Behavioural Public Policy (BPP) is often portrayed as a battleground between correcting biased decisions and working with people’s adaptive heuristics. In this blog post, Alejandro Hortal revisits the divide, arguing that the ‘rationality wars’ are far from new and that recognizing the plurality of rationality can support more effective and context-sensitive policy design today.
Why rationality wars matter for BPP
BPP rests on a deceptively simple question: how does human rationality work, and how should it work? This question is not merely theoretical. It shapes how we design interventions, often with real consequences for autonomy, welfare, and trust.
Across much of the behavioural sciences, there is broad agreement that human rationality is bounded. Yet there is far less consensus on what follows from this fact. Should bounded rationality be treated as a flaw to be corrected by reference to an idealized standard of logical coherence? Or should it be understood as an adaptive form of reasoning that sets its own normative benchmark?
This disagreement lies at the heart of what have come to be called the “rationality wars.” On one side, researchers inspired by Daniel Kahneman – the central reference point of this tradition – emphasize systematic biases and retain logical rationality (consistency, transitivity, Bayesian updating) as the normative ideal. On the other, Gerd Gigerenzer and scholars working on ecological rationality argue that heuristics are adaptive decision strategies whose rationality depends on their fit with uncertain environments.
In my recent article in Behavioural Public Policy, I argue that although they are often framed as empirical or methodological disputes, the rationality wars persist because different, epistemically grounded concepts of rationality are treated reductively as if they were exhaustive.
Rationality has always been plural
The idea that rationality comes in more than one form is a longstanding feature of philosophical reflection. Plato, by dividing the soul into reason (logos), spirit, and appetite, offered a descriptively plural account of human agency in which reason must govern motivation and desire in pursuit of the good. This structure is vividly illustrated in the charioteer allegory, where logos guides two horses – spirit and appetite – showing that rational agency consists not in eliminating desire but in coordinating it.
Normatively, however, Plato remained reductionist: logos alone has epistemic authority, and rational judgment is measured against an ideal standard oriented toward the world of Forms, only imperfectly instantiated in empirical life. From a contemporary perspective, this position can be read as anticipating the idea that logical rationality performs best in stable, idealized domains, while its application to contingent and uncertain material contexts – what we now describe as “big worlds” – yields only partial understanding.
Aristotle, by contrast, pluralized rationality both descriptively and normatively. He distinguished among different rational excellences, each appropriate to a distinct domain: nous (the intuitive grasp of first principles), epistēmē (demonstrative knowledge), sophia (the union of nous and epistēmē, concerned with necessary and eternal truths), technē (productive rationality), and phronēsis (practical rationality).
Phronēsis governs action in contingent and uncertain contexts. It is not the application of abstract rules but the exercise of judgment shaped by experience, habituation, and sensitivity to context. In contemporary terms, this is precisely the kind of rationality required to navigate today’s complex “big worlds.” By contrast, sophia belongs to the contemplative life (theōria) and concerns objects that do not admit of being otherwise, making it ill-suited to guiding action, decision-making, or public policy.
Aristotle’s pluralism extends even to his biological works, where he uses phronimos to describe animals exhibiting adaptive, situation-responsive behavior. Although such cases lack full deliberation, they reveal an early view of rationality as graded, embodied, and domain-sensitive rather than reducible to a single logical standard. This commitment to plural rational norms recurs throughout the history of philosophy, with thinkers such as Immanuel Kant explicitly distinguishing between theoretical and practical reason, each governed by different normative standards appropriate to its domain.
Why the ‘war’ persists
One way to make sense of rationality is to think about it as we think about love. Psychology, neuroscience, anthropology, sociology, law, and religion all speak about love using different concepts grounded in distinct methods and aims, without disagreeing about the same thing. Imagine a cultural anthropologist insisting that love is a socially embedded practice that cannot be reduced to oxytocin, and a chemist replying that love is nothing but a neurochemical process. The conflict would not reflect a genuine disagreement about love, but a category confusion.
Rationality works in the same way: economics, psychology, evolutionary theory, and public policy rely on different concepts of rationality, and the ‘wars’ arise only when one is treated reductively as if it exhausted rationality as such.
If plural conceptions of rationality have been recognized for centuries, why do ‘rationality wars’ persist? The answer lies in such reductionism. In his theory about sciences as categorical closures, the Spanish philosopher Gustavo Bueno distinguishes between concepts – discipline-specific tools shaped by particular scientific aims and methods – and ideas, which emerge through conceptual friction across disciplines. Confusion arises when a concept is mistaken for the idea itself.
Figure 1: Concepts and ideas
Source: Author
Rationality functions as a distinct concept in economics, cognitive psychology, and evolutionary psychology. Problems arise when one such concept is elevated to a universal idea and used to judge or dismiss the others. For example, when logical rationality is treated as a general standard rather than a domain-specific one, ecological or practical forms appear deficient; or when adaptive success is treated as the sole benchmark, logical coherence seems irrelevant.
This tension is sharpened by the fact that economics and evolutionary psychology, while both concerned with behavior and decision-making, operate under different epistemic contexts. Economics typically evaluates choices against formal, context-independent standards, whereas evolutionary psychology assesses rationality in terms of adaptive success in specific environments.
What this means for BPP
For BPP practitioners, these distinctions have direct implications. Consider retirement savings policy: defaults inspired by logical rationality aim to correct deviations from an abstract optimum, while boost-based interventions grounded in ecological rationality strengthen adaptive heuristics suited to uncertain labor markets—such as simple rules of thumb to save at least the employer match and increase contributions when income rises. Effective BPP therefore depends not on choosing sides, but on matching the epistemic field to the appropriate rationality standard and policy context.
Figure 2: Three concepts of rationality and their potential applications
Source: Author
Because BPP draws simultaneously on economics, psychology, and evolutionary theory, it cannot rely on a monolithic conception of rationality. Nudges, boosts, institutional design, and regulation all rest on different rationality assumptions, which are not interchangeable. The rationality wars dissolve once we recognize that logical and ecological rationality answer questions from different epistemic frameworks rather than compete for dominance. The real danger lies not in pluralism, but in reductionism.
Read the full article in the Political Economy section of Behavioural Public Policy.
Alejandro Hortal is a philosopher and behavioral public policy scholar whose work explores how insights from behavioral economics intersect with ethics, rationality, and culture in the design of public policy. He is Senior Lecturer at the University of North Carolina at Greensboro and Visiting faculty at Wake Forest University. His research focuses on nudges, autonomy, virtue ethics, and the ethical limits of behaviorally informed interventions, with applications to areas such as corruption, public health, sustainability, and gender-based violence.