Jon Elster, France, Unraveling of an Absolutist Regime
Rational choice theory & the Ancien Régime [Chapter 1, Vol. 1]
MGH: I noticed this morning that our valued loyal Social Science Files subscriber, Jon Elster, has a new book out in April 2023, volume 2 of his projected 3-volume series …
America before 1787: The Unraveling of a Colonial Regime [link to publisher site]
Therefore …. It is time to exhibit part of Chapter 1 of Volume 1, ‘sans’ highlighting, to whet the appetite! The chapter exceeds email length so I omit the first section (titled ‘The Nature of the Ancien Régime’) and a couple of paragraphs towards the end.
Jon Elster wrote:
Chapter one
Introduction
Mechanisms: The Importance of Choice
The argument of the book is driven by two concerns: the quest for causality and the quest for agency (methodological individualism). Jointly, these concerns imply a focus on choice as the key explanatory variable. A negative implication is that the mere effect of an institution cannot serve to explain it. It may well be true that this or that institution of the ancien régime served as a safety valve by keeping discontent at manageable levels, but that fact does not explain why it exists. Tocqueville, who mostly adhered to the principles of methodological individualism (without indicating that he had any methodology), violated them when he wrote that after the last Estates-General in 1614, the last before the Revolution,
The . . . desire to escape the tutelage of the estates led to the attribution to the parlements of most of their political prerogatives. There was a need to appear to provide new guarantees in place of those that had been eliminated because the French, who will put up rather patiently with absolute power as long as it is not oppressive, never like the sight of it, and it is always wise to raise some apparent barriers in front of it, barriers that cannot stop it but nevertheless hide it a little.
As an historical analysis of the origins of the politicized parlements, this statement is sheer fantasy. As a comment on the effects of the politicization it may or may not be true, but it is not absurd.
While compelling as a first principle, methodological individualism is often too demanding in practice. We rarely have the documentary evidence that we would need to identify the beliefs and motivations of, say, each peasant in an insurrection or each magistrate in a judicial strike. Nor can we assume that an occasional explicit statement by an agent is reliable and representative. It is also difficult to consistently avoid the cardinal sin of inferring mental states from the actions they are supposed to explain. Historians are forced to triangulate many sources to impute motivations and beliefs. When drawing on their work, I cannot but trust their intimate knowledge that comes from a lifetime in the archives.
To this caveat, I shall add another one. Beliefs and desires are intrinsically hard to identify with precision, for three reasons stated by James Madison in The Federalist # 37.
First, there is indistinctness of the object: it is not always clear whether the agents possessed stable beliefs and desires that they used as premises for action. A riot that is cut short because the participants go home for dinner (chapter 2) probably did not stem from a deep and strong conviction. Did the French peasantry really “believe” in rumors about an impending tax on children? Second, there is imperfec- tion of the organ of conception: “The faculties of the mind itself have never yet been distinguished and defined, with satisfactory precision, by all the efforts of the most acute and metaphysical philosophers” (Madison). How can we determine whether a reform triggered a desire for more reforms or a belief that more reforms were imminent? Can we tell for sure whether an act of aggression was motivated by anger, by envy, or by hatred? Finally, there is inadequateness of the vehicle of ideas: “no language is so copious as to supply words and phrases for every complex idea, or so correct as not to include many equivocally denoting different ideas” (Madison). In his journal, the bookseller Hardy is constantly struggling to find the words for the exact degree of credibility of the rumors he is reporting. In other words, there may not be a fact of the matter; even if there is, we may not be able to grasp it; even if we could, we may not be able to state it unambiguously. As Hegel notes somewhere, language always says both more and less than what the speaker or writer intended. Yet as Eliot’s Sweeney says, “I gotta use words when I talk to you.”
The focus on choices—their antecedents and their consequences— puts me in a different camp from many historians of the ancien régime. I shall pay little attention to the intellectual and cultural preconditions, or alleged preconditions, of the Revolution. The influence of the Enlighten- ment on political events and social movements was possibly strong, but it was certainly diffuse, often too diffuse to provide a causal, individual- level explanation of specific choices and decisions. Although, as I said, desacralization of the king and decline in religious fervor coincided with the Enlightenment, the causal relations are opaque. Depending on one’s definitions, desacralization and decline in religious fervor may even be constitutive of the Enlightenment.
When parish priests deserted en masse to the third estate in June 1789, when soldiers refused orders to shoot on the people in July 1789, and when women intruded on the privacy of the royal couple in October 1789, they certainly exhibited lack of the traditional deference towards their superiors. Yet the actions of the parish priests may have been due mainly to the fact that for the first time in the history of the Estates-General they formed a majority of the clergy, those of the soldiers to the unwise decisions of the July conspirators, and those of the women to the popular attitude towards Marie-Antoinette. I shall cover these events in Volume 3. Here, I am only suggesting that these specific facts, demonstrable independently of the events they are intended to explain (at least in part), are more probative than more diffuse and general tendencies.
The relevance of discourse analysis is limited in a context where the vast majority of the people was illiterate. We do not know to what extent discontent percolated down from the educated elites or, on the contrary, whether popular unrest provided the ferment for more articulate state- ments by lawyers and wealthy commoners. There is more evidence about intra-elite discourse. We may, for instance, follow the semantic trans- formations of terms like “credit” from the sixteenth century to the eighteenth, reflecting the changing relations between the monarchs and the nobles. Yet it is only these relations that have causal efficacy, not their verbal expressions.
For many social scientists, the privileged form of choice is rational choice. While often useful, rational-choice explanations of behavior can- not, in my opinion, claim any privilege, except on the (irrelevant) grounds that they lend themselves to mathematical modeling. Let me first state the standard rational-choice model of action and then generalize it. The stan- dard model and the general model involve many of the same variables: desires (preferences, motivations), beliefs, and information. The general model also introduces emotions.
The basic elements of the standard rational-choice model are shown in Figure 1.1. The arrows stand both for causal relations and for optimality relations. The desire, for instance, is both what makes the action optimal and what causes it. A rational choice is the joint product of the agent’s desires and her well-informed beliefs: doing as well as (she rationally believes) she can. For the beliefs to be rational, she has to collect an optimal amount—not too little, nor too much—of information, and then process it rationally in light of prior, more general beliefs. In this model, there is no room for motivated belief formation—it excludes a direct causal influence of desires on beliefs.
Figure 1.1. The standard model of rational choice
The standard model of rational choice can obviously explain a great many actions. To take an example, since peasants knew from experience that prompt payment of taxes in a given year would lead to higher impositions in the next year, they rationally procrastinated in paying, even if they had to pay a fine for late payment. Knowing that the cabinet noir would open and read their correspondence, many high-placed individuals took rational precautions, Turgot by sending his mail by private courier, Madame de Sévigné by using pseudonyms when referring to the royal household, and Saint-Simon by including only anodyne matters. Bondholders rationally demanded higher interest rates for risky loans. Such examples could be multiplied indefinitely.
Game theory—which ought really to be named “the theory of interdependent decisions”—is a special case of rational-choice theory. It arises when each of two or several agents need to form rational expectations about what the others will do, in order to respond optimally to their choices. In an early treatment, Nicolas de Montmort provided a numerical example that seems to lead to an infinite regress of “I think that he thinks that I think . . . ,” and added, “Such questions are very simple, but I believe they are unsolvable. If that is the case, it is a great pity, for this problem often arises in ordinary life, as when each of two people who have some business together wants to adjust his behavior to that of the other.” Some game theorists claim that this insolvability has since been overcome, since the idea of an equilibrium, in which each agent chooses the best response to the best-response choices of all others, allows one to short-circuit the infinite regress. However, in Montmort’s example, this “solution” requires the agents to use a chance-wheel that assigns calculable probabilities to each of the possible responses (I simplify), a requirement that makes the “solution” devoid of empirical interest. In my view, this is also the case for many of the other solution concepts that game theorists have proposed. Be this as it may, for my purposes in this book the relevant question is whether agents can be assumed to know each other’s preferences. In a two- person game involving the provision of a public good, for instance, each agent may not know whether the other prefers mutual cooperation over being the unilateral non-cooperator. There may be two equilibria, one good (mutual cooperation) and one bad (mutual non-cooperation), but if the agents do not know each other’s preferences none of the equilibria will stand out as the solution. [Later in the book] I suggest that some of the perverse features of préséance can be understood in this perspective.
When the rational-choice model fails, it is either because of indeterminacy—the model does not tell the agent what to do—or because of irrationality—the agent does not do what the model tells her to do.
Indeterminacy arises largely because of uncertainty. Because of the unreliable flow of information to the royal government, rational belief formation about the capacity of the economy was virtually impossible. Conversely, because the government shrouded both its revenues and expenses in secrecy, other agents, such as the parlements, had no basis for forming a rational judgment about the need for new taxes. In fact, the government itself often did not know the size of its debt. Also, as we shall see, game-theoretic situations can generate uncertainty, for instance if agents are unable to predict whether others agents are likely to cooperate.
Irrationality arises when the processing of information is subject to either a “cold” (unmotivated) bias or a “hot” (motivated) bias. Cold bias is a recent, revolutionary idea developed by psychologists and behavioral economists over the last half-century. They have demonstrated by experiments and field studies that both the ancient and classical moralists (see below) were wrong when implicitly assuming that irrationality was always due to passion. Zero-sum mercantilist policies, such as those advocated by Colbert, may be due to a cold cognitive bias. Although I cannot cite specific instances, I am confident that officials in the ancien régime (as officials everywhere) were occasionally subject to the sunk-cost fallacy (throwing good money after bad) or to the recency effect (paying more attention to new information than to older, equally relevant data).
I shall understand hot bias mainly as emotional bias, although it can also have other sources. When emotion shapes cognition, it can do so in two main ways. First, as La Fontaine wrote, “Each believes easily what he fears and what he hopes.” The second half of that phrase refers to garden- variety wishful thinking, such as the belief of the peasantry that a temporary relief from a tax would be a permanent one or that relief from one tax implied the relief from all others.
The first half of the phrase is more puzzling: why would people form beliefs that are both unsupported by evidence (as in wishful thinking) and are contrary to what they would like to be true? John Stuart Mill proposed an ambiguous answer: “[T]he most common case of [Bias] is that in which we are biased by our wishes; but the liability is almost as great to the undue adoption of a conclusion which is disagreeable to us as of one which is agreeable, if it be of a nature to bring into action any of the stronger passions. Indeed, it is a psychological law, deducible from the most general laws of the mental constitution of man, that any strong passion renders us credulous as to the existence of objects suitable to excite it.” I leave it to readers to sort out the relation between the two statements I have italicized. Independent of “general laws,” however, the existence of fear-based rumors provides indisputable proof that we easily believe what we fear. Social psychologists have also found that fear induces both stronger risk-aversion (compared to the agent’s non-emotional state) and more pessimistic risk- assessments (compared to those of a neutral observer).
Anger and, more conjecturally, enthusiasm have the opposite effects on both dimensions. It follows that when we observe agents engaging in highly risky behavior—magistrates ordering officials not to execute a royal edict (chapter 4) or urban consumers pillaging bakeries for bread (chapter 3)—the explanation may either be a cognitive one (underestimat- ing the risk) or a motivational one (assessing the risk correctly and accepting it). In the first case, the appropriate term may be foolhardiness; in the sec- ond, courage. In practice, as Madison warned, we may not be able to tell.
Emotions can also shape cognition by their urgency, by which I mean a desire to act sooner rather than later. (A more vivid term is “inaction- aversion.”) Although this tendency can be rational in the face of an acute danger or need, it is also observed in situations where there would be nothing to lose and possibly something to gain from taking the time to gather more information. In On Anger (I. xi), Seneca asked:
How else did Fabius restore the broken forces of the state but by know- ing how to loiter, to put off, and to wait—things of which angry men know nothing? The state, which was standing then in the utmost extremity, had surely perished if Fabius had ventured to do all that anger prompted. But he took into consideration the well-being of the state, and, estimating strength, of which nothing now could be lost without the loss of all, he buried all thought of resentment and revenge and was concerned only with expediency and the fitting opportunity; he conquered anger before he conquered Hannibal (my italics).
In the ancien régime, urgency on the part of the government was mainly a rational response to the constant need for money to fund wars, pay creditors, and pay for luxury at the Court. By contrast, crowds provide examples of urgency-caused irrationality, when they attack alleged male- factors on the basis of suspicions without taking the time to verify them. There was but a short step from an accusation to a conviction that it was justified. The suspicions and the anger they generated were, in turn, shaped by the tendency to believe what one fears.
Figure 1.2 shows how emotions can be integrated in a more general model of choice, which deviates from the standard model in two ways. On the one hand, the general model contrasts with the standard model in allowing for a causal influence from desires to beliefs. In belief formation, causality and optimality can diverge. On the other hand, the general model expands the standard model in allowing for a causal influence from beliefs to desires, mediated by emotions. In the standard model, desires are the unmoved movers in the machinery of action, whereas the general model allows us to go a step further back in the causal chain. In the formation of desires, the question of optimality and causality diverging from each other does not arise, since the notion of an optimal desire is not well defined.
Figure 1.2. A general model of choice
Emotions are triggered by beliefs, but can be reinforced by perceptions. Reading in the tax documents the list of duties levied by tax officials may cause anger, but seeing the “black-robed counselors” in the streets may turn the anger into fury. In this book, I refer to beliefs as a shorthand for both cognitions and perceptions as causes of emotion. Emotions differ among each other mainly by the beliefs that trigger them and by the desires for action they trigger. To illustrate, consider the difference between (what Descartes called) anger and indignation. A feels anger when he believes that B has harmed him unjustly, and indignation when he believes that B has harmed a third person C unjustly. Both anger and indignation make A want to harm B, but experiments confirm the intuition that third-party punishments will be substantially weaker than second-party ones.
In these experiments, carried out by Ernst Fehr and his cooperators, the intensity of anger and indignation is measured by how much the subjects are willing to harm themselves in order to harm the target of their emotion. In the ancien régime, harming another person in anger could be risky or costly; leaders of rebellions were often executed. Indignation, expressed as speaking truth to power, did not impose any material harm on the targets, whether it was the king or his officials, but if expressed publicly it might undermine, or be thought to undermine, their legitimacy. Criticizing the regime in public could, however, involve punishment of the critic, but mainly in the form of censorship, exile, or imprisonment. In the eighteenth century, these reactions were sometimes seen as badges of honor rather than as punishments.
Anger also differs from hatred, in that the former rests on the belief that the target person performed a bad action while the latter rests on the belief that he or she has a bad character. According to Aristotle, these two emotions also differ in their desires for action: an angry person wants his offender to suffer, while a person who feels hatred wants the other person to disappear from the face of the earth. Contempt, too, is based on beliefs about character, triggering avoidance behavior. One can propose similar characterizations in terms of triggering beliefs and triggered actions for most of the twenty-odd emotions that can be robustly distinguished from each other. Whereas I believe most of them are found in all societies, the situations that trigger them vary immensely. What is shameful in one soci- ety, such as obesity, can be a source of pride and envy in another.
The main emotions I discuss in this book are negative: fear, anxiety, envy, anger, indignation, resentment, hatred, disappointment, shame, and contempt. Sometimes a reform can trigger a positive emotion of hope that more reforms will be coming (and disappointment when they are not). I shall discuss such episodes in Volume 3. Below I shall also point to some paradoxical emotional reactions, as when the discovery that a fear was groundless triggered anger rather than relief, and that a reform made people more dissatisfied rather than less. Moreover, I shall discuss mechanisms by which the non-emotional state of hunger was transmuted into anger.
For the ancient moralists, reason was the antonym of passion (or emotion). They never defined, however, what they meant by reason. Did they refer to any dispassionate motivation, or did they also require reason to be disinterested? Modern writers, notably the French moralists of the seventeenth century and the authors of The Federalist Papers, were more explicit. Instead of the dyad reason-passion, they proposed a triad: reason-interest-passion. Reason and interest are both dispassionate, but only reason is also disinterested. As I shall use this triad of motivations to characterize the agents of the ancien régime, I need to say a bit more about each of them.
Reason, as I understand it here, is the rational pursuit of the long-term public good. It is not merely a matter of efficiency or raison d’état, at least if one holds (as I do) that social justice is part of the conception of the pub- lic good. I shall devote considerable space to the question whether officials from the king downward were concerned with the welfare of the French and not merely with the glory of France (and their own).
Interest is the pursuit of the good of a proper subset of society, be it a group or an individual. In this book, I consider mainly material interests, while noting that an interest in salvation could also be motivating. As many moralists have observed, the pursuit of long-term interest may mimic or simulate the pursuit of the public interest, since people are sometimes willing to cooperate in situations that invite free-riding if—but only if—they expect that others will reciprocate. An example from Hume shows the difficulty:
Your corn is ripe [in August]; mine will be so [in September]. It is profit- able for us both, that I should labour with you [in August], and that you should aid me [in September]. I have no kindness for you, and know you have as little for me. I will not, therefore, take any pains upon your account; and should I labour with you upon my own account, in expectation of a return, I know I should be disappointed, and that I should in vain depend upon your gratitude. Here then I leave you to labour alone: You treat me in the same manner. The seasons change; and both of us lose our harvests for want of mutual confidence and security.
There exists, however, a solution in terms of long-term self-interest that, in terms of the example, can be spelled out as follows: I have an incentive to cooperate with you in August 2020, because I know that your desire to receive my cooperation in August 2021 will make you keep your promise to help me in September 2020. When fighting factions of magistrates or of nobles kept a temporary truce during the harvest, they may have been acting under “the shadow of the future.” While long-term thinking can thus mimic reason’s efficiency dimension, it cannot mimic the social justice dimension in the case of orphans, the elderly, the disabled, or the very poor—those who cannot reciprocate for any assistance they might receive. Even when reciprocation would have been feasible— one peasant abstaining from denouncing his neighbors to the tax collector in the expectation that they will do the same—the short time horizon induced by acute material scarcity could make denunciation rational for self-interested agents. Moreover, nobody could be certain that their neighbor would abide by the tacit contract, even if all might have been willing to do so were they confident about the others.
Passion, or emotion, is the pursuit of goals motivated by episodic or standing emotions, some of which I have enumerated. Often, an emotion triggers only an action tendency that never leads to a full-blown action, because the tendency is counterbalanced by another emotion, by self- interest, by social norms, or by moral norms. The action tendency may also remain unfulfilled because the emotion runs out of steam (many emotions have a “short half-life”). Some repressive actions by the monarchy, whether against the peasantry or against subversive writers, subsided after a short while. Sometimes, the government counted on this mechanism, when it imposed an edict by force on the first day of the judicial vacations, hoping that tempers would cool over the summer.
Above, I have followed a tradition initiated by Aristotle, which focuses on the effects of emotions on action. I shall also pay attention to another tradition, initiated by Seneca and continued by La Rochefoucauld, which focuses on the effects of emotion on other mental states, which may or may not then become precursors of action. Seneca wrote, “Those whom they injure, they also hate.” A French proverb says, “Who has offended can never forgive.” These and many other reactions are due to the insidious operation of amour-propre, self-love, or egocentricity. An institution to which I belong must be an important one since I belong to it. (The magistrates in the parlements were massively subject to this institutional pride- fulness.) My opinion must be correct since it is mine. While amour-propre is universal—nobody likes to be contradicted—the nobility of the ancien régime exhibited it to an unusual degree, as demonstrated and denounced by Pascal, La Rochefoucauld, Pierre Nicole, and La Bruyère. In later chapters I provide cases in which disagreements were taken as insults, with a chilling effect on discussions.
There are variations in intensity both within and across emotions. Envy has to be stronger when it causes an agent to harm himself in order to harm another person (“black envy”) than when it merely causes him to harm the other at no cost to himself (“white envy”). As noted, for a given offense the emotion aroused in the person who has been offended is stronger—and induces a more severe punishment—than the emotion the same action arouses in a third-party observer. It is more difficult to compare the intensity of different emotions when they are triggered by different actions. I suggest, however, that by and large, emotions of interaction are stronger, and induce more radical behavior, than emotions of com- parison, just as “having power over someone” is more consequential than “having more power than someone.” In one sense, these comparisons are meaningless. “Having a bit of power over someone” can obviously be less consequential than “having a lot more power than someone.” I shall not try to address this question in the abstract, but only make the factual claim that in the relations between the bourgeoisie and the nobility in the ancien régime, the bourgeoisie’s resentment of contempt was stronger than the envy of privileges that Tocqueville, for instance, emphasized. Face-to-face contempt leaves deeper marks than “envy at a distance.” For the nobles, having power over their tenants, in the form of seigneurial justice, was more important than enjoying tax exemptions. Within the comparison- based emotions, horizontal jealousies among towns and provinces seem to have mattered more than vertical envy between classes.
I have been writing as if the relation between a cognitive antecedent and the triggering of an emotion, as well as the relation between the emotion and the subsequent action, are always one-to-one. I believe that another person deliberately hurt me, I get angry, and I retaliate. While this pattern is frequently observed, the relations can also be one-many. Consider for instance Tocqueville’s claim that the tax exemptions of the nobles generated envy in the bourgeoisie. These might also, however, have generated anger towards the king for granting and maintaining these exemptions, just as a child that is not offered the ice cream her sister received may react to the injustice of their parents rather than against the better fortune of her sibling. To be sure, the offended party might experience both emotions— anger towards one agent and envy towards another. In 1789 fear of brigands co-existed with anger towards those who were believed to have sent them on their way. In other cases, people might experience two emotions towards the same agent. Thus if an autocratic ruler enacts strong repressive measures, he might induce both fear and hatred in the public, the former triggering compliance and the latter rebellion. The net effect can go either way. Commenting on the persecution of heretics under Henry VIII, Hume writes that “those severe executions which in another disposition of men’s minds, would have sufficed to suppress [the new doctrine], now served only to diffuse it the more among the people, and to inspire them with horror against the unrelenting persecutors.” As we shall see, the persecution of Protestants under Louis XIV illustrates the same effect.
Mechanisms: Interaction
Explanations of social phenomena need micro-foundations. In the previous section I have attempted to sketch what they might be. However, even if per impossibile we could decipher the motivations and beliefs of all individuals, we would need to supplement the micro-foundations with the interaction mechanisms that explain the workings of the social system at a larger scale. In this book, the two main ways in which actions by individuals coalesce to yield outcomes that could not have been brought about by any one of them are social movements and the activities of formal decision-making bodies. As shown by the relation between the crowds in Paris and the parlement, the two processes can influence or shape each other.
The dynamics of social movements depend not on any given motivation of the participants, but on the motivational mix. It also depends on a range of factual beliefs. Since these movements are forms of collective action, they raise the free-rider problem: why do participants choose to engage in risky or costly actions when the personal benefits from their participation is, in most cases, vanishingly small? Why not stay on the sidelines and let others take the risks or incur the costs? If everybody thinks along those lines, however, they will all stay on the sidelines. As we shall see shortly, the free-rider problem is not the only obstacle to collective action.
Some individuals may perhaps initiate or join a social movement because they hope to emerge as leaders if it succeeds, but such cases, if they exist at all, seem rare. It is safe to say that in the ancien régime nobody joined a social movement to get a private good of this kind. People were mobilized for a collective good—lower prices, higher wages, abolition of unjust institutions or practices—from which all would benefit, whether they were leaders, followers, or bystanders. The question then, to repeat, is why not everybody were bystanders (a somewhat incongruous idea).
Since material interest is always directed towards private goods or the avoidance of private harms, it might seem that we can rule it out as a motivation. Modern theories of collective action allow for it, however, in the form of selective benefits to participants or selective punishment of non- participants. To explain why workers join trade unions, for instance, one can cite the facts that some of the union dues go to fund summer camps for members and that non-members will be denied the wage increases negotiated by the union. Once again, I do not see anything like this happening in the ancien régime. In any case, the movement that leads to the formation of a union or a comparable organization cannot itself rely on these motivations. Also, social movements in the ancien régime did not lead to durable organizations or associations: they would have been crushed.
It might seem more accurate to say that the social movements were ephemeral, either crushed or dissolved when their demands were met, yet this perspective is also somewhat anachronistic. Most movements were crushed, and few demands were met, but the movements had a nuisance value that caused authorities and property-owners to try not to trigger them. In preindustrial England, urban food riots caused by the high prices of bread invariably ended in failure—producing nothing but “a few ruined mills and victims on the gallows,” as the historian of these movements writes. Yet by virtue of their nuisance value the rebellions had a long- term success in making the government and the propertied classes behave more moderately than they would have done otherwise. I believe the same mechanism was at work in preindustrial France. In times of hunger, the privileged might provide relief to the poor, not out of humanitarian concerns, but for fear of riots.
I am not implying that the participants in these movements believed that their cause was hopeless. In some cases, they may not have cared about the prospects of success, since they “had only one death to die.” In other cases, the movement was sustained by the wishful belief that the king would respond to their protestations by abolishing the abuses. In still other cases, they may at least have counted on personal impunity, by the mechanism of safety in numbers (see below). Also, in some cases the risk of punishment for participation may have been dominated by the risk of ostracism for non-participation (see below).
On this background, let me sketch a stylized “snowball model” of how social movements may originate, develop, or perhaps fail. Its purpose is not to provide explanations of actual movements, but to offer some of the nuts and bolts that can enter into such explanations.
I shall assume that at least some participants are moved by their perception of a public good—abolition of abuses, social justice, reducing high prices created by speculators and hoarders—that they want to bring about. I do not imply that they are moved by reason, defined earlier as a disinterested and dispassionate concern for the public good. On the contrary, their belief that the status quo is unjust triggers anger, which induces a higher propensity for risk-taking behavior, as explained earlier. By virtue of the strength of their motivation these individuals act as first movers. They may be saints, heroes, or just slightly mad. As Tocqueville noted, “the same energy that impels a man to rebel in a violent way against a common error almost always carrie[s] him beyond the bounds of reason.”
To pursue this line of argument, I shall rely on a typology of norms that I have developed elsewhere [references], distinguishing among moral norms, quasi-moral norms, and social norms. Moral norms are unconditional, in the sense that their force does not depend on what other people do. This is the motivation of first movers. Quasi-moral norms are conditional, in the sense that their force depends on the agent observing what others are doing. Social norms, too, are conditional, in the sense that their force depends on the agent knowing that she is being observed by others. Apart from the first movers, who follow moral norms, most other participants are motivated by quasi-moral and social norms, or, in the words of Yves-Marie Bercé, by “solidarity and fear.”
The Madisonian caveats stated above obviously apply to this typology. In many cases we cannot tell what motivates the participants in collective action. In fact, the agents themselves may not always have had a clear idea, but simply gathered “weapons in hand, without knowing why.” In a crowded situation where everyone is both observing and being observed, it may be impossible or meaningless—there may not be a “fact of the matter”—to distinguish quasi-moral from social norms. Yet one fact seems clear: solidarity cannot be the sole motivation—what would it mean to feel solidarity with other agents also motivated only by solidarity? Some of the agents towards whom one shows solidarity must have initiated or joined the movement for other reasons. …
… Turning now to the role of beliefs in social movements, consider first the beliefs that potential participants have about each other. In the analysis above, I made the simplifying assumption that (apart from first movers) individuals will join a movement on the basis of their observation of what others have done (or are doing). However, their beliefs about what others will do may be just as important. In a state of pluralistic ignorance—a phenomenon first conceptualized by Tocqueville—each person may believe, falsely, that she is the only one (or one of the few) to be willing to join a movement. As a result, all stay home and nothing happens. The fact that nobody turns out confirms their (false) belief. By the nature of the case, it is hard to tell how often pluralistic ignorance prevents people from rebelling, but I believe it must be a frequent occurrence, not least because the authorities often work to keep people ignorant. In an early intuition of pluralistic ignorance, Seneca wrote, “A proposal was once made in the senate to distinguish slaves from free men by their dress; it then became apparent how great would be the impending danger if our slaves should begin to count our number” (On Clemency XXIV; my italics). In the ancien régime, as in autocracies everywhere, spontaneous assemblies above a certain size were forbidden.
Beliefs in the form of rumors were a vital piece in the machinery of social movements. The object of the rumor could be the hoarding of grain, the approach of brigands or of soldiers, an impend- ing tax reform that would justify the refusal to pay taxes, the kidnapping of children, or the deliberate starving of the people to reduce the size of the population. The rumors might be circulated for their entertainment value, as one may watch horror movies without believing in their reality. Even when they were taken seriously and used as premises for action, their source might be a hawker offering sensationalist rumors to attract buyers, an egocentric trying to make himself important, or just somebody trying to stir up trouble. Also, as noted earlier, the step from accusation to conviction might be short. Many hesitated to express disbelief in the rumor, lest they be accused of cowardice or complicity with the authorities. Sometimes, in Lefebvre’s succinct formulation, “the people scared itself ” (se fai- sait peur à lui-même), as in an episode from 1703 when “the tocsin would sound; each village would send a runner to neighboring villages and ask for their help; detachments arriving to help would be taken for enemies and, without further ado one would announce that the harm is done.”
The propagation and especially the magnification of rumors remain ill- understood. Hans Christian Andersen’s story in which one feather turns into five hens in the telling and retelling does not exaggerate the multiply- ing effect in the transmission of rumor, but the mechanisms are hard to nail down. Montaigne, who was both an observer of the devastating effects of rumor during the wars of religion and an acute psychologist, offered what may have been the first analysis, relying on the tendency of the purveyors of rumors to fill in gaps in the narrative:
The distance is greater from nothing to the minutest thing than it is from the minutest thing to the biggest. Now when the first people who drank their fill from the original oddity come to spread their tale abroad, they can tell by the opposition which they arouse what it is that others find it difficult to accept; they then stop up the chinks with some false piece of oakum. At first the individual error creates the public one: then, in its turn, the public error creates the individual one. And so it passes from hand to hand, the whole fabric is padded out and reshaped, so that the most far-off witness is better informed about it than the closest one, and the last to be told more convinced than the first. It is a natural progression. For whoever believes anything reckons that it is a work of charity to convince someone else of it; and to do this he is not afraid to add, out of his own invention, whatever his story needs to overcome the resistance.
A rough impression is that rumors inspired by anger or fear were more frequent than those inspired by hope. The object of anger-inspired rumors could be some flagrant injustice, such as hoarding of grain to raise prices and starve the people. The object of fear-inspired rumors could be the arrival of brigands, as in the Great Fear of 1789. In that episode, fear of brigands co-existed with anger towards those who had hired or commissioned them. The object of hope-inspired rumors was usually an impending alleviation of the burdens of the people. My impression— again a very rough one—is that hope-inspired rumors were for the most part pipe-dreams and not used as premises for action, an exception being the tendency to refuse to pay taxes that people thought were about to be abolished anyway. Anger and fear were more likely to inspire rumors that could trigger violent action.
I turn now to the second interaction mechanism at work in the ancien régime, the activities of formal decision-making bodies. These include the meetings of the upper clergy every five years, the irregular meetings of the Estates-General, the regular meetings every year, two years, or three years of the provincial Estates, and the ongoing activities of the magistrates in the parlements, interrupted only by the judicial vacations. At the local level, there were innumerable village assemblies. Concerning the parlements, I limit myself for now to their political activities, that is to their protestations (remonstrances) to royal edicts. Their judicial functions will concern me [in a later chapter].
Strictly speaking, almost by definition no bodies in an absolute monarchy could possess hard decision-making competence, in the sense of being capable of imposing decisions that could not be overruled by the king and that might even constrain his behavior. They might not even be masters of their own decision-making rules, as shown by the meta-decision by the finance minister Bertin in 1762 that henceforward decisions by the unruly Estates of Brittany would no longer require the unanimous agreement of all three estates, but that two of them would be able to outvote a third. Yet the deliberating bodies of the ancien régime were far from being sham bodies, such as parliaments under most Communist regimes.
As in many other cases of collective decision-making, deliberating bodies in the ancien régime proceeded by voting, arguing, and bargaining.
All these bodies, or sections of them, voted on proposals in votes that could be very close. In the assembly of the clergy in 1755, “the bishops were divided over the issue whether a refusal to accept the bull Unigenitus was a mortal sin or merely a sin in a serious matter. 16 bishops were of the first opinion and 17 of the second.” Although the parlements aimed for unanimity in their decisions, consensus was achieved only because “those who had opted for the least popular proposal had an opportunity to adhere to one of the others, a process that was repeated until unanimity had been achieved.” As I document in [a later chapter], the initial votes could be close. In one dramatic decision in 1764, the parlement of Brittany refused by 13 votes to 12 to register without protestations (purement et simplement) a royal declaration. However, even though we may have the numbers of votes pro and con, we rarely have the names of the voters, as strict adher- ence to methodological individualism would require.
The Estates-General, unlike the provincial Estates, never adopted a system of voting by estates, by which each of them would cast a vote and then reach a decision by majority or unanimity. Yet within each estate, the votes of the various provincial deputations were aggregated by majority rule to yield an opinion. Strangely (for us), the vote of an estate figured only as an output of its internal deliberation, never as an input to a vote by the Estates-General. In this respect, the Estates-General differed from other nested political systems, such as the Continental Congress where the vote of a state deputation was both the output of its internal votes and an input to the voting of the larger body. One reason why the Estates-General never developed voting rules by estates was the indeterminacy of voting within each estate, since there were competing proposals for how to define the voting subunits of each estate.
From the records available to us (that is, known to me) it is hard to assess the importance of arguing (persuasion without the use of threats or promises). Georges Picot seems too sanguine when he refers, with regard to the Estates-General of 1560, to an “[i]nevitable effect of deliberation among men: how bad their plans and how terrible their passions might be, it substitutes moral force for material force and the power of reason- ing for violence.” One can assume, nevertheless, that as in assemblies everywhere, arguing sometimes led to the elimination of a proposal when shown to be Pareto-inferior (not preferred by anyone) to another proposal. Also, as in assemblies everywhere, some participants must have been sensitive to arguments ad hominem, in the sense in which Locke used that expression: “to press a man with consequences drawn from his own principles or concessions,” to which we may add consequences drawn from his past behavior. These remarks apply equally to the deliberations of the parlements.
With respect to bargaining, we can go beyond these generalities. In the Estates-General and in the provincial Estates, the three estates negotiated not only with each other, but also with the royal government, demanding concessions on a number of issues in exchange for consent to taxation. In fact, they could bargain with each other about which demands to make when bargaining with the king, much as industry unions bargain with each other about which collective demands to make when bargaining with the employers’ union. The main issue was always the allocation of the tax burden. Because the Estates-General were an assembly of provinces as much as an assembly of estates, bargaining over taxation also took place among the former. Overall, the process was skewed in favor of the kings, by virtue of their stronger bargaining power. They could dismiss the deputies, close the doors to their meeting rooms, or refuse to pay their costs.
Even more important, when the kings made concessions to the Estates-General, the latter had no permanent machinery that could enforce a tax strike if the kings did not keep their promises, which they virtually never did. “One cannot find in our whole history a single king who scrupulously limited taxation to what the Estates-General had authorized.” The provincial Estates were more successful in bargaining with the government over their “free gift,” although it is hard to nail down the exact sources of their bargaining power. Also, the clergy could hold the kings to their promises by virtue of the fact that its assemblies met regularly, at five-year intervals, and could refuse the free gift if the king had not kept the promises he made to the previous assembly. As in the example adapted from Hume, the promises were enforceable because of repeated and open-ended interactions at fixed dates relatively close in time.
In some cases the government negotiated with the parlement of Paris before issuing its edicts, to ensure that they would be regis- tered without remonstrances, but it is not clear whether this was a regular or an occasional process. Since the Premier Président, unlike the other magistrates, was directly beholden to the king, he had the opportunity to act as a mediator by informing the king about what was acceptable to the parlement and vice versa. These remarks do not apply, however, to the period between 1673 and 1715, when the parlements were largely reduced to rubber-stamping bodies, or to the period between 1771 and 1774 when they briefly lost many of their powers. They apply to some extent, but not fully, to the period between 1715 and 1750, when the parlements could make remonstrances, but not publish them. The reemergence of the courts after 1750 as powerful political bodies occurred after the regular publication of their protestations against royal edicts created a two-way interaction between courts and crowds, mainly in Paris. In the struggles between the king and the parlements, publication now conferred a valuable bargaining tool on the latter.
The resistance of the parlements to the royal government was limited by collective action problems within and between the courts. Before the magistrates in a parlement were sent into exile, they received lettres de cachet in the early hours of the morning and were given the choice between compliance and resistance. Since each would be unaware of the choices of the others, they might comply even though they would have resisted if assured that others would resist too. Another collective action problem—known by game theorists as “The Battle of the Sexes”—arose when problems of préséance made the magistrates reluctant to confer the presidency to one among themselves in cases where the normal president recused himself. The same situation could arise among the parlements, when the provincial courts were reluctant to let Paris take the lead in a common front (the union des classes), against the king.
I conclude this Introduction by noting that two-way interaction between courts and crowds was mainly an urban phenomenon. Social movements in the countryside must have shaped the deliberations of provincial parlements and Estates, but with some exceptions these bodies did not use public opinion as a resonance board in their struggles with the royal administration. A two-way interaction between Paris and the countryside was not established until the summer of 1789. By contrast, in the last forty years before the Revolution the publicly stated opposition of the parlement of Paris to Louis XV and to Louis XVI contributed decisively to the undermining of royal authority. In one of the many counterfactual questions that arise in any discussion of the demise of the ancien régime, we can ask whether it might have survived if Louis XVI had not resurrected the parlements when he succeeded to the throne in 1774. [END]
The Source:
Jon Elster*, France before 1789: The Unraveling of an Absolutist Regime, Princeton University Press 2020
*Social Science Files subscriber and regular reader since March 2022
Evolutions of social order from the earliest humans to the present day and future machine age.
As a subscriber you have full access to the searchable Social Science Archives.
.. providing quality tools for Social Science since 2022 ..