The target article by John et al. provides a compelling framework for understanding proxy failures across biological and social systems. My contribution to the debate is specifically stressing how a key source of proxy failures in economics stems from a mismatch between the evolutionary function of biological reward systems and traditional conceptions of utility maximization. A clear illustration of this contradiction is the massive, unprecedented production and marketing of super-addictive goods that exploit vulnerabilities in dopaminergic reinforcement learning circuits, regarded by many as a legitimate (and even welfare improving) market response to individual utility-maximizing choices. This resonates with the example of neural proxy failure in Section 4.1 of the target article.
The notion of pleasure seeking as the implicit goal guiding human behavior is deeply embedded in classical utilitarian foundations of economics, from Bentham to Jevons (Sigot, Reference Sigot2002). However, as discussed by the authors, the dopaminergic system does not simply encode the hedonic value of pleasure. Rather, phasic dopamine provides a reward prediction error signal that enables reinforcement learning about stimuli and actions (Glimcher, Reference Glimcher2011). Products engineered to artificially amplify dopamine release, through ingredients targeting vulnerable nodes in mesolimbic reward circuitry, lead to maladaptive overvaluation of associated representations. This parallels neural proxy failure examples involving drugs of abuse.
Crucially, the evolutionary function of dopamine is to provide a scalar approximation of expected net benefits over cumulated rewards associated with stimuli and actions (Pasquereau & Turner, Reference Pasquereau and Turner2013). But the open-ended pursuit of artificial dopamine release, enabled by modern production and marketing techniques, violates an implicit assumption of the neural valuation system evolved in resource-limited environments. The result is uncontrolled dopamine-seeking behavior at the expense of organismic regulation and health. This evolutionary mismatch, which disrupts the proxy relation between dopamine and net benefits of actions, might underlie the pursuit and supply of unprecedented varieties of super-addictive synthetic products in the modern economy.
The solution requires integrating core biological principles such as allostatic regulation into formal and conceptual models of economic decision making. The organism's implicit goal is survival and health. But a utilitarian characterization of economic incentives prescribes the maximization of pleasure-related proxies such as utility. As the target article suggests, proxy failures may be mitigated by directly incentivizing adaptive goals such as sustainability and wellbeing, rather than easily hackable proxies with ambiguous adaptive value. Overcoming such dysfunction requires reconceptualizing the motivational foundations of economics in terms of allostatic, reward-predictive principles of life regulation (Friston, Reference Friston2010; Sennesh et al., Reference Sennesh, Theriault, Brooks, van de Meent, Feldman Barrett and Quigley2022), not mechanistic pleasure seeking – which does not amount of course to denying the adaptive value of pleasure signals (Berridge & Kringelbach, Reference Berridge and Kringelbach2015).
The mismatch between dopaminergic valuation systems evolved in a radically different environment and the modern economy leads to systematic proxy failures. But the problem runs deeper than specific hackable mechanisms. Utilitarian characterizations assume that economic decisions maximize a utility function that confuses pleasure for the organism's implicit goal. We must reconceptualize human motivation based on neurobiological principles of reward prediction and allostasis. Only then can we mitigate the proxy failure inherent in folk notions of “human desires” guiding behaviors. The solution is not correcting human decisional failures by means of paternalistic nudging interventions, but getting the biology right.
A biologically grounded approach to economic decision making may then move from the following building blocks.
Predictive regulation and allostasis: Organisms do not seek to simply maximize pleasure, but rather aim to maintain stability and homeostasis in the face of a changing environment. The brain uses interoceptive, hormonal, and sensory signals to predictively regulate physiological needs before they arise (Sterling, Reference Sterling2012). Allostasis refers to the continuous readjustment of optimal setpoints to minimize long-term cumulative costs arising from tradeoffs between competing demands (Schulkin & Sterling, Reference Schulkin and Sterling2019). This principle of stability through change is fundamentally at odds with notions of rational utility maximization.
Reward prediction errors, not pleasure: Dopamine neurons signal discrepancies between expected and received reward, that is, reward prediction errors (RPEs). Rather than encoding pleasure per se, phasic dopamine enables learning by stamping in stimulus associations that reduce RPEs over time (Abler, Walter, Erk, Kammerer, & Spitzer, Reference Abler, Walter, Erk, Kammerer and Spitzer2006). Drugs hijack this system by artificially inducing hyper-dopaminergic states. The evolutionary function is reward-based learning, not pleasure seeking.
Multicriterion optimization: The complex dynamics of allostatic regulation imply brain activity optimizes multiple constraints, not a single utility measure. Interdependent processes spanning molecular, cellular, network, and behavioral levels interact to mediate adaptive tradeoffs (Sterling, Reference Sterling2020). Multicriterial decision models better capture the conditional heuristics employed by biological systems than simplifying assumptions of utility maximization (Greco, Ehrgott, & Figueira, Reference Greco, Ehrgott and Figueira2016).
Motivational heterogeneity: Humans exhibit a diversity of social and cultural motives not adequately accounted for in economic models, such as meaning, autonomy, and connectedness. Intrinsically motivated behavior is guided by goal states rather than pleasure states (Ryan & Deci, Reference Ryan and Deci2000). Traditional notions of utility conflate hedonic experience with fundamentally distinct drivers of behavior.
In summary, utilitarian characterizations of economic decision making fail to capture the complex, conditional, and heterogeneous motivations underlying human behavior as shaped by the predictive, multicriterial drivers of biological regulation. Unless economic models start to acknowledge that humans have bodies and a biology with its own adaptive logic and tradeoffs, economic policies will be systematically exposed to, and systematic generators of, proxy failures.
You have
Access
The target article by John et al. provides a compelling framework for understanding proxy failures across biological and social systems. My contribution to the debate is specifically stressing how a key source of proxy failures in economics stems from a mismatch between the evolutionary function of biological reward systems and traditional conceptions of utility maximization. A clear illustration of this contradiction is the massive, unprecedented production and marketing of super-addictive goods that exploit vulnerabilities in dopaminergic reinforcement learning circuits, regarded by many as a legitimate (and even welfare improving) market response to individual utility-maximizing choices. This resonates with the example of neural proxy failure in Section 4.1 of the target article.
The notion of pleasure seeking as the implicit goal guiding human behavior is deeply embedded in classical utilitarian foundations of economics, from Bentham to Jevons (Sigot, Reference Sigot2002). However, as discussed by the authors, the dopaminergic system does not simply encode the hedonic value of pleasure. Rather, phasic dopamine provides a reward prediction error signal that enables reinforcement learning about stimuli and actions (Glimcher, Reference Glimcher2011). Products engineered to artificially amplify dopamine release, through ingredients targeting vulnerable nodes in mesolimbic reward circuitry, lead to maladaptive overvaluation of associated representations. This parallels neural proxy failure examples involving drugs of abuse.
Crucially, the evolutionary function of dopamine is to provide a scalar approximation of expected net benefits over cumulated rewards associated with stimuli and actions (Pasquereau & Turner, Reference Pasquereau and Turner2013). But the open-ended pursuit of artificial dopamine release, enabled by modern production and marketing techniques, violates an implicit assumption of the neural valuation system evolved in resource-limited environments. The result is uncontrolled dopamine-seeking behavior at the expense of organismic regulation and health. This evolutionary mismatch, which disrupts the proxy relation between dopamine and net benefits of actions, might underlie the pursuit and supply of unprecedented varieties of super-addictive synthetic products in the modern economy.
The solution requires integrating core biological principles such as allostatic regulation into formal and conceptual models of economic decision making. The organism's implicit goal is survival and health. But a utilitarian characterization of economic incentives prescribes the maximization of pleasure-related proxies such as utility. As the target article suggests, proxy failures may be mitigated by directly incentivizing adaptive goals such as sustainability and wellbeing, rather than easily hackable proxies with ambiguous adaptive value. Overcoming such dysfunction requires reconceptualizing the motivational foundations of economics in terms of allostatic, reward-predictive principles of life regulation (Friston, Reference Friston2010; Sennesh et al., Reference Sennesh, Theriault, Brooks, van de Meent, Feldman Barrett and Quigley2022), not mechanistic pleasure seeking – which does not amount of course to denying the adaptive value of pleasure signals (Berridge & Kringelbach, Reference Berridge and Kringelbach2015).
The mismatch between dopaminergic valuation systems evolved in a radically different environment and the modern economy leads to systematic proxy failures. But the problem runs deeper than specific hackable mechanisms. Utilitarian characterizations assume that economic decisions maximize a utility function that confuses pleasure for the organism's implicit goal. We must reconceptualize human motivation based on neurobiological principles of reward prediction and allostasis. Only then can we mitigate the proxy failure inherent in folk notions of “human desires” guiding behaviors. The solution is not correcting human decisional failures by means of paternalistic nudging interventions, but getting the biology right.
A biologically grounded approach to economic decision making may then move from the following building blocks.
Predictive regulation and allostasis: Organisms do not seek to simply maximize pleasure, but rather aim to maintain stability and homeostasis in the face of a changing environment. The brain uses interoceptive, hormonal, and sensory signals to predictively regulate physiological needs before they arise (Sterling, Reference Sterling2012). Allostasis refers to the continuous readjustment of optimal setpoints to minimize long-term cumulative costs arising from tradeoffs between competing demands (Schulkin & Sterling, Reference Schulkin and Sterling2019). This principle of stability through change is fundamentally at odds with notions of rational utility maximization.
Reward prediction errors, not pleasure: Dopamine neurons signal discrepancies between expected and received reward, that is, reward prediction errors (RPEs). Rather than encoding pleasure per se, phasic dopamine enables learning by stamping in stimulus associations that reduce RPEs over time (Abler, Walter, Erk, Kammerer, & Spitzer, Reference Abler, Walter, Erk, Kammerer and Spitzer2006). Drugs hijack this system by artificially inducing hyper-dopaminergic states. The evolutionary function is reward-based learning, not pleasure seeking.
Multicriterion optimization: The complex dynamics of allostatic regulation imply brain activity optimizes multiple constraints, not a single utility measure. Interdependent processes spanning molecular, cellular, network, and behavioral levels interact to mediate adaptive tradeoffs (Sterling, Reference Sterling2020). Multicriterial decision models better capture the conditional heuristics employed by biological systems than simplifying assumptions of utility maximization (Greco, Ehrgott, & Figueira, Reference Greco, Ehrgott and Figueira2016).
Motivational heterogeneity: Humans exhibit a diversity of social and cultural motives not adequately accounted for in economic models, such as meaning, autonomy, and connectedness. Intrinsically motivated behavior is guided by goal states rather than pleasure states (Ryan & Deci, Reference Ryan and Deci2000). Traditional notions of utility conflate hedonic experience with fundamentally distinct drivers of behavior.
In summary, utilitarian characterizations of economic decision making fail to capture the complex, conditional, and heterogeneous motivations underlying human behavior as shaped by the predictive, multicriterial drivers of biological regulation. Unless economic models start to acknowledge that humans have bodies and a biology with its own adaptive logic and tradeoffs, economic policies will be systematically exposed to, and systematic generators of, proxy failures.
Financial support
This research received no specific grant from any funding agency, commercial, or not-for-profit sectors.
Competing interest
None.