Branco, Manuel Castelo dan Lúcia Lima Rodriguez. 2008

7/27/2019 Branco, Manuel Castelo dan Lcia Lima Rodriguez. 2008.

1/18

Factors Influencing Social Responsibility Disclosure by Portuguese CompaniesAuthor(s): Manuel Castelo Branco and Lcia Lima RodriguesReviewed work(s):Source: Journal of Business Ethics, Vol. 83, No. 4 (Dec., 2008), pp. 685-701Published by: SpringerStable URL: http://www.jstor.org/stable/25482407 .

Accessed: 02/03/2013 09:21

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of

content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms

of scholarship. For more information about JSTOR, please contact [email protected].

.

Springeris collaborating with JSTOR to digitize, preserve and extend access toJournal of Business Ethics.

http://www.jstor.org
http://www.jstor.org/action/showPublisher?publisherCode=springerhttp://www.jstor.org/stable/25482407?origin=JSTOR-pdfhttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/stable/25482407?origin=JSTOR-pdfhttp://www.jstor.org/action/showPublisher?publisherCode=springer


2/18

Journal of Business Ethics (2008) 83:685-701DOI 10.1007/sl0551-007-9658-z? Springer 2008

Factors Influencing Social ResponsibilityDisclosure by Portuguese Companies

Manuel Castelo BrancoLucia Lima Rodrigues

ABSTRACT. This study compares the Internet (corporate web pages) and annual reports as media of socialresponsibiHty disclosure (SRD) and analyses what influences disclosure. It examines SRD on the Internet byPortuguese Hsted companies in 2004 and compares theInternet and 2003 annual reports as disclosure media. Theresults are interpreted through the lens of a multi-theoretical framework. According to the framework adopted,companies disclose social responsibiHty information topresent a sociaUy responsible image so that they canlegitimise their behaviours to their stakeholder groups andinfluence the external perception of reputation. Resultssuggest that a theoretical framework combining legiti

macy theory and a resource-based perspective provides anexplanatory basis for SRD by Portuguese listedcompanies.

KEY WORDS: annual reports, internet, legitimacytheory, resource-based perspectives, social responsibilitydisclosure, Portugal

Introduction

Most of the empirical studies on social responsibilitydisclosure (SRD) have focused on the annual report,which is considered to be the most important toolused by companies to communicate with theirstakeholders (see, for example, Gray et al., 1995b;Neu et al., 1998). However, the Internet has becomean important medium through which companies candisclose information of different natures, and thussome recent studies have been made analysing com

panies' web pages as a SRD medium (see, forexample, Frost et al., 2005; Patten, 2002a; Patten andCrampton, 2004; WiUiams and Pei, 1999). Exploration of companies' web pages as a SRD medium isnow as essential as the exploration of annual reports tounderstand SRD disclosure practices.

The purpose of this study is to understand SRD,both on the Internet and in annual reports, bydeveloping and testing a series of hypotheses. Thenature of SRD in annual reports and on theInternet by a sample of companies with shares listedon the Portuguese Stock Exchange {Euronext

-Lisbon) is analysed. Using content analysis, SRD isclassified in terms of theme (environment, humanresources, products and customers and communityinvolvement).

Companies are considered to engage in corporatesocial responsibility (CSR) activities and disclosurebecause of two different kinds of motivations. Somecompanies expect that having good relations withtheir stakeholders wiU lead to increased financialreturns by assisting in developing valuable intangibleassets (resources and capabilities). These assets canbe sources of competitive advantage because theycan differentiate a company from its competitors.

Other companies engage in CSR activities anddisclosure to conform to stakeholder norms andexpectations about how operations should be conducted, thus constituting mainly a legitimacyinstrument used by a company to demonstrate itsadherence to such norms and expectations. Although some companies engage in CSR activitiesand disclosure because their managers' personalvalues are aligned with CSR values, this aspect wiUnot be explored in this study.

Whereas the first kind of motivations may beexplored through a resource-based perspective analytical lens (see, for example, Branco and Rodrigues,2006; Hasseldine et al., 2005; Toms, 2002), thesecond kind is consistent with social and politicaltheory explanations, in particular legitimacy theory(see, for example, Deegan, 2002; Deephouse and

Carter, 2005; Neu et al., 1998; Patten andCrampton, 2004; Zimmerman and Zeitz, 2002).

This content downloaded on Sat, 2 Mar 2013 09:21:16 AMAll use subject to JSTOR Terms and Conditions
http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp


3/18

686anuel Castelo Branco and Lucia Lima RodriguesThe results are interpreted through the lens of amulti-theoretical framework which combines these

two perspectives, according to which companiesdisclose social responsibiHty information mainly topresent a sociaUy responsible image so that they canlegitimise their behaviours to stakeholder groups andinfluence the external perception of reputation.

Companies with a higher visibility seem to exhibitgreater concern to improve corporate image throughSRD. Results suggest that the framework proposed

may be an explanation of SRD by Portuguese listedcompanies.

This paper examines empirical evidence fromPortugal for two reasons. First, we want to add tothe scarce research on SRD by Portuguese compa

nies by providing new empirical data. Most of thepresent literature is based in Anglo-Saxon countriesand evidence should be added about other geographic, cultural and institutional contexts. In contrast to the understanding of SRD from commonlaw EngHsh-speaking countries (Australia, Canada,

UK, USA), the determinants of SRD in ContinentalEurope, particularly in Portugal, are stiU relativelyunknown. Second, we want to analyse if there arereasons to expect that listed companies in lessdeveloped countries, such as Portugal, wiU behave ina different manner than companies in more developed countries.

According to Lopes and Rodrigues (2007, p. 30),Portugal is one of the least developed countries inthe euro-area and a smaU OECD country. It presentsspecific features regarding its capital market, com

panies' financing structure and corporate governancesystems, providing for a different institutional settingfrom most developed and capital market-orientedcountries, where most of the SRD studies havebeen made. In particular, the degree of familyownership is significant and financing policies arebank oriented.

Notwithstanding the particular characteristics ofPortugal, the results of this study suggest that factorswhich influence SRD practices of Portuguese listedcompanies are not significantly different than those

which influence SRD of companies in moredeveloped countries. This is consistent with theresults of Cormier and Magnan (2003), which leadthem to suggest that the similitude in the way in

which disclosure strategies are determined, irrespective of

agiven country's socio-cultural envi

ronment, is "an iUustration of the strong impact ofglobalised stock markets on fostering convergence incorporate practices" (op. eit., p. 58).In the foUowing section, the theoretical frame

work used is presented. Thereafter foUow sectionson hypotheses development, methodology, resultsand discussion. FinaUy, some conclusions are drawn.

Theoretical framework

Two major influences on companies' SRD areacknowledged in this study: those related to thesocio-political context within which companiesoperate, and those related to economic incentives.The theoretical framework adopted incorporatesboth influences, by adopting institutional theoryperspectives, specificaUy legitimacy theory (see, forexample, Deegan, 2002; Deephouse and Carter,2005; Neu et al., 1998; Patten and Crampton, 2004;Zimmerman and Zeitz, 2002), and resource-basedperspectives (see, for example, Branco and

Rodrigues, 2006; Hasseldine et al., 2005; Toms,2002). Some authors provide important studies in

which similar combinations are attempted (see, forexample, Bansal, 2005).

In this study, companies are considered to engagein some form of stakeholder management, driven bytwo different kinds of motivations. Some companiesbelieve that being seen as sociaUy responsible wiUbring them a competitive advantage, aUowing themto achieve better economic results. They expect thathaving good relations with their stakeholders wiUlead to increased financial returns by assisting indeveloping valuable intangible assets (resources andcapabilities) which can be sources of competitiveadvantage because such assets can differentiate acompany from its competitors. These motivationsare consistent with a resource-based perspectiveanalytical lens.

Other companies engage in CSR activities anddisclosure because of external pressures. They eitherconform to what other companies do, because theybelieve that not doing sowould harm them in termsof their profitability and survival, or respond todiscrediting events, which they believe to be detri

mental to their profitability and survival and must beaddressed tomitigate their effects. CSR activities anddisclosure appear

as mechanisms thesecompanies

use



4/18

Factors Influencing SRD by Portuguese Companies 687

to act and be seen acting within the bounds of whatis considered acceptable according to the expectations of stakeholders on how their operations shouldbe conducted. Social responsibility activities anddisclosure constitute mainly a legitimacy instrumentused by a company to demonstrate its adherence tosuch expectations. These motivations are consistent

with social and poHtical theory explanations, inparticular legitimacy theory.

From a resource-based perspective the benefits ofCSR are, to a great extent, related to their effect oncorporate reputation (Branco and Rodrigues, 2006).

Companies with a good social responsibility reputation are able to improve relations with externalactors such as customers, investors, bankers, suppliersand competitors. They also attract better employeesor increase current employees' motivation and

morale as weU as their commitment and loyalty tothe company, which in turn may improve financial

outcomes. Disclosure of information on a company'sbehaviours and outcomes regarding social responsibility helps to build a positive image with stakeholders.

SRD is particularly important in enhancing theeffects of CSR on corporate reputation. Itmight beconsidered a signal of improved social and environmental conduct and hence reputation in thosefields because disclosure influences the externalperception of reputation. It wiU be difficult forcompanies investing in social responsibility activities,likely to create positive reputation, to realise thevalue of such reputation without making associateddisclosures (Hasseldine et al., 2005; Toms, 2002).

Probably the most important weakness ofresource-based perspectives is related to the lack ofunderstanding they provide on the influence that therelationships between a company and its environ

ment have on the company's success (Branco andRodrigues, 2006). This is why in this study aresource-based perspective is combined with socialand political theories, in particular legitimacy andstakeholder theories. However, these theories areconsidered complementary rather than alternative oropposite (Gray et al., 1995a, p. 52).

The institutional perspective of legitimacy theoryis one of the dominant theories in SRD research(see, for example, Deegan, 2002; Patten and

Crampton, 2004; Neu et al., 1998). The analyticalfocus of legitimacy theory's institutional perspective

is on social legitimacy, which refers to the acceptance of a company by its social environment, andexternal constituents. Companies consider theexpectations of various social constituents in theirbehaviour to achieve social legitimacy. Legitimacy"is a social judgment of appropriateness, acceptance,and desirability" (Zimmerman and Zeitz, 2002,p. 418). The importance of social legitimacy comesfrom the theoretical assumption that companies areembedded in the social environment in which theyoperate, and that their performance and expectationsare affected by the environment. The company'ssuccess, even survival, is determined by thisinterface.

SRD is seen, from such a perspective, as one ofthe strategies used by companies to seek acceptanceand approval of their activities from society. It is seenas an important tool in corporate legitimation strategies. It is used to establish or maintain the legiti

macy of the company because it may influencepublic opinion and public poUcy. Legitimacy theorysuggests that SRD provides an important way ofcommunicating with stakeholders, to convince themthat the company is fulfilling their expectations(even when actual corporate behaviour remains at

variance with some of these expectations).One problem regarding attempts to combine

different bodies of theory to explain organisationalbehaviour is that they are often incommensurable orincompatible in some important aspects. The theo

ries often focus on different core concepts. A multitheoretical framework should focus on commoncore concepts.

Legitimacy theory and resource-based perspectives are believed to be useful as they can be conceived as using what CampbeU et al. (2003, p. 559)call "the stakeholder metanarrative". Thus, theseperspectives can be explored by using stakeholdertheory insights. On the other hand, organisationallegitimacy and organisational reputation are considered to have similar antecedents, social constructionprocesses and consequences (Deephouse and Carter,2005). This study refers to these two interrelatedconcepts: that of legitimacy, which is explored froman institutional perspective; and that of reputation,which is explored from a resource-based perspective.For the purposes of this study, the fundamentalaspect is that legitimacy requires a reputation that

must be retained. It requiresa

company to convince



5/18

688anuel Castelo Branco and Lucia Lima Rodriguesits relevant pubHcs that its activities are congruent

with their values. Thus, reputation and legitimacyare inextricably Hnked, and in this study thedistinction between the two wiU not be exploredfurther.

Development of hypothesesIn what foUows, explanations for SFJ} based on thetheoretical framework presented in the previoussection are developed by selecting the most relevantfactors influencing SRD. To analyse the usefulnessof the theoretical framework proposed above, thisstudy adopts the strategy of examining a sample ofcompanies and using a variety of proxies for acompany's social visibility related to its characteristics and media exposure. Variables are chosen torepresent particular aspects of social visibility, and ineach case, an expectation regarding its relationship toSRD is stated based on prior literature.

International experienceInternational experience is developed by operatingin, and depending upon, foreign markets (Bansal,2005). The importance of international experienceas a determinant of SRD can be explained from theperspective of social and political theories (Choi,1999) and a perspective which is resource based(Bansal, 2005).

The manner inwhich the role of a company, andits stakeholders, is defined in a country, wiUundoubtedly affect SRD practices. Operating inforeign markets requires companies to consider national differences in customer needs, which areinfluenced by the culture and customs of thatcountry. Companies are exposed also to a greaterextent to the laws, rules and regulations governingtrade within different countries.

In less developed countries one would expect thata company which does a larger amount of businessabroad is exposed to a broader spectrum of stakeholder influences and to the international community scrutiny. Given the trend of pro-socialresponsibUity international initiatives, such exposureis likely to lead to more proactive corporate initiatives with respect to the social responsibility issues.

HI: There wiU be a positive relationship betweenthe degree of international activity and SRD.

Company size

SRD is related to corporate size, with larger companies disclosing more than smaUer ones (see, forexample, Adams et al., 1998; Archel, 2003; Neuet al., 1998; Patten, 1991; Purushothaman et al.,2000). Size is also used commonly as a proxy forpublic visibility. Larger companies are more susceptible to scrutiny from stakeholder groups sincethey are highly visible to external groups and morevulnerable to adverse reactions among them; andlarger companies,

on average, are more diversifiedacross geographical and product markets, thus having larger and more diverse stakeholder groups(Brammer and Pavelin, 2004a, p. 704). It is also

more likely that larger, more visible companies wiUconsider social responsibility activities and disclosureas a way of enhancing corporate reputation.

H2: There will be a positive relationship betweensize and SRD.

Industry affiliationAnother commonly used proxy for social visibility isindustry affiliation. This was found to be related toSRD by legitimacy theory studies. Industries withhigh pubUc visibility, or a potentiaUy more important environmental impact, or having less favourablepublic images were found to disclose more socialresponsibility information than their counterparts(see, for example, Adams et al., 1998; Archel, 2003;

Clarke and Gibson-Sweet, 1999; Patten, 1991).There are reasons to suspect that industry affilia

tion is related to certain categories of SRD. Companies in industries with larger potentialenvironmental impact are more likely to provideenvironmental information, and companies inindustries with high visibility among final consumersare more likely to consider important issues ofcommunity involvement and disclose informationrelated to such involvement (Clarke and GibsonSweet, 1999).



6/18


Thus, this study suggests that the classifications ofindustries should be refined to provide more reHabletests. Two types of proxies for social exposure related toindustry affiHation which were proposed in previousstudies are used: "consumer proximity" (see, forexample, CampbeU et al., 2006; Clarke and GibsonSweet, 1999) and "environmental sensitivity" (see, forexample, Archel, 2003; Patten, 2002b). The differentproxies for social exposure are beHeved to be related todifferent SRD categories: community disclosure isexpected to be related positively with a measure ofproximity to the final consumer, whereas environ

mental disclosure is expected to be related positivelywith ameasure of environmental sensitivity.

Consumer proximityThe nearer a company is to the individual consumer,the more probable is its name to be known to most

members of the general public, and hence, thegreater wiU be its social visibility. Thus, it is hypothesised that community involvement disclosure isassociated with the measure of a company's proximity to the final consumer.

H3a: There wiU be a positive relationship betweencommunity involvement disclosure and theconsumer proximity measure.

Environmental sensitivityCompanies in industries that have a larger potentialimpact on the environment are considered to besubject to greater pressures with respect to environmental concerns than companies in industrieswith less risk in terms of environmental impact.

Therefore, companies in environmentaUy sensitiveindustries are more likely to disclose environmentalinformation than companies in less environmentaUysensitive industries.

H3b: There wiU be a positive relationship betweenenvironmental disclosure and the environ

mental sensitivity measure.

Media exposureSeveral studies suggest that individual companies'

media exposure, which is used as a proxy for social

visibility, is likely to be associated to higher levels ofSRD (Bansal, 2005; Brammer and PaveUn, 2004b,2007, in press; Bewley and Li, 2000; Cormier et al.,2004, 2005). The total amount of media coverageraises companies' visibiUty, making them the objectof further public attention and scrutiny (Bansal,2005, p. 203).H4: There wiU be a positive relationship between

SRD and the media exposure measure.

Control variables

Control variables, which are designed to account forother potential influences on SRD practices andhave been analysed in the SRD literature, areintroduced. Prior researchers argue that socialresponsibility activities and disclosure are dependenton the availability of financial resources within acompany (for example, Brammer and Pavelin, 2007,in press; Roberts, 1992). FoUowing Brammer and

Pavelin (2007, in press), profitability and leverage areused in this study to capture the availability offinancial resources within a company. These variables are included as control variables.

ProfitabilitySeveral empirical studies have concluded that profitability does not appear to be a significant determinantof SRD (for example, Archel, 2003; Brammer andPavelin, 2007, in press; Patten, 1991; Purushothamanet al., 2000). From a legitimacy theory perspective,profitability can be considered to be related positivelyor negatively to SRD (Neu et al., 1998). On the otherhand, from a stakeholder perspective (Roberts, 1992),economic performance is expected to be associatedpositively with social responsibility activities and disclosure. In view ofthe existence of these results anddifferent interpretations, the association between thisvariable and SRD is tested without making any a priori assumption about the sign of such association (see,for example, Archel, 2003; Bewley and Li, 2000;Purushothaman et al., 2000).

LeverageThe power of creditors as a stakeholder groupdepends upon the degree to which

acompany relies



7/18

690anuel Castelo Branco and Lucia Lima Rodrigueson debt financing (Roberts, 1992). Noting a lack ofconclusiveness in the studies which explore thisrelationship, Purushothaman et al. (2000, p. 112)point out that companies with high leverage mayhave closer relationships with their creditors and useother means to disclose social responsibiHty information. Thus, in this study the association betweenthis variable and SPJD is tested without making anya priori assumption about the sign of such association.

Methods

Empirical models

The statistical analysis conducted in this studyincludes the use of multiple Hnear regression modelsto analyse the relationship between total SRD andeach one of its categories, both in annual reports andon the Internet, and the influencing factors referredto in the previous section. Ten models are examined. The models aU use the same influencing factorsdiscussed above. However, there are ten differentdependent variables: total SRD in annual reports andon the Internet (SRDAR and SPJD I), environ

mental disclosure in annual reports and on theInternet (EDAR and ED I), human resources disclosure in annual reports and on the Internet(HFJ3AR and HRDI), products and customers

disclosure in annual reports and on the Internet(PCDAR and PCDI), community involvement

disclosure in annual reports and on the Internet(CIDAR andCIDI).

The approach adopted in the empirical analysis issummarised by the foUowing general form of the

models:SPJD disclosure index =/ (international experi

ence, size, media exposure, consumer proximity,environmental sensitivity, control variables)

The general form of the models examined is thusstated as:

DISC, = 60l + Bi/IE, + B2lS, + B3lME, + fi4,CP,+ B5lES, + Be/Profit, + fi7lLev, + u,

where, for company i: DISC,: SRD index(SRDAR,; SRDI,; EDAR,; EDI,; HRDAR,;

HRDI,; PCDAR,; PCDI,; CIDAR,; CIDI,); IE,:international experience; S,: size; ME,:

media

exposure; CP,: consumer proximity; ES,: environmental sensitivity; Profit,: profitabdity; Lev,: lever

age; u,: error term.

Operationalisation of variablesInternational experienceFoUowing Bansal (2005), international experience is

measured by the percentage of sales outside Portugalto total sales as reported in the segment data of thefinancial statements (see also Choi, 1999).

Company sizeBecause there are no theoretical reasons whichmight clearly justify choosing a particular measure of

size (Hackston and Milne, 1996, p. 87), the measureused in this study is total assets, as reported on thebalance sheet (Brammer and Pavelin, 2004b; Haniffaand Cooke, 2005).

Industry affiliationBecause it is suggested in this study that the classifications of industries should be refined to providemore reliable tests ofthe theoretical framework used,two types of proxies for social exposure related toindustry affiliation are used: "consumer proximity"and "environmental sensitivity". The differentproxies for social exposure are believed to be relatedto different SRD categories: community disclosure isexpected to be related positively to a measure ofconsumer proximity, whereas environmental disclosure is expected to be related positively to a

measure of environmental sensitivity.

Consumer proximity. In this study, a binary measure(high profile and low profile) is used. High-profilecompanies are those that are better known to thefinal consumer and whose names are expected to beknown to most members of the general public.Based on prior literature, high-profile companies areidentified as those in the foUowing sectors: house

hold goods and textiles, beverages, food and drugretailers, telecommunication services, electricity, gasdistribution, water and banks. AU others are considered "low profile". A one/zero variable is used todesignate companies from these industries: one if thecompany is from a high-profile sector, and zero if itis from a

low-profilesector.



8/18


Environmental sensitivity. In this study, "more sensitive" industries are considered to be those with more

risk of being criticised in environmental mattersbecause of their activities involving higher risk ofenvironmental impact (such has natural resourcedepletion or poUution). Thus, based on prior literature, the foUowing "more sensitive" sectors areidentified: mining, oil and gas, chemicals, construction and budding materials, forestry and paper,steel and other metals, electricity, gas distributionand water. All others are considered as "less sensitive". A one/zero variable is used to designatecompanies from these industries: one if the companyis from amore sensitive industry and zero if it is froma less sensitive industry.

Media exposureTo develop a measure of the companies' mediaexposure, the number of articles in two Portuguesenewspapers was counted. Company exposure was

measured by perusing "Jornal de Noticias" and"Expresso", for the period between 1 January 2001and 31 December 2003. The search facilities presenton the web pages of these newspapers were used. Asearch was carried out for each company, using as a

keyword, the name of the company. The searchresults were examined to exclude articles that did notrelate specifically to social responsibility issues.

Control variablesProfitability and leverage are used in this study tocapture the availability of financial resources withinthe company. These two variables are used as control

variables.

Profitability. When measuring corporate performanceone can use accounting or market-based measures.In contrast with accounting-based measures, marketbased measures are less subject to bias by managerial

manipulation and they do not rely on past performance (McGuire et al., 1988, p. 859). However,they are based on investors' viewpoints on companyperformance, thus ignoring other stakeholder groups(ibid.). This is the main reason for adopting anaccounting-based variable in this study.

Thus, return on assets (ROA) is used as ameasurefor economic performance (Belkaoui and Karpik,1989; Bewley and Li, 2000; Brammer and Pavelin,2007, in press; Cormier et al., 2004; Patten, 1991).

ROA ismeasured by the ratio of Net income/totalassets (Belkaoui and Karpik, 1989).

Leverage. Leverage ismeasured by the ratio of Totaldebt/total assets (see, for example, Belkaoui and

Karpik, 1989; Brammer and Pavelin, 2007, inpress).

SampleThe sample used in this study comprises listedcompanies, as they are more likely to disclose socialresponsibility information and have aweb page thatprovides SRD. To be included in the sample for thisstudy, a company had to:

have its shares listed on the Portuguese StockExchange (Euronext - Lisbon) by the end of2003,have its 2003 annual report available for review, andhave an accessible corporate web page onthe Internet by August 2004.

The initial sample included aU companies listed onEuronext - Lisbon at 31 December 2003. From theinitial 57 listed companies (50 of them listed on the

main market and seven on the second market), afinal sample of 49 companies was identified, as described in Table I.

The companies included in the sample are classified according to sectors using the FTSE GlobalClassification System. This classification systemcomprises the several sectors which are considered in

Table II. Construction and building materials is thesector which presents the largest number of companies (8 companies and around 16% of the total).Banks and Forestry and paper are the sectors whichfoUow in terms degree of importance (each of them

with 6 and about 24% taken together).

Data collection

To measure the level of social responsibility information disclosed by sample companies, this studyuses "content analysis". This technique consists of

classifying the information disclosed into several



9/18

692anuel Castelo Branco and Lucia Lima RodriguesTABLE I

Identification of the sample

Description Main market Second market OveraU

Companies Hsted on Euronext Lisbon at 31 December 2004 50 7 7LessCompanies not subject to Portuguese law 2 2Companies without annual report for 2003 1 1Companies without web page 12Companies with web page under construction or maintenance 3 0Final sample 43 49

TABLE IINature of SRD by sectors

Companies Environmental Human resources Products and Communitydisclosure disclosure consumers involvement

disclosure disclosure

Annual Web Annual Web Annual Web Annual Webreports pages reports pages reports pages reports pages

Sector n % n % n % n % n % n % n % n %Automobiles and parts 0 00 01100 0 01100 0 00 00 0Banks 3 50 2 33 6 100 2 33 6 100 5 83 5 83 5 83Beverages 1 50 2 100 2 100 2 100 2 100 1 50 0 0 1 50Chemicals 2 100 2 100 2 100 2 100 1 50 2 100 1 50 2 100Construction and buUding materials 6 75 3 38 7 88 4 50 6 75 6 75 3 38 2 25Electricity 1 100 1 100 1 100 1 100 1 100 1 100 1 100 1 100Electronic and electrical equipment 1 100 0 0 1 100 1 100 1 100 1 100 0 0 0 0Engineering and machinery 1 100 1 100 1 100 1 100 0 0 1 100 1 100 1 100Food and drug retailers 2 67 2 67 3 100 2 67 2 67 1 33 1 33 2 67Food producers and processors 0 00 02100 1 50 1 50 0 00 00 0Forestry and paper 2 33 3 50 3 50 3 50 2 33 1 17 1 17 1 17Household goods and textiles 0 0 0 0 1 100 1 100 0 0 1 100 0 0 1 100Leisure, entertainment and hotels 1 33 1 33 3 100 0 0 1 33 0 0 0 0 0 0

Media and photography 0 00 02 67 0 0133 0 00 01 33Software and computer services 0 0 0 0 4 100 2 50 1 25 3 75 0 0 0 0Telecommunication services 2 100 2 100 2 100 2 100 1 50 2 100 1 50 2 100Transport 1 33 1 33 3 100 1 33 2 67 1 33 0 0 1 33Total 23 47 20 41 44 90 25 51 29 59 26 53 14 29 20 41%: Percentage of disclosing companies in the sector (except for the final line in the tablewhere disclosing companies as apercentage of total sample is reported).

categories of items which capture the aspects of social responsibiUty one wants to analyse.

The simplest form of content analysis consists ofdetecting the presence or absence of information

(see, for example, Haniffa and Cooke, 2005; Patten,2002b; Purushothaman et al., 2000). One of the

main shortcomings of this form of content analysis isthat it does not aUow the measurement ofthe extent



10/18


of information disclosure and, therefore, the codeddata do not reflect the emphasis that companiesattach to each information item (Zeghal and Ahmed,1990, p. 42). However, the number of differenttopics discussed is considered as a reasonable measureof management's wiUingness to provide socialresponsibility information in general (Bewley and Li,2000, p. 206). On the other hand, we consider it tobe a more appropriate method than counting ofsentences, words or proportion of pages when one iscomparing such different media of disclosure asannual reports and web pages.

Thus, the analysis of the SRD ismade using anequal-weighted index, that is, a scoring system

which assigns a point for each SRD theme pertaining to any of the categories considered. Disclosurescores for each company are added and not weighted, because it is assumed that each item of disclosureis equally important.

Listed companies' 2003 annual reports were analysed. Only the sections ofthe annual report where thedisclosure of social responsibility information is voluntary were analysed, namely the chairman's report orletter to the shareholders and the management report.

Each of the companies' web pages was accessedand analysed during the month of August 2004. Theentire web pages were examined. All links werefollowed, but for the following exclusions:

neither on-line copies of the annual report(Patten and Crampton, 2004) nor on-linecopies of social and/or environmental reports, where available, were included in the

web page analysis;links to external press release disclosures

were also not followed (but press releases ofthe companies were examined for SRD)(Patten and Crampton, 2004);links to company publications such as newsletters or product catalogues were not followed.

The reason for the exclusions referred to in thepreceding paragraph is the purpose of coUectingsegregated data on the two media analysed (Frostet al., 2005, p. 91). This is done because the focus ison the comparison of the social responsibilityinformation that companies choose to highlight on

their web pages with similar information disclosedon their annual reports.

Several empirical studies in the area were of greatutility in developing the SRD index used (see, forexample, Archel, 2003; Adams et al., 1998; Grayet al., 1995b, Hackston and Milne, 1996; Patten,1991; Purushothaman et al., 2000; WiUiams and Pei,1999). SRD refers in this study to disclosures in the

following four categories:

environmental;human resources;products and consumers;community involvement.

Environmental disclosure comprises disclosuresrelating to environmental policies, environmental

management system and environmental awards(including ISO 14001 and Eco Management and

Audit Scheme - EMAS), the environmental impactsof products and processes, environment-relatedexpenditures, the environmental benefits of products,conservation of natural resources and recyclingactivities, and disclosures concerning energyefficiency. Human resources disclosure covers suchissues as employee numbers and remuneration,employee share ownership, employee consultation,training and education, employment of minorities or

women, and trade union information. Products andconsumers disclosure encompasses disclosures relatedto product quality (for example, third-party recognition for the quality of the company's products) andconsumer relations (for example, customer complaints) .Community involvement disclosure includesdisclosures relating to sponsorship (e.g. of art exhibits), asweU as charitable donations and activities.

The total maximum score is of 30. The maximumscore for each of the categories considered is of 11 forenvironmental disclosure, 9 for human resources disclosure, 5 for products and consumers disclosure and 5for community involvement disclosure (see Table III).To avoid penaHsing companies for not disclosingitems considered irrelevant to them, these items wereexcluded. This is the case with Banks and Software andcomputer services sectors, particularly regarding some ofthe environmental disclosure items (poUution arisingfrom use of product, discussion of specific environ

mental laws and regulations, prevention or repair of



11/18

694Manuel Castelo Branco and Lucia Lima RodriguesTABLE III

SPJD areas

Categories and items of disclosure Annual reports Web pagesn % n %

Environmental disclosureEnvironmental policies or company concern for the environment 16 32.657 34.69Environmental management, systems and audit 14 28.57 100.41PoUution from business operations 11 22.45 9 18.37PoUution arising from use of product 4 8.16 9 18.37Discussion of specific environmental laws and regulations 4 8.16 1.04Prevention or repair of damage to the environment 3 6.12 5 10.20Conservation of natural resources and recycling activities 112.45 86.33SustainabUity 11 22.45 12 24.49Environmental aesthetics 4 8.16 3 6.12Conservation of energy in the conduct of business operations 1020.41 74.29Energy efficiency of products 1 2.04 2 4.08

Human resources disclosureEmployee Health and Safety 16 32.65 9 18.37Employment of minorities or women 2 4.08 0 0.00Employee training 29 59.18 5 10.20Employee assistance/benefits 6 12.24 4 8.16Employee remuneration 26 53.06 5 10.20Employee profiles 32 65.31 19 38.78Employee share purchase schemes 14 28.57 3 6.12Employee morale 10 20.41 6 12.24Industrial relations 7 14.29 0 0.00Products and consumers disclosureProduct safety 6 12.24 7 4.29Product quality 25 51.02 23 46.94Disclosing of consumer safety practices 1 2.04 4 8.16Consumer complaints/satisfaction 12 24.49 10 20.41Provision for disabled, aged, and difficult-to-reach consumers 3.12 50.20Community involvement disclosureCharitable donations and activities 10 20.41 14 28.57Support for education 14 28.57 13 26.53Support for the arts and culture 12 24.49 14 28.57Support for pubUc health 3 6.12 5 10.20Sponsoring sporting or recreational projects 10 20.41 13 26.53

%: Disclosing companies as a percentage of total sample.

damage to the environment, environmental aestheticsand energy efficiency of products) and some of theproducts and consumers disclosure items (safety andcustomer safety practices). The same is thought to be thecase with companies from the Leisure, entertainmentand hotels sectors, but only regarding environmentaldisclosure items.

Thus, the disclosure score indexes are constructedto take into account these considerations:

This index expresses the level of disclosure for acompany j, where N is the maximum number ofrelevant items a company may disclose and d{ is equalto 1 if the indicator i is disclosed, and 0 otherwise.



12/18


When the disclosure score index is equal to 0, itindicates that company / does not disclose any item.Index values equal to i = 1,..., m, mean that a levelof disclosure is provided, and m, is the maximumnumber of indicators d{ disclosed by a company j.

In the case of Banks and Software and computerservices sectors, environmental information totalscore is 6; and for products and consumers, the totalscore is 3 (SRD total score of 23). In the case ofcompanies from the Leisure, entertainment andhotels sector, environmental information total scoreis 6 (SRD total score of 25).

By comparison with other sectors, the financialservices sector has significantly lower direct environmental impact. This is used by some authors as anargument to exclude banks and finance companieseven in studies which analyse all the various components of SRD (Archel, 2003). Simpson andKohers (2002, p. 101) characterise the bankingindustry as having a limited direct poUution of theenvironment and a relatively homogeneous production process where product safety and employeesafety are minimal concerns.

In the case of banks, which represent an importantpart of the sample used in this study, we have considered that environmental disclosure comprisesdisclosures relating to environmental policies, environmental management systems and environmentalawards (including ISO 14001 and Eco Managementand Audit Scheme ? EMAS), lending and investmentpolicies (considered as poUution from businessoperations), conservation of natural resources andrecycling activities, sustainability and disclosuresconcerning energy efficiency. PoUution arising fromuse of product and energy efficiency of products arenot concerns that banks have and, as far as we know,there are no relevant specific environmental laws andregulations pertaining to the banking sector.

Results and discussion

Descriptive analysisResults suggest that companies prefer the annualreport as an SRD medium. 11 companies do notpresent social responsibility information on theInternet, whereas in the annual report the same

happens with only 5 companies. For the annual

report 44% of companies disclose 3 or 4 categories ofsocial responsibility information, and 45% only disclose information related to one or two of the categories considered. On the Internet, 38% of thecompanies disclose 3 or 4 categories of socialresponsibility information, and 39% only discloseinformation related to one or two of the categoriesconsidered.

Results in Table II indicate that the kind of socialresponsibility information that more companies disclose in their annual reports is human resourcesinformation (90%), foUowed by products and consumers information (59%) and environmentalinformation (47%). On the Internet, the kind ofinformation that more companies provide is products and consumers information (53%), foUowed byhuman resources information (51%).

Comparison of the information disclosed on theInternet with similar information disclosed in theannual reports in Table II indicates that communityinvolvement information is not disclosed as frequently both on the Internet and the annual reports.Community involvement disclosure is the onlycategory regarding which the Internet is the preferred media of disclosure by companies (40% of thecompanies disclose this information on the Internet,

whereas only 29% of them use the annual report todo so).

These results are similar to those reported byClarke and Gibson-Sweet (1999) and can be interpreted in a similar way. For example, Banks andTelecommunication services are sectors with a highvisibility among consumers, and community relations disclosure is an important part of the SRD

made by companies in these sectors. As expected,while few banks disclose environmental information,the percentage of retailers and telecommunicationservices

disclosingsuch information is higher

compared to banks.

On the other hand, as Clarke and Gibson-Sweet(1999) suggest, some industries have a larger potential impact on the environment but are not asclose to the final consumer, and the public is lessaware of them. A company less weU known to thepublic, and involved in activities with larger potential impact on the environment, would have lessreason to justify its existence to society by means ofcommunity disclosures than a better known one.This seems to be the case of companies in the



13/18

696 anuel Castelo Branco and Lucia Lima Rodrigues

Chemicals, Construction and building materials andForestry and paper sectors: a larger percentage ofthem disclose more environmental information thancommunity involvement information.

What seems more difficult to explain (and was notexpected) are the results for environmental disclosures by companies in some environmentaUy sensitive sectors, such as Construction and building

materials and Food producers and processors. Thesecompanies do not disclose more SPJD than companies from other sectors, as might be expected.However, the fact that the companies included inthe sample are listed on the second market, and arethus less visible, may explain the lack of disclosure, atleast in part.

Comparison of the information disclosed on theInternet with similar information disclosed in theannual reports in Tables II and III indicates thatenvironmental information and human resourcesinformation are more evident in annual reports thanon the Internet, whereas the reverse is the case withcommunity involvement information. The difference only seems to be relevant in the case of humanresources information (the annual report is the preferred medium of disclosure) and communityinvolvement information (the Internet is the preferred medium of disclosure). With respect toproducts and consumers information it is difficult tosay whether it has a stronger presence in annualreports or on the Internet.

As Zeghal and Ahmed (1990) argue, the choice ofamedium for information disclosure is dependent onthe target public for whom the message is intended.

Because annual reports are directed at investors and

human resources are an important resource, it isnatural for investors to be interested in it. On theother hand, because company web pages are aimedat a broader public, including consumers, it is naturalfor companies to give prominence to communityinvolvement and products/consumers information.Table IV presents the results of the test on thedifference between SRD in the annual reports andon the Internet. Both the Wilcoxon-signed ranktest and the paired sample t-test indicate significantdifferences in total SRD and human resources disclosure (more disclosure in annual reports). Although the difference in community involvementdisclosure is not statisticaUy significant, there is animportant difference between the Internet and annual reports as disclosure media (more disclosure onthe Internet).

Analysis ofthe main results

An analysis of the Kolmogorov-Smirnov (K-SLiUiefors) and the Shapiro-Wilk normality test statistics suggests that dependent variables and continuous independent variables are not distributednormaUy. To bring the variables closer to normaUtyfor the purpose of the regression analysis, thedependent and independent continuous variables aretransformed by computing normal scores using Vander Waerden's transformation (Haniffa and Cooke,2005). The regression analysis is performed with thetransformed variables.

Multiple regression is used to test the hypothesesdeveloped above. However, before conducting the

TABLE IVPaired sample f-tests and Wilcoxon-signed rank test for SRD in annual reports and on the Internet

Paired sample t-test Wilcoxon test

Mean difference t Sig. (2-taUed) Z Asymp. Sig. (2-taUed)Total SRD.063 3.452 0.001 -2.997a 0.003

Environmental disclosure 0.020 0.901 0.372 -0.316a.752Human resources disclosure 0.206 6.949 0.000 -5.132a.000Products and consumers disclosure -0.008 -0.186 0.853 -1.333b .844Community involvement disclosure -0.041 -1.183 0.243 -1.183b0.183

aBased on positive ranks (disclosure in annual reports > disclosure on the Internet).Based on negative ranks (disclosure in annual reports < disclosure on the Internet).



14/18


analysis several diagnostic tests, such as normalitytests, autocorrelation tests, heteroscedasticity testsand multicoUinearity tests, are performed to ensurevaHd conclusions are drawn based on the multipleregression results. If the tests are not satisfied thencorrective procedures are performed. The JarqueBera normaHty test is performed on the residuals ofaU the models.

The possible existence of multicoUinearity is tested based on the correlation matrix incorporating aUthe independent variables (transformed data) asweUas computing the variance inflation factor (VIF).

Results indicate that multicoUinearity is unlikely tobe a problem. In addition, results suggest that innone of the regressions the highest VIF is above 3,confirming that there is no need to be concernedabout the correlation between the independentvariables.

To test for unequal variances, White's generalheteroscedasticity test is performed on each set ofresults. All chi-squared test statistics are not significant at a five percent level. Thus, the tests suggestthat a widespread heteroscedasticity problem doesnot exist in the data and no corrective procedure isundertaken to combat its presence in the data.

Multiple regression analysis is used for multivariate testing ofthe hypotheses. Each ofthe dependentvariables, SRD in annual reports and on the Internet,is regressed against the transformed independentvariables. The categorical variables are also included.Table V presents the results ofthe regression modelspertaining to total SRD.

Table V reports the results of regressing theindependent variables on the dependent variabletotal SRD. The F values for the two models aresignificant at 0.01 level. This suggests that theindependent variables considered, when takentogether, explain total SRD and its categories takenindividuaUy. However, this does not mean that eachof the independent variables contributes to theexplanation of the dependent variables.

The adjusted R2,s suggest that approximately 34%(in the case of annual reports) and 37% (in the case ofthe Internet) of the variation in the SRD scoresbetween the companies can be explained by theindependent variables included in the regression

models. Only two of the independent variables aresignificant in each model: size in both, mediaexposure in the case of annual reports and leveragein the case of the Internet. The coefficients of total

TABLEVResults of the regression models for total SRD

Independent variables Disclosure media: Annual reports Disclosure media: Web pagesCoefficient estimate Coefficient estimate

(Constant) 0.109 -0.201International experience 0.027.146

Leverage -0.256 -0.332**Profitability -0.091 0.227Size 0.490** 0.404**

Media exposure 0.376*.224Environmental sensitivity ?0.21.181Consumer proximity ?0.151 .435

R2 = 0.432; Adj. R2 = 0.335; R2 = 0.465; Adj. R2 = 0.374;Durbin-Watson = 1.762 Durbin-Watson = 1.937F = 4.460; p = 0.001 F = 5.100; p = 0.000

White heterosced. test: White heterosced. test:Obs * R2 = 36.933; p = 0.292 Obs * R2 = 28.025; p = 0.713Jarque-Bera test: JB = 2.631; Jarque-Bera test: JB = 0.601;p = 0.268 p = 0.740

*Significant at the 0.10 level (2-tailed).**Significant at the 0.05 level (2-tailed).



15/18

698 anuel Castelo Branco and Lucia Lima Rodriguesassets and media exposure are positive. This indicates, as hypothesised, that as the value of thesevariables increases so does a company's SRD score.Leverage is significant in the case of total SRD onthe Internet, presenting a negative coefficient whichsuggests that the higher the leverage in a company,the lower its SRD.

Thus, at an aggregated level, the supportedresearch hypotheses in the case of SRD in annualreports are those related to size (H2) and mediaexposure (H4), whereas in the case of SRD on theInternet the supported hypothesis is the one relatedto size (H2).

Regarding the results of regressing the independent variables on each category of SRD, from theeight regression models only those which have hu

man resources disclosure as a dependent variable arenon-significant at the 1% level. In the case of humanresources disclosure in annual reports the regressionis significant at the 2% level, whereas in the case ofsimilar disclosure on the Internet the regression is

non-significant.The explanatory power of the regression ranges

from 7.5% for the human resources disclosure onthe Internet to 49.5% for community involvementdisclosure on the Internet. As for the importance ofthe independent variables in explaining variationbetween companies' disclosure, the size variable issignificant with positive coefficients in almost aUthe regression models. The exceptions are the

models which have as dependent variables productsand consumers disclosure in annual reports andhuman resources disclosure on the Internet. The

media exposure variable is significant with positivecoefficients only when products and consumersdisclosure is the independent variable (both inannual reports and on the Internet). Consumerproximity is significant with a positive coefficientin the case of community involvement disclosureon the Internet, which leads to a conclusion that

H3a is accepted in the case of disclosure on theInternet. In addition, leverage is significant with a

negative coefficient in the case of human resourcesdisclosure on the Internet, and profitability is significant with a positive coefficient in the case ofproducts and consumers disclosure in annualreports.

Table VI summarises the results on hypothesistesting.

Consistent with previous studies size and mediaexposure, which are considered as proxies for socialvisibUity, have in general a positive relationship withtotal SRD. These results are consistent with theexpectations resulting from the theoretical frame

work proposed and with previous SRD studies.The non-significant relation between SRD andinternational experience in both media of disclosureconsidered is an unexpected result. These results areconsistent with those of Choi (1999), who analysedenvironmental disclosure practices of Korean listed

TABLE VISummary of the results from the hypotheses testing

VariablesHypotheses Annual reportsWebages

International experience Positive relation Non-significant Non-significantCompanysizeositive relation Positive relation Positive relation

Non-significant: products Non-significant: humanand consumers resources

Media exposure Positive relation Positive relation: products Positive relation: productsand consumers and total and consumersSRD

Non-significant Non-significantEnvironmental visibility Positive relation Non-significant Non-significantConsumer proximity Positive relation Non-significant Positive relation:

community involvementNon-significant



16/18

Factors Influencing SLID by Portuguese Companies 699

companies and did not find a significant associationbetween international experience (which he calls"foreign customers' influence") and disclosure.

Regarding consumer proximity, a significantpositive relation is only discernible in the case ofInternet community involvement disclosure.Although a positive relation in the case of annualreports' community involvement disclosure is alsohypothesised, these results are considered to beconsistent with the theoretical framework proposed.Because company web pages are aimed at a broaderpublic than annual reports, it is natural for companiesto give prominence to the Internet as a media ofdisclosing their community involvement activities.

Environmental visibility is not a factor whichexplains the differences in environmental disclosureamong companies. This is an unexpected finding.The theoretical framework proposed leads to theexpectation that higher environmental visibility isassociated to higher levels of disclosure, and thefindings of previous SRD studies are consistentwith such expectation. These findings are a signthat companies with a more limited environmentalimpact are also disclosing environmental information. For example, banks are increasingly disclosingthis kind of information. This is probably explainedby the fact that in the last two decades the focus hasbeen on environmental responsibility and disclosure.

Portuguese firms seem to be quite sensitive topublic perceptions, as proxied by their media visibility and their size, when determining their SRDstrategies. These findings are well documented in theliterature, both in Anglo-Saxon countries (see, forexample, Bewley and Li, 2000; Neu et al., 1998;Patten, 1991) and in continental Europe (Cormierand Magnan, 2003; Cormier et al., 2005). Theseresults are also consistent with literature from othercountries (Choi, 1999; Purushothaman et al., 2000).

Cormier and Magnan (2003, p. 58), who haveanalysed French firms' environmental reportingpractices, suggest that "corporate disclosure strategiesseem to be determined in a similar way, irrespectiveof a given country's socio-cultural environment".These authors consider that this "is an iUustrationof the strong impact of globalised stock marketson fostering convergence in corporate practices"(ibid.).

Concluding remarks

This study analyses some factors which influenceSRD by a sample of companies listed on thePortuguese Stock Exchange (Euronext - Lisbon),using a theoretical framework which combineslegitimacy theory and a resource-based perspective.According to this framework, managers increasinglyneed to consider SRD as a signal of improved socialand environmental conduct in those fields becausedisclosure influences the external perception ofreputation. By demonstrating that they operate inaccordance with social and ethical criteria, companies can build reputation, whereas failing to do socan be a source of reputational risk.

Portuguese companies attribute greater importance to annual reports as disclosure media than tothe Internet. Noticeable differences are related to the

much higher presence of human resources information in annual reports than on the Internet andthe higher presence of community involvementinformation on the Internet than in annual reports.

These results are probably related to the fact thatannual reports are directed at investors and it isnatural for investors to be interested in humanresources. On the other hand, because company webpages are aimed at a broader public, it is natural forcompanies to give prominence to communityinvolvement information.

Evidence seems to suggest that companies withhigher visibiUty exhibit greater concern to improvecorporate image through SRD both on the Internetand in annual reports. In addition, in sectors with ahigh visibiUty among consumers there is greater concern for community involvement activities and disclosure. There is thus some support for the use of acombination of legitimacy theory with resource-basedperspectives to explain SRD by Portuguese Ustedcompanies.

An interesting result is related to the seeming lackof significant difference between the factors influencing SRD practices of Portuguese listed companies when compared to companies from moredeveloped countries. We consider that there is noreason to expect that companies in the less developed capital markets wiU behave in a significantlydifferent manner than companies inmore developedcapital markets.



17/18

700anuel Castelo Branco and Lucia Lima RodriguesWe interpret the findings as a result of the con

vergence in corporate practices which is promotedby the impact of globaHsed stock markets (Cormierand Magnan, 2003, p. 58) and has as consequence aseeming lack of importance of general contextualfactors in determining disclosure practices of Hstedcompanies. Although further analyses are required tovalidate this claim, it is a promising avenue for futureresearch (ibid).

Furthermore, we believe that such similitude inthe way inwhich disclosure practices are determinedis not as Hkely to happen in the case of unlistedcompanies. SPJD practices of listed companies areless subject to general contextual factors than thoseof unhsted companies. An interesting possibleextension of this study would be to use a sample ofcompanies which are not listed, including smaU and

medium-sized companies.Other possible extensions of this study, which are

not mutuaUy exclusive, may be envisaged to addnew insights to the analysis of SRD by companies.One such possible extension is related to the use ofmore refined content analysis procedures. Anotherpossible extension is an in-depth analysis of categories of SPJD, which very likely would involvevariations to the theoretical framework proposed.FinaUy, the use of a larger sample would be animportant way of adding new insights to the analysisof SPJD by Portuguese companies.

Notes

Only the results of the White heteroscedasticity testusing cross terms are reported.2 Detailed results of regressing the independent variables on each category of SPJD are avaUable from theauthors on request.

References

Adams, C. A., W.-Y. HiU and C. B. Roberts: 1998,'Corporate Social Reporting Practices in Western

Europe: Legitimating Corporate Behaviour', TheBritish Accounting Review 30(1), 1?21.

Archel, P.: 2003, 'La divulgacion de la informacion socialy medioambiental de la gran empresa espafiola en el

periodo1994?1998: situation actual y perspectives',

Revista Espafiola de Financiacion y Contabilidad 117,571-601.

Bansal, P.: 2005, 'Evolving SustainabiUty: A LongitudinalStudy of Corporate Sustainable Development', Strategic

Management Journal 26(3), 197-218.Belkaoui, A. and P. G. Karpik: 1989, 'Determinants of

the Corporate Decision to Disclose Social Information', Accounting, Auditing and Accountability Journal2(1), 36-51.

Bewley, K. and Y. Li: 2000, 'Disclosure of Environmental Information by Canadian ManufacturingCompanies: A Voluntary Disclosure Perspective',Advances in Environmental Accounting and Management 1,201-226.

Branco, M. C. and L. L. Rodrigues: 2006, 'CorporateSocial Responsibility and Resource Based Perspectives', Journal ofBusiness Ethics 69(2), 111-132.

Brammer, S. and S. Pavelin: 2004a, 'Buddinga Good

Reputation', European Management Journal 22(6), 704713.

Brammer, S. and S. Pavelin: 2004b, 'Voluntary SocialDisclosures by Large UK Companies', Business Ethics:A European Review 13(2/3), 86-99.

Brammer, S. and S. Pavelin: 2007, 'Factors Influencingthe Quality of Corporate Environmental Disclosure',Business Strategy and theEnvironment (in press).

CampbeU, D., B. Craven and P. Shrives: 2003, 'Voluntary Social Reporting in Three FTSE Sectors: A

Comment on Perception and Legitimacy', Accounting,Auditing andAccountability Journal 16(4), 558-581.

Campbell, D., G. Moore and P. Shrives: 2006, 'CrossSectional Effects in Community Disclosure', Accounting,Auditing andAccountability Journal 19(1), 96?114.

Choi, J.-S.: 1999, 'An Investigation of the Initial Voluntary Environmental Disclosures Made in KoreanSemi-Annual Financial Reports', Pacific AccountingReview 11(1), 73-102.

Clarke, J. and M. Gibson-Sweet: 1999, 'The Use ofCorporate Social Disclosures in the Management ofReputation and Legitimacy: A Cross Sectoral AnalysisofUK Top 100 Companies', Business Ethics: A EuropeanReview 8(1), 5-13.

Cormier, D., I. M. Gordon and M. Magnan: 2004,'Corporate Environmental Disclosure: Contrasting

Management's Perceptions with Reality', Journal ofBusiness Ethics 49(2), 143-165.Cormier, D. and M. Magnan: 2003, 'Environmental

Reporting Management: A Continental EuropeanPerspective', Journal of Accounting and Public Policy22(1), 43-62.

Cormier, D., M. Magnan and B. Van Velthoven: 2005,'Environmental Disclosure Quality in Large German

Companies: Economic Incentives, Public Pressures or



18/18


Institutional Conditions?', European Accounting Review14(1), 3-39.

Deegan, O: 2002, 'The Legitimising Effect of Social andEnvironmental Disclosures - A Theoretical Foundation', Accounting, Auditing and Accountability Journal15(3), 282-311.

Deephouse, D. L. and S. M. Carter: 2005, 'An Examination of Differences Between Organizational Legitimacy and Organizational Reputation', Journal ofManagement Studies 42(2), 329-360.

Frost, G, S. Jones, J. Loftus and S. Van Der Laan: 2005,'A Survey of SustainabUity Reporting Practices of

Australian Reporting Entities', Australian AccountingReview 15(1), 89-96.Gray, R., R. Kouhy and S. Lavers: 1995a, 'Corporate Socialand Environmental Reporting: A Review of the Literature and A Longitudinal Study of UK Disclosure',

Accounting, Auditing andAccountabilityJournal 8(2), 47-77.Gray, R., R. Kouhy and S. Lavers: 1995b, 'Methodological Themes: Constructing a Research Database ofSocial and Environmental Reporting by UK Companies', Accounting, Auditing and Accountability Journal8(2), 78-101.

Hackston, D. and M.J. MUne: 1996, 'Some Determinantsof Social and Environmental Disclosures in NewZealand Companies', Accounting, Auditing and AccountabilityJournal 9(1), 77-108.

Haniffa, R. M. and T. E. Cooke: 2005, 'The Impact ofCulture and Governance on Corporate SocialReporting', Journal ofAccounting andPublic Policy 24(5),391-430.

Hasseldine, J., A. I. Salama and J. S. Toms: 2005,'Quantity Versus Quality: The Impact of Environ

mental Disclosures on the Reputations of UK PLCS',The British Accounting Review 37(2), 231-248.Lopes, P. T. and L. L. Rodrigues: 2007, 'Accounting for

Financial Instruments: An Analysis of the Determinants of Disclosure in the Portuguese StockExchange', The International Journal of Accounting42(1), 25-56.

McGuire, J. B., A. Sundgren and T. Schneeweis: 1988,'Corporate Social Responsibility and Firm Financial

Performance', Academy of Management Journal 31(4),854-872.

Neu, D., H. Warsame and K. PedweU: 1998, 'ManagingPublic Impressions: Environmental Disclosures inAnnual Reports', Accounting, Organizations and Society23(3), 265-282.

Patten, D. M.: 1991, 'Exposure, Legitimacy, and SocialDisclosure', Journal ofAccounting and Public Policy 10(4),297-308.

Patten, D. M.: 2002a, 'Give or Take on the Internet: AnExamination of the Disclosure Practices of Insurance

FirmWeb Innovators', Journal ofBusiness Ethics 36(3),247-259.

Patten, D. M.: 2002b, 'Media Exposure, PubUc PolicyPressure, and Environmental Disclosure: An Examination of the Impact of Tri Data Availability',

Accounting Forum 26(2), 152-171.Patten, D. M. and W. Crampton: 2004, 'Legitimacy andthe Internet: An Examination of Corporate Web Page

Environmental Disclosures', Advances in EnvironmentalAccounting and Management 2, 31-57.

Purushothaman, M., G. Tower, R. Hancock andR. Taplin: 2000, 'Determinants of Corporate SocialReporting Practices of Listed Singapore Companies',Pacific Accounting Review 12(2), 101-133.

Roberts, R. W.: 1992, 'Determinants of Corporate SocialResponsibility Disclosure: An Application of Stakeholder Theory', Accounting, Organizations and Society17(6), 595-612.Simpson, W. G. and T. Kohers: 2002, 'The Link Between

Corporate Social and Financial Performance: Evidencefrom the Banking Industry', Journal of Business Ethics35(2), 97-109.

Toms, J. S.: 2002, 'Company Resources, Quality Signalsand the Determinants of Corporate Environmental

Reputation: Some UK Evidence', The BritishAccounting Review 34(3), 257-282.

WiUiams, S. M. and C.-A. H. W. Pei: 1999, 'CorporateSocial Disclosures by Listed Companies on Their WebPages: An International Comparison', The International

Journal ofAccounting 34(3), 389-419.Zeghal, D. and S. A. Ahmed: 1990, 'Comparison ofSocial Responsibility Information Disclosure MediaUsed by Canadian Firms', Accounting, Auditing and

Accountability Journal 3(1), 38-53.Zimmerman, M. A. and G. J. Zeitz: 2002, 'BeyondSurvival: Achieving New Venture Growth by Build

ing Legitimacy', Academy ofManagement Review 27(3),414-431.

Manuel Castelo BrancoFaculty of Economics,

University of Porto,Rua Dr. Roberto Frias,4200-464 Porto, PortugalE-mail: mcbranco@fep. up .pt

Lucia Lima RodriguesSchool of Economics andManagement,University ofMinho,Campus de Gualtar,

4710-057 Braga, PortugalE-mail: lrodrigues@eeg. uminho.pt

Documents

Branco, Manuel Castelo dan Lúcia Lima Rodriguez. 2008