Corporate Financial Performance and Corporate Social Performance : Methodological Development and the Theoretical Contribution of Empirical Studies

One of the roles of accounting is to provide information on business performance, either through financial accounting indicators or otherwise. Theoretical-empirical studies on the relationship between Corporate Financial Performance (CFP) and Corporate Social Performance (CSP) have increased in recent years, indicating the development of this research field. However, the contribution to the theory by empirical studies is made in an incremental manner, given that each study normally focuses on a particular aspect of the theory. Therefore, it is periodically necessary to conduct an analysis to evaluate how the aggregation of empirical studies has contributed to the evolution of the theory. Designing such an analysis was the objective of the present study. The theoretical framework covered the following: stakeholder theory, the relationship between CSP and CFP, good management theory, and slack resource theory. This research covered a 15-year period (1996 to 2010), and the data collection employed a search tool for the following databases: Ebsco, Proquest, and ISI. The sampling process obtained a set of 58 exclusively theoretical-empirical and quantitative articles that test the CSP-CFP relationship. The main results in the theoretical field reinforce the proposed positive relationship between CSP and CFP and good management theory and demonstrate a deficiency in the explanation of the temporal lag in the causal relationship between CSP and CFP as well as deficiencies in the description of the CSP construct. These results suggest future studies to research the temporal lag in the causal relationship between CSP and CFP and the possible reasons that the positive association between CSP and CFP has not been assumed in some empirical studies.


InTRoDuCTIon
One of the roles of accounting is to produce information on business performance (Gaspareto, 2004).This performance may be measured from the perspective of monetary values, normally using financial-accounting information, and/or from the perspective of non-monetary information (Hendriksen & Van Breda, 1999).The measurement of performance by non-monetary indicators is more recent and has increased in acceptance, particularly as a function of the contemporary concern regarding the social action of organizations (Oliveira, De Luca, Ponte, & Pontes Junior, 2009).The theme of the present study fits within this context of accounting, focusing on Corporate Financial Performance (CFP) and Corporate Social Performance (CSP).
There is a perception that studies related to stakeholder theory, more specifically with regard to theoreticalempirical studies on the relationships of CFP and CSP, have increased in recent decades.However, this process of theoretical development through empirical contributions is normally performed in a precise manner, with each study pointing to a specific problem in the theory or presenting a possible explanation of a theoretical gap.Periodically, therefore, it is convenient to develop studies that seek to analyze, in an aggregate manner, the evolution of a research field.In this regard, the objective of the present study was to investigate the evolution of the conceptual and methodological aspects of these theoretical-empirical studies, with the goal of presenting a consolidation of the advances that have been gained.
In particular, the present study focuses on the aspects of the relationship between CFP and CSP.For example, several studies have investigated which independent variables among those that typically serve as a proxy for CSP (customers, employees, suppliers, government, environment, diversity, and community) are neutral, positively, or negati-vely correlated with the variable CFP.However, once again, what is the synthesis of these results?Furthermore, several studies have specifically investigated which control variables could be employed in this relationship, e.g., stakeholder management (Berman, Wicks, Kotha, & Jones, 1999), company size (Orlitzky, 2001), industry (Waddock & Graves, 1997), and board composition (Shao, 2010).What is the set of control variables that have been tested and are potentially appropriate for this relationship?Some questions were designed to guide this research.In the theoretical field, these questions were used to facilitate the responses to the following questions: Do the empirical results reinforce stakeholder theory?What potential inconsistency does stakeholder theory display compared to the observed empirical results?What advances in stakeholder theory may be supported by empirical evidence?
In the area of empirical research, this study seeks to answer to the following questions: What are the most used variables to measure Financial Performance and Social Performance?How is the causal relationship between the two variables treated?What control variables are considered in this relationship?What are the most used statistical techniques?What are the sources and form of data collection employed?Has there been, in fact, an evolution in the volume of these publications?
Notably, there is a series of theoretical and methodological gaps to be researched.From the theoretical perspective, the main gaps are related to the variables that influence the CSP/CFP relationship and their causal and temporal relationship.From the methodological point of view, the main gaps are related to the most accepted forms of measuring CSP and CFP and to the main test methods employed.

Theory of the Firm and Stakeholder Theory.
The theory of the firm began with the emergence of economic science in the 18 th century, which took as its main reference the study of Adam Smith on the origins of the wealth of nations.Throughout the 20 th century, several studies, in particular that of Coase (1937) on transaction costs, enriched and developed the theory of the firm.
As with other microeconomic theories, the theory of the firm establishes profit maximization as its goal.However, profit maximization requires a more specific definition, such as if it is accounting or economic and short or longterm, among other characteristics.Jensen (2001) clarifies that, for economists, the objective of the firm should be to seek maximization of the long-term market value, resulting from the ability to generate cash over time.Maximizing the company's value maximizes the shareholder's wealth.
The main framework for stakeholder theory has normally been attributed to the work of Freeman from 1984(Froo-man, 1999;Barnett, 2007).According to Coombs and Gilley (2005), the most employed definition for the term stakeholder in the literature is that proposed by Freeman (1984) in which the stakeholder is any individual or group that may affect the achievement of the organization goals or that is affected by the process of searching for these objectives.Freeman (1984) also states that stakeholders are groups that have a legitimate right regarding the organization.However, variations in the definition of the term have been recorded, some of which are broader, such as that of Mitchell, Agle, and Wood (1997), and others that are less comprehensive, such as that of Clarkson (1995), according to which the company should consider the primary stakeholders of greater importance and relegate secondary stakeholders to the background.Some dimensions are attributed to stakeholder theory.For Donaldson and Preston (1995), this theory exhibits three dimensions: descriptive, instrumental, and normative.The descriptive dimension is revealed when the company uses the model to represent and understand its relationships and roles in external and internal environments; the instrumental dimension is evidenced when the model is used as a management tool for administrators; and the normative dimension emerges when management recognizes the interests of all the stakeholders, bestowing them with intrinsic importance.The normative dimension of this theory supports the objective function of the company.Initially, this dimension was based on Clarkson (1988), who indicated that the purpose of a company is creating and distributing wealth to the primary stakeholders.Subsequently, Evan and Freeman (1993) formulated the most commonly referenced definition, according to which the real purpose of a company is to serve as a vehicle to coordinate the interests of the stakeholders.In the descriptive dimension, the theory should explain how the managers can identify and classify the stakeholders to manage them.In this regard, Mitchell, Agle, and Wood (1997) state that characterizing only primary and secondary stakeholders is not sufficient for this purpose, and they establish a typology to reveal the salience of the stakeholders, classifying them according to the three following attributes: power, legitimacy, and the urgency of their claims.Through the instrumental dimension, the theory postulates that the firm will have superior performance if the normative precepts are employed (Berman et al., 1999).Jones (1995), for example, suggests that companies with relationships supported by trust with their stakeholders develop a competitive advantage over other companies.
One of the main counter positions of the two theories is related to the objective function of the firm.The theory of the firm proposes that the objective of the company is the maximization of shareholder wealth, while the stakeholder theory indicates that the objective of the company is to coordinate the interests of the stakeholders.There are several arguments to support each side.From the point of view of the firm, according to Sundaram and Inkpen (2004), the other parties associated with the company that are not shareholders (customers, employees, suppliers, and customers) are protected and benefit from contracts and legislation, which is not the case for shareholders.Meanwhile, according to stakeholder theory and Campbell (1997), it is not possible for companies to survive without delivering value to important stakeholders, and in this regard, although the shareholders have some different rights from the other stakeholders, this does not provide them with an imbalanced right to receive company benefits.This controversy constitutes a broad debate in academia, with no consensus (Silveira, Yoshinaga, & Borba, 2005;Marcon, Bandeira-De-Mello, & Alberton, 2008;Boaventura, Cardoso, Silva, & Silva, 2009).
The consequences of these counter positions are obvious in empirical studies that analyze CSP and CFP.More specifically, these studies reveal that there is an alignment of the firm's objective with CFP, according to the theory of the firm, and with CSP, according to stakeholder theory.

Corporate Financial Performance: Concept
and Measurement.
Although the definition of CFP is not debated in the li-terature, there is disagreement with respect to the best way to measure CFP (Cochran & Wood, 1984).According to Orlitzky, Schmidt, and Rynes (2003), a survey of the literature reveals that CFP has been basically measured in three forms: market, accounting, and survey measurements.Orlitzky, Schmidt, and Rynes (2003) further explain that the first approach reflects the degree of satisfaction of the shareholders; the second captures an idea of the internal efficiency of the company; and the last provides a subjective estimation of its financial performance.
It is worth indicating that, just as there is a relationship between CSP and stakeholder theory, there is an association between CFP and the theory of the firm, given that seeking to maximize CFP is linked to the objective of the firm.
In contrast to the variables that have been proposed to measure CSP, for which a precise definition by which to evaluate the fulfillment of the demand of each stakeholder is not available, the variables employed to measure CFP are supported in the literature by precise forms with which to measure them.

Corporate Social Performance: Concept and
Measurement.
The CSP concept, according to Ullmann (1985), refers to how an organization responds to social demands, a concept originally proposed by Strand (1983).Surroca and Tribó (2008) clarify that the concept of CSP is related to stakeholder theory because seeking maximization of CSP is linked to the objective of meeting the interests of the stakeholders.Waddock and Graves (1997) operationalize the concept of CSP, explaining that CSP constitutes a multidimensional construct whose behavior varies as a function of its inputs (e.g., investments in control pollution and other environmental strategies), processing (e.g., treatment given to minorities or the nature of manufactured products), and outputs (e.g., relationships with the community or philanthropic programs).
For the development of theoretical-empirical research, it is necessary to operationalize the means of measuring CSP.The measurement of CSP should consider how the demands of the various stakeholders of an organization are met, that is, CSP is conceptually an aggregate variable of an indirect observation, a proxy.There is an operational problem in the theoretical-empirical research that measures CSP, which has two central points: which stakeholders should be evaluated and how to evaluate the fulfillment of demands.The lack of standardization of which stakeholders should be considered as well as of the definition of the variables to measure the fulfillment of each stakeholder constitutes a relevant limitation of empirical studies on CSP.
The concept of a stakeholder proposed by Freeman (1984) subsequently found in the CSP literature to refer to an individual or a group has been extended to society and the environment (Carroll & Buchholtz, 2000;Irwin, 2002).Considering the environment as a stakeholder implies an extension of the original concept of stakeholder theory.Turcotte, Bellefeuille, and Hond (2007) explain that this extension is acceptable because environmental matters are of interest to several stakeholders, such as customers, investors, non-governmental organizations, and the government.The companies, in turn, respond to the issues as a function of the pressure generated from these stakeholders.Thus, it is possible to consider the environment as a CSP variable.

The Relationship between Financial
Performance and Social Performance.
Based on the theory of the firm and stakeholder theory, conceptual propositions are derived from a positive, negative, and neutral relationship between financial and social performance.
The proposal of a positive association is usually founded on arguments from stakeholder theory, as Bird, Hall, Moment' e, and Reggiani (2007) explain.An example of this type of argument is that of Alexander and Buchholz (1982), who argue that companies in which managers engage in activities that promote social performance will obtain better financial performance than other companies.There are also other arguments for the\is positive association, such as the presence of tension between the explicit costs of a company (such as payment of debentures) and their implicit costs (such as environmental costs).As a result, according to Cornell and Shapiro (1987), companies that seek to reduce their implicit costs by means of socially irresponsible actions will incur greater explicit costs, resulting in a competitive disadvantage.
The proposition of a negative association is normally defined by researchers of the neoclassical economic school of thought (Bird et al., 2007).The perspective of these researchers, such as Aupperle, Carrol, and Hatfield (1985), is that companies that engage in socially responsible behavior are at a competitive disadvantage because they incur costs that could be avoided or transferred to other agents (e.g., customers or the government).In other words, according to this line of reasoning, there are few economic benefits for socially responsible behavior and many costs associated with this behavior; thus, this type of behavior results in a decrease in the financial performance of the company (Friedman, 1970).
The proposition of a neutral association assumes that the relationship between CSP and CFP is either non-existent or non-linear.The lack of existence of such a relationship is advocated by Ullmann (1985), who states that there are so many factors or variables that influence the relationship between CSP and CFP that even if a relationship existed, the relationship could not be detected due to the problems associated with measurement in empirical studies of the relationship between CSP and CFP.Furthermore, some have proposed a non-linear relationship between CSP and CFP.Barnett and Salomon (2006) find evidence of a curvilinear relationship between CSP and CFP, in which the greatest returns on CFP are associated with the smallest and largest investments in CSP.

Stakeholder Management.
Stakeholder management constitutes one of the main principles of stakeholder theory, specifically in the theory's instrumental dimension.According to Freeman (1984, p.53), stakeholder management is summarized as the organization's ability to identify who its stakeholders are and their respective interests, objectives, and ability to influence the organization, to understand the processes that may be used by the organization to relate to these stakeholders, and to deduce what decisions best allow the stakeholder interests to be aligned with the organization's processes.
Invariably, organizations assume an extensive role of obligations with multiple stakeholders whose demands may not be met in their entirety.In this particular regard, Harrison and John (1996) state that stakeholder management may minimize the negative effects of conflicts of interests among stakeholders.According to Berman et al. (1999), stakeholder management is part of a company's strategy, and their empirical studies demonstrate that stakeholder management constitutes a variable that influences CFP.
According to Freeman (1984), companies that build better relationships with their primary stakeholders are likely to obtain greater returns.For example, companies seen as socially responsible have a greater ability to recruit qualified employees (Greening & Turban, 2000).Another example, indicated by Godfrey (2005), would be that companies with socially responsible activities build moral capital among their stakeholders, which promotes a certain type of safety against a loss of the company's reputation during problematic periods.The majority of empirical studies find results that attest to this positive association (Benson & Davidson, 2010).However, it should be noted that there are studies, such as studies by Meznar, Nigh, and Kwok (1994), that record a negative relationship, and others studies have not found any relationship (Bird et al., 2007).

Causal relationship between CSP and CFP.
Another aspect of this discussion is the causal relationship between CSP and CFP.In short, CSP is indicated as a dependent function of CFP, or vice versa.The idea of a virtuous circle is also derived from this discussion, that is, companies with good financial performance invest in social responsibility and with this obtain a greater return, which allows them to reinvest in social responsibility, and so forth.More specifically, Waddock and Graves (1997), who advocate a causal relationship between CSP and CFP, have proposed two theories: good management theory and slack resource theory.
According to good management theory, CSP occurs first.According to this theory, the company that is perceived by its stakeholders as having a good reputation, through a market mechanism, will more easily achieve superior financial performance.According to slack resource theory, the company needs to have positive financial performance before investing in social responsibility.Resources derived from good financial performance are necessary for CSP to occur.

Consolidation Studies of Empirical
Research.
Periodically, studies have appeared in the literature with objectives similar to those of the present study, that is, studies that seek to consolidate the knowledge accumulated by empirical research in a particular field of knowledge.With respect to CSP and CFP, the contributions of some studies are discussed as follows.Ullmann (1985) evaluated 13 empirical studies, including case and quantitative studies, published between 1970 and 1984, and did not find a trend in the results of the analyzed studies.According to Ullmann (1985), the reasons for this result were the following: the lack of a theoretical foundation, an inappropriate definition of terms, and deficiencies in the empirical data.Ten years later, Wood and Jones (1995) analyzed 50 empirical studies, including case and qualitative studies, published between 1970 and 1994.According to Wood and Jones (1995), many empirical studies lacked a theoretical foundation, with problems occurring with stakeholder mismatching (a mismatch between the relevant studied stakeholder and the respective measurement variable) and an apparently ambiguous relationship between CSP and CFP.
According to Ullmann (1985) and Wood and Jones (1995), from the beginning of the 1970s to 1994, empirical research lacked a theoretical foundation, and the results regarding a positive or negative relationship between CSP and CFP were inconclusive.However, soon afterward, Orlitzky (2001) published a meta-analysis study in which 20 quantitative empirical studies published between 1975 and 1997 were analyzed.This study found a positive relationship between CSP and CFP.The majority of the studies evaluated by Orlitzky (2001), i.e., 65%, were published after 1990, that is, during a period significantly later than the studies by Ullmann (1985) and Wood and Jones (1995).Orlitzky, Schmidt, and Rynes (2003) analyzed, through meta-analysis, 52 quantitative studies between 1972 and 1997 with the objective of providing accumulated and integrated statistics on the CSP-CFP relationship, evaluating the predictive validity of the instrumental dimension of stakeholder theory in the CSP-CFP relationship, and analyzing several moderators of this relationship.The results indicate a positive relationship between CSP and CFP in which CSP is more closely related to accounting metrics than market ones and that reputation is more closely associated with CFP than the other CSP variables.The study further revealed that between 15% and 100% of variations in the results of the empirical studies analyzed are explained by stakeholder mismatching, sampling errors, and measurement errors.Margolis and Walsh (2003) investigated the empirical evidence of the CSP-CFP relationship from 127 empirical articles published between 1972 and 2002.Of these articles, 109 assume that CFP is dependent on CSP, and 54 of these reveal a positive relationship, 7 a negative relationship, 28 a non-significant relationship, and 20 no relationship.
The studies by Orlitzky, Schmidt, and Rynes (2003) and Margolis and Walsh (2003) established a reference that the majority of empirical studies attest to a positive relationship between CSP and CFP.More recently, Beurden and Gössling (2008) evaluated 34 quantitative studies published between 1990 and 2007 by a meta-analysis and also found that there is a predominance of a positive relationship between CSP and CFP in the empirical studies.
The lack of a theoretical foundation for the empirical studies observed by Ullmann (1985) and Wood and Jones (1995) has apparently been overcome; this deficiency was not commented on in the subsequent studies of Margolis and Walsh (2003), Orlitzky, Schmidt, and Rynes (2003), and Beurden and Gössling (2008).This finding may be supported by theoretical advancements in stakeholder theory observed mainly in the short time between 1995 and 1997, particularly the following: descriptive, instrumental, and normative perspectives (1995), instrumental stakeholder theory (1995), the salience of stakeholders (1997), good management theory (1997), and slack resources theory (1997).

METhoDology
The concept of meta-analysis may have at least two definitions.According to Beurden and Gössling (2008), a meta-analysis may evaluate several studies with the use of descriptive statistics, in which each evalua-ted study constitutes a unit of analysis.As defined by Orlitzky, Schmidt, and Rynes (2003), a meta-analysis consists of a statistical technique capable of correcting for a particular number of previous studies, with their respective data sets, sampling errors, and measurement errors.The present study is classified according to the first definition of a meta-analysis.

Sampling.
The search tools of the following three databases were used to select the articles in the sample: Ebsco, Proquest, and ISI Web of Science.The selection was performed in two stages.The first stage was broader because the search of the periodical database was not restricted with an established impact factor or more restrictive selection filters.In this stage, the Ebsco and Proquest search tools were employed.In the second stage, which was more restrictive with respect to periodical coverage, the ISI search tool was used to identify periodicals with greater impact and was more specific and less restrictive in terms of filters.Both stages were restricted to academic articles in English.
For the first selection stage, in both databases, the following search criteria were defined: ("stakeholder theory") in the Text AND (variable) in the Text AND ("financial performance" OR "firm performance" OR "corporate performance") in the Abstract.The Ebsco database returned 98 articles from this search, and Proquest returned 92.Combined, these two sets comprised 158 articles.
In the second stage, other leading periodicals were selected in the field of management and other specific areas.The search was deepened in the second stage.In this phase, the following search criteria were applied: Combined, the articles selected in the two stages produced a sample of 198 articles.The desired sample consists of empirical articles that investigate the relationship between CSP and CFP that have been published in the last 15 years, more specifically, between 1996 and 2010.In this regard, based on this initial set of articles, the following exclusion criteria were applied to find the desired sample: the study of another topic other than the CSP-CFP relationship, a definition to measure the CSP variable that was inadequate for the theoretical framework, a definition to measure the CFP variable that was inadequate for the theoretical framework, theoretical essays, case studies or multi-case studies (case studies are more appropriate to explore a field of study than to validate results for large populations), articles prior to 1996, and articles after 2010.These exclusion criteria permitted the identification of a sample of 58 articles, which are reported in Appendix 1.

Analysis.
The articles in the selected sample were examined in depth to extract the factors that influenced the relationship between CSP and CFP.Specifically, factors that influenced the following relationships were considered: the independent variables employed to make up CSP and CFP, the causal relationship between CSP and CFP, the temporal lag in the causal relationship, and the variables that influence the relationship between CSP and CFP.
The aforementioned examination was summarized in a matrix, the synthesis of which is found in

RESulTS
The data analysis is divided into two sections: 1) the methodology employed in the studies -this has the purpose of analyzing how the quantitative empirical research is being developed in this field from the perspective of methods; 2) contributions to the theory -this seeks to reveal how these empirical studies, in an aggregate manner, are contributing to the development of the theory.

Methodology Employed in the Studies. a) Variables used to measure Financial Performance
The variables most used to measure financial performance are presented in Figure 1 and include, in decreasing order of frequency, the following: ROA -return on assets; ROE -return on equity; sales growth; ROS -return on sales; contribution margin; Tobin's Q; market share; risk of the firm; ROCE -return on capital employed; operational profit; cash flow; and earnings per share.
The variable ROA, an accounting variable by nature, is the variable most used to measure financial performance.However, ROA should be used with caution because this variable represents short-term performance and does not reflect long-term performance.The market variables and variables associated with long-term performance (CAR -cumulative abnormal return, RAP -risk adjusted performance) are used, but less significantly than variables of an accounting nature, as can be observed in Figure 1.Variables Measuring Financial Performance b) Stakeholders considered in the measurement of Social Performance CSP consists of a measure that evaluates the performance of an organization in attending to the interests of the organization's stakeholders; consequently, CSP is an aggregate measurement of indirect observations.In particular, it is of interest to identify the stakeholders that are considered in the composition of CSP in empirical research.The stakeholders that are most considered when measuring social performance are presented in Figure 2. A predominance of the following stakeholders can be observed: environment, employees, community, customers, suppliers, and shareholders.

Figure 2
Variables for the Measurement of Social Performance It should be emphasized that there is no standard way of measuring the fulfillment of interests for each stakeholder.This lack of standardization is one of the important limitations of the empirical forms of measuring CSP.

c) Control variables and intervening variables considered in the CSP/CFP relationship
The analyzed studies considered a broad range of variables that influence the CSP/CFP relationship.These are control variables or intervening variables, depending on the statistical model used.The identified variables include the following: company size (normally employed as a logarithmic function of the number of employees, assets, or sales), life cycle, sales growth, indebtedness, financial leverage, corporate type, company risk, research and development (R&D), age, region, year, intensity of the competition, work intensity, industry, and propaganda.However, the most widely employed variables are industry and size.The size of the firm is measured in different ways and was used in 34% of the articles in the sample; the industrial sector was used in 14%, financial leverage in 9%, and R&D in 7%.
Size is a relevant variable because there is evidence that small firms do not explicitly exhibit behavior that expresses "social responsibility".The sector is also a relevant variable because different sectors have different levels of investment in R&D, and consequently, the result may be confusing; furthermore, depending on its characteristics, the sector coincides, or not, with unsolved problems of a social nature (Waddock & Graves, 1997).

d) Treatment of the causal relationship between CSP
and CFP In the large majority of empirical studies of the CSP-CFP relationship, the CSP function is the independent variable, and the CFP function is the dependent variable.This condition of dependence is a conceptual option of the researcher a priori to the establishment of his hypotheses and not of the statistical results found.Table 1 contains the consolidated results of the analysis of this relationship.

e) Statistical techniques employed
The statistical techniques most employed are the correlation technique (88% of studies) and multivariate regression (81% of studies).In several studies, the authors employed more than one regression model, combining the independent variables available in the database and/ or collected by means of questionnaires.Table 2 presents the distribution of the frequency of use of the statistical techniques in the analyzed studies.f) Sources and data collection methods employed The data sources used in the studies include both primary and secondary sources.There is a predominance of the usage of secondary sources, that is, private databases of specialized companies and of institutions that work with other data but also possess data related to CSP and CFP, as is the case for Dow Jones.Table 3 presents the frequency distribution of the use of different databases used in the studies.Table 3 shows that the item "other databases" also has a relatively high frequency.To understand this frequency, the studies that made broad use of the data source were identified.In more recent years (2007,2008,2009), there was an increase in the publication of articles analyzing the CSP-CFP relationship in locations such as Jakarta, Guyana, Indonesia, Spain, China, and Australia.In these studies, the respective authors use specific databases that possess data that refer to the study's scope of interest.

g) Evolution of the volume of these publications
The historical series confirms a trend in the increase in publications exploring the CSP-CFP relationship.This trend is patterned by the increase in empirical studies, as shown in Figure 3, which presents the evolution of the volume of publications.During the 15-year period analyzed, 58.7% (34 articles) refer to the previous 3 years and 77.6% (45 articles) refer to the previous 5 years.Practically, the entire set of the 58 articles analyzed focused on some particular aspect of the CSP and CFP relationship.However, invariably, independent of the specific question being investigated, it was revealed that this relationship would be positive, negative, or null.It is worth noting that some articles, upon developing more than one test, thus found more than one type of existing relationship.
The majority of articles, 38 (65.5%), found a positive relationship; 11 articles (19.0%) found a negative relationship, and 18 (31.0%)found a neutral or inexistent relationship.Thus, the empirical observations reinforce stakeholder theory, in spite of the caveats.A synthesis of the main results of the analysis is shown in Table 4, which shows that the statistical analysis demonstrated a positive relationship (+) between the variables CSP and CFP.However, some studies did not find a relationship between these variables and are indicated by the letter "N", while other studies found a negative (-) relationship between these variables.In some cases, the variation in terms of the relationship between CSP and CFP was obtained only from changes in the independent variables or the control variables in the statistical model being tested.

b) Inconsistencies in Stakeholder Theory
Although there is a minority of empirical studies where the results do not reinforce stakeholder theory, there is a theoretical gap in the explanation of these cases.In several cases, the authors seek to indicate potential reasons for this result; however, these reasons seem to form only a set of exceptions and not a systematized explanation that can be incorporated into the theory.
Another aspect that lacks a more sustainable explanation is the set of variables that should compose the CSP construct.In the evaluated studies, a reasonable variation of elements that compose the CSP can be observed.In addition, with regard to some of these variables, such as the variable of the environment, some studies did not observe a positive relationship with CFP (Moore & Robson, 2002;Makni, Fracoeur, & Bellavance, 2009;Bouslah, Zali, Turcotte, & Kooli, 2010).In short, the definition of a set of variables that compose the CSP lacks a more encompassing explanation from the theory.
A third point that emerges from the empirical studies and that does not find a consolidated explanation in the theory is related to the time inherent to the causal relationship between CSP and CFP.Just 22% (13 articles) of the studies analyzed the CSP-CFP relationship considering the reference year and the previous year (year prior to the reference year), meaning that the social performance is reflected in the financial performance of the previous year.The majority of empirical studies assume a synchronous relationship between CSP and CFP, while some understand that a certain period of time must pass before the effect of one variable can be reflected in the other (Salama, 2005;Callan & Thomas, 2009).Although there are arguments to justify one design over the other, a general explanation for this aspect is still not incorporated into the theory.

c) Advances in Stakeholder Theory
During the studied period, important theoretical proposals emerged that were the target of empirical studies in search of verification.The first case to mention is related to good management theory and slack resource theory (Waddock & Graves, 1997), which aim to explain the interdependent relationship between CSP and CFP.In this regard, the majority of researchers are inclined to opt for good management theory rather than slack resource theory as the conceptual assumption; however, as highlighted by Gujarati (2006, p.17), the causal relationship between variables is a conceptual option of the researcher.
Another theoretical proposal supported by the empirical studies is related to the instrumental dimension of the stakeholder theory (Donaldson & Preston, 1995), which supports the concept of stakeholder management.Several empirical studies support this theoretical proposal, such as those by Berman et al. (1999) and Van der Laan, Van Ees, and Van Witteloostuijn (2008).

FInAl ConSIDERATIonS
As indicated by Atkinsons, Banker, Kaplan, and Young (2000), providing information on financial performance and other (non-financial) factors is a function of managerial accounting.In this regard, the contribution of studies on the relationship between the financial and social performance of companies is of accounting interest because it may provide support for decision making.
The results of this analysis demonstrate that, in fact, there has been an increase in the publication of empiricalquantitative articles in recent years because, over the 15 years analyzed (1996 to 2010), 58.7% (34 articles) refer to the last 3 years (2008 to 2010).The last study that reviewed empirical articles on this topic, that of Beurden and Gössling (2008), did not include the majority of these studies because only articles published after 2007 were analyzed, further justifying the current study.
The association between CSP and CFP proved to be predominantly positive and aligned with previous studies of this nature, such as those by Orlitzky (2001), Margolis and Walsh (2003), Orlitzky, Schmidt, and Rynes (2003), and Beurden and Gössling (2008).This evidence, which contrasts with older surveys, such as those of Ullmann (1985) and Wood and Jones (1995), is not sufficient to completely convince the scientific community of this relationship.
Another relevant point is that the lack of a theoretical foundation in empirical studies that was indicated by Ullmann (1985) and Wood and Jones (1995) has been overcome, given that this lack of a theoretical foundation was indicated as only a deficiency in subsequent studies by Margolis and Walsh (2003), Orlitzky, Schmidt, and Rynes (2003), and Beurden and Gössling (2008).The present stu-dy corroborates this assertion based on the theoretical advances in stakeholder theory observed mainly in the short span of time between 1995 and 1997, in particular, descriptive, instrumental, and normative perspectives (1995), instrumental stakeholder theory (1995), stakeholder salience (1997), good management theory (1997), and slack resources theory (1997), which provide conceptual support for the theoretical-empirical research analyzed in the period from 1996 to 2010.
As a suggestion for future studies to continue this research, a statistical meta-analysis study should be developed, as per Orlitzky, Schmidt, and Rynes (2003), to investigate aspects of this research field that remain obscure.Examples of these aspects are the following: the common characteristics of studies that indicate a negative association between CSP and CFP, the search for explanations for the behavior of some variables of debatable behavior in the construct of CSP (e.g., the environment), or explanations for the interference of the temporal lag in the causal relationship between CSP and CFP.
It is expected that the results of the current study will contribute to future studies in the field on both the methodological and theoretical fronts.In particular, methodological studies will benefit from the identification of the most accepted definitions in the specialized literature of the variables that comprise CSP and CFP as well as the identification of the main statistical methods employed in the research.With respect to the theoretical front, the following questions were addressed: what variables influence the CSP/CFP relationship, and what is the state of the art in the concept of the causal and temporal relationship between CSP and CFP?
(stakeholder) in the Text AND (performance) in the Text.The following periodicals were searched: Academy of Management Journal; Accounting, Organizations and Society; Environmental Management; Journal of Banking & Finance; Journal of the Academy of Marketing Science; and Strategic Management Journal.ISI returned 46 articles from this search.

Figure 1
Figure 1Variables Measuring Financial Performance

Figure 3
Figure 3Historical Series of CSP-CFP Publications

Table 1
CSP Function in the CSP-CFP Relationship

Table 3
Databases Used

Table 4
Synthesis of the Results Obtained from the Sample of Empirical Articles Analyzed