Cambridge University Press
978-0-521-87440-3 - Partisan Families - The Social Logic of Bounded Partisanship in Germany and Britain - by Alan S. Zuckerman, Josip Dasović and Jennifer Fitzgerald
The Social Logic of Partisanship:
A Theoretical Excursion
[A]nd man became a living being, – with a soul able to speak and reason.
Targum Onkeles on the phrase l’nefesh hayah in Genesis 2:7
[M]an alone of all the animals is furnished with the faculty of language.
Aristotle, The Politics, II: 10 (Barker 1962, 6)
As a social being the person needs to be capable of reading messages from other persons, or responding to these and of composing intelligible messages to send out.…This person has to have beliefs about how the world is and how it works, ontological knowledge, and knowledge about how other persons behave.…Equipped with the wherewithal to make choices, the rational social being will apply choice to dealings with other persons and will develop strategies for manipulating and controlling them and for escaping unwanted control.
Douglas and Ney (1998, 89)
Partisanship is a socially derived choice. In this statement, wedraw together two divergent theoretical sources: a socialized understanding of humans and a perspective that emphasizes persons as beings that seek to make reasoned decisions. Our perspective denies that social groups subsume individuals; people do more than simply act as reflections of their social categories and locations. As important, persons are also not “free floating atoms” (to recall Marx’s criticism of liberal theory). People reason, but they do not necessarily maximize expected utility or make the objectively correct choices. Instead, they employ subjective or bounded rationality, seeking to do the best they can with the intellectual and informational means available to them. Learning from others aids decisions, but learning from others is also what people do, as people.
Purposefully as well as inadvertently, people send and receive political messages. They influence those who receive their decisions and are, in turn, swayed by people whose messages they receive. The frequency of interaction affects the probability of influence: as more messages are sent and received, the level of influence shared increases. The level of shared trust affects political communication as well. Families, it follows, play a major role in political choices. These relationships are probabilities, not determined certainties. They apply to everyone. Politicians, whose political choices are made in contests with clear goals and rules, operate under an additional set of decision rules. We return to the distinction between citizens and politicians in the final chapter (Meehl 1977 and Riker 1982 provide the classic basis for this distinction). Here, we reason about citizens who are not politicians, and in the empirical analyses, we examine how they influence each other’s choices about political parties.
THE SOCIAL LOGIC OF PARTISAN CHOICE
Inherently social, people live their lives by interacting with others, by considering and anticipating the behavior of others, and by influencing each other. Numerous social mechanisms account for this principle. Some emphasize the simple number of interactions: the greater the number, the more likely there is to be influence. Physical propinquity matters in Tobler’s Law, “Everything is related to everything else, but near things are more related than far things” (cited in Miller 2004). In turn, the ancient principle of “like to like” (i.e., similar persons are drawn to each other) and its obverse (i.e., “opposites attract”) also draw attention to the nature of the interactions.
Consider also the following general statements taken from diverse research traditions:
A social being has one prime need – to communicate. Because it is a social being, everything in its genetic inheritance, especially its intelligence, must be equipped to read the signals and to signal back to beings of its kind. (Douglas and Ney 1998, 46)
Among social species we are unique in our plasticity.…Behavioral rules (including social norms and conventions) make social interaction predictable, so that interdependent individuals can influence one another in response to the influence they receive, hereby carving out locally stable patterns of interaction. In short, rules are not simply analytic shortcuts that lower the cognitive costs of decision-making. They are the grammar that structures our social life. (Macy 1998, 221)
The capacity of people to engage in complex cooperation, and to seek and give social support in particular, might be a defining characteristic of our species. (Cunningham and Barbee 2000, 274)
Individuals learn appropriate social behavior from observing each other. This proposition emphasizes the interconnected nature of social life, underlining the centrality of observing and copying others as people perceive, evaluate, and make decisions about how to act. We view behavior as more correct in a given situation to the degree that we see others performing it. (Cialdini, cited in Axelrod 1997b, 58)
Conformity is inherited. It is inherited socially, because social norms develop that reinforce conformity –glory [sic!] prestige, pride in the group, patriotism.…It is inherited genetically because social learning provides such an evolutionary advantage to the individual. (Jones 2001, 117)
Consider as well the recent research that has located in the dorsal striatum section of the brain the physical location for these kinds of expectations and learning rules (Berns et al. 2005 and King-Cassas et al. 2005). Diverse theories agree that humans are social beings necessarily living social lives.
Social animals though they be, humans do not move in herds (though one should not minimize the importance of cue-giving and taking among animals). Even as they communicate, persons stand apart from even the closest associates, spouses, and parents; each and all always retain the ability to accept or reject cues, to adhere to or ignore other people’s expectations and preferences. More, members of social groups disagree, sometimes sending conflicting messages to each other, sometimes fighting. Also, persons belong to many social groups, whose signals vary in the extent to which they are consonant with each other. Even as a person is influenced by members of a group, he or she always chooses whether or not to follow others, and the more that the members of a person’s groups disagree with each other, the more that he or she must pick whom to follow. The fundamental need to communicate does not remove the ability to stand apart from others and to ignore or disagree with them, and, therefore, the need to decide to agree or to disagree. And so, political influence, like other social phenomena, is best understood as a probabilistic relationship.
Seeking to account for the extent and ways that individuals learn from each other, an assortment of research traditions underlines the variability of social conformity. Some make this argument on theoretical grounds, and here too the overlap between divergent theories underscores the power of the point (for Freudian theories see Alford 1994; Bion 1961; for rational choice theories, see Lichbach 1996; 1997; Olson 1965). Some, like Mary Douglas, join theory and a synthesis of anthropological studies of small-scale communities (Douglas 1986; Douglas and Ney 1998). Political scientists display consistent evidence of political diversity in discussion networks and families (Huckfeldt, Johnson, and Sprague 2004; 2005; Stoker and Jennings 2005; Zuckerman and Kotler-Berkowitz 1998; Zuckerman, Fitzgerald, and Dasović 2005). Even as some argue for the existence of a powerful drive to cultural agreement (Axelrod 1997a; 1997b), classic theories (Simmel 1955), and recent work using agent-based models (Johnson and Huckfeldt 2005) sustain the principle of opinion diversity within members of social networks. Social conformity may not be assumed; individuals are not subsumed by their social ties; surrounded and influenced by others, they make their own decisions; relationships of social influence are probabilities.
How do they decide? The principles of bounded or subjective rationality guide the way. People seek to make what they perceive to be the correct decision, even if they do not succeed and even if the decision is not objectively correct (see Boudon 1992; 1998; Gigerenzer and Selten 2001a; Jones 2001; and for the classic sources on bounded rationality see Simon 1965 ; 1999). They seek strong reasons for what they do (Boudon 1992; 1998).
Usually eschewing complex calculations, they apply useful rules of thumb – or usually correct theories or understandings or decision heuristics, which are conditioned by their social circumstances and their intellectual abilities. Many of these decision rules involve taking cues from members of one’s immediate social environment. Many do not. Here, we emphasize those that draw on principles of social learning for decision heuristics.
We begin with the teachings of Kenneth Arrow. In large publics, he argues, the rationality assumption does not apply. Individual rationality is not just a property of each person. It is linked to the macro characteristics of the economy, the presence of economic equilibrium, perfect competition, and “completeness of markets:”
When these assumptions fail, the very concept of rationality becomes threatened, because perceptions of others and, in particular, their rationality become part of one’s own. (1986, S389)
Consider also the observations of other economists and decision theorists.
[I]n social species, imitation and social learning can be seen as mechanisms that enable fast learning and obviate the need for individual calculations of expected utilities.
Social norms can be seen as fast and frugal behavioral mechanisms that dispense with individual cost-benefit computations and decision-making. (Gigerenzer and Selten 2001a, 9–10)
Social imitation can help make decisions with limited time and knowledge. Heuristics such as “eat what your peers eat” and “prefer mates picked by others” can speed up decision making by reducing the time spent on information search. (Selten 2001, 48)
Imitate if Better, assumes that individuals imitate all others who are more successful than themselves, and stick with their current strategy otherwise.
Proportional Imitation – which dictates imitating those who are more successful than oneself with a probability that is proportional to the difference between the observed and current degrees of success. (Group Report in Gigerenzer and Selten 2001a, 175)
The behavioral decisions made by animals are to a large part influenced by what other animals are doing. Social learning is not restricted to humans, or clever animals, but is a fundamental feature of vertebrate life. In an array of different contexts, numerous animals adopt a “do-what-others-do strategy,” and in the process, learn an appropriate behavior. (Laland 2001, 233)
Most people most of the time eschew optimization and frequently apply social learning as a strategy for making choices.
How do boundedly rational persons engage in social learning? They are influenced by persons from whom they take other cues; successful learning in the past induces the expectation of successful learning in the present and future. When cue-givers disagree with each other, boundedly rational persons choose alternative strategies. They may follow the net preferences of these persons; they may weight them on an easily constructed scale; they may follow the most knowledgeable, the most trusted, the most frequently seen, or they may apply some other decision heuristic. No single strategy always applies. In the empirical portions of this book, we will show that this variation even characterizes families and households.
Joining these two themes implies two explanatory mechanisms. 1. People learn from others because that is what social beings do. 2. People learn from others because it is an effective and frequently used decision heuristic. Both general propositions justify the claim that partisanship and electoral choice are socially derived choices, such that the preferences of others affect the decision to support and vote for a political party. They imply too that family members have a particularly strong influence on each other’s political decisions. Note as well the limits of these claims. They imply no more than people sometimes learn from others and sometimes do not. The relationship is probabilistic. Like other social mechanisms, these are sometimes true theories (Elster 1998; Hernes 1998).
From whom are persons especially likely to take political cues? Individuals are especially likely to follow those persons from whom they take other cues, who know more than they do, those on whom they depend, whom they trust, with whom they regularly interact, and whom they perceive as being like themselves. They are especially likely to follow those to whom they feel accountable (Tetlock and Lerner 1999). Conversely, they are especially likely to reject the political cues of persons who know less than they do, whom they usually ignore, perceive as different from themselves, whom they do not trust, and who are strangers to them, and to whom they do not feel accountable. More concretely, in the following chapters, we show that family members are more likely to influence partisan choice than, for example, membership in a trade union or religious congregation. Note that we do not emphasize the importance of asymmetries in political knowledge or expertise as sources of political influence. In our view, choosing a political party is not an optimizing decision for which additional knowledge might affect the outcome. Trust and frequency of interaction underpin the political influence that affects this choice.
Consider, however, that the influence of social intimates on partisanship does not determine that they incline towards the same party; each or all may opt not to prefer any party at all. Indeed, in the extreme case, when persons who live together spend all their time together, they are social and political isolates and, therefore, are likely to support no party. Interacting only with each other, they only know each other; they take interest in no one else. Persons with ties outside the household have opportunities to absorb political messages from those with whom they work, play, pray, and share membership in social organizations. In turn, the cues that move across these weak ties affect partisan preferences and then are transmitted to those with whom they share strong ties.
POLITICAL LEARNING IN POLITICAL SCIENCE
Our understanding of the dynamics of political choice overlaps with and departs from other studies that also accept the principles of social learning. Consider first how we stand in relation to Achen’s (2002) application of rational choice theory to partisan socialization within families. Like Achen, we accept the principle that persons are not passive receptacles of political messages. Like Achen, we maintain that persons learn from those around them. We differ, however, with regard to the process of political learning. For Achen, children learn from their parents by following a complex set of steps. Children, like all persons, support a political party, “when they believe that its future course of benefits exceeds that of the other alternatives. When the voters expect that a party will benefit them in the future, they will be said to ‘identify’ with that party” (Ibid, 153). Each party is perceived to offer “a stream of benefits (cardinal utilities) that varies randomly around a constant mean” (Ibid, 153–4). Furthermore, children, like their parents and everyone else, take their understanding of desired political benefits from their social positions (Ibid, 154). And so Achen asks and answers the fundamental question of political socialization: “Why should children decide to think like their parents? The answer given is that parent and child will often occupy similar positions in the social structure, and thus parental experience is likely to be relevant to the child’s future life” (Ibid, 154–5). Seeking to maximize their expected benefits and supposing that they will occupy the same social positions as their parents, young people take their parents’ partisanship.
As noted, we eschew the claim that people choose parties by calculating costs and benefits: choosing a party carries no personal outcome; benefiting no one, partisanship suggests no interests, and the person who can make the required complex calculations of optimizing choices can easily recognize that these are not expectations about outcomes but hopes and wishes. More generally, as citizens consider the political parties they lack a “macroscopic driving force” (Aumann 2005) that guides their decisions. Choosing a political party is nothing like making money in the market place or winning in chess; for that matter voting too bears none of the critical characteristics that imply the strategic importance of optimization. Because partisanship and electoral behavior lack clear goals, the classic principles of rational choice theory do not apply. Furthermore, our analysis also shows that social class is a much weaker predictor of partisan choice and consistency than Achen’s assumptions allow.
Children learn partisanship from their parents, we suggest, not because they expect to benefit from the choice and not because they and their parents share the same social location, and so it is smart to learn from their parents. Rather, interacting frequently with their parents, they take many cues, among them political ones, from them. Parents are generally trusted (indeed usually beloved) persons, with whom young persons interact on a regular basis. And so, when young people offer a positive answer to a question about party preference, they usually echo what they have heard at home. In Chapter 5, we show that they do so in specific ways: almost always ignoring or rejecting the party that their parents do not support and varying the rate by which they accept their parents’ partisan preferences. Furthermore, as children enter and move through their third decade, obtaining the right to vote, social relationships within households imply that parents also learn from their children. In Chapters 5 and 6, we show that the exchange of partisan influence is strongest between mothers and children.
Indeed, one of the questions that helps to define Achen’s approach further elucidates our differences, and, we suggest, highlights the strength of our approach. After denying that children are passive recipients of their parents’ wisdom, he continues, “Parents are rarely able to influence their teenage children’s hairdos, clothing styles, tastes in popular music, or even more important decisions like the choice of a life partner. Why should party identification be any different? Put more precisely, why do teenagers implicitly accept their parents’ advice about political parties while they avoid taking it on a great many other topics?” (Achen 2002, 152). Achen’s model answers these questions. We are less surprised than Achen that young persons are more likely to accept their parents’ partisanship than that of their friends. After all, parents are more likely to have partisan preferences than are the youngster’s peers. Parents are also more likely to display their political preferences in words and gestures than are young persons, among whom political discussions are not frequent; teenagers rarely exchange political cues. And why suggest, as does Achen, that partisanship is unique? In Chapter 5, we show that children usually take the religion and the social class preference of their parents, but are also more likely to deny an affiliation or identification with any of the relevant social groups. There is nothing unique about partisanship, and partisan preference is not like the choice of clothes or hairdos; it is more like the choice of religion. In Chapter 5, we summarize a substantial literature in social psychology that maintains that parents usually matter more than friends, and we elaborate our argument on political learning in households.
Consider also how we stand in relation to others who apply decision heuristics to political choice. Like Fowler (2005), we expect each person’s decision to vote to influence and to be affected by the turnout of others in their social networks. Also like Fowler, we emphasize exchanges in families and households, rather than learning from those who are very knowledgeable about or interested in politics (Downs 1957; Huckfeldt, Johnson, and Sprague 2004), the media (Mutz 1998), or political institutions (Lupia and McCubbins 2000). We agree with Lupia and McCubbins that social learning requires trust, but we believe that it is both more parsimonious and more accurate to look to loved ones – parents, spouses, children, friends, and workmates – than to political institutions for sources of trusted information. The political preferences of one’s social intimates provide the lens through which a person perceives the political parties and the general political world (Schmitt-Beck 2003). All these analyses differ from those that apply decision rules on partisanship that are internal to the actor, such as ideology (Downs 1957; Boudon 1992), perceptions of the incumbent (Boudon 1992; Sniderman, Brody, and Tetlock 1991, 180), and assessments of relative feelings about groups – likeability or evaluations of desert, the desert heuristic (Boudon 1992, 23; Sniderman, Brody, and Tetlock 1991, 87–8). We emphasize personal ties, especially frequency of interaction, past learning, and trust as the factors that enable persons to apply heuristics to political choice.
In this volume, we present the results of several analyses that demonstrate the importance of social factors as predictors of partisanship. Both immediate social circles (household partners and parents and children) and more distant social contexts (social class and religious membership and identifications) as well as the political interest of the respondent and others in the social network influence whether or not a party is chosen, the party named, and the consistency of that selection. This evidence illustrates the power of the social logic of partisanship.
AN INTELLECTUAL HISTORY OF THE SOCIAL LOGIC OF POLITICS
Our analysis of partisanship1 rests on what we call the social logic of poli-tics. In this section, we recount an intellectual history in order to provide theoretical underpinning for a set of related propositions: (1) The social logic of politics has an ancient pedigree that supports modern social science and is part of most people’s intuitive understanding of social life. The principles sustained the first generation of studies of partisanship. (2) The reasons to move away from these principles were weak when first offered and are no longer tenable at all. (3) There are strong theoretical reasons to return to the social logic of politics, in general, and as applied to the analysis of partisanship in particular.
Students of politics have always known that the immediate social circumstances of people’s lives affect their political perceptions, preferences, and behavior. Recurrently, however, issues of theory and method induce them to follow other paths, and, recurrently too, they come back to this perspective.
Consider first the intellectual pedigree. Relevant stories, principles, themes, and statements abound in the Bible, Greek classics, and foundational works of the medieval period. The Bible opens (and see the reference to Onkeles’s commentary at the start of the Preface) with stories that center the principles of human life on the family. Aristotle begins the Politics with a discussion of the household, citing Hesiod: “First, house, and wife, and ox to draw the plow” (Barker 1962, 4). In more abstract terms, Aristotle defines the family as the nucleus of the polity, a claim that resonates through centuries of Christian political thought. Even a cursory look at these literatures finds the family at the heart of personal and collective life.
More directly linked to the task at hand, these literatures recognize that people live their lives by learning from those around them. Ancient proverbs are presented as analytical truths and behavioral rules that lead to a good life. One set of principles involves the benefits obtained from good friends and the need to avoid bad associates. Another set involves the selection of colleagues: similar persons associate with each other; but sometimes they repel each other, and different people may find themselves drawn together as well.
Here we present Aristotle’s views, highlighting critical statements and propositions.2 Aristotle explores the importance of friends in the following passage from the Nicomechean Ethics:
No one would choose to live without friends, even if he had all other goods.… Friends help young men avoid error; to older people they give the care and help needed to supplement the failing powers of actions which infirmity brings in its train; and to those in their prime they give the opportunity to perform noble actions.…Friends enhance our ability to think and to act.…There are, however, several controversial points about friendship. Some people define it as a kind of likeness, and say that friends are those like us; hence according to them, the proverb: “Like to like,” “Birds of a feather flock together.” Others, on the contrary, hold that all similar individuals are mutually opposed. (Aristotle, Nicomechean Ethics 1155a, 214–6)
The friendship of base people becomes wicked, because unsteady as they are, they share in base pursuits, and by becoming like one another they become wicked. But the friendship of good men is good, and it increases with [the frequency of] meetings. Also, it seems, they become better as they are active together and correct one another: from the mould of the other each takes the imprints of the traits he likes, whence the saying: “Noble things from noble people.” (Ibid, 1172a, 271–2)
These proverbs did not originate with Aristotle. Rather, he cites Homer, Empedocles, Euripides, and Heraclites. Indeed, these proverbs indicate that even Aristotle considered the aphorisms “like to like,” “birds of a feather flock together,” and “opposites attract” to be the wisdom of the ancients.
Centuries later, Maimonides (1135–1204) and Aquinas (1225 or 1227–74) extend these principles:
It is natural to man to be influenced by the beliefs and practices of his friends and neighbors and to behave like the practices of his community. Therefore, a man should attach himself to the righteous and sit always among the wise, so that he may learn from their deeds. He should also stay away from evil people who walk in darkness, so as not to learn from them. (Maimonides, Mishneh Torah, 6:1, our translation)3
Friends become better by working together and loving each other. For one receives from the other an example of virtuous work which is at the same time pleasing to him. Hence it is proverbial that man adopts noble deeds from noble men. (Aquinas, Commentary on the Nicomechean Ethics 2:5)
Are friends necessary for happiness? Response: If the question refers to the happiness of the present life then, as the Philosopher says, the happy man needs friends, not because they are useful, since he is able to get along without help, nor for the enjoyment, since the activity of virtue furnishes him with complete joy, but for good activity, that is so that he may do good to them and delight in seeing them do good and be helped by them in doing his own good deeds. For, in order to do well, whether in the works of the active life or in the activity of the contemplative life, man needs the help of friends. (Aquinas, Treatise on Happiness, 12)
© Cambridge University Press