Corporate Social Responsibility Disclosure in Small and Medium-sized Entities and Large Companies

Purpose – Corporate Social Responsibility (CSR) literature has focused mainly on larger firms. Only recently has discussion of the engagement of Small and Medium-sized Enterprises (SMEs) in CSR emerged in research studies. Here we contribute to that


Introduction
Corporate Social Responsibility (CSR) is "an umbrella term to describe the complex and multi-faceted relationships between business and society [as a way] to account for the economic, social, and environmental impacts of business activity" (Jamali et al., 2015, p.1).Despite a large volume of research on CSR and small and medium sized firms (SMEs), little is known about the type and extent of social responsibility reporting in SMEs (Lepoutre and Heene 2006;Morsing and Perrini 2009;Baumann-Pauly et al., 2013).This shortcoming seems to have arisen because of the premise that CSR predominantly affects, and has been developed in, large corporations (Jenkins, 2004).Such a view is understandable because larger firms have greater operational impact, attract greater media interest and public attention, and have the necessary resources to engage in CSR (Graafland et al., 2003;Lepoutre and Heene, 2006;Baumann-Pauly et al., 2013).For their part, SMEs are "frequently seen as a problem within the CSR debate because of their failure to engage with it" (Jenkins 2004, p. 52).We adopt the definition of SME that is used in the EU: that is, enterprises which employ fewer than 250 persons and which have an annual turnover not exceeding EUR 50 million, and/or an annual balance sheet total not exceeding EUR 43 million (EC, 2003).
Large firms differ from small firms in the way they view their social responsibility.This difference affects the content, nature and extent of their CSR activities.Recent research has concluded that it is not possible to understand CSR in SMEs through the simple application of methods applied in large corporations (Spence and Lozano, 2000;Graafland et al., 2003;Jenkins, 2004;Morsing and Perrini, 2009).Such research highlights the need to study CSR in SMEs to overcome the limited amount of quantitative and qualitative information available regarding CSR in SMEs (Perrini et al., 2007).In particular, comparative research involving SMEs and large companies is scarce (Blombäck and Wigren, 2009;Vo, 2011;Baumann-Pauly et al., 2013).
We contribute by investigating whether there are significant differences in CSR disclosure (CSRD) practices between SMEs and large Portuguese companies.
Additionally, we describe the CSR practices undertaken by SMEs and large companies with a view to identifying the most important targeted stakeholder.Using Stakeholder Theory and a multi-stakeholder approach, we seek to determine whether different business characteristics (such size, industry affiliation, third party assurance verification and profitability) affect CSRD.We find empirical support for the proposition that SMEs are not less advanced than large companies in CSRD practices.
We find also that the factors of size, industrial affiliation, and report assurance, explain different business approaches to CSR engagement and disclosure -but that levels of profitability do not.
In the following section we demonstrate the importance of SMEs in the Portuguese economy.Section 3 uses a stakeholder theory perspective to present a theoretical framework of the relationship between SMEs and CSRD.Section 4 develops hypotheses.Section 5 describes the research method.Section 6 discusses results.

The importance of SMEs in Portugal
SMEs are the main force driving European economic growth.In the European Union [EU] in 2011, they represented 99.8% of all companies, accounted for two out of every three jobs (67.4%), contributed 58.1% of the total value added by non-financial business economy, and were responsible for 29.6% of employment and 21.2% of value added (EC, 2013).In 2011, 92.2% of SMEs were micro-enterprises (with less than 10 employees) (EC, 2013).
SMEs are a key component in the Portuguese economy and one of the most important sectors of society.They are recognized broadly as the primary driver of economic development and as a powerful promoter of employment, social cohesion and quality of life (Spence and Rutherfoord, 2003;Abreu et al., 2005;Santos, 2011).
They account for 99.9% of all companies, 78.3% of employment and 67.6% of value added -much higher proportions than the EU average (EC, 2013).In Portugal, microenterprises represent 94.1% of all companies and account for 39.2% of total employment (ten percentage points above the EU average) (EC, 2013).
SMEs are less oriented to the services sector in Portugal than in all other EU countries.SMEs in the manufacturing, construction and trade sectors contribute to a higher percentage of employment and added value than in all EU countries (EC, 2013).Some of these differences are explainable by cultural and institutional preferences for self-employment and/or family-run businesses.
CSR reporting practices are voluntary in Portugal.Nonetheless, larger Portuguese companies have been found to be in the "leading pack" quadrant in an international evaluation of the quality and maturity of business CSR communications (KPMG, 2011).

Theoretical Framework
Although the literature regarding CSR in SMEs has been growing recently, it is fragmented and underdeveloped (Vo, 2011), with no "coherent theory" (Lepoutre and Heene, 2006, p. 257) and no "consolidated and generally accepted model to investigate [CSR/SME] relationships" (Russo and Perrini, 2010, p. 209).These omissions are undesirable because SMEs have very strong social importance and impact in the global economy.SMEs create employment opportunities, drive economic growth, promote private sector development, encourage equitable distribution of income (Inyang, 2013), and promote innovation and entrepreneurial spirit (Wilkinson, 1999).
The aggregate investment of SMEs in CSR is significant, and should not be underestimated (Quinn 1997, Jenkins, 2004;2009).
SMEs are not merely miniatures or clones of large companies (Tilley, 2000;Williamson et al., 2006).They differ intrinsically in terms of nature, practices and extent of CSR activities (Spence and Lozano, 2000) and in "the amount of resources available, strategies, drivers, importance of managerial values, level of involvement and stakeholder prioritization" (Coppa and Sriramesh, 2013, p. 32).SMEs differ among themselves too.They are heterogeneous in size, mission, industry sector, national context, historical development and institutional structures (Jenkins, 2004;Perrini et al., 2007).
Prior studies of barriers to the development of CSR activities by SMEs identified the constraints of time and financial resources (Vives, 2006) and diminished levels of human, technical and organizational resources to implement social responsibility policy (Lepoutre and Heene, 2006;Vo, 2011).There is strong support for the view that SMEs implement CSR policies and practices as a consequence of an owner/manager's ethical orientation, personal values, attitudes and choices (Quinn, 1997;Tilley, 2000;Jenkins, 2004;Jamali et al., 2009;Nielsen and Thomsen, 2009;Vo, 2011).Thus, owner/managers are found to be the change agents, often influenced by family pressure (Jenkins, 2004;Vives, 2006).They have the opportunity and power to carry out socially responsible activities.
Other factors influencing the commitment of SMEs to CSR include the desire to develop community and employee relationships, and company image and reputation (Spence et al., 2003;Jenkins, 2006;Russo and Perrini, 2010).SMEs also engage in CSR to promote a climate of trust with internal stakeholders (Graafland et al., 2003;Perrini, 2006); and to comply with related CSR legislation (Murillo and Lozano, 2006;Williamson et al., 2006).SMEs may also wish to promote long-term survival and business performance (Spence et al., 2003;Williamson et al., 2006); and/or to reduce costs and increase efficiency (Jenkins, 2009;Russo and Tencati, 2009).
Stakeholders who are "closer" to a firm have the most influence on SMEs.Usually, their interests are accorded priority when deciding CSR activities (Jenkins, 2004(Jenkins, , 2006;;Murillo and Lozano, 2006;Lepoutre and Heene, 2006;Vives, 2006;Spence, 2007;Perrini et al., 2007).Thus, the local community, environment, employees and customers are the key stakeholders for SMEs.They are the principal driver for SMEs to engage in CSR.The interests of each of these stakeholders are discussed, in turn, below.
Customers are a central stakeholder for any company, large or small.Continued loyalty of customers is vital for long-term survival of SMEs (Jenkins, 2004(Jenkins, , 2006;;Murillo and Lozano, 2006;Vives, 2006;Lepoutre and Heene, 2006;Spence, 2007).Customers are intuitive and natural stakeholders.In recent decades, they have increased their demand for socially responsible behavior by business entities (Santos, 2011).
Employees are fundamental stakeholders for SMEs too (Spence and Rutherfoord, 2003;Longo et al., 2005;Brammer and Pavelin, 2006a;Vives, 2006;Perrini et al., 2007;Jenkins, 2009).They are crucial to the business success of SMEs (Vo, 2011).Through CSR, SMEs can increase their ability to attract and retain better employees (Branco and Rodrigues, 2006).In smaller firms, because of the opportunity for closer labor relations, the dialogue between managers and employers is often continuous and informal.
The Community is another important stakeholder for SMEs (Longo et al., 2005;Brammer and Pavelin, 2006a;Perrini et al., 2007;Santos, 2011).SMEs have a very strong connection with their local community (Vo, 2011;Spence, 2016).Often, the owner-manager and family, employees and customers, all live in the same community.
SMEs provide goods and services, create jobs and help develop the local economy.
SMEs fulfill their CSR obligations to the local community by promoting projects that improve the quality of life of citizens and increase social and economic development (Jenkins, 2006).
Concern for Environment is also an important factor encouraging SMEs to engage in CSR activities.Many SMEs and companies with larger environmental impacts (such as printing, manufacturing and engineering firms) are committed to environmental improvements (Sharma, 2000Williamson andLynch-Wood, 2001;Williamson et al., 2006).Through the implementation of environmental practices, they can promote substantial reductions in waste and reduce costs (Mathieu and Reynaud, 2005).
CSRD is a direct expression of companies' commitment to social and environmental subjects (Calace, 2014).Stakeholders demand different information from business entities and thus impel CSR reporting practice (Gray et al., 1995;Deegan, 2006).Interest regarding a specific issue can differ between each stakeholder or between different companies (Freeman, 1984).Thus, CSRD in SMEs can differ considerably from CSRD in large companies (Jenkins, 2004(Jenkins, , 2006(Jenkins, , 2009)).Large firms use formal instruments to organize CSR activities, such as codes of conduct, certification procedures, and report guidelines (Graafland et al., 2003;Russo and Tencati, 2009).SMEs usually have less developed CSR strategies and are prone to make greater use of informal channels to communicate CSR activities (Graafland et al., 2003;Murillo and Lozano, 2006;Spence, 2007;Nielsen and Thomsen, 2009;Baumann-Pauly et al., 2013).
Stakeholder theory has been used to understand the engagement of SMEs in CSR activity (Graafland et al., 2003;Jenkins, 2004Jenkins, , 2006;;Vives, 2006;Murillo and Lozano, 2006;Perrini, 2006;Spence, 2007).Although stakeholder theory was developed primarily for large firms, it can apply to all types of companies (Jenkins, 2004).All companies have economic, legal, environmental and social responsibilities.
To promote survival, SMEs must consider the needs of their most important stakeholders and obtain their support.Intuitively, stakeholder theory is appealing for analysing CSR in SMEs and large companies (Jenkins, 2006).This is because both of these types of entity have similar stakeholder relationships and the same purposes for stakeholder management: risk reduction and company image improvement.Like large companies, SMEs perceive potential benefits of engaging in CSR.They try to manage the expectations of their stakeholders.However, the way this occurs can differ distinctly t between SMEs and large companies (Jamali et al., 2009).
Stakeholders can be divided into primary stakeholders (those "without whose continuing participation the corporation cannot survive as a going concern"); and secondary stakeholders ("those who influence or affect, or are influenced or affected by the corporation, but they are not engaged in transactions with the corporation and are not essential for its survival" (Clarkson 1995, p. 106).SMEs must have immediate concern for their primary stakeholders.These concerns may, or may not, differ from those of high-priority stakeholders identified by large companies.The question arising is: "To which stakeholder group(s) should companies respond?" Although some important stakeholders are common to SMEs and large companies, SMEs are likely to be influenced differently by stakeholders, especially those who are, or are not, vital for large companies.Different stakeholder groups present different and often conflicting needs.The interests and requirements of the same stakeholder may differ between large and small companies.For this reason, SMEs and large companies may assign different attributes to the same stakeholder, depending on the importance of each stakeholder to them.This brief overview highlights the need for comparative studies on CSR and CSRD between SMEs and large companies.Blombäck and Wigren (2009)  These studies do not produce convincing data about the actual implementation of CSR.
To capture the level of implementation of CSR in companies it is necessary to develop an assessment framework that accounts for the specific and distinct characteristics of small and large firms (Russo and Perrini, 2010;Baumann-Pauly et al., 2013).For this purpose, we develop a framework to evaluate the CSRD of SMEs by identifying CSR indicators (of customers, employees, environment, community and society) that are applicable to firms of all sizes.

Hypotheses
Researchers are not unanimous in their view of the influence of firm size on CSRD (Lepoutre and Heene, 2006).A focus only on size as a determinant of CSRD is insufficient (Baumann-Pauly et al., 2013) and "runs the risk of stimulating an unnuanced CSR discourse" (Blombäck and Wigren, 2009, p. 255).The present study provides nuance by looking through the dual lenses provided by SMEs and large companies, with a view to improving understanding of CSR practices.We add nuance by analysing the relationship between CSRD and business characteristics such industry affiliation, third party verification, and profitability.

Size (H1)
Large companies, especially listed companies, attract more attention from the general public, more extensive media coverage, and greater disclosure pressure from prominent stakeholders (Deegan and Gordon, 1996;Archel, 2003;Haniffa and Cooke, 2005;Artiach et al., 2010;Branco et al., 2014).Thus, as a consequence of greater public presence, large companies devote more attention to transparency and accountability (Jenkins, 2004).They understand the relevance of identifying the relationships with multi-stakeholders (Perrini, 2006).Increasingly, large companies have sufficient resources to establish a CSR committee, contract experts in business communication, and use CSRD to promote their image and avoid stakeholder pressures.
Despite SMEs generally being less visible and subject to less pressure from stakeholders, they have some characteristics (such as flexibility, adaptability and less hierarchical management structures) that help them engage in CSR activities (Jenkins, 2009).Through their smaller size, and their closeness to stakeholders, SMEs are wellplaced to establish informal relations with stakeholders (Nielsen and Thomsen, 2009) and targeted CSRD (Murillo and Lozano, 2006;Perrini, 2006;Spence, 2007).CSRD in SMEs can be seen as a tool to improve image (Longo et al., 2005;Jenkins, 2006).In Portugal, obligations to engage in CSRD do not depend on size.We test the following hypothesis: H1: Firm size is related to the level of CSRD.

CSRD Assurance (H2)
Assurance of CSRD enhances information credibility and is an important driver of the quality of CSRD (Simnett et al., 2009;Kolk and Perego, 2010;Pflugrath et al., 2011;KPMG, 2013;Branco et al., 2014;Gomes et al., 2015).Companies with higher visibility want to demonstrate to stakeholders that their CSR information is credible, and is not presented merely as a marketing exercise.The CSR reporting and assurance practices of Portuguese companies are well-integrated and widespread (KPMG, 2011(KPMG, , 2013;;Branco et al., 2014;Gomes et al., 2015).But these prior Portuguese studies do not consider SMEs.Here we address this shortcoming by investigating whether voluntary assurance is a factor that influences the quality of CSRD by SMEs and large companies.
We test the following hypothesis: H2: The assurance of the CSR reports is related to the level of CSRD.

Industry Affiliation (H3 and H4)
The industry sector in which a firm operates is important in understanding its CSRD (Hackston and Milne 1996;Archel, 2003;Jenkins, 2004;Branco and Rodrigues, 2008;Sweeney and Coughlan, 2008;Branco et al., 2014).Companies operating in industries "closer to the consumer" exercise greater care with their public image, since they are more visible, and are exposed to stronger stakeholder pressure and expectations.
Companies whose activities have greater environmental impact are especially concerned about environmental matters (Jenkins, 2006;Perrini et al., 2007;Avram and Kühne, 2008).These companies are pressured by specific stakeholders tp provide them with information about the magnitude of these impacts.The management of environmental aspects (including by SMEs) offers the possibility of reducing costs, developing innovation, and enhancing competitive advantages (Sharma, 2000Williamson andLynch-Wood, 2001;;Mathieu and Reynaud, 2005).
The effect on CSR of two proxies related to company visibility (proximity to consumers, and environmental sensitivity) has been studied by Branco and Rodrigues (2005;2008) and Dias et al. (2016) in reports of Portuguese listed companies.Their results reveal that industry classification based on "consumer proximity" explains different CSRD practices.Regardless of size, and depending on the nature of business, companies in the same (or similar) industries experience similar pressures and develop similar stakeholder management strategies.We contribute to understanding by testing the effectiveness of "consumer proximity" and "environmental sensitivity" in a sample that includes large companies (listed and non-listed) and SMEs.We test the following two hypotheses: H3: Consumer Proximity is related to the level of CSRD.
H4: Environmental Sensitivity is related to the level of CSRD.
Other empirical studies have concluded that profitability does not appear to be a significant determinant of CSRD (for example, Archel, 2003).We test the following hypothesis: H5: Profitability is related to the level of CSRD.

Sample and data
Initially, we selected all Portuguese companies which we could identify as having issued a CSR report in 2011.These companies were identified by systematic searches of the following company databases: BCSD Portugal (www.bcsdportugal.org),GRI Sustainability Disclosure (www.globalreporting.org)and the Corporate Register (www.corporateregister.com).We excluded eight large financial companies so as to not distort comparability between large companies and SMEs because there were no financial companies among the SMEs.The remaining sample comprised 57 reports on CSR activity by 13 Portuguese SMEs (unlisted companies) and 44 large Portuguese companies (19 listed and 25 unlisted) in 2011.CSRD data were collected from annual "combined reports" i (13) and stand-alone reports (44).Information pertaining to financial statements was obtained from annual reports and from the Sabi -Bureau van Dijk Database (www.bvdinfo.com).

Dependent Variable
CSRD was measured using a disclosure index: that is, "a series of pre-selected items which, when scored, provide a measure that indicates a level of disclosure in the specific context for which the index was devised" (Guthrie and Abeysekera, 2006, p. 118).Content analysis (Krippendorf, 1980) was used to codify written text into various groups or categories in terms of CSR dimensions previously identified.Such a framework was sufficiently similar to allow a reasonable comparison between large companies and SMEs (Baumann-Pauly et al., 2013).
To capture the extent of CSRD in each dimension, we identified the items most frequently used by researchers of CSRD in SMEs (Perrini et al, 2007;Balluchi and Furlotti, 2013;Madueño et al., 2016).The dimensions we analysed are wellestablished in research on CSRD in SMEs.They are applicable to all companies that disclose CSR matters, regardless of size.We devised a CSRD checklist with 25 CSR indicators partitioned into the five dimensions mentioned (see Table 3).
Based on the multi-stakeholder perspective, we consider these five dimensions reflect the principal CSR strategies of SMEs and large firms.As with previous similar studies (Hackston and Milne, 1996;Archel, 2003;Haniffa and Cooke, 2005;Branco and Rodrigues, 2008;Dias, 2009;Dias et al. 2016), the level of voluntary CSRD was measured using a disclosure index.We calculated a general index, and five CSRD indexes by CSR dimension, as follows: If a company reported all CSR items, the maximum score obtainable is 25 (10 for environmental disclosure, 3 for customer disclosure, 5 for employee disclosure, 3 for community disclosure, and 4 for society disclosure).
Consumer Proximity (ConP) was assigned the value 1 in the case of a consumer proximity company (household goods and textiles, beverages, food and drug retailers, telecommunications services, electricity, gas and water distribution), and 0 otherwise (Deegan and Gordon, 1996;Archel, 2003;Branco and Rodrigues, 2008;Dias et al., 2016).
Environmental Sensitivity (EnvS) was assigned the value 1 if company activities have strong environmental impact (mining, oil and gas, chemicals, construction and building materials, forestry and paper, steel and other metals, transport, electricity, gas and water distribution), and 0 otherwise (Hackston and Milne, 1996;Deegan and Gordon, 1996).
Assured Report (AssuR) was assigned the value of 1 if the company has its CSR report assured, and 0 otherwise.
To test the relationship between explanatory variables and CSR disclosure we estimated the following linear multiple regression model:

Descriptive Statistics
Table 1 shows the frequencies of the categorical variables by company type.Of the 57 firms studied, 52.6% have their CSR report audited, 36.8% are closer-toconsumer firms, and 70.2% have high environmental impact.In the SME group, only one SME does not belong to industries considered more environmentally sensitive.
In Customers, Community and Society dimensions at least one company did not report any indicator (minimum 0).Table 3 reports the frequency of disclosure for each of the 25 indicators comprising the CSRD indexes, by company type. - Insert Table 3 about here - -------------------------------------The total CSRD index obtained (0.74) is high when compared with the index of less than 0.50 reported in previous studies on CSRD in Portugal (Branco and Rodrigues, 2008;Dias, 2009;Dias et al., 2016).However, this result is not comparable since unlike other studies, the sample only considers companies that reported CSR issues.SMEs (0.71) reported slightly less information than large companies (0.75).
Overall, the employee dimension of CSR is the most reported (0.94), followed by customers (0.77) and environment (0.76) dimensions.Matters related to community (0.61) and society (0.50) are less reported.Only one single indicator ("workforce by employment type, contract or category") was reported by all companies.The indicators "packaging concerns" (0.40), "incidents related to discrimination" (0.42), "procedures for local hiring" (0.47) and "information related to public policy positions" (0.49) were disclosed by less than half of the companies.
To verify whether there are significant differences between groups, a one-way analysis of variance was performed, with the disclosure indexes as dependent variable and company type (large companies versus SMEs) as a factor.The results (Table 4) reveal no significant differences between CSRD and company type in any of the indexes, suggesting that SMEs and large companies have similar CSR reporting patterns.

Multiple Regression Analysis
To test for multicollinearity in the regression models we used Tolerance and Variance Inflation Factors.This revealed an absence of multi-collinearity in the regression models.iii All CSRD indexes have different significant variables.Table 6 summarises the results of the multiple regression analysis. - For ID Environment (R2 is 0.372, adjusted R2 is 0.310), three variables are statistically significant: Size (Size) at the 0.01 level, andd Assured Report (AssuR) and Environmental Sensitivity (EnvS) at the 0.05 level.
Regarding ID Customers (R2 is 0.435, adjusted R2 is 0.380), two variables are statistically significant: Assured Report (AssuR) and Consumer Proximity (ConsP), both at the 0.001 level.The same two variables are statistically significant with Community (R2 is 0.369, adjusted R2 is 0.308): Assured Report (AssuR) at the 0.001 level and Consumer Proximity (ConsP) at the 0.01 level.The model for ID Employees (R2 is 0.181, adjusted R2 is 0.101) presents the only positive significant variable: Assured Report (AssuR) at the 0.01 level.As R2 is low (0.181), the model does not reveal statistical significance.This can be justified on the grounds that when any company decides to published a CSR report, its employees are a natural and evident stakeholder in all companies.Thus, the indicators considered in ID Employees are those most disclosed.
For ID Society (R2 is 0.363, adjusted R2 is 0.301) three variables are statistically significant: Size (Size) at 0.001 level, Assured Report (AssuR) and Consumer Proximity (ConsP) both at the 0.01 level.
Finally, for CSRD global index (R2 is 0.473, adjusted R2 is 0.421), three variables are statistically significant at the 0.001 level: Size, Assured Report and Consumer Proximity.
Across all CSRD dimensions, there is a positive relationship between assured reports and the level of CSRD.This finding accords with previous research (Simnett et al., 2009;Kolk and Perego, 2010;Faisal et al., 2012;Branco et al., 2014;Gomes et al., 2015) and supports H2.Companies whose CSR report is assured, disclose more information than those with an unassured report.In addition to promoting the credibility of the reported information (Pflugrath et al., 2011, Branco et al., 2014), the assurance of CSR reports is a means to increase business image and reputation (Simnett et al., 2009;Branco et al., 2014).Size (Size) shows a positive and significant influence on ID CSRD, ID Environment and ID Society, consistent with previous research (Hackston and Milne, 1996;Haniffa and Cooke, 2005;Perrini et al., 2007;Michelon and Parbonetti, 2012), supporting H1.This suggests that larger companies report more CSR information (ID CSRD), especially in the Environment and Society dimensions.Large companies have human and financial resources to implement environmental management and reporting systems.
At the same time, due to their visibility, large companies communicate to a wider audience and adopt disclosure strategies to conform to society's expectations (Deegan, 2002) and promote their social image.
The absence of significance between Size and the remaining CSRD dimensions (Customers, Employees and Community), together with the lack of significant differences between SME and large companies on CSRD, suggests similar behaviour in these CSR dimensions among different sized companies.Customers, Employees and Community are the company stakeholders that deserve the same attention both for SME and large companies (Preuss and Perschke, 2010;Madueño et al., 2016).
These results accord with stakeholder theory.Despite the differences between SMEs and large companies, both dedicate similar attention to the most important stakeholders and use CSRD to enhance their reputation and public image.However, the resources applied in stakeholder management, and the relationship that is established, is naturally distinct in small and large companies (Jamali et al., 2009).
Stakeholders considered "primary" to large enterprises can be considered by small companies as secondary.Our results show that Environment and Society are primary stakeholders to large companies, but are "secondary" stakeholders to smaller companies.
Consumer Proximity (ConsP) is significant in ID CSRD, ID Customers, ID Community, and ID Society.These results are consistent with those obtained by Hackston and Milne (1996), Haniffa and Cooke (2005), Branco and Rodrigues (2008), Sweeney and Coughlan (2008), Faisal et al. (2012), and Dias et al. (2016).They support H3 partially and suggest that companies closer to consumers (except in environment and employee matters) report more CSR information.Companies closer to consumers have, as immediate stakeholders, Customers, Community and Society.Due to their public exposure, these companies know it is vital to the success of their activities to maintain a good image.For that reason, they report CSR information which meets the expectations of the nearest stakeholders.
Environmental Sensitivity (EnvS) is significant only for ID Environment: that is, companies with higher environment impact disclose more environment information.Such results accord with previous studies of the impact of industry affiliation on environmental disclosure (Gray et al., 1995;Frost and Wilmshurst, 2000;Deegan et al., 2002;Brammer and Pavelin, 2006b).However, they are not supported by previous studies on CSRD of Portuguese companies (Branco andRodrigues, 2008, Dias et al., 2016).The divergent results are explained by the fact that, contrary to previous research, the sample used is 70% composed of companies considered to have high environmental sensitivity and includes SMEs and unlisted large companies.
Since differences were found in only one of the dimensions studied, H4 is not supported.Thus, we conclude that industry classification by Environmental Sensitivity does not explain differences in CSRD.However, these companies assume the environmental risks of their activities and develop environmental management systems that provide adequate information about CSRD.Thus, through CSRD they respond to stakeholders' wishes to know how companies manage environmental risks (Patten, 1992).
The industry affiliation results can be understood using a stakeholder theory perspective.The business sector in which a firm operates is an essential determinant of the main stakeholders of each company.Companies with high environmental impact report on more environment issues, while companies closer to consumers disclose more CSR information related to consumers, community and society.
Companies manage the expectations of their stakeholders and give priority to stakeholders who are closer to, or are more affected by, their activities (Blombäck and Wigren, 2009).
The variable Profitability (Prof) is not significant in explaining CSRD or any specific CSR dimension.Thus, it is not possible confirm hypothesis H5.

Conclusions
This study extends knowledge by comparing the CSRD practices of SMEs and large (listed and unlisted) Portuguese companies, using a perspective that takes into account the particularities, and other fundamental business characteristics, of SMEs.
The CSRD practices of SMEs and large companies in Portugal do not differ significantly.SMEs reporting on non-financial aspects of business performance (such as CSR) engage in a process that requires significant resources.However, CSRD is also a structured way of presenting a SME's CSR involvement, promoting reputation, and managing dialogue with influential stakeholders (Calace, 2014).
Information related to employees is the most reported item of CSRD, followed by information related to customers and environment.Community and society matters are the least reported.In the period in which the study was conducted (2011), companies prioritised CSR activities that were related directly to the business maintenance and achievement of economic results.Employees, customers and environment are immediate and natural stakeholders of all companies, and are incorporated into their daily management.
Regardless of size, Portuguese companies recognize the importance of establishing appropriate relationships with customers and the community.All firms seem to be aware of the importance of employees to organizational success.However, with regard to environment and society matters, SMEs (when compared with large companies) have a lower incidence of reporting.Larger companies disclose CSR matters to a wider audience and they are particularly advanced in making extensive environment and society disclosures.
Industry affiliation, especially when considered in the context of Consumer Proximity, has a great impact on CSRD.Companies closer to consumers disclose more information related to Customers, Community and Society.However, the industry classification based on Environmental Sensitivity only justified greater disclosure of Environmental issues.
A highly important result is the positive influence of assurance in all CSR dimensions analysed.Besides enhancing credibility, assurance of the CSR report drives system improvements and extended CSR disclosures.
We contribute to the literature by building a replicable assessment framework of common CSR practices in SMEs and large companies, using a multi-stakeholder perspective.The research method developed offers a promising starting point for future studies of CSRD in distinctly different-sized companies.Further, we show that smaller firms are not necessarily less advanced in organizing CSR than larger firms.
Although SMEs have characteristics that distinguish them from large companies, they have similar obligations to stakeholders and similar patterns of CSRD.We also show that different stakeholders have different levels of priority to large companies and SMEs.The assurance of CSR reports is a very strong driver of improved CSR communication.
The results must be read mindful that this is a small sample, cross-sectional study that considers companies from a single small European country.For this reason, the findings may not be readily generalizable.Future research could beneficially be directed to replicating this study in other countries and time periods with a view to verifying, refuting and/or extending the findings.Additionally, qualitative research though interviews would help to reinforce and extend the results.
argue the need to enhance comprehension by exploring the width of CSR and adopting the same theoretical framework to analyse CSR in small and large firms.Typically, existing assessment frameworks have been based on survey data in which results are often biased towards what is considered socially desirable (Baumann-Pauly et al., 2013).
by group (one for each CSR dimension / one general index) ej = Attribute analysis (1 if disclosure item is found, and 0 if not found) e = Maximum number of items (by CSR dimension or general).

-
All disclosure indexes are correlated positively with each other and with the variable Assured Report (AssuR).Size is correlated positively with ID CSRD, ID Environment and ID Society.Consumer Proximity (ConP) is correlated positively with ID CSRD, ID Costumers, ID Community and ID Society.Consumer Proximity and Environmental Sensitivity are correlated negatively.Profitability is correlated positively with Assured Report and ID Environment.