Corporate social performance (CSP) and corporate financial performance (CFP) have been studied widely in the last fifteen years. These major efforts in research and publication might have been expected to produce aligned results but is it not the case; this paper aims to study why there is no consensus among them. By using the bibliometric data obtained from Scopus, this paper looks at four bibliometric indicators (journals, authors, countries, areas) for published work. We look for papers between 1977 and 2018 and identify a total of 371. The findings reveal that there are at least five key journals publishing in CSP-CFP relationship. There is a lack of co-occurrence between authors, regarding areas, two distinct perspectives are detected: one based on sustainability and social responsibility and the other linked more closely to marketing, and both perspectives can be connected via stakeholder theory. It contributes using bibliometric approach, it shows a reason to align the results around the relationship between CSP-CFP and open the windows to create a theory. Finally, the findings of this paper provide insights to the researchers on the development of social responsibility. The disparity of results in this line enriches analysis but does not outcome in a single line of contribution to science, thus decreasing the potential for channelling studies into a single theory. This only seems possible if contributions can be focused on the nexus that unites them: stakeholder theory.
Almost all types of business activity have to tackle social and financial issues. Historically, priority was given to financial issues, but in the late 20th century, social issues also began to be acknowledged and taken into consideration (Carroll, 1979; Wartick & Cochran 1985; Waddock & Mahon, 1991; Wood, 1991; Crowther & Aras, 2008; Venturelli et al., 2019). Numerous empirical studies have been conducted that bring together the nexus between corporate social performance and corporate financial performance (referred to hereafter as the CSP-CFP link). However, those studies have produced very different results and as yet no consensus has been reached;in the words of Sroufe&Gopalakrishna-Remani (Sroufe & Gopalakrishna-Remani, 2019: 334), “the relationship between social sustainability performance and financial performance represents one of the most questioned areas of sustainability business practices”. Indeed, there have been major efforts to analyse the CSP-CFP link in academic literature (Cochran & Wood 1984; Preston & O'Bannon, 1997; Waddock & Graves, 1997; Orlitzky, Schmidt & Rynes, 2003; Van de Velde et al., 2005; Isaksson & Woodside, 2016; Zhao & Murrell, 2016; Lin et al., 2019), “yet the results remain equivocal” as acknowledged by Hussain, Rigoni & Cavezzali (2018: 1198). A minimum level of consensus on results would enable the link to be better understood and thus help find potential ways of improving both areas at the same time.
Conventionally, the relevant literature has focused on differences in concepts, methods, measurements and sample size and composition in seeking explanations for the inconsistencies found in results concerning the CSP-CFP link (Orlitzky et al., 2003; Griffin & Mahon, 1997; Van Beurden & Gossling, 2008; Fu & Jia, 2012). This paper seeks to outline which areas should be worked to contribute to reaching a minimum level of consensus from a new approach: a bibliometric perspective (see Low & Siegel, 2019) that can provide an overview of the problem.In social sciences, the lack of consensus in the results usually has two main bases (Hicks, 1999): on the one hand, the different perspectives used in the analysis of the subject under study and, on the other, the different realities analysed in different geographical areas. The first basis has been covered by three variables: the journals that set the lines of argument and perspectives and establish when significant progress is made in each area; the authors, as the reference point to be followed; and the areas of research and scopes of study that reveal the different views and arguments concerning the same phenomena. For the second basis, we used the geographical area from which papers come, which may explain the cultural alignment that exists and enable the case histories from each country to be integrated (figure 1). These four bibliometric indicators are key in all research for constructing a unified theoretical approach and drawing consistent practical implications.
All these factors affect the scientific context, because we not only analyse the actual situation but also look at academic literature to continue making contributions, explaining and theorising on phenomena. The goal of our paper is, therefore, to outline the reasons why there is a lack of consensus in the CSP-CFP link and thus contribute to academic efforts to draw up a single theory to explain that link. We do this by applying bibliometric analysis. This provides an overall image of the subject under study and identifies statistical patterns by examining contents in such a way as to provide an illustrative overview of the different key perspectives for study. This bibliometric methodology and the use of network maps enables us to use conceptual analysis to objectively bring to light the precise scientific structure of the discipline under study, providing a comprehensive image of the scope of current research. An outline of how that research has developed is thus provided scientifically. This can be combined with an assessment of the results obtained to help bring to light and identify future directions for research by synthesising emerging areas (Feng, Zhu & Lai, 2017).
The paper is structured as follows: Section 2 describes social performance as a key factor for sustainability and looks at trends in research into the CSP-CFP link. Section 3 sets out the methodology used, with the bibliometric analysis, the research variables and the associated search. Section 4 presents an analysis of the results, Section 5 draws conclusions and finally Section 6 gives a list of bibliographical references.
2The link between corporate social performance and financial performance: a key factor for sustainabilityThe level of sensitivity towards social matters is increasing in society, and firms are expected to take on ever greater responsibility for bringing about sustainable development. Defining this new role is a major challenge for firms, as they need to find ways of encompassing both CSP and CFP. To factor the principles of sustainability into their business strategies, firms must be clear as to the nature of this link (Epstein & Roy, 2003).
Figure 2 shows the extent to which the CSP-CFP link has been investigated compared to research in the area of business, management and accounting (BMA) in general. Figure 2 compares trends in publications on the CSP-CFP link and sustainability.
As shown in Figure 2, in the past 25 years there has been a significant, largely parallel increase in publications in the field of Business, Management and Accounting (BMA) in general and those on the CSP-CFP link. For the first 10 years the trend was irregular for publications on the CSP-CFP link, but from 2005 onwards there has been sustainable, exponential growth which has outstripped that in general BMA publications in relative terms. The lack of regularity noted in publications on the CSP-CFP link in the first few years can be explained precisely in terms of its link with sustainability (Figure 3).
Having previously shown erratic growth, research into the CSP-CFP link began to grow exponentially around 2005, precisely when matters of sustainability also began to draw significantly more attention. Research into BMA in general dates back further, but there was parallel growth in regard to the CSP-CFP link and sustainability. Since 2017, studies on sustainability have outstripped those on the CSP-CFP link per se, in relative terms. Currently, the numbers of publications in the three areas (BMA, sustainability and the CSP-CFP link) are similar in relative terms, and there is some crossover between the three lines. It can thus be concluded that there is a tendency to include sustainability as an integrated topic, i.e., for more than 10 years now it has not been possible to discuss management without taking sustainability into account, and nor can a sustainable business be maintained without a balanced CSP-CFP link.
In spite of the importance of the CSP-CFP link as indicated above, the results remain equivocal.But how can it be that no consensus has yet been reached on these issues after fifteen years in which the scientific output has been truly huge?. Here we seek to outline an answer to that question, taking a global perspective that enables us to analyse how the topic has been worked on, via a bibliometric review. For the research community to reach a consensus certain basic pre-requisites must be met, even before there is any discussion of the concepts, methods and scales of measurement used. Researchers must work in a continued, collaborative fashion, there must be a meeting point in the journals where their results are published and if the topic is investigated by more than one discipline then the specific features of each one must be known and considered. Here, we look at whether those pre-requisites are met in the case in hand.
3MethodThis section starts by explaining the research method used (bibliometric analysis) and then sets out the research variables and search methods.
3.1Bibliometric analysisBibliometric analysis can make valuable contributions to the structuring of an increasingly broad knowledge set (De Bakker, Groenewegen & Den Hond, 2005). It is a tool suited to helping experts to enhance their knowledge of certain topics of study (Agle, Mitchell & Sonnenfeld, 1999; Low & Siegel, 2019). However, in this study, the bibliometric analysis has not only been used as a counting technique, but also as a method for measuring the evolution of a scientific domain, i.e, we observed the trend over time for the subject under study, which enabled us to identify distinct research trends and the key factor linking them, and, as a result, to contribute to the field of scientometrics (Hood & Wilson, 2001).
The bibliometric analysis in the present study was conducted with VOSviewer (Valenzuela et al., 2017; Vallaster et al., 2019; Sinkovics, 2016), a program developed to build bibliometric maps. The idea of VOS mapping is to minimise a weighted sum of squared Euclidian distances between all the pairs of elements via an optimisation process. Its main advantage is the high quality of its visual representation (Van Eck &Waltman, 2014).
Normalisation, mapping and grouping were carried out on the terms indicated by Van Eck &Waltman (2014):
(I) Normalisation:
VOSviewer uses normalisation of the strength of association (Van Eck & Waltman, 2009) to normalise the differences between nodes in the number of edges that they have with other nodes. The nodes (i.e., i and j) are in this paper the authors, countries and keywords. Taking Sij as the similarity of nodes i and j, and allowing aij to denote the weight of the edge between nodes i and j, where aij= 0 if there is no edge between the two nodes. Given that VOSviewer treats all networks as non-directional, aij = aji in all cases. Normalisation of the strength of association built up a normalised network in which the weight of the aged between nodes i and j is given by
whereki(kj) indicates the total weight of all the edges of node i (node j) and m gives the total weight of all the edges of the network. In mathematical terms,
Normalisation of the strength of association provides the connecting arcs that facilitate the construction of networks for investigating the CSP-CFP link according to the geographical origins of authors. Similarly, it makes it possible to compose networks of co-occurrence of keywords according to areas of study.
(II) Mapping:
The VOS mapping technique is used by VOSviewer to place the nodes of the network in a 2-dimensional space [26]. VOSviewer applies a variant of the SMACOF algorithm [27] to minimise the function
subject to the constraintwhere n indicates the number of nodes in a network, xi gives the location of node i in a 2-dimensional space, and ||xi − xj|| gives the Euclidian distances between nodes i and j.
The weighted sum of the squared distances between all the pairs of elements is minimised. The squared distance between a pair of elements is weighted according to the similarity between the elements. Finally, to avoid trivial solutions in which all elements have the same location, the constraint is imposed that the average distance between two elements must be one (Van Eck & Waltman, 2010).
This mapping method proves appropriate for establishing the locations of the main authors of papers on the CSP-CFPlink in the relevant literature according to volume of citations.
(III) Grouping:
The grouping technique used by VOSvieweris based on a variant of the modularity function introduced by Newman & Girvan (2004) and Newman (2004) to group the nodes in a network, enabling nodes to be assigned to clusters by maximising the function
whereci denotes the cluster to which node iis assigned, δ(ci,cj) denotes a function with a value of 1 if ci = cj and 0 otherwise, and γ gives a resolution parameter that determines the level of detail of the grouping. The larger γ is, the more groupings are obtained.
This grouping technique is particularly useful for identifying research into the CSP-CFPlink carried out collaboratively by authors from different countries, and thus enables cooperation between countries to be compared. Similarly, assignment to clusters also enables the different areas of knowledge from which CSP-CFP link is being studied to be determined and the various foundations on which those studies are based to be recognised.
The techniques for mapping and grouping in clusters used by VOSviewer provide a unified approach for mapping and grouping nodes in a network, facilitated by their complimentary nature (Waltman, Van Eck & Noyons, 2010).
3.2Research variables & the search for themTable 1 shows the technical research datasheet that sets out the problem under study and the indicators used to attain the purpose of this analysis. It also gives the origins of the data used (database, terms, sample, analysis period and date of implementation).
Research datasheet.
Source: Compiled by author.
Our investigation began with a bibliographical search in the Scopus database owned by Elsevier, because its coverage is broader than that of the Web of Science (almost 99% of the journals catalogued with Journal Citation Report are also in Scopus). This makes it easier to obtain maps of networks in emerging research fields which are as yet small.
The records considered are those that contained both the terms “Social Performance” and “Financial Performance” in their titles, abstracts or in the keywords of their meta data. The search was restricted to the fields of scientific knowledge in the database. This search resulted in a total of 371 records. The Scopus search was conducted in April 2019, so publications dated later than 2018 were ruled out because we wished to use only data on complete years. Once these records were identified, we cleaned up the terms and consolidated them. The process was carried out rigorously with the aim of not impacting on the results. The words classed as different variants of the same semantic term referring to an identical concept were grouped into a single term. It was also necessary to correct the differences in spelling, identify singular and plural forms, and combine full terms with abbreviations. Finally, generic terms needed to be ruled out. This led to the elimination of 16 terms, leaving 23 terms for the analysis. 17 authors identified as duplicates were also eliminated, leaving 739 authors.
4Analysis of resultsResults are shown in accordance with the research variables listed in Table 1.
4.1Are papers being published in journals with markedly different profiles?Our analysis of results published in the form of papers in journals included in Scopus database extended not only to the names of the journals in which they were published but also to the type of journals involved. Table 2 lists the main scientific journals by volume of publication in which eight or more papers concerned with the topic under study here have been published.
Main journals in terms of volume of publication.
| Journal | Publishing House | Purpose | Topics | Papers |
|---|---|---|---|---|
| Journal of Business Ethics | Springer | To publish contributions on ethical issues related to business | Ethical issues related to business | 36 |
| Business and Society | SAGE | To contribute to knowledge on corporate social performance, corporate social responsibility and corporate sustainability | Business and society more explicitly | 16 |
| Journal of Cleaner Production | Elsevier | To contribute to corporate sustainability and corporate social responsibility | The main perspective is that of environmental sustainability | 11 |
| Social Responsibility Journal | Emerald | To publish research in the areas of social responsibility, sustainability and governance | Social topics. Most closely aligned with the social perspective | 11 |
| Strategic Management Journal | John Wiley | To contribute to knowledge of strategic management | Questions in strategic management | 10 |
| Sustainability | MDPI | To publish studies related to sustainability and sustainable development | Sustainability issues; the main perspective is that of environmental sustainability | 8 |
| Corporate Reputation Review | Palgrave Macmillan | To publish research on corporate communication, corporate social responsibility communication, corporate identity, and organizational identity | More focused on marketing, which makes it distinct from but complementary to the previous journals | 8 |
Source: Compiled by author based on the Scopus database.
Despite publishing papers with topics in common, all these journals belong, curiously, to different publishing houses (Springer, Sage, Emerald, Wiley, Elsevier, MDPI, Palgrave) and they have no editorial board members in common. This denotes how widely dispersed the perspectives for analysis and communication are in regard to this topic. On the one hand this is positive, as it enables the same phenomenon on to be analysed from different perspectives, but on the other hand it is negative because it becomes more difficult to find a common line and overall contributions.
We found 328 papers with less than 100 citations and only 43 with more than 100, among them four papers with over 500 citations. Considering the number of citations per year instead of the total citations, the most cited articles remain the same (Table 3).This can be seen as an indicator of impact on and dissemination in the scientific community. The most widely cited paper, with a total of 2235 citations (approx. 100 citations per year), is “The corporate social performance-financial performance link” by Waddock & Graves, published in 1997 in the Strategic Management Journal (h-index = 6.42 & SJR = 8.006).
The most widely cited authors are all professors at US universities (Wallace E. Carroll School of Management in Boston, the Arizona School of Management and the Loyola University in Chicago, among others) and, as shown in Table 4, their publications date from before the turn of the millennium. Specifically, the first two papers were published in the Strategic Management Journal, the third, by Griffin & Mahon, in Business & Society and the fourth in the Academy of Management Journal. None of the most widely cited papers appear in the journal dedicated specifically to sustainability, maybe because it is a current topic and those are more recent journals. This indicates that their authors started out from a more overall perspective as regards the link between financial and social matters.
Main papers in terms of number of citations.
| Authors | Year | Paper | Citations |
|---|---|---|---|
| Waddock & Graves | 1997 | The corporate social performance-financial performance link | 2235(101.6) |
| McWilliams & Siegel | 2000 | Corporate social responsibility and financial performance: Correlation or misspecification? | 1182(62.2) |
| Griffin & Mahon | 1997 | The corporate social performance and corporate financial performance debate: Twenty-five years of incomparable research | 1104(45.6) |
| Agle, Mitchell & Sonnenfeld | 1999 | Who matters to CEOs? An investigation of stakeholder attributes and salience, corporate performance, and CEO values | 785(39.3) |
Note: Brackets provide the average number of citations calculated considering the total years from the publication data.
Source: Compiled by author based on the Scopus database.
Thus, the profiles of the journals are very different in their aim and scope, in the topics tackled, in their publishing houses, etc. This means that the CSP-CFPlink is tackled from different theoretical frameworks and the practical implications sought are applicable also in different areas. The fact that citations are mostly from just a few papers does not help to integrate previous results into future research, especially when the main citations come from work published before the turn of the millennium, at a time when sustainability was not yet properly developed. All this makes it harder to reach a minimum consensus on results.
4.2Are authors working in a collaborative and continuous way?Research furthers knowledge on the basis of a minimal level of consensus between different researchers, which leads to recognition by the scientific community of the results obtained. There are at least three variables which may be determinant in enabling a satisfactory consensus to be reached:
- -
Continuity over time, i.e., that the subject under study is not a passing issue but is of sufficient interest to society for researchers to maintain discussions long enough to work through arguments and results from different authors and reach a consensus.
- -
Continuity in terms of the topics studied by authors, i.e., that the topics are not merely touched on briefly by authors but comprise research lines which are maintained with input over time, leading to several publications by the same authors.
- -
Collaboration between researchers: the number of citations received is a good indicator of recognition by other authors of research work and its integration into subsequent studies.
In regard to continuity over time, Figure 4 shows a map of the trend in papers on the CSP-CFPlink from 2005 to the present day. A chromatic scale is used in which purple represents older publications, and shades from green to yellow are used to symbolise increasingly recent publications.
It can be seen that the CSP-CFPlink has been viewed as important by authors for at least the past 15 years. This has given rise to positive, stable continuity over time. Indeed, the topic continues to be of interest to the scientific community.
In regard to continuity in the topics studied by authors, Table 5 shows the number of publications per author on the CSP-CFPlink.
This table reveals that few authors have conducted continued, stable research into the subject studied here. Only two have five publications, four have four and 19 have three. By contrast, there are 653 authors who have worked on the CSP-CFPlink on a one-off basis and published a single paper. This wide spread of authors, making up small collaborative research groups, and the limited connections between them makes it difficult to achieve any consensus in the results obtained: continuity of work by authors and of arguments over time is conducive to agreement.
In regard to the third variable, Figure 5 maps out the main authors in terms of volume of citations in this field. The size in which their names appear represents the quality of citations that they have. There are few papers with a significant number of citations: most have very few, which means that their contributions are not included in subsequent research. This could be indicative of widely scattered interests in the scientific community, which could in turn lead to clear problems in adopting standards through common consensus.
The different colours denote numbers of publications: grey means a low number, while the other colours represent larger numbers. As indicated in regard to the variable on continuity in the topics studied by authors, there are few authors with more than one publication. However this figure also shows that the situation is worse in that the most widely cited publications are by authors who have not studied the topic in question consistently over time: i.e., they have merely published occasional papers on it. The authors with most publications as shown by the colour code have few citations.
Overall, results concerning the CSP-CFPlink have been published steadily over the past 15 years, which should have enabled comparisons between them, arguments and counter-arguments to have been made by the scientific community. But researchers themselves have not remained steady over that time: very few have published a number of papers on the topic and fed into the discussion on the CSP-CFPlink. Instead, one-off publications have predominated which make it hard to reach any consensus. Moreover, the publications most widely cited are precisely one-off papers on which there has been no continuity. This means that it is extremely difficult for a reasonable consensus to be reached among researchers.
4.3Do authors from different countries work together? If so, from what countries?The CSP-CFPlink has been investigated by authors from countries in very different circumstances. Table 6 lists those countries where there have been at least five publications. The list is shown in order of the number of publications. It is topped by the USA, followed by the UK with Spain just behind in third place. However, the ranking changes when the number of citations for each country is considered, with Spain dropping to fifth place behind Canada and Australia. This means that Spain is in a privileged position as regards scientific output, as the top-ranked non-English-speaking country, but on average its papers are cited 20 times compared to 100 for the USA and around 84 for the UK. Even so, Spain is above the average, which is around 16 citations per paper.
Documents published with citations per country.
Source: Compiled by author with Vosviewer based on the Scopus database. * The total strength of the connection is based on a quantitative index that represents the sum of the frequency of co-authorship in one country compared to the rest.
For a consensus to be reached on results, it is essential for authors from different countries (and therefore widely differing circumstances and contexts) work in collaboration. This section analyses whether such collaborations are actually taking place in regard to research into the CSP-CFPlink, and if so between which countries.
The 20 countries analysed can be grouped into five clusters depending on co-authorship criteria. “Cluster” here means a number of vectors grouped in line with a given criterion, which in this case is co-authorship of papers published on the CSP-CFPlink. Clusters bring together in the same group papers which share common properties, so that knowledge of groups provides a summarised description of a complex set of multidimensional data (Van Eck & Waltman, 2010; 2011).
Figure 6 shows a network map that brings together the top 20 countries in terms of number of publications, revealing 53 links and a total link strength of 92.
This clustering of countries on the basis of collaboration between authors (Table 7) can be explained through the confluence of geographical, language and historical criteria based on the internationalisation of students.
Clusters by countries depending on co-authorship criteria.
| Cluster 1 | Cluster 2 | Cluster 3 |
|---|---|---|
| France | Spain | Finland |
| Germany | Italy | Norway |
| Switzerland | Belgium | |
| Netherlands | ||
| Canada | ||
| Indonesia |
| Cluster 4 | Cluster 5 | |
|---|---|---|
| China | United States | |
| South Korea | United Kingdom | |
| Hong Kong | India | |
| Singapore | Brazil | |
| Australia |
Source: Compiled by author with Vosviewer based on the Scopus database.
The predominant criterion is geographical. This criterion is closely linked to the language criterion and to proximity. Cluster 1 includes central European countries, while Cluster 2 comprises countries from southern Europe. Cluster 3 contains four countries from Northern Europe (plus 2 countries which do not share the same geographical criterion). Cluster 4 contains Southeast Asian countries (along with Australia, due to different criteria). The fifth and last cluster is not based on geography but on a mix of other factors: first of all, the USA and the UK are the top two countries for research into this topic, and working together enables them to maintain their leading position. They are also both English-speaking countries, which makes for a smoother relationship. The language criterion is particularly significant in this fifth cluster: English is also an official language in India, which was a British colony until the mid-20th century. A similar historical link can be observed in Cluster 3: Indonesia has major historical links with both the Netherlands and Canada. Indonesia and the Netherlands have a close relationship also based on their colonial past. The Dutch government also made a major effort at the turn of the 21st century to strengthen economic, political and personal links with Indonesia; indeed, the Netherlands is currently Indonesia's biggest trading partner in Europe. Indonesia has also maintained diplomatic relations with Canada, which also belongs to Cluster 3, for 70 years. This creates a bond between countries which is reflected in the field of research.
The last criterion considered is the internationalisation of students, given that emigration rates among tertiary graduates are higher than total emigration rates. It is also post-graduate and PhD students, i.e., those most closely linked with research, who have the highest mobility rates. This criterion is clearly applicable to Cluster 4: Australia hosts university students from many countries (it is third in the G20 ranking of countries with most foreign students, behind only the US and the UK), with students from Asia (especially South-East Asia and China) accounting for 87%, which makes them a highly important group (OECD, 2019). Student internationalisation is one of the main ways of forging research links between universities in different countries, especially since it involves mostly postgraduate and Ph.D. students this may also account for the link between Brazil and the USA in Cluster 5: the US hosts around 1 million foreign students per year and the number of Brazilian students at US universities has grown steadily and significantly in recent years. This factor is not only important in relations between emerging and advanced countries but also between one advanced country and another, given the active promotion of student exchange, internationalisation and mobility programmes.
These clusters could also be related to institutional frameworks such as National Business Systems. The National Business Systems approach is based on the idea that different forms of capitalism institutionalize particular economic rules; the strategies, structures and competitiveness of firms are shaped according to it. Therefore, societies have historically developed institutions that reflected their own particular way of organizing economic relations. Social commitment was one of the axes in the emergence of these institutional configurations. (Matten & Moon, 2008; Kang & Moon, 2012).
Thus, cooperation between authors from different countries is widespread, and the links between them are varied and sound.
4.4Is work being done in different knowledge areas with different foundations?Table 8 shows the frequency of occurrence and the total link strength of the 23 terms obtained from 371 records. Total link strength is a quantitative index that represents the sum of the frequency of co-occurrence of a term compared to the rest.
Frequency & strength of link for key words.
Source: Compiled by author with Vosviewer based on the Scopus database.
CSP, CFP and CSR are the concepts that appear most frequently and with the greatest link strength. The question is what framework brings them together. The terms “stakeholder” and“stakeholder theory” appear with a relatively high frequency and a very high strength of link. Thiscan be interpreted as an initial sign of a link between numerous authors.Various authors have used the stakeholder framework for analysing and assessing the link between corporate social performance and financial performance (Clarkson, 1995; Ruf et al., 2001; Wood & Jones, 1995; Rowley & Berman, 2000; Neville, Bell & Mengüç, 2005). However, other theoretical approaches have also been used to look at that link, such as Legitimacy Theory and Institutional Theory.The results presented in this section enable us to discern whether the stakeholder theory approach carries any explanatory weight in regard to the CSP-CFPlink.
The map of key words shown in Figure 7 illustrates the most significant concepts associated with academic literature and the strength of the links between them. The size with which key words are shown denotes their frequency of occurrence, and the thickness of the links denotes the number of times that each pair of words appears together. The distance between terms denotes the strength of the link between them, i.e., the closer together to terms are, the closer the link between them is. Links are established on the basis of co-occurrence in scientific publications. The greater the number of publications in which to key words coexist, the stronger the link between them is and therefore the closer they are together. Each key word is represented by a node or a circle, the diameter and size of which indicate the number of publications that have the relevant term in their titles, abstracts or key words.
This shows the state-of-the-art in this knowledge area so as to provide an overview of the field of investigation, revealing how the different sub-fields are interconnected, with a view to finding potential opportunities for filling gaps in research in those sub-fields.
Once the key words are placed on the map, the next step is to identify the clusters or groups of related terms. Terms are assigned to groups in accordance with their co-occurrence in publications. Those terms that coincide often are strongly linked and are therefore assigned to the same cluster or group. By contrast, those for which there is little or no co-occurrence are assigned to different groups. A group made up of terms of the same colour denotes a research line within which different research topics can also be identified.
The network map in Figure 7 shows a green cluster associated with social responsibility and sustainability, which is very close to a second cluster in blue associated with governance and micro-finance. Both these clusters are characteristic of management. They are some distance away from the third compact, homogenous cluster in red, which brings together topics concerned with marketing and reputation and has interaction links with the other two clusters via stakeholders (hence the importance of the link strength shown in Table 7).
Bibliometric analysis reveals that stakeholder theory is the predominant approach when it comes to explaining the CSP-CFPlink, though not the only one. According to the network map in Figure 7, obtained from the bibliometric analysis, stakeholder theory is the only linkage between the clusters, which makes it the most suitable framework.
Figure 8 shows the trend over time in the subject under study, as an aid in understanding research trends from past to present. The colour of each node denotes the average year of publication of all the documents that include the relevant term. The chromatic scale a tribute colours to nodes in such a way that cold colours such as purple denote research activities with average dates of publication longer ago than those of warm colours, with red denoting the other extreme, i.e., the most recent publications (Ranjbar-Sahraei & Negenborn, 2017).
It can be seen that the marketing cluster dates from the 1990s, while the other two clusters date from the last 10 years, thus revealing the trend in research topics on this matter.
The density map shown below depicts the distribution and expansion of elements on a surface, with areas of concentration highlighted. The key word co-occurrence density map in Figure 9 shows no single central core (such cores would appear in red or orange) on which the different expressions could hinge. Instead, two separate focal points clear, one of them stronger than the other. This means that the research area is still in primary state, with a field of study open to infinite possibilities for development.
The yellow area on the density graph in Figure 9 denotes a higher density, while the blue area denotes lower density and green denotes an intermediate level. Two clusters can be seen on the graph, some distance apart and with different densities. Each cluster represents a perspective from which the topic is tackled, so two distinct research tendencies can be observed: one linked to social responsibility and sustainability and the other two marketing and reputation. This suggests that although there is no consensus, there is a clear trend in the perspective from which the topic is analysed, from marketing towards social matters and sustainability. This is backed up by the fact that the main journals in which papers on this topic are published are concerned with social issues and sustainability, though the papers most widely cited appear in journals which deal purely with strategy and management. The shared nexus or point of inflection is the inclusion of stakeholder analysis from different knowledge areas (beginning in the 1990s) a significant shift has been observed from the area of marketing towards social issues within business management. Therefore, stakeholder theory is the key connection between these different perspectives that study the relationship between social performance and financial performance. Accordingly, these results show a tendency to move towards strategic analysis of the connection between financial and social results based on sustainability seen from an overall rather than merely environmental perspective.
5ConclusionsThe field of business, management and accounting sustainability and the link between CSP-CFP have grown in parallel, significantly and largely over recent decades. Sustainability is increasingly treated as an integrated issue in BMA and, in the same way; a balanced CSP-CFP is an intrinsic part of sustainability. Concern for sustainability has increased over the last 15 years, particularly in two of its three facets: economic and social issues have been the subject of multiple research projects. However, in the results, no minimal consensus has been reached concerning the CSP-CFPlink. The bibliometric analysis conducted here identifies areas where work is needed to remedy this inconsistency in results. To date there have been no studies analysing issues that could affect the link between corporate social and financial performance (journals, authors, origin of papers and areas) from a bibliometric perspective. The originality of the analysis conducted here lies in the fact that it condenses existing theoretical and empirical studies to investigate the trend in the literature on the link between social performance and financial performance as shown by the Scopus database between 1977 and 2018. It does this by identifying the main publications and authors, the key research areas and changes in those areas over time.
The results show broad dispersion at least in four aspects: The first is the widely differing visions of academic journals, based on very different profiles, which tackle the topic from different theoretical frameworks and seek practical implications applicable also to different areas. Thus, the CSP-CFP link can be seen in terms of a single phenomenon being explained from different perspectives with opposing theories and unconnected explanations. The second factor is a lack of co-occurrence between authors: theyhave not worked continuously or indeed in collaboration on the subject (as shown by the erratic citationsof authors), and this has reduced the opportunity for joint contributions. Very few researchers have published more than one paper on the topic and provided input for discussion on the CSP-CFPlink. Rather, there has been a predominance of one-off publications that are unlikely to lead to a consensus.Thirdly, authors from different countriesmaintain links based on cultural similarities, geographical proximity, diplomatic relations and/or the provision of joint programmes emerging from the internationalisation of postgraduate and Ph.D. students, and thus of the contributions made. These vehicles effectively foster cooperation between authors from different countries. The fourth and last factor is that key word co-occurrence maps reveal that there are two coexisting lines of research (areas): one focused on sustainability and corporate social responsibility and the other on marketing, with stakeholder theory as a nexus between them. Thus, the differentprofiles of the journals in which research into the CSP-CFPlink is published, a lack of continuity in the topic on the part of authors and the different knowledge areas from which the topic is tackled all make it harder to reach a consensus in the results reported.
This study thus contributes to contemporary discussion on the link between CSP-CFP. The disparity of results that emerges from different perspectives enriches the analysis but does not lead to a single contribution and is not accepted by the scientific community. Contributions are made along different lines which cannot be brought together unless decisions are made based on bibliometric analyses such as this one. Consequently, our study reveals the need for joint efforts to produce a single line for benchmark contributions. The only way to continue integrating scientific studies seems to be to continue using the nexus that already exists, i.e., stakeholder theory. If this is done, future studies can make joint, significant contributions to scientific literature. We also recommend focusing on studies that entail more long-standing collaboration between authors; this would be especially enriching if the authors in question come from countries with different cultures, contexts and perspectives. Only consensus and interconnection in a single line will enable significant contributions to be made to the relevant theory.
In spite of these contributions, the limitations of this study must be taken into account. Firstly, there is a limitation inherent in the methodology used to tackle the problem: the analysis of bibliometric maps entails a simplified version of reality, which may lead to suppositions, which cannot be generalised; furthermore, the bibliometric analysis does not provide evidence of causality. Secondly, the analysis is conducted at a particular moment in time, but for its effects to be seen it needs to be reviewed in the future to check whether a trend towards cohesion has been attained in contributions along these lines. Finally, what is relevant in scientific terms is the specific content and underlying contributions of the papers published on the link between social and financial performance.
One future research direction is therefore to develop a global social performance measurement for all companies. Once such consensus is reached it will be useful to establish what factors affect social value creation without cost for companies but with profit, e.g. the positive social impact of using green products, the inclusion of women as directors and the generating of participative governance decisions. Last but not least, the social purpose of company management for stakeholders is becoming increasingly important (for example Business Roundtable states that companies in the 21st century are focused on generating long-term value for all stakeholders). Therefore, it would be highly useful to analyze the CSP-CFP relationship comparing stakeholder-oriented companies and non-stakeholder oriented companies to clarify whether the stakeholder view is a good starting point for securing the consensus of expert researchers on the topic. Some active academic networking will also be necessary to create links between authors and journals to align their contributions and obtain consensus to contribute significantly to society on this topic.
