presence$in$more$countries$means$that$there$is$higher$ complexity$an$MNE$needs$to$deal$with$as$well$as$more$different$cultures.$On$the$other$hand,$ negative$association$of$the$degree$of$internationalization$and$HDI$could$mean$that$the$ companies$represented$in$the$sample$entered$more$developing$markets$as$their$experience$ with$internationalization$increased.$This$would$be$in$line$with$the$Johannes$and$Vahlne$theory$ saying$that$the$company$first$goes$to$markets$that$are$similar$to$the$home$country$and$ progresses$to$more$distant$markets$afterwards$as$it$gains$more$experience.$$Variable 1.* 2.* 3.* 4.* 5.* 6.* 7.* 8.* 9.* 10.* 11.* 12.* 13.* 14.* 1.* KLD*Strengths 1 2.* KLD*Concerns .419* 1 3.* Industry .068 E.330* 1 4.* Size .395* .572* E.021 1 5.* Risk .058 .168* E.062 .142** 1
6.* R&D*expenditure .192* E.040 .137** .043 E.091 1
7.* AdverNsing*expenditure .143** E.103 .262* .001 E.018 .058 1
8.* Degree*of*InternaNonalizaNon .325* .086 .030 .110 .080 .291* .071 1
9.* HDI E.192* E.155** .099 E.085 E.121 E.079 .020 E.604* 1
10.* Cultural*Distance .202* .019 .100 E.007 .041 .254* .078 .587* E.561* 1
11.* Low*PD,*high*IND,*low*MAS .260* .031 .062 .110 .059 .282* .107 .842* E.355* .465* 1
12.* ROA E.014 E.148** .020 E.145** E.094 E.017 .260* .046 E.032 .152** .028 1
13.* ROS .037 E.015 .089 .026 .000 E.017 .129** .051 E.062 .100 .081 .487* 1
14.* ROE E.006 .023 E.046 E.015 .093 .042 E.039 .013 .061 E.013 .054 .104 .029 1
*p<0.01 **p<0.05 CorrelaNon*Matrix
To$see$how$much$each$individual$variable$contributes$to$the$overall$model,$the$ regression$analysis$was$run.$Since$there$are$two$different$predictors$(KLD$strengths$and$KLD$ concerns),$three$different$measures$of$financial$performance$(ROA,$ROS,$ROE)$and$four$ moderating$variables$(degree$of$internationalization,$HDI,$cultural$distance,$and$presence$in$low$ PD,$high$IND,$low$MAS$countries),$26$different$models$were$run.$$ $
KLD$Strengths$as$predictor$and$ROA$as$dependent$variable$
In$the$first$analysis,$the$effect$of$KLD$strengths$on$ROA$was$investigated.$To$proceed$ with$the$regression$analysis,$the$check$for$independence$of$observations$was$made.$Durbin$ Watson$test$was$performed.$The$value$of$2.01$ensured$that$the$residuals$were$independent.$ Secondly,$case$wise$diagnostics$revealed$that$there$were$seven$outliers$with$values$of$ standardized$residuals$over$3$and$below$b3.$All$these$cases$were$deleted$to$make$sure$there$ were$no$errors$in$the$results.$Check$for$leverage$points$was$also$performed,$where$the$cases$ with$values$over$0.2$would$be$deleted.$No$leverage$points$were$detected.$Finally,$check$for$ influential$points$was$made$using$Cook’s$Distance$values.$Values$over$1$would$be$deleted.$There$ were$no$influential$points$detected.$$ Plots$were$constructed$to$check$whether$independent$variables$were$relatively$linearly$ related$to$the$dependent$variable.$Check$for$homoscedasticity$was$performed$as$well.$$$ $ $ $ $$ $ $ $ $ The$standardized$residuals$were$distributed$relatively$normally$as$seen$on$the$histogram.$The$Pb P$plot$also$confirms$the$normality$of$distribution.$
$
$ In$the$model$1,$the$effect$of$control$variables$(industry,$size,$risk,$R&D$expenditure,$ advertising$expenditure)$was$investigated,$model$2$added$the$degree$of$internationalization$ variable,$model$3$added$the$HDI$variable,$in$the$model$4$the$cultural$distance$variable$was$ added,$and$model$5$tested$the$interaction$between$KLD$Strengths$and$degree$of$ internationalization,$KLD$Strengths$and$HDI,$and$KLD$Strengths$and$cultural$distance.$$
$
Models$1$to$5$were$all$overall$significant.$Model$1$F(6,232)$=$9.55,$(p<0.01);$Model$2$ F(7,231)$=$8.30,$(p<0.01);$Model$3$F(8,230)$=$7.35,$(p<0.01);$Model$4$F(9,229)$=$6.52,$(p<0.01);$ and$Model$6$F(12,226)$=$5.20$(p<0.01).$In$all$five$models$there$were$no$problems$with$ multicollinearity.$All$tolerance$values$were$over$0.1$and$VIF$values$below$10.$Constant,$size,$and$significant$in$the$first$four$models.$None$of$the$predicted$moderating$variables$appeared$to$be$ significant.$Interaction$between$variables$was$not$significant$either.$The$biggest$variance$was$ explained$in$Model$5$(R$square$=$0.216).$Adjusted$R$square$was$biggest$in$the$Model$3$with$the$ value$of$0.176.$$ All$five$models$were$significant,$however,$none$of$the$predicted$moderating$variables$ was$found$to$have$an$effect$on$ROA.$KLD$strengths$and$advertising$expenditure$had$a$positive$ coefficient,$however,$the$effect$was$very$small.$Industry,$size,$and$risk$had$a$negative$ coefficient.$The$analysis$suggested$that$risk$had$the$highest$influence$in$the$CSP$and$CFP$ relationship.$$ $
KLD$Concerns$as$predictor$and$ROA$as$dependent$variable$
The$next$analysis$compared$KLD$concerns$as$a$predictor$and$ROA$as$dependent$variable.$This$ analysis$resulted$in$the$Durbin$Watson$value$of$1.93.$Casewise$diagnostics$and$check$for$ leverage$values$and$influential$points$was$done.$As$a$result,$10$cases$were$deleted.$$
$
$
$$ The$standardized$residuals$were$relatively$normally$distributed$as$seen$on$the$histogram$and$Pb P$plot.$
$
$
$
Model$6$tested$only$the$effect$of$control$variables,$Model$7$added$the$degree$of$ internationalization,$Model$8$added$HDI,$Model$9$featured$also$cultural$distance,$and$Model$10$ tested$interactions$between$KLD$Concerns$and$the$degree$of$internationalization,$KLD$Concerns$ and$HDI,$and$KLD$Concerns$and$cultural$distance.$$
$
All$models$were$overall$significant.$Model$6$F(6,233)$=$7.78,$(p<0.01);$Model$7$F(7,232)$ =$7.16,$(p<0.01);$Model$8$F(8,231)$=$6.4,$(p<0.01);$Model$9$F(9,230)$=$5.76,$(p<0.01);$and$Model$ 10$F(12,227)$=$4.5.$All$tolerance$and$VIF$values$were$within$the$limits.$In$all$models,$only$ constant,$size,$and$advertising$expenditure$were$significant.$Model$10$explained$19.2%$variance$ in$ROA$(R$square$=$0.192).$The$highest$adjusted$r$square$was$in$Model$7$and$8$with$the$value$of$ 0.153.$$$ Size$had$again$a$negative$coefficient$and$advertising$expenditure$had$a$positive$ coefficient.$The$predicted$moderating$variables$did$not$appear$to$have$any$effect$in$these$ models.$Furthermore,$KLD$concerns$did$not$appear$to$be$significant$either.$$ $Independent'Variable Model'1 Model'2 Model'3 Model'4 Model'5 Constant .319*' (.031) .320*' (.031) .368*' (.063) .358*' (.077) .347*' (.077) KLD'Strengths' .003*' (.001) .002**' (.001) .002**' (.001) .002**' (.001) .002' (.001) Industry H3.950EH6' (.000) H3.896EH6' (.000) H3.681EH6' (.000) H3.774EH6' (.000) H4.251EH6' (.000) Size H.047*' (.007) H.047*' (.008) H.048*' (.008) H.047*' (.008) H.045*' (.008) Risk H.046**' (.024) H.047**' (.024) H.049**' (.024) H.049**' (.024) H.049**' (.024) R&D'Expenditure H.009' (.007) H.010' (.007) H.010' (.007) H.010' (.008) H.009' (.008) AdverQsing'Expenditure .017**' (.007) .016**' (.007) .017**' (.007) .017**' (.007) .017*' (.007) Degree'of'InternaQonalizaQon .000' (.000) 4.977EH5' (.000) 0.0000317' (.000) .000' (.000) HDI H.002' (.002) H.001' (.002) H.001' (.002) Cultural'Distance .003' (.010) .005' (.011) KLD'Strengths'*'Degree'of'InternaQonalizaQon 8.657EH5' (.000) KLD'Strengths'*'HDI .000' (.003) KLD'Strengths'*'Cultural'Distance .000' (.001) Adjusted'R'square .177 .177 .176 .173 .175 *p<0.01 **p<0.05 Coefficients'and'Standard'Errors;'ROA'as'dependent'variable
Independent'Variable Model'6 Model'7 Model'8 Model'9 Model'10 Constant .288*' (.032) .297*' (.033) .360*' (.068) .322*' (.082) .316*' (.083) KLD'Concerns .001' (.002) .002' (.002) .002' (.002) .001' (.002) .002' (.002) Industry D2.704ED6' (.000) D2.359ED6' (.000) D2.192ED6' (.000) D3.340ED6' (.000) D4.054ED6' (.000) Size D.039*' (.009) D.042*' (.009) D.042*' (.009) D.041*' (.009) D.041*' (.009) Risk D.050' (.025) D.053' (.025) D.055' (.025) D.054' (.025) D.056' (.025) R&D'Expenditure D.007' (.007) D.011' (.008) D.010' (.008) D.011' (.008) D.012' (.008) AdverPsing'Expenditure .024*' (.007) .023*' (.007) .024*' (.007) .024*' (.007) .024*' (.007) Degree'of'InternaPonalizaPon .000' (.000) .000' (.000) .000' (.000) .000' (.000) HDI D.002' (.002) D.001' (.002) D.001' (.002) Cultural'Distance .009' (.010) .009' (.011) KLD'Concerns'*'Degree'of'InternaPonalizaPon .000' (.000) KLD'Concerns'*'HDI D.003' (.004) KLD'Concerns'*'Cultural'Distance .000' (.001) Adjusted'R'square .145 .153 .153 .152 .149 *p<0.01 **p<0.05 Coefficients'and'Standard'Errors;'ROA'as'dependent'variable
KLD$Strengths$as$predictor$and$ROS$as$dependent$variable$
The$next$analysis$tested$KLD$strengths$as$a$predictor$and$ROS$as$a$measure$of$financial$ performance.$Casewise$diagnostics$and$the$check$for$leverage$and$influential$points$resulted$in$ the$reduction$of$the$sample$to$243.$The$Durbin$Watson$value$was$2.18.$$ $ As$seen$on$the$histogram$and$PbP$plot,$the$standardized$residuals$were$relatively$normally$ distributed.$$ As$in$previous$analyses,$the$first$model$(Model$11)$tested$only$the$influence$of$control$ variables.$Model$12$added$the$degree$of$internationalization,$Model$13$added$HDI,$and$Model$ 14$included$cultural$distance$as$well.$$ All$four$models$were$significant.$Model$11$F(6,236)$=$5.63,$(p<0.01);$Model$12$F(7,235)$ =$5,$(p<0.01);$Model$13$F(8,234)$=$4.59;$and$Model$14$F(9,233)$=$4.07,$(p<0.01).$All$tolerance$ and$VIF$values$were$within$the$limit.$In$Model$9,$only$industry$was$significant$(p<0.01).$In$the$ Model$11$to$13$KLD$strengths$also$became$significant$(p<0.05).$The$strongest$models$in$terms$of$variance$explained$were$Models$13$and$14$both$explaining$13.6%$of$variance$in$ROS$(R$square$=$ 0.136).$Adjusted$r$square$was$highest$in$Model$13$with$the$value$of$0.106.$$ As$opposed$to$the$previous$analysis,$coefficients$had$different$signs$here.$KLD$strengths$ had$a$negative$coefficient$and$industry$had$a$positive$coefficient.$The$effect$of$these$variables$ was,$however,$very$small.$$
KLD$Concerns$as$predictor$and$ROS$as$dependent$variable$$
This$analysis$tested$the$relationship$between$KLD$concerns$as$an$independent$variable$and$ROS$ as$a$measure$of$financial$performance.$The$Durbin$Watson$test$revealed$the$value$of$2.17.$$ $ $ The$histogram$and$PbP$plot$confirmed$relatively$normal$distribution$of$standardized$residuals.$$ Model$15$tested$only$the$influence$of$control$variables,$Model$16$added$degree$of$ internationalization,$Model$17$added$HDI,$and$Model$18$also$included$cultural$distance$as$ additional$variable.$$Models$15$to$18$were$all$significant$(p<0.01).$Model$15$F(6,236)$=$5.64,$(p<0.01);$Model$ 16$F(7,235)$=$4.86,$(p<0.01);$Model$17$F(8,234)$=$4.43,$(p<0.01);$and$Model$18$F(9,233)$=$3.93,$ (p<0.01).$All$tolerance$values$were$over$0.1$and$VIF$values$below$10.$In$all$cases,$only$industry$ was$found$significant$(p<0.01)$with$a$negative$coefficient$sign.$$ The$comparison$of$KLD$strengths$and$concerns$to$ROS$did$not$reveal$any$significant$ results.$Only$KLD$strengths$and$industry$in$the$first$analysis$and$industry$in$the$second$analysis$ were$significant.$$ $
Independent'Variable Model'11 Model'12 Model'13 Model'14 Model'15 Model'16 Model'17 Model'18 Constant :.003' (.145) .003' (.145) :.313' (.285) :.344' (.339) :.009' (.146) .003' (.148) :.278' (.284) :.314 '(.339) KLD'Strengths :.008' (.005) :.009**' (.005) :.009**' (.005) :.009**' (.005) KLD'Concerns :.014' (.008) :.014' (.008) :.013' (.008) :.013' (.008) Industry 6.317E:005*' (.000) 6.346E:005*' (.000) 6.185E:005*' (.000) 6.154E:005*' (.000) 5.175E:005*' (.000) 5.214E:005*' (.000) 5.105E:005*' (.000) 5.058E:005*' (.000) Size :.007' (.035) :.011' (.035) :.010 '(.035) :.009' (.036) .008' (.040) .002' (.041) .002' (.041) .003' (.042) Risk .093' (.111) .086' (.111) .097' (.111) .097' (.112) .107' (.111) .103' (.112) .111' (.112) .111' (.112) R&D'Expenditure .044' (.034) .033' (.035) .029' (.036) .029' (.036) .032' (.033) .026' (.035) .022' (.035) .021' (.036) AdverRsing'Expenditure .045' (.033) .044' (.033) .042' (.033) .042' (.033) .035' (.032) .034' (.032) .032' (.032) .032' (.033) Degree'of'InternaRonalizaRon .001' (.001) .002' (.001) .002' (.001) .000' (.001) .001' (.001) .001' (.001) HDI .011' (.008) .011' (.009) .010' (.009) .010' (.009) Cultural'Distance .008' (.048) .009' (.049) Adjusted'R'square .103 .104 .106 .102 .103 .101 .102 .098 *p<0.01 **p<0.05 Coefficients'and'Standard'Errors;'ROS'as'dependent'variable
KLD$Strengths$as$predictor$and$ROE$as$dependent$variable$$
This$analysis$compared$KLD$strengths$with$ROE$as$a$measure$of$financial$performance.$Casewise$ diagnostics$detected$20$outliers$that$were$deleted.$The$Durbin$Watson$test$resulted$in$the$value$ of$2.09.$$ $ As$seen$on$the$histogram$and$PbP$plot,$the$standardized$residuals$were$relatively$normally$ distributed.$$ As$in$previous$models,$Model$19$only$tested$the$effect$of$control$variables.$Model$18$ included$degree$of$internationalization,$Model$20$added$HDI,$and$Model$21$added$cultural$ distance$as$well.$$ All$models$in$this$analysis$were$significant$(p<0.01).$Model$19$F(6,223)$=$8.94,$(p<0.01);$ Model$20$F(7,222)$=$7.74,$(p<0.01);$Model$21$F(8,221)$=$6.76,$(p<0.01);$Model$22$F(9,220)$=$6,$ (p<0.01).$All$tolerance$and$VIF$values$fell$within$the$acceptable$range.$In$all$models,$the$ constant,$KLD$strengths,$size,$and$risk$were$significant$at$(p<0.01),$advertising$expenditure$was$significant$at$(p<0.05).$Constant,$risk,$and$advertising$expenditure$had$a$positive$coefficient.$Size$ had$a$negative$coefficient.$$The$biggest$variance$explained$in$ROE$was$in$Models$21$and$22$(R$ square$=$0.197).$The$model$with$the$highest$adjusted$R$square$(0.172)$was$Model$19,$where$ only$KLD$strengths$and$control$variables$were$represented.$The$moderating$variables$did$not$ appear$to$have$any$effect.$$ $
KLD$Concerns$as$predictor$and$ROE$as$dependent$variable$
The$final$analysis$compared$KLD$concerns$with$ROE.$As$in$the$previous$analysis,$the$ sample$was$reduced$to$230$due$to$outliers.$The$Durbin$Watson$value$was$2.05.$$ In$line$with$the$previous$pattern,$Model$23$only$tested$the$effect$of$control$variables,$ Model$24$added$degree$of$internationalization,$Model$25$featured$HDI,$and$Model$26$included$ cultural$distance$as$well.$$ $The$histogram$and$PbP$plot$confirm$relatively$normal$distribution$of$standardized$residuals.$$ All$models$in$the$analysis$were$significant$(p<0.01).$Model$23$F(6,223)$=$6.69,$(p<0.01);$ Model$24$F(7,222)$=$6.1,$(p<0.01);$Model$25$F(8,221)$=$5.33,$(p<0.01);$Model$26$F(9,220)$=$4.72,$ (p<0.01).$All$tolerance$and$VIF$values$were$within$the$limit.$In$Model$23,$the$constant,$risk,$and$ advertising$expenditure$were$significant$at$(p<0.01).$In$the$subsequent$models,$size$also$became$ significant$at$(p<0.05).$Models$25$and$26$explained$the$biggest$variance$in$ROE$(R$square$=$ 0.162).$Model$24$had$the$highest$adjusted$r$square$with$the$value$of$0.135.$$ $$
Independent'Variable Model'19 Model'20 Model'21 Model'22 Model'23 Model'24 Model'25 Model'26 Constant .704*' (.138) .709*' (.138) .805*' (.281) .865*' (.345) .547*' (.142) .587*' (.144) .664**' (.284) .702**' (.351) KLD'Strengths .015*' (.004)' .015*' (.004)' .015*' (.004) .015*' (.005)' KLD'Concerns .007' (.008) .009' (.008) .008' (.008) .008' (.008) Industry I1.725EI005' (.000) I1.697EI005' (.000) I1.653EI005' (.000) I1.603EI005' (.000) I9.215EI006' (.000) I7.969EI006' (.000) I7.773EI006' (.000) I7.352EI006' (.000) Size I.113*' (.034)' I.116*' (.034)' I.117*' (.034) I.119*' (.035) I.073' (.040) I.090**' (.041) I.089**' (.041) I.091**' (.042) Risk .649*' (.113) .644*' (.113) .644*' (.113) .644*' (.114) .647*' (.116) .636*' (.116) .636*' (.116) .636*' (.116) R&D'Expenditure I.014' (.032)' I.022' (.034)' I.021' (.034) I.020' (.034) .014' (.032) I.004' (.034) I.003' (.034) I.002' (.034) AdverRsing'Expenditure .065**' (.031) .064**' (.031)' .065**' (.031)' .065**' (.031) .086*' (.031) .083*' (.031) .084*' (.031) .084*' (.031) Degree'of'InternaRonalizaRon .001' (.001) .000' (.001) .001' (.001) .001' (.001) .001' (.001) .001' (.001) HDI I.003' (.008) I.004' (.009) I.003' (.009) I.003' (.009) Cultural'Distance I.014' (.046) I.009' (.047) Adjusted'R'square .172 .171 .168 .164 .130 .135 .131 .128 *p<0.01 **p<0.05 Coefficients'and'Standard'Errors;'ROE'as'dependent'variable
Testing$the$presence$in$low$PD,$high$IND,$low$MAS$countries$instead$of$the$degree$of$ internationalization$in$all$models$did$not$bring$different$results.$We$can$conclude,$based$on$the$ analyses,$that$the$moderating$variables$were$not$significant$in$the$models$and$therefore$do$not$ play$any$role$the$CSP$and$CFP$relationship.$Interaction$between$variables$was$not$significant$ either.$All$hypotheses$can$be$rejected.$$ What$was$interesting$to$see$was$that$in$all$cases$KLD$strengths$variable$was$significant$ together$with$most$control$variables,$as$opposed$to$KLD$concerns.$The$results$therefore$imply$ that$companies$that$get$involved$in$CSR$yield$higher$financial$performance,$but$the$fact$whether$ these$companies$also$do$bad$would$not$matter.$Perhaps$doing$good$matters$the$most$and$ translates$immediately$into$better$financial$performance,$even$though$there$can$be$some$areas$ of$social$responsibility$where$those$companies$lack$behind.$$
Discussion,$Conclusion$
This$study$tested$the$CSP$and$CFP$relationship$that$had$been$covered$in$many$studies$ with$inconsistent$results.$This$research$proposed$that$the$relationship$is$moderated$by$the$ degree$of$internationalization$of$a$company,$the$level$of$development$of$the$countries$in$which$ the$company$operates,$cultural$distance,$and$the$fact$whether$the$firm$operates$in$markets$with$ low$PD,$high$IND,$and$low$MAS.$The$underlying$assumption$of$this$theory$was$that$there$is$ higher$complexity$the$company$needs$to$deal$with$as$it$increases$its$presence$abroad.$This$ would$be$due$to$multiple$institutional$environments,$different$cultural$values,$and$stakeholder$ demands.$$ The$hypotheses$were$first$tested$using$correlation$matrix$to$see$whether$there$was$any$ association$between$variables$and$hierarchical$regression$analysis$to$find$out$how$the$proposed$ variables$contributed$to$the$overall$variance$in$the$dependent$variable.$The$study$employed$ three$different$measures$of$financial$performance$–$ROA,$ROS,$and$ROE.$Even$though,$the$study$ did$not$find$any$support$for$the$hypotheses,$since$the$moderating$variables$were$not$found$ significant,$it$did$reveal$that$KLD$strengths$variable,$ie.$positive$actions$of$CSR,$was$found$to$ have$a$slight$positive$effect$on$the$financial$performance.$The$fact$whether$a$company$that$was$ involved$in$CSR$also$lagged$behind$in$some$areas$of$social$responsibility$was$not$found$to$have$ any$influence$on$the$financial$performance.$$ $Limitations$
The$study$had$a$few$limitations.$The$first$limitation$to$mention$was$the$sample$size.$To$ be$able$to$draw$stronger$conclusions,$researchers$should$replicate$this$study$in$the$future$using$ larger$sample$size.$Secondly,$there$were$incomplete$data$on$Hofstede$cultural$dimensions$that$ may$have$caused$minor$errors.$Third,$the$degree$of$internationalization$was$operationalized$by$ looking$at$the$foreign$levelbone$subsidiaries$of$each$company$as$of$2015.$There$may$have$been$ minor$differences$in$the$amount$of$subsidiaries$between$2008b2015.$Furthermore,$looking$ deeper$in$the$organizational$structure$might$have$revealed$more$foreign$markets.$Fourth,$the$ study$used$KLD$data$as$the$measure$of$CSP.$Even$though,$this$has$been$a$common$practice$in$ the$previous$research,$the$measure$might$not$be$perfect.$It$treats$all$aspects$of$CSR$equally.$ Furthermore,$it$only$rates$companies$based$on$whether$or$not$they$are$involved$in$certain$areas$ of$CSR,$but$it$does$not$say$anything$about$the$impact$of$each$company.$Some$companies$may$ invest$the$way$more$resources$in$CSR$than$others.$Finally,$this$study$used$accountingbbased$ methods$as$a$measure$of$financial$performance.$These$methods$may$be$subject$to$different$ accounting$standards$and$managerial$styles$that$can$make$them$slightly$biased.$Including$ marketbbased$financial$performance$in$the$study$may$have$brought$stronger$results.$The$ financial$data$in$this$study$was$from$2009b2012,$right$in$the$middle$of$the$global$economic$ slowdown.$This$may$have$also$made$the$results$biased,$since$with$shrinking$economic$growth$it$ is$easy$for$companies$to$justify$cuts$on$CSR$activities.$$ $Implications$and$Future$Research$
The$implication$of$this$study$for$businesses$is$that$there$is$a$direct$link$between$social$ responsibility$and$financial$performance.$This$phenomenon$has$been$discussed$various$times$ with$advocates$and$opponents$both$outlining$arguments$for$and$against$it.$However,$in$today’s$ world$where$consumers$are$well$informed,$companies$are$easily$held$responsible$for$their$ negative$actions.$It$is$in$the$interest$of$every$company$to$invest$in$CSR$not$only$to$eliminate$ waste$and$increase$efficiencies,$but$also$as$a$way$of$enhancing$brand$reputation$and$risk$ management.$CSR$is$becoming$a$competitive$advantage$in$many$industries$where$companies$ that$offer$generic$products$try$to$differentiate$themselves$in$this$aspect.$$ This$research$was$one$of$the$first$of$its$kind$testing$the$role$of$internationalization$and$ culture$in$the$CSP$and$CFP$relationship.$To$confirm$the$results,$it$should$be$duplicated$in$the$ future.$To$broaden$the$applicability$of$results,$the$researchers$should$look$at$companies$from$ other$geographical$regions,$such$as$the$EU,$South$America,$or$Asia$to$see$how$cultural$values$ and$different$corporate$governance$systems$influence$the$CSP$and$CFP$relationship.$An$ interesting$observation$would$be$to$look$at$companies$in$the$developing$countries.$$ Secondly,$future$studies$should$take$into$account$measures$of$both$accountingbbased$ and$marketbbased$financial$performance.$The$future$research$should$be$over$longer$period$of$ time$and$might$need$to$take$into$account$economic$growth.$$ Third,$KLD$data$has$been$used$in$the$latest$research$as$the$main$indicator$of$corporate$ social$performance.$To$make$the$research$more$diverse$and$better$understand$what$aspects$of$ CSR$play$major$role$in$the$relationship$with$financial$performance,$other$sources$of$data$should$ be$used.$Future$research$could$also$try$to$measure$different$aspects$of$CSR$separately,$such$as$the$effect$of$better$environmental$performance$or$good$corporate$governance$on$financial$ performance,$since$different$aspects$of$CSR$are$often$claimed$to$have$different$importance.$$ Finally,$much$of$the$research$on$CSP$and$CFP$focuses$on$large$publicly$traded$ companies.$Since$these$companies$have$much$larger$resources$than$small$and$mediumbsized$ companies$and$are$much$more$visible$to$the$public,$they$have$better$means$and$higher$need$to$ invest$in$CSR$to$maintain$their$reputation.$Future$studies$should$look$at$small$and$mediumbsized$ enterprises$and$test$whether$investing$in$CSR$activities$results$in$higher$financial$performance.$$ $
Acknowledgements$
$ $
