Available in Adobe Acrobat PDF format

According to its advocates, social investing (also known as socially responsible investing or ethical investing) is a fast growing phenomenon that represents "nearly one out of eight dollars under professional management in the United States." Its central tenet is that clients can "invest for their own futures and a better world at the same time" by buying stock in companies that pass social screens and avoiding investments in companies that do not. According to this general approach, publicly-traded companies can be "objectively" rated using data gathered by social investing researchers, most notably by Kinder, Lydenberg, & Domini (KLD). Some academicians incorporate these ratings into their research as measures of Corporate Social Performance (CSP). In contrast, this study contends that social investing principles are problematic, and the data and ratings generated by social investment researchers are hopelessly flawed. Social investment advocates rely on sketchy, highly selective research and pseudo-objective ratings that belie the complexity of modern corporations and economies. Social screening in general and KLD's ratings in particular are tainted by anachronistic, contradictory, idiosyncratic, and ideologically constructed notions of corporate social responsibility. Representations of the growing financial impact and competitive performance of social investing are questionable. No social research organization or "socially responsible" mutual fund has yet presented a coherent case for why its criteria are ethical or socially responsible or better at effecting social change. This study concludes that the general approach of SI advocates, including by academic researchers, is one of vindication of the true believer, not investigation.

The Historical Context: The Legacy of Religious-Inspired Social Screens

Social screening has its primary origins in conservative religious principles. The first known consumer screens date to the 18th century, when the Quakers withdrew their business from companies involved in alcohol, tobacco or gambling--so-called sinful behavior. The Quakers also incorporated what is believed to be the first issue-specific screen resulting in boycotts of companies tied to the slave trade. (For historical overview, see Domini, 2001; Waddock 2000; Hutton, D'Antonio, Johnsen, 1998). In what is believed to be the first externally screened US investment, the Pioneer Fund in 1928 incorporated these negative or sin screens, excluding companies involved in tobacco or alcohol. (See SocialFunds.com.)

The social activism of the 1960s spurred the development of additional negative screens based on overtly ideological and political sentiments. In 1968, a pension fund in Boston asked a young securities analyst, Alice Tepper Marlin, to compile a "peace portfolio" of corporations with the least involvement in supplying armaments for the war in Vietnam. Hundreds of church and community groups asked for her report. The Interfaith Center on Corporate Responsibility composed of hundreds of religious members from varying faiths was formed in 1971. That same year, a Methodist group launched the Pax World Fund, which included a negative screen on military contracting along with alcohol and gambling.

The 1970s and 80s witnessed the blossoming of a concurrent movement known as green or ethical consumerism. Self-appointed watchdog organizations such as the Council on Economic Priorities (CEP), founded by Tepper Marlin, began rating companies on a variety of hot-button social issues including animal rights, nuclear energy, and gay and feminist issues. In the spirit of the times, many activist consumer groups criticized "multinational" corporations often using the pejorative "Corporate America." (In contrast, entrepreneurial businesses such as Ben & Jerry's and The Body Shop, although beset by ethical and operational problems, were often romanticized as "socially responsible", in part because their founders mouthed anti-corporate rhetoric (Entine, 1995b).)

The stated goals of most green consumer groups was not to encourage stakeholder-centric behavior as generally understood by academicians and practitioners but to promote pet social causes and "shop for a better world" (as reflected in the CEP guidebook series "Shopping for a Better World"). Many of social investing's most prominent researchers, including Steven Lydenberg, a principal at Kinder, Lydenberg, & Domini (KLD), cut their eyeteeth as researchers at ideologically-charged CEP.

These two essentially anti-corporate currents--socially conservative sin notions promoted by religious-based mutual funds and the vaguely liberal, consumerist brand of Sixties activist ideology--coalesced during the 1980s to form the core of what is today called social investing. The catalyzing event was the boycott of apartheid South Africa, which further politicized corporate critics and tapped into the sentiments of newly affluent baby boomers sympathetic to a green romantic, anti-corporate pitch. Portfolio management firms such as Franklin Research and Development (now Trillium Asset Management) and social investment companies such as Calvert were formed. By the late 80s, socio-religious screens conflated with populist but shifting activist concerns to form the hodgepodge that today constitutes the principles of the liberal wing of social investing.

Fungible Standards of Social Investing

Social investing research has attracted widespread interest among academicians who have been struggling for years to find reliable and objective ways to measure corporate social responsibility. All social research rating systems rely to some degree on perceptual measures, often translated into numerical rankings. In part because it evaluates thousands of companies including all of the corporations in the S&P 500, the database launched by KLD in 1990 has emerged as a popular and favorite academic benchmark.

KLD originally established eight categories, later expanded to ten, constructed around negative or "exclusionary" screens: nuclear power, alcohol, gambling, tobacco, and military contracting, with other negative activities such as insensitivity to gays swept into an "other" category. Companies that fail this initial sin screen are branded as not socially responsible and summarily excluded from various KLD social indices. After sweeping for negative concerns, KLD then evaluates companies in qualitative areas such as community relations, workforce diversity, employee relations, environment, non-US operations, and product safety. Using a process of data collection, analysis, and ranking that it does not explain in detail, one or more KLD staffers assign numerical ratings for each company in each category, which become the basis for its hierarchical rankings. The highest rated companies are included in one or another index, including the Domini Social Index 400. Launched in May 1990, the DSI was the first benchmark for equity portfolios subject to multiple social screens and today remains the most prominent.

Many corporate social performance researchers (Rawley and Berman, 2000; Hutton, D'Antonio and Johnsen, 1998; Griffin and Mahon, 1997; Woods and Jones, 1995) including its most fervent advocates (Kurtz, 2000; Waddock and Graves, 1997a, Waddock and Graves, 1997b) have pointed out the methodological and conceptual problems of linking CSP research to social investment data.1 Although there is general agreement that CSP is a multidimensional construct, there are no agreed upon standard or theoretical rationale or way to aggregate and therefore compare multiple dimensions across or within industries. Berman and Wicks (1998) go so far as to argue that CSP is not a theoretically viable construct," noting that it is a "complex collection of factors that do not maintain the same meaning across contexts" and "must be defined according to social context" (p. 407). Compounding such concerns, researchers often subjectively pick and choose which categories suit their personal notions of CSP and adjust the data using idiosyncratic formulae (Graves and Waddock, 1994; Turban and Greening, 1997).

Still, many academic researchers rely on this data. Among the characterizations: KLD has developed a "consistent, largely objective, set of screening criteria" (Waddock and Graves, 1994, p. 1038); "the database . . . objectively rates firms" (Johnson and Greening, 1999, p. 566); and the rating system is "carefully constructed" and reflects "mainstream social concerns" (Sauer, 1997, p. 141). Even some social investing skeptics have referred to the data as "largely objective" (Griffin and Mahon, 1997, p. 10).

Just how objective and carefully constructed are these ratings? In an astonishing assertion Waddock and Graves (1997b) have written that "although the limitations of KLD's data need to be recognized, these data do represent the only currently-available externally-based assessment of the S&P 500 corporations over a consistent range of stakeholder arenas, over time, by a set of observers with no immediate 'stake' in any given company." The fact that KLD's ratings are the "only currently-available" data begs the critical question: is the data reliable and a fair measure of CSP? The quality of conclusions depends upon the credibility of the data. What are the standards of data collection and analysis? What conceptual and operational definitions are being used? Do social investment data provide a reasonable measure of corporate social responsibility?

Methodological Issues

CSP research data suffers from a lack of reliability. There are fundamental questions about the quality of the research data. KLD and other research groups review thousands of companies with skimpy resources. Research relies upon often unreliable, anecdotal, and highly interpretable data. Overworked and undertrained junior staffers draw on government data banks, journalistic sources, and information supplied by companies, collating whatever information they select as relevant. The task is daunting and highly subjective. This data are then given a patina of objectively by being turned into hierarchical numbers.

The Odwalla controversy offers a notable anecdotal example of the problematic nature of such data. By 1996, Odwalla was considered a model of social responsibility. Researchers at Adams, Harkness & Hill extolled the company for "respecting the fruit" and for its "intense caring [which] translate[s] into a superior product. . . .We believe that no other juice company takes these steps to ensure the quality of its products." (Patsky, Galle and Negron, 1996, p. 48) At the time, KLD gave it glowing ratings and as a result Domini included Odwalla in its index. Amy Domini, a founder of KLD and president of Domini Investments, orchestrated the appearance of Odwalla founder Greg Steltenpohl at that year's Social Investment Forum conference where she lavishly praised him and his company.

One month after the conference, one child died and at least seventy people were injured after drinking Odwalla apple juice tainted by poisonous bacteria. Investigations by the New York Times and documents introduced in various court cases found Odwalla criminally negligent and derelict for ignoring a known pattern of quality, safety, and health problems, and culpable for withholding that information from the public (Drew and Belluck, 1998). Even after these revelations, Domini showed her emotional attachment to her friend's company when she wrote: "Odwalla has been compared to Johnson and Johnson (Tylenol) in rapid recall and voluntary assumption of blame, and the reporters give management high marks for action" (Domini, 1998). After more embarrassing revelations and adverse court rulings, Steltenpohl was forced to resign and the company, in financial disarray, was acquired (Entine, 1999). As it turns out, Adams, Harkness & Hill and KLD used highly selective data and based their conclusions primarily on company representations. Even after details of its criminal conduct became known, Domini and KLD's Kinder declined to revise their high ratings of their friend's company.

The Body Shop case offers another dramatic example of the methodological problems besetting CSP research. In the early 1990s, every social investment company including Franklin Research and Development gave the British cosmetics company high grades. However, after Business Ethics magazine published an article that raised sustentative concerns about its performance (Entine, 1994), Franklin did more research (Bavaria, Becker and Billenness, 1994), lowered its ratings across the board, and divested itself of its Body Shop shares. As Franklin's research director noted, "We learned a hard lesson: never take a company's representations at face value. This is not a science but an art."2 The Odwalla and Body Shop affairs underscore the personal and ideological biases that challenge the credibility of CSP research.

CSP research uses arbitrary standards. Can social performance attributes be measured and compared, from company to company and across industries, in a meaningful way? The use of exclusionary screens, which otherwise seems straightforward, highlights the subjective and often arbitrary nature of such ratings (Hutton, D'Antonio, Johnsen, 1998). Each fund sets cut-off points to determine whether a company is excluded. For example, Washington Mutual Fund, with $40 billion in assets, screens out companies that derive 50 percent or more of revenue from tobacco or alcohol. Why 50 percent? Beats them. Ethical Funds, a Canadian social research firm, screens out companies that make packaging for the tobacco industry--unless tobacco-related production represents less than 20 percent of business. How does EF verify that figure? It can't. Why a 20 percent cutoff? "Well, that's just the percentage we set," said its founder, David Shuttleworth. (Entine, 1995a, p. 45) KLD screens out companies that derive more than 2 percent of revenues from weapons sales; derive 15 percent of their revenue from adult entertainment; own 20 percent of another company with abortion involvement; or are 50 percent owned by a company with alcohol involvement. Putting aside the prickly issue of how it determines such things, what's the meaningfulness of these arbitrary percentages?

CSP research ignores aspects of corporate activity that are not easily measurable and is a priori biased against some industries that are more transparent. In practice, all current social screens eschew complexity. While banks do not release details on their portfolios and the technology industry is notoriously secretive, basic industries churn out reams of environmental data. That makes them easy targets for simplistic screens. Social research rating systems that do not use a "best in sector" approach (KLD does not) severely limit as a matter of principle investments in chemicals, natural resource, energy, and mining--companies and industries with large, negative environmental impacts.

Most research does not extend beyond the first-level of corporate activity, which can result in a distorted picture of a company's practices. For example, bank portfolios are not screened for relationships with companies involved in tobacco, nuclear power, or weapons production. The loan lists of banks are littered with investments in companies and industries that are excluded by liberal-oriented screens, but are not identified. Such superficial standards do confer a decided advantage to social funds: they conveniently allow social investors to screen in fast growth stocks of banks, financial institutions, communications firms, and technology companies even though they all do a high percentage of their business with the military and other sin-related businesses. Researchers do not examine layered business models (such as franchising) or the kind of complex business structures increasingly common in modern corporations (such as the dense corporate structure of Enron). Simplistic screens often miss egregious coporate behavior. Consider Microsoft, by far the favorite stock in social portfolios and 6 percent of the Domini Social Index. It has been found guilty of antitrust violations by both the federal government and the European Union and remains embroiled in other consumer antitrust suits. Yet it consistently gets high marks from social researchers because antitrust violations and their impact on stakeholders, most notably consumers, are not screened.

First-level screens make a mockery of the complexity of modern corporations and underscore the methodological limitations of niche CSP research. In their defense, social funds say they don't have the sophistication to go beyond a superficial analysis. "It's just too complicated, with too many variables," admits Jerry Dodson, who heads Parnassus, a leading social fund that nonetheless relies on incomplete data to make its investment choices.3 Social researchers pleaded poverty and limited research capabilities when it turned out that many of the recently disgraced companies including Anderson, Enron, WorldCom, Adelphia, Tyco, and Tenet Healthcare were favorites of social funds.

The numerical ratings used by CSP researchers create an illusion of objectivity. Consumer activists such as the Council on Economic Priorities pioneered the rating systems used by social researchers. In its "Shopping for a Better World" guide, CEP graded companies like school children, from A-to-E on a range of social issues. For instance, CEP awarded a company an A in the category "Women" or "Women's Advancement" if it had 10 percent or more female representation on the board of directors and lesser grades for lower representation. Did CEP examine the quality or contributions of board members before assigning grades? No. As with many of its ratings, it was purely a litmus test.

KLD adopted this litmus test formula. So, for instance, KLD gives high marks to companies in which women, minorities, or the disabled hold four or more board seats. What is the empirical basis for such criteria? There is none and none is offered. Other categories are even more subjective. Companies get high grades for implementing "innovative hiring programs" involving the disabled. What's innovative? Who crunches subjective perceptions to come up with the hard numbers? Behind KLD's ratings is a socially responsible Svengali-either an individual or group-who assigns numerical ratings. These numbers represent the perceptions of the ratings makers, colored by personal and ideological biases. The decision about what data to collect and the translation of this information into numbers render social investment data subjective and unreliable. Although numbers do not by themselves confer objectivity or meaningfulness, social researchers often cite these grades as if they represent an objective judgment and incorporate them in elaborate analysis.

Conceptual Issues

As has been noted, some academic researchers mimic the propaganda line of social investing advocates who equate aspects of social investing with CSP. This is problematic. While corporate social performance reflects a substantive definition of behavior that assumes that some choices are more ethical or socially responsible than others, social investing is burdened by a historical reliance on a client-centered procedural definition that requires only that the investor believes he is acting ethically (Mackenzie, 1997).

Social investing advocates have never made the case that buying and selling stocks based on social criteria promotes reform or is in any way more ethical or socially responsible than mainstream investing. According to cant, clients can "invest for their own futures and a better world at the same time" (Gravitz, 2002, p. 1). However, in postmodernist fashion, there are no agreed upon standards about what constitutes a "better world" or which companies are more ethical and responsible. Each individual gets to decide for herself what is a "better world" and which companies are deemed ethical. There are literally hundreds of funds and investment strategies with different ideological colorings and varying definitions of socially responsible and ethical corporate behavior. Social investing principles run the gamut from ultra liberal to hard conservative, from pacifist to militarist. Perception not ethical performance drives social investing. "Our common goal," stated Joan Bavaria of Trillium at a SIF gathering in 1997, "is to give people the information they need to make their own determinations about what they want to invest in."4 According to Cliff Feigenbaum, editor of Greenmoney Journal, the movement seeks to "match your capital to the companies that best represent your moral and ethical values." Social investing is based on investor feelings - the "heart" according to social investor advocates who talk about aligning investments with one's beliefs - not explicitly on company behavior (Feigenbaum, 2002; Domini, 2001).

As a result of this definitional ambiguity, social investing has devolved into an exercise in tactics and liberal sentimentality, focusing on litmus screens and pandering to perceptions, rather than sorting estimable or questionable corporate behavior. Favored investments--consumer products, media, health care, and business services--reflect a consumerist "shopping for a better world" esthete that people's purchasing habits define strong corporate social performance.

This confusion is reflected in the contradictory description of social investing provided by the SIF, the industry trade and advocacy group, in its trend reports. Under the heading "social investing defined," the SIF claims "social investment requires investment managers to overlay a qualitative analysis of corporate policies, practices, and impacts onto the traditional quantitative analysis of profit potential." In other words, the SIF conflates social investing, which is built around exclusionary and social litmus screens, with the very different stakeholder-centric notion of corporate social responsibility. It reinforces this fungible definition of social investing by quoting the Prince of Wales Business Leader Forum to the effect that "Corporate Social Responsibility means open and transparent business practices" (SIF Report, 2001, p. 8).

That is not the definition of social responsibility, however, used by the most prominent CSP research organizations. As practiced, social investing underplays or ignores most stakeholder practices. Most social investors focus only on a few sin issues. More than 40percent of socially responsible mutual fund dollars are invested through one fund manager, American, which is not a SIF member. America's two social funds do not screen for stakeholder or corporate governance issues. Rather, they use a very soft screen restricting investments in companies with a majority of their business in tobacco and alcohol.

Not one of the SIF's 73-member funds screen for the array of corporate practices touted by the SIF in its trend report. KLD president Peter Kinder has a long-standing personal bias against corporate governance related screens. "Focusing on non-issues such as independent boards, transparency and the like makes it easy to avoid taking stands on real issues of corporate accountability," he has written. Kinder claims that social issues are far better measures of corporate responsibility (Kinder, 1997). While KLD underplays corporate practice screens, Calvert historically ignored them altogether until recently. After the recent spate of corporate scandals, Julie Gorte of Calvert was asked why Calvert had never screened for accounting or governance issues. "This is a meteor," she said, acknowledging that Calvert had long ignored a fundamental measure of social responsibility. "We're still measuring the depth of the crater" (Entine, 2002). Although it has since introduced limited corporate practice screens, Calvert and the rest of the industry pay short shrift to the range of CSP standards that the SIF report claims are the centerpiece of social investing.

Social investors favor litmus tests on trendy social issues with an outsized reliance on sin screens. Consider the most popular screen on weaponry. Kinder hyperbolically refers to defense contractors as "merchants of death." Many social researchers and academics are so convinced that defense production is not ethical that they do not even attempt to justify this sentiment. "In the view of social investors," wrote KLD's Steven Lydenberg and Karen Paul of Florida International University in a paper and presentation at an annual meeting of the International Association of Business and Society, "full scale nuclear way poses a risk to the survival of life on Earth. Companies that manufacture nuclear weapons increase that risk" (Paul and Lydenberg, 1997). Sandra Waddock and co-author Neil Smith toed this ideological line when they ranted against ". . . individually harmful products (such as cigarettes, nuclear equipment, or military army) . . ." (Waddock and Smith, 2000, p. 78).

The taboo screen on armaments sidesteps the far more important social policy issue: what are wasteful or unnecessarily provocative defense expenditures. There is certainly no evidence that a reflexive ban on military production promotes the presumably desirable goal of preventing war. History suggests that absolutist pacifist ideology can have horrific unintended consequences and often leave countries more vulnerable. In the 1930s, ultra-pacifists and Nazi sympathizers formed an alliance in an attempt to keep the U.S. out of World War II. Holocaust survivors or supporters of Israel might find a screen on military production in the name of ethics offensive. It is certainly arguable that military production can act as a deterrent to aggression. Some historians believe that the military escalation undertaken by the US in the 1980s was a key factor in bankrupting the Soviet empire and hastening the end of the Cold War. It's particularly ironic to note that according to public opinion surveys, the liberal community (and presumably liberal-minded social investors and academicians) has overwhelmingly supported two recent wars: the overthrow of Slobodan Milosevic's repressive regime to stop genocide and bring Western values to the Balkans, and the military response in Afghanistan in the wake of 9/11, both on humanitarian grounds. It's dishonest, hypocritical and logically and morally indefensible for social investors and academic researchers to personally back select humanitarian military interventions while simultaneously endorsing a reflexive exclusionary screen on all military production.

Social screens are based on idiosyncratic biases that differ dramatically in conceit and implementation. As a result of the procedural bias of social investing, there is no unanimity about what values constitute social investing. Some social investors and funds focus entirely on one or two issues such as animal rights or tobacco. Many social investors and funds are faith-based, drawing on varying Protestant, Catholic, Islamic or Jewish beliefs. Clearly, a religious fundamentalist in Iowa does not share many social and political values with a gay activist in New York. Some funds filter out companies that profess support for abortion rights and others screen-in these same companies. The conservative Timothy Plan excludes Disney, a favorite target of right wing activists, for its allegedly lax moral standards. In the past, Disney has been a favorite investment of more liberal social investment funds such as KLD. One person's taboo is another person's sacred cow.

Compounding these differences, social researchers with similar ideologies often apply different conceits or methodologies. As a result, researchers screening for the same issue and with the same purported social goal may sometimes reach differing conclusions. For example, while Boston-based US Trust was praising Gillette in the early 1990s for being "extremely proactive in disclosing...its animal testing policies," concluding that it is "quiet about its corporate social responsibility achievements and forthcoming in areas of social controversy," Gillette was screened out of numerous funds including Domini Society Equity because it engaged in animal testing (Entine, 1997).

KLD, as other CSP research organizations, offers little explanation as to why certain categories are selected and others excluded. Indefinable "social justice" issues take precedence over classic stakeholders. Investors, vendors and franchisees are given second-class citizenship in comparison to ambiguous ideals like "peace" and "fighting oppressive regimes." Companies that create high-paying jobs, for example, get few of the social credits of companies that pass flowery social mission or ethic statements-like Enron. Oddly, there is little focus on a company's responsiveness to its customers, arguably its most important stakeholder. Although KLD developed a category called "product quality," the use of that screen as a surrogate for customer relations by Waddock and Graves (2003) is dubious at best.

By elevating trendy social issues, mission statements, and sin screens over broader stakeholder measures, liberal social investors often ignore behavior with significant social and financial implications. As an example, Levi Strauss, the jeans and clothing company, basked in a tremendous amount of good press in the 1980s and 90s because of its ethical pronouncements on a variety of trendy issues. Yet it made increasingly shoddy products, treated piece workers poorly, and eventually saw its CEO mismanage the company into the ground. It was ultimately forced to retrench, shutter its American facilities and flee to low-wage havens. Thousands of its employees no longer have jobs and the company still has not recovered. In essence, Levi Strauss' high-minded ethical stands--the actual operating centerpiece of liberal social investing--proved worthless in real terms.

The arbitrary nature of the categories used by social researchers can be illustrated by comparing the dominant liberal social investing model with the rating system developed for black trade unions in South Africa in the early 1990s. Consultants Richard Adams and Jayanti Durai surveyed members about how to invest their pension fund, which then totaled $175 million (Adams and Durai, 1995). Rather than relying on the socio-religious criteria popular in North America and the UK, the workers devised their own guidelines. Litmus-test social sin issues such as armaments manufacturing and alcohol production were scrapped. Union members focused on product quality, working conditions, benefits, safety, and equal opportunity as weighted measures of CSP. Worker-related issues represented more than 75 percent of the screening criteria.

South Africa's Community Growth Fund

(In order of importance based on 100 points)
Job creation policy  14
Industrial relations      14
Conditions of employment     13
Training  7
Equal opportunity for women     7
Health and safety       6
Type of product       5
Privatization       5
Profit retention       5
Affirmative action      5
Location (share of foreign assets)    4
Environmental management     4
Disclosure  4
Worker participating  4
Political profile       2
Social spending  1

Unlike the liberal model with its procedural focus on perception and its reliance on exclusionary screens, the workers' model focuses on behavior and stakeholder rights.

The Financial Footprint of Social Investing

Every two years since 1995, the Social Investment Forum (KLD and its sibling Domini Investments, Calvert Group, Parnassus, Citizens and other prominent liberal-oriented social investing groups and mutual funds are key members) has issued a report on industry trends including estimates of the assets of investors who use social screens and shareholder advocacy. According to the SIF's most recent report, as of June 30, 2001 "nearly one out of eight dollars under professional management in the United States is involved in socially responsible investing"; $2.03 trillion, 12 percent of the $19.9 trillion in investment assets under management, are socially screened; these assets "grew 1.5 times faster than all U.S. managed portfolio assets" over the previous two years; and "socially screened separate accounts grew by nearly 40 percent" (SIF Report 2001, pp.4-5). Citing the spate of corporate scandals, Citizens Advisor, Inc. founder Sophia Collier claimed that "socially responsible investing is about the come into fashion," predicting it will double in three to five years (Ackermann, 2002).

Academic researchers and journalists credulously report these purported trends, often reproducing SIF statistics and newsreleases almost verbatim (Waddock, Bodwell and Graves, 2002; Sauer, 1997). These are certainly astounding claims, worthy of academic study. But are they accurate and in context? How does the SIF define social responsibility and thereby determine which assets are considered socially responsible?

The SIF counts all assets that by its definition are "screened, involved in shareholder advocacy, or are directed to community investing." Under the heading "What Was Counted," the study notes that "an institution was considered to engage in socially responsible investing" if it "utilizes one or more social screens" or if the institution sponsors or co-sponsors shareholder resolutions on issues the SIF considers socially responsible [emphasis added]--even if the client or investment manager does not consider those investments social investments (SIF Report, p. 31-32). What does this mean in practice? Any institution, investment manager, or investor that screens on almost any issue as part of a "formal" policy is by SIF definition engaging in social investing. This definition is problematic. Consider the recent proliferation of governance screens on such things as accounting issues, board makeup, and pension fund liabilities. Under SIF criteria, a pension fund that screens for a majority independent board or the compensation of executives has that investment counted as a "social investment." Yet almost every responsible investment advisor now reviews corporate governance issues as a matter of course. According to SIF's boundary-less policy, their investments could be counted as social investments--if the screening is deemed formal. With such ambiguous standards, there are almost no barriers to entry into the SIF's social investment club.

There are also glaring computational mistakes in the SIF report. It stated that "assets in socially screened mutual funds stayed steady from 1999 to 2001, despite a substantial market downturn," and set the total at $153 billion (2001 SIF Report, p. 12). That is inaccurate. Socially screened mutual fund investments totaled only $144.5 billion. SIF misstated the Calvert Large Cap Growth Fund by a factor of ten - an $8.5 billion error.5 Using the corrected figure, the 1percent drop in mutual fund social investments turns out to be a 6.2percent fall off. According to the Investment Company Institute, the mutual fund market grew by 14.36 percent over that over two year period. So while the amount of money being invested overall in mutual funds was soaring, dollars were being drained from the most prominent social investment funds. It is disconcerting to note that these erroneous data have already been incorporated in academic research papers, including the winner of the 2002 SIF Moskowitz Prize, "Evidence on Ethical Mutual Fund Performance and Investment Style" (Bauer, Otten and Koadijk, 2002), which was also published by the Journal of Investing.

How large is the financial footprint of the liberal wing of the social investment movement, which garners most of the media attention? If the mutual fund market is the measure, it is tiny and shrinking. Of the 73 SIF member funds, only 16 broadly screen for social issues and hold assets over $100 million, accounting for less than $8 billion. The 73 liberal-minded funds in total hold only $10.11 billion in total investments, or 0.16percent of the $6,555.1 billion mutual fund market.

Since the puncture of the market and technology bubble, tech-weighted funds including most liberal focused social funds that placed outsized sector bets on technology have cratered and continue to underperform. While there has been an overall rise in mutual fund investments of 2.5percent over the two-year period, Domini Social Equity, the largest SI fund with $1.296 billion in assets as of December 31, 2002 has shrunk by 22percent; Pax World has dropped 14percent; and Calvert, the third largest fund, is off 25percent. Not one of the largest social funds tracked by the SIF showed an increase, with the overall drop approximately 20percent.

We are faced with the apparent anomaly that while social investing advocates claim dramatic growth, investments in socially focused mutual funds are contracting relative to the overall size of the mutual fund pie. It seems counterintuitive to claim that interest in social investing by individuals is soaring while social investment dollars in mutual funds, which are generally believed to reflect individual investment interests rather than institutional sentiment, now represent the tiniest fraction of fund investments.

Market Performance of Social Investments

Do companies rated as more socially responsible perform better financially? Do socially responsible stocks outperform those that are lesser rated? It is not surprising that social fund advocates would be tempted to selectively present performance data to make their case that social funds "outperform" mainstream investment styles (Camejo, 2002). After all, they are true believers, and often their livelihood depends on reinforcing this propaganda point. It is disconcerting, however, to find that so many supposedly independent academic researchers echo this view, sometimes without critically reviewing the performance data and the premises on which it is based.

Consider John Guerard, Jr.'s article "Is Socially Responsible Investing Too Costly". He asserts "there is a growing literature in academic and professional investment journals that suggests socially responsible investing might produce higher risk-adjusted portfolio returns than merely using all available stocks in the equity universe" (Guerard, Jr., 1997, p. 26). However, he then goes on to contradict this polemical conclusion, citing studies that mostly show either no statistical difference or relative underperformance by social funds. Similarly, Waddock, Bodwell and Graves (2002) have written that "significant evidence from a large and growing body of academic research suggests at a minimum a neutral, and quite likely a positive, relationship between responsible corporate practices and corporate financial performance." Yet, those who have reviewed the performance analysis literature, including Waddock, find no such consistent evidence that socially screened companies and funds outperform (Waddock, 2000; Sauer, 1997; Griffin and Mahon, 1997).

The few studies that purport to show either a neutral or positive social investing or financial performance effect tout the KLD data as a reliable measure of CSP practices (see Waddock and Graves, 2003; Waddock and Graves, 1997a; Waddock and Graves, 1997b; Guerard, Jr., 1997; Kurtz, 1997). Also, many studies and popular articles have favorably compared the performance of the KLD-linked Domini Social Index (DSI) or the Domini Social Equity Fund with the S&P 500 (Feigenbaum 2002; Guerard, Jr. and Stone, 2002; Waddock, 2002; Sauer, 1997; Young, 1996). There are significant problems with this assumption. Although the DSI counts as its benchmark the S&P 500, it is not a passive large cap index fund and is not optimized to any broader benchmark. The DSI only includes about half of the top 500 stocks and weights them as it chooses and arbitrarily. While some academic researchers incorrectly state that Domini maintains a "passive fund" (see "How We Did the Study" in Waddock and Graves, 2003, p. 8), like all social investing funds, it is actively managed.

Researchers also often overlook that there are a variety of factors that can convincingly explain the competitive performance of social investments, most notably: (1) the concentration of investments in a narrow range of sectors; and (2) the time frames that these researchers arbitrarily select to compare social investment funds with mainstream funds. The KLD database (and as a result the DSI) is biased towards larger growth companies concentrated in a few key sectors-technology and financials most prominently. While the S&P has 20.5percent of its holdings in financials, KLD/DSI hold more than 25percent; S&P holds 14.3percent in technology stocks while KLD/DSI hold 18.75percent6 (and as much as a third of its portfolio in recent years). KLD/DSI underweight as a matter of principle infrastructure, manufacturing and energy companies claiming these are more likely to have adverse environmental impacts.

Almost every academic study that purports to show a neutral or positive correlation pre-dates the current bear market during which large growth stocks and technology companies outperformed. For example, Guerard, Jr. has used various time frames between 1984-1997 (Guerard, Jr., 1997; Guerard, Jr. and Stone, 2002). Waddock's social investment study covered a ten-year period from 1986-1996 (Waddock, 2000). The 2002 Moskowitz Prize paper by a team of Dutch researchers incorporates data through 2000 (Bauer, Otten and Koadijk, 2002). These time frames are never theoretically defended; in fact, they present a very distorted picture of social performance. Even a minor change in dates selected for market analysis can singificantly alter the results. For example, in the early 1990s, when sin stocks including military companies outperformed, a sin portfolio yielded a significantly higher return than screened funds (Naber, 2001). Later in the decade, KLD and other social funds prospered with sizable bets in financials, technology, health care and communications.

The period beginning in the early 1980s and ending in 2000 corresponds to a historically unprecedented extended bull market in which growth companies and sector investments favored by social funds outperformed. It is no wonder that tech-heavy social funds performed competitively during this period; according to Morningstar, with the bubble at its largest during 2000, more than 90percent of tech mutual funds rated five stars (McDonald, 2003). During the 1970s and early 1980s large company growth stocks including technology companies that are prominent in the DSI performed relatively poorly for an extended period. Neither the DSI, nor academic studies using DSI data, have backtested for this period to fairly evaluate performance in varying investment climates.

The mediocre performance by social funds during the current bear market, when many of social investing's favorite sectors were out of favor, underscores the limited value of academic studies using narrow and arbitrary time periods. The Domini Social Equity Fund (DSEF), which is based on KLD's ratings, has been a notable recent laggard. It rates a below average "2" in Morningstar's Ratings and a "C" over 3 years by Lipper.7 It trails the S&P 500 index funds that it claims in its prospectus to track over the 3-and 5-year periods. However, comparing it to the S&P 500, which many social investing advocates and academic researchers choose to do, understates its even poorer performance relative to its peers. According to the Domini Fund's web page, the DSEF is a large blend (Morningstar) or large-cap core (Lipper) fund. Against these indices, Domini underperforms by about 15percent over three years independent of fees, which would increase the disparity.

The KLD/DSI stock mix would have significantly underperformed the market even during the go-go 1990s if they had screened these once outperforming stocks for key stakeholder issues such as corporate ethics. If only one stock, Microsoft, which represents more than 6percent of the indices, had been screened out because of gross consumer anti-trust violations, the performance data would have looked markedly less robust. Entire groups of questionable stocks were screened in only because their problem area--pension fund shenanigans and shifty accounting--are too arcane for the social investment community's porous screens. For example, Verizon, the US's largest local phone company and the DSI's seventh largest holding at 2.4percent, reported a strong fiscal 2002 with profits of $389 million. Only those investors who dug into the small print at the back of the document learned that Verizon's reported earnings included $2.7 billion in gains from its pension fund investments. One problem: as Verizon knew and the KLD/DSI research did not pick up, those profits did not really exist. The company pension fund actually lost $3.1 billion, a footnote on page 58 of the 68-page report revealed, and $15 billion since 2000. These are not just abstract accounting issues. Verizon's pension fund fiasco has vaporized more than 30,000 jobs including 25percent of the union workforce. Even today, with the full details of the misrepresentations (that has so devastated key stakeholders including investors, workers and local communities) public knowledge, Verizon remains a KLD/DSI favorite. Numerous other companies considered favorites by social researchers including Cummins Engine, American Airlines and Delphi Automotive face a raft of ethical questions about underfunding of their pension funds or unilaterally trying to change their pension fund schemes.

In summary, there is no evidence to believe that the principles of social investing, rather than active-fund portfolio management and sector bets, explain the competitive performance of some social investments during the bull market. Although it is certainly possible and even intuitively likely that CSP is positively correlated with both stock performance and financial performance, the data generated by social researchers is too methodologically flawed and theoretically suspect to be of serious use for academic research. 

Discussion

While social investment advocates claim that "social investors [are] consciously . . . working to build a more just and sustainable economy" (SIF Report, 2001, p. 8), there is an unbridgeable gap between rhetoric and reality. Perception and anti-corporate ideology, not corporate reform, drives the movement. Social investing remains trapped in an anachronistic paradigm based on a bouillabaisse of problematic conservative religious principles and dated liberal social notions. The rating systems that have evolved are based on superficial litmus screens judging social propriety and social conventions. They lack a coherent moral or ethical perspective. They confuse social liberalism with ethical behavior. They systematically underplay key aspects of corporate behavior including corporate governance and transparency and overplay facile moral notions. They reflect the personal ideologies of those who conceived the ratings and collect the data. "We expect corporations to act irresponsibly," Peter Kinder has written. Why? Because corporations have as their explicit purpose the insulation of management and employees and shareholders from responsibility for their actions" (Kinder, 1997).  Social researchers then convert this ideological stew into "objective" ratings.

Social investing also suffers what is called the fallacy of reification - the conversion of a complex and multi-faceted concept, in this case corporate social responsibility, into a putative real entity--a number or rating. That "objective" system serves an organizing purpose in much the same way that the word "intelligence" is shorthand for complex mental functions. But just as the intricate concept of intelligence cannot be reduced to one simple number--IQ caricatures the complexity of mental processes-the contradictory and ambiguous ratings now in vogue cannot define CSP. This is pseudo-science, on the level of astrological research. Social researchers who rely on this data caricature the complexity of business to suit their ideological beliefs.

Business ethics requires decision-makers to weigh the impact of their ethical choices on a number of stakeholders. Rarely do such decisions lend themselves to litmus tests or reductionist numerical ratings. Social investing principles and databases are hopelessly flawed proxies of corporate social responsibility and CSP. As Rowley and Berman (2000) argue, the social consequences of corporate behavior--stakeholder actions--are more attractive criteria on which to build a model of CSP. For social investing to evolve into more than just a symbolic exercise, it is imperative that researchers develop verifiable standards beyond mere social agendas and base them instead on the social consequences of corporate behavior.

Despite hyperbolic claims by its advocates, there is no serious research that the mix of conservative religious beliefs and liberal social notions yield a set of principles that result in superior financial or stock performance. More sophisticated and nuanced research models may indeed show that some aspects of social investing positively correlate with stock and financial performance. However, that conclusion is unwarranted based on current data. Studies that claim this are unreliable and misleading. Considering social investor's historically ambivalent attitude toward business, it is not surprising that the movement and its enablers in academia remain indifferent to financial and investment principles. As a result, although social investing attracts outsized media coverage, its financial footprint and influence remain modest. Social investing has drawn very little interest from institutional investors except from religious and union-based pension funds and some mainstream funds that view it as a potential brand with appeal to niche investors.

Although social investing has demonstrated that it can be an effective brand marketing strategy, it does not fulfill its stated rationale of promoting systematic investment in "good" companies and effecting positive change--to "do good". Moreover, by implicitly encouraging the belief that the intentions of a business can be judged distinct from the economic impact of a company, social investing may promote corporate behavior that is neither socially progressive nor ethical and may result in adverse consequences to stakeholders.

References

Ackermann, M. (July 25, 2002). Socially Responsible Fund Boom: Maybe. American Banker, 167(141), 8.

Adams, R. and Durai, J. (September 1995). South Africa's Community Growth Fund. New Consumer.

Bauer, R. and Otten, R. and Koadijk, K. (2002) Ethical Mutual Performance and Investment Style. 2002 Moskowitz Award.

Bavaria, J., Becker, E. and Billenness, B. (September 15, 1994). Body Shop Scrutinized: Faces Allegations on Social Performance. Insight: Investing For A Better World. Franklin Research and Development.

Berman, S. L. and Wicks, A. C. (October 1999). Does Stakeholder Orientation matter? The Relationship Between Stakeholder Management Models and Firm Financial Performance. Academy of Management Journal, 42(5): 488-508.

Camejo, P. (ed.). (September 2002). The SRI Advantage: Why Socially Responsible Investing Has Outperformed Financially. Gabriola Island, BC, Canada: New Society Publishers.

Dodson, Jerry. (November 12, 1997). Email to Social Investment Forum Mailing List.

Domini, A. L. (January 5, 1998). Email to Social Investment Forum Mailing List.

Domini, A. L. 2001. Socially Responsible Investing : Making a Difference and Making Money. Chicago: Dearborn Trade Publishing.

Drew, C. and Belluck. P. (January 4, 1998). Deadly Bacteria a New Threat to Fruit and Produce in the U.S.: Odwalla-A Costly Strategy, a Slump in Sales. New York Times, 1+.

Entine, J. (October 14, 2002). The Backlash of Social Investing. Chicago Tribune.

Entine, J. (January/February 1999). The Odwalla Affair-Reassessing Corporate Social Responsibility. At Work, 8(1): 7-11.

Entine, J. (January 1997). Vivisecting the Anti-Vivesectionist Movement. Drug and Cosmetic Industry, 160(1): 38-41.

Entine, J. (1996). Ethical Investing. At Work, 5(5): 16-19.

Entine, J. (October 1995a). In Search of Saintly Stock Picks. Toronto Globe and Mail Report on Business Magazine, 12(4): 45.

Entine, J. (October 1995b). Rain-forest Chic. Toronto Globe and Mail Report on Business Magazine, 12(4): 40+.

Entine, J. (September/October 1994). Shattered Image: Is The Body Shop Too Good to Be True? Business Ethics, 8(5): 23-28.

Feigenbaum, C. (November 2002). Taking Stock: Investing with Your Head-and Your Heart. Vegetarian Times, 303: 78+.

Graves, S. B. and Waddock, S. A. 37 (1994). Institutional Owners and Corporate Social Performance. Academy of Management Journal, 37(4): 1034-46.

Gravitz, A. (2002). News Release: 2002 SRI Trends Report. www.socialinvest.org.

Griffin, J. J. and Mahon, J. F. (March 1997). The Corporate Social Performance and Corporate Financial Performance Debate: Twenty-Five Years of Incomparable Research. Business & Society. 36(1): 5-31.

Guerard, Jr., J. B. (February 17, 1997). Is Socially Responsible Investing Too Costly? Pensions and Investments, 25(4): 26+.

Guerard, Jr., J. B. (1997). Is There a Cost to Being Socially Responsible in Investing? Journal of Investing, 6(2): 11-18.

Guerard, Jr., J. B. (Winter 1997). Additional Evidence on the Cost of Being Socially Responsible in Investing. Journal of Investing, 6(4): 31-35.

Guerard, Jr., J. B. and Stone, B. K. (September 16, 2002). Social Screening Does Not Harm Performance. Pensions & Investments, 30(19): 30-31.

Hutton, B. R., and D'Antonio, L., Johnsen, T. (September 1998). Socially Responsible Investing. Business & Society, 37(3): 281-304.

Johnson, R. A, and Greening, D. W. (October 1999). The Effects of Corporate Governance and Institutional Ownership on Corporate Social Performance. Academy of Management Journal 42(5): 564-577.

Kinder, P. (November 13, 1997). Email to Social Investment Forum Mailing List.

Kurtz, L. (2000). SRI Studies.

Kurtz, L. (Spring 1997). The Impact of Social Screening on Growth-Oriented Investment Strategies, The Journal of Performance Measurement, 65-71.

Kurtz, L. and DiBartolomeo, D. (Fall 1996). Socially Screened Portfolios: An Attribution Analysis of Relative Performance. Journal of Investing, 35-41.

McDonald, I. (January 15, 2003). The Stars in the Sky Flicker, And Fund Stars Do the Same. Wall Street Journal, C1+.

Mackenzie, C. (1997). Ethical Investment and the Challenge of Corporate Reform. Ph.D thesis.

Naber, M. (Winter 2001). Catholic Investing: The Effects of Scrrenson Financial Returns. Journal of Investing, 10(4): 58-65.

Patsky, M. W. and Galle, K., Negron, P. (September 23, 1996) Healthy Returns: How Healthier Lifestyles Create Investment Opportunities. Adams, Harkness & Hill, Inc.

Paul, K. and Lydenberg, S. (March 9, 1997). Stakeholder Theory and Socially Responsible Investing: Toward a Convergence of Theory and Practice. 1997 IABS Meeting, Sandestin, FL.

Rowley, T. and Berman, S. (December 2000). A Brand New Brand of Corporate Social Performance. Business & Society 39(4): 397-418.

Sauer, D. A. (1997). The Impact of Social-Responsibility Screens on Investment Performance: Evidence frm the Domini 400 Social Index and Domini Equity Mutual Fund. Review of Financial Economics, 6(2): 137-149.

Turban, D.B., and Greening, D.W. (1997). Corporate Social Performance and Organizational Attractiveness to Prospective Employees. Academy of Management Journal, 40: 658-672.

2001 SIF Report on Socially Responsible Investing Trends in the United States. November 28, 2001.

2001 SRI Trends Report News Release. November 28, 2001. www.socialinvest.org

Waddock, S. A. (September 2000). Performance Characteristics of Social and Traditional Investments. Journal of Investing, 9(2): 27-38.

Waddock, S. A., and Bodwell, C., Graves, S. B. 2 (May 2002). Responsibility: The New Business Imperative. Academy of Management Executive, 16(2): 132-148.

Waddock, S. A. and Graves, S. B. (1994). The Corporate Social Performance-Financial Performance Link. Paper presented at the national meeting of the Academy of Management, Dallas, TX, August.

Waddock, S. A. and Graves, S. B. (1997a). The Corporate Social Performance-Financial Performance Link. Strategic Management Journal. 18(4): 303-319.

Waddock, S. A. and Graves, S. B. (September 1997b). Quality of Management and Quality of Stakeholder Relations: Are They Synonymous? Business and Society, 36(3): 250-279.

Waddock, S. A. and Graves, S. B. (Winter 1997c). Finding the Link Between Stakeholder Relations and Quality of Management. Journal of Investing, 6(4): 20-24.

Waddock, S. A. and Graves, S. B. (2003) Beyond Built to Last . . . Stakeholder Relations in 'Built to Last' Companies (in forthcoming issue of Business and Society Review, courtesy of authors)

Waddock, S. A. and Smith, N. (Winter 2000). Corporate Responsibility Audits: Doing Well by Doing Good. Sloan Management Review, 41(2): 75-83.

Wood, D.J. and Jones, R. E. (July 1995). Stakeholder Mismatching: A theoretical Problem in Empirical Research on Corporate Social Performance. International Journal of Organizational Analysis, 3(3): 229-267.

Young, L. (December 26, 1996). Socially Responsible Stock Index Leads S&P 500." Dow Jones News Service.

Jon Entine is an adjunct fellow at AEI.