Institutional Pressures of CSR: Evidence from India

The aim of the study is to identify important pressures driving corporate social responsibility practices by companies in India. Institutional pressures are conceptualised as a second order hierarchical component model (reflective formative type) and its multiple dimensions as first order reflective constructs forming second-order components. The model has its roots in institutional theory as applied to CSR - linking government, industry association, NGOs, media, local communities and peer groups for overall institutional pressure. An exploratory study was conducted among top level and middle level CSR professionals of companies in S&P BSE 500 index to empirically validate constructs. Partial least squares structural equation modelling (PLS-SEM) was used for data analysis. Results indicate that institutional pressures can be measured in a reflective-formative type HCM (Hierarchical Component Model) and local community, peers, media, government and industry association are important triggers for CSR in India. The study provides evidence that stakeholders are exerting institutional pressures on corporates. Further, a new multidimensional institutional pressures formative construct to understand the pressures of CSR in India is proposed which can be used by researchers in future, by policy makers to make sector specific CSR laws and by corporates for training. Abstract


Introduction
In order to harness project management expertise of the corporates to create value in the development sector, the Companies Act, 2013 has made corporate social responsibility (CSR) obligatary on many listed companies operating in the country (Jain, Aguilera & Jamali, 2017). Researchers have tried to understand the drivers of CSR in India -under non-mandated spending era (Arevalo & Aravind, 2011;Dhanesh, 2015). None of the empirical studies on CSR drivers in India has used any theoretical framework. Researchers across the world are increasingly using the framework of institutional theory (Brammer Jackson & Matten, 2012) to understand institutional pressures due to which organisations are undertaking CSR activities. This study uses framework of Institutional theory Keywords: Institutional theory, CSR, India, PLS-SEM (Campbell, 2007;DiMaggio and Powell, 1991;Scott, 2001) to unravel the types of pressures that firms in India are experiencing now to engage in CSR. Accordingly, the objectives of present study are twofold: a.Conceptualizing and proposing a new multidimensional institutional pressures constructs to understand the drivers or pressures of CSR in India. We have adopted a reflective-formative type Hierarchical Component Model (HCM) to do so. b. Empirically validating the formative constructs of CSR. It uses institutional theory to understand the drivers of CSR. Such a study is important to know which stakeholders are exerting institutional pressures on the corporate and how expenditure made on CSR activities by the corporates can ensure better social impact. At the methodological level by using a multivariate data analysis technique (PLS-SEM), a new multidimensional institutional pressures formative construct has been conceptualised and proposed which can be used by researchers in future, by policy makers to make sector specific CSR laws and by the corporate for training for employees. At the empirical level, the study discusses the drivers of CSR and emphasis on the need to collaborate with various institutions exerting pressures in India. Previous studies on drivers of CSR have not been rooted in any theoretical framework and S&P BSE 500 companies have never been taken as a population to draw a sample for studying issues related to CSR.
To achieve the objectives of the study, the paper is structured as follows. First, based on institutional theory, we develop theoretically grounded formative constructs -institutional pressures of CSR-manifested through its various dimensions as reflective constructs with the help of SMART PLS-3 software. This study contributes to the literature on institutional pressures on CSR at appropriate context that authors find suitable. Next, we discuss our methods and subsequently, report the results of PLS-SEM estimations.

Evolving Concept of CSR in the literature a) CSR in India :Guiding Force
Many researchers have adopted notion of dharma or "righteous conduct" to understand reasons for companies taking up CSR (Das, 2009;Dhanesh, 2015;Sharma & Talwar, 2005). According to a study by Pricewaterhouse Coopers (2013), the focus and form of CSR in India has moved from philanthropy to strategic.

b) Institutional Theory:
Institutional theory proposes that the corporate behaviour including its social decision-making depends on the institutions within which it operates (Angus-Leppan, Metcalf & Benn, 2010;Jones, 1999) and creates institutional pressures (Campbell, 2007) for CSR. Organisations in order to gain legitimacy, responde to institutional pressures (Carpenter & Feroz, 2001) and net effect is homogenization of organizational structures or isomorphism (DiMaggio & Powell, 1983).

C) Sources of Institutional pressures:
The present study is using DiMaggio's and Powell's idea of isomorphism. Our framework consists of six institutional pressures, as identified by Campbell (2007), influencing CSR behaviour-Government, industry associations, media, NGOs, local community and peers. As suggested in Figure 1, the proposed framework of the present study examines the relationship of institutional pressures of CSR with its every individual constituent to determine the extent of influence of each pressure group.

(i) Local community
Local community is the community residing in geography in which the corporate is located/ operates ( Galaskiewicz, 1997;Marquis, Glynn & Davis, 2007). These communities have acquired the ability to exert pressure on the corporate (Husted, Jamali & Saffar, 2016;Imbun, 2007).

(ii) Peer pressure
The degree of competition in an industry has a significant effect on CSR policies of the firms (Adeyeye, 2011;Dupire & M'Zali, 2016). Managers prefer to shared frames of references and imitate the activities of the peers in order to cope up with uncertainty in business environment (DiMaggio & Powell, 1983).

(iii) Media
Campbell's (2007) have argued that firms whose behaviour is monitored closely by media are more likely to behave in socially responsible ways. The influence of media as an important institutional pressure determining CSR behaviour of firms (Aharonson & Bort, 2015;Chandler, 2014).

(v) Government
In order to promote and encourage responsible business practices, governments over the last two decades, have become important institutional pressures driving CSR policies of companies (Knudsen & Brown, 2015) (vi) Industry Association Business associations or industry associations are defined by Barnett (2013, p. 214) as ''pool their resources and coordinate their efforts so that they may 'speak with one voice' on matters of shared interest''. They are important actors from institutional perspective, as they are means for industry self-regulation (Campbell, 2007;Wooten & Hoffman, 2016).

Research Methodology Method
Sample and Data Sources: Purposive sampling was used to collect data for the study. Top level and middle level of managers of the companies in S & P BSE 500 index (as on 31st December, 2016) handling CSR portfolio for their respective companies were contacted. To overcome access barriers, the target executives were contacted through a professional networking site (LinkedIn) and approached at various CSR conferences as well. Questionnaires were personally administered to executives based out of Delhi NCR, Mumbai, Pune, Hyderabad and Kolkata. Respondents from other cities filled the questionnaires mailed to them after purpose of study was explained to them in detail during telephonic conversation. The electronic survey collection method, though novel, is in line with such methods being adopted world over (Hair, Hult, Ringle & Sarstedt, 2017). Out of 507 executives contacted, 160 usable responses were collected after at least two reminders and follow ups in most of the cases. The results indicated that such difference does not exist between both the groups.
To determine the required sample size for the application of PLS-SEM, the often citied 10 times rule was applied (Barclay, Higgins & Thompson, 1995). The largest number of arrows pointing to the construct is six. Thus, 6x10=60 represents the minimum sample size for meaningful analysis.

Questionnaire Development
The authors, while conducting literature review, did not come across any instrument directly measuring external institutional pressures of CSR. The scale developed by Zuo, Schwartz& Wu (2017) was adapted for the purpose of our study. Six academic experts and ten experts from different industries heading CSR function of their respective organisations validated the questionnaire. Changes were made in the questionnaire as per suggestions received. Respondents were asked to record their responses on a seven-point Likert scale based on the extent of their agreement or disagreement with each of the statements.

Results and Discussion
Assessment of Institutional pressures as higher order formative construct(HOC) For measuring institutional pressures as HOC, content specification; indicator specification; and indicator collinearity were determined. (Diamantopoulos & Winklhofer, 2001). Figure 1 shows all the contents of the formative construct along with its six dimensions. As suggested by Gudergan, Ringle, Wende, and Will (2008), we applied the confirmatory tetrad test using CTA-PLS to determine empirically if data support the formative measurement of institutional pressures. Results of the CTA-PLS indicate that all the tetrads are significantly different from zero, thus, statistical support for formative nature of institutional pressures is obtained. To check the indicators' collinearity, Table 1 gives the institutional pressures' VIF values which are uniformly below the recommended threshold value of 5 ( Table 1)

Assessment of reflective lower order constructs
To assess reflective constructs, it is important to determine their internal consistency by calculating composite reliability; to evaluate convergent validity by considering the outer loadings of the indicators of constructs and the average variance extracted (AVE); and to establish their discriminant validity (Hair et al. 2017). Table 2 summarise the construct properties. All the LOCs are internally consistent as composite reliability for them exceeds the threshold limit of 0.7  Figure 2 shows that the outer loadings for most of the indicators is higher than 0.7, thus establishing indicator reliability. Four indicators with outer loading between 0.4 and 0.7 were not considered for removal as the composite reliability was above suggested threshold limit of 0.7 and AVE was also either more than 0.5 or very close to 0.5 for all these reflective constructs (Hair et al. 2017). All Heterotrait -Monotrait (HTMT) values, as can been seen in Table 3, are lower than the recommended limit of 0.85. To assess the statistical significance of path coefficient estimates, bootstrapping procedure was used (Hair, Ringle & Sarstedt, 2011). To ensure sufficiency, as per suggestion made by Preacher and Hayes (2008), 5000 subsamples were drawn. Confidence interval of the HTMT statistic does not include 1 in Table 4. Thus, HTMT results indicate absence of discriminant validity issues in reflective lower order constructs.

Approach to assessment of Model
To assess the reflective-formative HCM of institutional pressures, the present work follows repeated indicator approach wherein institutional pressures of CSR -a HOC -represents all the manifest variables of pressures exerted by the underlying lower -order constructs (LOCs) (Tenenhaus, Vinzi, Chatelin & Lauro, 2005). A graphical representation of the hierarchical components model in SMART PLS 3 is shown in figure 2. We find from Table 5 that all the drivers have similar and significant total effects on the formative construct of institutional pressures.

Conclusions
In order to integrate CSR into the ccorporate strategy, the practitioners need a metric which has empirical roots. It will help them to design CSR initiatives and also to train the people in the organisations. A feasible metric to determine which pressure groups are important has not been introduced by academia to assist the practitioners. This metric will also help the Government to introduce better CSR legislation rooted in the perceptual biases of CSR managers and subsequently, improve the efficiency of money being spent by the corporates on socio-development activities. The present work is a modest attempt in this direction. Our data analysis gave notable findings.
First, we found support for measuring institutional pressures of CSR in a reflective-formative type hierarchical component model (Ringle, Sarstedt & Straub, 2012). Second, the multidimensional approach to measurement of institutional pressures of CSR helps us to identify the key pressures. We found that all the reflective constructs representing pressure groups' have significant total effect and, thus, have "equal relevance for forming the HOC" (Hair, Sarstedt, Ringle & Gudergan, 2018, p.62). Total effect is nothing but the path coefficients. Thirdly, the study revealed that the indicators representing drivers -local community (LC1 and LC2); peer pressure (P1, P2 and P3); media (M1, M2 and M3); government (G1, G2 and G3) and industry association (IA1, IA2 and IA3) are statistically significant at 0.001% level. These findings have important implications. These may be used by stakeholders to play a bigger role in the present mandated CSR era as they can, based on our study, have a fair idea of their role in influencing CSR decisions of the corporates.
In our study indicator named LC3 ("Local communities do not have much influence on my company's recruitment and vendor selection policies") is not significant even at 10% level though LC1 and LC2 are significant and local community as a whole is an important pressure group. The pressure exerted by local communities differs from industry to industry depending upon businesses' embeddedness in the community. Consequently, firms in different industries adopt different CSR approaches (Boutin-Dufresne & Sacaris, 2004). The Act requires CSR initiatives to be carried out in local area and does not make any industry wise distinction. Our findings suggest that law makers need to relook at this requirement to accommodate practitioners' views and should take into cognizance of the fact of different degree of embeddedness of industry in local community before drafting legal provisions.
Next, we examine significance of peer pressure. Sood and Arora (2006) have shown that peer pressure is one of the important pressures driving CSR in India. Our results indicated the same.
Media plays an important role in providing visibility to CSR initiatives (Prabhakar & Mishra, 2013). Our sample consists of top companies in India. Consequently, media visibility is important to them. Hence, media's pressure is also found to be statistically significant.
Results show that NGO dimension is not statistically significant. This can be attributed to the Companies Act, 2013 which has envisaged NGOs as implementation partners . It is important, therefore, for corporate managers to look into capacity building for NGOs especially those working at the grass root and local level, material/ financial resources; and ensuring organization and project sustainability.
The Governments across the world are encouraging corporates to take up CSR activities to mitigate to some extent social and environmental problems caused by the business (Midttun, 2005).To make greater impact, CSR initiatives of the corporates are being aligned with Government developmental goals and partnerships entered into.
There are more than fifty industry associations in our country. Industry association is an important pressure group for CSR in India as revealed by data analysis. This is because these associations have been playing 'developmental' role (Sinha, 2005). Thus, data analysis results of the study have statistical, empirical and literature support.

Limitations and future research
Our study has certain limitations which future researchers can take as the starting point for their work. Data for our work are collected from the cross section of the industry. To validate our institutional pressure measure, the researchers in future can carry industry specific research by using the instrument in different settings. Due to time and financial constraints our sample size, though sufficient, was small. PLS-SEM is capable of and suitable for analysis of small sample sizes (Henseler, Ringle & Sinkovics, 2009 1.017 G1 -Government at all levels -local, State and Centre -helps us to identify CSR activities our company should undertake.
1.205 G2 -Our CSR activities are aligned with the social and development goals of the Government.
1.168 G3 -If any company fails to engage in proper CSR behaviour as desired by the Government, it may influence its business interests adversely.
1.037 IA1 -Industry association of which our company is a member has a common CSR agenda for its members. 1.071 IA2 -Industry association(s), of which our company is a member, influences our CSR activities.
1.199 IA3 -Our company takes up CSR initiatives to ensure adherence to industrial association's focus area. 1.160 Source : Authors' calculations by using SMART-PLS 3