3. Power and democracy (en inglés) (2015)Apunte Español
Vista previa del texto
Power and democracy
Concepts and historical perspectives on power
The challenges for a unitary social psychology of power are clear. In order to understand the effects of power on those
who hold it and those who are subject to it, we must have some agreement about what it is. However, every discipline has
failed to agree upon a unified definition of power.
Social psychology has recently seen a vigorous expansion of research on how “power” affects a number of different processes and outcomes, including social attention and perception, social cognition, affect and emotion, action orientation, propensity for risk and influence of dispositions versus situations. Indeed, power is a single label used to refer to a range of intrapsychic and interpersonal phenomena. Thus, to begin to integrate across these various lines of inquiry, it is important to begin considering in greater depth what power is and how we speak of it. First, a social psychology of power must clearly be concerned with how individuals and individuals within social units experience power. At the same time, it is important to recognize that social power is an explicitly relational construct. One person cannot have power without others to be subordinate.
What is power and where does it come from? We begin with an overview of definitions of social power from diverse sources. The definitions are grouped by conceptual similarity rather than chronology; they proceed from quantitative capacity views, to consent-based views, to identitybased views.
Ø Quantitative capacity perspectives Many conceptions of social power treat it as a personal characteristic, emphasizing the individual’s possession of a certain amount of power. To use the broadest possible definition, power is the ability to get what one wants, or “the production of intended effects” (Russell, 1938); that is, one has power if one is able to obtain desired outcomes and to make things happen the way one wants. Weber (1946) extended and refined the definition, making it more explicitly social: in Weber’s view, power is “the probability that one actor within a social relationship will be in a position to carry out his own will despite resistance, regardless of the basis on which this probability rests”.
What all of these approaches share is a view of power as a quantitative capacity that inheres in an individual. It is as if we could measure a certain physical volume (say, 250 cc) of power residing within the person. If power were quantitative, then we might argue that when someone with 250 ccs of power meets someone with 200 cc, the first person is powerful. But when that person meets someone with 300 cc of power, then the new person is powerful. To identify power, we need only list the resources and attributes of various individuals (or institutions), and the relative ranking of the lists determines power (Hindess, 1996).
Social psychological theories of social power often reflect the quantitative-capacity perspective: they posit that the relative dependence of two or more parties constitutes the primary driver of power differences. Specifically, the amount of Party A’s dependence on Party B is compared with the amount of Party B’s dependence on Party A, and the individual with the lower net dependence is considered the high-power party. Thibaut and Kelley (1959), who developed the Party A–Party B comparison just presented, argued that there are two forms of power, fate control (control over the outcomes of another person) and behavior control (control over another person’s actions). French and Raven (1959) outlined a list of the resources that may confer power to an actor; that is, if one has the right characteristics, position, knowledge, or ability to inflict pleasure or pain, then others may (willingly or not) become dependent on that person given the resource holder’s capacity to affect them. Again, this framework suggests that the relative amount of these resources defines the individual’s power. Both fate control and the French and Raven (1959) power bases have led directly to today’s modal power definition: that social power reflects “outcome control”. All these views conceive of social power as separate from actual influence: it is potential, not necessarily realized; it is consequential, not merely controlling; it requires intent.
Another influential dependence-based approach to power is the resource dependence theory (Pfeffer & Salancik, 1978). The theory argues that Group or Person A has power when A possesses a resource on which B is dependent, and B cannot substitute a different resource for the one possessed by A. Like several other theories, resource dependence theory (and its interpersonal equivalent, strategic contingency theory) emphasizes the dependence of the powerless party.
A challenge for all of the dependence-based theories is that they tend to foster a pure quantitative–capacity perspective on power effects. For example, Kipnis (1972) found more “corrupt” patterns of interpersonal behavior for power holders who were given reward and punishment powers relative to those with only legitimate power.
This suggests that there may be important qualitative differences from one manifestation of power to the next, and these may affect how others react to that power. However, dependence theories, which posit that each of these Power and democracy bases— reward, punishment, and legitimacy—may be a separate and independent source of power, might suggest that the differences are driven by amount of power. The flaws in such a conception are obvious: power is indeed based on one’s balance of resources and characteristics, but it also depends on factors such as others’ resources and characteristics, one’s ability to exercise power effectively, situational constraints, and the scope of power possessed; that is, power is a relative, and not absolute, capacity. It is better regarded as a property of relationships than a property of individuals given that the same individuals may enjoy very high power in some contexts and rather low power in others.
However, psychological studies of power have tended to obscure the relationality inherent in the dependencebased approaches. For example, Fiske’s (1993) Power As Control model posits that organizational power holders do not depend on subordinates for their own outcomes, even though power relations in many organizations are actually reciprocal, in that power holders’ compensation is directly tied to subordinates’ performance. This approach is very well suited to isolating effects of power from other relational features. However, in contrast to this common social psychological approach, some basic definitions of power from other disciplines hold that social power is only interesting in situations of mutual dependence—that is, when one party may have access to resources that give rise to authority or dominance, but still relies on the other party for satisfaction of its own outcomes.
Ø Consent One way of resolving many problems with a quantitative capacity view of power has been to conceive of power as a function of the consent of subordinates. In short, the consent view argues that powerful individuals rule by legitimate right. Such a right may have been granted, originally, by God or other divinity; by successful exercise of force; by tradition; or by the blessing of some institution. However, to endure, power must not only be endowed to the holder but also ratified by the support (the consent) of those who will be subordinate to it. It is the subjects’ consent that provides authority for the power holder to act—that is, consent ensures the capacity to act, and thus constitutes the power itself.
Power as right is strongly associated with Locke, who believed that sovereigns had power as a function of the consent of their subjects. Once consent was withdrawn, then power was nullified. Later theorists such as Hamilton and Biggart (1985) have argued that the power holder and subordinate are bound in a dialectical relationship, in which each is obligated to show obedience: the subordinate to the power holder’s commands, and the power holder to the role demands and expectations associated with power. If the power holder should refuse to fulfill those role expectations (e.g., if a boss refused to make decisions, did not exert discipline, or avoided stating his or her opinion), then subordinates would likely withhold consent and refuse to carry out orders. As a consequence, the power structure would implode.
Both the Lockean concept of consent and the Hamilton and Biggart (1985) model posit a largely implicit process of establishing legitimacy based on moral right, duty, or responsibility. However, consent-based views of power may also be framed in terms more akin to social exchange. For example, Hollander (1958) proposed his concept of idiosyncrasy credits to account for why individuals are able to gain the freedom to violate group norms as they gain power. Hollander proposed that competence in attaining group goals and conformity to group norms demonstrate the member’s utility and fealty; such a member accumulates virtual “credits” that can be “redeemed” for the privilege of contravening the norms. This covaries with the accumulation of power and influence opportunities in the group. Some recent work has explored the idea that observers tend to be biased toward seeing power holders as unconstrained and dispositionally motivated (Overbeck et al., 2006). Individuals placed in positions of power may tend to see themselves as highly constrained, whether by others’ expectations or by their own shortcomings (Orizio, 2002). This disconnect between the assumptions in our manipulations of power and the subjective experience of holding power may lead to limited and fragmentary conclusions.
On the other hand, social psychologists do talk extensively about legitimacy. The concept of legitimacy is clearly related to consent. It is distinct insofar as legitimacy reflects the end state of acceptance of a system and its embedded hierarchies as morally right (Zelditch, 2001), and consent is the operational step that provides a means to that end. The intersection of legitimacy and power has played an important role in social psychological literatures such as those on destructive obedience (Milgram, 1969), the psychology of imprisonment (Haney, Banks, & Zimbardo, 1973) and system justification (Jost & Banaji, 1994).
Legitimacy is also an important part of the Social Identity Theory (SIT) tradition in social psychology. According to SIT, individuals identify with groups that can offer them a positive social identity. Groups vary in status and thereby in social power, and members of groups can adopt various strategies to deal with their own group’s disadvantages—for example, by devaluing the dimension on which they are disadvantaged. The emphasis on group-based social structure is relevant to social power, since groups inevitably develop power and status Power and democracy hierarchies, and since the intergroup context makes group organization, mobilization, and direction necessary. Thus, it is not surprising that SIT, and its companion Self-Categorization Theory (SCT; Turner, 1985), which stresses that individuals use signals of comparative fit with their groups to determine how strongly to align with the group, have given rise to two innovative theories of social power— theories that emphasize the role of identity rather than dependence or consent.
Ø Identity-based theories Turner’s Three Process Theory extended his Self-Categorization Theory to the domain of social power, arguing that shared group identity creates power through the rise of social influence. Individuals come together to form a group and the process of coordinating and unifying their self-interests creates influence that allows the group to act as an entitative unit. As members of the group, then, individuals gain the ability to act on the world and change their environment and other people: they gain social power; that is, rather than social power providing someone the means to influence others, it is the group’s possession of influence that confers social power upon an individual.
Simon and Oakes (2006) extended SIT and SCT even further. Instead of a coercive force characterized by a focus on controlling outputs, they proposed that social power can be seen as a constructive force characterized by the recruitment of human agency to channel inputs of energy and resources toward meeting the power holder’s goals.
They argued that power relations typically arise from identity, where conflicting identities between two parties lead to conflict and coercion, and shared superordinate identities lead to influence based on consensual understandings.
The SIT tradition, with its emphasis on relations among groups as a driver of hierarchies, has given rise to a final important perspective on power—an explicitly structural theory, Social Dominance Theory (SDT; Sidanius & Pratto, 1999). SDT holds that the higher one’s individual preference for social dominance, the more one tends to endorse inequalities in privilege and opportunity as a function of group-based differences in social power. SDT argues that group-based hierarchy is universal and inevitable. Individuals vary in the degree to which they endorse this form of social organization, but in general appear most comfortable when they know their place in the hierarchy. Even members of groups with little social power may be satisfied if they have the certainty and security of a clear (low) hierarchical position. SDT is perhaps the most macro level of all the social psychological approaches to power.
SIT and SDT are prominent and influential within social psychology; as such, it might be argued that psychology already has a record of examining “power that is embedded in and works through the social structure and norms of a community”. The study of power as a function of social structure and norms is most often ceded to sociology and political science. But it would not be accurate to argue that psychology has missed these subtle conceptions of power entirely. For example, these perspectives are echoed by Kelman’s (1958) notion of internalization—the subordinate’s completely embracing the superior’s wants as his or her own—and French and Raven’s (1959) notion of referent power—the subordinate’s desire to fulfill the superior’s wants to emulate or to please the superior.
Similarly, literature on conformity (Asch, 1951), persuasion (Petty & Cacioppo, 1984), influence (Cialdini, 1993), and even role models (Lockwood & Kunda, 1997) has examined areas that are relevant to these implicit perspectives.
Ø Distinguishing personal and social power To return to the first and simplest definition presented, if power most broadly refers to “the production of intended effects” (Russell, 1938), then one might conceive of a “power” that is not relational at all. A longtime tradition in the power literature distinguishes power over, or the control or domination of others, from power to, the ability to carry out action. Accordingly, we can distinguish between the interpersonal exercise of power (power over, or social power), and one’s own agentic capacity (power to, or personal power).
The concept of personal power has its roots psychologically in White’s conception of competence. White (1959) posited competence as the individual’s mastery of inanimate objects and of the self—in short, power, but without any relational component.
Works on locus of control, illusions of control, control deprivation and motivation, outcome dependency, learned helplessness, and self- efficacy, all focus on the domain of personal power, but the construct is typically treated as control rather than power. As such, there has not been an explicit effort toward developing a unified psychology of personal power. Though this perspective on power remains largely unexplored in the growing social psychological literature on power, Van Dijke and Poppe (2006) argue that many effects that appear to involve social power may instead reflect individual strivings for personal power. In two studies, they have found evidence that participants were motivated to decrease dependence on others but not particularly motivated to increase their Power and democracy power over others. They argue that many findings that appear to suggest a motive for social power may actually reflect desire for personal power.
In some Eastern philosophical traditions, it is argued that self- control is a greater source of power than dominating others, because it allows true mastery and achievement.
The purpose and exercise of social power Having explored what power is, it next makes sense to ask what power is for: how is it used, and to what end? Such questions also have important implications for how a social psychology of power develops. In particular, this chapter continues to focus on social power. Answers tend to fall into two categories: dominance perspectives, which emphasize a more sinister use aimed at coercion and exploitation, and functionalist perspectives, which argue for a constructive use aimed at mutual benefit.
Ø Dominance perspectives In the popular imagination, the powerful are viewed with distrust. It is believed that “power corrupts” (Lord Acton, 1865), that power goes to one’s head, that the powerful are willing to hurt others to get what they want. In short, power is often viewed in sinister terms, as a force of domination and coercion whose aim is to exploit other people, and whose attainment is “desired as an end in itself” (Russell, 1938) rather than as a means of accomplishing any substantive, consensually desirable objective.
According to Lenski (1966), the dominance view is characterized by a belief that the nature of human beings is essentially negative and corrupt. When social groups arise, some individual is likely to seize power for the self, using coercion to ensure his or her ascendancy and to secure resources and privileges. This leads to conflict, with the subordinate members oppressed and disadvantaged by not only the power holder’s initial coercion but also the system he or she establishes (using inheritance, nepotism, or fraud) to ensure that inequality endures. This dominance view holds that power is negative irrespective of how it is exercised. Russell (1938) distinguished naked power, in which physical force prevails, from traditional power, based in legitimacy and propaganda. But both forms of power can embody the dominance view: physical force can be used to dominate and injure; propaganda can be used to manipulate people to support ideas and actions that are contrary to their own interests.
What is required by the dominance view is that the power holder acts for self- or group- enrichment, at the expense of others, with power for its own sake as the ultimate goal. Power is seen as a temptation that must ultimately corrupt even the best intentions of its holder, if it is sought for its own sake. Indeed, in many cases, power’s bad reputation is merited. Social psychological research has shown power to be associated with failure to recognize others’ points of view, with self- serving behavior, with hostile teasing and aggression, with an increased likelihood to sexually harass, and with domestic abuse.
In short, the dominance perspective is easy to accept. However, not all power holders act abusively or coercively; not all are motivated by the desire to control others; and not all seek power as an end in itself. Power can also be seen as a universal, necessary, and even inevitable force. Without power, no collection of people would be able to accomplish any end. This functionalist perspective, periodically quite well developed by students of power, has received less attention from social psychologists, but it deserves a closer look.
Ø Functionalist perspectives The functionalist view of social power argues that human beings live and congregate in groups; groups naturally require organization and coordination. Direction is needed to ensure that the group meets its goals and does not waste resources or opportunities; such needs give rise to the emergence of power. The functionalist view holds that groups invest power in one or a few individuals to ensure the success of the entire group.
The leading proponents of this view argued that hierarchical inequality arises not from a sinister seizing of privilege by self- serving powermongers, but rather from the needs of social groups and societies to order and govern themselves so they may attain outcomes valued by the entire group. Groups have a vested interest in putting the most capable people in positions where their skills can benefit the group.
It appears that increasing attention is being paid to this perspective. Simon and Oakes (2006) refocused on functional power in their identity model of power: “Power should be seen not only as a conflictual coercive force but also as a consensual productive and organizing force. . . . Power has to do with the power holder’s capacity to “recruit agency” by other free social actors. . . . Power holders exert their power by getting others to have the desires they want them to have and by manipulating social identities” Power and democracy Though recognizing that power can also, at times, be associated with domination, Simon and Oakes argue that the functional perspective deserves much greater weight than it typically receives from social psychologists. Along the same lines, Overbeck, Correll and Park (2005) argued that power is assigned in a reciprocal process open to influence by both the individual seeking power and the group needing a leader. In general, power is available to all of the (few) individuals who seek it. However, some groups may have relatively more power seekers, and competition for power may ensue. In that case, power is invested in the person who pairs individual desire for status with a stronger value for interpersonal cooperation, that is, those who can best satisfy the group’s needs.
It should also be noted that all of the perspectives reviewed to this point focus on culturally Western conceptions of power. East Asian concepts of power comport with the functionalist perspective, in that power is seen to carry responsibility rather than privilege. The Confucian, Taoist, and Buddhist traditions, for example, discuss control over self and impulses, a subordination of personal interests to the good of others, an orientation toward one’s group, and a duty to hierarchical relationships. These beliefs can be manifested in actions, such as some Japanese CEOs’ taking such extreme personal responsibility for their companies’ failures that they commit suicide.
What we know, what we don’t know Theorists have tried and failed repeatedly to create an all- encompassing definition of What Power Is. In this way, systematic study can create a more comprehensive body of work that helps us understand power as it has been conceived by social psychology and beyond.
Ø Findings from Social Psychology The earliest work that addressed power in social psychology focused on the behavioral effects of power on those subject to it. For example, Milgram’s (1969) studies of obedience to authority established that ordinary individuals look to those with power for direction, and expect the powerful to assume responsibility for actions. More recently, social psychological inquiry has focused on the effects of power on those who have it—that is, on how powerful people think, act, choose, and judge, either within themselves or in relation to (generally low-power) others. Two extremely influential bodies of work launched this recent flurry of research: • Fiske’s (1993) Power As Control (PAC) model of power and social perception has prompted energetic follow-up study of power’s effects on stereotyping, individuation, and degree of flexibility in attention.
• A decade later, Keltner et al. (2003) published a theory linking social power with behavioral approach, and lack of power with behavioral inhibition. They argued for approach and inhibition differences in the behaviors, cognitions, and affective experiences of powerful and powerless people. From the ensuing research, a few consistent patterns have emerged.
o o o o Power appears to foster a strong orientation to rewards and opportunities (Keltner et al., 2003), which can depress perceptions of potential risks or disadvantages (Anderson & Galinsky, 2006).
Powerful people show a predisposition for action rather than inaction or watchful deliberation.
Power is associated with strong goal orientation— perhaps by definition, given that power is associated with getting things done, and this directly connotes goal achievement. As such, powerful people are more sensitive and responsive to goals than are the powerless (Overbeck & Park, 2006), and they allocate scarce attentional resources to goal-relevant, but not to goal-irrelevant, objects (Guinote, 2007). This instrumental use of attention appears to extend to the domain of social perception, such that powerful perceivers attend to others when that serves their own goals and interests but are much less likely to attend when it does not.
Power appears to foster the expression of individual dispositions and preferences, whereas powerlessness may foster accommodation to the situation and to others’ expectations.
The coherence among these findings—the degree to which they seem to reflect a consistent difference in the psychology of the powerful and powerless—may be related to two sets of fundamental dimensions that appear to characterize human psychology more broadly. Voluminous evidence supports the notion of two fundamental dimensions in the domain of attitudes and perception: one is variously labeled agency, competence, dominance, or self- profitability, and the other, communality, sociability, warmth, or other. In the domain of behavior, the two fundamental dimensions are approach, on the one hand, and avoidance or inhibition, on the other. Power and status seem to map fairly cleanly onto both sets of dimensions, with high power associated with agency and approach, and low power associated with sociability and avoidance.
Ø Findings from other disciplines Power and democracy These emphasize macro processes, examining how power is distributed in societies and political entities, and are virtually mute on individual-level effects of power. Power scholars seem to agree that societies tend toward greater inequality; that is, as any society arises, there also arise identifiable economic, religious, and political elites. These elites’ interests may converge or compete, but as their members interact, an established order forms and constitutes a power structure that enjoys disproportionate privilege and share of resources.
Once an initial power structure has arisen, that structure must maintain its legitimacy in order to persist. Often, however, elites use oppression and coercion to impose their will, and this unjust exercise of power elicits resistance— sometimes, ultimately, revolution—from the oppressed. Two noteworthy sequelae may result.
1) First, in the case in which the oppressed succeed in gaining power, a cycle of oppression may be launched given that the underclass has not learned a functional model of exercising power and so it will tend to enact its new, powerful role by engaging in oppression of the new underclass.
2) Second, Russell (1938) argued that political systems tend to follow a predictable pattern of revolution; consolidation of power by the victors marked by overt, physical enforcement; subsequent movement toward legitimacy as a means of avoiding the need for constant physical enforcement and of ensuring stable, consensual order; and finally, as opposing interests gained strength, revolution again.
Much attention has been paid to the ways distributive systems are created and maintained; who gains resources and privileges; and what obligations are exacted from them in return. Those who control means of production, land and resources, and channels of persuasion are seen as most likely to enjoy privilege and power. Relatedly, social network theories have posited that power is determined by position within bargaining networks: being central in a network or occupying the position that connects two separate networks can help an individual or a group to maximize influence.
Functional theorists argued that powerful elites’ advantage in these systems of distribution oblige them to provide for the needs of the broader society. The repeated pairing of the trappings of position, and the behaviors associated with that position, imbues the elite with an internalized disposition suited to holding and exercising power, and this fosters the self-perpetuation of social stratification. Those at the bottom of a society may also internalize their role expectations and environmental affordances, though to much more disadvantageous effect. Powerlessness tends to foster responses that the powerful see as maladaptive and that, indeed, tend to work against their ability to further their own interests, such as competition with fellow proletariat and even slovenliness, self-directed harm, and neglect.
As for system maintenance, several theorists have noted that the more subtle the exercise of power, the more potentially effective it is. It has often been argued that social approval and censure work more effectively than laws and formal enforcement in effecting behavioral control.
Ø What we still need to learn • The literature is typically mute on the issue of whether power in a given study is coercive and dominancebased, or functional and constructive.
• Manipulations of social power have tended to reflect the quantitative-capacity view of power that compares one person with “more power” with another person with “less power.” These particulars have so far left a great range of questions unaddressed.
• The power literature offers little sense for how power changes over time. Intuition suggests that people who are new to power are uncomfortable with it and strive to remain undifferentiated from their previous peers. But those who have long held power may become accustomed to it, may grow to see their use of power as a natural right, and may respond quite differently in terms of their emotions, cognitions, actions.
• Power is self-perpetuating • All groups and societies ultimately create power orders • Perhaps more interesting, it might prompt greater attention to the processes and responses of the powerless: why do they consent? How do they ensure that their own needs are met? Do concepts such as noblesse oblige and Raven et al.’s (1998) legitimate dependence—the subordinate’s claim upon the powerful by virtue of his or her own weakness that demands care— account for their willingness to support a system that perpetuates their own disadvantage? • In most work on social power, we tend to focus on the power holder, whether as perceiver or target.
Largely ignored in recent years are the perspectives of the powerless.
Power and democracy Psychological underpinnings of democracy. A selective review of research on political tolerance, interpersonal trust and social capital 1.
Introduction This section reviews the literature on the psychological foundations of democracy. The expectations for citizens in a democracy differ from those for citizens living with other forms of government in many ways. The first is that citizens must tolerate their fellow citizens' efforts to participate in politics, even if the latter's views diverge sharply from the norm. The second is that citizens can and do participate in their own governance.
Democratic values and support for civil liberties In his mid-l9th-century analysis of American democracy and American character, Democracy in America, Alexis de Tocqueville devoted an entire chapter to what he called "principal causes which tend to maintain the democratic republic in the United States". He discussed three sets of causes, including situation and context, law, and manners and customs of the people. Of these, he considered the last to be the most important. Although de Tocqueville was imprecise in his analysis of the exact nature of the manners and customs so essential to American democracy, he quite explicitly argued that most analysts had considerably overemphasized the importance of situation and law.
In the 1950s and 1960s, modern political scientists elaborated on de Tocqueville's claim. These theorists concluded that, for democracy to exist, certain attitudes must be widespread. In short, there should be a consensus on the procedural norms by which substantive matters are negotiated, as well as on fundamental values such as liberty, equality, and individualism. More generally, empirical theorists of the era argued that the development of democratic political institutions depends critically upon the political culture, by which they meant that citizens must have a particular orientation toward political life.
Ø Support for civil liberties: evidence from early survey research One of the first such studies was conducted by Samuel Stouffer (1955), who administered two extensive national surveys of over 2400 cases each. He used national scientific probability sampling methods to study Americans' support for the civil liberties of socialists, atheists, Communists, and fellow travelers. Stouffer studied an additional sample of 1500 local community leaders. Stouffer discovered that most Americans –in fact an overwhelming majority– did not support the civil liberties of the left-wing groups he studied. One of the more important factors that Stouffer identified as explaining citizens' willingness to take away the rights of left-wing groups was the perception of threat. Americans felt threatened at a fairly high level by these unpopular groups. This threat and its consequent feelings of fear and anxiety appeared to drive most citizens' attitudes and beliefs. Stouffer also discovered that, among his sample groups, community leaders were far more supportive of these groups' rights than were ordinary citizens.
Stouffer's research called into question two fundamental assumptions of democratic theorists: 1) The first assumption that was undermined was that U.S. citizens had internalized properly the “rules of the game” (democracy). It was clear that they had not. Stouffer found that overwhelming majorities of citizens were willing to take away rights-particularly those of communists and atheists-to free speech, to participate fully in the political process, and so on. In other words, most citizens were willing to apply double standards, allowing mainstream groups one set of rights while restricting the rights of more extreme or unpopular leftwing political groups.
2) The second assumption called into question was that a national consensus on how to apply the democratic rules of the game was a prerequisite to having a democracy. Rather than question the extent to which the United States was a democracy, the more common conclusion from this research was that a democracy could obviously be sustained despite the significant shortcomings of its citizens.
But, if true, how exactly was democracy being sustained? Shortly after Stouffer's watershed analysis, two classic studies offered answers to this question. Prothro & Grigg (1960) studied local samples in Tallahassee, Florida, and Ann Arbor, Michigan. They asked respondents to express their agreement or disagreement with some general principles of democracy, including majority rule and minority rights, and with some specific applications of these principles, such as restricting voting to well-informed citizens, preventing a duly elected "Negro" from holding office, and so on. Prothro & Grigg found that there was a broad consensus endorsing the principles of democracy but that this consensus disappeared when applying it to specific controversial cases. They did find, however, a much greater level of consensus in the application of democratic principles among more educated/wealthy citizens.
Power and democracy Herbert McClosky (1964) conducted a study of national political party activists and political party supporters among the general public. Just as Prothro & Grigg had found, McClosky reported a much higher level of consensus about the democratic rules of the game (free speech, minority rights, and so on) among party activists than among ordinary citizens. The activists not only exhibited a higher level of support for the rules and principles of democracy, they also more consistently applied these principles to specific situations that challenged citizens' commitment to the principles.
These studies offered a possible resolution to the dilemma. There existed "carriers of the creed" who protected the democratic system from the majority of citizens who did not fully understand or support it. Despite the obvious fact that most citizens did not possess the correct underlying attitudes and understandings to sustain a fully functioning democratic system, the educated and the politically active did. Most citizens who failed to internalize the attitudes and habits that undergird democratic practices were generally inactive politically. Since most ordinary citizens did not act politically based upon their (negative) attitudes, at least not in concert, democracy was rather easily sustained by an activist class with the proper psychological attitudes, values, and predispositions.
In the 1970s, Nunn et al (1978) updated Stouffer's classic analysis, using essentially the same questionnaire Stouffer had used some twenty years earlier. They found that the American public had become far more accepting of the civil liberties of the unpopular groups studied by Stouffer. They attributed much of the change to improvements in education, demographic changes in the electorate, cohort changes… But most importantly, they concluded that ordinary citizens exhibited a substantial increase in support for the application of democratic principles. They argued that we need no longer rely on political leaders to be the carriers of the democratic creed.
Ø Support for civil liberties vs. political tolerance In the early 1980s, Sullivan et al (1982) reviewed the conceptual underpinnings of research on civil liberties and public opinion. They argued that, historically, the concept of tolerance evolved from efforts to moderate the harmful and often violent effects of religious conflict. The idea of religious tolerance was promoted as a mechanism to allow religions to "put up with," or tolerate, other religions that they disliked or even hated vehemently. So too did the concept of political tolerance evolve as a way to live with one's ideological and political enemies. Political opponents need not be eliminated physically or even politically. One need not like or support one's opponents and their ideas, but one ought at least to put up with, or tolerate, them. We were urged to learn to agree about how to disagree.
As a consequence of this conceptualization of political tolerance, Sullivan et al realized that measuring tolerance required a two-step measurement procedure.
1) Researchers must establish that an individual has political objections and/or negative feelings about a political group 2) Then, they need to measure the extent to which the individual supports or opposes the political rights of that group.
As a result, Sullivan et al created a methodology that attempted to provide individualized target groups for each respondent. In one variation of that method, they asked respondents to identify the group or groups in politics that they "like the least" and then asked respondents how far they would support the civil liberties (right to free speech, right to run for office, and so on) of their least-liked group. Using this methodology, Sullivan et al (1982) found that the American public was far less tolerant than had been indicated by the Nunn et al (1978) update of Stouffer. It appears that the primary change identified by Nunn and his colleagues was not in actual levels of political tolerance, but rather in the level of objection to these left-wing groups. Communists, socialists, and atheists were no longer as salient as they had been, and citizens did not object nearly so strenuously to them as they had in the 1950s.
However, that did not mean that other unpopular groups-often racist or right-wing-were tolerated.
The determinants of political intolerance Several studies continue to confirm that political "elites" are more supportive of civil liberties –and even more generally tolerant– than are ordinary citizens. Thus one obvious "cause" of differing levels of tolerance and intolerance is political expertise and participation. But other important factors have been identified, and they are discussed below.
Ø Political intolerance and perceptions of threat Many studies find that one of the factors with the strongest direct relationship to levels of political tolerance is threat perception. More recent work has confirmed both a dispositional role for threat perceptions, and a short-term effect due to the current information environment. Some people have a predisposition to be easily threatened and thus Power and democracy are very sensitive and responsive to potential threats in the political environment. They are less tolerant than individuals who are calmer and more easily reassured.
For many citizens, information that describes the nature and activities of potentially unpopular groups has a profound impact on their level of tolerance toward these groups and their ideas and activities. If the information environment portrays such groups as violating normative expectations with regard to orderly behavior and proper procedures, many citizens –even those not particularly predisposed toward intolerance– will refuse to tolerate the group and its activities. If the information environment portrays them as behaving properly and in an orderly fashion, then far more people-often a majority-will tolerate the group and its activities, despite the group's unpopular and/or extremist image (Marcus et al 1995).
Nelson et al (1997) tested for framing effects on tolerance judgments. One frame highlighted free-speech concerns and the other highlighted public order and safety. Using news reports of a KKK rally in Ohio, they experimentally manipulated the frame placed around the same set of facts (that the rally would have implications for both free speech and public order). They found that the frame placed around the information affected tolerance judgments.
In the public-order condition, attitudes toward public order significantly predicted tolerance for the KKK's activities, but attitudes toward civil liberties were insignificant. Conversely, in the free-speech condition, attitudes toward public order were insignificant, whereas attitudes toward civil liberties were highly predictive of tolerance of the KKK. In both studies, Nelson et al found that the frame affected the importance attached to specific values. The importance attached to these competing values significantly predicted tolerance.
• • Variation in the importance of public order showed a significant negative relationship to tolerance.
For those in the freedom-of-speech condition, the importance attached to freedom of speech shows a significant positive relationship with tolerance.
These results show how framing by the mass media affects how people think about tolerance and thus their final opinions on tolerance. Thus aggregate levels of intolerance are somewhat malleable, depending on how political elites and the media portray those with less popular ideas.
Ø Political intolerance and the internalization of democratic values Prothro & Grigg and McClosky not only noted the lack of consensus on fundamental democratic values, but they also emphasized that most citizens were inconsistent and hypocritical. Most citizens expressed a strong belief in democratic values but they were unwilling to apply these values to groups/ideologies they found objectionable.
Based on these analyses, it would be logical to conclude that the ideas and ideals of democracy operate at a level disconnected from actual practices. In other words, ideas did not guide behavior in this very important instance.
More recent research, however, does demonstrate that the internalization of democratic norms can make a considerable difference in determining how tolerant an individual will be when her or his tolerance is sorely tested.
The early studies reviewed above simply noted the aggregate disjunction between abstract beliefs and their concrete application. Later studies measured individuals' degree of support for the norms of minority rights, freedom of speech, and so on. They then related individuals' scores on these scales to their scale scores on items designed to measure applications of these norms to unpopular groups. The findings are clear: the more completely an individual has internalized and believes strongly in democratic norms, the more likely the individual is to tolerate groups and ideas that he or she finds to be obnoxious. It turns out, then, that threat perceptions and strength of belief in the more abstract norms of democracy are both very strong predictors of applied tolerance judgments.
Ø Political intolerance and personality In addition to threat perceptions and democratic norms, one additional potent set of variables that explains citizens' beliefs in democratic values and their levels of applied political tolerance is personality. Stouffer (1955) found that certain types of individuals –those who believed in stern child-rearing techniques and those who tended to be pessimistic– were much less tolerant of ideological nonconformity than were others, whose child-rearing views were more permissive and who were very optimistic.
McClosky & Brill (1983) found that several personality characteristics predicted applied tolerance scores quite well, including measures of misanthropy, anomie, self-esteem, flexibility, and so on. Sullivan et al (1982) evaluated the impact on political tolerance of a number of personality factors and found that the strongest relationship was for a factor they labeled "psychological insecurity”. Gibson (1987) also found that measures of dogmatism and trust predicted levels of political tolerance.
Power and democracy Finally, Marcus et al (1995) used Costa & McCrae's measures and found that neuroticism, extroversion, and openness to experience all predict levels of political tolerance fairly accurately. Those individuals higher on neuroticism and extroversion are less tolerant, while those high on openness to experience are more tolerant. Openness to experience appears to be the most powerful predictor. Their analysis went on to show that certain types of individuals are more responsive to contemporary information about how threatening unpopular groups are.
Individuals high in neuroticism are most responsive, apparently because of their high level of anxiety, which makes them more sensitive and responsive to perceptions of threat (Marcus et al 1995).
Cross-national studies of political tolerance American political sociologists and political psychologists have studied and identified the more precise connections among underlying social customs, psychological attitudes and predispositions, and support for democratic values and their application. It is, however, difficult if not impossible to make broader claims about psychological prerequisites for democratic attitudes and systems without conducting cross-national, comparative studies.
As a prolegomenon to these broad and difficult issues, scholars have conducted a number of surveys about civil liberties and political tolerance in other countries. Sullivan et al (1985) replicated their US study with samples in Israel and New Zealand. Using the least-liked group methodology described above, they found that New Zealanders were much more tolerant than Americans and Israelis, attributing that in part to the more isolated and peaceful political context. They tested a theory created to identify the underlying factors influencing tolerance that should be most sensitive to the political context, and those that should be the most context-invariant. In particular, the theory predicted that perceptions of threat should have the same impact on tolerance judgments, regardless of political context. The context may determine the overall level of perceived threat, but not the consequences for tolerance given any particular level of threat. Although levels of perceived threat might be much higher in Israel than in New Zealand, citizens in both countries who perceive a strong threat from their political enemies should be much less tolerant than those who do not. The findings of this study confirm this expectation.
Similarly, the impact of personality on political tolerance should be independent of context. Highly dogmatic, authoritarian personalities are expected to be intolerant –unsupportive of the civil liberties of groups they dislike– regardless of where they live. The results of this study also strongly confirm this expectation and indicate that there is a strong relationship between psychological insecurity and dogmatism on the one hand, and intolerance on the other hand.
Finally, the impact on civil-liberties judgments of internalizing democratic values and norms should vary by context, depending in large part on the cultural meaning and interpretation given to these values. Unlike threat perceptions and personality characteristics, whose consequences may be almost automatic, the role played by democracy is mediated by how that concept is understood in the political culture of the country. For example, in the United States, the emphasis in the civil liberties arena is placed on individual rights such as the individual's freedom of speech and the rights of the minority to participate fully in the political process. Other democratic nations –including Israel, for example– place a much greater emphasis on majority rule and equality. Both of these political cultures might emphasize democracy, but they may also emphasize quite different understandings of what that concept means. As a consequence, the impact of a strong internalization of democratic values on tolerance judgments should vary by political context. The findings of this study are consistent with this argument. The internalization of democratic values and norms has a very strong relationship with tolerance in the United States but not in Israel.
Ø Tolerance in an emerging democracy Gibson (1992) applied the least-liked group methodology to the study of tolerance in the former Soviet Union. He found that –like citizens in long-standing democracies– Russians were quite supportive of the general norms of democracy. However, when asked to apply these norms to disliked groups, they were intolerant –much more so than citizens surveyed in longer-standing democracies. Importantly, however, Gibson found that democratic beliefs and attitudes formed a single, more general attitude cluster that conceivably could form a foundation for the longerterm development of a democratic political culture.
Gibson et al (1992) report findings that tend to confirm many of the findings about the determinants of political tolerance that were reviewed earlier. By far the most significant predictor of intolerance among Russian citizens is, once again, perceived threat. As the theory described above leads one to expect, the influence of personality is also unchanged by this new political context. Gibson & Duch (1993) find that dogmatism increases intolerance in the former Soviet Union. Additionally, commitment to abstract democratic norms decreased intolerance, just as it had in New Zealand and the United States.
This commitment was put to a strong test by the attempted coup in 1991. Gibson (1997) administered a survey that asked respondents whether they had acted in some way to support or oppose the attempted takeover of the Soviet Power and democracy Union. He found that commitment to democratic institutions and processes, including minority rights and political tolerance, was one of the strongest influences on people's decisions to act against the coup. Gibson (1997) found that these psychological and interpersonal factors were more predictive than a cost-benefit analysis derived from rational-choice theory. Interestingly, he found that commitment to democratic norms and processes mobilized those people who acted against the coup.
Attitudes and political participation In addition to the importance of citizens tolerating the efforts of opponents to influence the political system, democracies also need relatively high levels of political participation by citizens. For this reason, the comparative study of democracy has been as concerned with mass participation as with tolerance. The study of the psychological sources of this aspect of democracy has occurred in two main waves. The first wave began and largely ended with Almond & Verba's (1963) Civic Culture, a landmark study of political attitudes in five countries. The second wave is the relatively recent renaissance of political-culture studies by Inglehart (1977) and Putnam (1993).
Ø The civic culture Almond & Verba (1963) compared national survey data from two stable democracies with high levels of democratic openness (the United States and Great Britain) to national surveys from three countries with shorter continuous democratic rule (West Germany, Italy, and Mexico) and found three main differences.
1) Citizens of the United States and the United Kingdom reported higher levels of interpersonal trust.
2) Citizens of the United States and the United Kingdom reported more pride in their political institutions.
3) Citizens of the U.S. and the United Kingdom showed more widespread feelings of political competence.
Almond & Verba (1963) named this collection of attitudes political culture, defined as a "psychological orientation toward social objects. When we speak of the political culture of a society, we refer to the political system as internalized in the cognitions, feelings, and evaluations of its population." Despite its impressive collection of survey data, The Civic Culture was widely criticized.
• The Civic Culture did not explain how variation in interpersonal trust, pride in national institutions, and political efficacy can lead to national differences in the actual practice of politics.
• The authors were not able to show, for example, how different levels of interpersonal trust affect the institutional arrangements of a country or change the way political processes operate.
• The notion that a set of attitudes supports democracy is plausible, but The Civic Culture did not succeed in explaining how these attitudes enter into the practice of democratic politics.
• Additionally, some scholars objected to the authors' assumption that the United States and the United Kingdom should be held up as models against which other democracies should be judged. These critics argued that the ethnocentrism of this approach biased Almond & Verba's analysis because it led them to assume that the patterns found in the United States and United Kingdom were essential to democracy in general, rather than specific to the particular forms of democracy found in those two countries.
Almond & Verba's research could not conclusively demonstrate the direction of causality between political culture and democratic governance because their data were drawn from a single cross-sectional sample of citizens within each country. They did not specify a single causal direction; rather, they described political culture as both a cause and a consequence of democratic governance and economic development. Scholars continue to debate whether the attitudes associated with the civic culture cause transitions to democratic regimes, ensure that democratic regimes persist over time, or actually are a result of the pre-existence of democratic regimes.
In response, Inglehart (1997) argues that it is a mistake to ignore culture completely or to believe that democracy follows directly on economic development. He notes that many countries are wealthy without being democratic, while many countries with democratic constitutions are not robust democracies in practice (notably the former Soviet Union and Weimar Germany). In his own research, he found that there is an association between high aggregate levels of life satisfaction and the stability of democracy. He argues that this measure shows that when there is diffuse support for a regime (one of the properties of the civic culture), democracy will persist. Inglehart counters that it is well known that politics is not a high priority for most people. Very few people say that politics directly affects their life satisfaction; family, work, friends, and leisure are far more important. Thus it is equally impossible to demonstrate that democratic institutions are the primary cause of observed variation in life satisfaction or its link to democratic stability.
Ø Culture shift Power and democracy Ronald Inglehart (1977) argues that economic development can lead to changes in peoples' psychological orientations and preferences. This in turn creates attitudes and expectations that favor democracy over other, less participatory forms of government and sustain democracy once it begins to develop in a country.
He argues that economic development causes cultural change that is conducive to democracy. The growth of a service sector is particularly critical because careers in this sector of the economy demand skills that are particularly conducive to democracy. According to Inglehart, there are certain key indicators that illustrate the impact of culture and attitudes on democracy. The citizens of countries that have the cultural resources to sustain democracy also exhibit higher levels of interpersonal trust and satisfaction with their lives in general.
INTERPERSONAL TRUST AND LIFE SATISFACTION: Inglehart shows that interpersonal trust and subjective well-being (whether people are satisfied with their lives in general) significantly affect the duration and level of democracy. Interpersonal trust is important because it lends credibility to the concept of a loyal opposition. When people do not trust their fellow citizens, elections and transitions of power appear to be far more dangerous. Citizens may fear that losing an election will mean losing all access to political power. Without trust, it is easy to imagine that one's opponents would not cede power after losing an election and might even resort to force to stay in power. When one believes that this is the political situation, democratic compromise is a dangerous option. Conversely, when people trust their fellow citizens in general, the stakes do not seem to be so high. Losing an election one year does not mean the end of access to political power. Citizens with higher interpersonal trust have reason to believe that their side may be the victor in a later election. Inglehart's data fit this explanation: levels of interpersonal trust consistently predict the existence and stability of democratic regimes. Inglehart also argues that life satisfaction is particularly critical in democracies.
Authoritarian regimes can use coercion to remain in power; democracies persist only if citizens are generally contented.
COGNITIVE MOBILIZATION AND POSTMATERIALISM: When there is sufficient economic development to provide for the basic material needs of the population, to create a varied economy with a strong service sector and to educate a high proportion of its population, societies experience several changes that are conducive to democracy (Inglehart 1997). As service sectors and educational opportunities broaden, two psychological changes particularly stand out: cognitive mobilization and a shift from materialist to postmaterialist values.
Basically, cognitive mobilization is an increase in the skills and the motivation required to engage in decisionmaking, including political decision-making. It is a byproduct of economic development because higher levels of education and service-sector employment raise a population's level of cognitive mobilization. Inglehart argues that industrial work centers on routine and changes little from day to day; it does not place demands on people's decision-making skills. On the other hand, service-sector occupations require specialized knowledge. People in these occupations have more autonomy and must exercise their own judgment when making decisions on the job. Thus, when economic changes increase the percentage of a country's population in service occupations and decrease the percentage in industrial occupations, the total cognitive mobilization of the population increases. The importance of cognitive mobilization is one reason why his model places economic development in a causally prior position to a sense of subjective wellbeing and interpersonal trust.
Cognitive mobilization and widespread material and physical security lead, in turn, to a shift from material values to postmaterial values. Inglehart defines materialist values as emphasizing "economic and physical security" and defines postmaterialist values as emphasizing "individual self-expression and quality of life concerns." When materialists dominate a country's discourse, public opinion will be concerned with inflation, economic growth, and other issues that bear directly on citizens' material wellbeing. Postmaterialists, on the other hand, are more concerned with personal autonomy, environmental issues, and quality of life concerns besides those directly linked to material possessions. Inglehart argues that the postmaterialist desire for self-expression and autonomy predisposes them toward democratic institutions and practices. Indeed, he argues that the same cultural changes that lead to postmaterialism also decrease economic growth.
Ø Making democracy work The quality of democratic decision-making is the heart of Robert Putnam's (1993) landmark book, Making Democracy Work: Civic Traditions in Modern Italy. Putnam found substantial variation in the performance of Italian governmental institutions across different regions, both in terms of differences between northern and southern Italy and among regions within the north and south. His explanation for this variation centered on the concept of social capital. Putnam defines social capital as "features of social organization, such as trust, norms, and networks that Power and democracy can improve the efficiency of society by facilitating coordinated actions". When people are embedded in dense social networks, possess norms of generalized reciprocity, and have high levels of interpersonal trust, they are more likely to be able to effectively organize and act collectively.
CIVIC VIRTUE: Putnam (1993) finds that the regions in Italy that have higher levels of interpersonal trust and more voluntary associations also have citizens who read newspapers, are interested in public affairs, and believe other citizens will "act fairly and obey the law." Moreover, public officials in these high social capital regions are more likely to believe in political equality and mass control of government and to show more willingness to compromise with their opponents than are politicians in regions with less social capital. In the former regions, public and collective action problems are more likely to be communicated to government and also to be solved. The citizens report higher satisfaction with their lives. Putnam's performance measures show that the regional governments are more responsive and deliver services and information more effectively in regions with high social capital.
In sharp contrast, regions with lower social capital are populated with citizens who are more likely to believe that public affairs are not part of their lives. Corruption is common and expected. There is much less participation in voluntary associations. Finally, public officials fear compromise and are more likely to believe that corruption is neither rare nor surprising. Putnam hypothesizes that since compromise and communication are keys to effective collective action, the collection of attitudes and habits that he finds in the civic regions leads to good government.
ASSOCIATIONS: One of the key features of the more civic-oriented regions is the large number of voluntary associations.
Associations include choral societies, model railroad clubs, sports teams, or any group where people assemble face to face to pursue common goals. Putnam asserts that people acquire trust and the skills required to act collectively through participation in voluntary associations, as well as a sense of public-spiritedness and "shared responsibility for collective endeavors". The associations need not be political to build interpersonal trust and lead to increased civic engagement. However, this is not to say that all organizations build social capital. Their impact depends upon how they are structured. As long as the members of the association meet with each other as equals, Putnam's evidence suggests that they create social capital.
Even though social capital is inherently interpersonal, it manifests in the attitudes and behaviors of individuals.
Horizontally organized (where all members are basically equal) associations build social capital. Vertically organized (where power relations are hierarchical) associations do not build social capital and may, in fact, undermine it. This is a consequence of the form of interactions in each type of organization.
• In vertically structured organizations, more powerful members can defect from agreements with less powerful members of the group with impunity. There are no sanctions available to enforce norms of reciprocity.
• Horizontal organizations do not accomplish their goals through command and obedience. Instead, the members accomplish their goals through cooperation. Repeated interactions with fellow members build a history of successful collaboration, which in turn builds norms of generalized reciprocity and interpersonal trust. Putnam finds that horizontal associations are more common in areas with good government, whereas vertically organized organizations are more common in regions lacking good government.
Ø Individual-level evidence Brehm & Rahn (1997) explain social capital in the United States with an individual-level analysis. They assert that "the phenomenon of social capital manifests itself in individuals as a tight reciprocal relationship between levels of civic engagement and interpersonal trust." They model a three-way causal nexus among interpersonal trust, civic engagement (belonging to voluntary associations), and confidence in American national institutions.
• They find significant relationships in both directions between interpersonal trust and civic engagement.
• The path from civic engagement to interpersonal trust is much stronger than the path from trust to civic engagement.
• Their study finds stronger evidence that engagement builds trust than that interpersonal trust leads people to join voluntary associations.
• They have found that individual-level attributes, such as interpersonal trust, predict the other phenomena (civic engagement and confidence in government) found in Putnam's theories.
Ø Trust and associations Power and democracy Stolle surveyed the members of voluntary associations in Germany and Sweden, concentrating on horizontally structured associations because of Putnam's prediction on the sources of social capital. Her analysis of the variation in members' interpersonal trust generated some surprising results. She finds no evidence of a linear relationship between the time that individuals have spent in an association and their level of interpersonal trust.
• Those who have been in an association for one year are more trusting than those who have been members for only a short time (less than a year). Members who have been in the association for long periods of time (more than 5 years) actually show declining levels of interpersonal trust.
• It appears that joining an association boosts trust, but then the effect subsides for members who stay in the group for long periods. This may be explained by the effect of intragroup trust on generalized interpersonal trust.
Stolle hypothesizes that when members have strong bonds to the other members of their association (measured by their trust toward other members of their association), their trust will not generalize to others in general. As hypothesized, she finds a negative relationship between in-group trust and generalized trust. Thus, it appears that when people are loosely bound to an association, their trust for their fellow members generalizes, but when they are tightly bound, they are more likely to trust only their fellow members. She also found that diversity, as measured by variation in socio-economic status, does not have significant effects on generalized trust. However, when diversity is measured as the share of "foreigners" in an association, diversity shows a positive relationship with generalized trust.
Putnam (1993) insists that horizontally organized associations increase social capital but that vertical associations do not. To test this hypothesis, Stolle included a variable called engagement in her analyses. The engagement questions asked the respondent whether they had ever done "a responsibility task" for their association and whether they had ever "organized a project." From these questions, she built a mean score for each organization. Higher levels mean that more members have had more organizational experience; more breadth in responsibility and leadership should indicate a horizontal organization. The mean score for each organization was then assigned to each member of the organization; it indicated the breadth of involvement of a given group's members. As expected, she found that engagement is positively related to generalized trust.
Ø Projection of trust Orbell & Dawes ask how trust survives in a society even though it seems that high trusters are setting themselves up to be taken advantage of. They base their theory of cooperators' advantage on projection, arguing that people project their own trustworthiness onto others; people use information about themselves as a heuristic to form beliefs about what they can expect from others. An individual who honors their agreements expects others to do the same.
Orbell & Dawes (1991) expand the normal configuration of the prisoners' dilemma by also giving the players the option of whether to elect to play a game or not. Thus, in their game, a player decides whether to enter into a partnership with another player and also decides whether to cooperate or defect with their partner. They show that since cooperators (who expect their partners to be trustworthy) will enter into more partnerships than those who do not trust (who expect others to defect from agreements), there is an evolutionary advantage in trusting. Thus, even though expecting people to be trustworthy is riskier than avoiding relationships, the greater gains make it an advantageous strategy.