The effect of participation in controversial business on stakeholders ’ management

The main objective of this research was to provide knowledge about the effect that participation in controversial business can have on primary stakeholder management. The panel data technique was used to estimate the models. The sample is composed of US firms and covers a five-year period from 2003 to 2007. In total the sample collects information of 2460 company years, representing 800 firms and an average of 3.07 years per firm. The findings reveal that participation in controversial business negatively affects primary stakeholders’ management. Thus, firms that are involved in controversial businesses do not have the legitimacy argument to invest on primary stakeholder management, unlike a company that is not involved in controversial businesses. However, these results can not be generalized to every type of controversial business, since the results allow us to observe the difference between each effect on primary stakeholder management and suggest that the controversial business variable is not a homogenous construct.


INTRODUCTION
The empirical evidence of an increasing number of studies that examine the effect of corporate social performance on financial performance, are mixed at best, due to differences in sample coverage, research designs, firms" performances and corporate social performance measures (Margolis and Walsh, 2003;Orlitzky et al., 2003;Wood, 2010).Furthermore, corporate social performance has no precise definition in the literature, which increases the difficulty of researching this field.Even though there is no precise definition, many researchers have adopted corporate social performance as a broad construct comprising primary stakeholder management and non-participation in controversial business (Social issue participation) (Hillman and Keim, 2001).Clarkson´s (1994) conceptualization of primary stakeholders was adopted, which defines them as those who

*Corresponding
author.E-mail: rosamariamouraleite@gmail.com.Tel: +55 67 3345-3568 "bear some form of risk as a result of having invested some form of capital, human or financial, something of value, in a firm" (1994: 5).To define non-participation in controversial business, Moura-Leite et al. (2011) concepttualization is used.Non-participation in controversial business is referred as social issue participation by Hillman and Keim (2001).Hillman and Keim"s terminology was not used in order to avoid semantic problems.Social issue participation refers to non-participation in certain activities that can be seen as undesirable or improper in society.Hillman and Keim"s terminology can mislead a casual researcher.For the purpose of this study, the terminology "participation in controversial business" was used, which means that firms do engage in businesses considered to be undesirable by a segment of society (Moura-Leite et al., 2011).The common forms of participation in controversial business may include: involvement in "sin" industries (alcohol, tobacco, gambling, firearms, the military weapons and nuclear power), which means for example: engaging in production, supporting these products or services and participating in the ownership of "controversial companies" (KLD, 2008).
To examine the effect of corporate social performance on financial performance, previous researchers have measured corporate social performance (primary stakeholder management and participation in controversial business) as only one construct (Ruf et al., 2001;Waddock and Graves, 1997) or as solely the primary stakeholder management construct (Berman et al., 1999;Bird et al., 2001;Doh et al., 2010;Kacperczyk, 2009;Padgett and Galan, 2010;Strike et al., 2006;Surroca et al., 2010).The participation in controversial business construct has therefore been analyzed in only a handful of studies (Hillman and Keim, 2001;Moura-Leite et al., 2011).
Besides the effect of firms" participation in controversial business on financial performance, some recent studies emphasize the effect of its participation on risk and return on publicly traded firms (Hong and Kacperczyk, 2009;Schwartz, 2003).Other researches address the effect of a specific controversial activities on society (Banerjee and Bonnefous, 2011;Byrne, 2010;Colle and York, 2009;Hancock et al., 2009;Jones et al., 2009;Palazzo and Richter, 2005;Pratten and Walton, 2009;and Cai et al., 2011) investigating the impact of stakeholder management on firm value for firms in controversial industries.To my knowledge, there is no research that addresses the effect that participation on controversial business has on firms" stakeholder management, even though it is an important topic, since understanding this effect, will allow a better comprehension of how controversial businesses affect primary stakeholder management practices (Cai et al., 2011), which will consequently reflect on business social impact (Schwartz, 2003).
The main objective of this research is to provide knowledge on the effect that participation in controversial business can have on primary stakeholder management.Accordingly, the stakeholder theory perspective and some insights from organizational legitimacy perspective were adopted.In addition, to better understand the effect that firms" participation in controversial business has on stakeholder management, we also analyzed the effect that each type of controversial activity has on primary stakeholders management, thereby allowing us to observe the difference between each effect on primary stakeholders management and suggest that the controversial business variable is not a homogenous construct.
This article adds to the efforts that researchers have made to contribute to the literature on how the effects of participation in controversial business have on firms" performances.There are, moreover, many practical concerns of interest to various collectives as a result of this study.For example, it can provide non-governmental organizations, governmental and regulatory institutions with an indicator that assists the explanation of social performance variation levels in companies involved in controversial business, and also can help improve tools designed to promote social responsibility activities in companies involved in this types of businesses.
Furthermore, this research can provide assistance regarding the selection criteria adopted currently for social investment portfolios.
The remainder of the paper is organized as follows: reviews of the literature, empirical evidence and the hypotheses are presented; description of the data and estimation method used in this analysis; the results are discussed and, finally, the conclusions.

THEORETICAL BACKGROUND
The stakeholder theory widens the focus of management further, using corporate resources not only for the sole benefit of shareholders, but for the benefit of a much wider group of stakeholders (Freeman, 1984).According to Donaldson and Preston (1995:68), stakeholder theory posits a model of the firm in which ""all persons or groups with legitimate interests in participating in an enterprise do so to obtain benefits, and there is no prima facie priority of one set of interests and benefits over another"".Hillman and Keim (2001) assert that from this theory´s perspective, corporate social performance is assessed in terms of a firm meeting the demands of multiple stakeholders who can constitute intangible, socially complex resources that may enhance firms" ability to outperform competitors in the long term.
In the same line, Berman et al. (1999) conclude that firms improve their social performance, by identifying and managing the impact on primary stakeholders, because there are discernible benefits associated with having a reputation for being a good corporate citizen.Having a good social performance, plays a role in attracting, retaining, and motivating employees, thereby increasing productivity (Ali et al., 2010;Backhaus et al., 2002;Riordan et al., 1997;Turban and Greening, 1997).It can also attract customers that may prefer to purchase and use products or services from companies that are known to be strong social performers (Brown and Dacin, 1997;Ruf et al., 2001;Senthikumar et al., 2011).Additionally, suppliers may choose to be associated preferably with strong social performers, and hesitate to associate themselves with weak social performers (Godfrey, 2005).Likewise, bankers may favour companies with a positive stakeholder evaluation (Scholtens, 2006).On the other hand, investors may prefer to do business with companies recording a strong stakeholder acceptation because their cash flows may be perceived to be less risky (Ghoul et al., 2011;Godfrey, 2005;Graves and Waddock, 1994;Hong and Kacperczyk, 2009).
Reviewing the literature mentioned in the last paragraph, it can be deduced that having a good stakeholder management pays off.From this point of view, firms that participate in controversial businesses could face difficulties with their stakeholders and their firm´s performance could be negatively affected (Hillman and Keim, 2001).For example, it is possible that firms that participate in controversial businesses could have more difficulty attracting, retaining and motivating employees, compared to firms that do not participate in controversial activities.Firm"s customer relationship can also be affected and sales could suffer from costumer boycotts.Additionally, suppliers may choose not be associated with a controversial company, since this could affect their reputation.Another possible difficulty is a negative standing with bankers and investors that may choose not to work with controversial companies.
While social actions for primary stakeholders normally aim at gaining public respect and good reputation, companies that participate in controversial businesses have to accept that they are fighting on a different legitimacy battlefield (Hillman and Keim, 2001).Suchman (1995:574) contends that ""legitimacy is a generalized perception or assumption that the actions of an entity are desirable, proper or appropriate within some socially constructed system of norms, values, beliefs, and definitions"".Attending to stakeholder expectations, usually aims at legitimizing a corporation"s activities and increases corporate acceptance (Hillman and Keim, 2001).
According to Campbell et al. (2003), a controversial company is, in the view of some constituencies, structurally illegitimate.And it is probable that legitimacy can never be attained in the view of some constituencies and here the objective cannot be to restore something which it never had.In such a case, the aim of actions directed towards stakeholders might be simply to convince society that they are not all that bad.Furthermore, it might come about that such firms are so structurally illegitimate that social actions would have no effective value because of the contempt in which they are held by some publics.

Participation in Controversial Business and Primary Stakeholder Management
A good corporate behaviour goes beyond the compliance with the legal rules of society.Corporations have to take into account the common good and improve societal welfare (Kok et al., 2001).Branco and Rodrígues (2007) assert that primary stakeholders" expectations of companies are inextricably intertwined with society"s views or expectations.Controversial businesses impact negatively on general society (Hillman and Keim, 2001), subsequently controversial companies have a distinct primary stakeholder management motivation when compared to companies with no participation in controversial businesses.
A primary stakeholders" management preserves societal acceptance (Weaver et al., 1999).Therefore, corporate legitimacy has been described as the yardstick for the primary stakeholder management (Wartick and Cochran, 1985).For example, according to Ashforth and Gibbs (1990), tobacco industry finds itself on the lowest level of public acceptance and the lower the perceived Moura-Leite and Padgett 11955 legitimacy of a corporation the more skeptical its legitimization attempts will be observed by its relevant publics.In addition, Palazzo and Richter (2005) analysis demonstrates that mainstream corporate social responsibilities efforts will hardly contribute to legitimize tobacco companies.They affirm that tobacco companies see their social responsibility efforts exposed to a much greater scrutiny and a much higher level of negative expectations than others companies.They also demonstrated that the negative expectation works against mainstream social responsibilities aspects and they are partly confirmed by the seeming continuation of former patterns of behavior.
Likewise, the decision to avoid participation in controversial businesses for firms like "Body Shop", "Ben and Jerry"s", and "Johnson and Johnson" is related to their corporate social strategies and they are clearly doing so with the conviction that it is instrumental for the firms" economic performance, as it gives them a com-petitive advantage (Barney et al., 2001).In other words, by avoiding participation in controversial business they obtain a positive impact on competitive advantage and become a form of instrumental stakeholder management (Moura-Leite et al., 2011).
Based on these evidences, it is assumed that highly controversial businesses harming the common good are probably more significant than the benefit to some stakeholders.As a consequence, corporations that operate in those businesses have least motivation to invest on primary stakeholder management.Therefore, the following hypothesis is proposed: Hypothesis 1: Participation in controversial business negatively affects primary stakeholder management.
In view of the fact that firms need to consider social norms because they influence primary stakeholder expectations.Byrne (2010) asserts that the degree of unethical violation of social norm depends on the type of controversial business.It is relevant to observe the diverse effects between the controversial businesses.There are some few previous researches which address this topic.Schwartz (2003) examined the characteristics of some controversial activities and appoints that the arguments vary to classify an activity as controversial or unethical.Ghoul et al. (2011) results show that participation only in some controversial activities increases firm"s cost of equity.
The predominant controversial business adopted by social/ethical investment organizations" will be briefly reviewed, in the next paragraphs.The intention is not to provide an in-depth ethical analysis of each controversial activity, which would be beyond the scope of this paper.Instead the intention is merely to raise some business characteristics that may affect firms" behaviour.
The tobacco and alcoholic drinks businesses are associated with highly visible social issues.They are thought to produce large social externalities, since both are potentially addictive and dangerous (Brammer and Pavelin, 2004).In relation to tobacco, according to National Institute on Drug Abuse (NIDA) of the United States of America ( 2009), its use is the leading preventable cause of death in the country.In addition, the health care costs in 2008 attributable directly to smoking are more than $96 billion.It is important to point out that this total is well below the total cost, since it does not include burn care form medical care costs associated with diseases caused by secondhand smoke and the cost of loss in productivity due to smoking effects.In addition to these impacts it is relevant to stress that tobacco nicotine i is an addictive substance, which generates compulsive drug seeking and abuse, even in the face of negative health consequences (NIDA, 2009).
Like cigarettes, alcohol is potentially addictive and has been associated with many types of diseases (World Health Organization, 2012).Alcohol has also been connected with other related social problems such as drunken driving, birth defects, sexual assault, and alcoholism-related abuse.However, unlike tobacco, alcohol does not necessarily kill the user, or even shorten the user"s life.Used in moderation however, Colle and York (2009) defend that some types of alcohol may actually lead to improvements in one"s health, especially to heart disease and high cholesterol-related conditions, and highlight the positive effects of culture and value of the winemaking business in traditional producing countries.According to Schwartz (2003) despite these counter-arguments, in general the provision of alcohol products is socially controversial, since it can bring social problems.
Gambling is another activity that many argue is controversial.For many people it is simply a hobby which combines an element of risk with the possibility of financial reward, however, for a few, it may lead to debilitating problems that harm themselves and their society (Young, 2006).Gambling can become a severe addiction and it results can include financial ruin, abuse, crime and suicide (Pratten and Walton, 2009).The casino industry, has begun to introduce responsible gaming measures to diminish the negative social impact of gambling, but according to Pratten and Walton´s (2009) research conclusions, much more could be done.Hong and Kacperczyk (2009) asserts that hence, someone could argue that gambling has become more socially acceptable in the recent years, there are plenty of American surveys that indicate that public opinion still regards gambling as sinful behavior.
Related to military and firearms the controversy is based on the concern over offensive or aggressive weapons (Schwartz, 2003).Weapons put into the wrong hands can lead to great human suffering (Byrne, 2010).Arms proliferation breeds violence and fresh demands for additional weaponry, encouraging internal and inter-state conflict.In addition, weapons expenditures limit social and economic investment options in many countries, especially on the poorest ones (Bourne et al., 2004).However, Schwartz (2003) emphasizes that there are some circumstances where the use of weapons is a strategic instrument of national policy that has worked in various countries well in peace and war.In these cases, the companies" involved in this controversial business is in fact providing goods of critical importance to the wellbeing, safety and security.Nevertheless, military and firearms can be seen as positive only in these cases.
Another polemic business is nuclear power; it is one of the most controversial sources of energy due to its affiliation with nuclear weapons, the danger of nuclear radiations, the difficulty of disposing of radioactive waste and the possibility of nuclear accidents (Hodgson, 1999).Nuclear accidents, such as experienced during an earthquake and tsunami in Fukushima -Japan in March 2011, increased resistance of the use of nuclear energy.Nonetheless, many countries depend of nuclear power plants to fulfill their electric power demand in their countries.Nuclear energy provides 19.6 percent of the United States' electricity and is its number one source of emission-free electricity (Nuclear Energy Institute, 2012).
Given the aforementioned diversity of controversial business characteristics the following hypothesis is proposed: Hypothesis 2: The impact that business participation in each controversial activity has on primary stakeholders management differs among each controversial activity.

Data and sample
The data were drawn from two sources.Control data were taken from the Thomson Datastream, while the stakeholder management and participation in controversial business data were provided by Kinder, Lydenberg, and Domini (KLD) Research and Analytics, Inc. KLD Stats database use screens to monitor corporate social performance.Positive screens indicate a firm"s strengths and negative screens indicate its weaknesses or concerns.Each screen can be summarized in a binary variable, which reflects whether the firm meets that particular criterion.The screens are summarized in groups of corresponding items referring to a general theme.Seven themes are identified: community, corporate governance, diversity (to proxy for minorities), the natural environment, human rights, employee relations and product quality (to proxy for customers).In addition to the seven major issue areas, KLD data provide information on participation in controversial business issues, which include involvement with alcohol, gambling, firearms, the military, nuclear power and tobacco.
The sample composed of US firms is unbalanced and covers a five-year period from 2003 to 2007 ii .After merging both databases, the sample consists of 1543 firms and 5032 observations.500 firms and 1599 observations were discarded because of missing values related to the control variable R&D investment.Also, 234 firms and 933 observations were discarded because a 1-year lag was used on the CBI measure, the independent variable, and control variables.In addition, 9 firms and 40 observations were discarded when building the industry control variable.The final sample, therefore, is a panel data of 2460 company years, representing 800 firms and an average of 3.07 years per firm.The significant differences between the larger sampling frame and the final sample was tested and did not find any significant difference.

Measures
To measure primary stakeholder management, the dependent variable, the five KLD dimensions consistently reported between 2003 and 2007 were used: community, diversity (to proxy for minorities), the natural environment, employee relations and product quality (to proxy for customers).These dimensions have been selected because they reflect corporate attention to primary stakeholders with an impact on a firm"s survival (Clarkson, 1995) and exert a considerable influence on corporate strategy (Berman et al., 1999).
KLD"s five dimensions representing primary stakeholder management are designed as a binary system, which indicates the presence or absence of strengths and concern regarding numerous attributes of each firm"s social actions.For each strength or concern, a score of 1 indicates the presence of this attribute in the dimension and 0 indicates its absence.All the strengths in each dimension are rated on a scale ranging from 0 to +2, whereas all the concerns in each dimension are rated from -2 to 0. Then the strengths scale to the concerns scale for each dimension was added, with the result being a new scale ranging from -2 to +2.The primary stakeholder management measure was built giving equal weights to the five dimensions cited above (Hillman and Keim, 2001;Padgett and Galan, 2010).
In order to measure non-participation in controversial business, the explanatory variable for testing Hypothesis 1, the KLD dimensions concerning firm operations related to alcohol, tobacco, gambling, firearms, nuclear power and the military were used (for a breakdown of controversial business issues, please refer to the appendix A).These controversial business issue ratings differ from the stakeholder management ratings described earlier.The only type of rating for these issues is a concern rating.For each concern, a 1 is given to indicate the involvement of the concern and 0 to indicate avoidance; also each category adopted from the KLD is given equal importance to construct the measure of participation in controversial business.For testing Hypothesis 2, we used each category adopted from KLD (alcohol, tobacco, gambling, firearms, nuclear power and the military) separately.
Control variables were used based on the degree to which they may influence the effect that controversial business involvement have on primary stakeholder management.Company size was used because previous articles have suggested it is closely related to stakeholder management (Udayasankar, 2008;Waddock and Graves, 1997), and it is measured as the number of employees (Berman et al., 1999;Padgett and Galan, 2010;Surroca et al., 2010), defined on a log scale.Risk is another factor used as a control variable, since several studies have found that firms with proactive stakeholder management (Wood, 1991) tend to anticipate and reduce potential sources of business risk.β (beta) -a standard indicator of market-based risk (Hillman and Keim, 2001;Surroca et al., 2010) was adopted.In addition, Tobin"s Q is used as a control variable to financial performance effect on stakeholder management, mostly because of its ability to capture the value of long-term investments (Surroca et al., 2010).
Furthermore, McWilliams and Siegel (2000) show that investment in research and development (R&D) and advertising intensity is positively correlated with stakeholder management performance.To measure R&D, a proxy is used, calculated by dividing total expenditure in R&D by total sales (Padgett and Galan, 2010).To measure advertising intensity, in line with Brammer and Pavelin (2006), a dummy variable was constructed, on the basis of the company"s presence in two thematic lists, "100 leading national advertisers" and the "most valuable brands".Both lists are available online (Interbrand, 2009;Advertising Age, 2009).
Previous research posits the importance of the nature of a firm"s Moura-Leite and Padgett 11957 industry on its stakeholder management (Brammer and Millington, 2008;McWilliams and Siegel, 2000).However, there is a great deal of overlap between industry membership and a firm"s participation in controversial business.To readdress this concern, it was decided to test run the regression models with and without industry in the control variables.Industry has been operationalized in this paper by industry dummy variables measured according to the Standard Industry Classification (SIC) codes of Waddock and Graves (2007)

Estimation method
The panel data technique was used to estimate the models.Those were initially specified using the fixed-effects estimator.A Hausman test was carried out, which indicated correlation between individual effects and independent variables, which show that this fixedeffects model is better, suited than the between-effects one.As well as the individual effects added to control for the cross-reference units, also time dummies were included in the model.These temporal effects enable us to reduce a source of bias by capturing the events that all states were subjected to in a given year.An F test of significance was performed to assert the joint significance of the temporal dummies, and the result was that they contribute to the model overall"s significance.Furthermore, the Wooldridge and Modified Wald tests were applied to examine potential autocorrelation in the panel and heteroskedasticity problems in the fixed-effects equation.The result was positive in all cases.Consistent with Beck and Katz (1995), we have corrected both problems using panel corrected standard errors through a Prais-Winsten iv regression.Assuming the long-term effect of controversial business participation on stakeholder management (Cai et al., 2011), the models were estimated with a one-year lag on independent and control variables.Finally the models were estimated with a two-year lag on CSP variables and found a similar result.

RESULTS
Table 1 provides descriptive statistics and correlation matrices for the variables.It can be observed a significant correlation between the control variables Tobin"s q, employees, beta, R&D, advertising intensity, and the primary stakeholder management variable, participation in controversial business variable and most of the disaggregated participation controversial business variables (exception R&D).It can also be seen, as predicted by the literature, a significant and negative correlation between primary stakeholder management and participation in controversial business, but the correlation was not very high.
Table 2 presents the results of the regression analyses using an aggregated measure of participation in controversial business as explanatory variable for testing Hypothesis 1, while Table 3 shows the results when the disaggregated measures of involvement on controversial business was adopted for testing Hypothesis 2. The Model 01 column in Table 2 presents all of the control variable results for primary stakeholder manage-ment as dependent variable.Consistent with prior research, Tobin"s q was significant and positively related to Primary stakeholder management at p<0.001.Firm size was significant and positively related to Primary stakeholder management at p<0.001.Advertising intensity was significant and positively related to Primary stakeholder management at p<0.001.Nevertheless, a significant and positive effect of firms" size and R&D on Primary stakeholder management was expected, but the results demonstrated that both are not significant.The Model 03 column in Table 2 also presents all the control variable results for primary stakeholder management as dependent variable, without dummy industry included.
The Model 02 column in Table 2 presents Participation in controversial business, as the exploratory variable controlling for all the control variables.Participation in controversial business was significant and negatively related to Primary stakeholder management at p<0.001, and Model 04 column in Table 2 presents Participation in controversial business, as the exploratory variable controlling for the control variables without dummy industry included, the results for the Participation in controversial business variable were similar to Model 02, although at p<0.05.These results support Hypothesis 1, which shows that Participation in controversial business has a negative effect on Primary stakeholder management.
Table 3 presents disaggregated participation in controversial business variables, as the exploratory variables controlling for all the control variables in Model 01 column and without industry dummies included in Model 02.In Model 01 the Tobacco and Military business participation variables were significant and negatively related to Primary stakeholder management at p<0.001 and p<0.05, respectively; in addition, Gambling, Firearms, Alcohol and Nuclear power had no significance.In Model 02 only Tobacco business participation variable was significant and negatively related to Primary stakeholder management at p<0.001.These results support Hypothesis 2, which affirm that the impact that business participation in each controversial activity has on primary stakeholders management differs among each activity.

DISCUSSION
The purpose of this paper has been to provide knowledge on the effect that participation in controversial business can have on primary stakeholder management.To achieve this objective the stakeholder theory and some insights from organizational legitimacy perspective were adopted to support the first hypothesis, which  states that the participation in controversial business negatively affects primary stakeholder management.In addition, to improve the under-standing of the effect that firms" participation in controversial business has on stakeholder management, the effect that each type of controversial activity has on primary stakeholders" management was examined.
The results support the first hypothesis which states that participation in controversial business negatively affects primary stakeholders" management.Whereas involvement in controversial business generates negative social expectations, primary stakeholder"s management generates and preserves social acceptance (Hillman and Keim, 2001).Controversial company is, in the eyes of some constituencies, structurally illegitimate (Campbell et al., 2003).Subsequently the firm that is involved in controversial business does not have the legitimacy argument to invest on primary stakeholder management, unlike a company that is not involved in controversial business.
These results can not be generalized to every type of controversial business.On the second hypothesis results, it was found that not all controversial activities effect primary stakeholder management in the same way.This result suggests that the controversial business variable is not a homogenous construct.Furthermore, the results demonstrated that the tobacco and military business involvement variables are significant and negatively related to primary stakeholder management, when industry dummies are included in the model.When industry dummies are excluded from the model, only the tobacco business involvement variable is significant and negatively related to primary stakeholder management.Thus, involvement in the tobacco business negatively affects primary stakeholder management.This can be to a certain extent explained by the lower incentive that these firms have to attend their primary stakeholders.Palazzo and Richter (2005) analysis demonstrated that tobacco companies social responsibility efforts are exposed to a much greater scrutiny and a much higher level of negative expectations than other companies, and conclude that mainstream corporate social responsibilities efforts directed to firms stakeholders, will hardly contribute to legitimize tobacco companies.
Besides, the non-significant effect of alcohol, firearms and nuclear power activities on the results can be explained, in line with Brammer and Pavelin (2005) arguments, that some actions directed towards primary stakeholders may mitigate the negative reputational consequences of involvement in controversial business, thus, the involvement on these controversial activities does not affect firm"s primary stakeholder management in a negative way.
Similar to other empirical studies, this research has limitations that potentially restrict the generalizability of the results.The sample consists of only US firms; consequently, the results are logically restricted to the US market.Further research should use data from firms in other geographic locations, since the country, culture and values are key factors in the relationship between controversial activity and most measures of firms" performance (Schwartz, 2003).Thus, a comparative study could help to explain the impact of local characteristics in this relationship.
Furthermore, there are many other areas that could be considered controversial (Schwartz, 2003), since this research is limited to the data provided by the KLD database, due to the difficulty of accessing any other database that provides this information.Future researches should also focus on exploring with more details the relationship of each type of controversial business with primary stakeholder management, since it was found that they do not have the same effect.
In terms of academic contribution, this article adds to the efforts that previous researchers have made to participation in controversial business literature.The findings contribute providing some preliminary empirical evidence on the negative effect of participation in controversial business on primary stakeholder management in US firms.In addition, the results demonstrate that these negative effects differ by type of controversial activity.There are, moreover, many practical concerns of interest to various groups as a result of this study.For example, it can provide non-governmental organizations, governmental and regulatory institutions with an indicator that assists the explanation of social performance variation levels in companies involved in controversial business, and also can help improve tools designed to promote social responsibility activities in companies involved in this types of businesses.Furthermore, this research can provide assistance regarding the selection criteria adopted currently for social investment portfolios, since these results suggest that not all controversial companies may have a weak primary stakeholder management.

Table 1 .
Descriptive statistics and correlations matrix a .