(PDF) Does it Pay to Be Good...And Does it Matter? A Meta-Analysis of the Relationship between Corporate Social and Financial Performance

DOES IT PAY TO BE GOOD?

A META-ANALYSIS AND REDIRECTION OF RESEARCH ON THE RELATIONSHIP

BETWEEN CORPORATE SOCIAL AND FINANCIAL PERFORMANCE

Joshua D. Margolis

Harvard University

Boston, MA 02163-7819

Tel: (617) 495-6444

Fax: (617) 496-6568

email: jmargolis@hbs.edu

Hillary Anger Elfenbein

Haas School of Business

University of California

Berkeley, CA 94720-1900

Tel: (510) 643-9700

Fax: (510) 643-1412

email: hillary@post.harvard.edu

James P. Walsh

Ross School of Business

University of Michigan

Ann Arbor, MI 48109-1234

Tel: (734) 936-2768

Fax: (734) 936-0282

email: jpwalsh@umich.edu

July 26, 2007

Acknowledgments:

We are indebted to Desiree Schaan for her assistance with coding the articles. We also appreciate

insightful comments by Christopher Marquis and Nitin Nohria, research assistance by Alyssa

Bittner-Gibbs, and the collegiality of Janet Kiholm Smith and Daniel Turban for sharing

additional information about their published studies.

2

DOES IT PAY TO BE GOOD?

A META-ANALYSIS AND REDIRECTION OF RESEARCH ON THE RELATIONSHIP

BETWEEN CORPORATE SOCIAL AND FINANCIAL PERFORMANCE

Abstract

The empirical link between corporate social performance (CSP) and corporate financial

performance (CFP) has been steadily investigated for 35 years. We conduct a meta-analysis of

192 effects revealed in 167 studies. The overall effect is positive but small (mean r=.13, median

r=.08). Looking deeper, we analyze these effects across nine categories of CSP. We find that the

association is strongest for the analysis of the specific dimensions of charitable contributions,

revealed misdeeds, and environmental performance and when CSP is assessed more broadly

through observer perceptions and self-reported social performance. The association is weakest

for the specific dimensions of corporate policies and transparency and when CSP is assessed

more broadly through third-party audits and mutual fund screens. Although the results suggest no

financial penalty for CSP, they indicate at least as strong a link from prior CFP to subsequent

CSP as the reverse. We conclude that if future research on the link persists, it should meet a

number of minimum standards. Ideally, though, efforts to find a link should be redirected to

better understand why companies pursue CSP, the mechanisms connecting prior CFP to

subsequent CSP, and how companies manage the process of pursuing both CSP and CFP

simultaneously.

Text: 33 pages

Figures and Tables: 25 pages

References: 19 pages

3

It’s 8:30am on a Friday in July, and Carol B. Tomé is starting to sweat. The chief financial officer of Home Depot

Inc. isn’t getting ready to face a firing squad of investors or unveil troubled accounting at the home improvement

giant. Instead, she and 200 other Home Depot employees are helping to build a playground replete with swings,

slides, and a jungle gym at a local girls’ club in hardscrabble Marietta, Ga. … Is this any way to build shareholder

value at Home Depot, where the stock has been stuck near $43, down 35% from its all-time high? (Business Week,

2005)

Can a corporation create wealth and do it in a way that does not harm society, and, in the

best of all worlds, even redress social ills? The question of whether “doing good and doing well”

converge has waxed and waned over the past century (Morrissey, 1989; Wells, 2002), and it has

preoccupied thinkers for nearly 2000 years (Avi-Yonah, 2005). Some theories of the firm

emphasize reaching beyond a single-minded focus on wealth creation to attend broadly to

society’s needs, but the theory that now dominates legal and economic scholarship does not

(Allen, 1992; 1993). Commonly known as the “nexus of contracts” theory, it sees the firm as “a

legal fiction which serves as a focus for a complex process in which the conflicting objectives of

individuals (some of whom may ‘represent’ other organizations) are brought into equilibrium

within a framework of contractual relation” (Jensen and Meckling, 1976: 311). Even as

competing models of the firm gain influence, they must contend with this prevailing view (Blair

& Stout, 2006; Freeman, Wicks, & Parmar 2004), which may well continue to shape

assumptions about the firm for the foreseeable future (Hansmann & Kraakman, 2001).

Anyone who argues that the ultimate purpose of a firm involves anything more than

enhancing shareholder value must come to terms with this dominant theory. Attempts to mitigate

a firm’s ill effects on society or to fund projects that might directly benefit society are subjected

to a rigorous financial analysis. Indeed, the prevailing theory argues that society is best served if

these attempts can clear such a financial hurdle. In his appraisal of the longstanding controversy

regarding the purpose of the firm, Jensen (2002: 239) argued that “200 years’ worth of work in

economics and finance indicate that social welfare is maximized when all firms in an economy

maximize total firm value.” It is a tidy logic that puts the onus on corporate critics and social

advocates alike to show how a corporation’s social investment must benefit its shareholders.

Business Week’s (2005) skepticism about Home Depot’s community investment practices

certainly reflects this orientation.

4

This theory may be so influential now because it allows managers and regulators alike the

freedom to (relatively easily) restructure the firm’s assets to best meet the demands of global

competition. The globalization of the firms’ factor and product markets, and its implications for

management and corporate governance, is by now a very familiar story (Bradley, Schipani,

Sundaram, & Walsh, 1999; Jensen, 1993; Parker, 1996). As globalization ushered in a period of

hypercompetitive business practices (D’Aveni, 1994), companies have struggled to survive, well

enough thrive. To see the firm as a bundle of contracts facilitates change. After all, contracts can

be renegotiated, even if the social costs are high (Shleifer and Summers, 1988; Uchitelle,

Battenberg and Kochan, 2007). And the changes driven by this economic logic have been

enormous. The conglomerate merger wave of the 1960s was unraveled in the 1980s (Shleifer &

Vishny, 1991), as firms shed their unrelated business units (Comment & Jarrell, 1995) and

learned to leverage their “core competencies” (Prahalad and Hamel, 1990) to meet their new

competitive realities. Then in the 1990s and early 2000s companies combined anew, searching

for the scale economies and competitive advantages considered essential to prosper in a global

marketplace – even though the economic benefits have sometimes proven elusive (Moeller,

Schlingemann & Stulz, 2005).

Seen in this context, it is no surprise to discover that performance, and especially

corporate financial performance, became the dominant dependent variable in organizational

research over the past thirty years (Walsh, Weber, & Margolis, 2003). Even if performance was

ancillary to the topic at hand, it served to legitimate the work as academically credible and

practically relevant (Staw, 1984). Indeed, the study of organizations is marked by all manner of

attempts to link management practices to corporate financial performance. Work on strategy

(McGahan & Porter, 1997), research and development (Wieser, 2005) and human resource

management (Delery & Doty, 1996; Huselid, 1995), to name just a few, attempt to establish a

connection between corporate practices and their financial results. The work on corporate social

performance is no exception.

Scholars have been searching for a link between corporate social performance (CSP) and

corporate financial performance (CFP) for thirty-five years. If only doing good could be

connected to doing well, then companies might be persuaded to act more conscientiously,

whether in cleaning up their own questionable conduct (Campbell, 2006) or in redressing societal

ills (Porter & Kramer, 2006). A positive link between social and financial performance would

5

legitimize corporate social performance on economic grounds, grounds that matter so much these

days (Useem, 1996). It would license companies to pursue the good—even incurring additional

costs—in order to enhance their bottom line and at the same time contribute more broadly to the

well-being of society.

The influence of this economic reasoning was apparent in the very first empirical CSP-

CFP study. Bragdon and Marlin (1972) motivated their research by examining whether or not

virtue must be its own reward. They looked at this question from both a manager’s and an

investor’s perspective:

Proponents [of what they called the orthodox economic logic] argue that corporate

managers can either control pollution or maximize profits but that the former can be

accomplished only at the expense of the latter. From the investor’s perspective, this in

turn implies that he can either invest in a profitable company or a “good” company

(which protects its environment) but that no company is likely to be both. (Bragdon &

Marlin, 1972: 9).

These words were written on the heels of Friedman’s (1970) well-known criticism of a firm’s

corporate social responsibility initiatives. Friedman took direct aim at any firm that contemplated

such activity, considering such investments to be theft and political subversion. In his view,

executives were taking money that would otherwise go to the firm’s owners in order to pursue

objectives that the executives, under the sway of a minority of voices, selected in a manner

beyond the reach of accepted democratic political processes. But when Bragdon and Marlin

(1972: 17) found a positive CSP-CFP relationship, they could comfortably remove any conflict

by concluding, “[W]e hope that we have made a step in the direction of laying to rest the

economic model that poses the alternative.” If they only knew. Thirty-five years later, Nakao,

Amano, Matsumura, Genba, and Nakano (2007:107) were still investigating this very same

question: “to examine, by multiple linear regression analysis, whether environmental

performance has a significantly positive effect on financial performance.” One hundred and sixty

seven studies, investigating 192 CSP-CFP effects, have been conducted since 1972. Figure 1

profiles this steady research activity. Our goal is to take stock of this research stream and with a

meta-analysis, see if we can answer the question of whether it pays to be good.

--------------------------------

Insert Figure 1 about Here

--------------------------------

6

We are not the first to distill this longstanding line of research. It is a testimony to the

power of the question that sixteen reviews of CSP-CFP research have already assessed whether

doing good pays. The first review was published nineteen years ago (Aldag & Bartol, 1978).

Since then, another twelve appraisals and three formal meta-analyses have appeared in print.

They all try to keep pace with the heavy volume of work investigating the relationship between

CSP and CFP. Table 1 captures the reviews as they appeared through time, the number of CSP-

CFP studies each review examined, and the citations each has garnered over the years.

--------------------------------

Insert Table 1 about Here

--------------------------------

Taken together, scholars have turned to these reviews 1,445 times for guidance. The

value of a review is a function of the breadth of extant work considered and the insights the

authors bring to its evaluation. The rigor of the analysis certainly matters too. Today, both

consumers of literature reviews (Bies, Bartunek, Fort, & Zald, 2007) and research

methodologists (Rosenthal, 1991) see formal meta-analyses as more valuable than a scholar’s

idiosyncratic reading of a literature. Indeed, later views of the CSP-CFP literature criticize earlier

reviews on just this point (e.g., Orlitzky, Schmidt, & Rynes, 2003). Not surprisingly, the three

most recent reviews of this literature have employed meta-analysis: Orlitzky, et al. (2003)

analyzed 52 CSP-CFP studies; two years later Allouche and Laroche (2005) analyzed 82 CSP-

CFP studies; and most recently, Wu (2006) analyzed 121 studies, with 39 of them focused on the

CSP-CFP relationship. Our goal is to expand and deepen these efforts in two ways. First, we

offer a comprehensive appraisal of the 167 studies conducted to date and, second, we draw

implications both for the CSP-CFP relationship and for future research.

Our paper proceeds in five steps. First, we provide a backdrop to our meta-analysis,

describing theoretical approaches to CSP and the CSP-CFP connection. Second, we lay out the

methodology for our meta-analysis. Third, we present the results in two forms, in the aggregate

and then by the type of study. Fourth, we assess the implications of these results, interpreting

what the results do and do not indicate about the relationship between CSP and CFP. And fifth,

we identify two paths for future research, and suggest that a new one—rather than the well-

traveled familiar one—will best honor the enduring motivation and prodigious efforts behind

prior research.

7

Corporate Social Performance and the Quest for a Link to Corporate Financial

Performance

Despite years of theoretical and empirical attention, researchers have encountered

significant challenges in both specifying and operationally defining the CSP construct (Barnett,

2007; Clarkson, 1995; Frederick, 2006; McWilliams, Siegel & Wright, 2006; Wood, 1991;

Wood & Jones, 1995). Prior reviews of the CSP-CFP work often decry a range of theoretical and

methodological faults and, in so doing, promote a continuing research stream that might rectify

the problems.

To date, corporate social performance has been theoretically defined in two basic ways.

One approach casts social performance as a multidimensional construct, encompassing a

company’s efforts to fulfill multiple responsibilities — economic, legal, ethical, and

discretionary (Carroll, 1979, 1999) — or encompassing a company’s principles, processes of

response to rising issues, and observable practices and outcomes (Wartick & Cochran, 1985;

Wood, 1991). A second approach casts social performance as a function of how a company treats

its stakeholders (Campbell, 2007; Clarkson, 1995; Cooper, 2004; Post, Preston, & Sachs, 2002).

Although theorists attempt to distinguish corporate social performance from corporate social

responsibility (CSR), sometimes subsuming CSP under the umbrella of CSR and sometimes the

reverse (Barnett, 2007; Carroll, 1979, 1999; Wood, 1991), the terms corporate social

performance and corporate social responsibility (CSR)—or “socially responsible behavior”—are

often used interchangeably in empirical studies. Despite extensive theoretical development,

researchers have encountered significant challenges operationally defining the theoretical

construct of corporate social performance (Clarkson, 1995; Wood & Jones, 1995). As a result,

indicators and measures of CSP vary widely and tend to capture either a single specific

dimension, such as philanthropic contributions or pollution control, or broad appraisals of CSP

as a whole. The increasing influence of stakeholder theory on the study of CSP has corresponded

with increased use of Kinder Lydenberg Domini’s Socrates database, which rates companies

across dimensions that reflect attention to different stakeholder groups

(http://www.kld.com/research/socrates/index.html).

Just as theoretical elaboration of the CSP construct has coalesced around two main

models, so too have theoretical accounts of the link between CSP and CFP (Jones, 1995; Preston

& O’Bannon, 1997). One model treats CSP as a distinctive resource—a way of treating others,

8

for example, or a way of running the company’s operations—that substantively generates

benefits or reduces costs, both of which improve financial performance. Heightened benefits may

include employee effort that emerges from treating them well, or innovative products and access

to markets that emerge from aiding non-profit enterprises. Decreased costs may include the

avoidance of potential penalties and regulation as a result of clean and safe operations, less

contracting friction with stakeholders as a result of honest dealing, or lower material costs from

reduced levels of pollution and waste. For theories that fall within this first broad model, the

mechanism that turns CSP into CFP is the value-creating impact of the efforts to do good. Those

efforts have the effect of reducing costs or increasing revenues.

In contrast, another set of theoretical accounts suggests that the appeal of CSP, rather

than its substantive impact, generates financial returns. Independent of the actual effects of

efforts to do good, the second model suggests that the appearance of doing good (or the

perception among key stakeholders that a company is doing good) generates demand for and

commitment to the company’s stock, jobs, or products. The value-creating mechanism is the

appearance of CSP. That appearance increases demand and commitment, directly driving up the

stock price, for example, or indirectly reducing hiring costs by intensifying employee

commitment, or indirectly generating revenue by increasing the likelihood that consumers will

purchase the company’s products.

Although the mechanisms that connect CSP to CFP may both be at work, and thus the

two models may well overlap, the underlying mechanisms do indeed differ. Consider two ways

in which helping non-profit organizations might contribute to a company’s financial

performance. The first model suggests that when a company collaborates with non-profits, the

company may strike upon unforeseen markets or innovative products, which open new sources

of revenue (Kanter, 1999). The second model suggests that by collaborating with non-profits, a

company gains because the public develops a general impression that the company is a good

citizen, which makes people more likely to pursue the company’s products and jobs, or to permit

the company to expand without extensive oversight.

Theories of how CSP and CFP are connected, as well as the evolving definition of CSP

(Carroll, 1999), both help explain why studies of the CSP-CFP link have proliferated (see Figure

1). With multiple dimensions and many stakeholders treated as indicators of CSP, and with

evolving specifications of their link to CFP, each new study can promise to contribute to a

9

definitive assessment of the CSP-CFP relationship. Each new study promises to isolate a

different dimension of CSP, or to reflect an improved conceptualization of the construct or its

theoretical connection to CFP. The diversity of CSP variables suggests that it may be

inappropriate to lump all studies and their effects together. As a result, we also analyze CSP-CFP

effects across nine categories of CSP.

METHODS

Study Selection and Inclusion

Our review of research on corporate social and financial performance encompasses

studies from 1972 through 2007. We selected studies to include in the meta-analysis in five

ways. First, we collected articles covered in the sixteen prior reviews of the literature that are

listed in Table 1. Second, we searched the ABI/Inform, JSTOR, and EBSCO databases using the

keywords “social performance,” “social responsibility,” “socially responsible,” “charitable,”

“philanthropy,” and “environment.” Third, we manually checked the table of contents of seven

of the top journals in the management field (Academy of Management Journal, Administrative

Science Quarterly, Journal of Management, Journal of Organizational Behavior, Organization

Science, Organizational Studies, and Strategic Management Journal). Fourth, we learned of

some papers through informal efforts, such as inquiries with colleagues, suggestions from

seminar participants where we presented related work, and papers mentioned by colleagues.

Fifth, we identified articles that were referenced by studies found using the four earlier methods.

To be included in this review, a study had to satisfy the following three criteria. First, the

manuscript had to include a measure of CSP for individual firms. Because CSP has traditionally

been defined broadly and operationally defined in many different ways, we considered any

empirical research that fit past conceptualizations. Second, the manuscript had to include a

measure of CFP for individual firms, usually an accounting rate of return or a market measure of

performance. Third, the manuscript had to report an effect size for the association between CSP

and CFP at the firm level or provide enough information for us to calculate an equivalent effect

size for this association. A total of 167 studies satisfied these criteria.

Because the majority of studies reported a zero-order correlation as the relationship

between CSP and CFP, in the other cases we converted the reported effects into the equivalent of

an effect size r (Rosenthal & Rosnow, 1991). In the case of multivariate analyses, we used

standardized regression betas if reported, or calculated the effect size r if a t-test, F-test, or Z-test

10

statistic was provided.

1

In such cases, the resulting effect size r is the equivalent of a partial

correlation that accounts for the influence of any control variables that were used in the original

analysis. In the case of t-tests that compared groups differing in their levels of CSP, or when

authors provided information on means, standard deviations, and sample sizes that could be used

together to calculate a t-statistic, we also converted such effects into an effect size r. All values

were coded so that positive effects represent a financial benefit for high CSP and negative effects

represent a financial cost for high CSP. Thus, the studies that were included could be

summarized in terms of a single indicator of effect size, which enabled us to make direct

comparisons across different studies.

2

In order to be conservative with respect to estimating the

CSP-CFP association, some studies were included if the text mentioned that the association was

tested but not statistically significant, in which case the effect size was presumed to be zero. All

effect sizes were computed by the second author, or computed by a doctoral student in finance

and checked by the second author, with an inter-coder reliability of .95.

Coding Procedure

We attempted to code the primary attributes of the empirical studies that the original

authors of the empirical studies consistently reported. Either the second author or a doctoral

student in finance coded each study, and both coded a subset of 50 studies in order to confirm

sufficient inter-rater reliability for continuous measures and inter-rater agreement for categorical

measures. The following five characteristics of each study were coded.

Type of CSP. Studies vary in the indicator used to measure CSP, sometimes opting to

examine a specific dimension of CSP and sometimes opting for a broad appraisal of CSP. We

sorted the collection of studies into one of the nine categories below, with the first five

representing specific dimensions of CSP and the last four representing different approaches for

capturing CSP broadly. If a single study reported results using measures that fell into different

categories, we sorted each separate result into its most appropriate category, resulting in a total

of 192 effects in 167 studies. However, we did not double-count by sorting any effect into more

than one category. Inter-rater agreement for categorizing the type of CSP was .96. These are the

nine categories, with the first five representing specific dimensions of CSP and the last four

1

We used the formula r = sqrt [F/(F+df)] when F-test statistics were reported, r = sqrt [t^2/(t^2+df)] when t-test

statistics were reported, and r = sqrt [Z^2/N] when Z-test statistics were reported (Rosenthal & Rosnow, 1991).

2

Because it was necessary to express the results of each study using a common statistic, we unfortunately had to

exclude 12 articles that reported effects using unstandardized coefficients, in cases when these coefficients could not

be converted into an effect size r based on other reported information.

11

representing different approaches for appraising CSP more broadly.

(1) Charitable contributions. This included cash donations or the establishment of a

philanthropic foundation. When the authors provided specific information, we excluded in-kind

donations, given that these often serve instrumental purposes such as marketing or the disposal

of obsolete inventory (Seifert, Morris, & Bartkus, 2003).

(2) Corporate policies. These studies examine a range of corporate policies, such as

companies that divested from apartheid South Africa, firms that did business in apartheid South

Africa and signed the Sullivan Principles for fair treatment of citizens, banks that offered low

income loans, and defense contractors that agreed to a code of ethics.

(3) Environmental performance. This category includes measures of impact on the

environment, whether objective or self-reported. We coded as objective any information

indicative of corporate environmental practices assessed by or reported to third parties, such as

the toxic release inventory, fines paid, and energy reduction expenditures. Objective data

includes self-reported data that is under regulatory oversight by third parties (e.g., Superfund site

liabilities). Self-reported data includes company insiders’ subjective perceptions of their

environmental performance. If a misdeed involved environmental practices, we included the

effect in this category rather than in category four below. So too, if self-reported data referred to

environmental performance, we included the effect in this category rather than in category six

below.

(4) Revealed misdeeds. This includes the public announcement of arrests, fines, guilty

verdicts in lawsuits, involuntary recalls, and other actions that indicate socially irresponsible

behavior.

(5) Transparency. The release of information by a company itself in publicly available

documents, such as annual reports, is used as an indicator of a company’s CSP. This category

includes all studies that use the disclosure itself—rather than the substance of what is being

discussed in the disclosure—as the indicator of social performance. The underlying aim of these

studies is to determine whether transparency pays. When researchers treated the specific content

disclosed as the indicator of CSP, the study was coded in one of the above categories capturing

that specific content (such as a misdeed).

Four categories reflect different ways researchers attempt to capture companies’ CSP

more broadly, rather than specific dimensions of the construct. These four forms of broad

12

appraisal include:

(6) Self-reported social performance. One method for capturing a company’s social

performance more broadly used surveys that ask companies to report their own conduct in

response to journalists’ or researchers’ inquiries. The difference between this and the previous

category, transparency, is that the present category involves a researcher or media outlet

approaching the company for its self-report, rather than a voluntary and active decision on the

part of the company to release information. For example, companies are asked to rate the

importance of social responsibility and philanthropy (Goll & Rasheed, 2004). Self-reported

social performance related to the environment is included only in the environmental performance

category above.

(7) Observers’ perceptions. Two methods of assessing corporate social performance rely

on external observers. The first method relies upon observers’ intuitive impressions of a

company’s CSP. Observers include industry insiders, executives at other companies, business

school faculty members, and undergraduate business students. The most common form of

observer perceptions involves ratings from the Fortune magazine database of most admired

companies (60.0%).

(8) Third-party audits. The second method that uses observers to assess corporate social

performance involves the systematic assessment of data by investigators who evaluate a

company along a set of criteria. We refer to these as third-party audits. The most common

examples are the Kinder Lydenberg Domini (KLD) index, which evaluates companies on eight

dimensions, its precursor developed by the Council on Economic Priorities (CEP), and

equivalent organizations in other countries. Other examples include the U.S. Department of

Labor and Working Woman magazine, which both award recognition for companies whose labor

policies are deemed especially progressive. We also include the assessments of investment fund

managers, except in the case of assessments that yield a marketed investment vehicle, which we

categorize instead as screened mutual funds. In the case of audits that reported results about one

distinct category of CSP already listed above, notably the environment, we included those

studies only within the distinct category.

(9) Screened mutual funds. A growing number of studies examine the performance of

mutual funds that use screens to limit the companies included in the funds to those meeting

certain criteria of social performance. These screens are considered indicators of included

13

companies’ general CSP. We excluded those papers that tracked companies screened on the basis

of their industry membership (e.g., gambling, tobacco) rather than on a company-level basis. We

include those studies that compare entire stock performance indices, such as those comparing the

Domini 400 versus the Standard & Poor’s 500.

Type of CFP. We list the specific measures of financial performance examined by the

original authors. Further, we coded measures into two broad categories: accounting-based

measures of financial returns (e.g., Return on Assets, Return on Equity) versus market-based

measures of financial value (e.g., stock returns, market/book value ratio). A small number of

studies used measures of financial performance that did not fit this dichotomy (e.g., bond returns

in D’Antonio, Johnsen, & Hutton, 1997; observer ratings of “economic performance” in

Clarkson, 1988; dividend yields in Greening, 1995) and these were included in overall effects but

not listed in the breakdown by category. Inter-rater agreement for categorizing CFP measures

was .97.

Number of firms included. We recorded the total number of firms that were included in

a study’s sample. For two sets of studies, this was not always possible. First, for studies of the

stock market reaction to specific events, the authors often reported the number of events rather

than firms, and a given firm could generate more than one of the events. Second, studies of

screened mutual funds rarely listed the number of underlying securities. However, some studies

compared a specific portfolio of companies to a benchmark consisting of an entire marketplace,

in which case we noted as the number of firms the specific portfolio that the authors described.

Inter-rater reliability for determining the number of included firms was 1.00.

Timing of CSP and CFP measurements. We recorded the year or range of years for

both the CSP and CFP measures. Although many studies had a stated goal of examining the

influence of CSP on CFP (indicating a particular direction of causality), there are three main

choices regarding the timing of these measures: the measure of CSP precedes the measure of

CFP; the measure of CFP precedes the measure of CSP; or they are measured concurrently

(operationally defined as occurring within 12 months of each other). Event studies—in which

researchers observe the stock market reaction to discrete news announcements—were coded as

having the CSP measurement preceding CFP because the timing of both CSP and CFP were

specified precisely in such studies. We coded as concurrent those studies in which the

measurement of CSP and CFP were nested. For example, Alexander and Buchholz (1978)

14

measured CSP in 1971-1972 and CFP for the period 1970-1974. We also coded as concurrent

any studies of screened mutual funds that were actively managed, with the logic that fund

managers continually monitor the current social performance of firms included in their

portfolios. However, we coded CFP to precede CSP in those studies in which researchers

conducted retrospective analyses of the financial performance of stocks that were later included

in screened funds. Inter-rater agreement for the timing of CSP and CFP measures was .94.

Control variables. We noted whether control variables were incorporated into the

estimate of the CSP-CFP effect size. We coded for the most common control variables of

industry, firm size, and risk. Some studies are coded as having no control variables even though

the authors did include controls in the study because the effect size for the CSP-CFP value was

taken from a zero-order correlation matrix that did not account for the effect of control variables.

In addition, we coded the methodological attribute of whether effect sizes resulted from event

studies. These effects are coded as including all control variables because, in event studies, each

company serves as its own matched control when its stock price is compared before versus after

a news announcement.

Industries can vary in their social responsibility practices. Some industries may be

considered more “dirty” than others, such as heavy manufacturing or chemicals; some industries

may be growing versus declining; and stakeholders may vary in the degree of regulation and

scrutiny to which they subject different industries (Bowman & Haire, 1975; Griffin & Mahon,

1997; Spencer & Taylor, 1987). Reporting rules that apply to entire industries can promote

responsible behavior, but can also constrain it (i.e., mandating strict itemization for charitable

donations). We considered industry to be controlled either when it was explicitly entered as a

control variable in the authors’ original analyses, when it was incorporated into the research

design using samples matched on industry, or when the study sampled from within a single

industry.

Firm size is a worthwhile control variable because larger firms may have greater

resources for social investments, attract greater pressure to engage in CSP or—just the

opposite—succumb to a diffusion of responsibility (Wu, 2006). Wu’s (2006) recent review

regarding firm size indicated a small positive relationship between firm size and CSP and

between firm size and some measures of CFP. We considered firm size to be controlled either

when it was explicitly controlled in the original analyses or when the study sampled from firms

15

of similar sizes (e.g., the Fortune 500 focus on revenues, or assets, or the total number of

employees).

Firm risk is also an important factor to control because stable firms with lower risk

generally appear more likely to engage in CSP (Alexander & Buchholz, 1978; Brown & Perry,

1994; Chen and Metcalf, 1980; Cochran & Wood, 1984). Moreover, CSP has been linked to the

risk profile of firms (Orlitzky & Benjamin, 2001). Indeed, given the strong relationship between

risk and financial returns, O'Neill, Saunders, & McCarthy (1989) found that their CSP-CFP

correlations disappeared for risk-adjusted financial performance measures. We considered risk to

be controlled when it was included in the model explicitly as a control variable (e.g., regression

models or the CAPM financial model), when authors used a risk-adjusted measure of CFP or, in

the case of portfolio analyses, when the authors or mutual fund managers who constructed the

portfolio selected companies based on risk levels equivalent to a control sample or the larger

stock market.

Finally, we coded whether effect sizes resulted from “event studies”—in which the stock

price of a given company is observed before and after a specific event or announcement—

regardless of which of the nine specific types of CSP was represented by the event. These studies

are unique in that they are unusually precise because companies serve as their own matched

control and, when done correctly (McWilliams & Siegel, 1997), confounding events are

excluded. Event studies also isolate a specific mechanism for any association between CSP and

CFP, namely, the stock market’s reaction to news regarding a firm’s CSP. Inter-rater agreement

across all controls was .92.

RESULTS

Table 2 lists the studies included in this meta-analysis, the key attributes of each study,

and the effect size r for the CSP-CFP association. Table 3 summarizes the results of our

analyses, including the effect size overall, by timing of the CSP and CFP measures, and by type

of CFP measure, with all of these listed for all studies and separately for each CSP category.

Table 3 also lists the number of studies and total number of companies in each category of study,

the significance test for the size of the CSP-CFP effect (Rosenthal, 1991), and the results of a

heterogeneity analysis that indicates whether there are substantial differences in effect sizes

across the various studies (Hedges & Olkin, 1985; Rosenthal, 1991).

16

---------------------------------------

Insert Tables 2 and 3 about here

---------------------------------------

The diversity of CSP indicators and measures raises questions about whether these all

capture a single underlying construct. However, because of the widespread interest in corporate

efforts to do good and the contention that all indicators and measures of CSP capture, in some

way, the underlying propensity of a company to do good, we did calculate a single, cumulative

effect for all 192 effects. This analysis reveals a mean effect size of r=.132. As described above,

in the case of studies reporting the results of multivariate analyses, these effects are the

equivalent of a partial correlation that accounts for the effects of control variables. The median

effect size (r=.082) and weighted average effect size (r=.101), which accounts for the size of

each study, were lower than the mean effect size. That suggests that the overall mean is inflated

by large effect sizes from a small number of studies that used relatively smaller samples of

companies. When effect sizes are compared across the types of CFP measures, CSP generally

appears to predict accounting-based measures (r=.180) better than market-based measures

(r=.104). That said, market-based measures may be more appropriate for gauging the impact on

shareholder wealth (Mackey, Mackey, & Barney, 2007).

To examine the potential influence of moderator variables, we conducted additional

analyses. For results aggregated across all types of CSP, we did not find an influence on effect

sizes from including (or excluding) control variables, nor were there large aggregate differences

between studies in which the timing of CSP measures preceded, followed, or concurred with

measures of CFP. However, for each of the four studies that included all three types of timing

(Boyle, Higgins, & Rhee, 1997; McGuire, Sundgren, & Schneeweis, 1988; Preston & O’Bannon,

1997; Seifert, Morris, & Bartkus, 2003)—which arguably offer the most precise comparison—

there is a monotonic fall in the effect size from CFP→CSP (average r=.275) to concurrent

(average r=.120) to CSP→CFP (average r=.080). All four studies revealed this same pattern. A

binomial probability test suggests that this ordering of responses from the highest to lowest CSP-

CFP effect size is unlikely to occur by chance alone. Given that there are six different ways to

place the three types of timing in order from the highest to lowest, the chance of each study

showing this pattern is 1/6. The probability that all four studies would conform randomly to this

pattern is less than eight in ten thousand.

17

Event studies—which map precisely the stock market effects of releasing news regarding

CSP—appeared to have slightly larger effect sizes than those of conventional studies (r=.175

versus r=.118; median effects r=.189 versus r=.067). In event studies, the impact of a company’s

social performance is measured through a comparison of the stock market’s valuation of that

company’s stock preceding and following the announcement of positive or negative news. The

use of event-study methodology in research on the link between CSP and CFP has been

criticized (McWilliams & Siegel, 1997), most notably because these studies have used long

event windows, introducing the possibility that other events account for stock price movement,

and because, indeed, these studies have not adequately controlled for other confounding events

that could account for abnormal returns. Nonetheless, the consistency of our results for event

studies suggests that the market may read the announcement as new information indicating

future financial performance. The market may infer the company’s social performance to be an

indicator of the quality of management, to augur consumer demand for the firm’s products and

services as a result of the company’s social performance, to promise higher or lower costs

incurred from other stakeholders, or to provide greater insurance against damaging events

(Freedman & Stagliano, 1991; Peloza, 2006). Social performance may also attract demand for

the stock among investors with strong preferences for social responsibility (Mackey, et al., 2007)

or among market investors in general who believe that other investors will adjust their demand

for a firm's shares, for example if the company is now worthy of inclusion or exclusion from

socially screened funds. Alternatively, the market may simply assume that doing good generates

financial gains through some unspecified mechanism.

To investigate the relationship between CSP and CFP with greater precision, we sorted

studies into the nine categories described above and analyzed the cumulative effect within each

category. Effect sizes differed significantly across these categories, F(8, 183) = 4.12, p<.001. We

summarize the results by category in Table 3 and below, reporting the overall effects as well as

any noteworthy influence of moderator variables for those analyses with sufficient power to

warrant them.

Charitable Contributions

Thirteen studies examined the effect of corporate financial performance upon charitable

contributions. The average r was. 239. The effect was stronger when CFP was measured prior to

the philanthropy (r=.332) or after (r=.292) than when the two were measured concurrently

18

(r=.198). Studies using accounting measures of financial performance showed larger effects

(mean r=.281, median r=.203) than those using market-based measures (mean r=.147, median

r=.055). Taken together, these findings suggest that slack resources promote generosity towards

charitable endeavors (Seifert, et al., 2003). Companies are more able or willing, or they face

stiffer pressure, to donate when they do well.

Corporate Policies

Corporate policies are the one form of social responsibility without a significant

association to financial performance (r=.019, ns, weighted r=.038, median r=.005). However,

there is a trend in which prior financial performance does predict future socially responsible

policies (r=.111), but concurrent (r=.-.031) and CSP→CFP (r=.011) studies do not reveal any

meaningful trends.

Environmental Performance

A large sample of 44 studies examined environmental impact, including objective

measures such as the toxic release inventory, fines paid, and energy reduction expenditures, as

well as subjective perceptions of environmental performance. Effect sizes were larger for self-

reported (r=.190) than for objective environmental measures (r=.095).

Revealed Misdeeds

Announcements of negative events, such as regulatory violations, lawsuits, and fraud,

were the topic of 16 studies, generating an average r of .223, with a smaller r=.104 average when

weighted by sample size, due largely to the result of one outlier (Jarrell & Peltzman, 1985;

r=.563, N=22). These findings are consistent with Frooman’s (1997) earlier meta-analysis,

finding that the stock market reacts negatively to news announcements that a company has done

something socially irresponsible. Although companies are punished at the time misdeeds are

exposed (r=.227) and afterward (r=.239), a company’s financial performance—whether good or

bad—does not predict future revealed misdeeds (r=-.004). When misdeeds are revealed, the

market may interpret them as an indicator that the firm will incur greater costs from penalties or

from stakeholders less willing to cooperate, or will suffer lower revenues due to reduced future

demand for the firm’s products and services resulting from the company’s damaged reputation in

consumers’ eyes. Exposed misdeeds may also be read by the market as an indicator of poor

judgment among top managers, or of imminent decline in demand for the stock as investors with

a preference for responsible companies shy away from it.

19

Of course, this effect only captures the wealth effects for those caught doing some

misdeed. The wages of unrecognized sin may be quite handsome. Further, the effect is larger for

market-based measures (r=.239) versus accounting-based measures (r=.113), suggesting that the

mechanism for this effect is more likely to be the reaction of investors, rather than revealed

information about the health of corporate operations. This weak link to an accounting

performance measure is consistent with Staw and Szwajkowski’s (1975) work. They found that

the influence of having a munificent environment on corporate crime was related to the

munificence of the industry environment, rather than that of the individual firm.

Transparency

Fourteen studies examined the influence of transparency. The average effect size r was

.078 and median was r=.024. Effect sizes are larger for those studies that control for firm size

(r=.102 versus r=.034), but smaller for those studies that control for industry (r=.046 versus

r=.156). The slight positive trend in the mean but not median value can be attributed largely to

two studies (Anderson & Frankle, 1980; Verschoor, 1998) that did not control for industry.

Results were stronger for CSP measured before CFP (r=.191) than for CFP measured before CSP

(r=.079) or concurrent measurement (r=.029). Taken together, these results suggest that the

market reacts positively to company disclosures regarding socially responsible behavior.

Self-reported Social Performance

For nine studies using self-reported social performance, the average r was .210 (weighted

r=.128). The largest effects were found in one study in which executives were asked to report

both social and financial performance of the company in a single survey (Reimann, 1975, r =

.570, N=19) and in one study analyzing the relationship between responses of rank-and-file

employees and the company’s financial performance of the previous five years (Hansen &

Wernerfelt, 1989, r=.482, N=60). Effect sizes were smaller for studies that controlled for risk

(r=.039 versus r=.292), but larger for studies that controlled for industry (r=.312 versus r=.076).

The three of these studies that controlled for risk (Aupperle, Carroll, & Hatfield, 1985; O’Neill et

al., 1989; Starik, 1990) together reveal an average effect of r =.039. There did not appear to be

an overall influence based on timing, given that the CSP→CFP association of r=.272 was based

on a single study, versus r=.200 for CFP→CSP and r=.171 for concurrent measurement.

Observers’ Perceptions

20

For the 25 studies that use observers’ perceptions, the average r is .287. Studies using

accounting measures of financial performance reveal a stronger relationship (r=.320) than those

using market-based stock performance (r=.190). The effect size is smaller for studies that control

for risk (r=.167 versus r=.316), but larger for studies that control for industry (r=.440 versus

r=.131). When analyzed according to timing, the findings reveal a stronger relationship between

CFP that was measured prior to CSP (r=.328) than either concurrent CFP and CSP (r=.279) or

CSP measured prior to CFP (r=.157). In general, these studies show that there is a reasonably

sized relationship between corporate social performance, as measured by observers' perceptions,

and corporate financial performance. However, the results suggest that the strongest direction of

causality goes from CFP to CSP, which is consistent with the possibility that observer

perceptions are biased by a company’s recent financial performance (Brown & Perry, 1994).

Third-party Audits

The 28 studies that rely on third-party audits to assess CSP reveal an average r of .080

and an average r weighted by sample size of .037. Studies that used accounting measures of

financial performance had an average r of .114 versus .059 for those using market-based

measures of financial performance. When analyzed according to temporal direction, the findings

reveal a stronger relationship when CFP was measured prior to CSP (r=.142) compared to

concurrent measurement (r=.041) or measuring CSP prior to CFP (r=.096). Taken together, these

28 studies suggest a mild relationship between CSP and CFP, but this link is unduly influenced

by large effect sizes among studies with smaller samples (e.g., Shank, Manullang, & Hill, 2005;

r=.261, N=11) and appears to flow primarily from CFP to CSP, rather than the other way around.

Screened Mutual Funds

A growing number of studies examine the performance of mutual funds that use screens

to limit the companies included in the funds to those that meet certain criteria of social

performance. The increasing number and sophistication of these studies warrant a detailed

review by financial economists. Our analysis of 29 studies in this category reveals an average r

of .024 (median r=.021). Within this group of studies most effects were negligible, but there

were also several outliers showing gains for socially screened mutual funds relative to other

funds (e.g., Luck & Pilotte, 1993; r=.324) and others showing losses for screened funds relative

to comparative benchmarks (Schroder, 2003; r=-.515). Because the number of underlying

21

companies included in these funds was typically not reported, we could not conduct significance

testing nor calculate weighted means.

Summary Patterns

A formal meta-analysis like ours is preferred to a simple count of the positive, negative

and non-significant effects (Hunter & Schmidt, 1990). Nevertheless, when taken together, it is

interesting to observe that across all of the effects we coded from these studies, 58% are a non-

significant relationship, 27% a positive relationship, and 2% a negative relationship between

CSP and CFP. An additional 13% did not report sample size, so it was not possible to test for

significance. The meta-analytic results and the results of the vote counting procedure corroborate

each other in this instance—companies do not appear to suffer financially for their socially

responsible investments.

Critics of meta-analysis have argued that biases in the publication criteria of editors are

reflected in biased samples of studies used by meta-analytic researchers. In particular,

statistically significant results are more likely to be published (see Rosenthal, 1991). We address

this “file drawer problem” in two different ways. First, we gathered unpublished manuscripts

through informal efforts, as described above, in order to access studies from researchers’ file

drawers. Second, we computed a sensitivity analysis to measure just how many items must

languish in file drawers before the results of this meta-analysis would be affected. Using

Rosenthal’s (1991) formulas, we found that it would take at least 15,767 studies with an average

effect size of zero for the CSP-CFP association to no longer be statistically significant at p=.05.

That is, it would require over 82 times more null effects than we have here to render the current

results non-significant.

The sensitivity analysis gives us some confidence in these results. However, the absolute

size of this overall CSP-CFP effect is considered to be “small”: “small” effects are defined as

those around r=.10; effects are considered to be “medium” if they are about r=.30 and “large” if

they are greater than r=.50 (Rosenthal & Rosnow, 1991: 446). We did a second sensitivity

analysis to address the number of studies that would need to languish in file drawers to bring the

average effect up to the “medium” level. It would take at least 321 additional studies with a

medium-to-large average effect size of r=.40 for the overall CSP-CFP association to reach the

criterion for a medium effect size of r=.30. The 321 medium-to-large effects needed to boost the

present small effect to just barely medium-sized would amount to 1.67 times more effects than

22

the total body of 192 currently included. Using a more liberal criterion, it would take at least 130

additional effects with r=.40 to reach a moderate effect size of r=.24, or 68% more studies than

the current pool of available research from the last 35 years. To underscore just how unlikely this

might be, only 18 of 192 effects (9.4%) in our current dataset reach r=.40 or above.

DISCUSSION

After thirty-five years of research, the preponderance of evidence indicates a mildly

positive relationship between corporate social performance and corporate financial performance.

The overall average effect of r=.132 across all studies is statistically significant but, on an

absolute basis, it is small (Rosenthal & Rosnow, 1991), particularly considering the weighted

average of r=.101 and the median value of r=.082. These meta-analytic results lead to four broad

sets of implications.

Financial Impact of CSP

Companies do not seem to be richly rewarded for engaging in CSP. Friedman’s (1970)

concern about theft, however, may be misplaced: companies are not overtly penalized for CSP

investments. Penalties only accrue to firms that do wrong and perhaps only if they are caught. In

sum, the financial implications of CSP can be best understood as an interrelated set of three

findings.

First and most clear, revealed corporate misdeeds are costly to companies. Our analysis

of the financial impact of wrongdoing in 16 studies echoes Frooman’s (1997) earlier meta-

analytic result. Although the anecdotal evidence about recent scandals highlights just how grave

the consequences can be for companies and their executives who are caught doing wrong, it is

very difficult to estimate the likelihood that these kinds of misdeeds will be unearthed

(Schnatterly, 2003). Dubious firms may risk these sanctions because crime just might pay.

Second, on the other side of the ledger—doing good—our findings indicate that CSP

does not systematically destroy shareholder value. The overall effect of CSP on CFP is positive.

Only 2% of the individual studies reported a significant negative effect. Across our analyses by

CSP type, the average effects were nearly always positive and the occasional negative values

were negligible. There may well be less affirmative support for CSP’s positive financial impact

than there is for the negative financial impact of doing wrong, but managers who dedicate

corporate resources to social performance do not seem to be imposing a direct cost on their

shareholders. Companies can do good and do well, even if companies do not always do well by

23

doing good. This result provides some legitimacy for CSP when high-status public figures, such

as Kofi Annan (2001), so publicly call for CSP investments (Walsh, 2005).

Third, our findings suggest that CFP would seem to be an unlikely rationale or

justification for pursuing CSP. The small overall relationship between prior CSP and subsequent

CFP, the varied results across categories of CSP, and questions about causal direction all suggest

that more lucrative financial impact might attend investments other than CSP, providing better

returns on the next marginal dollar of corporate spending. Given these relatively low returns on

investment, unearthing alternative motivations for CSP warrants systematic inquiry, as we

suggest below.

Variation

CSP has come to encompass multiple dimensions, both in its theoretical specification and

in its empirical operational definition. Our findings indicate that those dimensions bear different

relationships to CFP. Relative to the overall effect size, the association is stronger for charitable

contributions, revealed misdeeds, self-reported social performance, and observer perceptions.

The CSP-CFP relationship is weaker for corporate policies, transparency, third-party audits, and

screened mutual funds.

Stronger results. In the case of charitable contributions, firms that performed well—and

particularly those with strong accounting performance—tended in the future to donate more

money and create more philanthropic foundations. For revealed misdeeds, we found no evidence

that poorly performing companies are more likely to engage in disreputable behavior. However,

when disreputable behavior is revealed to the public, it results in current and future penalties in

financial performance, particularly by the stock market. In the case of self-reported social

performance and observer perceptions, the CSP-CFP relationship may reflect a vulnerability to

halo biases such that CSP assessments are consistent with financial performance (Brown &

Perry, 1994).

Weaker results. Socially responsible corporate policies appear to be somewhat more

likely for companies that enjoy past financial success, but the presence of those policies does not

predict current or future financial success. Transparency appears to be valued by the market, but

third-party audits and screened mutual funds reveal effects of small magnitude, in particular

when CSP is measured first.

24

Signal in the noise. This mixed set of effects reveals just how complex the reality of the

CSP-CFP relationship may be, and just how difficult it is to measure and assess that relationship.

The complex reality emerges when results are considered in terms of the causal mechanisms they

suggest. Effects showing a positive relationship linking prior charitable contributions, revealed

misdeeds, and transparency to subsequent CFP suggest that it is the appearance of CSP, rather

than its substantive impact on a company’s operations, that affects subsequent financial

performance. Yet the larger effect sizes for observer perceptions and self-reported social

performance, compared to those for third-party audits and mutual fund screens, indicate that

appearances can be deceiving. Reporters and observers alike may succumb to biases that

confound CSP with CFP. The impact of environmental performance on CFP may result from the

attractiveness of the company to shareholders, customers, and employees, but it may also be a

function of the substantive reduction in costs produced by environmental performance. Moving

in the other causal direction, effects showing a positive relationship between prior financial

performance and subsequent charitable contributions and corporate polices suggest that wealthier

firms have the slack resources to engage in these practices, or that they encounter greater

pressure to do so. So too the link between prior CFP and self-reported social performance and

third-party audits suggests that more prosperous firms do more—or perceive that they do more—

–of what third-party auditors are likely to monitor.

The variation in results across types and measures of CSP may itself be the most

important signal to emerge from the 35 years of research on the connection between CSP and

CFP. That variation tells us how complex the relationship might be to unravel, which carries

important implications for how future research might make progress in wrestling with that

complexity. We turn to those implications below.

Direction of Causality

We find relatively consistent evidence that the link is as strong, if not stronger, when CFP

predicts subsequent CSP than the reverse causality, particularly for those studies including all

three types of measurements, and for the areas of charitable contributions, observer perceptions,

and third-party audits. While these results reinforce findings from two prior meta-analyses

(Allouche & Laroche, 2005; Orlitzky, et al., 2003), these findings tend to get overlooked.

Motivation for studying the link may revolve around efforts to establish the positive financial

effects of CSP, but the evidence of an association should direct our attention equally to

25

understanding how CFP ultimately gives rise to CSP, and not just the reverse. Although accounts

exist of why CFP makes subsequent CSP possible—slack resources or opportunism (Preston &

O’Bannon, 1997)—little has been written about the mechanics of how companies with strong

CFP end up engaging in greater CSP. That too has implications for future research, which we

elaborate below.

Assessing CSP

Beyond the relationship between CSP and CFP, our meta-analysis permits assessment of

CSP along three dimensions: its legitimacy, its value, and its effectiveness. First, is CSP

legitimate—is it a legitimate activity for society’s economic institutions? Despite some

normative opposition to the use of corporate resources to advance social purposes, our results

indicate that no damage is done to the purported owners of those resources. This means that CSP

cannot be delegitimized on economic grounds. Our findings may stop short of offering economic

grounds for a heavy investment in social performance, but by revealing no systematic negative

effects on CFP, our findings do suggest that it is not economically illegitimate for companies to

engage in CSP. It would seem that on economic grounds, the positive findings of this meta-

analysis—however mild and attenuated those findings might be—support the legitimacy of CSP.

Second, is CSP of value for companies? Is it worth their effort and investment? While

CSP may not transgress economic duties, it is open to question whether or not valuable benefits

accrue to companies that engage in CSP. The mild effect sizes for CSP open the possibility that

other areas of corporate activity are likely to have larger effects on financial performance. For

example, Wieser’s (2005) meta-analysis of research and development found an average 29%

return on research and development, with a lower bound of 7%. Nonetheless, we suspect that

well performing firms ignore CSP at their peril. Failure to invest in CSP can leave a company

hampered. As just one example, consider Wal-Mart’s late awakening to CSP. It has generated

enough opponents to stall its efforts to buy a bank and launch a credit card business (Leonhardt,

2006). Post, et al. (2002) spoke at length about how a firm needs society’s license to operate.

Given the positive CFP→CSP link, it may be that wealthy companies risk their “operator’s

license” if they avoid such investments (Campbell, 2007).

Third, is CSP effective? It may be legitimate for companies to concern themselves with

CSP, and the returns to companies may make some level of commitment worthwhile, but are

corporate efforts in social performance effective in achieving benefits for society? Here,

26

unfortunately, research remains meager, and CSP-CFP studies say little, leaving perhaps the

most fundamental questions unexplored: for whom are corporate efforts to do good effective and

for what purposes are company efforts effective?

In all, CSP proponents and opponents alike will find evidence for both joy and concern in

the implications of the results reported here. For proponents, the positive relationship found

across most categories of studies, no matter its magnitude, provides an economic defense for

CSP. Even CSP mavens are excited by the financial implications of these results. In an article

entitled “Holy Grail Found: Absolute, Definite Proof CSR Pays Off,” Kelly (2004:5) took stock

of the earlier findings (Orlitzky et al., 2003) and declared that socially responsible investors can

cash in on this knowledge: “Knowing that responsible companies outperform, savvy investors

have a head start in locating future winners before the broad market does.” Yet the results are not

strong and, across the nine categories, they often recede even further when the proper controls

are put in place or when only the effect of CSP upon subsequent CFP is examined. Rather than a

salubrious convergence of doing good and doing well, our meta-analytic results may indicate that

CSP advances neither objective. A mild effect size may be the product of corporate efforts to do

just enough CSP to avoid running afoul of social critics but not enough CSP to incur significant

costs that would incite economic critics.

For opponents, the small effect sizes place CSP investments in a suspect light. The

stronger relationship between preceding CFP and subsequent CSP, when combined with the

weak CSP→CFP result, suggests that such investments might be a waste of free cash flow.

These monies might be put to other more productive uses or returned to the shareholders if no

other positive Net Present Value investments are available. On the other hand, it may be that in

this era of intense corporate scrutiny, CSP investments do provide some latitude for the firm to

pursue its wealth objectives.

Ironically, 167 studies later, managers may be exactly where they were in 1972: seeking

criteria to judge when CSP makes sense and guidance about how to advance both CSP and CFP,

if they are both worthy of pursuit but not entirely consistent. The continuing quest to substantiate

or repudiate a link between CSP and CFP may be of little value. While the quest is seductive, it

may be time to let this particular question rest. There may be other aspects of the CSP-CFP

relationship that are now more important to investigate.

FUTURE DIRECTIONS

27

If fundamental tensions persist and major questions linger, what are the implications for

subsequent research? One option would be to drop the topic altogether. If CSP contributes little

economically, then those caught in the zeitgeist of investor capitalism might argue that it is no

longer worthy of attention. However, the mere fact of CSP should puzzle these proponents (see

Esrock & Leichty (2000) and Maignan & Ralston (2002) for documentation of companies’ CSP

activities). If CSP has limited financial impact, what explains companies’ investments in it? In

addition, some scholars will challenge the premise that CSP has limited economic value (Barnett,

2007; Orlitzky et al., 2003). They might argue that if only the relationship were better specified

and operationally defined, then stronger results would emerge or understanding would develop

regarding contingencies moderating the relationship. Therefore, we see two paths forward. The

well-traveled one may yield some additional insight, but we fear it will exact a high opportunity

cost. After all, there are only so many people who investigate these questions. It may be time to

take the path less traveled. This other path invites researchers to examine broader questions,

prompted by our meta-analytic results, about the relationship between the corporation and

society. We consider each path in turn.

The Well-Worn Path of Refinement

Virtually all past reviews of the CSP-CFP relationship call for more and better research

into the CSP → CFP relationship (Margolis & Walsh, 2001:20-24). We will not make that call

but we acknowledge that other scholars may be drawn to continue the quest. Future efforts to

examine the link should endeavor to do it well. Anyone who hopes to publish the 168

th

study

must meet four criteria. First, their data about CSP should consist of behavioral measures, such

as quantifiable outputs or third-party audits, and the assessment process for those must be clear

and open to validation. We suggest that researchers find alternatives to the convenient yet

difficult to validate measures such as the Fortune ratings of admired companies and company

insiders’ self-reported impressions. Second, the study must control for at least industry, risk and

size, if not R&D spending and advertising expenditures (McWilliams & Siegel, 2000). Third,

researchers need to assess CSP and CFP at different time periods; the direction of causality must

be theoretically articulated and empirically assessed. In our data, only 34% of effects (66 of 192)

featured measures of CSP that temporally preceded measures of CFP, surprisingly low if the

aspiration has been to establish a sequential link. Fourth, the CSP→CFP causal mechanisms

need to be articulated and tested. Too many studies speculate about mechanisms that explain

28

results or end with a call to investigate them. It is time to study mechanisms systematically. CSP

investments might help to recruit a high quality workforce (Backhaus, Stone, & Heiner, 2002),

attract a unique customer base (Sen & Bhattacharya, 2001), or provide insurance against some

unforeseen crisis (Schnietz & Epstein, 2005). CSP might bear upon CFP in some other way as

well. Although it is important to test the conditions under which CSP is more likely to contribute

to CFP (Barnett, 2007; Mackey et al., 2007; Rowley & Berman, 2000), it is as essential to

document the causal chain of connection. No matter how well measured the constructs, research

must move beyond simply assessing the magnitude of the CSP-CFP relationship. Research must

now show how CSP comes to bear upon CFP.

With these minimum standards in place, research on the link between CSP and CFP

should improve. But toward what end? Some scholars (Orlitzky et al., 2003) see merit in further

studies, especially those that examine the conditions under which CSP will influence CFP

(Barnett, 2007; Rowley & Berman, 2000). In contrast, we wonder whether ongoing research

efforts might be better devoted to other questions. Another set of CSP-CFP studies is unlikely to

change the general trend reported here, as our sensitivity analyses indicate, let alone convince the

opponents of the value of CSP (Tetlock, 2000).

A New Path of Exploration

Perhaps Bragdon and Marlin’s (1972) hope to stop sparring with economics can be

realized these 35 years later. The core dilemma may no longer be how to pursue social good

when it is seen to come at the expense of doing well. Globalization has turned up the competitive

heat on firms, but it has also brought them face-to-face with human misery of all kinds.

Corporations face public pressure to redress far-reaching societal problems (Margolis & Walsh,

2003) while keeping pace with market pressures to produce competitive products and financial

pressures to reduce costs and improve returns. The contemporary challenge facing managers and

scholars alike is therefore to learn how companies can navigate in a world that demands a firm

do good and do well. Understanding how companies endeavor to do both, side-by-side, might

best command scholarly attention in the years ahead. If the fundamental objective is to

understand the coexistence of CSP and CFP, then three compelling questions deserve as much

attention as any effort to determine if and when CSP pays: (1) Why do firms pursue CSP? (2)

How do companies pursue CSP? (3) How do firms pursue CSP and CFP simultaneously?

29

Why do firms pursue CSP? Consistent with past meta-analyses, we find that companies

with superior CFP are more likely to engage in CSP (Allouche & Laroche, 2005; Orlitzky et al.,

2003). What motivates these companies to engage in CSP? By understanding why and how the

firms more likely to engage in CSP—those high in CFP—do so and what benefits arise (for them

and for their intended beneficiaries), researchers may generate insight into why and how

companies in general should engage in CSP.

Research could begin by examining what propels companies that do well to attempt to do

good, perhaps even comparing them to other successful companies that do not do as much. At

least four motivations seem plausible: risk mitigation, external expectations, generalized

reciprocity, and guilt. As firms get bigger or more prosperous, reputation risks are more costly.

CSP may be a means of reducing risk—a means of buying reputation insurance (Peloza, 2006).

Indeed, well-known companies often find themselves to be targeted by social activists (Spar &

LaMure, 2003). Second, as firms become more prosperous and thus more prominent, external

expectations of their generosity may escalate, leading to an increase in appeals and pressure to

give. The CFP→CSP results we found also suggest that society may not be inclined to turn to

struggling firms for help. A firm’s first order of business is to create a high quality good or

service and sell it at a fair and profitable price. Only successful firms may be asked for additional

social investments (Campbell, 2007). Third, executives within a financially successful firm may

initiate CSP due to a sense of reciprocity. Much as successful individuals begin to assume they

owe something to those around them, so too may those who run successful companies (Frank,

2007). A fourth motivation puts a harder edge on reciprocity. Guilt, rather than gracious

reciprocity, may also propel companies that do well to endeavor to do good. Guilt has been

shown to arise from distress over inequity—when people “benefit more than others do” and feel

unduly rewarded—and it motivates efforts to reduce those inequities (Baumeister, Stillwell, &

Heatherton, 1994: 260). Guilt about reaping rewards without compensating others equitably may

trigger managers at firms high on CFP to engage in CSP.

As managers’ motivations gain attention from researchers, there is bound to be variance.

Why do some companies in an industry lead (e.g., Target) and others follow (e.g., Wal-Mart)?

Some firms resist these pressures altogether. Why do companies in an industry shy away from

these investments (e.g., ExxonMobil), while others trumpet their investments so loudly (e.g., BP

and Shell)? Much remains to be learned.

30

How do firms pursue CSP? Beyond the motivations to pursue CSP, systematic

understanding of how companies pursue CSP is essential. A first vein of research would

investigate how companies establish a general commitment to CSP. How does CSP gain traction

within companies? What rationale within the company do managers use first to “sell”

(Andersson & Bateman, 2000; Bansal, 2003) and subsequently to explain (Sonenshein, 2006)

their involvement in CSP? A second vein of research would examine the specific commitments

that companies make. Descriptive research needs to be done to catalogue the sorts of activities

and initiatives in which companies engage, documenting the methods companies use to engage

in CSP and unearthing the factors that may account for variance in corporate activities and

practices. Esrock and Leichty’s (2000) and Maignan and Ralston’s (2002) look at firms’ self-

presentations on their web pages is just a start. This descriptive research effort would set up

theory-building research into how companies pursue their CSP investments, augmenting the

early work that points to geographical and network influences on CSP choices (Galaskiewicz,

1997; Marquis, Glynn, & Davis, 2007).

Doing good and doing well. The mechanics of how companies engage in CSP implicate

important and often overlooked managerial questions. Thirty years ago, Merton (1976: 88)

wondered, “Does the successful business try first to profit or to serve?” It is a question, he

observed, that “must at one time or another plague every corporate executive.” The simple

answer “do both,” Merton recognized, “escapes the dilemma by swift flight from it,” begging the

question of “how to do both in appropriate scale.” The challenge for companies lies in doing well

and doing good (Margolis & Walsh, 2003; Paine, 2002; Porter & Kramer, 2006; Vogel, 2006),

whether it is finding ways for the two to converge or finding ways to manage the tensions, real or

only apparent, that managers experience in trying to do both.

It is essential for future research on CSP to investigate how organizations and managers

do both. What are the structures, strategies, processes, and practices that companies and the

individuals within them use that enable them to do both? Akin to research on ambidexterity

(Tushman & O’Reilly, 1997), which explores how companies pursue multiple and sometimes

competing objectives, future research can identify the organizational conditions and practices

that prove most effective for facilitating coexistence of efforts to do good and do well, and which

organizational attributes impede those endeavors. What matters are the organizational practices

that advance the impact of CSP investments not only upon the company itself (Bartel, 2001;

31

Porter & Kramer, 2006) but also upon the intended beneficiaries of those investments (Margolis

& Walsh, 2003). Research that investigates how to do good and do well can accomplish so much

more than simply extending the 35-year quest for a link between the two.

LIMITATIONS

Although we have tried to provide the most comprehensive review to date of empirical

research reporting on the relationship between CSP and CFP, several factors limit the

conclusions that we can draw. First, the meta-analysis is limited to the collection of studies that

are available for inclusion. That is, we can only examine what is and not what should be. Our

findings are qualified by all of the same limitations of the underlying empirical work that it

incorporates. We would have welcomed research with better measures, more control variables,

and a greater sensitivity to temporal sequencing. They just do not exist. We are also limited by

the possibility that our collection of studies excludes unpublished work, although our efforts to

obtain such work and the results of our sensitivity analysis mitigate this concern. Second, as we

discussed above, some studies had to be excluded from our analysis because it was not possible

to summarize their results in terms of a consistent effect size r. Third, we were prevented from

including significance tests on the influence of measurement timing or type of financial variable,

given the inconsistent overlapping nature of these variables. That is, in the case of financial

variables, some studies included only market measures, some included only accounting

measures, and some included both types. This made straightforward comparisons non-

independent and thus statistically problematic.

A further limitation of our study is also related to the statistical independence of data. We

were unable to include a number of advanced meta-analytic techniques—such as controlling for

unreliability in the effect size estimates—because many studies sampled from the same

underlying pool of companies. For example, large U.S. firms such as Johnson & Johnson or IBM

may have been included within dozens of our studies. Many studies used exactly the same data

sets, such as the Fortune 500 or Domini 400 firms. Some previous meta-analytic reviews of this

research area have even exacerbated this problem by including multiple effect sizes within a

single article as if they were separate studies, for example counting a study with five years of

data and three types of accounting measures as if it were 15 studies (Orlitzky et al., 2003; Wu,

2006). We did this only in the case of studies reporting distinct types of CSP, in which case we

reported studies as if they contained two effects or, at most, four (i.e., Griffin & Mahon (1997)

32

included charitable contributions, environmental performance, observers’ perceptions, and third-

party audits). Strictly speaking, meta-analysis assumes that each study represents an independent

sample (Judge, Thoresen, Bono, & Patton, 2001; Rosenthal, 1991) and, as such, the results of the

present paper—as well as other reviews of this topic—should be considered approximate and

descriptive rather than precise statistical tests. Fortunately, although violations of statistical

independence make significance tests highly suspect, the effect sizes themselves remain unbiased

estimates of the true CSP-CFP relationship (Rosenthal, 1991). The main goal of the present study

was to examine the absolute size of estimated effects rather than dwell on their relative

significance levels.

CONCLUSION

The sustained pursuit of a link between CSP and CFP may well reflect a deeper and

intensifying quest for meaning. That quest for meaning, Robert Merton (1976) observed,

becomes particularly fierce as societies achieve material security and organizations are asked to

deliver more than material welfare. Merton (1976: 88) suggested that members of society begin

to ask “affluence for what? and for whom, and what beyond affluence?” and business leaders

feel the repercussions:

The leaders of business in the morally more sensitive society of our time are coming to

recognize that they must pay the price of a growing commitment to the moral purposes of

the larger society. Acting in terms of an authentic moral commitment is not cost-free

(Merton, 1976: 86).

The steady flow of studies of the CSP-CFP relationship, even while sustaining a bridge to the

economic logic that has come to dominate the study of organizations, may also constitute a

response to two symptoms of the quest for meaning that Merton identified: the practical

symptom of business leaders’ need to justify the cost of “an authentic moral commitment” and

the scholarly symptom of researchers’ thirst for the deeper purpose that business serves for

society. Much as research on the financial impact of CSP may have been a harbinger of broader

efforts to identify the financial impact of other organizational practices, so too an underlying

quest for purpose and meaning beyond economic profitability—a quest that is now finding its

way into organizational scholarship broadly considered (Sandelands, 2003) and in such specific

areas as leadership (Podolny, Khurana, & Hill-Popper, 2005)—may have manifested itself first

in research on CSP. Efforts to identify the impact of CSP on CFP are, at least in part, efforts to

33

legitimize CSP and thereby create space for broader purposes in business activity—to establish

that business can be about doing good, not just doing well.

Whatever accounts for vigorous interest in the connection between CSP and CFP, the

justification of CSP rests on a range of considerations beyond CFP. The contribution any

corporate practice makes to economic welfare cannot alone justify that practice. Principles of

justice indicate that advancing economic welfare cannot justify the suspension or violation of

other rights and duties (Rawls, 1974), which have as strong a moral claim upon corporate

conduct as does the pursuit of its financial objectives. Ultimately, the merits of CSP, even merits

that transcend the link to financial performance, must be weighed.

The impact that organizations have on our lives, along with the meaningful purposes that

people (employees, customers, citizens, and investors) seek to pursue through them, implicates a

much larger question confronting organizational scholars. How do we live with organizations

that shape the distribution of costs and benefits, advantages and burdens within society? How do

we live with organizations that infuse our lives with meaning, or fail to? These kinds of

compelling questions might orient (some say must orient) future research on organizations

(Walsh, Meyer, & Schoonhoven, 2006). Demands for organizations with which we can live—

organizations that do well and do good—call not for the facile dismissal of companies’ economic

function. Rather, they call for careful inquiry into what companies do and can do to manage

these multiple demands. The demands and the challenge of meeting them will not recede with a

simple correlation between CSP and CFP, no matter its magnitude.

34

FIGURE 1

CSP-CFP Studies 1972-2007

1

5 5

4

5

3

5

7

8

9

8

19

28

15

14

6

16

9

0

5

10

15

20

25

30

1972-

73

1974-

75

1976-

77

1978-

79

1980-

81

1982-

83

1984-

85

1986-

87

1988-

89

1990-

91

1992-

93

1994-

95

1996-

97

1998-

99

2000-

01

2002-

03

2004-

05

2006-

07

Numb er of Art icles

35

TABLE 1

Prior Reviews of the CSP-CFP Relationship

Authors (Year)

Number of

Articles

Reviewed

Number of

Times

Cited by

Others*

Aldag and Bartol (1978) 10 14

Arlow and Gannon (1982) 7 57

Cochran and Wood (1984) 14 146

Aupperle, Carroll, and Hatfield (1985) 10 205

Wokutch and McKinney (1991) 20 13

Wood and Jones (1995) 34 150

Pava and Krausz (1996) 21 98

Griffin and Mahon (1997) 51 232

Preston and O'Bannon (1997) 8 66

Richardson, Welker, and Hutchinson (1999) 14 8

Roman, Hayibor, and Agle (1999) 46 82

Margolis and Walsh (2001) 95 96

Margolis and Walsh (2003) 127 134

Orlitzky, Schmidt, and Rynes (2003) 52 155

Allouche and Laroche (2005) 82 0

Wu (2006)

39

2

*Citation counts assessed using Google Scholar on July 26, 2007

36

TABLE 2

Attributes of 192 Effects from 167 Studies Included in Meta-Analyses of Association Between Corporate Social Performance and Corporate Financial Performance

Effect Size

Timing of CSP

measure Type of CFP

Study

# Study

N

Firms CSP Type CSP Measure CFP Measure

Sample of Companies and

Other Notes

Years

for CSP

Years

for CFP Overall

CFP -->

CSP Concurrent

CSP -->

CFP Accounting Market

Event

Study

Control

Variables

1 Alexander and

Buchholz

(1978)

40 Observer

perceptions

Surveys of students and

business people (data

from Vance, 1975)

Risk-adjusted

stock returns

Large national corporations

(Fortune 100 equivalent)

1971-2 1970-4

.078 .078 .078 No Risk, Size

2 Anderson and

Frankle (1980)

314 Transparency Beresford's Social

Involvement Disclosure

scale: Content analysis of

annual reports, i.e., self-

reported disclosures,

dichotomized into those

who did vs. did not make

disclosures in 1972

Stock returns Fortune 500 companies listed

on the NYSE with financial

data available for fiscal years

ending 12/31

1972 1972-3

.297 .297 .297 No Risk, Size

3 Asmundson and

Foerster (2001)

N/A Screened

mutual funds

Varies depending on the

fund

Fund returns Socially responsible equity

mutual funds of Canadian

equities with 5-year (4 total)

and 10-year (2 total) histories

1990-1999 1990-1999 .038 .038 .038 No Risk

4 Aupperle et al.

(1985)

192 Self-report Forced-choice self-report

survey based on Carroll's

(1979) concept of

concern for society

Risk-adjusted

ROA

Large firms in the Forbes 1981

directory whose CEOs

responded to self-report survey

1982 1977-82

.120 .120 .120 No Risk, Size

5 Aupperle et al.

(1985)

171 Corporate

policies

Use of social forecasting,

having a CSR committee

on the corporate board

Risk-adjusted

ROA

Large firms in the Forbes 1981

directory whose CEOs

responded to self-report survey

1982 1977-82

.044 .044 .044 No Risk, Size

6 Bagozzi,

Epstein and

Wisner (2001)

215 Environment

(self-report)

Survey to rate own

company relative to

others in the industry on

compliance with

environmental

regulations, limiting

environmental impact

beyond compliance, and

preventing and mitigating

environmental crises

(data from Judge &

Douglas, 1998)

Self-reported

ROI and earnings

growth relative

to others in

industry

Questionnaire sent to 725 U.S.

firms from the 1992 World

Environmental Directory,

randomly chosen among those

with a corporate environmental

officer listed. 30% response

rate, primarily from large firms

Survey

around

1992-1995

Survey

around

1992-1995

.244 . 244 .244 No Size,

Industry

7 Barnett &

Salomon (2006)

N/A Screened

mutual funds

Varies by fund Fund returns 67 Funds tracked by the Social

Investment Forum

1972-2000 1972-2000 -.009 -.009 -.009 No Risk

8 Barth and

McNichols

(1995)

850 Environment

(objective)

Estimates of firm cost

due to liabilities of

Superfund sites

Stock returns Firms with Superfund site data

from EPA filings and

Environmental Data Resources

Inc.

1982-1991 1982-1991 .091 .091 .091 No Size

9 Bauer, Derwall

and Otten

(2007)

88-

130

Screened

mutual funds

Varies by fund Fund returns Canadian ethical and

conventional mutual funds of

Canadian equities

1/94-1/03 1/94-1/03 -.025 -.025 -.025 No Risk, Size

37

Attributes of 192 Effects from 167 Studies Included in Meta-Analyses of Association Between Corporate Social Performance and Corporate Financial Performance

Effect Size

Timing of CSP

measure Type of CFP

Study

# Study

N

Firms CSP Type CSP Measure CFP Measure

Sample of Companies and

Other Notes

Years

for CSP

Years

for CFP Overall

CFP -->

CSP Concurrent

CSP -->

CFP Accounting Market

Event

Study

Control

Variables

10 Bauer, Koedijk

and Otten

(2005)

N/A Screened

mutual funds

Varies by fund Fund returns 103 ethical mutual equity funds

in U.S., U.K. and Germany,

each matched to 3 conventional

funds by fund age and size.

Effect "not significant",

presumed zero.

1/90-3/01 1/90-3/01 .000 .000 .000 No Risk, Size

11 Belkaoui (1976)

100 Transparency Whether pollution

expenditures were

disclosed in the Annual

Report

Risk-adjusted

stock returns

50 firms voluntarily reporting

pollution control information in

1970 annual reports, with 50

matched firms not reporting

1970 1969-71,

month

-12 to -1

-.044 -.095 .007 -.044 No Size,

Industry,

Risk

12 Belkaoui and

Karpik (1989)

23 Transparency Reputational scale

(among business people)

on social performance

from Business and

Society Review's

"Industry Rates Itself",

1972

ROA, stock

returns

Leading corporations surveyed

both by Business and Society

Review's 1972 "Industry Rates

Itself" and Ernst and Ernst

1973 survey of social

responsibility disclosure

1972

report

release

ROA 1973,

Returns

%1970-

1974

.168 -.104 .417 .417 -.104 No Size

13 Belkaoui and

Karpik (1989)

23 Observer

perceptions

Reputational scale

(among business people)

on social performance

from Business and

Society Review's

"Industry Rates Itself",

1972

ROA, stock

returns

Leading corporations surveyed

both by Business and Society

Review's 1972 "Industry Rates

Itself" and Ernst and Ernst

1973 survey of social

responsibility disclosure

1972

report

release

ROA 1973,

Returns

%1970-

1974

.072 -.225 .356 .356 -.225 No Size

14 Bello (2005) N/A Screened

mutual funds

Varies by fund Fund returns Socially repsonsible domestic

equity funds in Morningstar

database, each matched with

two conventional funds by net

asset size

1/1994 -

3/2001

1/1994 -

3/2001

.060 . 060 .060 No Risk, Size

15 Berman, Wicks,

Kotha, Jones

(1999)

81 Third-party

audit

KLD ratings in 5

categories: employee

relations, diversity, local

communities,

environment, product

safety/quality

ROA Top 100 firms on the 1996

Fortune 500 list that were

publicly traded 1991-1996

1991-1996 1991-1996 .089 .089 .089 No Size

16 Blacconiere and

Northcut (1997)

72 Environment

(self-report)

Annual Report

disclosures related to five

aspects of environmental

performance

Stock returns Chemical firms (defined by 2-

digit SIC code) with available

EPA Superfund data (liability

related to hazardous waste

sites)

1983-4

Annual

Reports

Events

1985 to

1986

day -1 to

+1

.189 .189 .189 Yes Company

is its own

control

17 Blacconiere and

Northcut (1997)

72 Environment

(objective)

Estimate of firm cost

liabilities for Superfund

sites

Stock returns Chemical firms (defined by 2-

digit SIC code) with available

EPA Superfund data (liability

related to hazardous waste

sites)

1983-4

EPA

reports

Events

1985 to

1986

day -1 to

+1

.260 .260 .260 Yes Company

is its own

control

38

Attributes of 192 Effects from 167 Studies Included in Meta-Analyses of Association Between Corporate Social Performance and Corporate Financial Performance

Effect Size

Timing of CSP

measure Type of CFP

Study

# Study

N

Firms CSP Type CSP Measure CFP Measure

Sample of Companies and

Other Notes

Years

for CSP

Years

for CFP Overall

CFP -->

CSP Concurrent

CSP -->

CFP Accounting Market

Event

Study

Control

Variables

18 Blacconiere and

Patten (1994)

47 Environment

(self-report)

Annual Report

disclosures related to five

aspects of environmental

performance

Stock returns Impact after Union Carbide

Bhopal accident on similar

companies (NYSE/ASE firms

with one of the same 4-digit

chemical SICs as Union

Carbide and at least 10% of

their revenues in chemical and

industrial gases)

firm 10K

report

prior to

day 0

(12/3/84)

to +4

.233 .233 .233 Yes Company

is its own

control

19 Blackburn el al.

(1994)

88 Third-party

audit

Council of Economic

Priorities ratings of

consumer product firms

ROA, stock

returns

Consumer product firms

examined by the Council on

Economic Priorities (1989)

1989 1989

-.014 -.014 -.045 .047 No

20 Bollen and

Cohen (2004)

416 Screened

mutual funds

Varies by fund Fund returns Equity mutual funds in the

CRSP U.S. Mutual Fund

Database with at least 2 years

of data, identified by a list from

the Social Investment Forum

1961-2002 1961-2002 .021 .021 .021 No Risk, Size

21 Bosch and Lee

(1994)

<=

121

Revealed

misdeeds

FDA disciplinary actions:

recalls of existing

products, bans of

products, warnings,

investigations, product

withdrawals, allegations

of false advertising

Stock returns Food and drug firms with FDA

decisions announced in the

Wall Street Journal

1962-1989 day -5 to

+5

.141 .141 .141 Yes Company

is its own

control

22 Bowman (1978)

46 Transparency Mention of CSP in

Annual Reports

ROS Computer industry, firms doing

business in only one SIC code

1974 1972-4

.190 .190 .190 No Industry

23 Bowman and

Haire (1975)

82 Transparency Content analysis of

annual reports

ROE 1973 Moody's list of firms in

food processing industry

1973 1969-73

.351 .351 .351 No Industry,

Size

24 Boyle, Higgins

and Rhee

(1997)

61 Corporate

policies

Defense contractor

signing of Packard

Commission agreement

to define a code of ethics

for the defense industry

to exclude defective

pricing, kickbacks, false

claims

Stock returns Top 100 defense contracting

organizations (in terms of

contract dollars awarded).

Contractors not publicly traded

or with confounding news

announcements dropped from

sample.

Companies listed in Appendix

7/3/1986 day -2 to

+2

-.041 .178 -.116 -.184 -.041 Yes Industry,

Risk

25 Bragdon and

Marlin (1972)

17 Environment

(objective)

Pollution control indices EPS growth,

ROE, ROC

Pulp and paper companies

studied by CEP

1970 1965-70

.211 .217 .203 .211 No Industry

26 Brammer and

Millington

(2004)

315 Charitable

donations

Corporate charitable

contributions

ROS Publicly listed UK

manufacturing and service

companies (finance companies

excluded)

1989-

1990,

1998-1999

1989-1990,

1998-1999

.485 . 485 .485 Yes Industry,

Size

27 Brammer,

Brooks, &

Pavelin (2006)

451 Environment

(objective)

Ethical Investment

Research Service (EIRIS)

audit of environmental

performance

Stock returns Firms appearing in the UK

FTSE All-share Index

weighted index for which data

were available, excluding

investment trusts

2002 2002-2005 -.076 -.076 -.076 No

39

Attributes of 192 Effects from 167 Studies Included in Meta-Analyses of Association Between Corporate Social Performance and Corporate Financial Performance

Effect Size

Timing of CSP

measure Type of CFP

Study

# Study

N

Firms CSP Type CSP Measure CFP Measure

Sample of Companies and

Other Notes

Years

for CSP

Years

for CFP Overall

CFP -->

CSP Concurrent

CSP -->

CFP Accounting Market

Event

Study

Control

Variables

28 Brammer,

Brooks, &

Pavelin (2006)

451 Third-party

audit

Ethical Investment

Research Service (EIRIS)

audit of environmental

performance

Stock returns Firms appearing in the UK

FTSE All-share Index

weighted index for which data

were available, excluding

investment trusts

2002 2002-2005 .002 .002 .002 No

29 Brown (1997) 108 Observer

perceptions

Fortune magazine ratings

of "responsibility to

community and

environment", controlling

for "halo effect" of past

financial performance;

comparison of top vs.

bottom quartile of ratings

Stock returns Fortune annual survey. Survey

data adjusted by Brown and

Perry (1994)

1982-1991

average

1982-1992

pooled

.046 .046 .046 No Size

30 Brown (1998) > 149 Observer

perceptions

Fortune magazine ratings

of "responsibility to

community and

environment", controlling

for "halo effect" of past

financial performance;

comparison of top vs.

bottom quartile of ratings

Stock returns Fortune annual survey for >4

years from 1982-1991. Survey

data adjusted by Brown and

Perry (1994)

firm

average

1982-1991

1984-1996

pooled

.054 .054 .054 No Size

31 Brown and

Perry (1994)

234 Observer

perceptions

Fortune magazine ratings

of "responsibility to

community and

environment", controlling

for "halo effect" of past

financial performance;

comparison of top vs.

bottom quartile of ratings

ROA, ratio of

market / book

value

Firms rated by Fortune 1991 1988-91

.440 .440 .420 . 460 No

32 Brown,

Helland, &

Smith (2006)

701 Charitable

donations

Whether a firm has a

charitable foundation

Net income Large firms listed in the

Corporate Giving Directory

1999 1999 .125 .125 .125 No Industry,

Size

33 Buehler and

Shetty (1976)

232 Self-report Questionnaire reported of

a structural change in

CSR, defined as change

in corporate policy

(mission statement) or

organization (i.e. develop

group within firm to

manage CSR) in terms of

urban affairs (urban

renewal, hiring of

minorities, etc.),

consumer affairs

(warranties, labeling,

quality, safety, etc.), or

the environment.

ROA Fortune 1250 largest US firms

(19% response rate); Note

authors list results as "not

significant" so coded as zero.

N/L,

presumabl

y late

1960s-

1972

%1967-

1972

.000 . 000 .000 No

34 Chauvin and

Guthrie (1994)

62 Third-party

audit

Announcement that firm

is on the annual list of

'Best Companies for

Working Mothers'

Stock returns Publicly traded firms on

Working Mother magazine's

list of 'best' employers

1986-1991 day -2 to

+2

.317 .317 .317 Yes Company

is its own

control

40

Attributes of 192 Effects from 167 Studies Included in Meta-Analyses of Association Between Corporate Social Performance and Corporate Financial Performance

Effect Size

Timing of CSP

measure Type of CFP

Study

# Study

N

Firms CSP Type CSP Measure CFP Measure

Sample of Companies and

Other Notes

Years

for CSP

Years

for CFP Overall

CFP -->

CSP Concurrent

CSP -->

CFP Accounting Market

Event

Study

Control

Variables

35 Chen and

Metcalf (1980)

18 Environment

(objective)

CEP measures of

environmental

performance

ROE, P/E ratio Pulp and paper industry firms

reviewed by CEP

1970 1968-73

.069 .158 -. 021 No Industry,

Size

36 Clarkson (1988)

32 Observer

perceptions

Evaluation by MBA

students of firm's Social

Orientation, categorized

into 4 groups: reactive,

defensive,

accommodative,

proactive

Observer-rated

"economic

performance"

relative to

industry during

last 5 years: loss,

below average,

average or above

average

Large Canadian firms 1983-1985 1983-1985

.874 .874 No Size,

industry

37 Clemens (2006) 76 Environment

(self-report)

Survey completed by

high-ranking executive

about environmental

performance,

investments, and

consciousness

Self-reported

profitability,

growth, and

ROA

Small private firms from a

2003 survey of steel industry

scrap yards (46% response

rate)

2003-2004 2003-2004 .400 .400 No Industry,

Size

38 Cochran and

Wood (1984)

39 Observer

perceptions

Company ratings of

social responsiveness

made by business

journalist Moskowitz

(1972)

Earnings/assets,

earnings/sales,

excess market

valuation

US industrial firms Composite

from 1972-

5

1970-4,

1975-9

.303 .296 .310 .181 .517 No Industry

39 Conine and

Madden (1986)

163 Observer

perceptions

Fortune magazine ratings

of "responsibility to

community and

environment"

Observer-rated

"value as a long-

term investment"

10 largest companies in each

industry included in Fortune

ratings

1983-5

(but

correlation

s are

within

year)

1983-5 (but

correlations

are within

year)

.715 .715 .715 No Size,

industry

40 Cowen et al.

(1987)

95 Transparency Number of Annual

Report disclosures,

assessed by Ernst &

Whinney survey

ROE Fortune 500 companies from

ten industries

1978 1976-78

-.060 -.060 -.060 No Industry,

Size

41 D'Antonio,

Johnsen and

Hutton (1997)

140 Screened

mutual funds

Domini 400 Social Stock

Index firms compared to

firms in Lehman Brothers

Corporate Bond Index

Difference in

bond returns

Domini 400 Social Stock Index

firms with bond issues listed in

the University of Wisconsin's

Fixed Income Database

5/90-3/96 5/90-3/96

.277 .277 No

42 Davidson and

Worrell (1988)

96 Revealed

misdeeds

Event of 5 types of

corporate crimes: bribery,

criminal fraud, tax

evasion, illegal political

contributions or antritrust

violations of price-fixing

and bid-rigging. Event

day is the Wall Street

Journal announcement

Stock returns Fortune's list of 800 largest

corporations

1970-1988 day -5 to

+5

.028 .028 .028 Yes Company

is its own

control

43 Davidson and

Worrell (1992)

31 Revealed

misdeeds

Comparison of impact for

product recalls that were

government-mandated vs.

voluntary

Stock returns Firms in non-automotive

industries making product

recall announcements in the

Wall Street Journal

-90 to 90

days,

between

1968-1987

-90 to 90

days

.406 .406 .406 Yes Company

is its own

control

41

Attributes of 192 Effects from 167 Studies Included in Meta-Analyses of Association Between Corporate Social Performance and Corporate Financial Performance

Effect Size

Timing of CSP

measure Type of CFP

Study

# Study

N

Firms CSP Type CSP Measure CFP Measure

Sample of Companies and

Other Notes

Years

for CSP

Years

for CFP Overall

CFP -->

CSP Concurrent

CSP -->

CFP Accounting Market

Event

Study

Control

Variables

44 Davidson,

Worrell and

Cheng (1994)

<= 47 Revealed

misdeeds

Announcement of OSHA

(Occupational Safety and

Health Administration)

penalty

Stock returns Firms with reported OSHA

violations in Wall Street

Journal, New York Times,

Washington Post and Los

Angeles Times

1979-1989 day -1 to

+1

.194 .194 .194 Yes Company

is its own

control

45 Derwall and

Koedijk (2005)

306 Screened

mutual funds

Varies by fund Bond returns Eight U.S. bond mutual funds

labelled by the Social

Investment Forum as being

socially responsible. Each fund

is matched to 5 conventional

funds in the CRSP Mutual

Fund database.

9/87-3/03 9/87-3/03 .054 .054 No Risk

46 Derwall,

Gunster, Bauer

and Koedijk

(2005)

62 Environment

(objective)

Environmental ratings by

Innovest Strategic Value

Advisors

Stock returns U.S. companies in Innovest

database (database rates 1200

firms globally)

7/95-12/03 7/95-12/03 .217 .217 .217 No Industry,

Risk, Size

47 DiBartolomeo

and Kurtz

(1999)

650 Screened

mutual funds

Domini 400 Social stock

index

Stock returns Domini 400 Social stock index

compared to S&P 500

5/90-1/99 5/90-1/99

.000 .000 .000 No Industry,

Risk

48 Diltz (1995) 159 Environment

(objective)

Ratings by the CEP

regarding environmental

performance; portfolio

pair constructed of good

vs. poor ratings

Stock returns All firms evaluated by Council

on Economic Priorities in

1991.

1991

report

release

1989-1991

.258 .258 .258 No Size, risk

49 Diltz (1995) 159 Charitable

donations

Ratings by the CEP

regarding charitable

giving; portfolio pair

constructed of good vs.

poor ratings

Stock returns All firms evaluated by Council

on Economic Priorities in

1991.

1991

report

release

1989-1991

.022 .022 .022 No Size, risk

50 Diltz (1995) 159 Third-party

audit

Ratings by the CEP in 9

categories: status of

women in management,

status of minority

members in management,

animal testing,

information disclosure,

community outreach,

South Africa, family

benefits, military, nuclear

safety; 9 portfolio pairs

constructed of good vs.

poor ratings

Stock returns All firms evaluated by Council

on Economic Priorities in

1991.

1991

report

release

1989-1991

-.003 -.003 -.003 No Size, risk

51 Dooley and

Lerner (1994)

86 Self-report CEO self-reported

orientation towards

employees, community,

and government

ROA Fortune 500 industrial and

service firms for which EPA

pollution data was available

and CEO completed survey

1989 N/L

.124 .124 No Industry,

Size

52 Dooley and

Lerner (1994)

86 Environment

(objective)

EPA data on total toxic

pollution release

ROA Fortune 500 industrial and

service firms for which EPA

pollution data was available

and CEO completed survey

1989 N/L

-.330 -.330 No Industry,

Size

42

Attributes of 192 Effects from 167 Studies Included in Meta-Analyses of Association Between Corporate Social Performance and Corporate Financial Performance

Effect Size

Timing of CSP

measure Type of CFP

Study

# Study

N

Firms CSP Type CSP Measure CFP Measure

Sample of Companies and

Other Notes

Years

for CSP

Years

for CFP Overall

CFP -->

CSP Concurrent

CSP -->

CFP Accounting Market

Event

Study

Control

Variables

53 Dowell, Hart

and Yeung

(2000)

89 Environment

(self-report)

Investor Responsibility

Research Center survey

of multinational

environmental policy into

3 categories: local (the

corporation adheres to

local standards only),

U.S. (applies U.S.

standards wherever it

does business), or

stringent global (firm's

internal standards higher

than any national

standard)

Tobin's q -

market value

over replacement

cost of tangible

assets

S&P 500 multinational

corporations - with

manufacturing or mining SIC

codes, and operations in

countries with GDP per capita

below $8000

1994-1997

panel data

1994-1997

panel data

.062 . 062 .062 No Industry,

Size

54 Fogler and Nutt

(1975)

9 Objective

environment

P/E ratio, Stock

returns

9 of the 12 pulp and paper

companies studied by Bradgon

& Marlin (1972)

1970 (year

preceding

1971 CEP

report)

P/E ratios

1971-2,

stock

returns 0

days before

vs. 10 days

after 1971

article

release

-.053 -.053 .000 -.105 No Industry

55 Fombrun and

Shanley (1990)

154 Charitable

donations

Data on charitable giving

and existence of a

foundation

ROIC, market-

to-book value

ratio, dividend

yield

Companies from Fortune's

1985 corporate reputation

survey with data available on

charitable giving and other

factors (154 of the 292 firms)

1984 1984-5

.055 .055 .085 .010 No Industry,

Size

56 Freedman and

Jaggi (1982)

109 Transparency Extent of environmental

self-report disclosures in

Annual Report,

regardless of valence

ROA and ROE, 2

operating ratios

Firms from high polluting

industries, excluding utilities -

chemical, oil refining, steel,

paper and pulp

1973 1973-4

-.194 -.194 -.194 No Industry

57 Freedman and

Jaggi (1986)

56 Transparency Extent of environmental

self-report disclosures in

Annual Report,

dichotomized into low

(not zero) vs. high

disclosure

Stock returns Firms from high polluting

industries, excluding utilities -

chemical, oil refining, steel,

paper and pulp

Annual

report

1973-1974

Same

annual

report

.023 .023 .023 No Industry

43

Attributes of 192 Effects from 167 Studies Included in Meta-Analyses of Association Between Corporate Social Performance and Corporate Financial Performance

Effect Size

Timing of CSP

measure Type of CFP

Study

# Study

N

Firms CSP Type CSP Measure CFP Measure

Sample of Companies and

Other Notes

Years

for CSP

Years

for CFP Overall

CFP -->

CSP Concurrent

CSP -->

CFP Accounting Market

Event

Study

Control

Variables

58 Freedman and

Stagliano

(1991)

27 Transparency Annual Report disclosure

in firm's 10K report

regarding potential costs

of stricter OSHA

standards for safety and

health of work

environment into 4

categories: disclosed that

standard would have little

impact on operations,

disclosed only non-

quantitative data about

effects, disclosed

quantitative data about

estimated costs, and no

disclosure

Stock returns Firms in Cotton Textile Mill

(SIC 2200) and Knitting Mill

(SIC 2250) industries

Annual

reports

presumabl

y around

year of

article

day 0 to +3

.129 .129 .129 Yes Company

is its own

control

59 Fry and Hock

(1976)

135 Transparency Analysis of Annual

Reports, where a

photograph relating to

CSR receives 0.5 points

and each paragraph 1

point, normalized by total

pages.

ROI Firms from 15 different

industry groups with sales from

$2M to over $1B. Results

listed as "not significant",

presumed zero.

Annual

reports

presumabl

y around

year of

article

Annual

reports

presumably

around

year of

article

.000 .000 .000 No

60 Galaskiewicz

(1997)

40 Charitable

donations

Corporate charitable

contributions

Firm's financial

performance

quartile ranking

based on ROS,

ROA, ROE

relative to

industry

Companies in Minneapolis-St

Paul metropolitan area with >=

200 employees

1979-

1981,

1987-1989

1979-1981,

1987-1989

.363 . 363 .363 No Industry,

Size

61 Galbreath

(2006)

38 Third-party

audit

Audit by Australian firm

Reputation Measurement

regarding employee

treatment and social

impact

ROA, ROE,

Mean value-

added

Australian top-100 companies

with financial data available

2000 2001 .014 .014 .056 -.069 No Size

62 Galbreath

(2006)

38 Environmental

(objective)

Audit by Australian firm

Reputation Measurement

regarding environmental

performance

ROA, ROE,

Mean value-

added

Australian top-100 companies

with financial data available

2000 2001 -.052 -.052 .076 -.310 No Size

63 Goll and

Rasheed (2004)

171 Self-report Response to three survey

questions adapted from

Aupperle (1984),

regarding the importance

they place on three

dimensions of

"discretionary social

responsibility"

ROA, ROS 645 largest manufacturing

firms in Business Week's top

1000 companies (1985)

surveyed with a 25% response

rate. Subsample analyzed had

70% or more of total sales in

one 4-digit SIC.

1985-1986 1985-1986 .221 .221 .221 No

64 Graves and

Waddock

(1994)

430 Third-party

audit

Kinder, Lydenberg,

Domini (KLD) social

audit

ROA, ROE S&P 500, for which

institutional ownership and

CSP data available

1990 1990

.090 .090 .090 No Size

44

Attributes of 192 Effects from 167 Studies Included in Meta-Analyses of Association Between Corporate Social Performance and Corporate Financial Performance

Effect Size

Timing of CSP

measure Type of CFP

Study

# Study

N

Firms CSP Type CSP Measure CFP Measure

Sample of Companies and

Other Notes

Years

for CSP

Years

for CFP Overall

CFP -->

CSP Concurrent

CSP -->

CFP Accounting Market

Event

Study

Control

Variables

65 Graves and

Waddock

(2000)

22 Observer

perceptions

Collins and Porras

identify 18 firms from

CEO surveys that have

reputation of being

"highly visionary", and

matched to other firms in

the same industry with

the same founding date.

ROE, ROA,

ROS, Stock

returns

Companies from sample used

in book "Built to Last" by

Collins and Porras, they are

identified as result of CEO

surveys and all founded before

1950, so they are likely all

large firms.

N/L 1989-1996 .417 .412 .433 No Industry,

Size

66 Greening

(1995)

102 Environment

(objective)

US department of energy

data on programs to

manage energy savings;

Survey of expert

stakeholders, including

regulators and public and

private-funded

environmental and

consumer groups

ROA, EPS,

dividend yield

Utility companies for which

government data available on

energy management activities,

and for which stakeholders

responded to a reputational

survey

Energy

data 1990-

91, survey

1992

.094 .104 .083 .130 No Industry

67 Gregory,

Matatko and

Luther (1997)

N/A Screened

mutual funds

Varies depending on the

fund

Fund returns 18 UK ethical mutual funds

matched to non-ethics screened

funds by fund type, age and

size

1986-1994 1986-1994 -.004 -.004 -.004 No Risk, Size

68 Griffin and

Mahon (1997)

7 Observer

perceptions

Fortune magazine ratings

of "responsibility to

community and

environment"

ROS, ROE, ROA

Large US firms in the chemical

industry

1992 1992

.198 .458 .072 .198 No Industry,

Size

69 Griffin and

Mahon (1997)

7 Charitable

donations

Corporate philanthropy ROS, ROE, ROA

Large US firms in the chemical

industry

1992 1992

.165 .533 -.017 -.017 No Industry,

Size

70 Griffin and

Mahon (1997)

7 Third-party

audit

Kinder, Lydenberg,

Domini (KLD) social

audit

ROS, ROE, ROA

Large US firms in the chemical

industry

1992 1992

.041 .164 -.020 No Industry,

Size

71 Griffin and

Mahon (1997)

7 Environment

(objective)

Toxics Release Inventory

(reverse-coded)

ROS, ROE, ROA

Large US firms in the chemical

industry

1992 1992

-.451 -.679 -.306 No Industry,

Size

72 Grossman and

Sharpe (1986)

N/A Corporate

policies

Firms on the 1984 IRRC

list of divesting from

South Africa versus those

doing business in South

Africa

Stock returns All firms listed on NYSE. Note

effect sizes become positive for

1976-1983, perhaps reflecting

changing political views

Data

published

in 1984,

presumed

continuous

over time

1960-1983

-.101 -.101 -.101 No Risk, Size

73 Guerard

(1997a)

1300 Third-party

audit

Kinder, Lydenberg,

Domini (KLD) social

audit dichotomized into

firms satisfying KLD

versus unscreened stocks

(criteria are military,

nuclear power, alcohol,

tobacco, gambling and

environment).

Vantage Global Advisors'

database. Results listed as "not

significant", presumed zero.

1987-1994 1987-1994

.000 .000 .000 No

45

Attributes of 192 Effects from 167 Studies Included in Meta-Analyses of Association Between Corporate Social Performance and Corporate Financial Performance

Effect Size

Timing of CSP

measure Type of CFP

Study

# Study

N

Firms CSP Type CSP Measure CFP Measure

Sample of Companies and

Other Notes

Years

for CSP

Years

for CFP Overall

CFP -->

CSP Concurrent

CSP -->

CFP Accounting Market

Event

Study

Control

Variables

74 Guerard

(1997b)

1200 Third-party

audit

Kinder, Lydenberg,

Domini (KLD) social

audit dichotomized into

firms satisfying KLD

versus unscreened stocks

(criteria are military,

nuclear power, alcohol,

tobacco, gambling and

environment).

Vantage Global Advisors'

database. Results listed as "not

significant", presumed zero.

1987-1996 1987-1996

.000 .000 .000 No

75 Gunthorpe

(1997)

60 Revealed

misdeeds

Announcement in the

Wall Street Journal of

illegal or unethical

behavior: that the firm or

it's senior management is

under investigation, the

object of a law suit, or

that an indictment has

been issued

Stock returns Firms reported on in Wall

Street Journal

1988-1992 day -5 to

+5

.087 .087 .087 Yes Company

is its own

control

76 Hall and Rieck

(1998)

<= 27 Charitable

donations

Article in the Wall Street

Journal that firm

announces a voluntary

charitable donation

Stock returns Firms reported on in Wall

Street Journal - confounding

events deleted from sample

1982-1995 day -5 to

+5

.444 .444 .444 Yes Company

is its own

control

77 Hall and Rieck

(1998)

<= 40 Environment

(objective)

Article in the Wall Street

Journal that firm

announces a voluntary

action to benefit the

environment

Stock returns Firms reported on in Wall

Street Journal - confounding

events deleted from sample

1982-1995 day -5 to

+5

-.067 -.067 -.067 Yes Company

is its own

control

78 Hall and Rieck

(1998)

<= 32 Corporate

policies

Article in the Wall Street

Journal that firm

announces a voluntary

social action such as

sanctioning women's

rights, offering child-care

services, etc.

Stock returns Firms reported on in Wall

Street Journal - confounding

events deleted from sample

1982-1995 day -5 to

+5

-.116 -.116 -.116 Yes Company

is its own

control

79 Hamilton

(1995)

50 Environment

(objective)

Media coverage on Toxic

Release Inventory (TRI)

pollution data

Stock returns Firms in EPA's first Toxic

Release Inventory (TRI) data

release of June 19, 1989 - we

use in calculations only a

subset of this sample - those

firms who received media

coverage on TRI data during

1989

6/19/1989 day 0 to +5

.351 .351 .351 Yes Company

is its own

control

80 Hamilton, Jo

and Statman

(1993)

N/A Screened

mutual funds

Varies depending on the

fund

Fund returns Socially responsible equity

mutual funds (32 total)

identified by Lipper Analytical

Services as of December 1990

compared to 320 randomly

selected non-SRI mutual funds

in Lipper database

1/81-12/90 1/81-12/90 .000 .000 .000 No Risk

46

Attributes of 192 Effects from 167 Studies Included in Meta-Analyses of Association Between Corporate Social Performance and Corporate Financial Performance

Effect Size

Timing of CSP

measure Type of CFP

Study

# Study

N

Firms CSP Type CSP Measure CFP Measure

Sample of Companies and

Other Notes

Years

for CSP

Years

for CFP Overall

CFP -->

CSP Concurrent

CSP -->

CFP Accounting Market

Event

Study

Control

Variables

81 Hansen and

Wernerfelt

(1989)

60 Self-report Survey Of Organizations

(SOO) survey items of

employee perceptions of

welfare and working

conditions

ROA Fortune 1000 firms across

industries

Varies,

during

1970s-80s

5 years

preceding

.482 .482 .482 No Industry,

Size

82 Harper and

Adams (1996)

<=

1307

Environment

(objective)

Notification that a firm is

a Potentially Responsible

Party (PRP) for a

Superfund site

Stock returns Publicly traded firms named by

the EPA as responsible

Superfund parties. Effect listed

as "not significant", presumed

zero.

1980 -

1989

day -5 to

+5

.000 .000 .000 Yes Company

is its own

control

83 Hart and Ahuja

(1996)

127 Objective

environment

Percentage change in

emissions efficiency

index (ratio of reported

emissions in pounds to

company revenues)

reported by the toxic

release inventory (TRI).

ROA, ROE, ROS

S&P 500 firms in

manufacturing, mining or

production. Concurrent effect

listed as "not significant",

presumed zero. CSP-CFP

effect listed as "positive and

significant".

1988-1989 1989 .000 .000 "positive

and

significant

"

.000 No Industry,

Size

84 Heinze (1976) 28 Observer

perceptions

Ratings by the National

Affiliation of Concerned

Business Students

(NACBS)

ROS, ROA 28 large US corporations that

had been selected for ratings by

the NACBS

1972 1972

.057 .057 .057 No Size

85 Herremans et

al. (1993)

76 Observer

perceptions

Fortune magazine ratings

of "responsibility to

community and

environment"

Stock returns Large US manufacturing firms

rated by Fortune that had

consistent ratings over 5 year

period; matched pairs good and

poor reputation in same

industry

1982-7 1982-7

.229 .229 .229 No Industry,

Risk, Size

86 Hersch (1991) <=

123

Revealed

misdeeds

Lawsuit filing that

charges firm with

violating equal

employment opportunity

law

Stock returns Firms reported on in Wall

Street Journal

1964-1986 day -2 to

+1

.022 .022 .022 Yes Company

is its own

control

87 Hickman, Teets

and Kohls

(1999)

N/A Screened

mutual funds

Varies depending on the

fund

Fund returns Six socially responsible equity

mutual funds in Morningstar

database with at least 4 years

of data

12/91-

11/95

12/91-

11/95

-.050 -.050 -.050 No

88 Hill,

Ainscough,

Shank, &

Manullang

(2007)

33 Third-party

audit

Companies appearing in

at least 2-3 screened

mutual funds

Stock returns Companies later held in

screened mutual funds in the

US, Asia, and Europe,

compared with market

benchmarks

N/L,

presumabl

y approx

2005

1995-2005 .075 .075 .075 No Risk, Size

47

Attributes of 192 Effects from 167 Studies Included in Meta-Analyses of Association Between Corporate Social Performance and Corporate Financial Performance

Effect Size

Timing of CSP

measure Type of CFP

Study

# Study

N

Firms CSP Type CSP Measure CFP Measure

Sample of Companies and

Other Notes

Years

for CSP

Years

for CFP Overall

CFP -->

CSP Concurrent

CSP -->

CFP Accounting Market

Event

Study

Control

Variables

89 Hillman and

Keim (2001)

308 Third-party

audit

KLD ratings across 9

categories, separates

categories into categories

representing CSR to

primary stakeholders

(employee relations,

diversity issues, product

issues, community

relations, and

environment) and

representing social issues

(alcohol/tobacco/gamblin

g, military, nuclear

power, international)

Market Value

Added (minus

debt and invested

equity)

Firms rated by KLD, which are

primarily from the S&P 500

1994 1995-1996 .001 .001 .001 No Industry,

Risk, Size

90 Hoffer, Pruitt

and Reilly

(1988)

3 Revealed

misdeeds

Automobile safety recalls Stock returns Big 3 auto fir ms 1975-1981 day -5 to

+5

.286 .286 .286 Yes Company

is its own

control

91 Hylton (1992) N/A Screened

mutual funds

Varies depending on the

fund

Fund returns "Better-known" social

investment mutual funds, with

data provided by the Social

Investment Forum

1987-1992 1987-1992

-.027 -.027 -.027 No Risk

92 Ingram (1978) 287 Transparency Disclosures across five

topic areas in Annual

Reports

Stock returns Large US firms in the Fortune

500 from 1970-76

1970-76 1970-76

.025 .025 .025 No Industry,

Risk, Size

93 Ingram and

Frazier (1983)

27 Transparency Prevalence of discussion

of environmental quality

control in the President's

letter of the Annual

Report

ROI Firms from metal

manufacturing and fabricating

industry. Effect listed as "not

significant", presumed zero.

1978 1978

.000 .000 .000 No Industry,

Risk, Size

94 Jarrell and

Peltzman

(1985)

22 Revealed

misdeeds

Product recall

announcements, which

are major safety recalls in

the case of automobile

companies

Stock returns Drug firms, Big three auto

firms (GM, Ford, Chrysler)

1974-1982

(drugs),

1967-1981

(auto)

day -4/-5 to

+5

.563 .563 .563 Yes Company

is its own

control

95 Johnson and

Greening

(1999)

252 Third-party

audit

KLD ratings of

responsible behavior

towards communities,

women and minorities,

employee relations, and

product quality

ROA, ROE, ROS

A random sample of firms

from KLD database (mostly

Fortune 1000 firms) - with

managerial, institutional

ownership data

1993 1991-1992 .098 .098 .098 No Industry,

Size

96 Johnson and

Greening

(1999)

252 Environment

(objective)

KLD ratings of

environmental

performance

ROA, ROE, ROS

A random sample of firms

from KLD database (mostly

Fortune 1000 firms) - with

managerial, institutional

ownership data

1993 1991-1992 .133 .133 .133 No Industry,

Size

48

Attributes of 192 Effects from 167 Studies Included in Meta-Analyses of Association Between Corporate Social Performance and Corporate Financial Performance

Effect Size

Timing of CSP

measure Type of CFP

Study

# Study

N

Firms CSP Type CSP Measure CFP Measure

Sample of Companies and

Other Notes

Years

for CSP

Years

for CFP Overall

CFP -->

CSP Concurrent

CSP -->

CFP Accounting Market

Event

Study

Control

Variables

97 Jones and

Murrell (2001)

51 Third-party

audit

Inclusion in Working

Mother list, selected

according to following

criteria: salaries in

relation to competitors,

advancement

opportunities available to

women, on-site child

care, familiy benefits

such as maternity leave,

job sharing, flextime

Stock returns Firms that made Working

Mother magazine's list of the

top family-friendly companies

for the first time between 1989-

1994

1989-1994 day -1 to

+1

.217 .217 .217 Yes Company

is its own

control

98 Jones and

Rubin (2001)

<= 73 Objective

environment

Report of a negative

environmental incident,

excluding events that

directly hurt the customer

were excluded

Stock returns Oil and electric utility firms

reported on in the Wall Street

Journal

1970-1992 day -1 to 0

.049 .049 .049 Yes Company

is its own

control

99 Judge and

Douglas (1998)

170 Environment

(self-report)

Self-report measures of

environmental

management: functional

coverage, amount of

resources provided,

integration into strategic

planning, environmental

performance

Self-reported

ROI, earnings

growth, and

change in market

share relative to

others in industry

Questionnaire sent to 725 U.S.

firms from the 1992 World

Environmental Directory,

randomly chosen among those

with a corporate environmental

officer listed. 30% response

rate, primarily from large firms

Survey

around

1992-1995

Survey

around

1992-1995

.173 . 173 .173 No Size,

Industry

100 Karpoff and

Lott (1993)

71 Revealed

misdeeds

Announcement of

corporate crimes or fraud

(against customers,

suppliers, employees or

investors) reported by the

Wall Street Journal

Stock returns Firms reported on in Wall

Street Journal

1978-1987 day -1 to 0

.266 .266 .266 Yes Company

is its own

control

101 Karpoff, Lee

and Vendrzyk

(1999)

98 Revealed

misdeeds

News announcement that

firm is suspected or

convicted of committing

procurement fraud

Stock returns U.S. defense contractors

(mostly large firms)

1983-1995 day -1 to 0

.207 .207 .207 Yes Company

is its own

control

102 Karpoff, Lott

and Wehrly

(2005)

<=

423

Environment

(objective)

Announcement in the

Wall Street Journal that a

firm is being investigated,

accused or has settled

charges of an

environmental violation

Stock returns Firms reported on in Wall

Street Journal

1980-2000 day -1 to 0

.252 .252 .252 Yes Company

is its own

control

103 Kedia and

Kuntz (1981)

30 Charitable

donations

Charitable donations as

% of NI

ROA Commercial banking firms 1977 1977

-.288 -.288 No Industry

104 Kedia and

Kuntz (1981)

30 Corporate

policies

Banks issuing low

income loans and

minority enterprise loans

ROA Commercial banking firms 1977 1977

.005 .005 No Industry

105 Khanna,

Quimio and

Bojilova (1998)

91 Environment

(objective)

Improvement over time

in Toxic Releases

Inventory (TRI)

emissions data

Stock returns U.S. chemical industry report

release

1989-1994

day 0 to +1

-.209 -.209 -.209 Yes Company

is its own

control

49

Attributes of 192 Effects from 167 Studies Included in Meta-Analyses of Association Between Corporate Social Performance and Corporate Financial Performance

Effect Size

Timing of CSP

measure Type of CFP

Study

# Study

N

Firms CSP Type CSP Measure CFP Measure

Sample of Companies and

Other Notes

Years

for CSP

Years

for CFP Overall

CFP -->

CSP Concurrent

CSP -->

CFP Accounting Market

Event

Study

Control

Variables

106 King and Lenox

(2001)

652 Environment

(objective)

Measures of

manufacturing emissions

from the EPA 's Toxic

Release Inventory (TRI)

data

Tobin's q (market

value over

replacement cost

of tangible

assets)

Publicly traded U.S.

manufacturing firms

1987-1996 1988-1997 .037 .037 . 037 No Industry,

Size

107 King and Lenox

(2002)

614 Environment

(objective)

Total facility emissions

of toxic chemicals from

EPA 's Toxic Release

Inventory (TRI) data;

note that measure is by

facility, not firm

ROA, Tobin's q

(market value

over replacement

cost of tangible

assets)

Publicly traded U.S.

manufacturing firms

1991-1996 1992-1997 .056 .056 .048 .063 No Industry,

Size

108 Klassen and

McLaughlin

(1996)

<=98 Environment

(objective)

NEXIS newswire search

of environmental events,

separated into positive

events (e.g., international

news media

environmental award)

and negative events (e.g.,

crises identifed with the

keywords "oil",

"chemical", "gas leak",

"explosion" along with

words "spill" and

"environment")

Stock returns NYSE/AMEX firms (see Notes

on CSR Measure).

Observations with

"confounding events" dropped.

1985 -

1991

day -1 to

+1

.438 .438 .438 Yes Company

is its own

control

109 Konar and

Cohen (1997)

128 Environment

(objective)

Whether first toxic

emissions data release

during July 1989 resulted

in a media report

Stock returns Manufacturing firms subject to

reporting requirements of toxic

emissions

7/19/1989 0 to 5 days

.264 .264 .264 Yes Company

is its own

control

110 Konar and

Cohen (2001)

233 Environment

(objective)

Two measures provided

by Investor

Responsibility Research

Center, the pounds of

toxic chemicals emitted

per dollar revenue of firm

(TRI data) and number of

environmental lawsuits

pending in 1989

Tobin's q (market

value over

replacement cost

of tangible

assets)

S&P 500 companies in

polluting industries, mostly

manufacturing

1988-1989 1989

.240 .240 .240 Yes Industry,

Size

111 Kreander, Gray,

Power and

Sinclair (2002)

N/A Screened

mutual funds

Ethical vs. non-ethical

mutual funds

Fund returns 40 European ethical investment

funds matched to 40

conventional funds by age,

country, size and geographic

investment universe

1996-1998 1996-1998 .010 .010 . 010 No Risk, Size

112 Kreander, Gray,

Power, &

Sinclair (2005)

N/A Screened

mutual funds

Varies by fund Fund returns 30 ethical mutual funds

compared with 30 matched

conventional funds

1995-2001 1995-2001 .000 .000 . 000 No Risk

113 Kumar, Lamb

and Wokutch

(2002)

87 Corporate

policies

Doing business in South

Africa upon the lifting of

investment sanctions

Stock returns KLD listing of firms with

equity interests in South Africa

at the time of lifting sanctions

9/24/1993 -30 to 50

days

.168 .168 .168 Yes Company

is its own

control

114 Kurtz and

DiBartolomeo

(1996)

650 Third-party

audit

Domini 400 Social stock

index

Stock returns Domini 400 Social stock index

compared to S&P 500

5/90-9/93 5/90-9/93

.000 .000 .000 No Industry,

Risk, Size

50

Attributes of 192 Effects from 167 Studies Included in Meta-Analyses of Association Between Corporate Social Performance and Corporate Financial Performance

Effect Size

Timing of CSP

measure Type of CFP

Study

# Study

N

Firms CSP Type CSP Measure CFP Measure

Sample of Companies and

Other Notes

Years

for CSP

Years

for CFP Overall

CFP -->

CSP Concurrent

CSP -->

CFP Accounting Market

Event

Study

Control

Variables

115 Laplante and

Lanoie (1994)

<=21 Environment

(objective)

Announcements of

environmental actions

against firms: violation of

environmental regulation

for which it is likely that

regulator will undertake

legal action, or

announcement that legal

action has been

undertaken

Stock returns Environmental events

published in Canadian

newspapers, typically firms in

pulp and paper, mining,

petroleum and chemical

industries

1982-1991 day -30 to

+30

-.026 -.026 -.026 Yes Company

is its own

control

116 Lashgari and

Gant (1989)

99 Corporate

policies

Compliance with the

Sullivan principles code

of conduct for fair labor

practices developed for

U.S. businesses operating

in South Africa

ROI, ROE Large, well-established firms

monitored for compliance by

Arthur D. Little, Inc.with

respect to the Sullivan

principles.

1977-1983 1977-1983

.036 .036 .036 No

117 Levy and Shatto

(1980)

55 Charitable

donations

Corporate gift giving Net Income (not

a financial ratio)

Large public utility firms 1976 1976

.720 .720 .720 No

118 Luck and

Pilotte (1993)

643 Screened

mutual funds

Domini 400 Social stock

index

Stock returns Domini 400 Social stock index

compared to S&P 500

5/90-9/92 5/90-9/92

.324 .324 .324 No Industry,

Risk, Size

119 Luo &

Bhattacharya

(2006)

113 Observer

perceptions

Fortune magazine ratings

of "responsibility to

community and

environment"

ROA, Stock

returns, Tobin's q

Companies rated by Fortune

that also have marketing data

available from ASCI.

2001-2003 2002-2004 .153 .153 .190 .135 No Size

120 Luther and

Matatko (1994)

N/A Screened

mutual funds

Varies depending on the

fund

Fund returns UK open-ended ethical-

screened mutual funds with 32

months history, with at least

80% of their equity portfolio

invested in the UK

8/89 - 3/92 8/89 - 3/92

.314 .314 .314 No Risk, Size

121 Luther, Matat ko

and Corner

(1992)

N/A Screened

mutual funds

Varies depending on the

fund

Fund returns UK open-ended mutual funds

identified by Ethical

Investment Research Service in

May 1990 with at least 2 years

of return data

5/88-6/90 5/88-6/90

-.084 -.084 .000 No Risk

122 Mahoney and

Roberts (2004)

214 Environment

(objective)

Canadian Social

Investment Database

(CSID) ratings of

environmental

performance, designed to

mirror the KLD Ratings

ROA, ROE, ROS

Firms in Toronto Stock

Exchange 300

1995-1999 1995-1999 .063 .063 .063 No

123 Mahoney and

Roberts (2004)

214 Third-party

audit

Canadian Social

Investment Database

(CSID) ratings of social

responsibility designed to

mirror the KLD ratings,

along the dimensions of

community, diversity,

employee relations,

international, product and

business practices

ROA, ROE, ROS

Firms in Toronto Stock

Exchange 300

1995-1999 1995-1999 .012 .012 .012 No

51

Attributes of 192 Effects from 167 Studies Included in Meta-Analyses of Association Between Corporate Social Performance and Corporate Financial Performance

Effect Size

Timing of CSP

measure Type of CFP

Study

# Study

N

Firms CSP Type CSP Measure CFP Measure

Sample of Companies and

Other Notes

Years

for CSP

Years

for CFP Overall

CFP -->

CSP Concurrent

CSP -->

CFP Accounting Market

Event

Study

Control

Variables

124 Mallin,

Saadouni and

Briston (1995)

N/A Screened

mutual funds

Varies depending on the

fund

Fund returns UK ethical funds identified by

Finstat, each matched to a non-

ethical fund with same

establishment date and fund

size

1986-1993

Funds

have

varying

sample

periods

1986-1993

Funds have

varying

sample

periods

-.232 -.232 -.232 No Risk

125 McGuire et al.

(1988)

98 Observer

perceptions

Fortune magazine ratings

of "responsibility to

community and

environment"

Stock returns,

ROA, and

growth in sales,

assets, and

income

Fortune 500 firms 1982,

1983,

1983-5

(reported)

1982-5

.107 .225 .056 . 039 .160 .000 No Risk, Size

126 McGuire,

Schneeweis and

Branch (1990)

131 Observer

perceptions

Fortune magazine ratings

of "responsibility to

community and

environment"

Stock returns,

ROA

Fortune 500 firms 1982 1977-

1981/2,

1982-

1984/5

.319 .503 . 109 No Risk, Size

127 McWillia ms

and Siegel

(1997)

22 Third-party

audit

U.S. Department of

Labor presents annual

award to firm with high-

quality affirmative action

policy (around 6 firms

receive awards per year)

Stock returns Firms receiving award from the

U.S. Department of labor

1986-1992 -10 to +10

days

-.211 -.211 -.211 Yes Company

is its own

control

128 McWillia ms

and Siegel

(1997)

21 Revealed

misdeeds

Major settlement

announcement by firm

found guilty of

discrimination against

underrepresented

minority employees

Stock returns Large U.S. firms 1986-1992 -10 to +10

days

.021 .021 .021 Yes Company

is its own

control

129 McWillia ms

and Siegel

(2000)

524 Screened

mutual funds

Domini 400 Social stock

Index

Stock returns Firms in the S&P 500 and

Domini 400 Social stock Index

1991-1996 1991-1996 .043 .043 . 043 No Industry,

Risk, Size

130 Menguc &

Ozanne (2005)

140 Environment

(self-report)

Survey based on 10

Valdez Principles of

environmental

management and

commitment to the

natural environment

Earnings Large Australian

manufacturing firms whose

CEO or equivalent executive

responded to survey (29%

response rate).

N/L 1999 .131 .131 No Industry,

Size

131 Meznar, Nigh

and Kwok

(1994)

39 Corporate

policies

Corporate divestiture

from South Africa

Stock returns U.S. corporations who publicly

announced withdrawal from

South Africa

1985-1989 day -30 to

+10

-.413 -.413 -.413 Yes Company

is its own

control

132 Mill (2006) N/A Screened

mutual funds

Range of ethical

screening criteria

Fund returns Family Charities Ethical Fund

in the UK after adopting social

screening versus 3

conventional control funds

3/1996-

3/2004

3/1996-

3/2004

.063 . 063 .063 No Size

133 Muoghalu,

Robison and

Glascock

(1990)

<=202 Environment

(objective)

Lawsuit or case

settlement involving toxic

or hazardous waste

materials

mismanagement

Stock returns Firms with environmental

cases announced in the Wall

Street Journal

1977-1986 day -5 to

+5

.077 .077 .077 Yes Company

is its own

control

52

Attributes of 192 Effects from 167 Studies Included in Meta-Analyses of Association Between Corporate Social Performance and Corporate Financial Performance

Effect Size

Timing of CSP

measure Type of CFP

Study

# Study

N

Firms CSP Type CSP Measure CFP Measure

Sample of Companies and

Other Notes

Years

for CSP

Years

for CFP Overall

CFP -->

CSP Concurrent

CSP -->

CFP Accounting Market

Event

Study

Control

Variables

134 Nakao, Amano,

Matsumura,

Genba, &

Nakano (2007)

278 Environmental

(self-report)

Nikkei Environmental

Management Survey,

including planning,

disclosure, education, and

management of waste and

energy

Tobin's q-1,

ROA

Large Japanese manufacturing

companies, excluding energy

and construction. Compares

top 30 vs. bottom 30 firms.

1999-2003 1999-2003 .068 .070 .065 .063 .072 No Industry,

Size

135 Nehrt (1996) 50 Environment

(objective)

Timing and intensity of

environmentally

beneficial investments in

paper manufacturing to

reduce chlorine content

Growth in net

income

Producers of chemical

bleached paper pulp in 8

countries, listed in Pulp and

Paper International's 1992

annual review of the largest

150 paper companies in the

world. The sample of 50 were

chemical paper pulp

manufacturers with at least

70% of their revenues in the

paper industry. 19 are from the

US.

1984-1991 1983-1991

.105 .105 .105 No Industry,

Size

136 Newgren et al.

(1985)

50 Self-report Survey of whether there

is a concerted effort to

identify and analyze the

social and political

environment

P/E ratio Largest corporations in five

industrial and five

nonindustrial classifications

1975 1975-80

.272 .272 .272 No Industry,

Size

137 O'Neill et al.

(1989)

<=157 Self-report Self-report measure using

forced-choice response to

rank priorities

ROA Companies from list of Fortune

magazine's most admired

companies. Note that multiple

directors may have responded

from the same firm, so the total

number of firms <=157.

N/L,

presume

1985

N/L,

presume

1985

-.043 -.038 -.047 No Risk, Size

138 Patten (1990) 54 Corporate

policies

Signing of the Sullivan

Principles for divestment

from South Africa

Stock returns 37 firms whose signing of

Sullivan principles about South

Africa was disclosed

1977 Day -4 to

+2

.053 .053 .053 Yes Company

is its own

control

139 Patten (1991) 156 Transparency Annual report disclosures

in 7 categories:

Environment, energy, fair

business practices, human

resources, community

involvement, products,

and other

ROA, ROE Sample drawn from 8 industry

classifications of Fortune 500

(petroleum refining, chemical,

forest and paper products,

electronics, industrial and farm

machinery, metal products,

computer, and rubber products)

1985 1980-5

.023 -.009 .056 .023 No Industry,

Size

140 Pava and

Krausz (1996)

106 Third-party

audit

Council for Economic

Priorities report

Market return,

P/E ratio,

market-to-book

value, ROA,

ROE, EPS

Firms rated by the CEP along

with matched control firms

1991

report

1985-8,

1989-91

.224 .357 . 081 No Industry,

Size

141 Peltzman

(1981)

<= 23 Revealed

misdeeds

Event is an initial

complaint by FTC of

false advertising

Stock returns Consumer goods firms with

major FTC cases concerning

false advertising

1962-1975 day -3 to

+1

.669 .669 .669 Yes Company

is its own

control

53

Attributes of 192 Effects from 167 Studies Included in Meta-Analyses of Association Between Corporate Social Performance and Corporate Financial Performance

Effect Size

Timing of CSP

measure Type of CFP

Study

# Study

N

Firms CSP Type CSP Measure CFP Measure

Sample of Companies and

Other Notes

Years

for CSP

Years

for CFP Overall

CFP -->

CSP Concurrent

CSP -->

CFP Accounting Market

Event

Study

Control

Variables

142 Posni koff

(1997)

40 Corporate

policies

Announcement of

disinvestment in South

Africa

Stock returns Companies who announced

disinvestment in South Africa

between 1980-1991, as

announced in the Wall Street

Journal and New York Times

1980-1991 day -1 to

+1

.541 .541 .541 Yes Company

is its own

control

143 Preston and

O'Bannon

(1997)

67 Observer

perceptions

Fortune magazine ratings

of "responsibility to

community and

environment"

ROA Fortune annual survey,

included companies that have

been rated in every survey

1982-1992

1982-1992 1982-1992

.370 .407 .373 . 330 .370 No Size

144 Preston and

Sapienza (1990)

108 Observer

perceptions

Fortune magazine ratings

of "responsibility to

community and

environment"

Stock returns 10 largest companies in

industries surveyed by Fortune.

Only companies surveyed in all

five years 1982 to 1986

included.

1982-1986 1977-1986

.190 .190 .190 No Size

145 Reichert,

Lockett and

Rao (1996)

83 Revealed

misdeeds

Announcement that

formal indictment has

been filed against firm for

major corporate crimes

Stock returns Firms reported on in Wall

Street Journal

1980-1990 day -10 to

+10

.064 .064 .064 Yes Company

is its own

control

146 Reimann (1975)

19 Self-report Self-report of value of

seven different

constituencies

Self-reported

survey of goal

achievement in

comparison to

other

organizations

along eight

dimensions

American manufacturers from

multiple industries, varying

from 200 to 4,000 employees

1970 1974

.570 .570 No Industry

147 Rennin gs,

Schroder and

Ziegler (2003)

153 Third-party

audit

Swiss bank's evaluations

of European corporations

on social and

environmental

performance. Evaluations

are both by industry

sector as a whole, and

individual firms relative

to their industries

Stock returns 300 European corporations

whose social and

environmental sustainability

has been evaluated by the

Swiss bank Sarasin & Cie.

Effect "not significant",

presumed zero.

data

released

2001

1/1996 -

8/2001

.000 .000 .000 No Industry,

Risk, Size

148 Rey & Nguyen

(2005)

N/A Screened

mutual funds

AMP Capital's

Sustainable Future

Australian Share Fund

inclusion, based on

stakeholder relationships

with employees,

customers, shareholders,

environment

Fund returns Australian publicly traded

companies

11/2001-

2004

1995-2004 .097 .097 .097 No Industry,

Risk, Size

149 Reyes and

Grieb (1998)

N/A Screened

mutual funds

Varies depending on the

fund

Fund returns Fifteen socially responsible

mutual funds from Wilson

Associates Capital Asset

Management System Database,

matched by investment type to

control funds

1986-1995 1986-1995

.034 .034 .034 No Industry,

Risk, Size

150 Roberts (1992) 80 Charitable

donations

Existence of a funded

foundation

ROE Large Fortune 500 companies

profiled by the CEP

1983-4 1981-84

.161 .161 .161 No Size

54

Attributes of 192 Effects from 167 Studies Included in Meta-Analyses of Association Between Corporate Social Performance and Corporate Financial Performance

Effect Size

Timing of CSP

measure Type of CFP

Study

# Study

N

Firms CSP Type CSP Measure CFP Measure

Sample of Companies and

Other Notes

Years

for CSP

Years

for CFP Overall

CFP -->

CSP Concurrent

CSP -->

CFP Accounting Market

Event

Study

Control

Variables

151 Roberts (1992) 80 Third-party

audit

Council on Economic

Priorities (CEP) ratings

ROE Large Fortune 500 companies

profiled by the CEP

1984-6 1981-84

.203 .203 .203 No Size

152 Rockness,

Schlachter and

Rockness

(1986)

21 Environment

(objective)

Government verifiable

self-reported

environmental data on

total cumulative tonnage

of waste disposed of at

on-premise sites and

number of EPA

Superfund sites at which

company has dumped

waste, both normalized

by company asset size

and number of plants.

ROS, ROA,

ROE, excess

market value

Chemical firms (if at least 40%

of sales were chemical) who

participated in a self-reported

congressional "Site Survey"

conducted in 1979

Firms listed Table 1

1979 1969, 1979  .174 .174 .227 .008 No Industry

153 Ruf,

Muralidhar,

Brown, Janney

and Paul (2001)

488 Third-party

audit

KLD ratings ROE, ROS KLD database 1990-1991 1990-1991 .043 -.015 .100 No

154 Russo and

Fouts (1997)

243 Environment

(objective)

Data on environmental

performance

ROA, growth in

sales

Large US firms rated by the

Franklin Research and

Development Corporation

(FRDC) on environmental

performance

1991, 1992

.145 .145 .145 No Industry

155 Sauer (1997) N/A Screened

mutual funds

Domini 400 Social stock

index

Stock returns S&P 500 firms, and firms in

Domini 400 Social stock index

1986-1994 1986-1994

.086 .086 .086 No Risk, Size

156 Schroder (2003)

N/A Screened

mutual funds

Varies by fund Fund returns 46 major SRI investment funds

from 10 SRI indices in the

U.S., Germany and

Switzerland with at least 30

months history

1990 -

9/2002

1990 -

9/2002

-.515 -.515 -.515 No Risk, Size

157 Seifert, Morris,

& Bartkus

(2003)

82 Charitable

donations

Charitable giving data

from the Foundation

Center

Cash flow, ROA,

ROE, ROS,

market-to-book

ratio, stock

returns

US publicly held firms with

charity information available

from the Foundation Center,

with matched pair of high

($5M+) versus low donors

N/L Year -1, 0,

+1

.144 .280 .150 .125 .128 .172 No Industry,

Size

158 Seifert, Morris,

& Bartkus

(2004)

157 Charitable

donations

Corporate philanthropy in

yearly cash and total

yearly charity via

foundations

Stock returns Fortune 1000 firms with

consistent charity data in the

Taft Corporate Giving

Directory and Foundation

Center's Foundation Directory

in 1997 or 1998, excluding

financial services and utilities

1997-1998 1997-1998 .151 .151 .245 .055 No Industry

159 Shane and

Spicer (1983)

72 Environment

(objective)

CEP reports on

environmental

performance

distinguishing low vs.

high polluters

Stock returns Industrial companies studied

by the CEP

1970-1977 day 0 to +1

.213 .213 .213 Yes Company

is its own

control

160 Shank (2005) 11 Third-party

audit

Companies appear in at

least 1/3 of screened

mutual funds

Stock returns Large companies appearing in

screened mutual funds,

compared to the NYSE

2003 1993-2003 .261 .261 .261 No Size

55

Attributes of 192 Effects from 167 Studies Included in Meta-Analyses of Association Between Corporate Social Performance and Corporate Financial Performance

Effect Size

Timing of CSP

measure Type of CFP

Study

# Study

N

Firms CSP Type CSP Measure CFP Measure

Sample of Companies and

Other Notes

Years

for CSP

Years

for CFP Overall

CFP -->

CSP Concurrent

CSP -->

CFP Accounting Market

Event

Study

Control

Variables

161 Shank (2005) N/A Screened

mutual funds

Varies by fund Fund returns Five large-cap mutual funds

listed by the Social Investment

Forum with 10-year history,

compared to the NYSE

1993-2003 1993-2003 .067 No Size

162 Simerly (1994) 110 Observer

perceptions

Fortune magazine ratings

of "responsibility to

community and

environment",

dichotomized into high

vs. low

EPS, share price,

market value,

ROE,

sales/equity,

ROI, sales rate

Fortune 500 firms within select

industries experiencing growth

in 86-88 and decline in 89-90

1988 1986-88

.233 .210 .255 .199 .315 No Industry,

Size

163 Simerly (1995) 84 Observer

perceptions

Fortune magazine ratings

of "responsibility to

community and

environment",

dichotomized into high

vs. low

ROE Fortune 500 firms, 42 pairs

matched within industries for

high and low CSR

1988 1988-90

.265 .265 .265 No Industry,

Size

164 Spencer and

Taylor (1987)

120 Observer

perceptions

Fortune magazine ratings

of "responsibility to

community and

environment"

ROA, ROS Fortune 500 firms in

manufacturing industries

1982 1978-1982

.149 .196 -.043 No Industry,

Size

165 Spicer (1978) 17 Environment

(objective)

Council on Economic

Priorities (CEP) ratings

of environmental

performance

ROE, P/E ratio Pulp and paper industry firms

reviewed by CEP

1970, 72 1968-73

.515 .515 .437 .586 No

166 Starik (1990) N/L Self-report Survey measure of

stakeholder

communication

management strategies

ROI, Revenue US investor-owned electrical

utility companies

1979-1988 1979-1988

.039 .039 .039 No Industry,

Risk, Size

167 Starik (1990) N/L Observer

perceptions

Survey measure of

stakeholder

communication

management strategies;

Reputation of utility as

rated by regulators,

consumer, utility industry

members, and other

respondents

ROI, Revenue US investor-owned electrical

utility companies

1979-1988 1979-1988

.095 .095 .095 No Industry,

Risk, Size

168 Statman (2000) N/A Screened

mutual funds

Inclusion in the Domini

Social Index

Stock returns Domini Social Index and S&P

500 firms, and socially

responsible equity mutual

funds (31 total) on

Morningstar's list as of

September 1998 matched to 62

conventional mutual funds by

asset size

5/90-9/98 5/90-9/98 .127 .127 .127 No Industry,

Risk, Size

56

Attributes of 192 Effects from 167 Studies Included in Meta-Analyses of Association Between Corporate Social Performance and Corporate Financial Performance

Effect Size

Timing of CSP

measure Type of CFP

Study

# Study

N

Firms CSP Type CSP Measure CFP Measure

Sample of Companies and

Other Notes

Years

for CSP

Years

for CFP Overall

CFP -->

CSP Concurrent

CSP -->

CFP Accounting Market

Event

Study

Control

Variables

169 Staw &

Szwajkowski

(1975)

500 Revealed

misdeeds

Violations of antitrust

laws and the Federal

Trade Commission Act

ROE, ROS Fortune 500 companies with

convictions or merit found to

trade litigation vs. others. CFP

aggregated for 5 years

preceding the trade complaint.

Positive relationship without

industry control variable.

1968-1972 1954-1971 -.004 -. 004 -.004 No Industry,

Size

170 Stevens (1984) 48 Environment

(objective)

CEP estimates of

required pollution

abatement costs

Stock returns Largest firms in four industries

among the five most polluting:

pulp and paper, pertroleum

refining, steel, electrical

utilities

CEP data

release

dates

1972-1977

-11 to 0

months

.490 .490 .490 Yes Industry,

Risk, Size

171 Sturdivant and

Ginter (1977)

28 Observer

perceptions

Company ratings of

social responsiveness

made by business

journalist Moskowitz

(1972)

EPS Large US firms rated by

Moskowitz (1973), which were

in industries in which there was

variance in Moskowitz ratings

1972-4 1964-74

.455 .455 .455 No Industry,

Size

172 Teoh, Welch

and Wazzan

(1999)

46 Corporate

policies

Divestiture from South

Africa

Stock returns Dow Jones News Retrieval

announcements of firms

divesting from South Africa

1983-1989 -1 to 1 days

-.002 -.002 -.002 Yes Company

is its own

control

173 Tichy, McGill,

St. Clair (1997)

10 Observer

perceptions

Fortune magazine ratings

of "responsibility to

community and

environment"

ROE Fortune annual survey

companies rated in top 3 of

firms at least once 1983-1996,

compared to S&P 500

1983-1996 1983-1996

.758 .758 .758 No Industry,

Size

174 Travers (1997) N/A Screened

mutual funds

Varies depending on the

fund

Fund returns Sample of international

socially responsible equity

funds matched with

unrestricted funds with the

same fund manager. Effect

"not significant", presumed

zero.

7/92-6/97 7/92-6/97

.000 .000 .000 No N/L

175 Turban and

Greening

(1996)

161 Observer

perceptions

Rating of corporate

reputation by

undergraduate

management students

ROA Large US firms rated both by

KLD and Fortune, excluding

those firms whose reputation

and attractiveness was not well

known to pilot test

undergraduate management

students and faculty

1993-94 N/L,

presume

concurrent

.250 .250 .250 No

176 Turban and

Greening

(1996)

161 Environment

(objective)

KLD ratings of

environmental

performance

ROA Large US firms rated both by

KLD and Fortune, excluding

those firms whose reputation

and attractiveness was not well

known to pilot test

undergraduate management

students and faculty

1993-94 N/L,

presume

concurrent

-.020 -.020 -.020 No

57

Attributes of 192 Effects from 167 Studies Included in Meta-Analyses of Association Between Corporate Social Performance and Corporate Financial Performance

Effect Size

Timing of CSP

measure Type of CFP

Study

# Study

N

Firms CSP Type CSP Measure CFP Measure

Sample of Companies and

Other Notes

Years

for CSP

Years

for CFP Overall

CFP -->

CSP Concurrent

CSP -->

CFP Accounting Market

Event

Study

Control

Variables

177 Turban and

Greening

(1996)

160 Third-party

audit

KLD ratings of

community relations,

employee relations,

product quality, and

treatment of women and

minorities

ROA Large US firms rated both by

KLD and Fortune, excluding

those firms whose reputation

and attractiveness was not well

known to pilot test

undergraduate management

students and faculty

1993-94 N/L,

presume

concurrent

.085 .085 .085 No

178 Van de Velde,

Vermeir, &

Corten (2005)

N/A Third-party

audit

CSP scores from the

European agency Vigeo

regarding treatment of

human resources, the

environment, customers

and suppliers, community

and society, and

corporate governance

Stock returns Companies rated on CSP by

Vigeo, compared to the

European market index MSCI

EMU. Effect "not significant",

presumed zero.

2000-2003 2000-2003 .000 .000 . 000 No Risk, Size

179 Vance (1975) 49 Observer

perceptions

Ratings by urban-affairs

and public-affairs

corporate staff members

and business students in

the National Affiliation

of Concerned Business

Students

Stock returns Large national corporations

(Fortune 100 equivalent)

1972 1974

-.460 -.460 -.460 No Size

180 Verschoor

(1998)

376 Transparency Indication in an Annual

Report of a commitment

to ethics or describes the

importance of an ethics

standard or code of

conduct

Business Week

ranking on

financial

performance,

based on

acounting

measures

S&P 500 1996 1996

.151 .151 .151 No

181 Waddoc k and

Graves (1997)

469 Third-party

audit

Kinder, Lydenberg &

Domini (KLD) ratings

ROA, ROE, ROS

S&P 500 1990 1989

.124 .100 .147 No Size

182 Waddoc k,

Graves and

Gorski (2000)

483 Third-party

audit

KLD screening of

socially responsible

companies

ROA, ROE S&P 500 companies high vs.

low in KLD screens

1996 1996

.017 .017 .017 No Size

183 Waddoc k,

Graves and

Gorski (2000)

483 Screened

mutual funds

S&P 500 firms that have

passed KLD screens vs.

S&P 500 firms that

haven't

Stock returns Companies meeting vs. not

meeting inclusion criteria for

KLD screening

1996 1987-1996

.053 .053 .053 No Size

184 Waldman,

Siegel and

Javidan (2004)

56 Third-party

audit

KLD ratings along 8

categories separated into

two factors of "social"

and "strategic" using

factor analysis (excluding

alcohol, tobacco,

gambling categories as

they were not available

the entire sample period)

ROE 95 U.S and 55 Canadian firms

randomly sampled from

publicly traded firms with

Sales above $1B and CEOs

with at least 2 years tenure,

with 28% response rate to a

survey on a different topic

1991-1992 1991-1992 .274 .225 .320 .274 No Industry,

Size

58

Attributes of 192 Effects from 167 Studies Included in Meta-Analyses of Association Between Corporate Social Performance and Corporate Financial Performance

Effect Size

Timing of CSP

measure Type of CFP

Study

# Study

N

Firms CSP Type CSP Measure CFP Measure

Sample of Companies and

Other Notes

Years

for CSP

Years

for CFP Overall

CFP -->

CSP Concurrent

CSP -->

CFP Accounting Market

Event

Study

Control

Variables

185 Wartick (1988) 252 Corporate

policies

Membership in the Issues

Management Association

representing a concerted

effort to identify and

analyze the social and

political environment

Survey rating of

CFP: long-term

investment value,

soundness of

financial

position, and

wise use of

corporation

assets

Large US firms rated by

Fortune magazine

1986 1986

.023 .023 No Industry,

Size

186 Wokutch and

Spencer (1987)

74 Charitable

donations

Archival data on

charitable giving

ROA, ROS Large manufacturing firms

rated by Fortune, for which

charitable and financial

information available

1982-84 1978-82

.323 .323 .323 No Industry,

Size

187 Wokutch and

Spencer (1987)

39 Revealed

misdeeds

Archival data on anti-

trust corporate crimes

ROA, ROS Large manufacturing firms

rated by Fortune, for which

charitable and financial

information available

1980-3 1978-82

.227 .227 .227 No Industry,

Size

188 Wright and

Ferris (1997)

31 Corporate

policies

Firm divestment from

South Africa as reported

by the Investor

Responsibility Research

Center

Stock returns Firms on IRRC South Africa

divestment list. Firms dropped

from sample if they did not

have good Sullivan ratings

prior to divestment.

Firms listed Table 1

1984-1990 day -10 to

+10

.005 .005 .005 Yes Company

is its own

control

189 Wright, Ferris,

Hiller

and Kroll

(1995)

34 Third-party

audit

U.S. Department of

Labor presents annual

award to firm with high-

quality affirmative action

policy (around 6 firms

receive awards per year)

Stock returns Department of Labor news

releases reported by the Wall

Street Journal Index and Dow

Jones News Retrieval Service

1986-1992 day -10 to

+10

.259 .259 .259 Yes Company

is its own

control

190 Wright, Ferris,

Hiller

and Kroll

(1995)

35 Revealed

misdeeds

Major settlement

announcement by firm

found guilty of

discrimination against

underrepresented

minority employees

Stock returns Department of Labor news

releases reported by the Wall

Street Journal Index and Dow

Jones News Retrieval Service

1986-1992 day -10 to

+10

.189 .189 .189 Yes Company

is its own

control

191 Yamashita, Sen

and Roberts

(1999)

49 Environment

(objective)

Fortune ratings of firms

on 20 environmental

categories. Compares

firms cited as "10

Leaders" and "10 Most

Improved" vs. "10

Laggards"

Stock returns 130 of America's largest

manufacturing companies

whose environmental

performance ranking was

reported on by Fortune

Event day

7/12/93,

data cover

1987-1993

day -24 to

+24

-.113 -.113 -.113 Yes Company

is its own

control

192 Yamashita, Sen

and Roberts

(1999)

49 Environment

(objective)

1995 Council for

Economic Priorities

(CEP) rankings of

environmental

performance

Stock returns,

earnings

49 companies rated both by the

CEP "Better Investment

Guide" (1991) and the CEP

Screen Service (1995)

1995 1986-1995 -.023 .145 -.090 -.090 .145 No Size

59

TABLE 3

Summary of Results from Meta-Analyses of 167 Studies of the Association Between Corporate Social Performance

and Corporate Financial Performance

Effect Size

Timing of CSP Measure Type of CFP Significance Test Heterogeneity Test

CSP Type Overall

CFP -->

CSP Concurrent

CSP -->

CFP Accounting Market Companies Z p-value chi-square Df p-value

Mean Values

Overall .132 . 148 .115 .140 .180 .104 27,848 16.07 <.001 742.26 166 <.001

N (192) (35) (110) (66) (75) (125)

Charitable contributions .239 .332 . 198 .292 .281 .147 1,881 6.97 <.001 83.99 12 <.001

N (13) (4) (10) (2) (10) (5)

Corporate policies .019 .111 -.031 .011 . 040 .015 942 .96 .17 23.22 11 .02

N (13) (2) (5) ( 8) (2) (9)

Environmental performance .112 -.051 .145 .106 .102 .121 8,195 8.15 <.001 140. 16 44 <.001

N (45) (5) (20) (22) ( 19) (32)

Objective .095 -.081 .117 .104 .088 .118 7, 108 6.38 <.001 126.59 36 <.001

N ( 37) (4) (15) (20) (15) (28)

Self-reported .190 . 070 .225 .127 .153 .140 1, 087 5.60 <.001 13.58 7 .06

N (8) ( 1) (5) (2) (4) (4)

Observer perceptions .287 .328 .279 .157 .320 .190 2,000 9.44 <.001 161.02 23 <.001

N (25) (6) (19) (7) (16) (15)

Revealed misdeeds .223 -.004 .227 .239 . 113 .239 1,373 5.02 <.001 51.58 15 <.001

N (16) (1) (1) (14) (2) (14)

Screened mutual funds .024 .053 .021 - - .014 3,271 - - - - -

N (29) (1) (27) (0) (0) (26)

Self-reported performance .210 .200 . 171 .272 .171 .272 967 4.78 <.001 38.94 7 <.001

N (9) (3) (5) (1) (6) (1)

Third-party audit .080 .142 .041 .096 .114 .059 7,386 4.12 <.001 41.23 26 <.001

N (28) (8) (14) (9) (11) (17)

Transparency .078 .079 .029 .191 .102 .056 1,833 3.12 <.001 39.19 13 <.001

N (14) (5) (9) ( 3) (9) (6)

Median

Overall .082 . 164 .055 .112 .133 .053

Charitable contributions .161 .302 . 137 .284 .203 .055

Corporate policies .005 .111 .005 .002 .040 -.002

Environmental performance .094 .133 .159 .060 .130 .075

Objective .077 . 139 .105 .052 .105 .070

Self-reported .181 . 070 .233 .127 .152 .130

Observer perceptions .229 .316 .153 .265 .258 .135

Revealed misdeeds .192 -.004 .227 .192 . 111 .192

Screened mutual funds .021 .053 .010 - - .005

Self-reported performance .124 .120 . 039 .272 .122 .272

Third-party audit .042 .132 .000 .100 .089 .001

Transparency .024 -.009 .023 .129 .023 .024

Weighted Mean

Overall .101 . 120 .102 .085 .140 .086

Charitable contributions .220 .273 . 213 .215 .254 .071

Corporate policies .038 .080 .006 .026 .041 .049

Environmental performance .090 .099 .137 .066 .088 .083

Objective .078 . 124 .114 .064 .062 .081

Self-reported .169 . 071 .223 .091 .149 .104

Observer perceptions .296 .338 .267 .162 .332 .267

Revealed misdeeds .104 -.004 .237 .165 . 012 .165

Screened mutual funds - - - - - -

Self-reported performance .128 .119 . 075 .285 .141 .285

Third-party audit .037 .102 .015 .081 .083 .011

Transparency .099 .047 .100 .096 .079 .127

Note: Weighted means, significance and heterogeneity tests include only those studies reporting the number of companies sampled.

Insufficient numbers of screened mutual fund studies provided data regarding number of companies.

Values in parentheses are the number of effects on which the above coefficient is based. Total effects = 192.

60

REFERENCES

Aldag, R. J., & Bartol, K. M. K. M. 1978. Empirical studies of corporate social performance and

policy: A survey of problems and results. Research in Corporate Social Performance

and Policy, 1: 165–199.

Alexander, G. J., & Buchholz, R. A. 1978. Corporate social performance and stock market

performance. Academy of Management Journal, 21: 479–486.

Allen, W. T. 1993. Contracts and communities in corporate law. Washington and Lee Law

Review, 50: 1395–1407.

Allen, W. T. 1992. Our schizophrenic conception of the corporation. Cardozo Law Review, 14:

261–280.

Allouche, J., & Laroche, P. 2005. A meta-analytical investigation of the relationship between

corporate social and financial performance. Revue de Gestion des Resources Humaines,

57 (July–September): 18–41.

Anderson, J. C., & Frankle, A. W. 1980. Voluntary social reporting: An isobeta portfolio

analysis. Accounting Review, 55: 467–479.

Andersson, L. M. & Bateman, T. S. 2000. Individual environmental initiative: Championing

natural environmental issues in U. S. business organizations. Academy of Management

Journal, 43: 548–570.

Annan, K. 2001. The ‘unparalleled nightmare’ of AIDS. Speech to the US Chamber of

Commerce, June 1, 2001.

Arlow, P., & Gannon, M. J. 1982. Social responsiveness, corporate structure, and economic

performance. Academy of Management Review, 7: 235–241.

Asmundson, P., & Foerster, S. R. 2001. Socially responsible investing: Better for your soul or

your bottom line? Canadian Investment Review, 14: 20–34.

Aupperle, K. E., Carroll, A. B., & Hatfield, J. D. 1985. An empirical examination of the

relationship between corporate social responsibility and profitability. Academy of

Management Journal, 28: 446–463.

Avi-Yonah, R. 2005. The cyclical transformations of the corporate form: A historical perspective

on corporate social responsibility. Delaware Journal of Corporate Law, 30: 767–818.

61

Backhaus, D. B., Stone, B. A., & Heiner, K. 2002. Exploring the relationship between corporate

social performance and employer attractiveness. Business and Society, 41: 292–318.

Bagozzi, R. P, Epstein, M. J., & Wisner, P. S. 2001. Evidence of the influence of management

control systems on environmental and financial performance. Working paper, Arizona

State University Research Forum.

Bansal, P. 2003. From issues to actions: The importance of individual concerns and

organizational values in responding to natural environmental issues. Organization

Science, 14: 51–527.

Barnett, M. L. 2007. Stakeholder influence capacity and the variability of financial returns to

corporate social responsibility. Academy of Management Review, 32: 794–816.

Barnett, M. L., & Salomon, R. M. 2006. Beyond dichotomy: The curvilinear relationship

between social responsibility and financial performance. Strategic Management Journal,

27: 1101–1122.

Bartel, C. A. 2001. Social comparisons in boundary-spanning work: Effects of community

outreach on members' organizational identity and identification. Administrative Science

Quarterly, 46: 379–413.

Barth, M. E., & McNichols, M. F. 1995. Estimation and market valuation of environmental

liabilities relating to Superfund sites. Journal of Accounting Research, 32: 177–209.

Bauer, R., Derwall, J., & Otten, R. 2007. The Ethical Mutual Fund Performance Debate: New

Evidence from Canada. Journal of Business Ethics, 70: 111–124.

Bauer, R., Koedijk, K., & Otten, R. 2005. International evidence on ethical mutual fund

performance and investment style. Journal of Banking and Finance, 29: 1751–1767.

Baumeister, R. F., Stillwell, A. M., & Heatherton, T. F. 1994. Guilt: An interpersonal approach.

Psychological Bulletin, 115: 243–267.

Belkaoui, A. 1976. The impact of the disclosure of the environmental effects of organizational

behavior on the market. Financial Management, 5 (4): 26–31.

Belkaoui, A., & Karpik, P. G. 1989. Determinants of the corporate decision to disclose social

information. Accounting, Auditing and Accountability Journal, 2: 36–51.

Bello, Z. Y. 2005. Socially responsible investing and portfolio diversification. Journal of

Financial Research, 28: 41–57.

62

Berman, S. L., Wicks, A. C., Kotha, S. & Jones, T. M. 1999. Does stakeholder orientation

matter? The relationship between stakeholder management models and firm financial

performance. Academy of Management Journal, 42: 488–506.

Bies, R. J., Bartunek, J. M., Fort, T. I., & Zald, M. N. 2007. Corporations as social change

agents: Individual, interpersonal, institutional, and environmental dynamics. Academy of

Management Review, 32: 788–793.

Blacconiere, W. G., & Northcut, W. D. 1997. Environmental information and market reactions to

environmental legislation. Journal of Accounting, Auditing and Finance, 12: 149–178.

Blacconiere, W. G., & Patten, D. M. 1994. Environmental disclosures and regulatory costs, and

changes in firm value. Journal of Accounting and Economics, 18: 149–178.

Blackburn, V. L., Doran, M., & Shrader, C. B. 1994. Investigating the dimensions of social

responsibility and the consequences for corporate financial performance. Journal of

Managerial Issue, 6: 195–212.

Blair, M. M., & Stout, L. A. 2006. Specific investment and corporate law. European Business

Organization Law Review, 7: 473–500.

Bollen, N. P. B., & Cohen, M. A. 2004. Mutual fund attributes and investor behavior. Working

paper, Vanderbilt University.

Bosch, J. C., & Lee, I. 1994. Wealth effects of Food and Drug Administration (FDA) decisions.

Managerial and Decision Economics, 15: 588–599.

Bowman, E. H. 1978. Strategy, annual reports, and alchemy. California Management Review,

20 (3): 64–71.

Bowman, E. H., & Haire, M. 1975. A strategic posture toward corporate social responsibility.

California Management Review, 18 (2): 49–58.

Boyle, E. J., Higgins, M. M., & Rhee, S. G. 1997. Stock market reaction to ethical initiatives of

defense contractors: Theory and evidence. Critical Perspectives on Accounting, 8: 541–

561.

Bradley, M., Schipani, C. A., Sundaram, A. K., & Walsh, J. P. 1999. The purposes and

accountability of the corporation in contemporary society: Corporate governance at a

crossroads. Law and Contemporary Problems, 62(3): 9–85.

Bragdon, J. H., Jr., & Marlin, J. A. T. 1972. Is pollution profitable? Risk Management, 19 (4):

9–18.

63

Brammer, S., Brooks, S., & Pavelin S. 2006. Corporate social performance and stock returns:

UK evidence from disaggregate measures. Financial Management, Autumn, 97–116.

Brammer, S. & Millington, A. 2004. The development of corporate charitable contributions in

the UK: A stakeholder analysis. Journal of Management Studies, 41: 1411–1434.

Brown, B. 1997. Stock market valuation of reputation for corporate social performance.

Corporate Reputation Review, 1: 76–80.

Brown, B. 1998. Do stock market investors reward reputation for corporate social performance?

Corporate Reputation Review, 1: 271–282.

Brown, B., & Perry, S. 1994. Removing the financial performance halo from Fortune’s ‘most

admired’ companies. Academy of Management Journal, 37 (5): 1347–1359.

Brown, W. O., Helland, E., & Smith J. K. 2006. Corporate philanthropic practices. Journal of

Corporate Finance, 12: 855–877.

Buehler, V. M. & Shetty, Y. K. 1976. Managerial response to social responsibility challenge.

Academy of Management Journal, 19: 66–78.

Business Week 2005. The debate over doing good. August 15: 76–78.

Campbell, J. L. 2007. Why would corporations behave in socially responsible ways? An

institutional theory of corporate social responsibility. Academy of Management Review,

32: 946–967.

Campbell, J. L. 2006. Institutional analysis and the paradox of corporate social responsibility.

American Behavioral Scientist, 49: 925–938.

Carroll, A. B. 1979. A three-dimensional conceptual model of corporate social performance.

Academy of Management Review, 4: 497–505.

Carroll, A. B. 1999. Corporate social responsibility: Evolution of a definitional construct.

Business and Society, 38: 268–295

Chauvin, K. W., & Guthrie, J. P. 1994. Labor market reputation and the value of the firm.

Managerial and Decision Economics, 15: 543–552.

Chen, K. H., & Metcalf, R. W. 1980. The relationship between pollution control record and

financial indicators revisited. Accounting Review, 55: 168–177.

Clarkson, M. B. E. 1988. Corporate social performance in Canada, 1976–86. Research in

Corporate Social Performance and Policy, 10: 241–265.

64

Clarkson, M. B. E. 1995. A stakeholder framework for analyzing and evaluating corporate social

performance. Academy of Management Review, 20: 92–117.

Clemens, B. 2006. Economic incentives and small firms: Does it pay to be green? Journal of

Business Research, 59: 492–500.

Cochran, P. L., & Wood, R. A. 1984. Corporate social responsibility and financial performance.

Academy of Management Journal, 27: 42–56.

Comment, R., & Jarrell, G. 1995. Corporate focus and stock returns. Journal of Financial

Economics, 37: 67–87.

Conine, T. E., Jr., & Madden, G. P. 1986. Corporate social responsibility and investment value:

The expectational relationship. In W.K. Guth (ed.), Handbook of Business Strategy:

1986/1987 Yearbook:18:1–18:9. Boston: Warrren Gorham and Lamont

Cooper, S. 2004. Corporate Social Performance: A Stakeholder Approach. Burlington, VT:

Ashgate.

Cowen, S. S., Ferreri, L. B., & Parker, L. D. 1987. The impact of corporate characteristics on

social responsibility disclosure: A typology and frequency-based analysis. Accounting,

Organizations and Society, 12: 111–222.

D’Antonio, L., Johnsen, T., & Hutton, R. B. 1997. Expanding socially screened portfolios: An

attribution analysis of bond performance. Journal of Investing, Winter: 79–86.

D’Aveni, R. A. 19943. Hypercompetition: Managing the Dynamics of Strategic Maneuvering.

New York: The Free Press.

Davidson, W. N., & Worrell, D. L. 1988. The impact of announcements of corporate illegalities

on shareholder returns. Academy of Management Journal, 31 (1): 195–200.

Davidson, W. N., & Worrell, D. L. 1992. The effect of product recall announcements on

shareholder wealth. Strategic Management Journal, 3 (6): 467–473.

Davidson, W. N., Worrell, D., & Cheng, L. T. W. 1994. The effectiveness of OSHA penalties: A

stock market-based test. Industrial Relations, 33, 283–296.

Delery, J. E. and Doty, H. 1996. Modes of theorizing in strategic human resource management:

Tests of universalistic, contingency, and configurational performance predictions.

Academy of Management Journal, 39: 802–835.

Derwall, J., & Koedijk, K. 2005. The performance of socially responsible bond funds. Working

paper, Erasmus University, Rotterdam, Netherlands.

65

Derwall, J., Gunster, N., Bauer, R., & Koedijk, K. 2005. The eco-efficiency premium puzzle.

Financial Analysts Journal, 61: 51–63.

DiBartolomeo, D., & Kurtz, L. 1999. Managing risk exposures of socially screened portfolios.

Technical report, Northfield Information Services.

Diltz, D. J. 1995. The private cost of socially responsible investing. Applied Financial

Economics, 5: 69–77.

Dooley, R. S., & Lerner, L. D. 1994. Pollution, profits, and stake-holders: The constraining

effect of economic performance on CEO concern with stakeholder expectations. Journal

of Business Ethics, 13: 701–711.

Dowell, G., Hart, S., & Yeung, B. 2000. Do corporate global environmental standards create or

destroy market value? Management Science, 46: 1059–1074.

Esrock, S. L. & Leichty, G. B. 2000. Organization of corporate web pages: Publics and

functions. Public Relations Review, 26: 327–344.

Fogler, H. R., & Nutt, F. 1975. A note on social responsibility and stock valuation. Academy of

Management Journal, 18: 155–160.

Fombrun, C., & Shanley, M. 1990. What’s in a name? Reputation building and corporate

strategy. Academy of Management Journal, 33: 233–258.

Frank, R. 2007. Richistan: A Journey through the American Wealth Boom and the Lives of

the New Rich. New York: Crown Publishers.

Frederick, W. C. 2006. Corporation, Be Good! The Story of Corporate Social Responsibility.

Indianapolis, IN: Dog Ear Publishing.

Freeman, R. E., Wicks, A. C., & Parmar, B. 2004. Stakeholder theory and “The Corporate

Objective Revisited.” Organization Science, 15: 364–369.

Freedman, M., & Jaggi, B. 1982. Pollution disclosures, pollution performance and economic

performance. Omega, 10: 167–176.

Freedman, M., & Jaggi, B. 1986. An analysis of the impact of corporate pollution disclosures

included in annual financial statements on investors’ decisions. In M. Neimark (ed.),

Advances in Public Interest Accounting, 1: 193–212. London: Elsevier.

Freedman, M., & Stagliano, A. J. 1991. Differences in social-cost disclosures: A market test of

investor reactions. Accounting, Auditing and Accountability Journal, 4: 68–83.

66

Friedman, M. 1970. The social responsibility of business is to increase its profits. New York

Times Magazine, September 13: 32–33, 122, 124, 126.

Frooman, J. 1997. Socially irresponsible and illegal behavior and shareholder wealth: A meta-

analysis of event studies. Business and Society, 36: 221–249.

Fry, F. L., & Hock

Does it Pay to Be Good...And Does it Matter? A Meta-Analysis of the Relationship between Corporate Social and Financial Performance

Abstract and Figures