Corporate social responsibility and firm financial performance: A literature review

Use your smartphone to scan this QR code and download this article ABSTRACT This paper aims to investigate the literature on Corporate Social Responsibility (CSR) to provide a comprehensive overview of whether CSR would make a difference to organisational financial outcomes. The paper also provides a closer focus on CSR research in Vietnam. Through an extensive analysis of 86 most recent empirical studies from 2015 to 2020, we found that the contribution of CSR to firm financial performance has received significant support from the literature. Yet the overall findings are still inconsistent, and the majority of evidence is mainly from developed countries. The current literature on CSR and firmperformance highlights some important issues, ranging from theoretical background, CSRmeasures, methodological issues, the need to consider intervening factors in CSR-firm performance relationship, and the need to extend this literature further in developing and emerging countries. The literature on CSR-firm performance research in Vietnam closely resembles these problems. Research in this country domain is still scarce in both quantity and quality, reflecting in a number of issues including the limited number of international publications, the absence of theory-driven research, and the less rigorous research design. Building on these findings, we recommend future research to (i) adopt the multi-theoretical approach for a more extensive view on whether and how CSR contributes to firm performance; (ii) obtain more rigorous methodological approaches to measure a wide range of CSR dimensions and address the issue of endogeneity in CSR-firm performance causal relationship; (iii) open the Pandora box to explore why and through which channels CSR can improve firm financial performance with the presence of situational factors; and (iv) build the literature withmore evidence from different country contexts and from developing and emerging countries.


INTRODUCTION
Over the past few decades, Corporate Social Responsibility (hereinafter CSR) has become a growing topic of interest to organisations, communities, researchers, and policy makers. CSR has been defined in multiple ways. For instance, it was originally portrayed as " the social responsibility of business encompasses the economic, legal, ethical, and discretionary expectations that society has of organisations at a given point in time" [ 1 , page 500]. In 2011, CSR was redefined by the European Commission as a broader concept focusing more on the responsibility and duty aspect rather than the voluntary acts of organisations. In that sense, CSR is described as "the responsibility of enterprises for their impacts on society" [European Commission 2011, cited in 2 , page 5]. Another definition of CSR is "context-specific organisational actions and policies that take in to account stakeholders' expectations and the triple bottom line of economic, social, and environmental performance" [ 3 , page 855].
Due to its perceived importance to business and society, Corporate Social Responsibility has become a topic of interest in both research and practice. In modern trends, it is believed that CSR is an essential success factor for organisations and helps to build positive corporate image and reputation in the eyes of consumers and stakeholders, especially with the shift to focus more on the environment and community of the society in recent years 4 . There is a common belief that CSR not only helps companies to improve their reputation, but also brings great economic outcomes. Theoretically, research on the question "Is good ethics good business?", both conceptual and empirical, is not new. A number of studies have reviewed this literature, emphasising that the relationship between CSR and firm financial performance does exist [4][5][6][7] . However, it is also highlighted by these studies that up to date, the overall finding of CSR-firm performance relationship is still inconsistent at best. More importantly, empirical reviews on CSR and firm financial performance tend to focus more on evidence from developed countries, leaving this question unanswered in the context of developing and emerging markets 6 . Therefore, it is needed to conduct a comprehensive review of CSR and firm financial performance relationship with wider evidence from both developed and developing countries for a more thorough understanding of this literature in different economic and institutional settings. This paper sets out to provide an overview of the most recent CSR-firm financial performance literature which capture evidence from both developed and developing countries, and with a closer view on CSR research in Vietnam. To achieve this purpose, we applied the common review techniques to generate a sample of empirical and theoretical papers on CSR and firm financial performance from 2015 to 2020 and used this sample as the foundation for our literature review. We structured our study as follow . The paper starts with an overview of our sample, followed by a comprehensive review of empirical evidence on CSR and firm financial performance in the global context and in Vietnam. On this basis, the paper addresses a number of important issues in the current literature, including CSR management approaches, theoretical background, the role of intervening factors, and methodological issues in CSR-firm financial performance relationship. The paper concludes with some implications and directions for future research.

LITERATURE REVIEW
In this study we adopted the common techniques used in literature review study to generate our data. First, we conducted a wide search for empirical evidence on CSR and firm financial performance on the most common databases, including Web of Science, ProQuest, Science Direct, and Google Scholar with a combination of these key terms: "CSR", "Corporate Social Responsibility", "Firm financial performance", "Firm value". Second, we filtered our search to only include empirical papers that directly address the impact of CSR on firm financial performance and narrowed down our search within 2015-2020 to capture the most recent evidence on this literature. Third, we did not exclude evidence from lower ranked journals, as we aimed to gather more evidence from emerging and developing countries (with the assumption that more studies on CSR-firm financial performance in developing countries are published in these journals). To support our analysis, we also included some discussion papers, case studies and previous literature review papers in our search. Our data collection procedure has resulted in a sample of 93 international publications on CSR and firm financial performance. Of which 86 are empirical studies and 7 are literature review, discussion papers and case studies. We also conducted a search for international publications on CSR in Vietnam following these above steps. We purposely widened the time range and search topics to cover as many studies in this literature as possible. Our search came up with 49 international publications on CSR in Vietnam in total and three publications specifically addressing CSR-firm financial performance relationship. Table 1 provides a summary of our sample. The list of all empirical studies included in the sample can be found in Appendix 1: Summary of all empirical studies on CSR and firm financial performance). It can be seen that evidence of CSR-firm financial performance relationship is still heavily skewed towards US and developed countries. Evidence from developing and emerging countries, particularly Vietnam, is still limited. Table 2 provides an overview of the empirical findings on CSR and firm financial performance relationship. It can be seen that different aspects of firm financial performance have been examined in the literature, including accounting performance 8 , market performance 9 , or both 10 . Some studies also use qualitative measures to capture firm financial performance based on management perceptions 11 . Empirically, the relationship between CSR and firm financial performance has received strong support from the literature with 61 out of 86 studies confirming the positive impact of CSR on organisational financial outcomes. For instance, CSR is found to increase firm accounting performance [12][13][14][15] ; market performance [16][17][18] ; or both 10,19 .

CSR-firm financial performance evidence in a global context
However, despite a large number of studies supporting the business case of CSR, overall, this literature is still inconclusive. It is evident that the impact of CSR on firm financial performance is not statistically significant 4 , especially after controlling for endogeneity 20 ; or negative [21][22][23] ; or mixed, depending on different dimensions of CSR [24][25][26] .

CSR-firm financial performance evidence in Vietnam
Vietnam is one example of a country that provides a unique cultural and institutional setting that can bring prior relevant CSR theories and findings to the test. The country offers an interesting and promising research domain with its unique political context   51,52 and governance 53 , (6) government responsiveness and regulatory compliance 54 , and (7) the business case for CSR, which will be discussed in detail below. Though still scarce in quantity, the extant studies focusing on organisational-level consequences of CSR have provided evidence to support the business impact of CSR, such that CSR engagement can bring benefits to organisations in Vietnam, including enhancing employee behaviour [55][56][57] and customer behaviour [58][59][60][61][62][63][64][65] , improving firm efficiency 66 , promoting knowledge sharing 67,68 , increasing firm engagement to foreign markets 69 , increasing employees' productivity 23 , increasing supplier performance 70 , improving brand performance and customer service [71][72][73] , and strengthening organisational competitiveness 74 .
CSR literature in Vietnam also provides some supporting evidence on the positive impact of CSR on firm financial performance, including the studies of Hoang (2005) 75 , Kabir and Thai (2017) 76 , and Xie, Jia (2017) 77 . However, the three notable empirical studies on CSR-firm financial performance in Vietnam either limit their focus only on listed companies, which hinders the generalisability of the results 76 , or face some potential issues with the research design, particularly the cross-sectional data and small sample sizes, which would threaten the reliability of the results 75,77 .

ISSUES AND IMPLICATIONS OF CSR-FIRM FINANCIAL PERFORMANCE LITERATURE
A review of the literature on CSR-firm financial performance above has implied that there is strong evidence to support the positive impact of CSR on firm financial performance, which is in accordance with previous reviews on this literature 5,78-80 . The sections below will provide a more detailed discussion on the theoretical background, CSR measurement approaches, intervening factors, and methodological approaches used in CSR-firm financial performance, and consequently, identify some important issues and implications for future research. Table 4 provides an overview of the use of theories in CSR-firm financial performance research. Out of 86 empirical studies, 46 (54%) used either a single theory or multiple theories to explain the logic behind CSR-firm financial performance relationship. However, theoretical background and assumptions were still absent in almost half of prior empirical research in this sample (40 studies). This is problematic as the lack of theory-driven empirical research in CSRfirm financial performance literature would prevent us from achieving a systematic and precise view of whether, how and why CSR practices can improve firm financial performance. The absence of solid theoretical frameworks might also threaten the rigorousness of the hypothesis development and research design in empirical study settings 81 . It can also be seen in Table 4 that CSR-firm financial performance empirical research utilises some main theories, including stakeholder theory, slack resource theory, resource-based view, institutional theory, and signalling theory to gauge the relationship between the two. Of which, stakeholder theory 82 is widely used in CSR-firm financial performance research and it is also the most popular theoretical foundation for the mainstream research on CSR. Stakeholder theory states that organisations are surrounded by networks of stakeholders -those who either affect or are affected by the decisions made by the organisation. Some important stakeholders include, but not limited to: customers, shareholders, employees, suppliers, communities, and environmentalists 82 . Stakeholders are akin to organisational effectiveness as they can control valuable resources to organisations. As such, organisational success depends on its capacity to manage the relationship with stakeholders, which is the essential tool for value generation 78,[83][84][85] . Despite being widely used in the literature, the stakeholder theory is heavily criticised for its lack of specificity, vacuousness, and impracticality. It fails to provide a specific framework on how and to what extent organisations should manage their relationship with different stakeholders 86 . As a result, it is recommended that future research on CSR-firm financial performance should build upon a multi-theoretical approach for a more complete and overarching view of this relationship 87 . Yet this line of research is still under-developed with only 12 out of 86 studies representing in the current sample.  (2) what aspect should or should not be included in this construct. Table 6 presents the use of mediators and moderators in the CSR-firm financial performance relationship. It can be seen that CSR-firm financial performance relationship is moderated by a wide range of factors, including both internal factors, for instance, competitive strategy 13,16 , innovativeness capability 107 , or firm risk 108 ; and external factors, for instance, competition intensity and market turbulence 109 , or institutional environment 77 . It can also be seen that the impact of CSR on firm financial performance varies in different types of organisations (state-owned or nonstate-owned, MNCs or domestic companies) 87,96 , in different phases during a crisis (pre/post-crisis) 17 , and with different levels of CSR performance (best in class/worst in class) 20 . What can be learned from this literature is that CSR and firm financial performance relationship are likely contextual, and therefore, it is critical for research in this domain to consider possible variables that potentially affect the direction and magnitude of CSR's contribution to firm financial outcomes. However, this line of research is still limited with only 27 out of 86 studies (30%) examining situational factors in their models. The literature also highlights an important point that research on CSRfirm financial performance has moved further to find explanations of why and how CSR contributes to firm financial performance. On this thought, several studies have identified the channels (mediators) through which CSR translates firm financial outcomes, for instance, marketing competence 109 , innovation 106 , management effectiveness 9 , or competitive advantage 14 . However, this research domain has remained modest with only 20 out of 86 (23%) studies including mediators in their models. Table 7 provides an overview of the research designs used in the most recent empirical studies on CSR and firm financial performance. It is evident that most of the studies used secondary data and panel data to gauge the relationship between CSR and performance and used Ordinary Least Squares (OLS), Generalised Least Squares (GLS) or Partial Least Squares (PLS) approach to analyse such relationship. Some studies on CSR-firm financial performance relationship have highlighted that endogeneity is an important methodological issue which might threaten the reliability of parameter estimation. Endogeneity refers to the problem where at least one independent variable is correlated with the error term, causing biases in the prediction of parameters in regression models 110 . This problem is mainly caused by three main factors, including (1) omitting variables, which  are important factors that might predict firm financial performance but are not included in the model; (2) reverse causality -the phenomenon when CSR can increase firm financial performance, but firm financial performance can also predict CSR 111 ; and (3) measurement error, when the proposed measurement approach does not fully capture CSR 79 . By default, this problem of endogeneity violates the assumption of the exogenous error term in the OLS regression approach and affects the consistency of parameter estimation under OLS. Consequently, it can be understood that parameter estimation based on the OLS approach is biased with the presence of endogeneity problem. One of the most common approaches to parameter estimation on panel data is the fixed effect/random effect least square 110 . However, the fixed effect/random least square approach is still prone to bias and inconsistency as (1) it only removes the unobserved time-constant effect but not unobserved timevariant effect 112 . (2) The models after transformation still need to satisfy all the assumptions of OLS 110 .

Methodological approaches in CSR -firm financial performance research
(3) For a panel with short periods (t<5), fixed effect estimation is biased 113 . And (4), fixed effect estimator does not allow observed time-constant variables (e.g. industry dummies, geographic locations) in the model as it cannot distinguish between observed time-constant and unobserved time-constant effects 110 . As such, it is needed for research on CSR and firm financial performance to address the problem of endogeneity by adopting more advanced and complex analytic procedures and tools, for instance, the instrumental variable approach (two stage/three stage least squares (2SLS/3SLS) or System Generalised Method of Moments (System GMM) 114 ) to build a solid research design that fully addresses endogeneity problem.
The idea of an instrumenting technique is to remove the proportion of the endogenous/predetermined X that is correlated with the error terms by, firstly, identify truly exogenous instrumental variables Z which are correlated with X and indirectly correlated with Y, then remove the part of X that is correlated with the disturbance term by regressing it against those instrumental variables. Next, the fitted values of X (conditioning on Z), which are now uncorrelated with the error terms, are used in the model with Y, to estimate β 115 . The instrumental variables can be external, in the case of 2SLS, or internal using deeper lagged levels (or lagged differences) of the endogenous/predetermined variables, in the case of Anderson-Hsiao's method 116 . As such, correlations between the regressor and the error terms are likely to be removed after the first differencing transformation and instrumental variable techniques. Yet it can be seen from Table 7 that the instrumental variable approach is still not widely used in CSR-firm financial performance research with only 14 out of 86 (16%) empirical studies adopting this approach.

IMPLICATIONS FOR FUTURE RESEARCH
Overall, the empirical literature on CSR and firm financial performance has provided strong support for this relationship with most studies in our sample confirming the positive impact of CSR on firm financial outcomes. This literature also highlights some important implications for future research. Firstly, it is critical for CSR-firm financial performance research to build a solid theoretical background and assumptions to explain the relationship between CSR and firm financial performance. The absence of theory and the dominance of a singletheoretic approach might be partly responsible for the inconclusiveness of evidence regarding the impact of CSR on firm financial performance 117 . It has been suggested that research needs to go beyond a singletheoretic approach in favour of a multi-theoretic approach that is capable of providing a more comprehensive and nuanced understanding of the wellsprings of the business impact of CSR. Secondly, it is important to acknowledge that the CSR concept is multi-dimensional in its nature and there is no best way to measure CSR. Even though the KLD index continues to show its popularity and reliability in measuring CSR, it is recommended for future research to take into consideration other dimensions of CSR which are not captured by KLD for a more wellrounded construct. Thirdly, as the positive impact of CSR and firm financial performance is strongly confirmed, future research should go further to investigate possible intermediate factors and the mechanisms behind this relationship 79 . Future studies focusing on why CSR improves performance with an emphasis on the processes, communication, and decision making within organisations 118-120 would significantly extend our knowledge in this literature. Fourthly, a very important implication from CSRfirm financial performance literature is the need for a rigorous research design to reveal the true causality between CSR and firm financial performance 79 . Future studies in this domain need to address endogeneity with more complex analytical approaches and need to take into consideration different aspects of firm financial performance. While accounting measure is a retrospective measure reflecting the internal efficiency of the company's operations and management control, market performance, for instance, Tobin's Q, is a prospective measure indicating the market's perception of the firm's ability to generate profit and shareholder returns in the future 121 . Therefore, it is suggested that future research should use both accounting measures and market measures to reflect two distinct yet complementary aspects of firm financial performance 121,122 . Lastly, the CSR-firm financial performance literature is inviting more evidence from different national and institutional contexts, particularly in developing and transitioning economies. The literature has pointed out that CSR-firm financial performance relationship is highly contextual. As such, the presence of comparative studies which examine this relationship in different cultural, political, and institutional settings would stand to extend our knowledge in this domain much further. While prior evidence has been heavily focused on Western countries, research on CSR-firm financial performance in developing countries has become an emerging field of study 123 . Particularly in Vietnam, this literature is still in its infancy. Given the unique national and institutional setting, and a growing concern about CSR engagement in Vietnam, future studies that provide a more comprehensive view of the relationship of CSR and firm financial performance in the Vietnamese context are both warranted and overdue. These studies may also serve to illuminate some specific aspects that have wider significance, as well as offer important points of difference from prior findings.

CONCLUSION
CSR has become an important factor and linked to various operational results of a company. Due to the rising attention of this topic in both academic and practice, this paper has summarised a number of the most recent empirical studies on the relationship between CSR and firm financial performance and offered a closer look into the CSR literature in the Vietnamese context. It is found that the contribution of CSR on firm financial performance has received strong support from literature, although the overall findings are still inconsistent. It is recommended that future research would probably benefit from the implications on the measurement of CSR, the intervening factors, the problem of endogeneity in research design, and the need to conduct research on CSR and firm financial performance in different cultural and institutional contexts. This paper is not without limitations. It attempts to provide an extensive review of CSR and firm financial performance literature, however, there are possibilities that some recent studies on CSR-firm performance were not captured in our study. Another limitation is that the review on CSR literature in Vietnam, though similar, might not characterise CSR literature in other emerging, developing, and transitioning economies. Nevertheless, future research might benefit from the implications and suggestions provided by this paper.

APPENDIX
The list of all empirical studies included in the sample can be found in the Table 8.