«13TH INTERNATIONAL PUBLIC RELATIONS RESEARCH CONFERENCE “Ethical Issues for Public Relations Practice in a Multicultural World” Holiday Inn ...»
Cotrill concluded a positive association existed between CSR and market share. Also, Roberts (1992) tested the stakeholder theory (McGuire et al. (1988) explained that, “Stakeholder theory suggests that a firm must satisfy not only stockholders and bondholders, but also those with less explicit, or implicit, claims” (p.854)) using financial accounting measures to conclude, once again, that a positive association existed between CSR and financial performance. Finally, Cochran and Wood (1984) used accounting data across five years and compared it to three categories of socially responsible corporations (best, honorable mention, and worst) identified by Moscowitz (1975).
Cochran and Wood concluded that with operating earnings/sales as the financial performance measure, firms with “best” rankings outperformed “honorable mention” firms, which, in turn, outperformed “worst” ranked firms. Therefore, a positive association between CSR and financial performance was recognized.
However, research and studies on the association between CSR and financial performance have not always revealed a positive association, and it is important to address some prominent studies that have concluded that the relationship between CSR and financial performance has been identified as negative or nonexistent. One of the earliest and most commonly cited studies concluding that the relationship between CSR and financial performance is negative was Vance (1975). Vance used Moskowitz’s (1975) social responsibility index in comparison to a percentage 255 change in stock prices. He concluded that a negative association existed between CSR and financial performance. Vance is one of very few researchers to find a negative association.
More studies have found that no association existed at all than that a negative association existed as posited by Vance (1975). For example, Bowman and Haire (1975) performed a content analysis, examining the annual reports of 82 firms in the food processing industry between 1969 and
1973. By examining the amounts of attention given to CSR in the firms’ annual reports, they divided the firms into upper, middle and no mention of CSR categories. Then the groups were compared to returns on equity percentages. Bowman and Haire (1975) concluded that the mean return on equity for firms in the middle group was 14.3 percent, while the mean return on equity for firms with no mention of CSR was 9.1 percent. Therefore, Bowman and Haire posited that at least a moderate amount of involvement in socially responsible programs increased financial performance. However, Bowman and Haire were unable to conclusively show that the upper-grouped firms produced a higher mean return on equity than firms in the middle group.
Likewise, Alexander and Buchholz (1978) used Moskowitz’s (1975) social responsibility index in comparison to market-based returns to conclude that “no significant relationship exists between social responsibility ratings and market-based returns” (Pava and Krausz 1995, 156), and Freedman and Jaggi (1982) examined the Council on Economic Priorities air and water pollution measures in comparison to various financial accounting measures to conclude that, “In general there is no association between pollution measures and financial performance” (Pava and Krausz 1995, 157).
While the majority of studies in CSR and financial performance have indicated a positive association, the literature clearly lacks an examination of purchase intentions in relation to CSR. It is important to recognize the relationship between consumers’ purchase intentions and organizations’ involvement in socially responsible programs in order to both fill this gap in the literature and position CSR activities as a strategic management function of public relations.
The Theory of Reasoned Action
In order to predict the importance CSR plays in consumer’s purchase intentions, it is necessary to review some basic information and the relevant research and studies in behavioral theory. Specifically, this study is foundationally based on Ajzen and Fishbein’s (1980) theory of reasoned action; however, the theory is complex and controversial. Therefore, it is important to specifically address attitudes as they are capable of predicting behavior through the theory of reasoned action.
Historically, attitudes have been assumed to be direct predictors of behavior, thus the emergence of attitude scales such as the Thurstone (1931), Likert (1932), Guttman (1944) and Osgood (1957) scales. After a review of the current literature, Allport (1935) concluded that, the social psychological study of attitudes has been one of the core areas of discipline for decades, and that, “No other term appears more frequently in the experimental and theoretical literature” (798).
Moreover, Armitage and Christian (2004) posit that although attitudes serve a great number of functions, attitudes as they serve to guide people’s behavior account for the vast majority of research on the topic. (1) LaPiere’s (1934) study is perhaps one of the first and most widely cited examples of discrepancy between attitudes and behavior. LaPiere toured the United States with a young Chinese couple during times of anti-Chinese sentiment. After visiting 250 different establishments, the group was denied service on only one occasion; however, when LaPiere subsequently wrote the same establishments, 118 of the 128 replies claimed that they would not accept members of the Chinese race as guests in their establishment. LaPiere concluded, “Only a verbal reaction to an entirely symbolic situation can be secured by the questionnaire. It may indicate what the responder would actually do when confronted with the situation symbolized in the question, but there is no assurance 256 that it will. […] If social attitudes are to be conceptualized as partially integrated habit sets which will become operative under specific circumstances and lead to a particular pattern of adjustment they must, in the main, be derived from a study of humans behaving in actual social situations. They must not be imputed on the basis of questionnaire data” (236-7). Likewise, Whitlow (1935) used questionnaires to predict the behavior of approximately 600 high school students over a six-year period. Whitlow concluded, “A particular student may behave in a way directly opposed to his attitude; or with respect to a particular item of conduct the behavior of the students in the aggregate may contradict their attitude. In other words, a professed attitude is not a criterion for predicting behavior” (491).
Following LaPiere (1934) and Whitlow’s (1935) conclusions, many researchers shifted their focus and concentrated on developing attitude measurements and theoretical positions on the measurement of attitudes, as they were capable of predicting behaviors. Perhaps, the most prominent of these studies is Wicker (1969). Wicker (1969) examined 42 studies (not necessarily using the theory of reasoned action) to determine that attitudes only generally correlate with behaviors, r =.15, and that corresponding correlations rarely exceed r =.30 (Armitage and Christian 2004, 2).
Moreover, Corey (1937) assessed attitudes toward cheating using, what was considered at the time, a highly reliable measure of attitudes toward behavior. Corey (1937) found only r =.02 between attitude toward cheating and overt behavior.
Triandis (1967) concluded that, “there is a gap between those who are primarily concerned with the measurement of attitudes and those who have written theoretically about the measurement of attitudes. The former frequently rest their case after providing us with a single score, whereas the latter make a large number of theoretical distinctions but do not provide us with precise and standard procedures for measurement” (228). In so much, theoretical research was further split into examining potential moderators versus variables that might mediate the attitude-behavior relationship.
Baron and Kenney (1986) defined the difference between moderator and mediator measurements. A moderator variable “partitions a focal independent variable into subgroups that establish its domains of maximal effectiveness in regard to a given dependent variable” (1173).
However, mediator refers to a variable “which represents the generative mechanism through which the focal independent variable is able to influence the dependent variable of interest” (1173).
Armitage and Christian (2004) add that “as far as they are aware, only one variable has been investigated in this regard, namely, behavioral intentions” (4).
Ajzen and Fishbein (1980) stated, “Behavioral intentions are regarded as a summary of the motivation required to perform a particular behavior, reflecting an individual’s decision to perform a particular behavior, reflecting an individual’s course of action, as well as an index of how hard people are willing to try and perform the behavior. The idea that behavioral intentions mediate the attitude-behavior relationship represents a significant move away from the traditional view of attitudes: rather than attitudes being related directly to behavior, attitudes only serve to direct behavior to the extent that they influence intentions” (Armitage and Christian 2004, 4-5).
Thus, deriving from Fishbein’s extension of Dulaney’s theory of propositional control (1967), researcher Martin Fishbein introduced the theory of reasoned action in 1967 (See Dulaney 1967 and Fishbein 1967). Originally termed the Fishbein model, the theory of reasoned action is based on the assumption that human beings are usually quite rational and make systematic use of information available to them with the ultimate goal being the prediction and understanding of behavior. The theory is founded on the idea that the influence of attitude on behavior is mediated
through behavioral intentions, and that behavioral intention is a function of two basic determinants:
attitude toward the behavior and subjective norms. A person’s attitude toward a behavior is basically the individual’s positive or negative evaluation toward performing the behavior, and subjective norms are the individual’s perception of the social pressures from relevant referents to perform (or not perform) the behavior.
257 Ajzen and Fishbein (1980) posit that a person’s beliefs underlie a person’s attitudes. Ajzen and Fishbein add that, “Although a person may hold a large number of beliefs about any given object, it appears that he can attend to only a relatively small number of beliefs—perhaps five to nine—at any given moment. According to our theory, these salient beliefs are the immediate determinants of the person’s attitude” (63).
Therefore, according to the theory of reasoned action, the first step in predicting behavior is elicitation of salient beliefs, and because elicitation usually produces sets of beliefs that differ from respondent to respondent, Ajzen and Fishbein (1980) recommend eliciting beliefs from a representative sample of the population and selecting the most frequently elicited beliefs to create a modal set for the population, termed “modal behavioral beliefs.” Next, the strength of beliefs is measured through weighting. For example, respondents are asked to indicate the likelihood that performing the behavior will result in a given outcome. According to the theory of reasoned action “a person’s attitude toward a behavior can be predicted by multiplying their evaluation of each of the behavior’s consequences by the strength of their belief that performing the behavior will lead to that consequence and then summing the products for the total set of beliefs” (Ajzen and Fishbein 1980, 67).
Subjective norms, which beliefs are also considered to underlie, must also be measured in order to predict behavior. Ajzen and Fishbein (1980) recommend assessing not only a person’s normative beliefs (a person’s belief that a specific referent thinks he should or should not perform a behavior) but also, the person’s motivation to comply with each of his or her referents or the weight of those referents.
Finally, Ajzen and Fishbein (1980) state, “We have argued that with the aid of appropriate elicitation and measurement procedures, it is possible to predict a person’s attitude toward a behavior from a weighted sum of his or her beliefs about performing the behavior and to predict his or her subjective norm from a weighted sum of his normative beliefs. Since attitude toward a behavior and subjective norm are the determinants of intention, it should theoretically be possible to predict intention directly from the two sets of beliefs” (76). However, in order to predict behavior, Ajzen and Fishbein (1980) argue that certain conditions must be met: the set of behavioral beliefs must predict the attitude toward the behavior, the set of normative beliefs are predictive of the subjective norm and the attitude toward the behavior and subjective norm must be shown to predict the intention. (76) Additionally, Ajzen and Fishbein (1980) contend that in order for behaviors to be predicted using the theory of reasoned action, each component above must correlate specifically in terms of target, context, time and action. For example, “an individual’s attitude toward exercising (action), to get fit (target) in the gym (context) in the next week (time) should be more closely related to a measure of behavior designed to tap exercising to get fit in the gym in the preceding week, than (say) an index of fitness” (Armitage and Christian 2004, 3).
In order to better understand the theory of reasoned action, the following example is offered by Armitage and Christian (2004): “Consider the following example of Gary’s intention to use a condom. Gary’s mother might want her son to use a condom every time he has sex with a new partner, but Gary is only likely to do so to the extent that he is motivated to comply with his mother (very little in this case). Similarly, Gary’s latest partner also wants Gary to use a condom everytime he has sex with her; in this case, however, Gary is very motivated to comply with his new partner and therefore is more likely to intend to use a condom. Within the theory of reasoned action, both behavioral and normative beliefs are summed to produce global measures of attitude and subjective norm, respectively” (5-6).