The role of self-interest in motivating human behavior has been extensively discussed throughout our intellectual history, with most major theories of human behavior touting it as the primary motivation behind human endeavors. The purpose of the present paper is to review how self-interest has been investigated in the social scientific literature and reveal several peculiar factors associated with this important construct. First, this paper will discuss how self-interest is widely accepted as a primary motivation for human behavior. A major reason for this wide acceptance may be how broadly self-interest is defined in different fields. As such, the various definitions of self-interest are presented. Then, the ways in which major theories in evolution, economics, political science, and psychology all hold self-interest to be the main driving factor in human behavior are discussed. Second, this paper will review empirical research demonstrating that despite the wide acceptance of self-interest as the primary motivator behind human behavior, the actual effect of self-interest on policy evaluations has a rather limited effect. Third, research suggests that people hold ambivalent feelings about the propriety of self-interested thinking, not always embracing it as an appropriate motive when given the opportunity. Thus, although self-interest is widely touted as the primary motivator behind all of human behavior, the literature reveals a complicated, nuanced picture of the psychology of self-interest, calling for a more rigorous discussion of definition, theory, and research on when and how self-interest operates, and on attitudes towards this ever important construct. Finally, theoretical implications and suggested future directions are discussed.
As stated, most academic and lay theories of human behavior hold that self-interest is the primary motivator behind all behaviors. Part of the reason for this widespread acceptance may be that self-interest is often broadly defined to include wide-ranging psychological processes that are arguably distinct from one another. This paper addresses three perspectives of self-interest that are commonly employed in the social sciences: 1) the motivation to maximize material resources, promote one's health and to minimize material harm and avoid harm to one's health, 2) the hedonistic motive to pursue feeling good and avoid feeling bad, and 3) the psychological impact of being involved in a particular outcome. See Sears and Funk (1991) for a comprehensive review of the operationalization of the self-interest construct.
The most commonly used definition of self-interest is the motive to maximize material resources and to minimize harm to one's wealth and health. When examining the literature, this is the definition of self-interest most commonly adopted by economists, psychologists, and political scientists (e.g., Mansbridge, 1990; Sears & Funk, 1991; Sen, 1977; Thaler, 2000). Though this definition appears to be straightforward, it encompasses a variety of behaviors that may not at first appear to be self-interested. For example, a resident may oppose paying a local tax designed to raise revenue for schools because he does not want to lose any money and because he does not have any children in the school system. However, if the resident supported the local tax because he feels he would personally benefit from an improved education system (e.g., because educated children may attract businesses to his community), then that behavior, too, would be considered self-interested. Another example could be quitting smoking in order to benefit one's health. Although quitting smoking is a long and painful process for most smokers, the motivation to improve one's health would make that behavior self-interested according to this definition of self-interest.
A second approach can be described as hedonism, the desire to promote pleasurable psychological states and to avoid negative affect. Such a definition means that any behavior that makes one feel subjectively good can be classified as self-interested, even if it does not result in a change in material wealth or affect one's health. Thus, smoking cigarettes and engaging in risky sexual behaviors can be construed as self-interested according to this perspective because of the pleasurable feelings they induce, even though they come with obvious expenses with respect to one's health. Similarly, the taxpayer who wants to pay taxes simply because it makes him feel good to help society would be self-interested because of the hedonistic benefits he accrues. Yet another behavior that is included in this definition is engaging in prosocial, or helping behaviors. Although at first glance self-sacrifice appears to be inherently selfless if the behavior results in a feeling of satisfaction on the part of the helper, many have argued that the behavior is self-interested (Cialdini,1991). According to this approach, donating to a charity is still considered a self-interested behavior because of the positive affect that results in donating, even though the behavior results in a material loss in wealth.
A third definition of self-interest in the psychological literature is known as personal involvement. Research and theory in social psychology holds that anything related to the self is regarded as special to the perceiver, and that the self exerts unique effects on cognition (see Greenwald, 1980). Attitudes that pertain to the self (attitudes about one's self, attitudes toward objects related to the self) tend to be especially strong and important to the attitude holder because of the potential consequences that such attitudes have for self-esteem (Greenwald, 1980; Krosnick, 1990). Personal involvement, or personal relevance, occurs whenever an attitude object has a direct effect on one's own life. For example, if a proposal to lower the drinking age to 18 were introduced, then all of those who are between the ages of 18–21 would be personally involved with that proposal because it would directly affect their lives. Research on dual process theories of attitude change demonstrates that personal involvement with an attitude object significantly biases information processing such that perceivers exert more effort when evaluating messages and pay closer attention to the content of arguments. Thus, according to this approach to self-interest, people are automatically considered to be self-interested if they are personally involved or affected.
As can be seen, these three perspectives on the self-interested motive are fairly diverse, probably contributing to the perception that all behaviors are self-interested. It is easy to see how all behaviors can be called self-interested if one is willing to change the definition of self-interest to explain a particular behavior. Sears and Funk (1991) note that such a broad definition of self-interest makes it nearly impossible to scientifically test; if self-interest is so broad as to include all motives, then all behaviors can be explained in terms of self-interest and the theory becomes impossible to falsify. At that point, the question of whether self-interest influences behaviors ceases to become a scientific one. Thus, a common definition of self-interest should be more thoroughly discussed and adopted among researchers to ensure shared understanding about this all-important construct.
The Primacy of Self-Interest in Explaining Human Behavior
Self-interest, in all of its incarnations, is widely accepted by most people as the ultimate motivation for all human behavior. According to basic principles of evolutionary fitness, the driving factor for all human behavior is to garner advantage over the environment (including the social environment) in order to maximize odds of survival and the propagation of genes (Darwin, 1859; Dawkins, 1976). Behavior driven by this motivation can be construed as self-interested because it results in accruing material advantage and resources. The classic example of evolutionary fitness involved the group of finches that Darwin observed and collected from the Galapagos Islands. Darwin (1859) theorized that though they differed in appearance, the birds must all have once been part of one species of bird that subtly changed over the years to adapt to their differing environments. Genetic mutations caused minor differences in appearance and those differences that allowed the organism to successfully secure food and therefore procreate were passed on genetically through offspring. This basic tenet, that organisms must garner an advantage over the environment, has inspired explanations for biological adaptations and social-scientific explanations for human social behavior that are the foundation for how self-interest drives human behavior.
For instance, evolutionary psychologists have examined how mate preferences differ between the two sexes as a function of maximizing the odds of procreation and survival. Women are said to prefer men who have the resources to take care of them during and after the gestational period, whereas men are said to prefer women who exhibit the physical characteristics associated with the potential to successfully bear children (e.g., large breasts, wide hips; Buss, 1989; 2004). Moreover, both men and women prefer mates who are physically symmetrical to reduce the odds of genetic abnormalities in offspring, and this preference is exaggerated for women when they are ovulating and most fertile (Gangestad & Thornhill, 1998). Thus, according to evolutionary approaches to human sexual behaviors, mate selection functions as a desire to propagate one's genes and to further ensure advantage over the environment.
As stated earlier, evolutionary psychologists have also examined the phenomenon of prosocial (i.e., helping) behaviors, which at first glance appear to be selfless. Evolutionary psychologists have pointed out that helping behaviors may improve helpers' wealth and health (see Dovidio & Penner, 2001). Human cultures around the world have incorporated norms of prosocial behaviors such that sacrificing one's well-being for the sake of others is expected; thus, according to this perspective helping someone else benefits one's own self-interest because it enhances one's status in a group for following the group norm to help (e.g., Cialdini, 1991). Moreover, the universal norm of reciprocity dictates that the person who has been helped is obligated to return the favor at some point in the future. Thus, although helping someone else may be costly in the short run, it may be tangibly beneficial in the long run because it may mean higher status in a group for following norms and helping, higher odds of successfully passing on genes, and because it invariably means that someone in your group owes you a favor in the future.
Cialdini and colleagues (Cialdini, 1991; Cialdini, Brown, Lewis, Luce, & Neuberg, 1997; Neuberg, Cialdini, Brown, Luce, & Sagarin, 1997) have also pointed out that helping behaviors serve to promote one's hedonistic self-interest. Violating social norms to help someone is distressing (i.e., normative distress; Cialdini, 1991), and helping someone in need eliminates such distress. Similarly, seeing someone in need arouses negative affect (i.e., reflexive distress, Cialdini, 1991) in the perceiver, and helping the target reduces this negative arousal. Complementing this process is the boost in positive affect that one experiences when helping another person in need. Additionally, Cialdini and his colleagues have proposed that helping someone in need is affected by perceived oneness, or self-other overlap (Cialdini, et al., 1997; Neuberg, et al., 1997). In such conditions, helping the other person is akin to helping one's self because of the perception of a shared identity.
Traditional theories of economics adopt the more generic definition of self-interest; traditional theories of economics assume that people are rational actors who are motivated to maximize their utility, or material wealth. In fact, traditional theories of economics equate rationality with the hyperlogical pursuit of material wealth (see Sears & Funk, 1991; Sen, 1977). According to this approach, people are characterized as homo economicus: fundamentally self-interested, pursuing tangible material rewards and acting to maximize gains and minimize losses. In fact, that people act out of self-interest is the basic premise of microeconomic explanations for all human behavior (within certain real-life constraints, e.g., Becker, 1992) and for macroeconomic explanations for the larger market (see Ostrom, 2010).
Moreover, traditional theories of political science explain voting as an avenue by which citizens maximize their material interests (Downs, 1957). The role of self-interest as a primary motivator has a long and rich history in European and American political science, with many political theorists regarding the conflict between individual interests and collective needs as the most basic problem that a democracy faces (see Mansbridge, 1990). Many political scientists have adopted a rational choice model of political science based on self-interest: self-interest is regarded as the basis for human behavior, and voting is an avenue by which citizens express self-interested concerns. According to this model, citizens vote for policies that materially help them and vote against policies that cause them material harm.
Although self-interest is widely accepted as the primary motivating force behind people's preferences and behaviors, there have been enough observed vagaries in human behavior to warrant modifications to the idea that self-interest alone affects behavior. With respect to helping behaviors, Batson and his colleagues (e.g., Batson, 1997; Batson & Shaw, 1991; Batson, Turk, Shaw, & Klein, 1995; Batson, Sager, Garst, Kang, Rubchinsky, & Dawson, 1997) have argued for the existence of true altruism in which people help others for purely selfless reasons, which is in stark contrast to Cialdini and colleagues' more egoistic and hedonistic arguments (see above). The precipitating event for true altruism, they argue, is experiencing empathy, which is similar to feelings of compassion and tenderness for another person, and which motivates the perceiver to help the target. According to empathy-altruism hypothesis, when a perceiver sees a target in need, s/he can experience empathy which, in turns, prompts the perceiver to help the target solely for the purpose of enhancing the target's welfare. In this ongoing debate, both camps concede that prosocial behaviors (like any behavior) are likely caused by multiple motivations (Batson & Shaw, 1991; Cialdini, 1991). A promising avenue for future research is one proposed by Loewenstein and Small (2007) to address how people think about victims. According to this framework, perceivers can experience sympathy upon seeing someone in distress, which is an automatic and somewhat immature response. Helping as a function of sympathy may be more hedonistic because it is automatic and the purpose is to reduce negative affect or promote positive affect. By contrast, they argue that a second response, a slower, more effortful response (“deliberation,” p. 112) results in helping that is more likely purely altruistic. Thus, it is possible that helping behaviors can be characterized as selfish or altruistic, depending on the thought process and the emotional process that preceded the helping event.
In the field of economics, there have been several approaches to explain imperfectly self-interested behavior. First, Kahneman and Tversky's (1979) work on prospect theory, heuristics and biases, led to some questioning about the rationality of the purely rational actor which, as stated, was assumed to be self-interested (Kahneman, 2003a, 2003b; Rabin, 1998; Rabin & Thaler, 2001, Thaler, 2000). Whereas economists traditionally assume that each person has stable preferences that are consistently expressed to maximize utility (i.e., self-interested concerns to maximize material wealth and health), research in psychology underscored a need to rethink this model of human choices and behavior (Rabin, 1998). In particular, research revealing that people value gains and losses differently (and therefore behave differently – more or less risky – when considering a gain or a loss), could not be explained by economists' traditional concave utility of wealth function (Rabin & Thaler, 2001). Rather, a psychological explanation by which people evaluate outcomes in relation to a reference point was more powerful. Losses threaten the status quo, so they are valued more than gains which the perceiver already knows s/he can get by without (Kahneman & Tversky, 1979; Rabin & Thaler, 2001; Tversky & Kahneman, 1991). Similarly, research in cognitive psychology has revealed a host of other biases in judgment (and subsequent behavior) that deviate significantly from perfect rationality (see Kahneman, 2003a, 2003b; Rabin, 1998). For example, whereas rational choice models assume that individuals have perfect knowledge off of which to make self-interested judgments and decisions, research on the availability heuristic strongly suggests that people do not use base rate information as rational choice models assume that they should. Rather, people judge the frequency of events based on how easily the event comes to mind, or how available the construct is mentally. These departures from perfect rationality have forced economists to consider actors as “bounded” in their rationality; that is, limited to the information that they have (Kahneman, 2003a, 2003b). However, a shift to “bounded” rationality from perfect rationality does not address self-interest as a motive per se; people may still be aiming to maximize their material self-interest, but may be doing so with imperfect human decision-making processes.
By contrast, the second major way in which economists have accounted for behavior that appears to be imperfectly self-interested is to reconceptualize the term “utility” so that it also includes hedonistic desires for pleasure and satisfaction (see above, “Defining Self-Interest”). In these newer economic models of behavior, people are still maximizing their utility when they behave in a way that acts against their economic self-interest if their behavior enhances some other personal satisfaction. For example, Akerloff and Kranton (2000; see also Akerloff, 2002, 2007) argue that one's identity, one's personal sense of self, affects behavior via the adoption of norms for how one should behave based on group membership. Different groups of people have different norms and prescriptions for the appropriate way to behave, and the adoption of those norms and conformity to those behaviors serves to maximize a different kind of utility. Because although some behaviors appear to go against one's economic self-interest, if the behavior conforms to the person's identity, or understanding for how s/he should behave, then it improves his/her hedonistic, not materialistic, “payoff.” The classic example that Akerloff and Kranton (2000) offer is that of the suburban housewife after Betty Friedan released The Feminine Mystique. Some could argue that many housewives could have financially benefited from joining Freidan in challenging the stereotypes and roles of women in Cold War American society. However, the fact that Friedan was challenging their accepted identity as women, their understanding of who women fundamentally are and how women should behave, led some housewives to eschew her cause. Because their identity as women was being devalued by Betty Friedan, their utility, too, was losing value. Similarly, consider the situation in which a man works in a traditionally male-dominated industry, like business accounting or coal mining. If a woman were to enter his workplace in his same position, then Akerloff and Kranton (2000) argue that his identity as a man is threatened, possibly resulting in his harassing the woman co-worker or preventing her promotion, as has happened in innumerable cases involving sex discrimination and hostile work environments (e.g., Price Waterhouse v. Hopkins, 1989; Jenson v. Eveleth Taconite Company, 1997). From a purely rational choice model, the act of harassing a co-worker does not optimize one's economic self-interest. However, it does serve to bolster his identity as a male, increasing his utility via personal satisfaction. Thus, according to Akerloff and Kranton (2000) a model of economic behavior that includes actors' identities is more explanatory than models that do not.
It is worth noting in both of these examples that the actors may have been influenced by their traditional self-interest as well. In the case of the man working in the male-dominated field, he may simply view a female co-worker as more threatening competition in the workplace because she is distinct, and thus engage in harassing behaviors. The Cold War housewife may not have joined Betty Friedan's movement because it may have harmed her material self-interest; by refusing to engage in traditional, conventional housewife behaviors, she may have risked her livelihood and perhaps even her safety. Thus, as stated earlier, it is possible to interpret many behaviors as being self-interested depending on the particular definition of self-interest one is employing.
The final way in which some economists have explained imperfectly self-interested behavior is to simply question whether all behavior can be attributed to the one self-interest motive at all. In particular Sen (1977, 1990, 1994, 1995) has advocated for the idea that people belong to social units and have over time developed goals for human behavior that are not self-interested: commitment and sympathy. Both commitment and sympathy involve the recognition of another person's goals and the willingness to act in a way that promotes those goals. The major difference is that sympathy involves an emotional reaction on the part of the perceiver whereas commitment does not. For example, consider the case of a man who was denied housing in a desirable neighborhood because of his race. If one hears about this incident and feels indignant and angry on behalf of the man, and helps the man in pursuing justice, then one has experienced sympathy. But if another hears about the incident and acknowledges that it is wrong, and that society would be a better place if racial discrimination in housing did not exist, and helps the man pursue justice, then that person is committed to the cause. Sen (1977, 1990, 1994, 1995) argues that commitment and sympathy fundamentally differ from self-interest in that the behaviors do not work to maximize one's own personal self-interest, but for the benefit of others. However, one can make the case that acting in the name of commitment or sympathy still operates to affect one's hedonistic self-interest in that the behaviors still satisfy the person's desired end, to help others (Pettit, 2005) and to reduce one's negative emotions upon seeing someone in need of help (Dovidio & Penner, 2001; Loewenstein & Small, 2007; Piliavin & Charng, 1990).
Thus, the concept of a rational self-interested actor has been challenged and modified in economics. Some scholars have pointed to biases in human cognition that cause people to account for information differently than normative theories of decision-making do (e.g., Kahneman, 2003a. 2003b; Rabin, 1998; Rabin & Thaler, 2001; Thaler, 1991). And others have pointed out the benefits of expanding the concept of utility to include hedonistic desires of conforming to norms of group membership, holding an identity, and experiencing emotions in relation to others (Akerloff, 2002; 2007; Akerloff & Kranton, 2000; Sen, 1977; 1990; 1994; 1995; Pettit, 2005). Still, however, in the field of economics there have been calls for a broader understanding of how boundedly rational people make decisions while accounting for biases in human cognition and the need for heuristics, the incorporation of the preferences of others, the incorporation of norms that change with different settings, and how trust in others affects these behaviors (Ostrom, 2010; Thaler, 2000).
In political science, the effect of self-interest on voting behaviors has also been questioned and empirical research has yielded complex, nuanced results. Broadly, there is good reason to believe that the economy affects politicians' approval ratings and tenure, but the individual-level, psychological mechanism by which this occurs has been debated (Feldman, 1982; Fiorina, 1978; Kinder, 1981; Kinder & Kiewiet, 1979; 1981). Although the simplest explanation would be that when people suffer economically, they vote against the incumbent, the data do not reveal this straightforward pattern (Kinder, 1981). Rather, it appears that people vote against the incumbent because of perceptions of sociotropic economic concerns – collective economic grievances that society faces as whole (Kinder, 1981) and not because of personal economic concerns per se. These concerns are best categorized as hedonistic concerns, then, because of the satisfaction voters feel in these situations to help society as a whole. Thus, it may be that people perceive broader economic troubles and, whether they are personally suffering from the economy or not, vote against the incumbent.
Additionally, Feldman (1984) argues that in order for people to vote in line with their self-interested concerns, they must first attribute their economic woes to the government's economic policies, which not all constituents are inclined to do. If someone perceives that he lost his job because his boss was incompetent and not because of the lagging economy, then he is less likely to punish the incumbent politician for his job loss. Conversely, an individual could vote for raising taxes in order to benefit the broader economy, if s/he thought that improving the broader economy would improve his/her personal self-interest, i.e., saw a direction relationship between the government's policies, the broader economy, and his/her personal material circumstances.
Research in symbolic politics reveals a similar pattern wherein people must be able to connect how they are affected by policies in order to evaluate them in accord with their self-interest. According to work in this area, symbolic beliefs and symbolic racism play much stronger roles than self-interest in most policy evaluations because of people's inability to connect how they are personally affected by most policies. Symbolic beliefs refer to affect-based, internalized values about the way the world should be and the way that people should behave (Eagly & Chaiken, 1993; Kinder & Sears, 1981; Sears, Lau, Tyler, & Allen, 1979). Sears argues that these symbolic beliefs are learned via conditioning at a young age, and this learning does not include a calculation of the costs and benefits of holding these attitudes; they are stable predispositions that are independent of material self-interest. One prominent example of symbolic beliefs is symbolic racism, a blend of anti-Black affect and the belief that Black people violate the values of the Protestant Work Ethic (Sears & Henry, 2003).
Research in symbolic politics consistently demonstrates that self-interest has a weak effect on social policy evaluations in most domains, but that related symbolic beliefs (including symbolic racism) are significantly more predictive (see Lau & Heldman, 2009; Sears & Funk, 1990, 1991). For example, when investigating Whites' attitudes toward busing Black students into primarily White schools to achieve racial integration in public schools, Sears, Lau, Tyler and Allen (1979) found that respondents' symbolically racist beliefs were stronger predictors of attitudes towards busing than whether the respondent would be personally affected by the busing policy. Here, self-interest is operationalized as personal involvement (see above, “Defining Self-Interest”). In the racially charged 1972 Los Angeles mayoral election, Kinder and Sears (1981) found that personal concerns about racial threats to “the good life” had little effect on candidate preference: respondents' concerns about neighborhood and school desegregation, Black violence, and economic competition (more traditional interpretation of self-interest) were weakly related to candidate preference. Rather, White voters' symbolically racist beliefs about Black people were much stronger predictors of their candidate choice, regardless of whether they would be personally affected by the candidate's policies. Again, it is possible that White people may have felt that racial integration or having a Black mayor would help their long-term material self-interest, or simply have felt personally satisfied at the prospect of racial integration.
Findings like these provide the basis for symbolic politics research, which investigates whether attitudes toward social policies are better predicted by self-interest or symbolic beliefs. As Sears and Funk (1990, 1991) and Lau and Heldman (2009) review, further research has revealed symbolic beliefs to be a significant predictor of a variety of attitudes towards social policies, independent of the effects of material self-interest. For example, political ideology (liberalism vs. conservatism) is a stronger predictor of support for government-provided health insurance or privatized health care than self-interest; whether respondents did not have self-interest or were underinsured was not nearly as strong a predictor (Lau & Heldman, 2009; Sears & Funk, 1990). Political ideology is also a stronger predictor of policies that would guarantee jobs for everyone, even among those who are unemployed (Lau & Heldman, 2009; Sears & Funk, 1990).
Why isn't self-interest as predictive of policy attitudes as classical political theories suggest? Sears and Funk (1990, 1991) argue that self-interest does not play a role in most people's political behavior because politics are distal for most people; in order for people to vote in accordance with their self-interest, they argue, they must first be able to recognize their stakes in the particular issue and then vote in accordance with their self-interest. For most people, however, the broader political arena tends to obfuscate matters in such a way that policy outcomes are less clear and people are less able to discern how they are materially affected by various policies. In these cases, Sears and Funk (1990, 1991) suggest that people use their symbolic beliefs about related issues to guide their evaluations of relevant policies.
When and how self-interest matters
Although symbolic beliefs play a more influential role in policy attitudes when the stakes are unclear, research in symbolic politics reveals that self-interest has a powerful effect when the costs of the outcome are clear. Because people have the capacity to draw the link between their self-interest and the policy in these situations, their attitudes toward the policy are strongly affected by their self-interest. Other work has similarly demonstrated that people must be able to draw causal connections between how policies work and the strength/extremity of their attitudes (Fernbach, Rogers, Fox, & Sloman, 2013). In particular, policies that affect voters' pocketbooks have shown clear self-interest effects, presumably because voters can easily understand that they are financially affected by those policies and easily calculate a cost-benefit analysis of the passage of those policies.
For instance, Sears and Funk (1990) review previous work in which public employees were more strongly opposed to proposed tax cuts. The public employees likely understood that tax cuts could affect their salaries and whether they would have jobs, and these concerns likely drove their negative evaluation of the tax cut policies. Because the costs of the policy were clear and substantial to public employees, their traditional self-interest concerns played a much stronger role in their attitudes toward that policy.
Taxation policies also have obvious financial effects on voters' pocketbooks and thus yield significant, consistent, and strong self-interest effects. As Sears and Funk (1990, 1991) review, beneficiaries of proposed tax cuts are much more supportive of the proposed policies than those who would not benefit; conversely, proposals for tax increases are strongly opposed by those who would suffer the most as a result. Again, Sears and Funk (1990, 1991) argue that because the monetary consequences of a tax proposal are obvious to voters in these situations, their self-interest plays a larger role in their attitudes toward that policy than it would if the costs of the policy were not clear.
Other findings in social and political psychology support the idea that self-interest matters when voters can clearly connect how they are affected by a policy. Green and Gerken (1989) reported great differences among current smokers, former smokers, and nonsmokers in their support of regulatory policies proposing smoking restrictions. Employing a hedonistic approach to self-interest, Green and Gerken (1989) found that the more respondents smoked, the less they supported policies that proposed the banning of public smoking or its restriction to “special areas.” Conversely, nonsmokers were much more supportive of raising cigarette taxes, irrespective of the amount proposed, and smokers were less supportive, and were increasingly less supportive as the proposed tax amount was increased because of the obvious material effects on their pocketbooks. Previous poll results suggested no differences in demographics or political party affiliations between smokers and nonsmokers. Green and Gerken argued that the reason for the effect of self-interest (i.e., smoking status) in their poll was because the effects of the proposed policies were unambiguous to those who would be affected (i.e., smokers). Similarly, Kim, Stark, and Borgida (2011) found that smokers were significantly more opposed to taxing tobacco products, but symbolic beliefs about the government were more predictive of other regulatory policies. It is interesting to note that in these cases, smokers were not affected by their traditional self-interest in the sense that these policies are designed to minimize damage to their health in the form of making smoking more difficult.
Additional research has explored whether people's self-interest can be enhanced by making it more cognitively accessible (Chong, Citrin, & Conley, 2001; Sears & Lau, 1983; Young, Thomsen, Borgida, Sullivan, & Aldrich, 1991). In this line of research, researchers have primed participants to consider their self-interest before exposing them to a policy and having participants evaluate that policy. Priming is a technique in psychology in which a concept is made more cognitively accessible by briefly exposing people to the concept, whether people are aware of it or not (see Bargh & Chartrand, 1999). Young, Thompsen, Borgida, Sullivan, and Aldrich, (1991) primed participants by exposing them to a conversation between two confederates about taxation. Participants were waiting for the experimental session to begin, and overheard two research confederates speaking. In this conversation, one confederate opined that the best way to evaluate a government program is to consider how it affects your self-interest, whereas his conversation partner argued that there were other matters to consider. In an ostensibly unrelated experiment, participants then indicated their support for hypothetical legislative proposals. The researchers found that participants who had been exposed to the conversation were more strongly influenced by self-interest concerns than those who had not been exposed to that conversation, and concluded that construct accessibility of self-interest facilitated self-interest effects.
Similarly, Chong, Citrin, and Conley (2001) surveyed adults and asked them to consider three policies that differed in how easily participants could calculate their objective stakes in the issue, and the objective size of the beneficiary group (Social Security reform, home mortgage taxes, and healthcare for unmarried partners or employees). However, before asking participants to indicate their support for the measures, Chong et al. (2001) asked some participants how they and their families would be affected by the measure, and asked other participants how society as a whole would be affected by the measure. They found that when participants were primed to think of the policy in self-interested terms, they were more self-interested in their evaluations of the various policies. These findings mimic the arguments proposed by Sears and Lau (1983), who argued that the demonstrated effects of self-interest on policies and political candidates were likely inflated due to the “politicization” of self-interested thinking in those surveys. It is important to note, however, that all of these modes of priming are confounded with social effects, and should be interpreted as priming with caution, as discussed below.
Some of the disagreements regarding self-interest are due to measurement and operationalization issues as discussed (see above, “Defining Self-Interest”). Crano (1997a, 1997b) argued that the observed lack of relationship between self-interest and attitudes toward busing may be due to an inappropriate definition, operationalization, and measure of self-interest; Sears categorized all White parents who had children in an affected school district as self-interested, and all who did not have children in such a school district as not self-interested, employing the personal involvement approach to defining self-interest. He then entered this variable as a predictor alongside symbolic beliefs in a regression model, and found that self-interest did not emerge as a significant predictor. The problem with this operationalization and these analyses, Crano pointed out, was that all White parents who were personally involved in the policies might not feel uniformly affected by them. That is, some White parents who were personally affected by busing policies may have supported them for other self-interested reasons. Instead, Crano argued for a more specific operationalization of personal involvement (“vested interest,” Crano, 1997a, p. 485). Crano argued that when people are affected by a policy and perceive it as subjectively important to them, there is greater correspondence between pre-existing attitudes and support for busing policies. Thus, Crano also employed the personal involvement approach to self-interest, arguing that those who are personally affected by a policy and feel it is more important think about the policy more and exhibit more consistency in their attitudes.
When self-interest is operationalized in this way Crano and his colleagues argue that self-interest exerts significant effects on attitude strength and self-interested behavior. In a variety of studies, they present evidence showing that their operationalization of self-interest biases attitude strength and significantly moderates the relationship between attitudes and related behaviors. For example, Sivacek and Crano (1982) showed that young adults who would actually be affected by a change in the legal drinking age were much more likely to sign up for activism opposing the policy. Similarly, there was a stronger correlation between attitudes toward comprehensive senior examinations and willingness to engage in various activities expressing opposition to the examinations among college students who felt they would be affected by the proposed comprehensive exams. Moreover, using the same national dataset as Sears, Crano (1997a) showed that the relationship between attitudes toward busing and intention to vote for a politician with congruent busing attitudes was stronger among those who would be personally involved in the busing policy. In other words, when people were affected by a busing policy because their own children would be bussed or would attend a school to which others would be bused, they were more likely to vote for a candidate whose views matched their own attitudes toward busing.
Additionally, Darke and Chaiken (2005) have provided evidence that self-interest exerts a biasing effect on cognitions. In four separate experimental studies, Darke and Chaiken (2005) showed that self-interest affected the direction of policy attitudes such that those who would suffer from a policy were opposed to that policy; students who believed that their university would adopt a form of testing on which they performed poorly were more opposed to the policy. The researchers also showed that perceptions of personal costs mediated the relationship between self-interest and policy attitudes, and that self-interested individuals devoted more cognitive resources to the policies. These studies provide compelling evidence that self-interest plays a significant role in policy evaluations, affecting both the valence of policy attitudes and increasing systematic processing of policy messages.
In sum, research in social and political psychology reveals complex findings surrounding the power of self-interest. Although self-interest is widely accepted as a primary motivator of human behavior, it appears to exert very limited effects on political thinking and behavior and in very specific circumstances: when costs and consequences are clear and when self-interest is cognitively accessible. In these cases, research demonstrates that self-interest can bias thinking and inspire people to engage in political action. Research regarding social norms and self-interest also reveals yet another complicated picture regarding self-interest.
Social Norms Surrounding Self-Interest
Research on the “norm of self-interest” establishes that people chronically overestimate the power of self-interest. According to work in this area, academic theories regarding self-interest as the basis for human behavior have become so well-known that everyone assumes that everyone is affected by their self-interest (Miller, 1999; Miller & Ratner, 1998; Ratner & Miller, 2001). Moreover, Miller and his colleagues argue that this pervasive belief has resulted in a “self-fulfilling prophecy” (Miller, 1999, p. 1053) causing people to pursue self-interested goals when they otherwise wouldn't. In other words, self-interest in and of itself is no longer a main motivation, but because of the primacy that self-interest is accorded in our society (via all of its definitions), people believe that self-interest strongly affects others' behaviors, and this belief spurs them to behave in more self-interested ways.
In fact, Miller and his colleagues argue that self-interest has taken on normative properties, having both descriptive and prescriptive components (Cialdini, Reno, & Kallgren, 1990). Descriptive norms have the power to induce a behavior by providing information about what is effective. If someone believes that others are behaving in a certain way, s/he may conclude that it is the most effective behavior in the given situation and behave in the same way because s/he personally accepts the effectiveness of that behavior. Thus, a descriptive norm of self-interest operates via the belief that everyone is acting on behalf of their own self-interest, which further prompts perceivers to also behave in self-interested ways. By contrast, prescriptive norms prescribe what ought to be. Prescriptive norms indicate which behaviors would curry social favor and which would result in social penalties. Prescriptive norms induce behaviors by leading the actor to believe that s/he may be penalized for not conforming to the norm. Thus, in the case of a prescriptive norm of self-interest, actors should perceive social pressure to act in self-interested ways, and conform for fear of social rejection or suspicion. When actors conform due to a prescriptive norm of self-interest, the effectiveness of the behavior is not necessarily privately accepted (see Cialdini, 2000).
The evidence for both types of norms of self-interest and the related psychological processes is equivocal. As evidence of a descriptive norm of self-interest, Miller and Ratner (1998) demonstrate very clearly that people believe others are motivated by their self-interest; participants overestimate the power of incentives in inspiring behavior and of group membership in predicting attitudes toward policies affecting that group. Thus, people think others are strongly motivated by two forms of self-interest, as described above (see “Defining Self-Interest”); monetary incentives and tangible wealth, and personal stakes refer to personal involvement in a policy. The belief that others are all motivated by self-interest is a major component in a descriptive norm of self-interest.
It is not clear, however, whether self-interested thinking is ever privately accepted as effective and desirable. Everyday experience tells us that to call someone “selfish” is pejorative; and common wisdom aside, there is some empirical evidence that challenges the assertion that self-interest is privately valued. In research on naïve realism (Pronin, Lin, & Ross, 2002), participants consistently indicate that motivational biases, including self-interested thinking, infects the thinking of others but not their own. (In their study it was not clear what form of self-interest their respondents were considering.) This finding raises the question of how much self-interest is privately accepted by participants as an appropriate, desirable way of thinking, given their apparent reluctance to embrace self-interested thinking on a survey, and their desire to contrast their own thinking with that of others.
As evidence that beliefs surrounding the prevalence of self-interest actually engender self-interested behavior, Frank, Gilovich, and Regan (1993) showed that students majoring in economics behave in a more traditionally self-interested manner than non-economics majors, and that senior economics majors behave in a more self-interested manner than freshman economics majors. Although it is possible that people who are more self-interested are simply attracted to the field of economics, these findings suggest that repeated exposure to theories of economics (which hold human beings as inherently self-interested), causes people to become more self-interested over time, in accord with the tenets of a descriptive norm of self-interest.
Additionally, it is possible that the research on cognitive priming of self-interest actually supports Miller's (1999) assertion that a descriptive norm of self-interest engenders self-interested thinking. As described earlier, research in symbolic politics suggests that cognitively priming one's self-interest can cause one to become more self-interested (Chong, Citrin, & Conley, 2001; Sears & Lau, 1983; Young, Thomsen, Borgida, Sullivan, & Aldrich, 1991). However, as noted earlier, the sources of these priming were social in nature, raising the question of whether the self-interest construct was primed, or a norm was activated. In the case of Chong, Citrin, and Conley (2001), participants were ostensibly primed by being asked by the interviewer how much they would be personally affected by a proposal before being asked to evaluate the proposal. Similarly, the research reviewed by Sears and Lau (1983) also implies a normative, not cognitive, process in which participants are asked in surveys how their self-interest is affected before evaluating various policies. And, lastly, the study conducted by Young, Thomsen, Borgida, Sullivan and Aldrich (1991) involved priming exposing participants to a confederate conversation. In doing so, the researchers may have established a norm of self-interest for policy evaluations; they may have cued participants to think that self-interested thinking was sanctioned. Researchers have traditionally interpreted these findings as evidence that the self-interest construct simply needs to be cognitively activated in order to have an effect. However, the fact that the researchers employed social, rather than purely cognitive means (e.g., subliminal priming), legitimately raises the possibility that these effects are due to activating a norm of self-interest among participants. In other words, the research demonstrating the effect of cognitively priming self-interest may actually demonstrate the effect of activating a descriptive norm of self-interest on participants' subsequent evaluations.
Thus, the evidence in support of the psychological processes behind a descriptive norm of self-interest is mixed. There is good reason to believe that people seem to think that others are strongly motivated by selfish concerns (Miller & Ratner, 1998; Miller, 1999; Pronin, Lin, & Ross (2002), which is a major component in a descriptive norm of self-interest. Moreover, research from the extant literature also supports the idea that at the very least, a descriptive norm of self-interest can be cued, causing actors to behave in more self-interested ways than they otherwise would. However, based on results from research on naïve realism, it is unclear whether self-interest is privately accepted as an appropriate motivation, given the finding that people report being less self-interested than others.
Evidence in support of a prescriptive norm of self-interest is even less clear. A prescriptive norm of self-interest would ostensibly operate by creating a belief that self-interested thinking is a socially desirable way of behaving, and that failure to behave selfishly would result in social penalties. Although Miller and colleagues demonstrate that people certainly have normative beliefs surrounding self-interested behavior, whether there is a prescriptive norm of self-interest that encourages self-interested behavior remains to be seen. Ratner and Miller (2001) show that men report feeling less comfortable acting on behalf of policies surrounding the availability of abortion because of a concern that their behaviors would elicit confusion in others because they are not personally affected by abortion policies). Moreover, they show that this perception is warranted; other participants indicated they would be more surprised and angry at the presence of a man at a meeting for the initiative than they would by a woman, ostensibly because of they are not personally affected by, or personally involved with, abortion policies.
Thus, Miller and his colleagues demonstrate that people are less likely to act on behalf of an issue because they perceive social consequences for pursuing goals that do not benefit groups to which they belong, that people report they would be more suspicious and angry at someone who is engaging in social action that does not benefit his/her group, and that these social penalties prevent people from engaging in social action that does not obviously benefit their group interests. (It is worth mentioning that this perspective provides a social-normative explanation for the findings reported by Crano and his colleagues (Crano, 1997a; Lehman & Crano, 2002; Sivacek & Crano, 1982), who found that self-interest significantly moderates the relationship between attitudes and behavior such that those who are self-interested in the outcome of a policy demonstrate stronger relationships between their policy attitudes and behaviors directed toward that policy.) However, the extent to which a prescriptive norm of self-interest operates to encourage self-interested behaviors is not at all clear. The evidence provided by Miller and colleagues supports the idea that people don't act for or against a policy that does not personally affect them. In other words, it is socially acceptable for people who are tangibly affected by a policy to engage in social action related to that policy, but it is not socially acceptable for someone who is not tangibly affected to behave in the same way, and this norm has an inhibitory effect, preventing unaffected people from engaging in social action.