The Moral Economy is a wonderful read for all those unfamiliar with the experimental tradition in economics. The book succeeds in introducing a readership of non-economists to state of the art research in experimental economics. Through a detailed discussion of the experiments conducted by Bowles and his colleagues, one gets a real sense of this new breed of economists as they tinker with the experimental set-up and make subjects across the globe play by the new experimental rules of the game. On top of that, Bowles has found an interesting—though not always convincing—way to embed the experimental results in a broader narrative on politics and society. It is because of the clarity of his narrative that it is possible to provide a systematic overview of the book’s content and critically reflect on a number of broader issues: the experimental style of reasoning, the relation between social scientific disciplines and the link between politics and expertise.
In the first two chapters, Bowles outlines how modern political philosophers and political economists have contributed to a perspective on politics and citizenship dubbed “the constitution for knaves.” The central feature of that constitution is the conception of individuals as fully self-interested and amoral creatures. The authors responsible for creating the constitution already knew this was too simple a view of man, but they were convinced that individuals should be governed on the assumption that they only cared for their own interests. In 1742, Scottish philosopher David Hume put this political philosophy of action in canonical form: “in contriving any system of government […] every man ought to be supposed to be a knave and to have no other end, in all his actions, than his private interest” [cited on 16]. In terms of the policy-making process, the “Humean legislator” hence favors the use of material incentives to steer the citizen-as-homo-economicus in a direction that is favorable to all. Although its construction takes us back to the 18th century and earlier, the constitution for knaves is not just a historical curiosum. Bowles takes it to be a veritable policy paradigm that contemporary economists wittingly or unwittingly support: “Even today many in my discipline combine a professed indifference to the nature of individual preferences with excessive confidence in the ability of clever incentives to induce even an entirely amoral and self-interested citizenry to act in the public interest” [36]. In light of its current predominance, Bowles’ explicit objective is to advance a new policy paradigm that creates synergy between conventional instruments—namely, incentives and punishments—and the social preferences of citizens [6-7].
On our way to that new policy paradigm, the Humean legislator is replaced by a sophisticated “Aristotelian” one. In chapters three to five, this imaginary figure visits economists around the world, takes a look at their experimental procedures and results and sees what (s)he can gain from them for the art of wielding power. The experiments vary slightly but all involve the transfer of money and information: one subject shares a given sum with another subject who can decline or accept the offer; all subjects have to contribute to a common pot from which they receive a predetermined share back into their pockets; they either transfer the money in a single instance or in a sequence of moves; and there might (not) be a third party observing their actions with the option of punishing the players. In general, the experimental results pose a challenge to the idea that people only—or even predominantly—act in a self-interested and amoral way. More specifically, they challenge an idea many economists subscribe to, namely that “incentives and morals are additively separable” [22]. In less technical terms, that separability assumption holds that the introduction of a material incentive (such as a fine) will do nothing—positively or negatively—with the individual’s pre-existing moral predispositions. If we grant, however, that people are influenced by situational cues, then fining a certain course of action might signal to them that they can consider that action in purely economic terms and neglect any moral views they might simultaneously hold. That is, they take the fine as a price that is paid for transgressing social norms—as demonstrated by the famous but somewhat worn example of the Haifa day care center where the number of parents that picked up their children too late increased from the moment latecomers were fined [4-6].
After traveling the globe looking for economic insights, the Aristotelian legislator comes to the conclusion that the relationship between material incentives and behavioral outcomes is not as simple and straightforward as his Humean colleague assumed. But where does that take us? Chapters six and seven explore the key dilemma faced by the legislator and outline his or her political mandate. To articulate the legislator’s dilemma, Bowles turns to the question whether it would be possible, in principle, to design a mechanism that allocates resources efficiently among fully self-interested individuals without forcing them to participate. After decades of tinkering with equations, economic theorists have come to the conclusion that no such mechanism is possible. In light of these impossibility results, the best thing a legislator can do is to ensure that incentives and social preferences work in tandem. Hence the legislator’s mandate is to find policy measures that crowd in these preferences instead of crowding them out.
Laboratory games
As the experiments are already published, the philosophical texts all have a canonical status and there is no claim to originality in their interpretation, the evaluative part of my review focuses on the way Bowles embeds the experimental results into a narrative that is meant to speak to a broader audience of social scientists and policy-makers. I will first discuss the experimental style of reasoning that has gained in status among economists over the past decades. To start with, the games that subjects are required to play are fairly odd to anyone unfamiliar with economic laboratory experiments. These experiments have names such as “public good games,” “ultimatum games,” “dictator games” and “third-party punishment games,” and do not refer to any recognizable real-world phenomena. One way to look at these experiments is to say they were not meant to do so in the first place. Instead, as Lee [2016] Footnote 1 demonstrates, the experimental set up reveals that laboratory experiments grew out of a very peculiar interest in the theory of economic mechanisms—not in economic reality per se. Put differently, laboratory experiments started to gain (some) legitimacy in the 1980s in a field that was quite hostile to them by bidding for the support of mechanism design theorists—a small but high-status group of mathematically trained economists at the time. They did so by mimicking the language in which mechanism design theorists expressed themselves. In a context of major budget cuts for the sciences under Reagan, mechanism design theorists returned the favor by emphasizing that the laboratory experiments might be taken as evidence that their theoretical endeavors were more than just that. That legacy is still visible in present-day laboratory experiments. First, the central place of mechanism design theory in chapter six shows a persistent interest in the matter. Second, the conceptual proximity is visible in the endless but often quite minor variations in the rules that determine the exchange of information and resources (mechanism) in a range of controlled settings (environment). As a nod to relevance, finally, we should somehow consider the experimental subjects as a “stand-in” for real actors or conflicts present in society at large: there are “employer-employee relationships” without recognizable employers or employees; there are “investors” who have to deal with “trustees”; and there are “citizens” who are confronted with a central allocation mechanism that is meant to represent the state.
Of course, economists know and acknowledge that external validity is a thorny epistemological issue in (economic) laboratory experiments [70-71]. But assuming that the imaginary legislator does not visit the laboratories to become a fully-fledged experimental economist it seems that (s)he has to constantly suppress that issue in order for these experiments to make any real-world sense. Although economists sometimes test the validity of their results beyond the confines of the laboratory, in many cases the legislator simply has to treat the results as insightful for real-life situations and, more importantly, as necessary for ruling wisely. But when the experiments become more artificial—the economist might speak of a “clever modification” here [94]—the question as to what they actually mean beyond the confines of the laboratory presents itself more and more forcefully. Take a German student who is now given the option—the clever modification—to fine another German student when (s)he considers that the latter was not generous enough in returning some of the money [87-88]. Whatever the conclusions one wishes to draw from experiments like these—I am sometimes tempted to stick to: “if you play a very strange game with (some) German students then this is what (some) German students do”—this seems to be very far removed from discovering man as he really is.
Experimental economics for sociologists?
Bowles sees the trespassing of disciplinary boundaries as part of his own educational trajectory and current scientific predisposition [xiii]. And indeed, he certainly appreciates the work of sociologists, anthropologists, philosophers and psychologists to a far greater extent than others in the field (though it must be added that they are predominantly evoked when they support the experimental findings or participate in experimental research). The peculiar path dependencies in economic laboratory experimenting discussed in the previous section lead to a second question: is this a useful and promising research strategy for studying politics and organizations that sociologists should pay attention to? Well, everything hinges on the idea that we can understand these phenomena better by simulating organizations and state institutions than by studying actual political and organizational behavior. Sometimes, the conclusions that emerge from these simulations are not that revolutionary, although Bowles is the first to acknowledge that economists are quite good at bringing “new” ideas to the fore that are already common sense elsewhere—like the idea that our identities matter [192]. Even the idea that playing on self-interest through incentives might lead to more self-interested behavior is not one to baffle a sociologist who never subscribed to the separability assumption in the first place. But there is also much to learn.
What is already more promising in terms of crossing disciplinary boundaries is Bowles’ transgression of a principle that has long enabled economists to keep their distance from sociology and psychology, namely that the preferences of individuals should be taken as given. As long as this principle is accepted, questions concerning the formation of our cultural and political dispositions and questions about the ways in which goods come to have value for us are simply irrelevant to economists [Bourdieu 1984; Hochschild 2016; Beckert and Aspers 2011] Footnote 2 . But when preferences are not simply given but (partly) depend on incentives instead—both in a more direct situational sense and in long-term processes of socialization—then it is more difficult to maintain that principle and the boundary-work it makes possible [65]. In doing so, Bowles opens up new terrain to explore the relationship between the products and policies offered to us, and the formation of our social, cultural and political values.
In addition, there is something more substantial to gain even for those who are somewhat skeptical about economic experiments. Because the laboratory is such a restricted setting, the monetary rewards are so explicit and the games so repetitive, economists can pay a lot of attention to articulating the mechanisms that might explain the behavioral responses to incentives [84-103]. These mechanisms all have to do with the information conveyed by attempts to steer human behavior in a different direction: the use of incentives might signal that one party distrusts the other; the appeal to self-interest might lead to the moral disengagement of the actors and thereby increase their self-interestedness; and the sheer attempt to wield power might lead to all kinds of adverse responses because actors feel their autonomy is compromised.
It is worthwhile further pursuing these mechanisms even for those who do not buy into the whole research program. Although they seem to be strange bedfellows—as their methods and vocabulary are rather different—it could be interesting, for instance, to bring Bowles’ insights into the debate on the performativity of economics [Callon 1998; MacKenzie et al., 2007; Pinch and Swedberg 2008] Footnote 3 . First, the explicit focus on political and managerial instruments is a useful corrective to the predominant focus on market devices in this line of work [Hirschman and Berman 2014] Footnote 4 . Moreover, Bowles’ focus on the behavioral assumptions embedded in the techniques that economists advocate to govern individuals and institutions is equally relevant for those who study the political role of economic experts. Finally, the literature on the performativity of economics is full of complexity and nuance [Healy 2017] Footnote 5 but would certainly benefit from a stronger focus on the mechanisms by which economics transforms the world: how do economic theories and devices perform the practices they pertain to depict?
Incentivizing wisely: economics and toolbox politics
The link between economic expertise and policy-making is central to the third and final broader issue. In his book, Bowles embraces a “toolbox” approach to politics in which laboratory experiments give policy-makers the tools to incentivize wisely. Bracketing the issue of external validity, how satisfying is this version of toolbox politics? And how far removed are we from actual policy-making processes? Of course, things already look slightly different when we substitute a real name—say, Trump—for the impersonal Aristotelian legislator. But even in cases where the aversion to expertise is less extreme, the question of what policy-makers could actually do with the “tools” offered in The Moral Economy is not always easy to answer. Let us take a closer look at some of the content of the legislator’s toolbox first: the legislator who tries to design the best subsidy for his or her citizens should find out when they respond best, and “happily adds the best-response functions […] to his tool kit” [59]; Aristotle’s legislator is also happy to see the effects of incentives on reciprocity, generosity and trust as “new additions to his toolbox” [76]; the legislator, moreover, “knows that good-will is important for a well-ordered firm or nation” [80]; and he finds out that employees sometimes challenge the low expectations of employers by working harder, and “adds it to his tool kit” [103]. Although these are all valuable reminders that citizens and employees might respond to policies in a variety of ways, I find it difficult to see anything resembling a toolbox here. More importantly, the emphasis shifts from laboratory experiments to information gathered from newspapers [202-203], surveys [193] and political campaigners [194] as soon as we get closer to real instances of policy-making. Bowles’ own emphasis on the political payoff of the economic experiments leads to a series of uneasy questions: if we need ordinary “factual” knowledge for actual policy-making then why should we take the long detour through the laboratory? How should real-world legislators actually distinguish cases of crowding in and crowding out, and discriminate a vicious from a virtuous cycle? And how much information about citizens do states have to gather—think of the dream of neuro-government [103-107]—in order to avoid the mechanisms of distrust, moral disengagement and compromised autonomy [206-207]?
Perhaps questions like these take us beyond the book’s more abstract purpose. And even pitched at this fairly abstract level, Bowles’ book is a timely intervention in a debate about the roles we play today as employees, citizens, experts, managers and politicians. For in the past three decades, incentives have come to dominate our attempts to wield power over individuals and institutions in diverse domains of economic and political life—from tax, insurance and social policy to public and private sector organizational reforms [Crouch, 2011; Dobbin and Jung, 2010; Fourcade and Khurana, 2013; Grant 2012; Prasad, 2012; Rodgers, 2011: 202-209; Sandel 2012; Streeck, 2014] Footnote 6 . With roots in the 19th century, “incentivization” has developed into a general and self-evident framework for understanding and acting upon human behavior [Dix 2014] Footnote 7 . Though partly a sign of this development, Bowles’ criticism of the simplistic behavioral assumptions that underpin the use of incentives makes The Moral Economy an excellent starting point for problematizing incentivization as a contemporary modality of power.
