Orbitofrontal signals for two-component choice options comply with indifference curves of Revealed Preference Theory

Abstract

Economic choice options contain multiple components and constitute vectorial bundles. The question arises how they are represented by single-dimensional, scalar neuronal signals that are suitable for economic decision-making. Revealed Preference Theory provides formalisms for establishing preference relations between such bundles, including convenient graphic indifference curves. During stochastic choice between bundles with the same two juice components, we identified neuronal signals for vectorial, multi-component bundles in the orbitofrontal cortex of monkeys. A scalar signal integrated the values from all bundle components in the structured manner of the Theory; it followed the behavioral indifference curves within their confidence limits, was indistinguishable between differently composed but equally revealed preferred bundles, predicted bundle choice and complied with an optimality axiom. Further, distinct signals in other neurons coded the option components separately but followed indifference curves as a population. These data demonstrate how scalar signals represent vectorial, multi-component choice options.

Values Encoded in Orbitofrontal Cortex Are Causally Related to Economic Choices

“…When subjects make choices, different groups of neurons encode the identities and values of offered and chosen goods (Padoa-Schioppa and Assad, 2006). Value encoding cells integrate multiple dimensions that characterize the options available in any given context (Roesch and Olson, 2005;Hare et al, 2008;Kennerley et al, 2009;Pastor-Bernier et al, 2019). Moreover, trial-by-trial variability in the activity of each group of neurons correlates with variability in choices (Padoa-Schioppa, 2013;Conen and Padoa-Schioppa, 2015).…”

Reward value revealed by auction in rhesus monkeys

“…We trained rhesus monkeys to move a joystick cursor on a computer monitor in order to place a bid for juice reward, paying from a water budget to obtain it. We chose these commodities because our animals are highly familiar with them and express meaningful, ordered preferences across them (Kobayashi and Schultz 2008;Stauffer et al 2014;Pastor-Bernier et al 2019). We found that the animals reliably expressed well-ranked, trial-by-trial estimates of subjective economic value for up to five juice volumes.…”

Experimentally revealed stochastic preferences for multicomponent choice options.

“…Future neuroeconomic work on underlying decision mechanisms should particularly benefit from the empirical estimation of whole maps of well-ordered ICs derived from multiple IPs that conform to predictive mathematical functions; thus, avoiding to test preferences for a few bundles with limited general validity. For example, different neurons in the orbitofrontal cortex of monkeys combine both bundle components into a common scalar neural signal or code each component separately (requiring later integration for contribution to the decision; Pastor-Bernier et al, 2019 ). The systematic ICs may also help to investigate neural underpinnings of specific theories, such as the switching of attentional processes between components conceptualized by multialternative decision field theory ( Roe, Busemeyer, & Townsend, 2001 ).…”

“…Correspondingly, the utility of a choice option can only be higher, lower, or equal to that of its alternative. Further, neural signals representing choice options can only vary along a single dimension at any given moment and, thus, are also scalar; their firing rate either increases, remains unchanged, or decreases (thus, constituting a distribution), even when encoding multidimensional variables ( Pastor-Bernier, Stasiak, & Schultz, 2019 ). Hence the question: how can single-dimensional preferences, utility, and neural signals concern multicomponent choice options?…”