Family firms and social responsibility: preliminary evidence from the S&P 500.

Little is known about the impact of family ownership and management on corporate social performance. Some scholars have suggested that family firms are not likely to act in a socially responsible manner, while others have indicated that socially responsible behavior on the part of the family firm

protects the family's assets. This preliminary study compares the degree to which family and nonfamily firms are socially responsible using data from 1991 to 2000 from the S&P 500. Two hundred sixty-one firms (202 nonfamily and 59 family) appeared in the S&P 500 for the 10-year period, Findings show that family firms are more socially responsible than nonfamily firms along several dimensions. This is likely due to family concern about image and reputation and a desire to protect family assets.

**********

An increasing number of scholars have turned their attention to studying the impact that a family, which owns and operates a business, might have on that firm's performance (Dyer, 2005). Studies comparing the performance of family firms and firms with no family connections have focused on such variables as return on assets (Anderson & Reeb, 2003), sales growth (Chrisman, Chua, & Litz, 2004; Daily & Dollinger, 1992; Gallo, Tapies, & Cappuyns, 2000), job satisfaction (Beehr, Drexler, & Faulkner, 1997), and innovation (Tanewski, Prajogo, & Sohal, 2003). Little is known, however, about the impact a family might have on the corporate social performance (CSP) of a firm that it controls and there appear to be conflicting positions in the literature regarding the subject. Some scholars such as Morck and Yeung (2004) have argued that family firms are highly self-interested and merely want to protect their own parochial interests. Thus, the families that own various enterprises would not be inclined to improve the broader societies in which their firms are embedded. Indeed, such family firms may foster corruption, which undermines public confidence as well as the legitimacy of public institutions in order to protect their own interests. Others, such as Godfrey (2005), have suggested that firms, including those owned by families, have incentives to be socially responsible to maintain a positive image, since a positive reputation in the minds of key stakeholders may serve as a form of social insurance, protecting the firm's (and family's) assets in times of crisis.

To explore this rather complex and uncertain relationship between family ownership and control and CSP, this preliminary study will:

1. briefly discuss the concept of CSP;

2. discuss those theories that help us understand why family-owned firms may be more or less socially responsible; and

3. determine what empirical support there is for these theories by comparing the CSP of family and nonfamily firms, using a sample of companies in the S&P 500 during the years 1991-2000.

What Is CSP?

While there have been numerous approaches to defining and studying CSP (e.g., Margolis & Walsh, 2003; McWilliams & Siegel, 2001; Wartick & Cochran, 1985; Whetten, Rands, & Godfrey, 2001), H.R. Bowen's seminal idea that businessmen should "pursue those policies, to make those decisions, or to follow those lines of action which are desirable in terms of the objectives and values of our society" seems to capture the essence of the concept (Bowen, 1953, p. 6). The concept of CSP suggests that a society's firms, whose very existence is predicated on societal sanction and support, have the obligation to be economically and socially responsible "citizens" and act in publicly responsible ways (Wartick & Cochran, 1985). Prior research on CSP suggests that firms might be rated as to their degree of social performance along two dimensions: (1) positive "social initiatives" (e.g., charitable giving), where the firm proactively tries to improve society, and (2) the firm's ability to avoid activities that might prove to be of "social concerns" (e.g., polluting the environment) (http:www.kld.com). Firms who fail to engage in positive social initiatives or who become the target of social concerns may face legal, economic, or social sanctions from their stakeholders and society in general (Godfrey, 2005).

While Bowen and others argue that there is a moral dimension underpinning socially responsible behavior (Donaldson, 1982; Rawls, 1971)--that corporations should "do good" because it is the right thing to do--some have suggested that socially responsible behavior is also the profitable thing to do (Waddock & Graves, 1997). However, previous reviews of the various studies on this topic by Bragdon and Marlin (1972) and Griffin and Mahon (1997), and more recently Margolis and Walsh (2003) and Orlitzky, Schmidt, and Rynes (2003) suggest a tenuous link between CSP and corporate financial performance. This raises the question: Should corporations attempt to benefit society even if there is no apparent financial gain for the corporation's shareholders? Friedman (1970) has argued that managers who attempt to "do good" rather than helping the firm "do well" are violating their fiduciary responsibility to shareholders. Hence, he argues that a firm's manager should only act in a socially responsible manner if it helps to maximize shareholder wealth. Proper behavior should be determined by the market, not by moral imperatives.

In addition to the moral and financial arguments related to why a firm might be socially responsible, which are prevalent in this literature, there appear to be other determinants of CSP as well. In this regard, we will now turn our attention to the question as to whether or not family control of a firm should affect the firm's proclivity to be socially responsible. Because there are legitimate competing theories on why family firms are likely to be more or less socially responsible than nonfamily firms, we will draw on three streams of theory to develop two sets of alternative hypotheses. These streams include: (1) self-interest; (2) identity, image, reputation, and identification; and (3) moral capital.

Why Family Firms May Not Be Socially Responsible

In his classic work, The Moral Basis of a Backward Society, Edward Banfield describes a phenomenon he calls "amoral familism," a term he used to characterize the families of Southern Italy in the 1950s (Banfield, 1958). Banfield, working as an anthropologist, noted that the villages he studied were afflicted with severe poverty--poor roads, substandard schools, and other infrastructure weaknesses, along with poor economic conditions in general. As he attempted to understand the underlying causes of such poverty, he came to the conclusion that the families in these communities were unable to cooperate with one another to build a better society. This was due, in large part, to the lack of trust between families and those on the "outside." Banfield (1958, p. 116) writes: "towards those who are not of the family, the reasonable attitude is suspicion. The parent knows that other families will envy and fear the success of his family and that they are likely to seek to do it injury. He must therefore fear them and be ready to do them injury in order that they may have less power to injure him and his." From such attitudes spring behaviors on the part of a family that are based on self-interest, with the outcomes being at the expense of, or even injury to, other families and the broader society. In the context of a family firm, this dynamic of amoral familism would suggest that owning-families would not likely be socially responsible, but would likely emphasize self-interest. The outcomes of such beliefs might be behaviors such as nepotism that could disadvantage company employees and other stakeholders, or competing in the marketplace in ways that could prove harmful to the greater social good (Rosenblatt, de Mik, Anderson, & Johnson, 1985; Schulze, Lubatkin, Dino, & Buchholtz, 2001).

There appears to be some empirical evidence that family-controlled firms may indeed be irresponsible social actors, causing significant social concerns. Morck and Yeung (2004) examined the concentration of family-controlled firms in 27 of the larger industrialized countries in the world. They correlated the concentration of family firms with various dimensions of societal progress: economic development, physical infrastructure, health care, education, quality of government, and social development (as defined by the degree of income inequality). As a result of their analysis, they conclude: "countries whose large firms are controlled by great mercantile families are more backward in a number of dimensions. They are poorer. They provide worse public goods--including worse infrastructure, worse healthcare, worse education, and more irresponsible macroeconomic policies. They are less egalitarian" (p. 395).

Morck and Yeung (2004) discuss possible explanations for these findings and conclude that the owning-families are more invested in protecting their own interests than they are in developing their countries' economies and, in a sense, helping their "neighbors." To protect their interests, these families become adept at bribing government officials. As Morck and Yeung note: "established wealthy families controlling substantial assets can pay corrupt officials up front for subsequent favors" (p. 401). Hence, these families become "political rent seekers" at the expense of the broader society.

Matthew Josephson, in his celebrated work The Robber Barons, highlights this indifference to public needs by wealthy families in the United States in the late 1800s (Josephson, 1934). Josephson describes the "robber barons" and their families, such as the Rockefellers and the Vanderbilts, as controlling substantial segments of the U.S. economy to the detriment of society. The robber barons were primarily interested in accumulating personal and family wealth and were known for their oppressive and heavy-handed tactics. They bribed government officials, engaged in secret deals that benefited themselves at the expense of the public, and employed Pinkerton detectives to quell labor unrest. When confronted with charges of undermining the public good, J.R Morgan replied: "I owe the public nothing"; William Vanderbilt was somewhat more blunt when he exclaimed: "The public be damned" (Josephson, 1934, pp. 441, 118).

Such attitudes on the part of powerful families suggest that family-controlled firms may not have the public's best interest at heart. They may be more interested in protecting their own interests than the interests of the public, and therefore would likely cause the firms that they lead to act in less socially responsible ways than those firms led by leaders with no such family connections.

In summary, if Morck and Yeung's (2004) theory is correct that family firms act in self-interested ways to the detriment of society, we should find:

Hypothesis 1: Family firms will significantly have fewer positive social initiatives than nonfamily firms; and

Hypothesis 2: Family firms will significantly have more social concerns than nonfamily firms.

Why Family Firms May Be Socially Responsible

Recent works by Whetten and Mackey (2005) and Godfrey (2005) suggest reasons why family-owned companies may, in fact, be more socially responsible than firms without family ties. Whetten and Mackey (2005) describe how one's corporate (or family) identity, one's need to create a positive image, and the desire to maintain a good reputation with stakeholders, may encourage a firm's leadership to act in a socially responsible fashion. Godfrey suggests that firms and their family owners may want to appear socially responsible to create "chits" that can be "cashed in" when needed as "insurance" to protect their assets. We will now discuss each of these theories in turn.

Identity, Image, Reputation, Identification, and Social Responsibility

There is growing interest in examining organizational identity, image, and reputation as determinants of CSP (Whetten & Mackey, 2005). While these terms have been discussed in some detail elsewhere (e.g., Davies, Chun, Vinhas da Silva, & Roper, 2001; Dutton, Dukerich, & Harquail, 1994; Gioia, Schultz, & Corley, 2000, Hatch & Schultz, 2000; Wartick, 2002; Whetten & Godfrey, 1998), we will briefly describe how they relate to the subject of family firm social performance.

Organizational identity has been conceptualized as the shared answers to the question, "Who are we as an organization?" that qualify as central, enduring, and distinctive elements of an organization (Albert & Whetten, 1985; Whetten & Mackey, 2002). As an explanation for CSP, the notion of organizational identity suggests that firms will engage in socially responsible business practices when to do otherwise is unthinkable--these concerns have been so central to who we are as an organization for so long that we cannot imagine the organization as anything but a good corporate citizen. Complementing organizational identity, organizational image is "what organizational agents want their external stakeholders to understand is most central, enduring, and distinctive about their organization" and organizational reputation is "a particular type of feedback, received by organizations from their stakeholders, concerning the credibility of the organization's identity claims" (Whetten & Mackey, 2002, p. 401). An organization's image is a conception of the organization, that is, intentionally projected to outsiders and its reputation is the conception of the organization by outsiders, that is, reflected back to organizational members. Image is often linked to corporate strategy, and reputation constitutes an important organizational asset (Fombrun, 1996). Scholars who posit a positive relationship between the public's perception of a firm as a socially responsible corporate citizen and the firm's financial performance tend to invoke image and reputation as the explanatory mechanisms (Whetten & Mackey, 2002).

Recently, Whetten and Mackey (2005) used these concepts to explain different levels of consistency in the social performance ratings given to S&P 500 firms during the 1990s. They concluded that firms who initiated at least one socially responsible business practice during their "founding era" were significantly more likely to receive consistently higher social performance ratings than similar firms whose corporate histories indicated that the first such practice was initiated sometime after the founder left the firm, or firms whose histories never mention these kinds of practices. In line with Stinchcombe's (1965) classic treatment of the long-term consequences of founding conditions, these results suggest a strong founder-effect explanation for consistency in contemporary CSP ratings.

The research on family firm start-ups indicates that founders leave a deep impression on the business organizations they create (Dyer, 1986; Schein, 1983). This research also suggests that family firm founders are likely to view their business operations as an extension of themselves--their identity, or self-view (Dyer, 1992; Kets de Vries, 1977; Schein, 1983). In most cases, their vision is to pass on a legacy to their posterity, not simply a sustainable income stream. Thus, they would likely view negative press pertaining to a bitter labor strike over health benefits, customer complaints about faulty products, or legal suits following an ecological disaster as indelible stains on themselves and as an extension on their company and their family name (Post, 1993).

There are also identity-based reasons to believe that families founding businesses are likely to see their corporate assets as a means for "doing good" in society--that more than wishing to keep their family name unstained by harmful business practices, they would actively use their material wealth and associated social capital to advance important social causes. A study of 3,000 families conducted several years ago by Stinnett and DeFrain (1985) found six attributes of what they labeled "strong families." One of those attributes was "spiritual wellness." The term spiritual wellness did not necessarily mean that these strong families were religious or had a belief in God, but that they had a sense of a greater good or power in life that gave their families purpose and meaning. It was not enough for these families to merely live together as an economic and social unit. The families felt a need to have a greater purpose that they could work toward and even sacrifice. Many of these families' greater purposes concerned helping others outside the family and benefiting their communities.

Post (1993) describes how the Saunders family, owners of Boston's Park Plaza Hotel and other properties, has been involved in numerous community initiatives consistent with the family's identity as a responsible corporate citizen:

   The Saunders family has been concerned with health issues and
   involved with Boston-area health institutions for many years. The
   family has actively supported local hospitals ... Roger Saunders
   has contributed substantial time and effort to the financial
   affairs of these [hospitals]; Nina Saunders also gave generously of
   her time and energy to these causes until her death in 1991. Gary
   and Jeffrey Saunders serve on the boards of local organizations
   concerned with health issues, and Todd Saunders also has a special
   interest in the American Cancer Society ... Issues of health and
   health care thus define an area in which the family has found an
   opportunity to express its spirit of responsibility ... At the Park
   Plaza Hotel, the environmental action program [e.g., protecting the
   health of guests, recycling, reducing water and energy consumption,
   etc.] emerged from the relationship between the Saunders family's
   values, and needs and the mission, strategies, and operations of
   the business (pp. 143, 144).

As a result of their efforts, in 1992, President George Bush gave the Saunders family and the Park Plaza Hotel the President's Environment and Conservation Challenge Award. The citation from the president noted that "This family-owned and -operated landmark property has successfully aligned business and environmental action" (Post, 1993, p. 131).

A related concept, member identification, helps us understand how individual-level psychological processes might account for family firm founders, shareholders, and managers steering their firms toward socially responsible business practices. Members are said to identify with their organization to the extent that they see an overlap between the identity of the organization and their individual identity (Foreman & Whetten, 2002; Pratt, 1998). A basic tenet of social identity theory is that individuals view themselves as extensions of the groups in which they hold memberships--particularly those groups with which a person is closely identified (Hogg, 2003; Hogg & Terry, 2001). This identification link between individual identity and group or organizational identity is relevant for our discussion on CSP because the universal human need for positive self-regard dictates that individuals prefer membership in groups that are generally viewed positively by outsiders (Baumeister, 1998). To avoid negative self-views, members who closely identify with a group that develops a bad reputation will, other things being equal, switch to groups with better reputations. The obvious problem family firm owners face is that in response to bad publicity they cannot, practically speaking, switch groups. Although they may sell off their shares in disgust, they cannot escape the fact that their family, in some sense, is the offending firm.

The length to which families will go to maintain a positive image in their communities is well documented in the family studies field (Reiss, 1981). Even highly dysfunctional families that experience child abuse or even incest tend to cover up such behavior by threatening family members and compelling them to secrecy. Identification theory's explanation for this self-serving practice is threefold: (1) Family members share a common need to view themselves positively (I am a good person); (2) they know they cannot switch families, if word of a shameful family activity "gets out," so (3) they band together to preserve the family's reputation.

In contrast to the high levels of identification that family firm founders, as well as subsequent family shareholders and managers, are likely to experience as extensions of the family business, there is evidence that their counterparts in nonfamily firms are less likely to so closely identify with their work organizations. Studies done over the years by Schein (1976, 1978) and others demonstrate that college graduates, particularly MBAs, regularly change their employment, with most staying in their jobs for only a few years. And, the average tenure of CEOs in major firms has continued to decline. A study conducted by the consulting firm Drake Beam Morin (DBM) reported in 2000, reviewed the tenure of CEOs in 476 of the world's largest companies (http://www.amgr.com/pdf/ o200.pdf). DBM noted that: (1) Nearly half of the CEOs in these firms had held their jobs for less than 3 years; (2) two-thirds of the companies had installed a new CEO in the past 5 years; and (3) CEO turnover is continuing to rise. The notion of lifetime employment is an anachronism--neither firms with ever-changing job requirements nor an increasingly mobile labor force wish to have their options curtailed by long-term employment relationships. Consistent with our earlier analysis, it follows that managers of nonfamily firms may find it much easier to buffer their personal self-view if their companies develop a bad reputation since they see their employment as a short-run proposition--hence, low commitment translates into low identification. Thus, they would have less of a personal stake in helping their firm avoid a bad reputation (Meek, Woodworth, & Dyer, 1988).

In summary, the literature regarding organizational identity, image, reputation, and identification suggests that compared with their nonfamily firm counterparts, family firm founders, shareholders, and managers are more likely to initiate a tradition of socially responsible business practices and to avoid harmful practices that can besmirch the image of the firm. These actions may be motivated by (1) family firm owners' and managers' shared intent to treat their firm as an extension of their personal commitment to both do well and do good as members of society, or (2) by their fear that a bad company reputation would soil the "good name" of their family and, in turn, reflect poorly on them individually.

Moral Capital as Insurance for Family Firms

Godfrey's (2005) argument begins with a simple premise: In large corporations, "harm" is bound to happen. Sooner or later, managers will likely underestimate the social consequences of a strategic action: An accountant might embezzle large sums of money, a subcontractor may engage in sweatshop activities in a remote corner of Asia, or a technological failure may cause an ecological disaster. Godfrey further notes that the same harmful incident may provoke more severe social sanctions toward one firm than another firm--varying from mild rebukes in the business press to a consumer boycott. His explanation for why some firms tend to receive more lenient social sanctions than other firms is that they have built up stores of goodwill with their key stakeholders. If stakeholders view a firm as having "bad character," seeing malicious intent in a harmful incident, they are more likely to apply severe social sanctions. In contrast, firms with a long and commendable record of "doing good" will likely be given the "benefit of the doubt" by stakeholders when a similar incident occurs. In brief, Godfrey (2005) argues that reserves of "goodwill," what he refers to as "positive moral capital," provide insurance-like protection--protecting a firm's underlying relational wealth and earnings streams against loss of economic value arising from the risks of business operations (Trieschmann & Gustavson, 1998).

The story of John H. Patterson, founder of National Cash Register (NCR), helps to illustrate Godfrey's point:

   NCR was facing stiff competition from other companies that were
   selling refurbished cash registers. To combat this threat,
   Patterson ordered Thomas Watson [the future leader of I.B.M.] and a
   few other managers to set up a company--to be funded by NCR--that
   would sell the secondhand cash registers at cost. This move
   effectively destroyed the competition, but it was obviously
   illegal. The competition filed suit, and Patterson and twenty-seven
   of his top managers--including Thomas Watson--were found guilty.
   Patterson appealed. In the interim, a terrible flood engulfed
   Dayton [Ohio, where the company was located]. Patterson mustered
   all the company's resources to feed the hungry and the homeless. In
   the wake of the public support that followed, Patterson won his
   appeal and was given a large victory parade in Dayton (Dyer, 1986,
   p. 63).

The implication from this story is that Patterson's efforts to help Dayton's flood victims proved to be an important factor in the final disposition of his legal troubles. By doing good, Patterson generated enough public goodwill that he was able to avoid becoming a convicted felon.

In the context of a family firm, much of an owning-family's current income stream and its accumulated wealth may be directly connected with the firm's profitability and the family's ownership stake in the firm. Thus, damage to the firm through loss of reputation, litigation, or other deleterious events, can have a significant negative impact on the family's wealth. To the extent that a family's financial well-being is tied to the firm, the family may be unwilling to risk the loss of that wealth. Therefore, the family may be more willing to be proactive in developing positive moral capital--and the "insurance" it provides--than are managers who have no such stake in the firm. In the course of developing positive moral capital, the family may therefore be more willing to act in socially responsible ways.

From this stream of theory, if a family firm is motivated to create positive moral capital or has a "socially responsible identity," the firm should be proactive in launching initiatives consistent with its identity or in attempting to create goodwill. Under these conditions we should find:

Hypothesis 3: Family firms will significantly have more positive social initiatives than nonfamily firms.

If, however, a family wants to maintain a positive family and firm image and have a strong reputation, it should strive to avoid any "social concerns" that could damage that image and reputation. Thus, if image and reputation are the primary determinants of socially responsible behavior, we should find:

Hypothesis 4: Family firms will significantly have fewer social concerns than nonfamily firms.

The Study

Our review of several bodies of scholarly knowledge yielded alternative predictions regarding the impact of family influence on CSP. To answer the question: Are family firms more socially responsible than nonfamily firms?, we needed a population of similar firms, both family and nonfamily, and standardized social performance measures for those firms. We chose for our study those firms that were ranked in the S&P 500 for each of 10 years, 1991-2000. These firms are large, publicly traded corporations which reflect a broad cross section of U.S. businesses. There were 261 firms for which we had complete data on their social performance that were listed in the S&P 500 for that 10-year period.

To differentiate the family from the nonfamily firms in this population, we relied on a BusinessWeek survey which tracked the financial performance of S&P 500 family and nonfamily firms for approximately the same 10-year period that we were interested in (BusinessWeek, 2003). A firm was designated by BusinessWeek as a "family firm" if members of the founding family had continued to remain as significant company shareholders and/or members of the founding family were still in senior management or held a seat on the board of directors. Using these criteria, 59 firms that appeared in the S&P 500 for the 10-year period were designated as "family firms" and 202 firms were designated as "nonfamily firms." The family firms represented diverse industries such as automobiles (Ford), chemicals (DuPont), publishing (Knight-Ridder), pharmaceuticals (Eli Lilly), and retail (Wal-Mart).

To ascertain CSP, we relied on data collected from the social performance rating service of Kinder, Lydenberg, and Domini (KLD), which, since 1991, has rated the social performance of all the S&P 500 firms. Over 40 peer-reviewed publications have based their findings on the KLD data (http://www.kld.com/research/universities/ universities_ratings.html). KLD evaluates a firm's social performance in seven categories of "positive initiatives" and 13 categories of "social concerns." (To make comparisons between initiatives and concerns, we only used the seven concern categories that were identical to the positive initiative categories.) Each category is defined by several items reflecting a type of initiative or concern. Appendix A describes the categories of social performance rated by KLD used in our study. In any given year, KLD rates S&P 500 firms in each of these categories. For positive initiatives, a company is given either a "0" or a "1" for each category. A "1" is given to a company in a particular category if KLD determines that the company has launched a positive initiative during the previous year. A "0" is awarded if there is no evidence of such initiatives. For the categories related to social concerns, a company is awarded a "-1" for a particular concern if it demonstrates that it has acted irresponsibly during the previous year, and a "0" is awarded if there is no evidence of negative social behavior. Thus, the KLD data yielded 36,540 observations in the 14 categories for the 261 firms during the 10 years we studied.

Firm scores from the KLD categories for the 10-year period were aggregated and t-tests were conducted to compare the social performance of the family and nonfamily firms across the 14 categories. We also aggregated the scores from each category to create a single "social initiative" score and a single "social concern" score. We deemed a two-tailed t-test to be most appropriate inasmuch as we wanted to determine if there were significant differences in the mean scores of the family and nonfamily firms (the null hypothesis being that the means are equal), and it is more conservative than the one-tailed test.

In addition to the t-tests, we also wanted to examine the relationship between CSP and family business influence in the presence of other variables. Hence, we conducted several regressions that also included performance variables (return on assets [ROA], return on sales [ROS], and Tobin's q) and control variables such as size (number of employees) and leverage (long term debt/common shares outstanding). To avoid the problem of multicollinearity with the performance variables, we tested three different models. We were not able to effectively control for industry since our sample did not include enough firms in each industry, even when we used two-digit Standard Industrial Classification (SIC) codes. (1)

To do our analysis, we used a random effects panel model given that our data were both cross-sectional and longitudinal and that random effects were found to be present in the data. If ordinary least squares were used when random effects were present in the data, a downward bias would have been introduced in the calculation of the standard errors, which might have resulted in certain parameter estimates being significant when they were not (Kennedy, 2003).

We determined that random effects were present in the data through the use of the Breusch-Pagan Lagrangian Multiplier (LM) test (Breusch & Pagan 1980) for random effects (test statistic = 4,563.66 ~ [chi square] (1), p < .000). The Breusch-Pagan LM test examines whether or not the firm-specific intercepts are different from each other (i.e., testing if fixed or random effects is preferred to ordinary least squares). Typically, Hausman's (1978) specification test is used to ensure that the model has random and not fixed effects. This test essentially compares a random effects model to a fixed effects model to see which one is a better fit. Since the variable of interest (family business) is time-invariant, fixed effects cannot be estimated without dropping the family business variable. Thus, the Hausman test comparing the random effects model with family business variable to the fixed effects model without family business variable is inappropriate--comparing apples to oranges. Since fixed effects are not a valid statistical option, random effects are used in the analysis. Theoretically, the random effects estimator is justified because we have no reason to believe that any unobserved variables captured by the composite error term are correlated with the regressors (Kennedy, 2003).

Results

The results of the t-tests are found in Table 1. Our comparison of the family and nonfamily "positive initiative" means showed no significant difference between the family and nonfamily firms in the seven categories and no significant difference in their aggregate means. The aggregate mean for the family firms for positive initiatives was 2.298 and the mean for the nonfamily firms was 2.320. For social concerns, however, there was a significant difference between the family and nonfamily firms in the aggregate mean scores of the seven categories. The mean score of the family firms for social concerns was -1.547 while the mean score for the nonfamily firms was -2.455 (the larger the mean, the more concerns). Moreover, in all seven categories family firms had fewer social concerns than did the nonfamily firms, and the differences in three of the concern categories were statistically significant: product concerns, environmental concerns, and employee concerns.

The regressions were also consistent with the findings from our t-tests. For social initiatives, the performance variables and firm size had significant relationships, but family business and leverage did not. For social concerns, however, family business had a significant relationship, as did firm size, leverage, and Tobin's q, but ROS and ROA did not. These findings are found in Tables 2 and 3.

These results do not support hypotheses 1 and 2, that family firms would have fewer positive initiatives and more social concerns than nonfamily firms. Thus, we can reject both hypotheses. Hypothesis 3, that family firms would have more positive initiatives, should also be rejected given that the means of both groups were virtually identical. However, hypothesis 4, that family firms would have fewer social concerns than nonfamily firms, is supported by our findings. Overall, family firms had fewer social concerns and several of the categories of social concerns noted significant differences between the two groups.

Discussion

Our data, although preliminary in nature, indicate that family firms and nonfamily firms behave similarly in regard to positive social initiatives, but family firms are more adept at avoiding social concerns than their nonfamily counterparts. Hence, these findings suggest that Morck and Yeung's (2004) view that family firms are socially irresponsible actors may indeed be wrong, or at least incomplete. Family firms, in general, do as well or better than the nonfamily firms according to KLD's assessment of social performance in their designated categories. However, it may also be true that national or cultural differences may account for the fact that family firms in the United States seem to be socially responsible actors. Family firms in less developed countries, where corruption is more prevalent than the United States, may be less inclined to be socially responsible, and moreover, the country's government and legal system may make it easier for family firms to undermine the public good. Moreover, the KLD data analyze specific socially responsible practices, while Morck and Yeung based their conclusions on the correlation between family firm concentration and the degree to which a country was able to effectively care for its citizens. We believe that the connection between family ownership and socially responsible behavior can truly be determined only through an examination of corporate practices, not through more general associations between the presence of family firms and negative social outcomes (Dyer, 2003). Future research should attempt to examine CSP practices in family firms across national boundaries to answer this question more definitively.

Our findings also suggest that a family firm's identity or its desire to create positive moral capital may not distinguish it from firms managed by those without family connections. In the aggregate, nonfamily firms were just as likely to engage in positive social initiatives as were the family firms. A family's desire to act consistently with a positive social identity or its desire to create "chits" as Godfrey (2005) has argued, may not be sufficient to motivate the family to use its wealth and resources to launch social initiatives. Additional work needs to be done to explore and understand the conditions that would encourage family firms to support CSP initiatives.

Finally, we discussed the importance that image and reputation may play in encouraging CSR Our findings appear to be consistent with this line of theorizing. Families that are concerned about their image and reputation would want to avoid being labeled socially irresponsible--which would be inconsistent with portraying a positive image and damage to their reputation. Socially irresponsible corporate actions related to environmental concerns such as hazardous waste and air pollution, employee concerns such as workplace safety, workforce reductions, and union relations, as well as product safety concerns, frequently make their way to media outlets such as newspapers and television. Such unfavorable publicity, like product recalls, can cause significant damage to the reputation of a firm and family and can undermine a well-cultivated image. Furthermore, social concerns along these dimensions often spawn high-profile lawsuits that can damage one's reputation and can cause the firm (and the family) to be convicted in the court of public opinion, regardless of the legal outcome. Such loss of reputation could have a negative impact on the family's wealth. Hence, our findings seem to support the idea that family firms may be socially responsible actors to protect their image and reputation. Managers in nonfamily enterprises are not likely to be as concerned about these factors since societal evaluations of the behavior of the corporations they manage would not likely have the same impact on their personal image, and their families' reputation. (2)

The conclusion that might be naturally drawn from our findings is that owners and managers, who see themselves and/or their families as personally identified with the firms they own and manage, may be more willing to encourage CSP than those owners and managers who believe they can toil and reap firm benefits in relative anonymity, and need not accept personal responsibility for the firm's poor behavior. A family that owns an enterprise with its "name on the building" may have greater difficulty distancing itself from the firm it controls and hence may feel a greater responsibility to ensure that the firm does nothing to damage the family's reputation. While it is possible that nonfamily owners and managers could also feel this sense of responsibility, it is less likely than those who have their names and families associated with a firm.

Conclusion

We have suggested that there are various theories that might account for the social performance of a family-owned business. Our study, comparing the social performance of family and nonfamily firms in the S&P 500 over a 10-year period, supports the thesis that family firms are more likely to be socially responsible actors than are firms without family involvement. This is likely due, in part, to the fact that families see their images and reputations as inextricably connected to the firms they own, and therefore will be unwilling to damage those reputations through irresponsible actions on the part of their firms. While the family-owned firms in our sample were no more likely to engage in positive social initiatives than were the nonfamily firms, the data suggest that family firms do avoid those actions that would cause them to be labeled as socially irresponsible and hence avoid negative publicity that would undermine their reputations.

Clearly, more work needs to be done to ascertain the differences in CSP practices in family firms across national boundaries to better test Morck's and Yeung's (2004) hypothesis that family firms are more interested in their well-being than the public good. Also, studies could be designed to determine under what conditions and to what extent family firms will be willing and able to develop positive moral capital along the lines suggested by Godfrey. Our study was also limited given the fact that we were not able to directly measure such constructs as "identity" and "reputation," and our definition of "family firm" was constrained by the BusinessWeek sample we used. With such limitations, we see our findings as preliminary in nature with further empirical work needed. Most importantly, additional work is needed to compare CSP within the population of family firms. For example, we might compare the CSP of privately held family firms with those publicly held family firms in our sample. By doing so, we may be able to better understand how and why family control of a firm may be a significant determinant of CSE

Appendix A

KLD Qualitative Screens for Social Initiatives and Concerns

Social Initiatives

Community Initiatives: Generous giving (1.5% NEBT to charity); Innovative giving; Support for housing; Support for education, Indigenous peoples relations; Other community strengths.

Diversity Initiatives: CEO a woman or minority; Promotion of women and minorities; Board of directors includes women, minorities, and/or disabled individuals; Family benefits (address work/family, childcare, flextime); Employment of the disabled

Employee Initiatives: Strong union relations; Cash profit sharing; Employee involvement (stock options, gain sharing, participation in management decision making), Strong retirement benefits; Other (employee safety record, noteworthy commitments to employee well-being)

Environmental Initiatives: Beneficial products and services; Pollution prevention; Recycling; Alternative fuels; Communications (signatory to CERES principles, publishes environmental report)

Non-U.S. Operations Initiatives: Investment in positive international community involvement; support for indigenous peoples; support for positive initiatives in countries such as South Africa

Product Initiatives: Quality program; R&D/Innovation; Benefits to the economically disadvantaged

Other Initiatives: Limited top management or board compensation; Ownership strength (owns or is owned by a socially responsible company by KLD standards)

Social Concerns

Community Concerns: Investment controversies; Negative economic impact (e.g., environmental contamination, water rights disputes, plant closings, etc.); Focus of strong community opposition

Diversity Concerns: Controversies (e.g., affirmative action); Nonrepresentation (women in senior management positions or on board)

Employee Concerns: Poor union relations; Safety controversies; Workforce reductions; Pension/benefits concerns

Environmental Concerns: Hazardous waste; Regulatory problems; Ozone depleting chemicals; Substantial emissions; Agricultural chemicals; Manufactures coal/oil; Environmental accidents

Non-U.S. Operations Concerns: Company involvement with South Africa, Northern Ireland, Burma, and Mexico; International labor; Indigenous peoples' relations

Product Concerns: Product safety; Marketing/contracting controversy; Antitrust violations; Other (affiliated franchises, product malfunction fines, etc.)

Other Concerns: High compensation to top management or board members; Tax disputes; Ownership concern (firm owns a company KLD has cited as having an area of social concern)

REFERENCES

Albert, S. & Whetten, D.A. (1985). Organizational identity. In L.L. Cummings & B.M. Staw (Eds.), Research in organizational behavior (Vol. 7, pp. 263-295). Greenwich, CT: JAI Press.

Anderson, R.C. & Reeb, D.M. (2003). Founding-family ownership and firm performance: Evidence from the S&P 500. Journal of Finance, 58(3), 1301-1328.

Banfield, E.C. (1958). The moral basis of a backward society. Glencoe, IL: Free Press.

Baumeister, R.F. (1998). The self. In D.T. Gilbert, S.T. Fiske, & G. Lindzey (Eds.), The handbook of social psychology (pp. 680-740). Boston: McGraw-Hill.

Beehr, T., Drexler, J.A., & Faulkner, S. (1997). Working in small family businesses: Empirical comparisons to non-family businesses. Journal of Organizational Behavior, 18(3), 297-312.

Bowen, H.R. (1953). Social responsibilities of the businessman. New York: Harper & Row.

Bragdon, J.H., Jr. & Marlin, J.A.T. (1972). Is pollution profitable? Risk Management, 19(4), 9-18.

Breusch, T.S. & Pagan, A.R. (1980). The Lagrange multiplier test and its application to model specification in econometrics. Review of Economic Studies, 47, 239-253.

BusinessWeek. (2003). Family Inc. November 10, 111-114.

Chrisman, J.J., Chua, J.H., & Litz, R.A. (2004). Comparing the agency cost of family and non-family firms. Entrepreneurship Theory and Practice, 28(4), 335-354.

Daily, C.M. & Dollinger, M.J. (1992). An empirical examination of ownership structure in family and professionally managed firms. Family Business Review, 5(2), 117-136.

Davies, G., Chun, R., Vinhas da Silva, R., & Roper, S. (2001). The personification metaphor as a measurement approach for corporate performance. Corporate Reputation Review, 4(1), 113-127.

Donaldson, T. (1982). Corporations and morality. Englewood Cliffs, NJ: Prentice-Hall.

Dutton, J.E., Dukerich, J.M., & Harquail, C.V. (1994). Organizational images and member identification. Administrative Science Quarterly, 39, 239-263.

Dyer, W.G., Jr. (1986). Cultural change in family firms: Anticipating and managing business and family transitions. San Francisco: Jossey-Bass.

Dyer, W.G., Jr. (1992). The entrepreneurial experience: Confronting career dilemmas of the start-up executive. San Francisco: Jossey-Bass.

Dyer, W.G., Jr. (2003). The family: The missing variable in organizational research. Entrepreneurship Theory and Practice, 27(4), 401-416.

Dyer, W.G., Jr. (forthcoming). Examining the "family effect" firm on firm performance. Family Business Review.

Fombrun, C.J. (1996). Reputation: Realizing value from the corporate image. Boston: Harvard Business School Press.

Foreman, P. & Whetten, D.A. (2002). Members' identification with multiple-identity organizations. Organization Science, 13(6), 618-635.

Friedman, M. (1970). The social responsibility of business is to increase its profits. New York Times Magazine, September 13, 32-33, 122, 124, 126.

Gallo, G.A., Tapies, J., & Cappuyns, K. (2000). Comparison of family and non-family business: Financial logic and personal preferences. "Chair of Family Business" IESE. Research Paper No. 406 BIS, University of Navarra, Barcelona.

Gioia, D., Schultz, M., & Corley, K. (2000). Organizational identity, image and adaptive instability. Academy of Management Review, 20, 63-81.

Godfrey, P.C. (2005). The relationship between corporate philanthropy and shareholder wealth: A risk management perspective. Academy of Management Review, 30(4), 777-798.

Griffin, J.J. & Mahon, J.E (1997). The corporate social performance and corporate financial performance debate: Twenty-five years of incomparable research. Business and Society, 36, 5-31.

Hatch, M.J. & Schultz, M. (2000). Scaling the tower of Babel: Relational differences between identity, image, and culture in organizations. In M. Schultz, M.J. Hatch, & M.H. Larsen (Eds.), The expressive organization: Linking identity, reputation, and the corporate brand (pp. 11-35). Oxford: Oxford University Press.

Hausman, J.A. (1978). Specification tests in econometrics. Econometrica, 46, 1251-1271.

Hogg, M.A. (2003). Social identity. In M.R. Leary & J.P. Tangney (Eds.), Handbook of self and identity (pp. 462-479). New York: Guilford Press.

Hogg, M.A. & Terry, D.J. (2001). Social identity processes in organizational contexts. Philadelphia: Psychology Press.

Josephson, M. (1934). The robber barons. New York: Harcourt, Brace, and Company.

Kennedy, P. (2003). A guide to econometrics (5th ed.). Cambridge, MA: MIT Press.

Kets de Vries, M.ER. (1977). The entrepreneurial personality: A person at the crossroads. Journal of Management Studies, 14, 34-57.

Margolis, J.D. & Walsh, J.P. (2003). Misery loves companies: Rethinking social initiatives by business. Administrative Science Quarterly, 48, 268-305.

McWilliams, A. & Siegel, D. (2001). Corporate social responsibility: A theory of the firm perspective. Academy of Management Review, 26, 117-127.

Meek, C., Woodworth, W.W., & Dyer, W.G., Jr. (1988). Managing by the numbers: Absentee ownership and the decline of American industry. Reading, MA: Addison-Wesley.

Morck, R. & Yeung, B. (2004). Family control and the rent-seeking society. Entrepreneurship Theory and Practice, 28(4), 391-409.

Orlitzky, M., Schmidt, F.L., & Rynes, S. (2003). Corporate social and financial performance: A meta-analysis. Organization Studies, 24, 403-411.

Post, J.E. (1993). The greening of the Park Plaza Hotel. Family Business Review, 6(2), 131-148.

Pratt, M.G. (1998). To be or not to be? Central questions in organizational identification. In D. Whetten & P. Godfrey (Eds.), Identity in organizations: Developing theory, through conversations (pp. 171-207). Thousand Oaks, CA: Sage Publications.

Rawls, J. (1971). A theory of justice. Cambridge, MA: Harvard University Press.

Reiss, D. (1981). The family's construction of reality. Cambridge, MA: Harvard University Press.

Rosenblatt, P.C., de Mik, L., Anderson, R.M., & Johnson, P.A. (1985). The family in business. San Francisco: Jossey-Bass.

Schein, E.H. (1976). The first job dilemma: An appraisal of why college graduates change jobs and what can be done about it. In J.B. Ritchie & P.H. Thompson (Eds.), Organization and people (pp. 4-11). St. Paul, MN: West.

Schein, E.H. (1978). Career dynamics: Matching individual and organizational needs. Reading, MA: Addison-Wesley.

Schein, E.H. (1983). The role of the founder in creating organization cultures. Organizational Dynamics, 12(1), 39-46.

Schulze, W.G., Lubatkin, M.H., Dino, R.N., & Buchholtz, A.K. (2001). Agency relationships in family firms: Theory and evidence. Organization Science, 12(2), 99-116.

Stinchcombe, A.L. (1965). Social structure and organizations. In J.G. March (Ed.), Handbook of organizations (pp. 142-193). Chicago: Rand McNally.

Stinnett, N. & DeFrain, J. (1985). The secrets of strong families. Boston: Little Brown.

Tanewski, G.A., Prajogo, D., & Sohal, A. (2003). Strategic orientation and innovation performance between family and non-family firms. World Conference of the International Council of Small Business, Monash University, Caulfield East, Australia.

Trieschmann, J.S. & Gustavson, S.G. (1998). Risk management and insurance (10th ed.). Cincinnati, OH: Southwestern Publishing.

Waddock, S.A. & Graves, S.B. (1997). The corporate social performance-financial performance link. Strategic Management Journal, 18, 303-39.

Wartick, S.L. (2002). Measuring corporate reputation: Definition and data. Business and Society, 41(4), 371-392.

Wartick, S.L. & Cochran, P.L. (1985). The evolution of the corporate social performance model. Academy of Management Review, I0, 758-769.

Whetten, D.A. & Godfrey, P.C. (1998). Identity in organizations: Developing theory through conversations. Thousand Oaks, CA: Sage.

Whetten, D.A. & Mackey, A. (2002). A social actor conception of organizational identity and its implications for the study of organizational reputation. Business and Society, 41(4), 393-414.

Whetten, D.A. & Mackey, A. (2005). An identity-congruence explanation of why firms would consistently engage in corporate social performance. Working paper. Brigham Young University, Provo, UT.

Whetten, D.A., Rands, G., & Godfrey, P.C. (2001). What are the responsibilities of business to society? In A. Pettigrew, H. Thomas, & R. Whittington (Eds.), Handbook of strategy and management (pp. 373-410). London: Sage Publications.

(1.) There were 46 industries but only 59 firms in our family firm sample. We did regression analysis that controlled for industry and the preliminary results showed that industry did not significantly affect our findings. However, the lack of degrees of freedom made it difficult to clearly determine the impact of industry on CSP. Future work (e.g., paired comparison studies) should be done to further clarify the relationship between industry and social responsibility.

(2.) Inasmuch as image and reputation appear to account for favorable CSR we decided to see if the family name in the firm's title might have an even stronger impact on a family's wish for its firm to be seen as being socially responsible. Of the 59 family firms in our sample, 33 had the family name (or part of its name, e.g., Wal-Mart) in the title. We compared the mean social initiative and social concern scores using a t-test (and did some additional tests using regression analysis). However, the results of this analysis were not statistically significant, with the social initiative and concern means of those firms with the family name in the title being 2.16 and -1.71 and the means of firms without family titles being 2.46 and -1.31, respectively. The results, although not significant, are slightly in the opposite direction of what we might have expected.

Please send correspondence to: W. Gibb Dyer, Jr., tel.: (801) 422-2666; e-mail: w_dyer@BYU.edu.

W. Gibb Dyer, Jr. is the O. Leslie Stone Professor and Academic Director of the Center for Economic Self-Reliance at the Marriott School of Management, Brigham Young University.

David A. Whetten is the Jack Wheatley Professor and Director of the Faculty Center at Brigham Young University.

The authors would like to thank Alison Mackey, Nile Hatch, David Bryce, Jim Cbrisman, and two anonymous reviewers for their help on earlier drafts of this article.

Table 1
Social Responsibility Mean Scores for
Family and Nonfamily Businesses

                                     Family    Nonfamily     t score

Social responsibility initiatives     2.298      2.320       -.086
  Community                            .469       .525       -.627
  Diversity                            .733       .726        .060
  Employee                             .513       .479        .355
  Environmental                        .242       .358      -1.970
  Non-U.S. operations                  .038       .019       1.145
  Product                              .264       .162       1.677
  Other                                .035       .050       -.743

Social responsibility concerns       -1.547     -2.455       3.731 **
  Community                           -.044      -.079       1.913
  Diversity                           -.140      -.165        .768
  Employee                            -.166      -.343       4.399 **
  Environmental                       -.418      -.761       2.614 *
  Non-U.S. operations                 -.150      -.161        .230
  Product                             -.218      -.514       4.397 **
  Other                               -.411      -.432        .384

* p < .01: ** p < .001.

Table 2
Corporate Social Performance Initiative
Regressions

Variables                Model 1        Model 2        Model 3

Family business           -.26           -.05           -.09
                          (.26)          (.26)          (.27)

Firm size                  .46 ***        .62 ***        .61 ***
                          (.06)          (.06)          (.06)

Leverage                   .46            .34            .53
                          (.37)          (.39)          (.40)

Tobin's q                  .26 ***
                          (.02)

Return on sales                          2.19 ***
                                         (.53)

Return on assets                                        1.39 *
                                                        (.64)

Constant                 -3.04 ***      -4.46 ***      -4.25 ***
                          (.67)          (.69)          (.69)

Wald                    266.45         107.86          94.24

Prob < [chi square]        .000           .000           .000

* p <.05; *** p < .001.

Dependent variable = total initiatives.

Standard errors are in parentheses.

Table 3
Corporate Social Responsibility Concern
Regressions

Variables                Model 1       Model 2       Model 3

Family business          1.02 ***       .97 ***       .95 ***
                         (.28)         (.29)         (.29)

Firm size                -.37 ***      -.44 ***      -.44 ***
                         (.06)         (.06)         (.06)

Leverage                -1.10 ***      -.92 *        -.79 **
                         (.37)         (.38)         (.39)

Tobin's q                -.10 ***
                         (.01)

Return on sales                         .08
                                       (.52)

Return on assets                                      .85
                                                     (.62)

Constant                 1.47 *        1.92 **       1.79 *
                         (.68)         (.69)         (.69)

Wald                    95.09         67.63         69.82

Prob < [chi square]       .00           .000          .000

* p < .05; ** p < .01; *** p < .001.

Dependent variable = total concerns.

Standard errors are in parentheses.