The role of good corporate governance

(1)

www.elsevier.es/brq

Business BRQ Research Quarterly

ARTICLE

Social responsibility and ﬁnancial performance:

The role of good corporate governance

Mercedes Rodriguez-Fernandez

UniversityofMalaga,29071Málaga,Spain

Received12January2015;accepted1August2015 Availableonline4September2015

JEL

CLASSIFICATION M14;

G34

KEYWORDS Corporatesocial responsibility;

Financial performance;

Goodcorporate governance;

Spanishlisted companies

Abstract Theobjectiveofthistheoretical---empiricalstudyistoinvestigatethebidirectional relationshipbetweenCorporateSocialResponsibilityandFinancialPerformanceinSpanishlisted companies.A completetheoreticalframework---basedonagency,stewardship,dependency resources,andstakeholdertheories---providesthebasisfortheconceptualmodel.Animpor- tantcontributionistheuseofasocial behavioralindexformedbyfour components:Global ReportingInitiativeparticipation,DowJonesSustainabilityIndexﬁrminclusion,GoodCorporate GovernanceRecommendationscompliance,andGlobalCompactsignee.

Theconclusionsdrawnfromtheempiricalstudyperformedonthecompaniesregisteredon theMadridStockExchangedemonstratepositiverelationshipsinbothdirections,namelythat thesocialisproﬁtableandthattheproﬁtableissocial,therebyoriginatingapositivefeedback virtuouscircle.

The results ofthisanalysis have practical applicationsinthe boardroom; they areproof thatallsocialpoliciesincrementfinancialresources,andviceversa,thatincreasedfinancial performancesleadtogreatersocialbenefits.Asaconsequence,thispaperencouragesallboard memberstoseriouslyweighinvestingfinancialresourcesindevelopingpoliciesthatboostthe levelsofsocial behaviorcomponentsinordertocontributeglobally totheimprovementof society.

Abbreviations:CG,corporategovernance;CSP,corporatesocial performance;CSR,corporatesocialresponsibility;DJSI,DowJones SustainabilityIndex;FP,ﬁnancialperformance;GC,globalcompact;

GRI,globalreportinitiative;R&D,ResearchandDevelopment;ROA, returnonassets;ROE,returnonequity;TOBINQ,Tobin’sQ;UCGG, uniﬁedcodeofgoodgovernance.

E-mailaddress:[email protected]

Introduction

Companyconcerns areincreasingly focused toward issues ofsocialcontent,allthewhile resolvingtomaximizeeco- nomicperformanceinordertosatisfyshareholdersandact inasociallyresponsible mannerfor thebeneﬁtof society asawhole.Social, economic andenvironmental concerns areforcingcompaniestointegratesystemsthat takeinto accounttheobservanceofthelawinallspheres,andalso

http://dx.doi.org/10.1016/j.brq.2015.08.001

(2)

focusonthecommongoodforsocietyingeneralandstake- holdersinparticular.

Shareholders, throughthegeneral assemblies,exercise theirroleindemandingethicalattitudesandbehaviorsat thecorporatelevel,therebyexercisingstronginﬂuenceon the formulation of strategies by the board of directors.

They require transparency, efficiency and efficacy onthe part of managers, in order to obtain economic benefits and,thus ensure the continuityof thecompany over the longterm,whilstdemandingthatsociallyresponsiblepoli- cies be integrated into the companies themselves (Pava andKrausz, 1996).Froman academicpointofview, there existsanincreasingdemandindevelopingbusinessethics--- byintegratingasresearchobjectivethedetectionofillicit businessescontrarytosocialrights(Byrne,2011).Business ethics is by no means a recent development, a present- day trendy topic; studies demonstrate that conformation toethicalstandards andprincipleshasbeen an issueper- sistingthroughtheagesandwithstandingthetestof time (Michalos,2008).

PresentlyCSRandbusinessethicsareintricatelyrelated from both an academic and practical perspective. We consider a variety of CSR deﬁnitions proposed in the literatureandbyseveralinstitutionsthatemphasizeavol- untaryinvolvementinthesolutionofcertainsocialissues;

socialresponsivenessisfundamentallymultidimensionaland embodiesalargeandvariedrangeofcorporatebehaviorin relationto itsresources,processesand outputs (Waddock andGraves,1997).

Amajorityofresearchtodateonthisthemehasfocused ontherelationship between Corporate Social Responsibil- ity(CSR) andFinancial Performance(FP).Generally,these ﬁndingsshowthisrelationshiptobepositive,howeverthere existsalackofhomogeneityintheresults.Thereasonsare twofold:(1) theabsenceof ageneral methodthat serves asyardstick for comparative studies, and (2) thereexists norigorousmethodofmeasuringreturnonCSR(Gjølberg, 2009).

Ourworkstrivestoﬁllthisexistinggapintheliterature;

andwiththisaimwesettwomainobjectives:First,todeter- mineifFinancialPerformancedependsonCorporateSocial Responsibility,representedasacombinedfunctionoffour distinct social variables: Global Reporting Initiative (GRI) participation, Dow Jones Sustainability Index (DJSI) firm inclusion,Good Corporate Governance (CG)recommenda- tionscompliance,andGlobalCompact(GC)signee.Second, totesttheinverserelationship,socialresponsibilitydepend- enceonfinancialperformance,usingaCSRindex,orSocial BehaviourIndex,thatincludesthepreviousfoursocialvaria- blesasequalweightedcomponentsaggregatedinaunique value.Inbothcasesfinancialperformanceisrepresentedby threefinancialvariables or ratios,namely ROA(return on assets),ROE(returnonequity)andTobin’sQ.

A further objective sets to reveal whether ﬁrms are interestedindevelopingCSRfromaneconomicstandpoint, basedonthefactthatitrepresentsanimportantengineof developmentand contributestoward improving socialand environmental protection. As collateral beneﬁt,investors andconsumerscanbetterevaluatecompaniesthattakeinto accountCSR actions.The projectionof animage ofsocial responsibility,byhelpingtoshapecustomersoftheentity, affectstheevaluationoftheservicereceived.Intheirrole

associalagents,companiesareexpectedtomeetcommit- mentsthatgobeyondstrictlybusinessmatters.

Ourstudy willbe centered onSpain, a representative memberof thegroup of developedcountries,where such a study has yet tobecarried out andwhere the increas- inginternationalizationofthecountry’slargestﬁrmsshould clearly illustratethe importance of adoptingfar reaching corporatesocialpolicies.

InsheddinglightonthekeyinterplaybetweenCSRandits FP,weexpecttheresultsofourworktogoalongwaytoward convincingcorporateboardsthatsocialpoliciesmustform anintegralpartofoverallcompanystrategy.Takenfroma stakeholders’perspective,thestudyshouldalsobeofvalue toallinterest groupsthat layclaim,or stake,in acorpo- ration’swellbeing.Finally,bothcorporationsandsocietyat large beneﬁtfromincreasedawareness in companysocial undertakings, reconciling at times differentviewpoints as to whether corporate proﬁts are adequately distributed amongstallstakeholders.

Thepaperisstructuredasfollows:‘Theoreticalframe- work’ section deals with the theoretical framework, presentingthemainadministrativetheoriesandconceptson whichthestudyisbased.WeintroduceCG,CSRandFPasthe threekeypillarssustainingourconceptualmodel.‘Proposed model and formulation of hypotheses’ section establishes the twobasic hypotheses that address the question asto whetherCSRexplainsFPandviceversa.In‘Empiricalstudy’

section,weempiricallytestourhypothesesusingtheSocial Behaviour Index for measuring CSR and the three ﬁnan- cialratiosrepresentingFP.Thelastsectionconsistsofthe conclusionandsuggestsfuturelinesofresearch.

Theoretical framework

Reviewofthemaintheoriesapplicabletothestudy The theoretical framework underlying this work includes a number of differenttheories. Theirdistinct approaches areallpertinentinsome measure.Wecandistinguish,on the one hand, the set of theories applicable tothe rela- tionbetweenCSR,FP,andCG---conformingtheconceptual modelof ourstudy---and,onthe otherhand,stakeholder theory --- the unique theory --- that supports the relation betweenCSRandFP.Theintegrationofthesediversecon- structsenrichestheliteratureandstrengthenstheproposed genericmodel.Fig.1illustratesthetheoreticalframework usedinthestudy.

Agencytheory(JensenandMeckling,1976;Fama,1980;

Fama and Jensen, 1983) establishes that the principal (shareholder)andtheagent(manager)haveopposinginter- eststhatmaytriggerconﬂictswhichwillinterferewiththe smooth running of the company. In contrast, stewardship theory offersan alternative view, whichstates thatthere existethicalandprofessionalmotivesthatwilloverrideand preventconﬂictsofinterestfromdeveloping betweenthe principalandagent(MuthandDonaldson,1998).Thislatter theoryassumesthatmanagersaregoodresourcemanagers (Donaldson,1990;DonaldsonandDavis,1991,1994)whowill achievegoodbusinesstrackrecordsthankstotheirefforts (Davisetal.,1997);inaddition,managers,ashonestpeople (DonaldsonandPreston,1995),endeavortonothinderthe

(3)

Coporate social responsibility, financial performance &

corporate governance

Coporate social responsibility &

financial performance

Conceptual model

Agency theory Stewardship theory

Stakeholder theory

Descriptive Instrumental Normative Resources dependenccy theory

Theoretical institutional perspective

Empirical study

Figure1 Theoreticalframework.

Source:Ownelaboration.

objectivesoftheshareholders(DonaldsonandDavis,1994) inordertopreservetheirreputation.Bothagencyandstew- ardshiptheories,intakingtheboardofdirectorsasprincipal andthe executive body asagent,come intoconﬂict with regard to the consideration of who is responsible for the policiesofsociallyresponsibleinvestment andtheactions ofCSR.

Resource dependencytheory analyzesthe relationships andinteractionsofthecompanieswithotheragents,valu- ingtheircontributionsonthebasisoftheextenttowhich theyfacilitatethemaximizationofitsperformance(Pfeffer, 1973; Pfeffer and Salancik, 1978).The board of directors plays a key rolein obtaining important resources for the company,suchasﬁnancialresourcesthatcanlaterbeear- marked for socially responsible investments and actions.

Also of interest, the approach offered by the theoretical institutionalperspectivedevelopedbyScott(2001),which holdsthat allsocialparticipantsseeklegitimacyandinso doinghelpdeveloplegitimateruleswithintheinstitutional environment(Judgeetal.,2010).Ifcompaniesﬁxastheir objectivethequestforlegitimacyovereconomicefﬁciency (Carver,2010)andifCGblendsinaneconomic,cultural,and socialcontext,thensocial welfareandthebalanceof the interest groupsmust take center stage(Hessand Warren, 2008;JohansonandÖstergren,2010).

Alloftheabovetheories,inconsideringCGasresponsible fortheactionsofCSRandthepoliciesofsociallyresponsi- bleinvestments,underlieandsupporttheinterrelationship between companyCSR andFP. However,stakeholder theory,anintegrativeandholisticperspective,asitconsiders society in general,serves asmain pillar for ourresearch.

Good CSR policy together with the appropriate behavior of theboard of directorswillimprove ﬁnancialproﬁtabil- ity,andfavorshareholders,employees,customers,suppliers andallotheragents likelytobeaffectedbythedecisions taken by the company. The academic debates surround- ingthisapproach(Freeman, 1984;DonaldsonandPreston, 1995;Donaldson,1999;JonesandWicks,1999;Prestonand Donaldson, 1999; Sternberg, 1999; Pesqueux and Damak- Ayadi,2005;KaufmanandEnglander,2011)havedeveloped over time. The descriptive, instrumental and normative aspectsofthetheoryareeverpresentinresearch.Although

quitedifferentfromeachother,aspointedoutbyDonaldson (1999), they are complementary and able to explain the interplaybetweenCSRandFP.

Thedescriptiveaspectprovidesanotionforthedeﬁni- tionofacompany;DonaldsonandPreston(1995)describeit asaconstellationofcooperativeandcompetitiveinterests withintrinsicvalue.Fromaninstrumentalpointofviewthe theoryprovidestheframeworkforexaminingthecompanies and analyzing the relationship between management and theachievementofperformanceobjectives(Surrocaetal., 2010);itadvocatesthatcompaniesestablishanorderofpri- orityamongst itsinterest groups andfavorthose whoare bestpositioned. Thus,thelevelofeffortinCSR exercised bycompaniesdependslargelyontherelativeimportanceof theirinterestgroups(Choietal.,2010).Thiscontrastswith thenormativeaspect ofthistheory, which focusesonthe legitimacyofthecompany’sinterestgroups andthevalue oftheirinterests,alwaysworthyofattentionregardlessof category(Kaufmanand Englander,2011).Consequently, it becomesimperativetointroducegoodCGrecommendations asanimportantelementofCSR.

Corporategovernanceanditsimplications

Corporate Governance arises as a result of the separa- tion between ownership of the business and its control inresponse toa system bywhich companiesaredirected andcontrolled(Cadbury,2000).Agencytheory(Jensenand Meckling,1976;Fama,1980;FamaandJensen,1983)pro- videstherationaleforthepossibleconﬂictthatcandevelop betweentheprincipal(shareholders) andtheagent(management).AsexplainedbyGuerrasMartínandNavasLópez (2004)theproblemofownercontrolonmanagementandthe mechanismsavailable toexercisethat controlisknownas CG;aspeciﬁcmechanismofgovernance,suchastheboard ofdirectors,playsarelevantroleindiscipliningandadvis- ingmanagementontakingthemost appropriatedecisions ateveryjunctureandforeachorganization(Cuervo,2002).

The board of directors must ensure the long-term viability of the company by maximizing proﬁtability for shareholders(DailyandDalton,1994)andharmonizingthe

(4)

interestsofthecompanywiththoseoftheinterestgroups (CoombsandGilley,2005).Thedecisionstakenbytheboard willleadtodistinctlevelsofFP,thepossibleimplementation ofCSRpolicies(Ingleyetal.,2011),andthedeploymentofa particularstrategyofsociallyresponsibleinvestment(Mill, 2006).

As a result of the increasing importance of ethical behaviorinbusinessandthedemandfortransparencyand information by shareholders, corporate codes of conduct areemerginginmanycountries(AmaeshiandAmao,2009;

Stiglbauer,2010; Mody and Mudoi, 2011). Spain published thedirectivesforCGinlistedcompanies,orOlivenciaCode (CNMV,1998), in response tothe social demandfor more efﬁciency, agility and transparency from the companies.

The AldamaReport (CNMV, 2003), basedon the Olivencia Codewaspublishedin2003;thiswasfollowedbytheCNMV (2006)approvalin 2006of theUniﬁedCode ofGoodGov- ernment(UCGG),requiringlistedcompaniestosubmittheir annual reports on CG in accordance with the recommen- dationsestablishedin article116 ofthe SecuritiesMarket Act.

TheUCGG(CNMV,2006)clearlylaysoutthesocialaspect the board of directors of listed company must abide by, expressingthatalldirectorsmusthavethecommonobjec- tive of defending the social interest, understanding it to be that which best meets the expectations of investors, althoughnotimplyingthattheseinterestsmustbepursed at‘‘anyprice’’.Amongthecorepoliciesandstrategiesthe boardofdirectorstakesonasmission,weemphasizethat relatedtothe approvalof CGpolicyandCSR (recommen- dations8.a.iv and8.a.v).Companycompliancewiththese andotherrecommendations(58intotal)isdisclosedbythe publicationof theannualreportonCG,speciﬁcallyinthe ﬁnalsection,whichdealswithcompliance,orlackthereof (listingmotives).ThetotalnumberofUCGGrecommenda- tionsbasicallyconstitutescorporateCSR‘‘accountability’’

beforesocietyingeneral.

CSR is invariably linked tothe seminal work of Bowen (1953),whichstatesthatacompany’ssocialandeconomic responsibilities are inseparable. The vision that incorpo- rates CSR to the business objective does not refer to whatcompaniesaresearchingfor, butrathertohowthey aregoing about it. Achievement includes matters related withsociety and theenvironmentin alldimensions (Davis and Blomstrom, 1975). CSR’s ﬁve fundamental principles accordingtotheCSRObservatory(2010)¹ are:compliance withlegality;universality,thatis,coverageofallareasof activity;theobligationtoacceptobjectiveethicalcommit- ments;itsmanifestationthroughgenerated impacts;and, ﬁnally, its orientation toward the satisfaction of interest groups.

CSRhasbeendeﬁnedinmanyways(Maak,2008)andits contenthasevolvedovertime(Argando˜naandVonWeltzien Hoivik,2009).However,allofthemmakereferencetoboth theimportanceofinterestgroupsandtheconcernforsocial andenvironmentalmatters(Maak andPless,2009;Luand Liu,2013).The Green Bookof the EuropeanCommunities

1TheCSRObservatorywascreatedin2004byagroupoforga- nizationsrepresentingSpanishcivilsocietywiththeobjectiveof favoringknowledgeandcompliancewithCSR.

Commission (2001) states that CSR consists not only of a company’s voluntary compliance withsocial and environmental issues, but also its respect for existing rules and regulationsin thosecountrieswhere itoperates; and,the participation in the developmentof thesewhere theydo notexist.

ForMcWilliamsandSiegel(2000)CSRconsistsofactions thatfavorsocialwellbeingbeyondtheinterestsofthecom- panyorthestipulationsrequiredbylaw.Hence,companies shoulddirecttheireffortstowardmanagementmodelswith agreatersocialcontent,notwithstandingassuringproﬁtabil- ity, the zealous respect for the law, standingon superior moral grounds, and defending collective interests. This concept is embraced by those organizations that believe

‘‘something must be given back’’ to society (Lindgreen etal.,2009).

CG and CSR aretwo concepts that have been studied separatelyinpreviousliterature,(BhimaniandSoonawalla, 2005)andtherelationshipsbetweenthetwogeneratebene- ficialsynergies(Jamalietal.,2008;Chanetal.,2013).From aSpanishUCGGperspective,thefunctionsoftheboard of directors includes defendingthe interestsof all shareholders,albeitalwaysrespectingthelaw,honoringthirdparty agreements, and abiding by all CSR policies. With regard totherelationshipbetweenthesetwoideas,HarjotoandJo (2011)makeaninterestingcontribution:theauthorsdiscov- eredthatinfullconsistencywiththehypothesisofconflict resolution,thechoiceofCSRstrategyispositivelyassociated withthecharacteristicsofCG.But,moreimportantly,after correcting the CSR measures for endogeneity, the results showedthatcommitmentwithCSRleadstoimprovedfinan- cialreturns.

The existing connection between CG and CSR has contributed to the development of regulations on CG, introduced in some European countries (Spain, UCGG;

UnitedKingdom,‘‘CadburyReport’’;France,‘‘TheViénot’’;

and, Germany ‘‘The Gerhard Cromme’’), and which has served to clarify the roles and responsibilities of the companies, the boards of directors, and theshareholders (Rodríguez-Fernández, 2015). In Spain, the publication of codes of conduct for listed companies has provided the opportunityforshareholderstoassumetheirroleasowners ofthecompany(AlbaredaandBalaguer,2009).Attitudesof fullcommitmentwithsocietyandtheenvironmenttogether withethicalcodesofconducthaveledtobusinessstrategies thatcovermanifoldCSRpolicies,addingyetmorerespon- sibilitiestotheboardofdirectors,everinvolvedwithsuch issuesashumanrights,briberyandcorruption,andglobal change (Elkington, 2006). In Asia, Welford (2007) revises theselinksandexpressesthatCSRpracticesareoftenbased ongoodstandardsofCGprovidingasolidCSRfoundation--- bycreatingvalueaddedrelationshipswithallstakeholders.

Companiesmust understandCSR managementasaway todevelopproper CG(SpenceandPerrini,2009).Byinte- grating CSR within the activities of companies, different norms, guidelines, management systems, and other standards have risen totheforefront. The implementation of management systemsallows for the development of CSR.

The Global Reporting Initiative (GRI) promotes the draft- ingofCSRreports,so-calledsustainabilityreports,andthe GlobalCompact,arestatementsofcommitmentswithsoci- ety,theenvironment,anddevelopment.

(5)

Albareda (2013) afﬁrms CSR reporting standards are convergingtowardhomogeneousguidelinesunderthepre- dominance of the GRI model; the consolidation of this standard is demonstrated by Boesso et al. (2013), who collected data --- in accordance with the GRI guideline --- relatedtoCSR.WilburnandWilburn(2013)examinedtheGRI reportingguidelinesanditsapplicabilitytoCSRprinciples, detailingitssuitabilityinallowingcompaniestoformulate CSRstrategiesandhelpingstakeholdersevaluatethosesame strategies.

With respect to the Global Compact commitments as part of the ongoing CSR drive, Knudsen (2011) finds that firmsfromcountrieswithinternationaleconomiesaremore willing to abide by the Global Compact specifications.

Cetindamar(2007),and Ruggie(2004) illustratehowcom- panies that have participated many years in the Global Compact regard their CSR involvement as having had a stronginﬂuenceontheirmarketperformanceandcreation ofvalue.

Finally,Strand(2013)demonstratesthatcompanieswith astrongerfocusonCSRarethreetimesmorelikelytobe includedinDowJonesSustainabilityIndex(DJSIintheNew YorkStockExchange).Lopezetal.(2007)intheirempirical analysisrelateinclusionintheDJSIwithactiveCSRpolicies.

Corporatesocialresponsibilityandﬁnancial performance

Fromatheoreticalperspective,stakeholdertheory(McGuire et al., 1988) sets the framework for the relationship between CSR and FP; interest groups claim company resources, and in so doingimplicitly require proper com- panybehavior, such asconsideration for the environment andconcernforfairandjustlaborrelations.Inthosecases inwhichthecompanydoesnotactwithsocialresponsibil- ity,resultantcostscouldbecomesignificantandrepresent a financial burden likely to reduce profits, leading to a less socially aware entity. In contrast, if companies that adoptsociallyresponsiblepoliciesaremoreprofitable,then sociallyresponsibleinvestmentswillprovideanincentivefor businessestoincreaseinvestmentsinCSR programs(Pava, 2008).

Numerous studies (Cochran and Wood, 1984; Aupperle et al., 1985; McGuire et al., 1988; Waddock and Graves, 1997; McWilliams and Siegel, 2000; Orlitzky etal., 2003;

Smith,2003;Ortasetal.,2014)testifytotheever-present dichotomy between CSR and FP; however, there exist no clear-cutconclusionsthat clarifythepositive,negativeor inexistentcorrelation.Thereasonslieintheimperfections ofthestudies(causedbyproblemsinmeasuringFPandCSR), theomissionof signiﬁcantlatentvariables intheformula- tionof themodels,the absenceof causalityanalysis, the lack ofrigorin themethodology,andbya shortfallin the theoryunderpinningthestudy(MargolisandWalsh, 2003).

Nonetheless,Stanwick andStanwick(1998)reviewedstud- iesthatexaminedtheeffectsofCSRonFPandconcluded thatthereexistsapositive,albeitweak,relationship.

Moreover,otherauthors(WoodandJones,1995;Akpinar et al., 2008) have argued about the existence of a

‘‘misalignment’’problemintheinterestgroupsasacause forthevarianceintheresults;thesolutionliesinidentifying

themajorinterestgroups mostimportanttothecompany.

In this regard, Alniacik et al. (2011) conclude that posi- tiveinformationoncompanyCSRleadstobothemployment desirabilityattheﬁrm,andtoanimprovementinpurchase andinvestmentintentions.Akpinaretal.’s(2008)contribu- tionsettledthequestion astowhetherallinterestgroups heldthesameimportance;hisstudy,basedoninterestgroup theory,concludedthattherelationship waspositiveifthe measurementofCSRtookintoaccounttherelativeimpor- tanceofeachinterestgroup.

Customers, employees, suppliers, shareholders, and societyasawholerepresentinterestgroupsforthecorpo- rations;however,instrumentaltheorypositsthatinvestors leantowardthosecompanieswithsuperiorsocialbehavior whenallotherfactorsremainconstant,andtheinformation onsocialresponsibilityisindependentlyavailable.Thethe- oryfurtherpostulates(Choi etal.,2010)thatthelevelof effortthecompaniesdedicatetothedifferentareasofCSR dependsonthe importancegiventothem by eachof the interestgroups.

Inthesamesense,BrammerandPavelin(2006)introduce theaggregateconceptofcorporatereputationthatreﬂects theperceptionsofahostofindividualstakeholders.Demon- stratingahighdegreeofsocialresponsibilitymaytherefore requireadiverse rangeofsocial activities, eachof which mayhaveaseparatelyidentiﬁableimpactuponreputation.

Furthermore, stakeholder groups have differing expecta- tionsregardingﬁrmbehavior(FombrunandShanley,1990), and the salience of each stakeholder group varies across industries.Therefore,the impactof CSRactivism onrep- utationisjointlycontingentuponwhichtypeofCSRactivity isundertaken.

Sternberg (1999) points out that the interest group approachpresentstwomajordrawbacks:ﬁrst,theneedto resolve the conﬂict between the values, objectives, and interests of the stakeholders; and second, the need to correctlyaccountforresponsibility,stressingthatinthetra- ditional corporation the directors are accountable to the shareholders,whereastheemployeesandotheragentsare responsible,throughthetoplevelexecutives,tothedirec- tors.However,thisdoctrineexplicitlyrejectsbothtypesof responsibilities.This rejection is oneof thedistinguishing featuresof the stakeholderapproach, which insteadpro- posesadiffuseandineffectivestructureofresponsibilities.

McWilliamsandSiegel(2000)reachaninterestingconclu- sion:thelackofcorrelationbetweenﬁnancialproﬁtability andCSRiscaused byerrorsinthestatistical analysesand thenon-inclusion of investment in the Research& Devel- opment(R&D)variable;thelattercorrelatesmodestlywith CSR.TheypointoutthatinvestmentinR&Dcorrelateswith bothFPandCSR;thiscorrelationisduetotherelationship betweeninvestmentinR&Dandinnovationofproductsand services.

The reviews of Choi et al. (2010) show the results to bemostlypositive,althoughsomearenegative,mixed, or uncorrelated.MargolisandWalsh(2003)reachedthesame conclusioninthereviewsof127studies,carriedoutbetween 1972and2002;theresultsshowedamostlypositivecorre- lationindependently ofwhetherCSRwastheindependent (109studies)orthedependentvariable(18studies).

A positive correlation was observed in those studies in which the instrumental theory of interest groups was

(6)

usedtosolvethe problemof misalignment. Based onthe work of Akpinar et al. (2008), and taking as reference the KEJI Index,² Choi et al. (2010) analyzed a sample of 1222Koreancompaniesbetween theyears2002 and2008 in order to perform their statistical analysis --- using two typesofindicestomeasurethesocialbehaviorofcorpora- tions:theycalculatedﬁrst,anequalweightedresponsibility index(‘‘Equal-weightedCSRIndex’’),andsecond,anindex weightedaccordingtotheimportanceoftheinterestgroups (‘‘Stakeholder-weightedCSRIndex’’).Theoutcomeshowed apositive relationship between FPand thesecond index, illustrating that when companies focus their CSR policies towardthoseinterestgroupsthatholdgreaterimportance forthecompany,ﬁnancialresultsimprove.

McGuire et al. (1988) introduced a time lag factor to furtherinvestigate therelationshipbetweenfinancialpro- fitability,over severalyears,andsocial behavior.Using as ameasure ofCSR a Fortunemagazine corporatebehavior index,they concludedthat CSR showed a higher correla- tionwithfinancialresultsofpreviousyears.CSRvs.financial resultsofsubsequentyearsdisplayedalowercorrelation.

Ofnote, Schuler andCording(2006), while recognizing the profound importance of the Corporate Social Perfor- mance(CSP)andcorporateFPlinkage,alerttotheunclear natureoftherelationship.Theysuggestthat(1)empirical shortcomingsmaydistort theCSP-FPrelationship, and(2) largedeﬁcienciesexistinthetheoreticalmodelsused(most assumeadirectlinkbetweenCSPandFP).However,indeﬁn- ingtheCSP-FPlinkanditsconstructsingreaterdetail,they advancehowCSPleadsFP.

Recently, Callan and Thomas (2009) respond to these issuesinan updatedstudy ofthis relationshipby examin- ingtwodifferentapproachestomeasuringCSR,controlling forkeyvariablesidentiﬁedintheliterature,andtestingfor thenon-linearityofcertainvariables.Theirmainconclusion assertsthatapositiveCSR---FPrelationshipexists.

Themainconclusions drawnafterthisextensivereview oftheexistingliteraturesupportstheselectionoftheCSR andFP variables usedto build the modelsof theempiri- calstudy.Inaccordancewiththeexposedstudiesoutlined above,andﬁllingthegapintheexistingSpanishliterature wherethebi-directionalCSR---FPrelationshiphasnotbeen previouslymeasured,andtakingintoaccountthatallcom- paniesmustsatisfyCSR,therebyaccountingtosocietyasa wholeassuggestedbystakeholdertheory,wewouldexpect greaterCSRtobepositivelyrelatedwithhigherlevelsofFP andviceversa.

Therefore,weproposethefollowinghypotheses:

Hypothesis1. CompaniesdisplayinggreaterCSRbehavior achievehigherﬁnancialproﬁtability.

Hypothesis 2. The most proﬁtable companies are those thatadoptsuperiorCSRbehavior.

Totestthetwohypothesesweproposetwocomplemen- tarymodels:thus, wetest thebi-directionalrelationships

2KEJI(KoreaEconomicJusticeInstituteIndex)referstoaKorean indexdevelopedbyoneoftheleadingNGO’sinthecountry.

Corporate governance

Corporate social responsibility

Financial performance Good corporate governance

recommendations

Model 1

Model 2

Figure2 Proposedconceptualmodel.

between CSR andFP;and,seektoclarify the cruxof the relationship.

Proposed model and formulation of hypotheses

Followingtheliteraturereview,weformulateaconceptual basemodel comprisedoftheinterdependentrelationships between the parameters previously outlined: CG, from which concrete actions in CSR arederived andgiven lev- elsofFParedetermined;andthemutualinterdependencies betweenCSRandFP.Afurtherelementofourmodelconsists ofcompliancewiththerecommendationsofGoodCorporate Governance,ataskoftheboardofdirectors.Seegraphical representationinFig.2.

Giventhisgenericconceptualmodelwewillderivesubset models1and2,andfocusontestingstatisticallytherela- tionshipbetween CSR andFP.Basing ourresearch,onthe onehandonStakeholdertheory,asdonebyMcGuireetal.

(1988)andKaragiorgos(2010),andontheotherhandonthe instrumentalapproach,followingAkpinaretal.(2008),Choi etal.(2010)andHarjotoandJo(2011),weexpecttoﬁnd apositiverelationshipbetweenCSRandFP---inagreement withthepublishedstudiesbyMargolisandWalsh(2003),Choi etal.(2010)andKaragiorgos(2010).

TheequationforModel1isasfollows:

FP=c+b1.GRI+b2.DJSI+b3.COMPLRECOM+b4.GC +b5.LNASSET+D

Thedependentvariable,FP,asusedbyotherauthorsin theirstudies(Guest,2009;JacklingandJohl,2009;Crespí, 2010),isrepresentedby:ROA(ReturnonAssets),equalto operatingproﬁtbeforedepreciationandprovisionsdivided bytotalassets;ROE(Returnonequity),theoperatingproﬁt beforedepreciationandprovisionsdividedbystockholders equity;andTobin’sQ,themarketvalueofthesharedivided byitsbookvalue(PerfectandWiles,1994).

(7)

Table1 VariablesusedinModel1ofthestudy.

Socialvariables (CorporateSocial Responsibilityused asindependent variables)

Financialvariables (Financial Performanceused asdependent variables)

Control variable

GRI ROA LNASSET

DJSI ROE

COMPLRECOM QTOBIN

GC

IndependentvariablesGRI,DJSI,COMPLRECOM,andGC, deﬁne the dimension of CSR. LNASSET is used as control variable in agreement withthe practice of other authors (McWilliamsandSiegel,2000;Choietal.,2010;Harjotoand Jo,2011)whichincludecompanysizebytakingthenatural logarithmoftheassets(seeTable1).

GRI(GlobalReportInitiative) indicatesthevaluationof thesustainabilityreportaccordingtoguideG-3.VariableGRI isincludedbecause(1)ofitsuniversalcharacter,and(2)it allowstheclassiﬁcationofcompanies.Itswidespreadusein theEuropeanUnionandincountriesoftheOECDmakesit arigorous and comprehensiveindicator (Hedberg andVon Malmborg,2003;Frias-Aceitunoetal.,2013;Legendreand Coderre,2013).

DJSI (Dow Jones Sustainability Index) indicates if the companybelongstotheDJSI.Itsuniversalityandreputation amongstthesustainabilityindicesmake itagoodvariable formeasuringCSR.

COMPL RECOM (ComplianceoftheRecommendationsof GoodCorporateGovernance)indicatestheextentofcompli- ance withthe recommendations of theUCGG. It givesan ideaoftheimportancethatcompaniesgranttooneoftheir interest groups, theshareholders (58 recommendationsin totalcan beseen in the Fsection of thereports of Good CorporateGovernanceofthelistedSpanishcompanies).

GC(GlobalCompact)indicatesifcompanieshavesigned theGlobalCompact.Thisvariablehasbeenselectedasone oftheCSR measuresbecausetheforum ofCSRexpertsof theMinistryofLabor³hasembracedit,togetherwithothers suchasGRI,assigniﬁcant.

Consequently,thefourselectedsocialvariablesarerep- resentative ofthe commitmentthatcompanieshave with society,theenvironment,andwithitsinterestgroupsingen- eral.Additionally,byincludingvariableCOMPLRECOM,we havestressedthespecialimportanceallstakeholdershold for the companies. Other rules andindices that exist for evaluatingCSRhavenotbeenconsidered,eitherbecauseof theinclusionofarelatedparameter,itslimitedacceptance, and/orthedifﬁcultyinﬁndingthecorrespondingdata.

FromModel1wederivethefollowingsub-hypotheses:

H1.1 ((GRI)). Companies obtaining a higher ratingin the GRIindexachievebetterﬁnancialresults.

3TheforumofCSRexperts,foundedin2005,consistedofrep- resentatives from the Public Administrations, Civil Society, and Universities.

H1.2((DJSI)). CompaniesincludedintheDJSIachievebet- terﬁnancialresults.

H1.3((COMPLRECOM)). Thegreaterthecompliancewith therecommendations ofgoodCG,thebettertheﬁnancial results.

H1.4 ((GC)). Companies that sign the Global Compact achievebetterﬁnancialresults.

Model 2, in agreement with the possible bidirectional natureoftherelationshipproposedbyanumberofauthors (McGuireetal.,1988;MargolisandWalsh,2003),questions whether the superior ﬁnancial returns of Spanish companies have an impact on their social behavior. Under this approach,thereversemodelaccountsfortheinherentdif- ﬁcultyofstatisticalmodelsthatincludeCGvariables;that is,weattemptavoidingthepossibleendogeneitybetween the dependent and independent variables (Shleifer and Wolfenzon,2002;Adamsetal.,2010).

The studies carriedout in other countries,which have servedasastandardforthiswork,havegenerallyusedCSR indicesgatheredbyindependentagenciesforevaluatingthe socialbehaviorofthecompanies.Choietal.(2010)develop their own CSR index, based on the KEJI ﬁnancial index.

ThestudyofMcGuireetal.(1988)isbasedontheranking compiledbyFortunemagazine(Fortunereputationrating).

Other works,such as(Margolis and Walsh, 2003; Gallego, 2006), are based on reports of environmental emissions, environmentalpractices,orsocialactionscarriedoutbythe companiesthemselves.Donkeretal.(2008)alsostudythe relationshipbetweenfinancialreturnandcorporateethics inlistedCanadiancompanies.HarjotoandJo(2011)devel- opedaproprietaryindexviatheassessmentoffivevariables thatcorrespondedtothecommunity,theenvironment,the products,thediversity,andemployment.Hence,webelieve thatourproposedindexisbothjustifiedandrepresentative ofthesocial,environmentalandethicalcomponentsofthe Spanishcompanies.

Inourcase theindex,similartothatusedbyHongand Andersen(2011),includes fourcomponents:the presenta- tionofcompanysustainabilityreportsaccordingtotheGRI model,companyinclusionintheDJSI,companycompliance withtherecommendations ofgoodCG,andthesigning of theGC.

Formodel2,weproposethefollowingequation:

INDEX=c+b1.FP+b2.LNASSET+D Wetestthefollowinghypothesis:

H2 ((ROE/ROA/QTOBIN)). Better ﬁnancial results lead to betterbehaviorinCSR.

The independent variable FP, ﬁnancial performance, takes on the values of the ratios ROA, ROE, and Tobin’s Q; and, the dependent variable, INDEX as proposed by theauthors,representsacompoundvalueconsistingofan equally weighted sum of the values of the four selected variables (GRI, DJSI, COMPLRECOM and GC). This index measurestheﬁrm’sCSRbehavior,asperformedinthework ofBeluandManescu(2013).LNASSET,whichmeasuresthe

(8)

Table2 VariablesusedinModel2ofthestudy.

Socialvariable (CorporateSocial Responsibility Indexusedas dependent variable)

Financialvariables (Financial Performanceused asindependent variables)

Control variable

INDEX ROA,ROE,QTOBIN LNASSET

sizeofthecompanies,hasbeenincludedasacontrolvari- able---similartoModel1(seeTable2).

Empirical study

Methodology,sampleanddatacollection

For the empirical analysis we formulate six multivariate regressionmodels:threeforModel1andanotherthreefor model 2. To carry out the statistical analysis we ran the econometricsoftware,Eviews5.0,widelyusedinempirical research.

ThesampleiscomposedofSpanishcompanieslistedon theMadridStockExchange intheyear2009.Theinforma- tionwasobtainedfromthewebpagesofTheMadridStock Exchange(http://www.bolsamadrid.es)⁴andTheSecurities Market National Commission(http://www.cnmv.es).⁵ Data fromBanks, Savings Banks andFinancial Institutionswere excludedbecausetheiraccountingsystemdiffersfromthat used by the majority of the companies; this would have caused a lack of homogeneity in the calculation of the financial ratios; additionally, other authors (Jackling and Johl, 2009) that have performed previous studies on the corporate-financiallinkhavedonelikewise.Thedataofthe financialvariableswereobtained fromthe SABI⁶ database andcheckedwiththeAMADEUS⁷database.Thefinalsample consistsofdatafrom121companies(seeAnnex1).

To assess a company’s social dimension the following aspectshavebeentakenintoaccount:

GRI:InaccordancewiththeGRIindexvaluation,wepro- posethefollowingnumericalrating:A+:1;A:0.9;B+:0.8;

B:0.7;C+:0.6;C:0.5;and,ifnoGRIIndex:0.Theinforma- tionwasgatheredfromtheGRIwebsite:http://database.

globalreporting.org.

DJSI:Value1ifthecompanybelongstotheDJSI,and0 otherwise.

COMPLRECOM: calculated by dividing the number of satisﬁedrecommendations bythetotalnumberof applied

4TheSpanishMarketsandStockmarketsCorp.(BME)integrates thedifferentcompaniesoverseeingandmanagingtheSpanishstock exchangesandﬁnancialsystem.

5TheSecuritiesMarketNationalCommission(CNMV)istheregu- latorinchargeofsupervisingandinspectingallactivityonthestock exchanges.

6SABI (IberianBalances Analysis System): database containing generalandﬁnancialinformationonSpanishcompanies.

7 Amadeus: database containing ﬁnancial information on both publicandprivateEuropeancompanies.

recommendations (58 in 2009). We suggest assigning the value1totherecommendationsmet,0.5tothepartlymet, and0pointstotheexplainedbutunmet.Thedatahavebeen obtainedfromparagraphF(degreeoffollow-uptotherec- ommendationsofCG)ofTheAnnualReportsonCGforthe companiesinthesample(year2009)inaccordancewiththe UCGG(2006).

GC:Takesvalue1ifthecompanyhassignedtheGlobal Compact, and 0 otherwise. The data proceed from The Global Compact Network, Spain webpage: http://www.

pactomundial.org.

Theindexvariable,employedinModel2,wascalculated usingtheequallyweightedsumofthefourvariables,asper- formedbyBeluandManescu(2013);allwereassignedthe sameweight(0.25)thatwasusedtoassessCSR.

Analysisandresults

Thestatisticalanalysisperformedontheinitialsamplepro- duced novalid result.The correlationsindicated verylow values,andnoneofthemodel’sproposedequationsoffered an explanation; hence, we decidedtoreduce thesample sizeandtestwithaspeciﬁcgroupofcompanies.

Weretestedby selecting asample group ofcompanies meetingspeciﬁccriteria;wediscoveredthatcompaniesdis- playingasustainabilityreportaccordingtoGRIdidpresent explanatorymodels.Byﬁlteringthosecompaniesthatmet thecriterion,the samplesizewasreducedto107compa- nies.

The descriptive statistics of the selected sample is includedinTable3.

Compliance with the recommendations of CG varies betweenamaximumvalueof100%andaminimumof73.6%, with amean of 89.3%. Only 6 companiescomply with all therecommendations.51.6percentofthecompaniesare includedintheDJSIand80%havesignedtheGC.Inregards totheGRIreport,itsmeanequals0.94.Finally,theIndex variablehasa minimumvalue ofa 35.9% andamaximum valueof99.6%,withthemeanat78.9%.Themaximumvalue, 99.6%,correspondstoasinglecompanythatsignedtheGC, followedtheGRItodraftthesustainabilityreport,belonged totheDJSI,andfulﬁlled98percentoftheUCGGrecommen- dations.

Asfortheﬁnancialvariables,themaximumandminimum valuesofROEare35.8and−105.40;forROA,theyare16.4 and−10.2;and,forTobin’sQtheseare5.89and0.69.The respectivemeanvaluesare:4.7,2.04,and1.85.

In Table 4, correlations between variables, we draw attentiontothehighcorrelationvaluesforGlobalCompact (0.4832),ROE(0.5072),andROA(0.4170);allinrelationto thecompliancewithGoodCorporateGovernancevariable.

Next, we analyzed whether the equation advanced in Model1wasexplanatoryforanyofthethreeproposedﬁnan- tialratios(seeTable5).

For the variable ROE, the best model included variables GRI and COMPLRECOM. The model was signiﬁcant (Prob. F<0.05) and explanatory; and, explained 43%

(Adj. R-squared 0.43)of the behavior of theindependent variable. Variables GRI and COMPLRECOM were signiﬁ- cant aswell (Prob. t<0.005). The model didnot present

(9)

Table3 Descriptivestatistics.

COMPLRECOM DJSI GRI GC INDEX ROE ROA QTOBIN

Mean 0.8934 0.5161 0.9387 0.8065 0.7887 4.7065 2.0419 1.8574

Median 0.8981 1.0000 1.0000 1.0000 0.9049 10.8000 2.2000 1.3000

Maximum 1.0000 1.0000 1.0000 1.0000 0.9955 35.8000 16.4000 5.8900

Minimum 0.7358 0.0000 0.5000 0.0000 0.3590 −105.4000 −10.2000 0.6900

St.deviation 0.6570 0.5080 0.1256 0.4016 0.2182 1.3534 27.4496 5.0058

Table4 Correlationsbetweenvariables.

COMPLRECOM DJSI GRI GC INDEX LNASSET ROE ROA QTOBIN

COMPLRECOM 1.0000

DJSI 0.1914 1.0000

GRI 0.1951 0.3555 1.0000

GC 0.4832 0.5060 0.2856 1.0000

INDEX 0.4371 0.8805 0.4970 0.8322 1.0000

LNASSET -0.0701 0.4952 0.2648 0.4594 0.5325 1.0000

ROE 0.5072 0.3605 0.5259 0.5090 0.5580 0.2243 1.0000

ROA 0.4170 0.3936 0.4245 0.3275 0.4724 0.1441 0.8211 1.0000

QTOBIN 0.1854 0.2303 0.2912 0.0157 0.1972 -0.1110 0.4552 0.6234 1.0000

problemsofmulticollinearity,⁸norheteroscedasticity⁹;the autocorrelation¹⁰ problem was corrected by including the termAR(1)(followingtheCochrane-Orcuttmodel).

SeveralestimateswerecarriedoutforvariableROA;and, aswithvariableROE,thebestresultswereofferedbythe modelthatincludedvariablesGRIandCOMPLRECOM.There were autocorrelation problems, however theywere over- come.Thetwovariableswerealsosigniﬁcant,butthefull model explained only 29.8% of the result,somewhat less thanthatobtainedforROE.

NosatisfactoryresultswereobtainedforTobin’sQvari- able.

IntheModel2analysis(seeTable6),themodelsforROE (Adj.R-squared50.2%),andforROA(Adj.R-squared44.8%) are signiﬁcant and explanatory (Prob. F<0.05); however thiswasnotthecasewhenconsideringTobins’Qvariable.

Autocorrelation problems were detected under both ROA and ROE; these were corrected by including variable AR (1),afterobservingwiththeBreusch-Godfreytestthatthe autocorrelation wasof first order.In both cases, the cor- rectioncanbe verifiedby checkingthedatain thetables (the introduced variableAR (1) is notsignificant asProb.

t<0.05).

Endogeneity hasbeen resolvedin both models1and 2 throughtheapplicationofHausman’stest(Hausman,1978;

HausmanandTaylor,1981).

8Multicollinearityistheexistenceofalinearrelationshipbetween theindependentvariables.

9Heteroscedasticityreferstotheunequalvariancebetweenthe variables.

10 AutocorrelationproblemswereresolvedbycontrastingROEand ROAdatafrom2008.

The hypothesesinitiallyposedareconfirmedaftercar- ryingout the statistical analysis onthose companiesthat met the requirement --- having drafted out the sustaina- bilityreport--- accordingtotheGRIguide.Thus,according to Model 1, hypothesis H1.1 and H1.3 hold for variables ROEandROA,andwe canaffirm that(1)thosecompanies obtaining better rating on sustainability according to the GRI Sustainability Reporting Guidelines obtained superior financialresults; and(2)thosecompaniesenjoying higher percentages of compliance with the recommendations of UCGGpresentbetterfinancialresults,asmeasuredbyROE andROA.

According to the results of Model 2, we can also con- ﬁrmthesecondhypothesisforvariablesROAandROE;and, canstatethatthosecompaniesachievingthebestresultsin termsofROAandROEobtaingreaterSocialBehaviourIndex valuesand,asaconsequence,aredistinguishedforadopting betteroverallCSRpolicies.

In none of the models did we ﬁnd a correlation for Tobin’s Q ratio, as exhibited in Table 7. In this respect, McGuireetal.(1988)hadadvisedofusingaccountingratios, especially ROA, ratherthan market or risk ratios, for FP variables,deemingthembetterpredictors.Wecanafﬁrm, asother authors (McGuire etal.,1988; CharloMolinaand MoyaClemente,2010;Choietal.,2010;Karagiorgos,2010;

HarjotoandJo,2011)havedone,thattheexpectedconclu- sion---positivesign inthestudiedrelationships-hasbeen reached.

Themainresults(asshowninTable7)are:

Model1 (direct relation):greater values for GRIand for COMPLRECOMleadtogreatervaluesofROAandROE.

Model2(inverserelation):greatervaluesforROAandROE implygreatervalueofSocialBehaviourIndex.

(10)

Table5 Model1results.

Variables ResultsoftheModel1forROE ResultsofModel1forROA

Coefﬁcient Std.error T-statistic Coefﬁcient Std.error T-statistic COMPLRECOM 229.307(0.0007)^*** 59.036 3.884 35.392(0.0073)^** 12.129 2.918

GRI 93.426(0.0030)^*** 28.440 3.284 14.634(0.0205)^** 5.917 2.473

LNASSET 5.940(0.0357)^** 2.678 2.219 0.702(0.2153) 0.552 1.271

C −342.940(0.0000) 62.890 −5.453 −45.591(0.0007) 12.795 −3.876

AR(1) 0.394(0.0505) 0.192 2.054 0.389(0.0550) 0.193 2.013

R-squared 0.5097 0.3954

AdjR-squared 0.4313 0.2986

Durbin---Watsonstatistic 2.1400 1.8504

F-statistic 6.4980 4.0866

Prob(F-Sta.) 0.000995 0.011033

Numberofﬁrms 107

*p≤0.1.

**p≤0.05.

*** p≤0.005.

Table6 Model2results.

Variables ResultsofModel2toindexandROE ResultsofModel2toindexandROA Coefﬁcient Std.error T-statistic Coefﬁcient Std.error T-statistic

ROE 0.003(0.0035)^*** 0.001 3.210 0.016(0.0155)^** 0.006 2.590

LNASSET 0.0580(0.109) 0.021 2.743 0.066(0.0050)^*** 0.022 3.066

C 0.247(0.2249) 0.199 1.243 0.147(0.4751) 0.203 0.725

AR(1) 0.343(0.0852) 0.192 1.789 0.341(0.0918) 0.195 1.751

R-squared 0.5539 0.5047

AdjR-squared 0.5024 0.4476

Durbin---Watsonstatistic 1.7600 1.8170

F-statistic 10.7605 8.8318

Prob(F-Sta) 0.000089 0.000332

Numberofﬁrms 107

*p≤0.1.

**p≤0.05.

*** p≤0.005.

Table7 Veriﬁcationofmodels’hypotheses.

Studyresults Model1 Model2

Hypothesis1.1(+GRIimplies+FP) VeriﬁedwithROAandROE Hypothesis1.2(+DJSIimplies+FP) Notveriﬁed

Hypothesis1.3(+COMPLRECOMimplies+FP) VeriﬁedwithROAandROE Hypothesis1.4(+GCimplies+FP) Notveriﬁed

Hypothesis2(+FPimplies+SocialBehaviourIndex) VeriﬁedwithROAandROE

Inspiteof thesepositiveresultswe acknowledgethere arecertainlimitationsinourresearch.First,thelimitedgeo- graphicalandtemporalscope.Second,asaconsequenceof the bidirectionalrelationship, a feedback circle is gener- ated between social and ﬁnancial variables,and it is not altogetherclearwhichcomeﬁrst.

Conclusions and future directions for research

This studyhasdemonstrated,asexpected,thatthe social isprofitable,andtheprofitableissocial,therebyforminga virtuouscircleassuggestedbySurrocaetal.(2010).Thatis, sociallyresponsiblepoliciestransformintohigherprofitsand

(11)

higher proﬁtstransform into socially responsible policies.

ThisbidirectionalrelationshipinCSR---FPhasprovenpositive inbothdirections.Hence,ineconomicterms,weaffirmthat for firms, ceteris paribus,increasing CSR outlays leads to animprovedFP,andwithal,firmsenjoyinggreaterfinancial strengthpresent animproved SocialBehaviourIndex.This generatespositivemutualfeedbackthatencouragesfirmsto (1)applyCSRpolicieswiththeirfinancialresources,and(2) verifyhowtheirCSRinvestmentsleadtoimprovedfinancial returns.

Furthermore, fromatheoretical pointof view, the lit- erature revision has introduced distinct theories: agency, stewardship, resources dependency and institutional perspective; in certain postulates they may be applied to themodelsofouranalysis.However,ourempiricalresults confirm that stakeholder theory provides the most solid foundationforthecompletestudy.TheCSR-FPbidirectional relationship demonstrates that all interest groups derive benefitsinsomewayoranother.Inaddition,thenormative andinstrumentalapproaches of thistheoryarealso satis- fiedbytheoutcome ofthiswork;theformer becausethe recommendationsofGoodCorporateGovernanceasanele- mentof CSR undoubtedlygeneratebenefits; thelatterby confirmingandconsolidatingtherelativeimportancegiven bythecompaniestothedifferentareasofCSR,afunction ofthepriorityassignedtoeachoftheinterestgroups.

Webelievethecontributionofthisworktobeimportant forthefollowingreasons:ﬁrst,becauseoftheoriginalcom- binationofvariablestoassessCSR,basedonthevaluation ofGRI,inclusionintheDJSIindex(DowJonesSustainability Index),compliancewiththerecommendationsofgoodCG, andwhetherthecompanyisasigneeoftheGlobalCompact;

second,becausethevariablereferringtoCG,albeitenjoying widespread backing, had never been used in an empiricalstudy inSpain; andﬁnally, becausewe performed the directandthereverseanalysis,posingthequestioninterms of whethertheﬁnancial resultsobtained bySpanish companies have an impact on their social behavior and vice versa.

An importantand novelconclusion inthis Spanishcase relates to the CG variable as no study had previously includedthecompliancewithgoodgovernmentasacompo- nentelementinmeasuringcompanyCSR.Fromthepositive correlationexistingbetweenthisvariableandtheratiosROE andROAwemayconcludethattheexistingmechanismsthat regulatetheoperationoftheboardsofdirectorscanserveas aguidetoshareholdersandinvestorsinmakingtheirinvest- ment decisions;likewise, sociallyresponsible investments will be bolstered by the ethical and respectful behaviors conformingtotheprinciplesoutlinedbyUCGG.

Although no single internationally implemented legal mechanism exists to regulate the behavior of company boards of directors, the Spanish UCGG possesses a large number of concordances between its recommendations and both the OECD’s Principles of CG, and other recommendations of the European Commission. Even if some recommendations are presently being challenged, as for examplethatthesameindividualactascompanyCEOwhile simultaneouslyholdingthepositionofchairmanoftheboard ofdirectors,weunderstandthatthisstudyinaddingtothe knowledgebaseattheinternationallevelfacilitatesfurther research---whentakingasvariableCG---ofCSRandFP.

In light of our results, company observance with the guidelineof theGRIis highlyrecommended.Applying the principlesoftheGRIwhenpresentingsustainabilityreports assuresﬁnancialadvantages,andbyextrapolationallstake- holdersstandtobeneﬁt.

From an academic standpoint, we would like to draw attentiontotheimportanceforcorporationstocomplywith therecommendationsofgoodCGandtheGRIguideline.We stresstherelevancefor theboardofdirectorstoconsider thatsocialpoliciesformanintegralpartofoverallcompany strategy.Boardsshouldfeelemboldenedtoadoptsuchpoli- ciesastheyfeedbackintoincreasedproﬁtabilityandboost overallcorporatevisibility.

Future research should aim to resolve which of the two variables, CSR or financial profitability, initiates the CSReconomicprofitabilitycycle,ormorespecifically,clarify if profitable companies invest in CSR and in so doing gainanadditionaladvantagethattranslatesintoimproved economic returns, or on thecontrary, companies exercis- ing sociallyresponsible behaviorsobtain better economic returns. As business strategies focus over the long-term, researchshouldincludea datasamplecomprisingat least fouryears.

A furtheraim, a consequence of the uniformity evolv- ingfromincreasedawarenessinthesocialresponsibilityof ﬁrms,shouldconsistincompilingatrueandrepresentative indexto measure CSR. The complexity involved in calcu- latingCSR complicates comparisonof thestudies; thus, a universalindexrepresentingsocialbehaviorwould helpin thehomogenizing of future empiricalanalyses. Moreover, and in agreement with the authors who have served as benchmarkinourdevelopmentofthetheoreticalapproach tothis research, we would like to emphasize that stake- holdertheoryas wellas itsinstrumentalapproachshould playasigniﬁcantroleintheformulationofthisindex.

Finally,thisstudyhasprovedthatpositivefinancialratios leadtoimprovedsocialbehavior;theinverserelationship, businesseswithamoresocialfocusobtainbetterfinancial results,alsoholds.Thismutualfeedbackbetweenthesocial andfinancialaspectsiskeytoanethicalbusinessbehavior.

Theboard of directors,infollowing therecommendations ofgoodCG,representsthestartingpointfortheapplication of CSR tobusiness decision-making as it strives toenrich societyasawhole.

Annex 1. Sample of listed companies

Abengoa,S.A.

AbertisInfraestructuras,S.A.

Acciona,S.A.

Acerinox,S.A.

Acs,ActividadesdeConst.yServiciosS.A.

AdolfoDominguez,S.A.

AﬁrmaGrupoInmobiliario,S.A.

Agrofruse,AgricoladeF.Secos,S.A.(AGF) AhorroFamiliar,S.A.(AHF)

Almirall,S.A.

Amper,S.A.(AMP)

Antena3deTelevision,S.A.

AycoGrupoInmobiliario,S.A.(AYC) AzkoyenS.A.

(12)

BarondeLey,S.A.

Befesa,MedioAmbiente,S.A.

BodegasRiojanas,S.A.

BolsasyMercadosEspa˜noles,Shmsf,S.A.

CampofrioFoodGroup,S.A.

CarteraIndustrialRea,S.A.

CementosPortlandValderrivas,S.A.

Cia.deInversionesCinsaS.A.

Cia.Espa˜noladePetroleos,S.A.

Cia.Levantina,EdiﬁcaciondeO.Publicas Cia.VinicoladelNortedeEspa˜na CieAutomotive,S.A.

ClinicaBaviera,S.A.

Codere,S.A.

Const.yAuxiliardeFerrocarrilesS.A.

CorporacionDermoestetica,S.A.

CorporacionFinancieraAlba,S.A.

CriteriaCaixacorp,S.A.

DinamiaCapitalPrivado,S.A.

DogiInternationalFabrics,S.A.

DuroFelguera,S.A.

EbroPuleva,S.A.

Elecnor,S.A.

Enagas,S.A.

Endesa,SociedadAnonima ErcrosS.A.

Espa˜noladelZinc,S.A.

FaesFarma,S.A.

Ferrovial,S.A.

FersaEnergiasRenovables,S.A.

Fluidra,S.A.

FomentodeConstr.yContratasS.A.

Funespa˜na,S.A.

GamesaCorporacionTecnologica,S.A.

GasNaturalSdg,S.A.

Gral.deAlquilerdeMaquinaria,S.A.

Grifols,S.A.

GrupoCatalanadeOccidenteS.A.

GrupoEmpresarialEnce,S.A.

GrupoEmpresarialSanJose,S.A.

GrupoTavex,S.A.

IberdrolaRenovables,S.A.

Iberdrola,S.A.

Iberia,LineasAereasdeEspa˜na,S.A.

IberpapelGestion,S.A.

Inbesos,S.A./Nyesavalorescorporacion,SA IndoInternacionalS.A.

IndraSistemas,S.A.,SerieA

IndustriadeDise˜noTextil,S.A.‘‘inditex’’

InmobiliariaColonial,S.A.

InmobiliariadelSur,S.A.

Inmolevante,S.A.(ILV) Inverﬁatc,S.A.(FIA)

InypsaInformesyProyectos,S.A.(INY) Jazztel,P.L.C.(JAZ)

LaboratoriosFarmaceuticosRovi,S.A.(ROVI) LingotesEspecialesS.A.

LiweEspa˜nola,S.A.(LIW) Mapfre,S.A.(MAP) Martinsa-fadesa,S.A.

MetrovacesaS.A.

MiquelyCostas&Miquel,S.A.

Montebalito,S.A.

NatraS.A.

Natraceutical,S.A.

NhHoteles,S.A.

NicolasCorreaS.A.

ObrasconHuarteLain,S.A.

PapelesyCartonesdeEuropa,S.A.(EUROPAC,PAC?) Pescanova,S.A.(PVA)

Prim,S.A.(PRM)

PromotoradeInformaciones,S.A.(PRISA,PRS) ProsegurS.A.,Cia.deSeguridad(PSG) PulevaBiotech,S.A.(BIO)

RealiaBusiness,S.A.(RLIA)

RedElectricaCorporacion,S.A.(REE) Renta4ServiciosdeInversion,S.A.(R4) RentaCorporacionRealEstate,S.A.(REN) RepsolYpf,S.A.(REP)

ReyalUrbis,S.A.(REY) Rusticas,S.A.(RTC)

SacyrVallehermoso,S.A.(SYV) SedadeBarcelona,S.A.(la)(SED) ServicePointSolutions,S.A.(SPS) Sniace(SNC)

SolMelia,S.A.(SOL)

SolariaEnergiayMedioAmbiente,S.A.(SLR) SosCorporacionAlimentaria,S.A.(SOS) SotograndeS.A.(STG)

TecnicasReunidas,S.A.(TRE)

Tecnocom,TelecomunicacionesyEnergia,S.(TEC) Telefonica,S.A.(TEF)

TestaInmueblesenRenta,S.A.(TST) Tubacex,S.A.(TUB)

TubosReunidos,S.A.(TRG)

UnionEuropeadeInversiones,S.A.(UEI) Unipapel,S.A.(UPL)

Uralita,S.A.(URA)

UrbarIngenieros,S.A.(UIN) UrbasGuadahermosa,S.A.(UBS)

VerticeTrescientosSesentaGrados,S.A.(VER) VidralaS.A.(VID)

Viscofan,S.A.(VIS) Vocento,S.A.(VOC) VuelingAirlines,S.A.(VLG) ZardoyaOtis,S.A.(ZOT) Zeltia,S.A.(ZEL)

References

Adams, R.B., Hermalin, B.E., Weisbach, M.S., 2010. The role ofboardsofdirectorsin corporategovernance:a conceptual frameworkandsurvey.J.Econ.Lit.48(1),58---107.

Akpinar,A.,Jiang,Y.,Gómez-Mejía,L.R.,Berrone,P.,Walls,J.L., 2008.StrategicuseofCSRasasignalforgoodmanagement,IE BusinessSchoolWorkingPaperWP,08-25.

AlbaredaVivó,L.,BalaguerFranch,M.R.,2009.Thechallengesof socially responsibleinvestmentamong institutional investors:

exploringthelinks betweencorporatepensionfundsandcor- porategovernance.Bus.Soc.Rev.114(1),31---57.

Albareda, L., 2013. CSR governance innovation: standard competition-collaboration dynamic. Corp. Gov. 13 (5), 551---568.