Behavioral Economics: Method, Norms, and Policy

In the field of public policy, for example, Richard Thaler and Cass Sunstein1 defend the stance of steering individuals into making choices that benefit their own long-term self-interest. People would retain the possibility of declining (opting out of) any state-mandated directive. Under such an arrangement, they write, “the goal should be to avoid arbitrary and harmful random effects and to produce a situation that tends to promote the well-being of people properly defined” (179, emphasis added).

The crucial question, of course, is who defines the well-being of people, and on what grounds. Classical liberals think that it must be a matter of free individual choices, while behavioral economists attribute the task to experts and public planners who allegedly know better what an objective measure of well-being is and how to achieve it.

We will not pursue the comparison between the two approaches extensively. The purpose here is only to point out some methodological and normative issues with behavioral economics, and their implications for public policy design, by looking at a text by one of its founders: Richard Thaler.

Thaler’s presidential address2 given at the American Economic Association in 2016 synthetizes his perspective on behavioral economics. There are several methodological aspects in the text that merit attention, namely, the reconstruction of neoclassical thought, the consideration of costs, the analysis of how markets work, and the interplay between the descriptive and normative approaches.

The reconstruction of neoclassical thought: cognition

In the neoclassical model of Homo Economicus, Thaler includes the following assumptions: well-defined preferences, unbiased beliefs and expectations, and “infinite cognitive abilities” and “infinite willpower” to make “optimal choices” (1578), and adds that in neoclassical theories agents act as if they understood the model (1593). He considers that model to be highly implausible given that optimization is a difficult task (1579). Following Amos Tversky and Daniel Kahneman’s experiments, he holds that “humans make judgments that are systematically biased” (1581) and that markets exacerbate those biases. For that reason, he endorses a theory of engineering, that is, a “set of practical enhancements” to better predict behavior and to correct such biases (1586, 1592).

Thaler partially misconstrues the cognitive premises of neoclassical authors. In The Economic Approach to Human Behavior, Gary Becker writes that “the economic approach does not assume that all participants in any market necessarily have complete information [nor that] decisions units are necessarily conscious of their efforts to maximize” (6-7). That said, participants do have an optimal amount of information to engage in market transactions on the basis of a calculus of utility, and they seek to minimize costs “that may not be easily ‘seen’ by outside observers” (Ibid. 14, 112). Thus, cognition is not perfect nor infinite but it is strong enough to allow market actors to find ways to choose the proper means to achieve competing ends.

Thaler seems to believe that either we are perfect rational maximizers or we are inclined to make cognitively-biased decisions mostly leading to undesired economic choices. But dichotomous approaches may be unfit logical instruments to explain reality. A model based on the premises of incomplete information, limited rationality, and unarticulated ways of doing economic calculations seems more adjusted to explain why free markets have worked better than all the engineering alternatives.

There has been a steady global growth in wealth in the past 250 years, as pointed out by Deirdre McCloskey3. The history of wealth shows that rationality in free markets—albeit a limited, fragile, and tacit rationality—has prevailed over cognitive biases and has led thousands of millions of people to more prosperous and fulfilling standards of life.

The consideration of costs

Thaler also addresses neoclassical views on costs. He casts doubts about firms’ ability to optimally calculate marginal costs and revenues (1580); he relates the notion of opportunity cost to rational choice models only (1584), and he writes that “unless the transaction costs can be measured the use of the concept is undisciplined” (1594). In other words, in his view firms seem to be quite unintelligent when it comes to calculating costs; the appeal to opportunity costs is limited to a single economic model, and the notion of transaction costs comes across as almost metaphysical. Such an understanding of costs is derived from his reconstruction of neoclassical cognitive premises. He infers that, because there is no such thing as perfect rationality, there can be no adequate calculation of costs. However, Thaler does not offer his own account of the calculation of costs in order to justify the possibility of engineering” individual economic decisions.

Other perspectives on costs can help us assess that possibility. As James M. Buchanan points out in Cost and Choice4, costs represent the “anticipated utility loss upon the sacrifice of a rejected alternative” (p.41). Costs are subjectively assessed by the agents involved in the particular economic exchanges, on the basis of information only they can possess and of their own values (Ibid. 25,44,76). This subjective approach to costs refutes the idea of a successful “engineering” of any individual decision when it comes to a cost-benefit calculus.

Along similar lines, Rosolino Candela points out that transaction costs are the costs of engaging in economic calculation,5 a crucial element to understanding how the market economy works. From this perspective, it seems equivocal to limit the concept of transaction costs to a unique and correct standard of measure, as Thaler suggests, since the costs of participating in market transactions can only be estimated by individual agents based on a diversity of factors that can never be detected by the engineers.

Last, and perhaps most importantly, Buchanan6 also calls attention to the concept of unintended consequences as the most distinctive characteristic of the social scientist (57). Applied to the subject under analysis, the unintended costs of economic choices must also be taken into account when proposing to regulate, steer, or engineer the economy.

Economic analysis cannot be impervious to costs. Yet Thaler overlooks these considerations. Take, for example, his proposal, “Save More Tomorrow,” which defends workers’ mandatory enrollment in retirement plans and the option to increase the proportion of pension contributions as their salaries rise (1596). Several questions arise in the face of such a proposal. Why induce people to save more and not buy more property or invest in business ventures? Who should estimate the opportunity costs and the unintended consequences of each decision? The answers to these questions require an overall cost-benefit analysis that can only be done by market agents. But Thaler assumes that higher levels of savings are better than investment alternatives. He does not weigh the transaction costs, opportunity costs, and unintended or unanticipated consequences associated with the Save More Tomorrow proposal.

How markets work

The core idea of behavioral economics is that individual economic choices in markets tend to fail in securing desired ends due to a number of cognitive biases. With regard to consumers, Thaler finds that learning is complex and people err at making important decisions (such as saving for retirement). Therefore, he infers, people need steering by experts and public planners. With regard to producers, Thaler holds that markets exploit generalized cognitive biases in order for firms to increase profits (1585). To illustrate this point, he examines the case of the 2008 financial crisis and posits that housing markets exacerbated behavioral biases and price bubbles created a global recession. He mentions that prior to the crisis the “lending requirements were unusually lax” and that borrowers were fooled by lenders (1586-89).

However, Thaler does not take into consideration the political and financial mechanisms behind the crisis, fueled by the Federal Reserve’s cheap-credit monetary policy that led to a housing-building and lending bubble. Several scholars have pointed out that an active federal housing subsidy policy helped weaken the standards and encouraged a mortgage scheme based on high-risk loans. The federal backing sent signals to the financial executives, who assumed the government’s implicit guarantee if something went wrong, thereby encouraging irresponsible behavior and neglecting precautions and controls to prevent it from happening.7

In his analysis, Thaler overlooks the arguments and evidence that might refute his reading, which attributes the causes of the crisis to foolish people and unprincipled lenders only and omits any critical reference to government regulations and subsidies. This is a partial examination of the real causes and consequences of the crisis, perhaps due to a bias against markets. Consequently, it weakens the case that experts are supposedly less exposed to the type of biases ascribed to the rest of the people.

Descriptive, predictive, and normative approaches

A few behavioral economists8 invoke Adam Smith’s ideas on the sentiments and thoughts that drive people in social interactions. Thaler sees in Smith the father of empirical economics; he proposes to stop arguing about theoretical principles or predictive approaches and turn to evidence-based economics (1597). The analysis of his methodological stance requires that we revisit the relation of theory to empirical observation, and the work of Karl Popper9 may be useful for such a task.

Popper finds that: 1) empirical observation alone is incomplete and subject to the dangers of misinterpretation if it is not guided by theories (which is “why induction doesn’t work” 41-42n8); 2) the rationality of science lies “solely in the critical approach—in an attitude which, of course, involves the critical use, among other arguments, of empirical evidence” (228-29), and 3) the theoretical social sciences should be mainly concerned with tracing “the unintended social repercussions of intentional human actions” (342).

Popper’s recommendations are useful for understanding some of the shortcomings of the behavioral economic methodology. First, the treatment of the 2008 crisis does not take into account all available factors and therefore shows an uncritical attitude that taints any observation with regard to the causes of the crisis, in particular, and to the dynamics of markets more generally. Second, as mentioned earlier, by skirting the consideration of the costs of economic engineering behavioral economics may be eluding a scientific responsibility as understood by Buchanan and Popper.

Last, Thaler assumes there is a unique right measure for economic indicators (such as the level of individual savings for retirement), that experts know what that measure is, and that it is legitimate for the government to induce people to adopt it. The latter is a judgment of value implicitly slid in an approach that claims to be only descriptive, which suggests that evidence-based economics is guided by an implicit normative theory after all.

From Method to Norms and Policy

The analysis of Thaler’s proposal to steer individuals into making the right choices must take into consideration questions about its normative value and its policy recommendations. From the normative point of view, should the people be oriented by the government, or rather should they be left free to make individual economic decisions? According to Adam Smith, individual liberty, security, and justice should never be obstructed nor destroyed by governments (An Inquiry into the Nature and Causes of the Wealth of Nations, 66410). This normative focus in Smith is highlighted by Ronald Coase in “Adam Smith’s View of Man”11: “It is because markets are thought to embody ideal justice, and not because they promote interests, that Smith relies on them” (57n70). Along similar lines, in Economic Enquiry and its Logic,12 Buchanan defends the Smithean policy position that seeks to shrink political intervention in the economy and widen the [individuals’] range for potential choice (15-16).

From a classical liberal stance, the proper business of government is not to nudge or steer individuals but to protect their rights. If behavioral economics finds it acceptable to restrict the range of individual liberty for the sake of policy goals, despite its claims to the contrary, it is an anti-liberal normative theory.

On the public policy front, some of the questions to address include the following: what design helps better mitigate the cognitive problems of the people, the experts, and the public planners? How are their cognitive biases different? What are the resulting costs of their decisions and who bears them? Behavioral economics presumes that politically-mandated solutions that deal with the consequences and biases of individual choices will be successful. But technocrats are also subject to overconfidence, miscalculation, preference for short-run results, errors in forecasting, and discretional decisions that they observe in the people. Behavioral economics tends to overlook these issues.

Bryan Caplan13 points out that out of 67 articles in behavioral economics with policy recommendations, 65 fail to assess the cognitive ability of policymakers and the unintended consequences of the proposed policies. This is problematic, for policymakers and regulators may believe that they are exempt from cognitive biases, or that they can predict the magnitude of the impact of regulations. Several other authors warn against the experts’ cognitive biases. For Kurt Weyland,14 “even well-trained, highly competent specialists are compelled to apply cognitive heuristics—and to incur the corresponding risks of distortions and biases,” and Slavisa Tasic15 observes that “the possibility of regulatory error is not seriously considered in policy formulation, and the concept of unanticipated negative ramifications is largely absent from policy analysis” (429). See also Roger Koppl16 on the bounded rationality of experts (167-168).

Conclusion: Back to Smith

Behavioral economists such as Thaler think that experts and planners can know what an objective measure of economic well-being is and how to achieve it by means of public policies. However, they overestimate their cognitive capacities to do so, and they tend to ignore the negative costs of those policies and the responsibilities that go with them. Thus, they fail to consider all the cognitive and moral aspects associated with their policy prescriptions.

Some behavioral economists (Thaler included) admire Adam Smith. The founder of modern economics sought to explain the dynamics of a complex reality, and he found in the system of natural liberty the best possible arrangement for the social order. He arrived at two conclusions worth highlighting here. First, “not only the prejudices of the publick, but what is much more unconquerable, the private interests of many individuals, irresistibly oppose [freedom of trade]” (An Inquiry into the Nature and Causes of the Wealth of Nations, 471). Second, he preferred the man of public spirit, who acknowledges the people’s common prejudices and errors and yet hopes to correct them with patience and persuasion, to the man of the system, who intends to manage people as if they were inert pieces on a chessboard (The Theory of Moral Sentiments,17 233).

Because Smith was more worried about the harmful actions of special interests and presumptuous bureaucrats than about the effects of people’s biases, he wrote in defense of a limited government and a free market economy. It seems that those behavioral economists who find a source of inspiration in Smith should look again at his recommendations.

Footnotes

[1] Thaler, Richard H. and Cass R. Sunstein. “Libertarian Paternalism.” The American Economic Review 93, no. 2 (May 2003): 175-179.

[2] Thaler, Richard. Misbehaving: The Making of Behavioral Economics. New York: W. W. Norton & Co., 2015.

[3] Interview with Deirdre McCloskey. By Marian L. Tupy. Human Progress (November 4, 2022).

[4] Buchanan, James M. Cost and Choice: An Inquiry in Economic Theory. Available online as Volume 6 of The Collected Works of James M. Buchanan at https://www.econlib.org/library/Buchanan/buchCv6.html, Library of Economics and Liberty.

[5] Candela, Rosolino. “Transaction Costs are the Costs of Engaging in Economic Calculation.” EconLog (June 3, 2020).

[6] Buchanan, James M. The Collected Works of James M. Buchanan, Volume 17: Moral Science and Moral Order. Indianapolis: Liberty Fund, Inc., 2001.

[7] See the analysis by R. G. Holcombe and B. Powell (comp.), Housing America: Building Out of a Crisis, New York and London: Routledge, 2017 [2009], pp. 6-7; L. White, “Fannie Mae, Freddie Mac, and Housing Finance: Why True Privatization Is Good Public Policy,” Policy Analysis, Cato Institute, No. 528, 2004, and Arnold King, “The 2008 Financial Crisis,” The Concise Encyclopedia of Economics.

[8] Ashraf, Nava, Colin F. Camerer, and George Loewenstein. 2005. “Adam Smith, Behavioral Economist.” Journal of Economic Perspectives, 19 (3): 131-145.

[9] Popper, Karl. Conjectures and Refutations: The Growth of Scientific Knowledge. New York: Routledge Classics, 2002.

[10] Smith, Adam. An Inquiry into the Nature and Causes of the Wealth of Nations, Volume I. Edited by R. H. Campbell and A. S. Skinner. Indianapolis: Liberty Fund, Inc., 1982. Available online at https://www.econlib.org/library/Smith/smWN.html, Library of Economics and Liberty.

[11] Coase, R. H. “Adam Smith’s View of Man.” The University of Chicago, Graduate School of Business, Selected Papers No. 50.

[12] Buchanan, James M. The Collected Works of James M. Buchanan, Volume 12: Economic Inquiry and Its Logic. Indianapolis: Liberty Fund, Inc., 2000.

[13] Caplan, Bryan. “Biased Behavioral Econ.” EconLog (April 14, 2016).

[14] Weyland, Kurt. Bounded Rationality and Policy Diffusion: Social Sector Reform in Latin America. Princeton: Princeton University Press, 2007.

[15] Tasic, Slavisa. “The Illusion of Regulatory Competence.” Critical Review, 21:4, 423-436.

[16] Koppl, R. (2018). “Expert Failure.” In Expert Failure (Cambridge Studies in Economics, Choice, and Society, p. i). Cambridge: Cambridge University Press.

[17] Smith, Adam. The Theory of Moral Sentiments. Edited by D. D. Raphael and A. L. Macfie. Indianapolis: Liberty Classics; Liberty Fund (Glasgow Edition), 1982. Available online at https://www.econlib.org/library/Smith/smMS.html, Library of Economics and Liberty.