Morality and political economy from the vantage point of economics

Abstract

Background

During the majority of the post–World War II period, politics was dominated by disagreement about economics. The emergence of the burgeoning field of political economy arguably partly reflected the longstanding primacy of economic incentives in the political process. In recent years, however, politics has become highly moralized. Deep divisions over values and identity have taken center stage, and disagreements over what is “right” and “wrong” occupy much of the public discussion. These developments raise the question how the academic study of political economy interacts with that of moral psychology.

At the origin of the social sciences, economics, politics and morality were deeply intertwined. Worldly philosophers such as Adam Smith, Jeremy Bentham, Karl Marx, or Alexis de Tocqueville routinely discussed the relationship between morality, economics, and the political process. Then, over time, economics and political science diverged from psychology, and cross-disciplinary exchange about morality became rare (the field of moral psychology in its modern form did not even exist until relatively recently). This divergence reflected differences in interests and methodological approaches. Psychologists largely focused on individual moral reasoning, rather than its interaction with broader social, economic, and political phenomena. Economists, meanwhile, were reluctant to engage with morality partly because they were intrinsically more interested in aggregate phenomena than in individual behavior and, hence, valued the analytical traction that insisting on Homo economicus provided them (moreover, economists desire to engage in ostensibly positive social science; while moral psychology is a positive discipline just like economics is, I have found that economists often implicitly associate the language “morality” with normative reasoning). The lack of sustained interdisciplinary exchange thus reflected at least partly a productive process of specialization.

Yet, more recently, the tide has started turning again. Both fields came to recognize that economics and morality influence each other in a rich bidirectional relationship, in a way that directly impacts political outcomes, too. This partial coming together had at least two—I suspect largely orthogonal—origins. In moral psychology, a new synthesis emerged that connected more strongly to a body of biological and cultural evolutionary theories of the origins of morality (1–4). This wave of work paved the way for bringing moral psychology closer to the “aggregative” social sciences because it embedded individual-level emotions and decision processes in a broader context. As part of this process, moral psychologists came to view morality both as a direct product of and an important determinant of politico-economic outcomes. To paraphrase Haidt (3), moral psychology is not only about what morality is, but also largely about what it does (how it impacts social and economic outcomes) and where it comes from (including the fundamental importance of economic incentives).

In economics and political economy, meanwhile, the behavioral economics and cultural economics movements gained traction. Economics became more micro and empirical in nature, and in the process came to appreciate the insights about individual-level decision making and social contexts that psychologists had amassed. As they have been for many years, economists continue to be reluctant—at least prima facie—to argue over morality. Yet because moral psychology is largely not about arguing over what is right or wrong but, instead, about studying both the origins and the social ramifications of heterogeneity in morality, economists discovered intellectual connections between their own interests and research in moral psychology.

A prime example for the conceptual overlap between moral psychology and economics is the fact—which for outsiders is often surprising—that one of the most foundational ideas in modern moral psychology is partly economic in nature. A rich body of theories posits that morality is ultimately functional and reflects incentives—that it evolved as a form of “psychological and biological police” to enforce prosocial behavior and cooperation, including in economic production and exchange (3–11). It is a particularly stark demonstration of the salience of functionalist thinking in contemporary moral psychology that psychologists sometimes even define the very concept of morality through its socioeconomic function of supporting social and economic cooperation (2, 12). While these ideas about the centrality of the problem of cooperation originated in the evolutionary sciences, rather than in economics, they nonetheless suggest an important role for economic incentives in shaping the structure of morality. After all, if morality evolved to solve a problem (cooperation) that often pertains to economic activity, then as the economic system changes, so should the functional response. Economic theorists such as Tabellini (13) took up these functional-evolutionary ideas and embedded them in mathematical analyses of the coevolution of morality and institutions, and their joint effect on aggregate income.

The relationship between moral psychology and political economy extends in the other direction as well. Moral psychologists argue not only that economics and politics shape morality, but also that morality shapes politico-economic outcomes. According to this perspective, the structure of morality exhibits large heterogeneity across individuals, in ways that shape political views and polarization, including on economic issues such as taxation and redistribution (14, 15). Psychological research in this tradition tends to view morality as a primitive that shapes and structures political views across a large number of policy dimensions. Politico-economic outcomes such as economic inequality and institutional dysfunction are then argued to be determined partly by the distribution of basic moral values in society.

In a nutshell, then, moral psychology posits that what determines economic and political outcomes is partly morality, and that what determines morality is partly economic and political incentives or events.

Economic imperialism?—complementarities in the social sciences

This raises the question: should economists be involved in psychologists’ business of studying the causes and consequences of morality, including in areas that psychologists have analyzed themselves? And might noneconomists (such as researchers in moral psychology or cultural evolution) be interested in what economists have to say? I believe that the answer to both questions is yes, for two reasons. First, viewed narrowly from the perspective of economics alone, it is necessary for some economists to engage in cross-disciplinary translation—to import ideas from neighboring fields and subject them to the methods and tests that economists have come to accept. After all, at this point, the social sciences are sufficiently segregated that most economists cannot be reasonably expected to follow work in moral psychology, also given that economists and psychologists tend to have different views on what constitutes “good” research designs.

More importantly, however, I believe that there is substantial value to economists engaging with moral psychology also from a broader social science perspective. Naturally, given their respective foci, the social sciences exhibit different strengths and weaknesses. Moral psychologists possess unique skills and methods to understand individual decision-making processes. This includes—but is not limited to—beautiful experimental paradigms, eye-opening vignettes, questionnaires that are often considerably richer than those of other social scientists, and an enormous creativity in formulating and structuring hypotheses about human nature.

One of the primary strengths of economics, on the other hand, is arguably its rich and sophisticated toolkit for juggling and rigorously analyzing large amounts of data. This, I believe, allows for a productive partial division of labor in the social sciences. In this vein, economists have recently developed ways to study the effect of economic incentives on morality, and that of morality on economic and political outcomes, in a broad range of settings and using a large set of techniques: historical data, natural language processing, online experiments, field experiments, real-world outcomes, global surveys, and more.

I argue that this constitutes a productive and valuable example of what is sometimes implicitly criticized as “economic imperialism.” Such imperialism tends to draw criticism from the broader social science community when economists work on questions that are in the traditional domain of other social sciences yet are blissfully ignorant of the work done by noneconomists. However, the work summarized subsequently does not seek to replace or reinvent the other social sciences. Rather, it explicitly builds on and engages with the work done by noneconomists. Indeed, the vast majority of the economics papers that I allude to subsequently do not contain truly original ideas or hypotheses—rather, they “only” deploy the economics toolkit to test psychological ideas at scale, or in the wild.

To illustrate these arguments, I now briefly discuss two case studies of how recent empirical economics research has contributed to the integration of moral psychology and political economy. In doing so, my objective is merely to illustrate—to selectively showcase economics research that I am personally interested in and that I believe speaks to the interests of a diverse social science audience. As a result, the case studies are not representative of the literature in any meaningful way—they only cover a small part of the burgeoning field of moral psychology, and they also do not do justice to the growing body of work in economics on the topic. For a (somewhat) more comprehensive review see, for example, Enke (16). For instance, I do not cover the large body of interdisciplinary work on social norms (17) in the political process, much of which engages with topics related to morality. Moreover, my focus on economics research does not imply a claim of economists’ superiority but simply follows from my desire to argue how economics can be useful for moral psychology and adjacent fields.

Case study 1: economic incentives and events shape morality

The broad theory that morality is a functional response to the problem of cooperation generates the key prediction that heterogeneity in moral values, beliefs, and emotions partly reflects heterogeneity in the structure of the economic environment. Economists have leveraged this idea primarily to study variation in morality along the universalism-particularism cleavage. A universalist morality emphasizes equal treatment and relationship-independent values, whereas a particularist morality highlights relationship- or group-specific principles such as loyalty or treating in-group members well. The key difference between universalism and particularism is, hence, not who is “more moral” but, instead, whether one's moral principles are primarily impersonal or tied to certain relationships. Psychologists have long documented large heterogeneity in universalism across individuals and cultures. Where do such differences come from?

Viewed through the lens of economic incentives, universalism may be conducive to success in environments that imply high relative benefits of impersonal, one-shot interactions such as in impersonal markets or large-scale institutions. In contrast, in environments that are characterized by large benefits to intensive, repeated in-group interactions, a particularist morality may be beneficial. Economists have studied this broad idea in three contexts: (1) the role of extended kinship relations, (2) the effects of market exposure, and (3) the impact of local ecology.

Enke (18) studies the effects of cross-cultural variation in the tightness and density of historical kinship networks on the structure of a society's moral system. The idea is that societies with tight extended families structure a large part of social and economic interactions around the family, clan, or lineage, producing high benefits to a particularist morality. Loose kinship systems, on the other hand, go hand in hand with more infrequent and anonymous interactions, rendering a universalist morality productive. Based on data on the structure of historical kin networks in the Ethnographic Atlas (19), Enke (18) shows strong links between kinship tightness and morality, in both historical and contemporary data. Tighter kin networks are associated with an entire moral system of less universalist values, a strong emphasis on personal punishment, and a high cultural salience of external shame (rather than internalized guilt). Schulz et al. (20) and Akbari et al. (21) present related results from different contexts and using different datasets. One interpretation of these results is that morality helpfully matches and supports the prevailing structure of economic and social interactions in society.

A second example that has a closely related flavor is research on the impacts of market exposure on morality (22). Similarly to the case of kinship structures, a prominent argument is that a universalist morality is more beneficial, and thus more likely to emerge, when market transactions among strangers take place. The question of how market exposure leads to medium- or long-run changes in the structure of morality cannot credibly be answered with laboratory experiments, which has lead psychologists and anthropologists to resort to lab-in-the-field experiments with a handful of contemporary small-scale societies [e.g., (23)]. Economists have contributed to this research agenda by devising empirical strategies that do not only rely on the peculiarities of a few small-scale societies. For instance, Enke (24) uses textual data on cultural folklore to construct quantitative indices of the moral values and market exposure of about 1,000 preindustrial ethnolinguistic groups. He finds that a high cultural salience of markets is strongly predictive of a more universalist morality as revealed in the stories that form a society's folklore. Instrumental variables analyses that leverage proximity to trade routes or the degree of ecological diversity as exogenous shifters of market exposure suggest that at least a part of the markets-universalism correlation reflects a causal effect of market exposure. Complementing this historical analysis, Agneman and Chevrot-Bianco (25) and Rustagi (26) present experiments with contemporary societies in which one part of the population was more strongly exposed to markets than another one. For example, Rustagi (26) leverages a natural experiment in Ethiopia, whereby some villages were (somewhat randomly) located in greater proximity to a historical market place. Rustagi conducts experiments with inhabitants of these villages and documents that villagers who reside closer to the market (and hence visit it more often) have developed stronger norms of generalized, impersonal cooperation. My reading of this body of work is that, while none of the suggested causal identification strategies is airtight, the evidence broadly supports the idea that markets foster a universalist morality.

The general idea that market incentives affect morality is also present in economic history contributions that study the determinants of intergroup hatred, though in this work researchers only observe downstream behavior rather than also deep values. Becker and Pascali (27) analyze the labor market incentives that may have fostered anti-Semitism in medieval Germany. They observe that, prior to the emergence of Protestantism, the presence of tight usury bans imposed by the Catholic Church effectively forced a productive division of labor, according to which Jews specialized in money lending and Christians in other occupations. However, with the advent of Lutheranism, Protestant Christians were allowed to engage in money lending, and hence had economic “incentives” to denounce and persecute Jews to reduce unwanted competition. Becker and Pascali (27) use an econometric differences-in-differences identification strategy to document that the Protestant Reformation indeed increased anti-Semitism, suggesting that labor market competition affects moral views. Jha (28) presents related evidence by showing how opportunities for mutually beneficial trade between medieval Hindus and Muslims—again, a form of market incentive—shaped ethnic tolerance.

A third category of papers that study the effect of economic incentives on morality focuses on the effects of ecology. The main idea in this literature is that some ecologies are conducive to intensive local cooperation—and should hence give rise to a particularist morality—while others are not [e.g., (29)]. Raz (30) studies this idea in the context of 19th century U.S. settlers. Local soil differs widely in its heterogeneity. When soil is very homogeneous, settlers can easily share best practices with their neighbors, and such opportunities for social learning may induce people to develop strong communal ties and morality. In contrast, when the soil is heterogeneous, such that even relatively immediate neighbors cannot learn much from each other, close-knit communities are less likely to emerge and a universalist morality may develop. Raz (30) documents that this is the case—areas with less heterogeneous soil historically developed stronger communities and still have more particularist moral values today.

Related arguments are pursued by Grosjean (31), Le Rossignol and Lowes (32), and Cao et al. (33). They study the effect of historical subsistence modes—which is largely determined by ecology—on morality and conflict. In particular, building on the “culture of honor” theory from social psychology, they investigate whether societies that largely relied on pastoralism (herding) in the past developed stronger in-group vs. out-group attitudes, a morality conducive to violence, and ultimately see more conflict today. By linking different historical and contemporaneous datasets, they document that this is indeed the case. For instance, whether or not an ethnic group's distant ancestors practiced herding is strongly predictive of the frequency of conflict, including of civil conflict, today. This suggests that the moral proclivities that are suggested by the “culture of honor” matter not only for small-scale aggressions, but also for economically meaningful conflict events such as civil wars.

A broad methodological takeaway that emerges from this entire body of work—on kinship systems, markets, and ecology—is that economists were able to make valuable contributions to a broader social science research agenda because of their entrepreneurial and technical skills in identifying and navigating data structures that are more complex than canonical laboratory experiments or surveys. Yet, this also highlights the strong complementarities that exist across the social sciences—almost all or even all of the empirical exercises summarized previously do not rely on original theorizing, but rather were explicitly motivated by theoretical frameworks and evidence in moral and evolutionary psychology, cultural evolution, and cultural anthropology.

Case study 2: morality shapes political behavior

As noted previously, politics has become more moralized. Yet, how can we as social scientists conceptualize and measure the heterogeneity in values that seems to underlie political conflict? The idea that values along the universalism-particularism cleavage affect the formation of policy views and voting behavior has been articulated by various moral and political psychologists [e.g., (3, 14, 34)]. This idea has intuitive appeal especially in the current political climate in which many policy issues that are intimately linked to the universalism-particularism divide appear very salient. This includes discussions about immigration, affirmative action and minority rights, environmental protection, national and transnational redistribution, and more. The evidence psychologists have brought to bear on this issue consists of various surveys and lab experimental games that document a link between universalism and left-wing ideology.

How could economists contribute to this discussion? Again, the answer is largely data and empirical techniques. Enke (35) leverages a combination of text analyses of political speech and large-scale psychological questionnaires to study how closely the “demand” for different types of morality by voters matches the “supply” of morality by politicians. He quantifies the moral appeal of recent presidential candidates based on campaign speeches, and then documents in a cross-county analysis that candidates garner higher vote shares precisely in those U.S. counties in which citizens’ moral values match the politician's espoused values more closely. Moral values not only explain across-party differences (universalist counties voting Democratic), but also shape how voters evaluate candidates from the same party.

While psychological questionnaires have the key advantage that they often effectively tap into people's deep moral values, they are also sometimes criticized on various grounds. First, by design, survey responses are not incentivized and may hence reflect “cheap talk,” rather than costly and real ecological decisions. Second, it is not always immediately obvious which precise concept any given moral psychology questionnaire attempts to elicit—questions can sometimes be vague, ambiguous, or even conflated with the political outcome that one ultimately wants to explain. Thus, a reasonable question is whether there is also evidence for a link between universalism and left-wing voting from natural field contexts. Enke et al. (36) take up this challenge by linking large-scale donations data and voting records. Based on a large dataset on millions of donations through an educational crowd-funding platform, they estimate each U.S. Congressional District's universalism in giving. The main idea is to analyze to whom (rather than how much) each district gives—primarily to schools that are relatively close (say, in the same state) or equally to schools that are located far away. It turns out that districts exhibit large heterogeneity on this dimension. Some districts donate almost exclusively to relatively close schools, while others’ giving is entirely invariant to distance. Most importantly, this donation-based measure of universalism is strongly correlated with local vote shares—districts that donate in a more universalist fashion tend to be more Democratic. While this result is purely correlational, the explanatory power of universalism for vote shares is very large, and much larger than that of economic variables such as local per-capita income, educational attainment, or unemployment rates.

Given these results, an immediate question is whether the link between universalism and left-wing voting and policy views generalizes to other countries. Until recently, a main impediment to answering this question was a scarcity of representative high-quality data. Enke et al. (37) and Cappelen et al. (38) implement multicountry studies to remedy this shortcoming. For instance, in one data collection through the professional infrastructure of the Gallup World Poll, Cappelen et al. (38) collect data on universalism from representative samples in each of 60 countries, for a total of 65,000 respondents. They measure universalism using hypothetical money allocation tasks such as “How would you split $1,000 between a friend and a stranger?” These data reveal that, in essentially all “Western” countries, universalism is strongly predictive of left-wing policy views. Importantly from the perspective of economists, universalism descriptively explains not only policy views on social or cultural issues, but also core economic topics related to taxation, redistribution, health care, and foreign aid.

While the case studies summarized previously mostly focus on the universalism-particularism divide, economists have also made various contributions to the empirical study of distributive justice—how subjective fairness views shape policy preferences on redistribution. A key characteristic of this body of work is the incentivized measurement of fairness views through large-scale experiments that often span representative samples in multiple countries. Almås et al. (39) implement online experiments to study differences in fairness views between Scandinavians and Americans, as well as their linkage to political views. Their choice experiments allow them to identify different fairness types—egalitarians, meritocrats, and libertarians—and to quantify their relative population shares. They find that Scandinavians and Americans are equally meritocratic, but there are more egalitarians and fewer libertarians in Scandinavia. Moreover, these fairness types are predictive of political views.

Similarly, Cappelen et al. (40) study the both economically and psychologically highly relevant distributive justice trade-off between false positives and false negatives. Any redistributive policy implicitly needs to trade off the risk of false positives (some people may receive transfers even though they do not deserve them) and false negatives (some people may not receive transfers even though they do deserve them). In a world of imperfect information, society cannot avoid both of these errors—when eligibility criteria are very tight, false negatives will arise, and when they are very loose, false positives will emerge. Cappelen et al. (40) experimentally elicit from large representative samples their fairness views in these contexts. They document that a majority of people prefer to avoid false negatives but that there is a strong partisan divide, with conservatives being considerably more likely to prioritize avoiding false positives.

Overall, these results suggest that a considerable fraction of political views on economically relevant topics appear to reflect individual differences in moral values and fairness views. Again, the previous is not intended to be a comprehensive review of the relevant economics literature. Rather, my objective is merely to offer a snapshot of the types of contributions economists have made to questions that are likely of central interest also to moral psychologists. In my opinion, what has enabled economists to make these contributions is typically that they explicitly build their investigations on the ideas and concepts suggested by psychologists or philosophers but study them either in the wild or at a larger scale.

Toward greater integration

I conclude with 5 thoughts on methods and topic areas. First, while this perspective has repeatedly argued for the value of economists’ empirical skills in conducting research on morality, I do not wish to be understood as offering the (in any case ridiculous) suggestion that moral psychologists drop the ball on their comparative advantage of creative hypothesis development, careful experimentation, vignettes, and computational modeling. Rather, in my opinion, we as a broader social science community should be ecstatic when psychologists, economists and political scientists converge on a common set of results, as has happened with the bulk of work on the universalism-particularism cleavage.

Second, while economists have made some progress on incorporating morality and fairness into their political economy analyses, they have so far only touched a small share of moral psychology. There are many ideas and literatures that I did not mention in this perspective that appear intuitively relevant for contemporary political economy, such as work on respect, honor, and social dominance. For example, given political economists’ growing interest in populist-authoritarian leaders, it may be useful to formalize and test at scale the theory of intergroup social dominance relations pioneered by Pratto et al. (41) and Sidanius and Pratto (42).

Third, all of the previous concerns empirical work. Yet, economics distinguishes itself from psychology not only through applied econometrics, but also through its sophisticated theoretical modeling of incentives, strategic interaction, and equilibria. Space constraints prevent me from discussing much of the rich literature in economic theory that interacts with moral psychology. This literature includes both work on how morals affect strategic and market interactions (43) and work on their evolutionary foundations (44, 45).

Fourth, one area in which morals have been actively discussed as restricting economic transactions is market design (46), a debate in which moral philosophers have actively participated (47, 48). Yet, beyond providing compelling case studies of repugnant transactions, we still lack an understanding of the deeper moral principles that constrain market transactions, and how to work around them (49).

To conclude, I believe that both economists and moral psychologists stand much to gain from more closely interacting. Encouragement is due not only for the economists—while we have certainly oftentimes been too ignorant of the key developments in moral psychology, moral psychologists have arguably been similarly complicit in paying scarce attention to the latest in the “dismal science.” The most compelling case for the broad proposition that politico-economic outcomes and morality influence each other is likely to emerge when interdisciplinary exchange takes place—not only when economists are inspired by work in moral psychology, but also when moral psychologists appreciate and build on the techniques and approaches suggested in economics.

Acknowledgments

The author is grateful to an associate editor, 2 reviewers, Fiery Cushman, and Josh Greene for valuable comments.

Funding

There was no funding for this work.

Preprint

A preprint of this manuscript was previously published on the National Bureau of Economic Research website: https://www.nber.org/papers/w32279.

Data Availability

There are no data underlying this work.

References

Author notes