Social Beings Ch. 9

SocialBeingsCh.9READING.docx

Home >Psychology homework help >Social Beings Ch. 9

CHAPTER 9 HELPING: PROSOCIAL BEHAVIOR

In Hattiesburg, Mississippi, an elderly washerwoman donated her lifelong savings of $150,000 for scholarships at the local college. Oseola McCarty had lived alone for decades. Who else but strangers would get her gift? (http://en.wikipedia.org/wiki/Oseola_McCarty)

Why did Ms. McCarty give her unselfish gift? Social psychologists study such prosocial behavior, examining its underlying causes—both dispositional and situational. Social psychologists also debate its core motives—egoistic, altruistic, collectivist, or principled. This chapter argues that these motives fit combinations of our core social motives. Debates have raged over the evolutionary significance of self-sacrificial behavior, as we will see, and the core motives shed some light on this contentious issue. We will examine how prosocial behavior can result from specific factors as varied as social learning, mood, attributed responsibility, empathy, identity, norms, and moral reasoning. But first, let's define our terms and organize these topics according to clusters of motivations.

WHAT IS PROSOCIAL BEHAVIOR?

No one would disagree that Oseola McCarty's gift represents prosocial behavior. But what about a fictional family whose son dies, and they take in his young widow? Blamed for his death and hideously disfigured by a botched suicide attempt, she is barely tolerated, but she is called Precious Auntie and lives with the family nonetheless. In keeping with ancient beliefs in traditional China, the son's ghost came to the grandmother

in a dream and warned that if Precious Auntie died, he and his ghost bride would roam the house and seek revenge on those who had not pitied her. Everyone knew there was nothing worse than a vengeful ghost. They caused rooms to stink like corpses. They turned bean curd rancid in a moment's breath. They let wild creatures climb over walls and gates. With a ghost in the house, you could never get a good night's sleep. (Tan, 2001, p. 175) 1

Given their beliefs, was caring for her a prosocial act, and if so, was it egoistic self-interest, empathic altruism, collective family loyalty, or principled morality? As we will see when we define these terms and the research they each encompass, cultures differ in their interpretations of responsibility, with some Eastern ones emphasizing responsibility (for certain ingroup members) and some Western ones emphasizing more personal factors such as liking and direct self-interest.

Conceptual Definitions

Prosocial behavior includes behavior intended to benefit others—behaviors such as helping, comforting, sharing, cooperating, reassuring, defending, and showing concern (Schroeder, Penner, Dovidio, & Piliavin, 1995, p. 15). Note several features of this definition: First, it includes the intent to help others, so acts that unintentionally help others do not count (for example, taking a new job may benefit another person who fills the old one, but that was probably not the intended consequence). However, acts intended to help, which may actually fail to help, would be included (for example, it would be useless, but still prosocial, to shovel snow off the front walk of someone who uses only the back door). Second, what actually benefits another person is socially defined, changing with time and place. Depending on the context, prosocial acts might include circumcision, foot-binding, piercing, scarring, tooth-pulling, or ruthless criticism. Finally, note that benefit to one or more others, including society, but not benefit to self, is key. The positive behavior is social and interpersonal, not self-directed in its intent.

Social psychologists often distinguish a subset of prosocial behaviors according to their motivations. Although prosocial behaviors are intended to benefit others, the underlying motivation might or might not be other-oriented. Altruism conceptually underlies the subset of prosocial behavior that is “motivated mainly out of a consideration of another's needs rather than one's own” (Piliavin & Charng, 1990, p. 30). As a motive, altruism involves self-sacrificial costs, absent “obvious, external rewards” (Batson, 1998, p. 282). Altruism thus involves concern for others' needs, independent of hoped reward or feared punishment outside the self (Grusec, 1991). By these definitions, then, the traditional family that took in their widowed daughter-in-law, for fear of her ghost, acted prosocially but not altruistically.

Operational Definitions

Prosocial behavior is good. Social psychologists, like everyone else, want to promote it, but to do so, they must first explain it. They have pursued two approaches, dispositional and situational. One possible explanation is that people behave prosocially because they have prosocial personalities. In this view, the Oseola McCarties of the world differ from the rest of us, and the trick is to socialize children to become more prosocial. Researchers pursuing this track use personality questionnaires, combining variables to predict helping, for example, a fellow student with severe stomach cramps (Gerbasi & Prentice, 2013; Staub, 1974; see Table 9.1 ).

TABLE 9.1 Personality Predictors of Helping Another Student

Positive predictors

Valuing helpfulness

Valuing equality

Taking responsibility for others, e.g.:

If a good friend of mine wanted to injure an enemy of his or hers, it would be my duty

to try to stop the friend.

Professional obligations can never justify neglecting the welfare of others.

I would be obligated to do a favor for a person who needed it, even though the friend

had not shown gratitude for past favors.

Endorsing social responsibility, e.g.:

I am the kind of person that people can count on.

I usually volunteer for special projects at school.

Cheating on examinations is not so bad as long as nobody ever knows. *

Moral reasoning (see Figure 9.5)

Negative predictors

Valuing a comfortable life

Valuing ambition

Valuing cleanliness

Machiavelli scale

The best way to handle people is to tell them what they want to hear.

It is hard to get ahead without cutting corners here and there.

Most people are basically good and kind. *

* Indicates a negatively worded item that researchers reverse-coded from others in the same set. Each questionnaire included other negatively worded items, to control for acquiescence response bias, but they are omitted here for clarity. Items listed are examples, not entire scale.

Another personality method, besides questionnaires, uses people's initial reactions to choosing cooperative or competitive options in experimental games. You might try this method before reading further ( Table 9.2 ). This test allows researchers to see what kind of social value orientation people spontaneously choose (Messick & McClintock, 1968), that is, what degree of rewards they generally prefer for themselves relative to a generalized other person. Three kinds emerge: individualistic (simple self-interest, also called max own), cooperative (helping self and the other simultaneously, called max both), or competitive(creating the biggest discrepancy of self over other, called max diff). When people choose consistently, they are classified as having one of these three orientations (Van Lange, De Bruin, Otten, & Joireman, 1997). Other individual differences involve empathy and perspective-taking (Eisenberg, 1991), described later.

TABLE 9.2 An Instrument to Measure Social Value Orientation

In this task we ask you to imagine that you have been randomly paired with another person, whom we will refer to simply as the “Other.” This other person is someone you do not know and you will not knowingly meet in the future. Both you and the “Other” person will be making choices by circling the letter A, B, or C. Your own choices will produce points for both yourself and the “Other” person. Likewise, the Other's choice will produce points for him/her and for you. Every point has value: the more points you receive, the better for you, and the more points the “Other” receives, the better for him/her.
For each of the choice situations, circle A, B, or C, depending on which column you prefer most.
		A	B	C			A	B	C
(1)	You get	480	540	480	(5)	You get	560	500	490
	Other gets	80	280	480		Other gets	300	500	90
(2)	You get	560	500	500	(6)	You get	500	500	570
	Other gets	300	500	100		Other gets	500	100	300
(3)	You get	520	520	580	(7)	You get	510	560	510
	Other gets	520	120	320		Other gets	510	300	110
(4)	You get	500	560	490	(8)	You get	550	500	500
	Other gets	100	300	490		Other gets	300	100	500

Participants are classified when they make six or more consistent choices. Prosocial choices are 1c, 2b, 3a, 4c, 5b, 6a, 7a, 8c; individualistic choices are 1b, 2a, 3c, 4b, 5a, 6c, 7b, 8a; and competitive choices are 1a, 2c, 3b, 4a, 5c, 6b, 7c, 8b.

In measuring the ways that personality predicts helping, researchers must consider measuring both personality and situation because they interact. In other words, different personalities are suited to different kinds of helping situations (Dovidio, Piliavin, Gaertner, Schroeder, & Clark, 1991). For example, in emergencies, the likely helpers are more impulsive, emotional, socially responsible, esteem oriented, and less oriented to safety and reality. These people and situations implicate helping based on sheer arousal. In long-term helping, the likely helpers are thoughtful about pros and cons, competent, confident, and responsible. These people and situations implicate helping based on a cost-benefit analysis. This distinction between people and situations suggests different research strategies about what variables to measure and manipulate, depending on the focus (e.g., emergency versus long term). About a third of Americans report long-term volunteering, and about a fifth report heroic acts (Zimbardo, Breckenridge, & Moghaddam, in press).

Most often, rather than measuring personalities or personalities in specific situations, social psychologists focus on situational causes. Experimental games as one operationalization of prosocial behavior can assess situational variables (as well as personality differences in social value orientation). Instead of merely choosing preferred outcomes, as in the social value orientation test of personality, participants actually play experimental games with strangers. This method developed as part of game theory (Colman, 1982; Luce & Raiffa, 1957). Normally the researcher sets up cooperative and competitive choices, as displayed in Figure 9.1 . (Notice that the payoff matrix in Chapter 8 's discussion of interdependence theory derives from this kind of work.) The cooperative choice helps the partner, and the competitive one hurts the partner. Thus, a cooperative choice benefits both parties only if both parties cooperate (both people gain 2). But if one party cooperates while the other competes, then the competitor gains disproportionately (3) to the partner (0). If both compete, both lose, relative to most other outcomes (both get 1). Researchers manipulate situational variables, such as degree of communication allowed or patterns of partner choices, to track the frequency of cooperative (prosocial) and competitive choices.

Figure 9.1 Payoff Matrix for an Experimental Game of Cooperation (Prosocial) and Competition (Self-interest) Choices

In examining situational theories of prosocial behavior, social psychologists have elaborately staged some of our field's most inventive operational definitions of helping. Many studies use emergency situations, based on the original classic, Darley and Latané's 1968 study of bystander intervention in emergencies. That study, modeled after a gruesome and very public murder on the streets of New York, examined why bystanders might fail to intervene—precisely because the presence of other bystanders diffuses each individual's sense of personal responsibility. (For present purposes, we describe just the method; we examine the theory later.) In the classic study, participants listened through earphones as a series of alleged other participants took turns discussing personal problems associated with college life. On his first speaking turn, one person mentioned hesitantly that he was prone to seizures. On his second turn, the victim began to have an audible epileptic seizure, over three full minutes. Experimenters timed how long it took participants to leave their experimental cubicle to help, if they helped at all.

As Table 9.3 indicates, participants' helping was a direct function of the number of other people available to help (more bystanders, less helping). We will return to this study, but the relevant point for now is that this study introduced the now-standard emergency-helping paradigm. Since that study, participants have overheard explosions and toppling bookcases, and they have encountered people needing a ride to the pharmacy, a call for the towing service, an ambulance, or help picking up spilled groceries. Some of the staged emergencies represent social psychology at its most theatrical. But the impact of the operationalized emergencies is important, for researchers need to know how people respond when they are startled, aroused, and genuinely worried.

TABLE 9.3 Effects of Group Size on Helping

Group Size	% Responding by End of Seizure	Time (sec.)
2 (Participant and victim)	85	52
3 (Participant, victim, and 1 other)	62	93
6 (Participant, victim, and 4 others)	31	166

Not all helping occurs under emergency, in the real world or in the laboratory. Some prosocial behavior is dogged and reliable over time. For example, research participants may respond to videotaped or written scenarios depicting victims of ongoing distress (e.g., family, relationship, or school troubles). Participants may report their willingness to donate hours, blood, or money for charity. In field studies, participants may report their actual time spent volunteering with AIDS patients or other community efforts.

Core Social Motives

Why do people help in emergencies or donate resources over time? Prosocial behavior explicitly reflects four types of social motivation (Batson, 1994), which arguably reflect our core social motives. (1) Egoism (self-enhancement and controlling outcomes) contrasts with (2) altruism (which both indicates and maintains trust in the social world). (3) Collectivism represents group belonging, and (4) principlism reflects socially shared understanding of the moral world by those belonging to particular groups, as well as a sense of control over how people should behave. A partly overlapping set of motives surfaces in the self-reports of people who volunteer to work with people with AIDS (Snyder & Omoto, 1992). Social psychologists debate all these motives as if they were competing, even mutually exclusive explanations for prosocial behavior, but arguably each sometimes contributes to actual prosocial behavior—often in combination—so it is helpful to unpack them. Self-interest and other-interest are independent orientations (Gerbasi & Prentice, 2013).

SELF-ENHANCING AND CONTROLLING

Egoism focuses on self-interest as a motive for prosocial behavior. Self-benefit is the ultimate goal. For social psychologists who endorse this view, people help only because it fosters their own survival or that of their genetic kin or is an accidental spillover from mechanisms evolved to help genetic kin. Similarly, people might help only in order to obtain personal rewards or avoid personal punishments, including self or public esteem. Or people might help merely in order to alleviate their own distress at watching someone else suffer. Or people might help because they mistake their self-interest (Burton-Chellew & West, 2013; but see Camerer, 2013). In any case, egoism combines motives for controlling outcomes (contingency between what you do and what you get) and self-enhancing (protecting self-esteem). The controlling motive reflects the self-interest benefits just mentioned. As an example of self-enhancement, people who volunteer to help people with AIDS sometimes cite what the researchers label esteem enhancement (“to feel better about myself”) as a motivation (Snyder & Omoto, 1992).

TRUSTING

In direct contrast to egoism, altruism describes a motive that makes people help because of genuine concern for others, to increase the welfare of others. They expect people to be responsible for each other. Recall that we defined the trusting motive to reflect a need to believe that the social world is a benevolent place. With altruism, people interpret the situation to be one that deserves help, and they experience empathy for the victim. Feeling the other person's pain motivates prosocial acts. The experience of individual sympathy and empathy distinguishes this motive from all the others. Apparently none of the AIDS volunteers explicitly cited simple altruism as their self-reported motive, perhaps because it would sound immodest. Regardless, altruism reflects the core social motive to trust, because it reflects the need to see people as fundamentally benevolent.

BELONGING

Collectivism identifies the group's welfare as a central motive. Here, maintaining the group and one's belonging to it motivates prosocial behavior. Helping ingroup others, following the group's rules about reciprocity, playing one's role, and adhering to norms all facilitate collective life. One helps one's ingroup in order to maintain one's ingroup identity. AIDS volunteers who cite community concern (“because of my concern and worry about the gay community”) illustrate collectivism. Note that the concern is for the group, not a specific individual, so this motive relates to the need to maintain belonging to a group. One's own inconvenience and comfort are sacrificed for a group one cherishes.

UNDERSTANDING (AND CONTROLLING)

Principlism is motivated to uphold moral standards. Impartial and universal, moral standards result from people's understanding of what is right and what is wrong. Principlism also relates to controlling outcomes, what people do and what they (deserve to) get. Thus, it implicates both understanding and controlling motives. People can reason at different moral levels, but principles or abstract standards guide prosocial behavior beyond self and group interest. Debatably, principlism reflects the core social motive of maintaining a coherent understanding of how the world ought to operate, perhaps mixed with a need for control, believing that certain ideas should guide one's own and others' behavior. As one example, AIDS volunteers who cite values (“because of my humanitarian obligation to help others”) illustrate principlism.

Not all the motives listed by the AIDS volunteers are covered here, because they do not typically appear in other research on prosocial behavior. For example, a direct application of our understanding motive might seem to be illustrated by AIDS volunteers who cite what the researchers label understanding (“to learn about how people cope with AIDS”), but a pure curiosity motive, which is not here elaborated to indicate altruism (e.g., if they had added, “so I can relieve the suffering of others”), seems simply self-serving. In any case, this curiosity motive has not figured prominently in research on motives for helping. AIDS volunteers who cite personal development (“to challenge myself and test my skills”) might be self-enhancing or might illustrate a possible role for the core social motive of controlling—efficacy and competence. The relevance of control and efficacy fits with dispositional findings that helpers often are high on general or task-specific competence, but the prosocial research has not explored this motive specifically, so we will not pursue it here. The four main motives in research on prosocial behavior thus are egotism, altruism, collectivism, and principlism, variously reflecting our core social motives.

Summary of Definitions and Motives

Prosocial behavior intends to benefit others. Social and personality psychologists have operationalized individual differences in prosocial behavior via questionnaire studies, finding that emergency helping relates to dispositional impulsivity, whereas long-term helping relates to dispositional social responsibility. Situational factors related to prosocial behavior have prompted compelling and creative experimental methods to simulate impactful situations comparable to real cases of need. Dependent measures include participants' rate and speed of helping. Other methods include laboratory scenario studies and real-world volunteering. When people act prosocially, they do so for a variety of motivational reasons (egotism, altruism, collectivism, and principlism, at least), which variously address our core social motives, egoism reflecting both controlling and self-enhancing, altruism creating and maintaining the motive to trust the social world, collectivism helping to maintain group belonging, and principlism reflecting moral understanding and controlling.

EGOISM HYPOTHESES: PURELY SELF-ENHANCING AND CONTROLLING

In the novel by Amy Tan, an American couple—who have been living together for years but temporarily separate when the woman goes to care for her ill elderly mother—discuss putting her into a nursing home:

She was both shocked at the expense and amazed that Art would be willing to pay that for three months, nearly twelve thousand dollars. She stared at him, openmouthed.

“It's worth it,” he whispered ….

Ruth exhaled heavily. “Listen, I'll pay half, and if it works out, I'll pay you back.”

“We already went through this. No halves and there is nothing to pay back. I have some money saved, and I want to do this. And I don't mean it as a condition for us getting back together or getting rid of your mother or any of that. It's not a condition for anything. It's not pressure for you to make a decision one way or the other. There are no expectations, no strings attached.” (Tan, 2001, p. 322) 2

In this instance, it would be easy to interpret Art's apparent gift as a manipulative, self-serving ploy, no matter what he says. Although subsequent events suggest this to be a truly selfless act, it conceivably could also be motivated by selfish ends. And even more so, the traditional, rural Chinese family who takes in their daughter-in-law, ostensibly to avoid the ghosts' revenge, would be most obviously interpreted as self-serving. But go back to Oseola McCarty. How can we possibly view her gift as self-interest? Maybe, being childless, she did it because the local branch of the state university held the closest thing to genetic relatives? Maybe she did it to get social rewards in a lonely life? Maybe she did it to feel good despite her arthritic pain?

Social psychologists hotly debate whether prosocial behavior is ever truly altruistic or all ultimately reduces to egoism (e.g., Batson, 1998; Hoffman, 1981; Piliavin & Charng, 1990; Schroeder, Penner, Dovidio, & Piliavin, 1995; Zaki & Mitchell, in press). Egoism seeks the ultimate goal of increasing one's own welfare. This of course fits a simple interpretation of one evolutionary perspective. According to this approach, purely altruistic self-sacrifice would pose a nonadaptive puzzle. Why should an organism diminish or eliminate its chances of passing on its genes by sacrificing itself for another organism? Surely, this argument runs, such genetic self-sacrifice would not be selected over time, because the altruistic genes would never be passed along. Some even argue that prosocial behavior is a mistake, correctable by learning one's accurately calculated self-interest (Burton-Chellew & West, 2013); others disagree (Camerer, 2013; Zaki & Mitchell, in press).

The simple analysis, it turns out, is deceptive, and the debate is more genuine, complex, and interesting. The proposed forms of egoistic motives include (1) kin selection (stemming from inclusive fitness), (2) social learning (including reciprocity and esteem seeking), and (3) mood protection, all forms of self-enhancement. And each could explain taking in your son's widow, subsidizing your partner's mother, and maybe even donating scholarships to strangers. Egoism is one of many convincing motives for prosocial behavior, but later we will see that it is not the only one.

Kin Selection

As noted in Chapter 1 , the unit of natural selection, evolutionary scientists point out, is the gene, not the individual, and this basic principle is called inclusive fitness (Hamilton, 1964; Wilson & Sober, 1994). Thus, kin selection might involve adaptive pressures promoting the survival of one's genetically related relatives. This suggests one plausible account for the evolution of prosocial behavior. If people predisposed to help their genetic relatives encourage the reproduction of such genes better than those who do not, people might be genetically predisposed to prosocial behavior. This does not mean that there is an “altruism gene,” merely that people (and other social animals) might have developed a propensity to respond to the needs of genetically related others, particularly those with reproductive potential. Kin selection could operate through a variety of motives, including pure egoism or even altruism. Evolutionary psychologists talk about people acting as if they are altruistically motivated, but with the underlying reality being genetic reproduction, in their view.

Evidence favoring the kin selection idea does suggest that people are more motivated to help genetically related others (Burnstein, Crandall, & Kitayama, 1994). As Figure 9.2 indicates, people report that they would particularly help kin, the more genetically related they are, especially under life-and-death circumstances. Women report receiving help from closer kin more than from distant kin (Essock-Vitale & McGuire, 1985). Genetic kin helping occurs for both Hong Kong Chinese and English participants (Ma, 1985).

Figure 9.2 Tendency to Help Kin, as a Function of Shared Genes

People report being more likely to try to save another person than an animal and close relatives over strangers, and this focus on people and on relatives is consistent with the genetic relatedness hypothesis. But their propensity to help an innocent person over a uniformed Nazi suggests that similarity or empathy is important as well (Petrinovich, O'Neill, & Jorgensen, 1993). The dimensions shown in the United States are important in Taiwan also (O'Neill & Petrinovich, 1998), again suggesting cross-cultural generality. Although kinship matters and, by extension, egoism (self-enhancing and controlling), other motives (similarity and empathy) evidently matter as well. For example, meta-analysis indicates that dependency is a cue for helping (Bornstein, 1994). Many social factors are confounded with genetic relatedness and could equally account for the results.

If the evolution of pro-kin behavior makes sense, then some social-psychological mechanisms would have to include recognizing kin and detecting their distress (Hoffman, 1981; Piliavin & Charng, 1990), as well as being motivated to act. With regard to recognizing kin, mothers recognize even their newborn infants by sight and odor (Porter, 1987). Adults report feeling closer to kin the more biologically related they are (Burnstein, Crandall, & Kitayama, 1994). People help people who are similar (Schroeder, Penner, Dovidio, & Piliavin, 1995), which may be a cue for recognizing genetic relatedness.

As for detecting distress, even newborns cry when they hear other infants cry. And very young children are distressed when another is distressed. Spontaneous communication of emotions (e.g., facial expressions) could mediate helping (Buck & Ginsburg, 1991). But the underlying motives for responding to another's distress could still include altruistic empathy, collectivism, or principlism, as well as egoism.

Even if people have evolved to behave more prosocially toward kin, they would have to be motivated to act. The immediate psychological mediators for behavior remain unclear. For example, people could help those to whom they feel emotionally close. This could represent either altruistic empathy or collectivism or an evolutionary spillover from a tendency to help genetic relatives. Self-reported probable helping in a series of hypothetical life-or-death dilemmas showed that emotional closeness partially accounted for the effects of genetic relatedness (Korchmaros & Kenny, 2001). That is, emotional closeness, not merely genetic relatedness, explains the kin-helping phenomenon. Besides emotional closeness, other possible mechanisms for kin favoritism include mere exposure (introduced in Chapters 6 and 7 ) and other mechanisms discussed in the rest of this chapter: positive mood, guilt, empathic distress, and attributions of responsibility (Cunningham, 1985–1986). Helping genetic relatives fits the kin selection idea, but it is silent on the psychological processes by which it would occur. And the data also fit other possible motives, for example, the ingroup we-ness of collectivism (Dovidio et al., 1991), discussed later in this chapter.

Social Learning

People might help other people for egoistic reasons other than genetic survival. For example, people might learn that prosocial behavior ultimately pays off for the self. Social learning (introduced in Chapter 6 with regard to attitudes) highlights the uniquely interpersonal processes in conditioning such social phenomena as attitudes, helping, and aggression. Processes that illustrate the prosocial kind of social learning primarily include reciprocity and social rewards. Both support egoism (self-enhancing and controlling) because what's-in-it-for-me constitutes the primary motive. People clearly do learn cooperation from reward and punishment, according to meta-analysis (Balliet & Van Lange, 2011).

RECIPROCITY

People might learn that if they supply help, they will receive help in return, the principle of reciprocity (introduced in Chapter 7 ). Reciprocity could benefit genetic survival, as a form of symbiosis (literally, living together). In so-called reciprocal altruism (Trivers, 1971), if people help others who help them, then the survival of both is more likely. Regardless of an evolutionary explanation, reciprocity could simply result from short-term social learning: Helping is rewarded by the reciprocal return of the favor in some form. Recall from Chapter 7 that people like others who like them. Similarly, people help others who help them.

In relatively impersonal relationships, equity (introduced in Chapter 8 ) prevails, whereby people hold equivalent their respective outcomes relative to inputs. Thus, for example, if one person is overhelped, that person may be able to reciprocate other benefits, such as deference and esteem instead of reciprocal helping. In exchange, people reciprocate benefit for benefit, creating an equivalence of outcomes. If you scan articles for the seminar one week, other students should scan them other weeks. Norms favoring pure exchange-based reciprocity are strong (Gouldner, 1960). Business school alumni, for instance, cite reciprocity as one of the main reasons for donating to their alma mater, even years later (Diamond & Kashyap, 1997). The community-oriented notion of “giving-back” as an adult for what you received growing up also fits the idea of social exchange and reciprocity.

Reciprocity clearly occurs in short-term interactions. When people play experimental games with strangers, normally the researcher sets up cooperative and competitive choices, introduced earlier ( Figure 9.1 ). Recall that the cooperative (prosocial) choice helps the partner and the competitive one hurts the partner. When people make their choices simultaneously on each trial, they eventually figure out that they both are better off if they both cooperate (Kelley, Thibaut, Radloff, & Mundy, 1962). Most relevant to considerations of reciprocal helping, research participants typically discover the most effective strategy, namely tit-for-tat (Pruitt, 1998). A tit-for-tat player cooperates only so long as the partner cooperates, but as soon as the partner defects to compete, the player also competes, until the partner cooperates again (Axelrod & Hamilton, 1981). This suggests reciprocity of both helping and hindering, a form of self-interested prosocial behavior.

Tit-for-tat is effective partly because it punishes cheaters. Cheater detection (introduced in Chapter 4 ; the idea that we have evolved sensitivities to notice the violation of social contracts) is important to evolutionary accounts of self-interested helping behavior. If the model holds true, the basic evolutionary principles are that (1) needs arise, (2) nearby others can help, and (3) cheating does not pay (Trivers, 1971). If cheating indeed does not pay, then evolutionary psychologists need to look for a detection and enforcement mechanism. This analysis led to the idea that people might be adapted to detect people who cheat on a social contract (Cosmides, 1989; Cosmides & Tooby, 1996; see Chapter 4 ). People who are chronically free-riders (letting others pay the cost of public goods, i.e., shared benefits, such as public television) are less helpful in general (Piliavin & Charng, 1990). If individuals detect cheaters and free-riders, they can minimize personal costs.

Yet another form of reciprocity is indirect reciprocity, whereby people help others mainly to develop an altruistic reputation, which then leads to third parties benefiting them in the future (Nowak, 2006). Indirect reciprocity is still egoistic, being incentive-driven. Individual differences in egoism-altruism drive people's motivation toward indirect reciprocity: Egoists help when future rewards from others are possible, while altruists are less affected by such incentives (Simpson & Willer, 2008). Thus, some people may help strategically, when they know their reputation will profit them in future encounters (Semmann, Krambeck, & Milinski, 2004). Indirect reciprocity is cognitively demanding because people need to track their own interactions but also the whole network of interactions, suggesting the evolution of norms in groups and abstract principles of morality (see those later sections). Indirect reciprocity depends on knowing someone else's reputation before helping (Nowak, 2006). Nevertheless, even five-year-olds understand that their reputation matters (Engelmann, Herrmann, & Tomasello, 2012).

A more advanced form of indirect reciprocity, network reciprocity, also can encourage cooperation among a limited number of immediate neighbors. These ideas also suggest a revived form of group selection, whereby a group of cooperators (mutually helpful people) survives better than a group of entirely selfish, so-called defectors. All these forms of social learning related to reciprocity and exchange fit the idea that people acquire enlightened self-interest, learning that selfish behavior pays off in the short term but not the long term. People believe this even in large-scale social dilemmas, such as pollution, overfishing, and other ecological issues (Baron, 1997). One could argue that these kinds of social learning fit ultimately egoistic motives, but one could perfectly well also argue that social norms establish the collective rules and that belonging instead motivates reciprocity.

SOCIAL REWARDS AND COSTS

People often have too narrow a view of rewards and costs. Helping can entail a range of costs, such as time, money, effort, pain, danger, embarrassment, disgust, disapproval, and uncertainty (Dovidio et al., 1991). But it can also entail a range of rewards, such as money, praise, recognition, gratitude, and relief. Tangible rewards clearly would undermine labeling a particular behavior as motivated by altruism, according to our earlier definition, whereas intangible rewards are more ambiguous. For example, cooperative behavior appears to be intrinsically rewarding; the same areas of the brain activate under cooperation as under rewards (Fehr & Camerer, 2007; Rilling et al., 2002). People benefit even more from giving social support than they do from receiving it (Brown, Nesse, Vinokur, & Smith, 2003). Even without self-reward, one might receive praise, approval, or social connection for helping; might that be an egoistic motive for otherwise apparently altruistic prosocial behavior?

People do weigh costs against rewards in deciding whether to help, except under the most impulsive helping. For example, in a review of 12 experiments, Dovidio (1984) found that helping decreased, as the cost increased. Furthermore, in a meta-analysis of 13 studies, helping increased, as the personal costs of not helping increased (Dovidio et al., 1991). If a cost-benefit analysis operates, perhaps helping is fundamentally self-interested because people do not help unless they are on balance rewarded.

People might learn that if they help, they will receive social rewards, such as public esteem. This constitutes a form of social learning because the rewards come from other people, mandated by their shared social groups. In one such egoistic view of social learning (Cialdini, Baumann, & Kenrick, 1981), children only rarely help, in part because it constitutes a material cost, but later they help more when they learn that social rewards accompany prosocial acts. Adults' helping results from internalized self-reward, in this model. And indeed, both adults and children do help more after prior helping has just been rewarded, especially if the rewards are social and not material (Grusec, 1991). Any reward seen as a bribe undermines intrinsic motivation to help (Fabes, Fultz, Eisenberg, May-Plumlee, & Christopher, 1989), as self-perception theory, noted in Chapter 3 , predicts. Thus, to increase intrinsic motivation to help, social rewards are likely to be more effective than material rewards. In addition to their own reward experiences, people also probably help more when they see someone else rewarded for helping, a process called vicarious conditioning or observational learning (introduced in Chapter 6 ), although not a lot of research has addressed this idea.

In another form of social learning, modeling, people imitate someone else who helps. In one classic study (Bryan & Test, 1967), drivers saw a female undergraduate standing near a Ford Mustang with a flat tire, along the side of the road. A spare tire leaned up against the car. Half the time, her car was preceded, one-quarter mile away, by an Oldsmobile with a flat tire, where another young woman watched a man changing the tire; half the time no model was present. The conditions with a model present or absent were compared for 2,000 passing motorists apiece, to see how many would stop to help. As the first line of Table 9.4 shows, many more people stopped to help when they had just seen someone else helping. Two other studies varied the presence or absence of a model on the number of donations to a Salvation Army kettle during the December holiday season. Again, seeing someone else donate made passersby more likely to donate as well.

TABLE 9.4 Naturalistic Studies of Models and Prosocial Behavior

	No. of helpers or donors
Setting	Model	No Model
Flat tire on highway, Los Angeles	58	35
Salvation Army kettle, Princeton	69	43
Salvation Army kettle, Trenton	84	56

Source: Data from Bryan & Test, 1967.

This effect of modeling has been demonstrated among adults reporting that they imitated their parents' prosocial behavior in settings as mundane as blood donation (Piliavin & Callero, 1991) and as historic as gentiles who rescued Jews from Nazi extermination (London, 1970) or the most fully committed whites who worked for civil rights in the southern United States (Rosenhan, 1970). Also consistent with a modeling account, prosocial behavior by family members clearly influences children's prosocial behavior (Radke-Yarrow & Zahn-Waxler, 1986). In contrast, people also imitate a model's selfish, exploiting behavior, even if it makes them indignant (Masor, Hornstein, & Tobin, 1973).

Mood Protection

People's moods reliably affect their likelihood of helping, with most good moods facilitating help and some bad moods also facilitating help, but for very different reasons (Cunningham, Steinberg, & Grev, 1980; Cunningham, Shaffer, Barbee, Wolff, & Kelley, 1990; Eisenberg, 1991). Both can be viewed as egoistic motives, if the primary cause of helping is maintaining or improving one's mood.

MAINTAINING GOOD MOODS

Good moods reliably help helping. Recall the example from Chapter 1 , in which finding a coin encouraged people to help a stranger pick up spilled papers (Isen & Levin, 1972). In 61 tests of the feeling good → helping hypothesis, a meta-analysis showed that the average study obtains a large effect (Carlson et al., 1988). Experimental manipulations of positive moods have given participants success experiences, money, gifts, pleasant music, or an instruction to think happy thoughts (Salovey, Mayer, & Rosenhan, 1991). The resulting prosocial behaviors have included picking up someone's dropped books and papers, searching for lost contact lenses, donating money or time, tutoring a needy student, donating blood, and participating in another experiment.

The large effect of cheerfulness on helping is up-and-down, which suggests some moderating variables, that is, those that increase or decrease the size of the effect. For example, the more positive the appeal for help, the more likely cheerful people are to help (Cunningham et al., 1980). Conversely, helping that requires attending to unpleasantness undermines the effects of feeling good on helping (Isen & Simmonds, 1978). A whole series of moderating variables suggests that people in a good mood will help mainly when it maintains their cheerful mood ( Table 9.5 ; Carlson et al., 1988). For example, the more the good mood improves people's social outlook (focus on the social community), the more likely the mood is to motivate helping (Holloway, Tucker, & Hornstein, 1977; Hornstein, LaKind, Frankel, & Manne, 1975). The more pleasant the task, the more people in good moods are helpful (Forest, Clark, Mills, & Isen, 1979). The more people's mood reflects self-satisfaction (i.e., their relative advantage in resources and good fortune), the more happy people help (Rosenhan, Salovey, & Hargis, 1981; Salovey et al., 1991).

TABLE 9.5 Variables that Increase the Impact of Good Moods on Helping

	Correlation with the
Variable	Mood-Helping Effect
Self as target of positive event (i.e., own relative advantage)	.49
Pleasantness of the helping task	.34
Social outlook (i.e., positive attitude toward community)	.31
Sustained helpfulness (i.e., burden over time)	−.37
Extreme high or low positive affect (i.e., deviation from moderate)	−.38
Guilt	−.44

Entries are partial correlations, which isolate the effect of the variable, controlling for other variables. The higher a positive correlation, the more the variable facilitates the effect of good moods on helping; negative numbers indicate interference with the effect. All listed variables are statistically significant, in the hypothesized direction.

In contrast, when people are happy because of someone else's good fortune, envy apparently sets in, and they are less helpful (Rosenhan et al., 1981; Salovey et al., 1991). Other kinds of helping that would damage a good mood also discourage helping: Feeling guilty decreases helping when people are cheerful (Cunningham et al., 1980). Burdensome helping that would entail an obligation sustained over time also dissuades cheerful people from helping (Carlson et al., 1988). Being in only a little bit of a good mood decreases helping because dealing with any negative experience (someone else's distress, one's own effort) would easily eliminate one's mood. And being in too good a mood decreases helping because when elated, one would not bother with someone else's problem (Carlson et al., 1988). Another way to think about these moderating variables is as boundary conditions, that is, situations in which the positive mood effect on helping does not occur, primarily cases that endanger one's mood.

Given these moderators, what might be the mediating variables—processes by which positive mood enhances helping? Mood maintenance (prolonging a good mood) fits most of the moderating variables. For example, when a model reports feeling good about helping, people are more likely to help. In one study (Hornstein, Fisch, & Holmes, 1968), Manhattan pedestrians encountered a man's wallet protruding out of an open envelope. The wallet obviously contained money, and wrapped around it was a letter from a previous finder reporting his feelings about being able to return the wallet. Participants were more likely to return the wallet intact, when the letter said:

Dear Mr. Erwin: I found your wallet which I am returning. Everything is here just as I found it. I must say that it has been a pleasure to be able to help somebody in the small things that make life nicer. It's really been no problem at all and I'm glad to be able to help.

They were also likely to help when the letter said nothing about feelings (both neutral and positive conditions would not interfere with a good mood). But they helped less when the letter said:

Dear Mr. Erwin: I found your wallet which I am returning. Everything is here just as I found it. I must say that taking responsibility for the wallet and having to return it has been a great inconvenience. I was quite annoyed at having to bother with the whole problem of returning it. I hope you appreciate the efforts I have gone through. 3

The results ( Table 9.6 ) fit the idea that people learn vicariously from the model's reported experience what their own likely experience will be. This fits an egoistic effort to maintain one's mood and to avoid bad feelings.

TABLE 9.6 Returned Wallets as a Function of Previous Finder's Experience

Finder's Reported Experience	Returned	Not Returned
Positive	14	6
Neutral	12	8
Negative	2	18

But the egoistic desire to hold onto one's good mood is not the only possible motive. For example, the research goes beyond egoistic mood maintenance in that a positive outlook must be specifically social (e.g., lives saved because of people's intervention) to enhance helping, whereas a nonsocial positive outlook (e.g., lives saved because of natural events) does not (Holloway et al., 1977). Perhaps, then, some of the mood maintenance data also fit a collectivist (group-oriented) motive, given the specifically social nature of good moods that enhance helping.

In line with this view of mood priming positive views of others, positive moods increase attraction to others, their perceived positive features, and overall positive evaluations (e.g., Isen, Shalker, Clark, & Karp, 1978). Positive moods specifically increase positive associations with helping situations and with people one expects to meet (Clark & Waddell, 1983). All this fits the idea that good moods improve people's social outlook, which in turn leads to helping. However, as noted, a social outlook could prime collectivism and belonging motives (or perhaps altruism and trusting motives), so this questions the assumption that the relationship between positive mood and helping must be egoistic.

Nonetheless, as an alternative to the egoism, collectivism, and altruism motives, the concomitance hypothesis (Manucia, Baumann, & Cialdini, 1984) holds that although a good mood makes people help, it is a side effect (mere concomitant), not intended to maintain one's mood. That is, feeling good makes people more optimistic, other people seem more attractive, and the positive features of situations more salient. In this view, any of the effects can facilitate helping, but the effect is indirect and not instrumentally directed toward mood maintenance. Alternatively, this hypothesis might simply explain the mechanisms or mediating variables.

Whatever the mediators and moderators, good moods do reliably lead to helping in many circumstances. In organizations, for example, various factors promote good moods, which in turn promote spontaneous acts of goodwill. Specific causes of good moods might include one's immediate work group and one's context, as well as dispositions and life history. Good moods in turn predict organizational spontaneity, which includes “helping co-workers, protecting the organization, making constructive suggestions, developing oneself, and spreading goodwill” (George & Brief, 1992, p. 310).

RELIEVING BAD MOODS

People sometimes help when they feel bad, but it depends on exactly how they feel bad. If they feel bad because of guilt (i.e., they feel personally responsible for an aversive outcome), that consistently encourages prosocial behavior (Carlson & Miller, 1987; Salovey et al., 1991). Notice that this differs from the way guilt decreases helping when people are originally in a good mood. Here, the point is that being in a bad mood, specifically because of guilt, increases help. When people believe they have transgressed (lied, cheated, broken equipment, harmed someone, touched museum displays, fed zoo animals), they presumably feel guilty and in a bad mood (guilt and mood are not always measured). Then, they are more likely to donate to charity, help an accident victim, pick up someone's dropped items, and volunteer for an experiment (Katzev, Edelsack, Steinmetz, Walker, & Wright, 1978; Kidd & Berkowitz, 1976; Riordan, Dunaway, Hass, James, & Kruger, 1984). If they are excused, pardoned, or cheered up after transgressing (no longer in a guilty, bad mood), they help less.

Guilt as a cause of prosocial behavior fits a model whereby people assume personal responsibility, but only when they are objectively self-aware (Duval & Wicklund, 1972). This concept refers to the idea that when people are self-focused, they compare themselves to salient ideals (and usually find themselves wanting), so they try to escape from self-awareness or correct their behavior to bring it closer to salient standards (Scheier & Carver, 1983). Participants become objectively self-aware when, for example, they see their images on television or fill out an autobiographical questionnaire (Duval, Duval, & Neely, 1979). If they are exposed immediately to a victim of disease or poverty, self-aware people are more likely to feel personally responsible and to volunteer help than they would otherwise ( Table 9.7 ). These data are consistent with the idea that guilt increases helping through self-awareness. A meta-analysis indicates that, indeed, high self-awareness, combined with salient values or requests for helping, increases the effect of bad moods (presumably guilt) on helping (Carlson & Miller, 1987).

TABLE 9.7 Mean Willingness to Help Victims of Sexually Transmitted Disease

	Immediate	Delayed
	(Self-Aware)	(Not Self-Aware)	Control
Item	Before	After	Before	After
Attribution of responsibility to self	9.0	10.3	6.9	7.9	6.5
Teach class	11.2	11.2	8.5	7.9	7.0
Volunteer work at clinic	7.6	8.9	7.6	6.0	8.2
Personal contribution	6.9	7.7	3.3	6.6	5.2
Combined helping	25.7	27.8	19.4	20.5	20.4

Numbers represent means for each condition. The higher the number, the greater the willingness to help.

Guilt as a cause of prosocial behavior also fits an attentional model, whereby attending to another person's misfortune increases empathy or increases awareness of the other person's needs (Thompson, Cowan, & Rosenhan, 1980). And indeed, a focus on the other person's misfortune, as opposed to one's own misfortune, increases the effect of negative moods on helping (Carlson & Miller, 1987). This fits an empathy-altruism motive. In much of the work on guilt, the meditating processes are not always specified but must be inferred from manipulations and outcomes.

Sadness (depressed feelings resulting from awareness of negative events) might seem to operate in the same way as guilt, but its effects on helping are more mixed. In one view, people rarely enjoy sadness and by adulthood have learned ways to avoid it. Building on this idea, people might use helping as a form of negative state relief because helping has acquired its own reward value. If one is brought up being praised for helping, then helping becomes an internalized reward, and it could relieve sadness. This fits the social learning framework of the previous section as well as the mood maintenance framework of this section.

The earliest proposal of negative-state-relief idea (Cialdini, Darby, & Vincent, 1973) involved participants induced to harm or witness harm against another person. A chair was rigged so that when it was pulled out from the desk, either the undergraduate experimenter or participant spilled three boxes of a graduate student's data, thereby putting the records completely out of order. This event—amplified by the experimenter's apparent distress—presumably put participants into a negative state (though this was not measured directly). Some of them then received an independent positive event (money or approval). Finally, all had a chance to help a fellow student by volunteering for an experiment. The witnesses and doers of harm helped only if they had received no positive event, so this fits the idea that they helped the other person in order to make themselves feel better. Those who had already been rewarded (and presumably felt better already) helped about as much as control subjects, who had never been exposed to the harm doing ( Table 9.8 ). Note that guilt alone is not a parsimonious explanation because (1) it was not measured and (2) people who had merely witnessed the misfortune (so probably did not feel guilty) behaved just as did those who had caused it.

TABLE 9.8 Negative State Relief Reduces Helping

	No Relief from Negative State	Relief from Negative State	No Harm
	No Task	Additional Task	Money	Approval
Harm doer	5.43	3.50	1.86	1.44
Witness	3.38	6.29	2.86	2.43
Neither					2.75

The means in the No-Relief and Relief conditions reflect nonoverlapping distributions and thus show a large effect.

The negative-state-relief model takes an explicitly self-serving and instrumental view of helping, and various findings are consistent with this view. People help the most when they believe their sad mood is labile (i.e., not fixed) and thus can be changed (Manucia et al., 1984). Negative moods cause helping only when the costs are low (and would not undermine mood enhancement; Weyant, 1978). And helping indeed is intrinsically rewarding (Weiss, Boyer, Lombardo, & Stich, 1973).

The negative-state-relief model is controversial. A meta-analysis (Carlson & Miller, 1987) found no support for some of the conditions that should encourage helping according to the model. According to the model, negative mood effects on helping should increase with age, as children learn to relieve their negative moods (Cialdini et al., 1981); they do not. Negative mood effects on helping should increase with the degree of bad mood; they do not. Negative mood effects on helping should decrease if the helping task itself is aversive (because it would undermine its role in mood enhancement); they do not. The failures to find results are disappointing for the model. Moreover, empathy causes people to help even when they know their mood will improve for other reasons, so that helping per se will not improve their mood (Batson et al., 1989). And some people continue to feel bad when their help fails, even though their failure may be fully justified and they could still self-reward their efforts, if mood enhancement were their main reason for helping (Batson & Weeks, 1996).

All this has led some researchers to question the generality of the negative-state-relief model (Batson, 1998; Eisenberg, 1991; Salovey et al., 1991), although others still defend it (Cialdini & Fultz, 1990; Schaller & Cialdini, 1988). Nevertheless, the results of the original studies remain, and researchers have found results consistent with the model.

In short, good moods and bad moods (when they do facilitate helping) might do so for entirely different reasons. In the case of good moods, people help because they see the good side of everything, including helping. In the case of a bad mood from guilt, they help out of personal responsibility and obligation (Cunningham et al., 1980). In the case of a bad mood from sadness, they might help to relieve sadness, at least some of the people, some of the time. Some evidence suggests that people in good moods, compared with neutral or bad ones, think the hardest about the affective consequences of their helping (Wegener & Petty, 1994).

Summary of Egoism Hypotheses

Egoistic motives for helping could include genetic kin selection, whereby people help others in direct proportion to their genetic relatedness, a phenomenon that occurs cross-culturally. Possible mechanisms might have evolved for helping as a way to perpetuate one's genes: recognizing kin and detecting distress, both of which people do well and early in life. However, some of the variables that mediate between detecting the distress of kin and coming to their aid are compatible with other theories. Emotional closeness and empathy could fit kin selection, but they could also fit more altruistic motives, throwing into doubt kin selection (sometimes interpreted to be pure egoism) as a sole motive.

Social learning is another mechanism proposed to reflect egoistic motives. People learn that helping other people pays off. So they might help others only because they expect to be helped in return. But the reciprocity principle also fits equity and exchange theories of social norms. In short-term, artificial experimental games, people quickly evolve tit-for-tat rules of reciprocity. People are alert to detect cheaters who fail to abide by these social rules. And people do believe in enlightened self-interest, namely that prosocial behavior pays off in the long run.

Besides reciprocity, people help when the social rewards outweigh the social costs. They learn these contingencies through vicarious conditioning or observational learning, but especially by imitating the behavior of others, in the process called modeling (all introduced in Chapter 6 ).

Finally, people might help others mainly in order to protect their own moods. Cheerful people do help more, perhaps to maintain their good moods (egoism) or perhaps because they see the good side of everything. When people feel good, they do see good in others and in the social outlook more generally. When people feel personally advantaged, they are more likely to help than when they are happy because of someone else's good fortune. And unpleasant helping does not increase under good moods. Although these findings fit an egoistic self-enhancement motive, they also fit altruistic, collectivist, and perhaps even principled helping.

People in bad moods clearly help when the mood is guilt. Feeling responsible for another's misfortune reliably increases helping, which may relieve guilt, a negative mood. When the bad mood is sadness, the results are more mixed, and researchers disagree. Sad people do help under some circumstances that seem designed to make themselves feel better, but sad people also help sometimes when it will not improve their mood. The negative-state-relief hypothesis continues to provoke controversy.

As a final note, the basic premise of this whole argument—that self-interest and other-interest (prosociality) are inversely correlated—may simply be wrong (Gerbasi & Prentice, 2013). Each contributes independently to behavior, depending on values accessible in context.

ALTRUISM HYPOTHESES: MAINTAINING TRUST IN THE WORLD AS BENEVOLENT

Why would Art offer to pay for a nursing home for his live-in partner Ruth's mother? Kin selection is unlikely, because he does not expect to have children with Ruth. Nevertheless, evolutionary psychologists would argue that the evolved mechanism for kin selection might spill over to surrogate relationships that mimic the conditions of kin selection. Social learning is also plausible, specifically, belief in reciprocity, although Art denies expecting anything in return. Mood might be an explanation, although he makes the decision over a period of time when he doesn't seem particularly happy, guilty, or sad. One could make parallel arguments about Ms. McCarty. Maybe, instead of all these motives that social psychologists have interpreted as egoistic, just maybe their motives are pure altruism.

Altruism, as introduced earlier, constitutes the motivation to help out of a concern for others and their welfare. In this view, people are motivated by other-interest. Altruism goes with a belief that people are responsible for each other and that they deserve help. As such, it fits with motives to believe the best of other people, the core social motive here termed trusting. Trust in others goes with a concern for others. Individual differences in empathy correlate with the tendency to trust others (Piliavin & Charng, 1990). People have a strong need to believe that the world is benevolent and will go to great lengths to restore that belief (Janoff-Bulman, 1992).

Evidence for altruism consists partly in showing that prosocial behavior is (1) based on processes demonstrating concern, such as attributions of responsibility to self, and (2) partly based on showing that people and situations high on empathy tend to facilitate helping. The best situational predictors of helping are the victim's unambiguous and severe need, attractive appearance, and similarity to the potential helper. Also, more bystanders present (as noted earlier) and being in a big city (with stimulus overload) are factors that reliably decrease helping (Batson, 1998). All these situational factors impact people's felt responsibility and empathy, consistent with altruistic motivations.

Attributions of Responsibility

In New York City's Queens borough, Kitty Genovese was repeatedly stabbed, over a period of 30 minutes, and died, screaming for help, as 38 of her neighbors watched from their homes, without intervening or even calling the police. Why?

People confronting other people's misfortunes are more likely to help if they attribute responsibility to themselves for alleviating the victims' distress, if they do not see others as responsible, and if they do not see the victims as responsible for their own distress. Research suggests that the neighbors did not attribute personal responsibility for helping, thought other neighbors might help, and may have thought the assault was a lovers' quarrel, thereby apparently implicating the victim as partly responsible. We address each factor in turn.

HELPERS' FELT RESPONSIBILITY

The classic studies in prosocial behavior are the series on bystander intervention, introduced earlier ( Table 9.3 ). The basic counterintuitive finding shows that the more bystanders present, the less likely a given person is to help (Darley & Latané, 1968; Latané & Darley, 1968). The explanation is that, the more people present, the more the witnesses experience a diffusion of responsibility. That is, people believe that others may take action, so they are not personally responsible. In the Genovese case, the neighbors apparently assumed that someone else would intervene if appropriate. The crucial process here is felt responsibility. What does it take for people to feel that they personally must take action on someone else's behalf? Several steps pave the way for bystander intervention. People must

· Notice the events

· Interpret the events as an emergency

· Feel personal responsibility for acting

· Consider what form of assistance is needed

· Implement action (Latané & Darley, 1970)

The first three steps are less likely the more other people are around and not reacting. In one study, male undergraduates sat in a room that gradually began to fill with smoke as they completed questionnaires. When alone, 75% reported the smoke, but when in groups of three participants, only 38% of participants responded, showing that people respond less to an emergency, the more of them there are. Consistent with the idea that it is the passivity of the other bystanders that inhibits people, when the participant was in a group of three with two confederates trained to be passive, only 10% reported the smoke. “A crowd can … force inaction on its members” (Latané & Darley, 1968, p. 217). If everyone else seems to interpret the event as not serious, then the apparent consensus favoring nonintervention will guide the onlooker. The irony, of course, is that as each person tries to appear calm, while covertly watching other people's reactions, everyone is fooled into thinking no one is worried. This exemplifies pluralistic ignorance (previously introduced in Chapter 2 ).

Bystanders have inhibited intervention for victims of seizures (Darley & Latané, 1968; Schwartz & Clausen, 1970), theft of belongings in a library (Shaffer, Rogel, & Hendrick, 1975), shoplifting in a supermarket (Bickman & Rosenbaum, 1977), and preventing a peer's driving while drunk (Rabow, Newcomb, Monto, & Hernandez, 1990). The interventions seem to turn on people's felt responsibility, which emerges from the decision-making stages just described (Bickman & Rosenbaum, 1977).

VICTIMS' ASSIGNED RESPONSIBILITY

Attributions of responsibility depend not only on potential helpers but also on the victims' apparent responsibility for their own plight. People are less likely to help others they perceive as responsible for their own misfortune. In an attributional theory of motivation (Weiner, 1980, 1985), causes can be internal or external, as in other attribution theories that contrast, respectively, dispositional and situational causes ( Chapter 3 ). In addition, causes can be controllable or uncontrollable, that is, subject to human volition. People are least likely to help someone whose misfortune is attributed to internal, controllable factors ( Table 9.9 ), that is, when it's their own fault. A student who has no notes to study because he didn't bother to attend class (internal, controllable) will receive less help than a noteless student who could not attend class for internal, uncontrollable reasons (broken leg) or for external reasons, either controllable (the TA did not bother to communicate a room change) or uncontrollable (the TA was unable to communicate a room change).

TABLE 9.9 Attributions of Responsibility to Victims

Locus	Controllable	Uncontrollable
Internal	Disgust, anger	Pity, sympathy
	No help	Help
External	Pity, sympathy	Pity, sympathy
	Help	Help

As the table indicates, different attributions elicit different emotions. Internal, controllable problems (the person's own fault) generate anger, disgust, and distaste, as well as neglect. External and uncontrollable misfortunes generate sympathy, pity, and concern, as well as helping. Recall from Chapter 3 that such attributions for poverty generate differing policy implications for degree of government aid (Zucker & Weiner, 1993). Similarly, physical stigmas (e.g., disabilities, either congenital or acquired) elicit pity and helping, because they are usually considered uncontrollable and external. In contrast, mental-behavioral stigmas (e.g., mental illness, alcoholism) elicit less pity, more anger, and neglect because they are considered internal and controllable (Weiner, Perry, & Magnusson, 1988). As another example, when researchers staged a person's collapse on a subway car, an apparently ill person (uncontrollable) received more help than an apparently drunk person (controllable) (Piliavin, Rodin, & Piliavin, 1969).

Political conservatives put more weight on perceived personal responsibility than liberals do, as noted. In one study, conservatives withheld assistance from people seen as personally responsible for their fate, across policy domains (e.g., prescription subsidies for people with AIDS, low-income housing, help to victims of natural disasters). Their reported motivations emphasize punishing people who violate social norms and deterring free-riders (Skitka, 1999; Skitka & Tetlock, 1993). Liberals tend to help everyone equally, their reported motivations aiming to avoid awkward social value trade-offs that place monetary value on people's lives. That is, compared with conservatives, they do not penalize people responsible for their own fate more than people not responsible. In a related vein, religious fundamentalists hold different attributions about outgroup members who do and do not contradict their values. They view single mothers and homosexuals as responsible for their predicament (e.g., unemployment) but do not hold students and Native Canadians responsible for the same predicament (Jackson & Esses, 1997). Regardless of where they stand, most people are morally outraged at putting a price tag on their sacred values, as we will see (Tetlock, Kristel, Elson, Green, & Lerner, 2000).

COMBINED RESPONSIBILITY FOR CAUSE AND CURE

People are concerned not just with causes but also with cures. Combining responsibility assigned to victims for causes and cures, along with responsibility felt by witnesses for causes and cures, creates four models of helping and coping (Brickman et al., 1982). The distinction between responsibility for a problem and its solution has several implications: who should act, how to perceive the victim, how to view human nature, and what the potential pathologies are ( Table 9.10 ). Helping-coping models tend to vary across helping professions. For example, a medical model absolves the victim of responsibility for either cause or cure: The victim is ill because humans are weak and must accept treatment by experts. In a moral model, espoused by some religions, the victim is responsible for both the problem and its solution: The victim is lazy and must strive with peers, who exhort each other because humans can be strong. In an enlightenment model, espoused by some religions and some cults, victims are responsible for the cause because humans are basically bad or sinful, so they are guilty and must submit to the discipline of authorities. Alcoholics Anonymous, for example, requires people to admit their responsibility for their drinking problem but also to admit they are powerless to restore themselves and must turn themselves over to a higher power, who will remove their shortcomings; the label enlightenment reflects the need to have a spiritual awakening, according to the Alcoholics Anonymous 12-step program. Finally, the compensatory model reflects Jesse Jackson's reputed motto: “You are not responsible for being down, but you are responsible for getting up.” Victims are deprived, but they must assert themselves, along with other subordinates who must mobilize because humans are basically good and can bring about change for the better. Although largely untested by social psychologists, other psychologists have applied these four models to understand people's attributions for individual unemployment, depression, alcoholism, self-poisoning, counseling, cancer care, elder care, and dissertation completion (author's PsycInfo search, 2003).

TABLE 9.10 Consequences of Attribution of Responsibility in Four Models of Helping and Coping

Attribution to Self of	Attribution to Self of Responsibility for Solution
Responsibility for Problem	High	Low
High	Moral model	Enlightenment model
Perception of self	Lazy	Guilty
Actions expected of self	Striving	Submission
Others besides self who must act	Peers	Authorities
Actions expected of others	Exhortation	Discipline
Implicit view of human nature	Strong	Bad
Potential pathology	Loneliness	Fanaticism
Low	Compensatory model	Medical model
Perception of self	Deprived	Ill
Actions expected of self	Assertion	Acceptance
Others besides self who must act	Subordinates	Experts
Actions expected of others	Mobilization	Treatment
Implicit view of human nature	Good	Weak
Potential pathology	Alienation	Dependency

The more severe the problem, the more people want to believe someone is responsible for it; this is termed the defensive attribution bias, introduced in Chapter 3 . People will go to considerable effort to figure out why a personal tragedy allegedly occurred. This aspect of the attribution process fits our core social motive of understanding. But more broadly, as noted, explaining negative events helps people's core motive to maintain trust in the benevolence of the world and especially other people (Janoff-Bulman, 1992). When people's assumptions about trust are shattered by tragedy, one of their main tasks is to rebuild that basic trust, by attributions of responsibility and by trying to relieve distress.

Empathy, Sympathy, and Altruism

As we have just seen, attributions of responsibility can result in prosocial feelings and actions. Other kinds of feelings also guide prosocial behavior. Empathyis an affective response that mimics another person's emotional state (Eisenberg, 2000). When people observe another person experiencing an aversive event, at least one of the same neural areas (i.e., the amygdala) activates as when they personally fear experiencing the event (Olsson, Nearing, & Phelps, 2007). But people also have to differentiate self and other, so other neural areas implicated in empathy also include the insula, the anterior cingulate cortex, and the right temporo-parietal region (Decety & Jackson, 2006). Among other evidence, all this implies a neural basis for empathy and social learning. Further, altruism in response to another can operate through empathy even in mammals and birds via matching emotional states (de Waal, 2008).

Usually, empathy turns into either sympathy or personal distress (Eisenberg, 2000). Sympathy entails feeling compassion or tenderness for another person's distress. Sometimes sympathy is interchangeably called empathy (Batson, 1987, 1998), and both are sources of altruistic motivations. However, the difference between truly feeling what the other person feels (empathy) and compassionately apprehending what another person feels (sympathy) seems useful. Knowing what someone feels is not the same as experiencing similar feelings. Indeed, people trying to take someone else's perspective often erroneously start with their own experience and adjust inadequately, to imagine the other's experience (Epley, Keysar, Van Boven, & Gilovich, 2004); most likely their well-intentioned feelings are inaccurate. In contrast to both sympathy and empathy stands personal distress, which entails one's own aversive feelings at perceiving another person's distress but not feeling what the other person is feeling. For example, one might be upset at seeing an unconscious accident victim, whereas the victim might be feeling nothing. Personal distress is self-oriented, whereas empathy and sympathy both are other-oriented.

EVIDENCE

Personal distress is empirically distinct from sympathy (and empathy): Distress leads to egoistic motivation, whereas sympathy leads to altruistic motivation. For example, the more personal distress people report (alarmed, grieved, upset, worried, disturbed, distressed, troubled, perturbed), the less they help when they can escape the situation. In contrast, the more sympathy they report (sympathetic, moved, compassionate, warm, soft-hearted, tender), the more they help, regardless of whether they can escape the situation or not (Batson, O'Quin, Fultz, Vanderplus, & Isen, 1983). Similarly, in both children and adults, sympathy and personal distress show distinct patterns of heart rate. Deceleration, which typically correlates with taking in information, interest, and an outward focus, correlates with sympathy. Acceleration, which typically correlates with cognitive problem solving and active coping, correlates with distress (Eisenberg et al., 1989). A variety of indicators show the same distinction. In both children and adults, empathy and sympathy (as indicated by facial expressions, behavior, and heart rate) correlate with prosocial behavior and prosocial dispositions. In contrast, personal distress indicators correlate either negatively or not at all with prosocial behavior (Eisenberg, 2000). Given the mixed effects of negative mood on helping (covered earlier), perhaps this is not surprising.

Assorted evidence fits the motivating role of empathy (Hoffman, 1981). These findings indicate that empathic arousal (a) precedes helping, (b) increases with the severity of the victim's distress, and (c) correlates with helping. For example, one study (Coke, Batson, & McDavis, 1978) tested the hypothesis that taking the other's perspective leads to empathy, which in turn increases helping. Undergraduates heard a newscast about a fellow undergraduate whose parents had recently died in a car crash. She was struggling to support her younger sibling while finishing college and needed help with childcare, transportation, and errands. Students listened to the newscast, trying either to imagine her feelings (empathy set) or to identify the media techniques (observation set). Further, they had been given a capsule just before the newscast and were told either that it would arouse them or that it would calm them. Participants in the empathy-arouse combination would think their distress was due to the pill and ignore it. In contrast, those in the empathy-calm combination would not be able to misattribute their arousal to the pill, and they should be especially likely to help if empathy motivates helping. Participants in the observe condition (whether in combination with the calming pill or arousing pill condition) should not especially help because they had failed to take the role of the other, so they would not feel empathy. Results supported the two-stage model of perspective taking and empathy leading to help. Altogether, a variety of evidence supports the role of empathy, consistent with the altruism hypothesis.

CONTROVERSY

However, the relationship between helping and sympathy (or empathy) still could result from egoistic or altruistic motivations (Batson, 1998; Piliavin & Charng, 1990). That is, empathy and sympathy could conceivably entail egoistic motives: reducing one's own empathic distress, avoiding empathy-specific punishments, gaining empathy-specific rewards, and self-other merging. Luckily for the empathy-altruism hypothesis, researchers have tackled each of these proposals and so far have found them wanting.

Reducing one's own empathic distress might seem to resemble the negative-state-relief model, discussed earlier as sadness, with the conclusion that much evidence fails to fit this idea. Most relevant here, sadness does not account for the effects of empathic concern on helping (Dovidio, Allen, & Schroeder, 1990). However, distinguish negative state relief, which refers to egoistic self-reward, from empathic distress. The latter creates a tension state and a drive to reduce that tension (recall drive reduction models from Chapter 6 ). As we have just seen, empathy does mediate helping, but the idea that acting on empathy reflects an egoistic motive to reduce distress seems unlikely, given the evidence showing that empathy and personal distress are distinct.

Empathy-specific punishments consist in one's experience that if one feels empathy and fails to help, then censure, embarrassment, shame, and guilt will follow as negative consequences (Tangney, Stuewig, & Mashek, 2007). Empathy-specific rewards include praise, gratitude, elevation, honor, and pride. In contrast to this punishment-reward view, recall that altruism entails helping purely in order to improve the other's welfare. In one set of studies that parallel one just described (Batson et al., 1988, Studies 2 & 5), students listened to an audiotaped interview of a fellow student whose parents and sister had recently been killed in a car crash, leaving her to struggle through her last year of school, as she attempted to support her two remaining siblings. When induced to feel empathy by focusing on how the target feels (as opposed to focusing on technical aspects of the audiotape), people helped more ( Table 9.11 , lines 1 and 2). As an indicator of the empathy process, the high-empathy helpers also were more distracted by victim-relevant words (loss, needy, adopt, tragic), as shown ( Table 9.11 , lines 3 and 4). And they were less distracted by words that were reward-relevant (nice, proud, honor, praise) or punishment-relevant (duty, guilt, shame, oblige). Also, the more participants were distracted by the victim-relevant (empathy) words, the more they helped, which was not true for reward and punishment-related words (last three lines of table). In short, the process was more consistent with empathy leading to victim-relevant thoughts, which in turn encourage helping, than with a process anchored in empathy-specific rewards and punishments.

TABLE 9.11 Empathy as a Predictor of Helping

	Low Empathy	High Empathy
Helping (percent)	35%	70%
Helping (amount)	.45	1.20
Responsiveness to victim-related words
Helpers	−29.11	47.32
Nonhelpers	−24.66	−44.22
Helping predicted by response to
Victim-related words	−.06	.62
Reward-relevant words	−.15	−.30
Punishment-related words	−.29	−.30

The first two lines come from Study 2; the remainder come from Study 5. For amount of helping, 0 = no help, 1 = 1 to 2 hours; 2 = 3 to 5 hours; 3 = 6 to 8 hours; 4 = 9 to 10 hours. Response to words was measured by the degree of interference (distraction, measured as response latency or time delay) that the words created when participants attempted to ignore their meaning and just name the font's color. Higher numbers indicate slower times to name the ink color and therefore greater interference (i.e., distraction) by the meaning of the words themselves.

A variant on the empathy-specific reward hypothesis relies on the notion of empathic-joy (Smith, Keating, & Stotland, 1989), defined as sensitivity to the emotional state of the victim, which causes relief, pleasure, and happiness when the victim feels those emotions, as a result of being helped. This differs from empathy-specific rewards because it focuses on experiencing what the other person experiences, the defining feature of empathy. It also differs from the negative-state-relief model because it emphasizes empathy per se and because it is an intrinsically positive experience, not just a contrast with a negative experience. But it resembles the negative-state-relief model in focusing on rewards to self. In this view, only empathic people will help and only when they can receive feedback that the victim feels better, which makes them experience empathic joy, the presumed goal of helping, in this model.

Indeed, in one study, participants had the opportunity to write advice to a first-year student (viewed on videotape) who felt so stressed and isolated that she considered dropping out. People high in self-reported empathy were far more likely to help (93%) when they knew they would receive feedback from her after she had read their advice, compared with high-empathy participants not expecting feedback (63%) or low-empathy participants in either feedback condition (53% in both cases). Similar effects occurred when instructions manipulated an empathy goal (“try to take the perspective of the person being interviewed”), compared with a detached observer goal (“watch all of this person's body movements … try to be objective”). Consistent with the empathy-altruism hypothesis, all the empathy goal participants helped more than did the observer-goal participants. But suggestive of the empathic-joy hypothesis, the empathy-goal participants especially helped when feedback was expected. This study's results seem to fit the empathic-joy hypothesis because none of the other models predicts that feedback from the victim should differentially affect high-empathy witnesses.

Nevertheless, the empathy effect is bigger than the feedback effect, and subsequent studies (Batson et al., 1991) failed to replicate the empathic-joy effect. Moreover, researchers reasoned that high-empathy participants might be interested in feedback from the victim, to see how she was doing, regardless of whether she would in fact improve and reward them with their own experience of empathic joy. Recall that the empathy-altruism hypothesis predicts that empathy creates an interest in the victim's welfare, regardless of how well she is likely to be doing. In contrast, the empathic-joy hypothesis predicts that her probability of improvement should predict the chance of empathic joy. Consistent with the empathy-altruism hypothesis, level of empathy did in fact predict higher rates of wanting to see how the victim was doing, regardless of how likely it was that she would indeed be better. These results contradict the empathic-joy hypothesis and support the empathy-altruism hypothesis.

Despite all the evidence favoring altruism, efforts to find egoism in altruism continue. The self-other merging hypothesis goes beyond empathic joy to propose that, under empathy, one's representation of the other person merges with the self. Recall from Chapter 8 the theory of relationships that defines intimacy in terms of self-other overlap. Extending that idea, any situation in which people take the perspective of another could involve merging of self and other (Davis, Conklin, & Luce, 1996); empathy certainly takes the other's perspective. Participants instructed to imagine the perspective of a videotaped target or to imagine themselves as the target both contrasted with participants instructed to observe neutrally the target's nonverbal behavior. The two role-taking groups both ascribed positive traits to the target person that overlapped with their own prior self-descriptions. These results fit the idea that empathy generates self-other overlap.

There ensued a controversy regarding the meaning of self-other overlap in helping behavior. If empathic concern leads to self-other overlap, does that mean that helping the other really equals helping the self? Is that behavior therefore egoistic, not altruistic? From one perspective, oneness or self-other overlap constitutes a nonaltruistic explanation consistent with kin selection ideas (Cialdini, Brown, Lewis, Luce, & Neuberg, 1997). Measuring oneness as the inclusion of other in the self scale ( Chapter 8 ) and as a self-reported usage of “we” to describe the relationship, researchers found that relationship closeness predicted both empathy and oneness but that only oneness, not empathy, predicted helping. Consistent with a kin selection idea, severity of need also predicted empathy and helping, but again helping was independent of empathy. (Note, however, that other studies have clearly linked empathy and helping, so this failure to replicate is puzzling.)

In response, defenders of the altruism hypothesis argued in part that the studies (a) did not manipulate empathy, only relationship closeness, and (b) used written scenarios and self-reports, not a real person and actual volunteering to help. Their own studies found that empathy did predict helping, without regard for group membership, and that self-other merging did not mediate (account for) the effect (Batson, Kobrynowicz, Dinnerstein, Kampf, & Wilson, 1997). The egoism camp replied that the offered helping was only superficial (letter-writing) and that severity of need was an empathy manipulation (Neuberg et al., 1997). Not surprisingly, the empathy theorist Batson (1997a, 1997b) disagreed, leaving readers to decide.

One resolution to the egoism-altruism controversy is that some situations and some people do genuinely demonstrate helping for purely selfless reasons, and others do genuinely demonstrate help for selfish reasons. Such a resolution supports the altruism perspective, which contends only that empathic altruism can sometimes motivate helping, regardless of frequency or prevalence. However, this resolution would not equally support the egoism perspective, which contends that altruism can never constitute the primary motive for helping; an egoistic motive must always underlie all seemingly prosocial behavior. Regardless of one's view of the weight of the evidence in this ongoing controversy, however, manipulated empathy and measured empathy do predict helping. What's more, people do differ in levels of dispositional empathy, as the next section shows.

INDIVIDUAL DIFFERENCES IN EMPATHY

Various approaches demonstrate that people vary in their inclination to be concerned for the welfare of others. Concerning empathy per se, one of its dimensions involves ascription of responsibility, perspective-taking, social responsibility, and sympathy, whereas another dimension involves personal distress and emotional intensity (Carlo, Eisenberg, Troyer, Switzer, & Speer, 1991). Everyone is more likely to help under strong situational constraints (in this study, an evocative need and a situation difficult for the observer to escape, as compared to less evocative and easy-escape conditions). Dispositional differences emerge when the situation is weaker: When the situation is easy to escape but the portrayal is evocative, then dispositional empathy predicts sympathy in the moment (feeling sympathetic, touched, soft-hearted, compassionate, sorrowful, and concerned for others), as well as helping. In parallel with results noted earlier, dispositional empathy but not personal distress predicts helping (Eisenberg, 2000). These dispositional empathy results have replicated in Japan (Misumi & Peterson, 1990).

In real-world settings, certain dispositional variables reliably predict real-world helping, such as volunteering to help people with AIDS (Penner & Finkelstein, 1998). For example, other-oriented empathy (concern for the welfare of others, satisfaction from being helpful, feeling responsible for others' welfare) predicts time spent volunteering. Other individual differences relevant to prosocial behavior include dominance, self-efficacy, self-confidence, and competence. (The motive to help in these cases relates to the core social motive of controlling, not altruism and trust.) As further evidence related to real-world helping, dispositional pity and empathy also distinguished 231 gentiles who saved Jews during World War II from 126 nonrescuers matched on age, sex, education, and location (Oliner & Oliner, 1988). ( Table 9.1 shows personality variables related and unrelated to helping.)

Children learn empathy and sympathy from parents who are sympathetic; such parents focus on understanding others' feelings, express little hostile emotion, and help children cope with their own negative feelings (Eisenberg, 2000). Cross-culturally, some results are similar. Hong Kong Chinese children's altruistic orientation results from families that are cohesive and harmonious, with little anger, aggression, and conflict (Ma & Leung, 1995). In the same setting, an orientation to work well with others (belonging) and to gain skill (controlling) predicts prosocial behavior, whereas a personal competence orientation (self-enhancement) does not (Cheung, Ma, & Shek, 1998).

In general, the tendency toward empathy correlates with the tendency to trust others (Piliavin & Charng, 1990), affirming the core social motive of trusting. Cooperators view cooperation (which entails trust) as good, whereas competitors view cooperation as weak. Cooperators especially expect cooperation more from honest (good) people, whereas competitors expect cooperation more from unintelligent (weak, stupid) people (Van Lange & Kuhlman, 1994). Although these results come from experimental games (noted earlier in this chapter), where cooperation often improves pay-off in the long run, their analogy to helping as another form of prosocial behavior is apt. Thus, those low on trust are likely to see trust as reflecting gullibility and not to appreciate its role in prosocial behavior. These individual differences in both trust and trustworthiness may have a biological basis in the hormone oxytocin, which, as the relationships chapter noted, also increases social bonding (Kosfeld, Heinrichs, Zak, Fischbacher, & Fehr, 2005; Zak, Kurzban, & Matzner, 2005).

Summary of Altruism Hypotheses

Altruism begins when people avoid diffusion of responsibility to others (bystanders or the victim) and go through the steps toward felt responsibility for another's misfortune. Moreover, people are more likely to feel pity and to help victims seen as not responsible for their own plight, that is, when attributing responsibility to external, uncontrollable causes. Helping varies by attributed responsibility for both cause and cure, with varying models in different helping professions.

Empathy means feeling another person's feelings, and it results in either sympathy (apprehending the other person's feelings), which promotes helping, or personal distress, which does not. Ongoing controversy pits the empathy-altruism hypothesis against egoistic alternatives. If empathic people help merely to relieve personal distress, that would fit the egoistic negative-state-relief model, but the evidence does not support that view. If empathic people help merely to avoid empathy-specific punishment, that also would be egoistic, but evidence contradicts that view as well. Alternatively, empathic people could help in order to obtain empathy-specific rewards or to experience empathic joy, but empathic people seem to help even in the absence of those rewards. Finally, egoistic and altruistic theories differ regarding self-other merging and the feeling of oneness. If empathic people's self merges with the other they help, does that make empathy selfish or selfless? Regardless of one's endorsement of either the egoistic or altruistic model, people and situations do differ in empathy and perhaps therefore also in the motivations to help. Empathy develops from a family environment that models empathy.

COLLECTIVISM: MAINTAINING GROUP BELONGING

Self-interest and other-interest, discussed so far, do not exhaust the possibilities for motives to explain prosocial behavior. Collective interest reflects a concern with the group's welfare. It not only differs from but can conflict with both self- and other-interest at the individual level (Batson et al., 1995). For example, in one study, students received 16 raffle tickets that they could distribute to themselves, to another individual group member, or to the group as a whole. Here's the catch: If they gave the tickets back to the group, their 16 tickets would become 24, which would then be divided equally among group members ( Figure 9.3 ). If everyone in the 4-person group donated their tickets back, the group resources would go from 64 to 96, and each group member would end up with 24. If the other 3 participants donated their tickets back, but the participant kept his or her own tickets, the participant would get the 16 original tickets plus 18, making 34 total, while other members would receive 18 each. But if the participant were the only one to donate tickets back, and other members kept theirs, the participant would get only 6 tickets and the others would get 22 each. These conditions set up a social dilemma, in which self-interest conflicts with group interest; they expand the two-person game dilemmas introduced earlier (see Figure 9.1 and Chapter 12 ).

Figure 9.3 Payoff Matrix for a Four-Person Social Dilemma Involving Prosocial Cooperation and Individualistic Competition

The prosocial possibilities had still another level. The social dilemma allowed participants further to choose to give half their own original tickets to one other student, a victim of a recent break-up from a long-term relationship, who needed “something good to happen to cheer me up.” Researchers predicted that empathy for this individual would undermine the group good, if empathic people donated some of their tickets directly to the victim, thereby leaving fewer for the group pot.

Researchers measured naturally occurring empathy after participants read the note, by self ratings on feeling sympathetic, warm, compassionate, soft-hearted, tender, and moved. (Another study obtained similar results by manipulating empathy.) As Table 9.12 shows, low-empathy participants were most likely to act in the collective interest, self-interest, or the two combined, while high-empathy participants were most likely to act in a combination of self- and other-interest, that is, to give half their tickets to the victim. Their other-oriented altruism at the individual level undermined the collective good (also see Batson, Thompson, Seuferling, Whitney, & Strongman, 1999).

TABLE 9.12 Allocation of Raffle Tickets in Self-Interest, Other-Interest, and Collective Interest

	Self-Reported Empathy
Allocation of Tickets	Low	High
Both halves to self (self-interest)	6	4
Both halves to group (collective interest)	9	5
Half to self; half to group (collective and self-interest)	7	5
Half to self; half to victim (other- and self-interest)	1	8
Total number of participants	23	22

Thus, the collective interest is separable from egoistic self- and altruistic other-interest. Collectivism reflects the motivation to benefit one's own group as a whole, and it is separate from individual-level empathy. For example, in a social dilemma such as the one just described, people have to balance the group good against individual good (both their own and other people's). When group members have a chance to discuss allocation, they are more likely to act for the collective welfare (Caporael, Dawes, Orbell, & van de Kragt, 1989). Although discussion could instill other egoistic motivations (social reinforcement, for example), the finding fits other data on collectivist motives. In Chapter 12 , we see other examples that differentiate personal (self), interpersonal (other), and collective (group) phenomena as distinct, but this chapter for now addresses just three collective phenomena of prosocial behavior, focusing in turn on the role of similarity and group identity, group norms, and individual differences in group prosocial orientation.

Similarity and Group Identity

Recall that when good moods specifically enhance people's social outlook—their positive feelings about other people—they are more likely to help. Consistent with the more specific idea that people can be motivated to help group welfare, people are more likely to help in a variety of circumstances that reinforce their sense of group identity. For example, people are more likely to help a stranger whose opinions are similar to their own. However, the effect seems to depend on developing a sense of we-ness (Flippen, Hornstein, Siegel, & Weitzman, 1996; Sole, Marton, & Hornstein, 1975; Wagner, Hornstein, & Holloway, 1982). Attraction seems to be less the issue than group identity, particularly under external threat, which increases group loyalty (Flippen et al., 1996; Sole et al., 1975). That is, people are more likely to help members of their own group when their group identity is salient.

In one indirect manipulation of groupiness, investigators found that reminding people of their own mortality caused them to be more helpful (Jonas, Schimel, Greenberg, & Pyszczynski, 2002). Typically, this kind of mortality salience (priming thoughts of one's own death) makes people affirm cultural values that will endure beyond their lifetimes. In the case of prosocial behavior, people reminded of their own mortality (thinking about their own death) are more helpful, but only to their own group, in this case, an American charity, not a foreign one.

People sometimes help those in their own group more than those in another group (Batson, 1998; Crosby, Bromley, & Saxe, 1980), particularly when they have an excuse (see Chapter 11 ; Frey & Gaertner, 1986; Gaertner & Dovidio, 1977; Gaertner, Dovidio, & Johnson, 1982). However, this bias toward one's own group vanishes when two previously separate groups merge to form a one-group identity. As Chapter 11 indicates, in one larger group, they start to help all group members equally (Dovidio, Gaertner, Validzic, Matoka, Johnson, & Frazier, 1997). As a fictional example, perhaps Art helped Ruth's mother because he was, in translating his solitary “I” into “we,” showing his identification of them as a family unit.

Group identity clearly facilitates helping. People differ in their degree of group-prosocial orientation, depending on group identification. For example, group empathy better predicts helping ingroup recipients, compared with outgroup ones, and perceived similarity among ingroup members strengthens the empathy → ingroup-helping relationship (Stürmer, Snyder, Kropp, & Siem, 2006; Stürmer, Snyder, & Omoto, 2005). The importance of identification on group-level helping also appears in evidence that neighborhood stability encourages community identification and pro-community behavior (Oishi et al., 2007).

People may vary in the size and type of group with which they identify. For example, men may tend to identify with the broad social matrix and help strangers more often, whereas women may identify more with close relationship and help intimate others more (Baumeister & Sommer, 1997). Both levels of identification can serve the social belonging motive, but the desire for connection may operate at different levels of group.

Scattered cultural comparisons suggest that Westerners may define a level of group that makes them appear more selfish with regard to helping their group than Easterners are. For example, Americans do not feel responsible even for siblings and colleagues if they don't happen to like them, whereas Asian Indians feel responsible regardless of liking (Miller & Bersoff, 1998). Similarly, Americans do not feel obligated to help either best friends or strangers even in cases of moderately serious need, whereas Indians feel obligated even to strangers (Miller, Bersoff, & Harwood, 1990). The exception, for Americans, comes with parents caring for their own children (regardless of liking) or anyone with a life-threatening need, whom they do feel obligated to help. Americans and independent cultures may view helping as best coming from internal sources, when it is spontaneous, but feel externally obligated when helping is forced on them by reciprocity, which seems too utilitarian (Miller & Bersoff, 1994). In contrast, Indians and interdependent cultures may view helping as a broadly applicable affirmative personal choice, whether driven by norms of reciprocity or not. In perhaps related findings, Japanese (interdependent) children gave equal say to a minority within their own classroom, whereas Australians (independent) followed self-interest and favored their own group (Mann, Radford, & Kanagawa, 1985). Again, the more collectivist culture shows a broader, less self-serving sense of their prosocial obligations. The relevant group to be helped depends on local definitions, but similarity and identity are important everywhere.

Norms

However defined, groups develop norms, unwritten rules for who should help whom, and when. The reciprocity norm, discussed earlier, is one example. Another is social responsibility, the idea that people are responsible for one another's welfare. As we saw earlier, though, the social responsibility norm carries several exceptions. Based on attributional evidence, potential helpers must feel personally responsible, must feel that others are not, must attribute the victim's fate to uncontrollable or external factors, and must attribute responsibility for the cure to people like themselves.

The effects of norms on helping are surprisingly difficult to detect (Batson, 1998). One reason may be that the salience of prosocial norms varies across situations, and behavior follows the norm only when it is salient (Cialdini, Kallgren, & Reno, 1991). For example, not littering (arguably a prosocial behavior) is influenced by the situational salience of anti-littering norms. Norms can be salient based on the behavior of other people (as in the modeling that is characteristic of social learning). For instance, people litter less in litter-free settings. The setting communicates a descriptive norm that says what people actually do, what is.

In addition, the setting can communicate an injunctive norm that says what ought to be. The development of social norms requires sanctions, and people will experience rewards and incur personal costs to sanction norm-violators, especially when the victim is an ingroup member; this might be termed altruistic punishment in support of moral norms (Bernhard, Fischbacher, & Fehr, 2006; de Quervain et al., 2004; Fehr & Fischbacher, 2004). Rewards and punishments do increase cooperation in social dilemmas and situations that pit short-term self-interest against longer-term collective interest (Balliet, Mukder, & Van Lange, 2011).

The impact of injunctive norms (e.g., moral norms) has been much less clear in research on prosocial behavior, as the chapter's final section on moral principles will show. However, injunctive norms can vary in salience as well. In one study, researchers (Cialdini et al., 1991) made antilittering injunctive norms salient by sweeping all the litter into a tidy pile. In another study, they primed the cognitive accessibility of the antilittering norm and other norms at varying proximity to it (e.g., recycling norms are close, but voting norms are far; Figure 9.4 ); they did this by giving pedestrians a handbill promoting the relevant message. Littering the handbill increased linearly as the norm departed further from the littering issue.

Figure 9.4 Littering as a Function of the Accessibility of Injunctive Antilittering Norms

Further evidence for the importance of cognitive accessibility comes from a study that also primed norms at varying proximity to helping norms (Harvey & Enzle, 1981). Participants read one of three stories: One story primed norms against harming another person (close to helping norms); another primed norms against property damage (farther from helping norms); and a third primed norms against speeding (farthest from helping norms). Participants then had an opportunity to help a professor by crossing out vowels on several pages of text. They were more likely to help, the closer the primed norm was to helping norms. Overall, the role of norms depends on their salience in the environment and hence their cognitive accessibility in people's minds.

The salience of norms varies not just with setting but also with age. As children develop, they become aware of prosocial norms. Some evidence fits the idea that children become increasingly aware of social expectations, which causes them to behave in a progressively more prosocial manner (Cialdini et al., 1981; Piliavin & Charng, 1990).

Individual Differences in Group Prosocial Orientation

People's perceptions of social norms vary, depending on their own degree of prosocial orientation to others. Social value orientation, measured as in Table 9.2 , differentiates among people with cooperative (prosocial), competitive, and individualistic orientations to others. Prosocial value orientation emphasizes both equalizing outcomes and enhancing joint outcomes (Van Lange, 1999). Competitive and individualist orientations are proself, the former by maximizing the difference between self and other and the latter by maximizing own outcome, regardless of the other person's outcome.

In general, people expect consensus from others, that is, agreement on their own social values. The false consensus effect ( Chapter 5 ) predicts this projecting of self onto others. Competitors expect a more competitive norm than cooperators do (Kelley & Stahelski, 1970). However, people's expectations of consensus with their own individual orientations decrease as the population goes from near (friends, classmates) to far (schoolmates, other peers) (Iedema & Poppe, 1999).

People with different prosocial orientations create different group dynamics for themselves, creating a self-fulfilling mirror-image world for themselves. A meta-analysis examined particularly difficult conflicts, in which negotiators highly resisted yielding to the other side. Prosocial negotiators were generally less contentious, engaged in more problem solving, and (ironically) ended up with higher joint outcomes than more egoistic negotiators (De Dreu, Weingart, & Kwon, 2000). For example, in one study, prosocial dyads negotiated by restructuring the problem, supporting each other, and managing the negotiation process so both perceived it to be fair. Egoistic (individualistic and competitive) dyads similarly exchanged information and concessions and avoided arguing. Mixed dyads argued about distribution of outcomes (i.e., who gets what) (Olekalns & Smith, 1999). Prosocial people see the other person as more fair and considerate. They demand less and make concessions more. In contrast, noncooperative people elicit less cooperation from partners, as well as tendencies to withdraw to lower levels of interdependence (Van Lange & Visser, 1999). The prosocial orientation is not anti-self but enhances the whole social experience for self and others.

Social value orientation moderates (i.e., influences the impact of) several situational variables already discussed. For example, a prosocial orientation facilitates reciprocity (Van Lange, 1999; Van Lange & Semin-Goossens, 1998). In contrast, an egoistic (proself) orientation facilitates perceived personal costs of helping (Cameron, Brown, & Chapman, 1998). And the impact of cooperative versus competitive value orientations is exaggerated by a focus on loss (De Dreu & McCusker, 1997). Thus, personality interacts with situation to predict prosocial behavior.

Another individual difference variable, less studied, is prosocial self-schema. (Recall from Chapter 5 that a self-schema is the most central aspect of one's self-concept.) Individual differences in prosocial self-schemas also predict prosocial behavior, but they again interact with the situation. That is, situational factors, including the salience of normative ideals (self-awareness) and salience of the opportunity to help, increase the impact of prosocial self-schemas (Froming, Nasby, & McManus, 1998). Over time, prosocial behaviors in childhood predict later popularity with peers and even academic achievement (Caprara, Barbaranelli, Pastorelli, Bandura, & Zimbardo, 2000). This fits the view that both social and academic goals at school determine learning (Covington, 2000).

People develop a prosocial orientation from different patterns of social interaction as they are growing up (Van Lange et al., 1997). Prosocial behavior results from secure attachment (introduced in Chapter 8 as one indicator of trusting), because it requires openness to others. Also, in keeping with the idea of building on past interaction patterns, people are more likely to be prosocial if they have had siblings (especially sisters!), presumably because they have had to cooperate with them. But, as just noted, people become more prosocial as they age, in any case. This may stem from the experience that a positive social value orientation fosters more positive social processes and outcomes than a pro-self orientation, an experience supported by the self-fulfilling nature of a cooperative, prosocial orientation, noted earlier.

Summary of Collectivism Hypotheses

Collective interest is distinct from both self-interest and individual other-interest. In social dilemmas, empathy for individual others may actually undermine collective interest. Prosocial behavior aimed at benefiting the group can result from group identity. A sense of we-ness or one-group identity encourages helping ingroup members. Group-oriented prosocial behavior also can result from salient norms. Norms can be descriptive (what people do) or injunctive (what people ought to do). Finally, people differ in their degree of group prosocial orientation. Prosocial, cooperative behavior (within limits) elicits like behavior from others, more constructive negotiating processes, and higher joint outcomes. Pro-self (competitive or individualistic) behavior elicits less constructive processes and outcomes.

PRINCIPLISM: MORAL UNDERSTANDING

Theoretically, people might engage in prosocial behavior for neither personal, interpersonal, nor group interest but simply because it is the right thing to do. Principlism is the motivation to uphold a moral stand (Batson, 1994, 1998). Moral principles might include impartial justice, the greatest good for the greatest number, or love thy neighbor. Principlism goes beyond self, other, and group interest.

Because where you stand depends on where you sit, people might make the most impartial judgments when they do not know their own relative position. That is, people's understanding of universal moral values takes them “behind the veil of ignorance” (Rawls, 1971) to behavior deemed right, without knowing whether one might be victim, perpetrator, or helper and whether an ingroup or outgroup member. Judgments of fairness without knowing one's own position reveal principles uncontaminated by self-interest or other biases. The Rawls veil of ignorance proposed that people should make moral choices without knowing how they, specific others, and their groups would fare.

Indeed, research participants making ability-based task assignments were more willing to subscribe to the principle of meritocracy when they could not be seduced by specific interests. In effect, they were more likely to endorse inequality that trades one group of people against another when they did not know their own position (Brickman, 1977). This supports the idea of a principled (abstract, impartial, though not necessarily consensual) form of moral reasoning. Extending this type of logic, social psychologists have used two approaches to moral reasoning, both of which involve an abstract understanding and an effort to control outcomes and sometimes a group-belonging motive, all as possible prosocial motivations.

Moral Reasoning

Moral reasoning reflects how people think about ethical dilemmas. In this view, opportunities for prosocial behavior are cognitive puzzles to be solved, and people can develop through several stages of sophistication in their type of moral reasoning. In the first well-known account (Colby & Kohlberg, 1987; Kohlberg, 1963), people can operate at three primary levels of moral reasoning ( Table 9.13 ). The preconventional level operates in an egoistic fashion, with concern only for short-term self-interest. For example, a person reasoning in this fashion would worry mainly about getting caught in a transgression; this fits the motive to control one's outcomes. The conventional level operates in a collectivist, belonging fashion, with concern for group standards. A person reasoning conventionally would worry about upholding the group's rules. The postconventional level operates according to abstract principles thought to be universal, consistent with the understanding motive. A person reasoning in a postconventional fashion would worry about equality, dignity, and justice, also consistent with an understanding motive.

TABLE 9.13 Levels of Moral Reasoning

Kohlberg (1963)	Eisenberg (1982)
Level	Concerns	Level	Concerns
Preconventional	Immediate consequences for self; own needs; reward and punishment	(1) Hedonistic	Self
		(2) Needs-oriented	Others' needs
Conventional	Obedience to social norms, rules, and laws; social order; approval and disapproval	(3) Stereotyped	Images of good and bad people; approval
		(4a) Empathic	Other's humanity
		(4b) Transitional	Larger society
Postconventional	Transcendent universal principles; generality beyond group	(5) Internalized	Values, norms, responsibility

Source: From descriptions by Eisenberg, 1982; Kohlberg, 1963.

The Kohlberg approach drew a lot of attention, both admiring and critical, primarily in developmental and educational psychology, so most of it lies beyond our scope. Some of the criticisms included disputes over the alleged universality of the scheme, its requirement of linear stages, its possible gender differences, and its exclusion of caring and altruism as moral bases of action (Batson, 1998; Eisenberg, 1991). In response, other partially overlapping schemes developed (e.g., Eisenberg, 1982; see Table 9.13 ).

Another tactic was to downplay the stage idea and study people's utilization of various moral judgment processes (Thoma, Rest, & Davison, 1991). This recognizes the flexibility of people's moral understanding. For example, consider the classic moral dilemma, whether Heinz should steal a costly drug to save his sick wife (part of the original Kohlberg dilemma and also in the Defining Issues Test, Figure 9.5 ). People can come to the same conclusion by different levels' reasoning processes (don't steal the drug because of possibly getting caught versus because it's wrong). Or they could come to different conclusions, even at the same level. For example, someone at a conventional Kohlbergian level might define the Heinz dilemma in terms of family duty (steal the drug) or duty to society (don't steal). A four-step model of moral behavior (Rest, 1983) breaks moral reasoning down to include the following:

· Interpreting the situation as a moral problem (utilizing empathy and perspective-taking)

· Applying relevant standards to action plans (judging morally right and wrong action)

· Evaluating how actions serve moral values (decision making and trade-offs)

· Executing and implementing action (using self-regulatory processes)

Figure 9.5 The Heinz Moral Dilemma from the Defining Issues Test

Level of moral reasoning predicts prosocial behavior, according to one meta-analysis of varied studies (Underwood & Moore, 1982). “In many ways, the data for moral reasoning are the most compelling we have seen, not for the magnitude of the relationship but for the generality of that relationship” across participant populations, helping measures, and moral reasoning measures (p. 158). However, levels of moral reasoning can be confounded with age, education, and social class, so these other variables may account for the impressive effect. Nevertheless, as noted earlier, but here as an example of principlism, gentiles who rescued Jews had higher ethical values than those who did not (Oliner & Oliner, 1988).

What predicts type of moral reasoning? As people's education increases, they use increasingly sophisticated forms of moral reasoning (Rest & Thoma, 1985). Moral reasoning also increases with age (Darley & Shultz, 1990; Eisenberg, 1991; Piliavin & Charng, 1990).

Moral reasoning cannot be reduced to equivalence with religious involvement because it depends on the type of religious motivation one has. For example, one could be motivated by religious beliefs about rewards and punishments, by social approval and disapproval from one's co-believers, or by impartial principles from one's religion (e.g., Batson, 1990). Similarly, moral reasoning cannot be reduced to cultural ideology (Narvaez, Getz, Rest, & Thoma, 1999). Finally, moral reasoning is distinct from empathy and may sometimes conflict with it (Batson et al, 1995).

An argument for the adaptive universality of moral reasoning is that humans have always had to solve the problem of maximizing gains from living in groups but also resolving conflicts of interest (Krebs, 2008). Moral reasoning could be strategic in managing cooperation, deference, and altruism, generating impartial principles for group living.

Reasoning or Rationale?

Whatever people's level of moral reasoning, people may deceive themselves about their own ethics. A series of studies demonstrates one version of moral hypocrisy (Batson, Thompson, & Chen, 2002; Batson, Thompson, Seuferling, Whitney, & Strongman, 1999). In the experiments, people were fully capable of convincing themselves that they had used a fair decision rule (coin flip) for allocating a more desirable task to themselves over another person. But in fact they could be observed falsifying the coin flip, despite clear instructions, prominent labels, and salient social standards. In the real world, likewise, people can claim one moral rationale but act on another (Carlsmith, Darley, & Robinson, 2002). When people are asked the appropriate moral basis for punishment by society, as in the legal justice system, they frequently cite utilitarian (controlling) motives, such as deterrence or incapacitation, that aim to prevent further harm doing. However, their theory differs from their practice. Empirically, people make choices that reflect revenge or retribution. They apparently focus on what the harm doer supposedly deserves, that is, on just desserts. People heed the dimensions of offenses that reflect just desserts (seriousness, morality) and not those that would reflect deterrence (frequency, ease of detection). Their decisions about penalties do not reflect their high-minded ideals about the operations of society, but rather punishment in proportion to harm, or as Kant put it, the criminal's apparent “internal wickedness.”

People's lack of moral self-insight suggests that moral reasoning may be a post hoc rationale, rather than a true underlying motive. Indeed, a growing set of evidence suggests that moral responses may often be relatively automatic and even more emotional or intuitive than rational and abstract. People's judgments about hypothetical dilemmas or their post hoc justifications for their moral decisions may not represent their actual, spontaneous moral decision making.

Intuition, not rationality, may best describe moral responses (Haidt, 2001, 2007): (a) dual-processes in social cognition suggest a role for automatic as well as deliberate response (see Chapter 4 ; Cushman & Greene, 2012); indeed, moral intuition is fast, automatic, and often affect-laden, preceding deliberate moral reasoning (Rand, Greene, & Nowak, 2012; Zaki & Mitchell, in press). (b) Moral actions link more closely to emotions than to reasoning (see also Tangney et al., 2007). (c) The role of motivation is well established elsewhere in social cognition; moral reasoning—much of it post hoc—often operates pragmatically to preserve one's reputation. And (d) moral communities preserve adaptive group solidarity, based on various norms: harm/care and fairness/reciprocity are most familiar to Americans, especially liberals, but ingroup/loyalty, authority/respect, and purity/sanctity also matter, for example, to conservative Americans and in more traditional societies. The intuitive, affective basis for moral reasoning questions not the presence but the origins of morality (Miller, 2008). In a provocative set of findings, some kinds of moral judgment clearly evoke emotional responses in the brain, rather than simply rational responses (Greene & Haidt, 2002; Greene, Sommerville, Nystrom, Darley, & Cohen, 2001). Moral reasoning may not be all it's cracked up to be.

Personal Norms and Values

People sometimes operate on principles dictated by personal norms: self-expectations for specific action in particular situations (Schwartz, 1977). People construct personal norms themselves, so they are internalized, although socially learned. When activated, personal norms are experienced as moral obligations. In this view, particular situations generate a feeling of personal moral obligation, which guides helping, unless people neutralize those feelings by denying their relevance. Personal norms do predict altruistic behavior, but only for people who are aware that their behavior has consequences for others and who do not deny responsibility for those consequences.

Personal values are standards that cut across specific situations. They are general beliefs about desirable ends; they guide evaluations and behavior; and they are ordered by importance (Schwartz, 1992). One psychological structure of human values considers biological needs, interpersonal coordination, and group welfare, in order to derive eight motivational domains of values (Schwartz & Bilsky, 1987). The values appear to have a similar structure across cultures ( Figure 9.6 ). The value of benevolence, closely linked to universalism, promotes concern for the welfare of others.

Figure 9.6 Psychological Structure of Universal Moral Values

People from communal (roughly, collectivist) societies prioritize different values than do people from contractual (i.e., exchange-oriented; roughly, individualist) societies. For example, people from relatively communal Taiwan value conformity more, whereas people from relatively contractual New Zealand value stimulation more (Schwartz, 1992). As noted earlier, Asian Indian adults prioritize responsibility to help colleagues and siblings based on moral values, but Americans prioritize helping by personal liking and disliking (Miller & Bersoff, 1998). The role of affect in moral reasoning surfaces again.

A more specific type of personal value relates directly to prosocial behavior. Belief in a just world (Lerner & Miller, 1978; Rubin & Peplau, 1973, 1975) holds that people get what they deserve, that the world is fundamentally a fair place. If a person experiences bad outcomes, the person must have done something bad to merit it. Believers in a just world will blame and derogate a victim, allowing them to maintain their belief that the world is fair. Earlier (in this chapter and in Chapter 3 ), we described differing attributions for poverty, finding that people who attribute poverty to the dispositions of poor people (i.e., blame them) show more anger, but those who attribute poverty to the situation of the poor people show more pity and are more likely to help them (Zucker & Weiner, 1993).

Similarly, believers in a just world will help people with an isolated or a temporary problem—presumably because helping can restore justice to their world—but they will not try to help people with pervasive or long-term problems—presumably because helping one person at one time will still leave a lot of inexplicable suffering (Miller, 1977). Believers in a just world also help others in order to make themselves more deserving of good outcomes (luck on an upcoming exam, for example; Zuckerman, 1975), clearly an egoistic motive in this case. More generally, they are protecting their own understanding that the world is a just place and that they can control their own outcomes in it. Some cultural differences in just world beliefs suggest that they can help people justify the status quo in an unjust society, for instance, during apartheid in South Africa (Furnham, 1985).

Most people hold personal values that matter deeply. Even thinking about violating such sacred values (e.g., putting a monetary value on human life) makes people angry. Racial egalitarians refuse to consider certain kinds of racial statistics, and Christian fundamentalists refuse to consider everyday explanations for biblical events (Tetlock et al., 2000). Whatever they are, the violation of sacred values makes people angry and motivated to reaffirm their own beliefs.

Many people's personal values include religious beliefs, although social psychologists rarely study the role of religion in people's lives. Religious beliefs of some kinds (internal, intrinsic, secure) benefit people more than other kinds (imposed, unexamined, tenuous). Religion can improve the outlook of groups that are marginalized and individuals who are under stress (Pargament, 2002). But religion can also be a source of stress, confusion, and sadness (Exline, 2002). Although religious involvement links to mental and physical health, various correlated variables might account for it: social support, health practices, self-esteem, and self-efficacy, as well as beliefs such as a sense of coherence (George, Ellison, & Larson, 2002). So far, the evidence does not clearly indicate whether any of these variables accounts for the sometimes salutary effects of religious involvement. Ironically, none of the relevant research focuses on prosocial behavior as a beneficial outcome of religious involvement.

Summary of Principlism Hypotheses

When people are motivated to help because of abstract moral principles that they perceive to be universal, this motivation can operate independently of egoism, altruism, and collectivism. Prosocial behavior can result from varying levels of moral reasoning, which range from controlling outcomes and promoting self-interest to more collective concerns to understanding abstract moral principles, although emotion and intuition may matter more. Specific personal norms can operate in a given situation, and more general moral values transcend situations.

CHAPTER SUMMARY

Return to the example of the washerwoman who donated her life savings for college scholarships. If prosocial behavior includes behavior intended to benefit others, certainly her act qualifies as prosocial. Social psychologists have studied prosocial behavior in experimental games by examining personality differences in strategy choices that reveal social value orientations (cooperative, competitive, individualist). People differ in empathy and perspective-taking as well.

Social psychologists focus more often on situational causes of prosocial behavior. One approach uses the experimental games in which people can choose to cooperate or compete, and their combined behavior determines their respective payoffs. Researchers then vary situational factors. Another situational approach designs socially realistic, highly impactful, and personally involving situations in which participants witness an emergency and researchers measure helping. Alternatively, participants may encounter a person with an ongoing need. Finally, social psychologists also study real-world helping situations, such as volunteering and blood donation.

Washerwoman McCarty's gift could have stemmed from many motivational sources. Core social motives underlying prosocial behavior can include self-enhancement, trusting, belonging, or understanding. In helping research, a continuing controversy debates these motives respectively as egoism, altruism, collectivism, and principlism.

The egoism hypothesis starts from an evolutionary puzzle: Why would people help others when it apparently makes no contribution to their own genetic survival? From an evolutionary perspective, the primary explanation rests on kin selection, and people do indeed help genetic kin more readily than others. Necessary adaptive ingredients would include recognizing kin and detecting distress, both of which are present in people from an early age. However, the evolutionary explanation is silent on the psychological mediators, which may include emotional closeness, mere exposure, or other mechanisms compatible with a variety of explanations.

Social learning potentially supports the egoism hypothesis, as people aim to obtain social rewards and avoid social punishments. One of the most robust forms of social learning is reciprocity, in which people help others who help them in return. Equity holds that mutual help is proportional (each person's cost-to-benefit ratio is the same). Under exchange, helping is returned equally. In experimental games, a strategy of tit-for-tat works well to reward the other's cooperation and punish competing defections, which results in higher joint payoffs. It also deters cheaters and other free-riders, who try to exploit others. According to enlightened self-interest, people often learn that selfish behavior does not pay off in the long run. People do receive a variety of social rewards for helping (approval, praise) and try to avoid social punishments for not helping (disapproval, censure). People learn vicariously, by observation, and they imitate models.

The third major leg of the egoism hypothesis rests on mood protection. Maintaining good moods nicely predicts prosocial behavior: Cheerful people do indeed reliably help others more. The good mood is more effective when based on one's own happiness rather than someone else's good fortune. Especially important is the mood's connection to improving one's general social outlook. Good moods accentuate the positive in perceptions of self, other people, the helping situation, and society. In contrast stands the concomitance hypothesis, which maintains that the link between good moods and helping is a mere by-product but not psychologically meaningful.

In contrast to good moods, bad moods have more variable effects. Guilt does consistently encourage helping, perhaps through attention to one's own position relative to the other person. Sadness tends to undermine helping. The negative-state-relief hypothesis holds that people help in order to relieve their personal distress at another's misfortune or because they have learned that helping relieves sadness in general.

None of the egoism hypotheses particularly explain Ms. McCarty's scholarship donation. The altruism hypothesis proposes that the primary motivation of helping can indeed be the other person's welfare, and her case could fit this analysis. People must first attribute responsibility for the victim's predicament because people do not pity or help another person unless the causes of the troubles are external, uncontrollable, or both. Next, people must experience felt responsibility, interpreting the situation as an emergency and attributing responsibility to themselves. The presence of passive others can undermine both processes. Moreover, either victim or others can be held responsible for the problem or its cause, leading to different models of the helping situation.

Empathy would provide the clearest evidence for altruism. Empathy, feeling another person's distress, does reliably increase prosocial behavior. Empathy is distinct from sympathy, acknowledging the other person's situation, and personal distress, being upset by it regardless of how the other is actually feeling. Only the sympathetic aspect of empathy facilitates helping. The negative-state-relief model would predict helping from personal distress, but the effects are controversial. Empathy-specific rewards (including empathic-joy) and punishments constitute another egoistic explanation for empathy effects on helping. A final egoistic argument asserts that empathy causes people to merge self with other into oneness, so that helping the other becomes helping the self. Empathy theorists disagree and muster impressive evidence for the role of altruism. Regardless, people individually high on empathy (but not personal distress) do help others more. They learn empathy from secure trusting relationships and are more likely to engage in a series of real-world prosocial behaviors. Ms. McCarty's donations clearly could result from empathic motives.

Her donations to the local university could also result from collectivism, the motive to improve group welfare. People help similar others more, and they help more when they have a sense of we-ness or one-group identity. Social dilemmas pit individual interests against group interest, but sometimes the group interest comes out on top.

Group norms may be descriptive (what people do) or injunctive (what people ought to do). Norms prominently include reciprocity and social responsibility, discussed earlier. But other prosocial norms may be more specific (e.g., not to litter), and their effect depends on how salient they are. People also differ in their orientation to others, and prosocial (cooperative) people inhabit different worlds than proself (competitive or individualist) people. Prosocial orientation involves trust. It elicits trust and cooperation, more constructive group processes, and higher outcomes all around.

The final proposed motivation for prosocial behavior is principlism, upholding moral standards. Without knowing Ms. McCarty's personal beliefs, one cannot know; however, the newspaper interviews make principlism seem likely. Abstract levels of moral reasoning predict helping for different reasons than other forms of moral reasoning, which rely on egoistic or collectivist motives. People's personal norms and values also predict helping. People have a varied set of potential motives for helping others. As we will see, their motives for hurting others are equally varied.

SUGGESTIONS FOR FURTHER READING

1. Batson, C. D. (1998). Altruism and prosocial behavior. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), Handbook of social psychology (4th ed., Vol. 2, pp. 283–316). New York: McGraw-Hill.

2. De Waal, F. B. M. (2008). Putting the altruism back into altruism: The evolution of empathy. Annual Review of Psychology, 59, 279–300.

3. Eisenberg, N. (2000). Emotion, regulation, and moral development. Annual Review of Psychology, 51, 665–697.

4. Haidt, J., & Kesebir, S. (2010). Morality. In S. T. Fiske, D. T. Gilbert, & G. Lindzey (Eds.), Handbook of social psychology (5th ed.). New York: Wiley.

5. Penner, L. A., Dovidio, J. F., Piliavin, J. A., & Schroeder, D. A. (2005). Prosocial behavior: Multi-level perspectives. Annual Review of Psychology, 56, 365–392.

6. Pyszczynski, T., Greenberg, J., Koole, S., & Solomon, S. (2010). Experimental existential psychology: Coping with the facts of life. In S. T. Fiske, D. T. Gilbert, & G. Lindzey (Eds.), Handbook of social psychology (5th ed.). New York: Wiley.

7. Tangney, J. P., Stuewig, J., & Mashek, D. J. (2007). Moral emotions and moral behavior. Annual Review of Psychology, 58, 345–372.

Notes