Introduction
Economic choice is the behavior observed when individuals make choices solely based on subjective preferences – for example when choosing between dishes on a restaurant menu, between possible financial investments, or between different career paths. This behavior is in intrinsically fascinating, and intimately related to philosophical questions such as free will and moral behavior. Since the 18th Century, generations of economists and psychologists accumulated a rich body of knowledge, identifying concepts and quantitative relationships that describe economic choices. Thus choice is a rare case of high cognitive function for which we have a formal and quantitative behavioral description. This body of knowledge can be used to guide and constrain research in neuroscience. Importantly, economic choice is relevant to a constellation of mental and neurological disorders, including frontotemporal dementia, major depression, and drug addiction.
Options available for choice in different situations can vary on many dimensions. For example, different flavors of ice cream evoke different sensory sensations and may be consumed immediately; different houses for sale may vary for their price, their size, the school district, and the distance from work; different financial investment may carry different degrees of risk, with returns available in a distant, or not-so-distant, future. How does the brain generate choices in the face of this enormous variability? Economic and psychological theories of choice behavior have a cornerstone in the concept of value. While choosing, individuals assign values to the available options; a decision is then made by comparing values. Hence, while options can vary on multiple dimensions, value represents a common unit of measure to make comparisons. From this perspective, understanding the neural mechanisms of economic choice amounts to describing how values are computed and compared in the brain.
Recent discoveries
Representation of goods and values in orbitofrontal cortex. Much of our research has been devoted to understanding how subjective values are computed and represented at the neuronal level. In a series of experiments, monkeys chose between different juices offered in variable amounts. The (subjective) relative value of the juices was inferred from choices and used to interpret neuronal activity. Neuronal recordings focused on the orbitofrontal cortex (OFC), where we found three groups of cells encoding the value of individual offers (offer value), the value of the chosen good (chosen value) and the binary choice outcome (chosen juice) (Padoa-Schioppa and Assad 2006; Padoa-Schioppa 2011). Our finding were replicated in a large number of studies, and value signals were also found in other brain regions. Until recently, however, it was unclear whether values encoded in OFC – or anywhere in the brain – are causally related to choices. This question is critical because values can guide a variety of mental functions such as associative learning, perceptual attention, and emotion. Using electrical stimulation, we demonstrated that offer values encoded in OFC are causal to choices (Ballesta, Shi, et al 2020).
Context adaptation and optimal coding. A central focus has been to understand how the neuronal representation in OFC depends – or does not depend – on the behavioral context of choice. We discovered several phenomena. (1) One study found that the activity of neurons encoding the value of a particular good does not depend on what other goods are offered in alternative. This trait, called “menu invariance”, is consequential because it implies choice transitivity (Padoa-Schioppa and Assad, 2008). (2) The number of different goods available for choice is very large, potentially infinite. How does the brain cope with this variability? In one study, we found that OFC neurons associated with one particular good in one session remapped and became associated with a different juice in a different session. Importantly, the functional role of each cell remained stable — e.g., offer value cells remained offer value cells. Thus, the same neural circuit could subserve decisions between a variety of goods. (3) Values available in different contexts can vary substantially – the same person might choose between goods worth a few dollars or between goods worth thousands of dollars. We discovered that the encoding of value in OFC adapts to the range of values available in any given condition – a phenomenon called range adaptation (Padoa-Schioppa 2009). We also showed that range adaptation in the encoding of offer values is optimal in the sense that it ensures maximal expected payoff (Rustichini et al 2017).
Decision mechanisms. There is broad consensus that values are explicitly represented at the neuronal level. In contrast, where in the brain and how exactly values are compared to make a decision remains matter of debate. Some scholars proposed that decisions are made in a motor representation; others proposed that decisions are a distributed process taking place across many brain areas. Notably, the cell groups identified in OFC capture both the input (offer value) and the output (chosen juice, chosen value) of the decision process. This consideration lead us to propose that these neurons constitute the building blocks of a decision circuit (Padoa-Schioppa and Conen, 2017). Several lines of evidence support this view. (1) Theoretical work shows that the cell groups in OFC are computationally sufficient for binary choices. In other words, one can build a biophysically realistic neural network that generates decisions and whose elements resemble the cell groups identified in OFC. (2) Trial-to-trial variability in choices is correlated with trial-to-trial variability in the firing rates of each cell group in OFC. (3) Experiments using electrical stimulation demonstrate that neuronal activity in OFC is causally involved in value comparison (Ballesta et al 2022).
Economic choices in mice. Economic decisions likely take place within OFC. However, this neural decision circuit is poorly understood. For example, it is unclear whether different groups of neurons identified in this area are excitatory or inhibitory, whether they correspond to different anatomical cell types, how they are connected with each other, and how they are connected with other brain regions. These questions are hard to tackle using traditional neurophysiology, but they can be addressed in principle using genetic tools. Hence, in collaboration with Tim Holy, we developed a mouse model of economic choice behavior. In a first study, we found that optogenetic inactivation of OFC disrupted choices (Kuwabara et al 2020). Furthermore, we identified in the mouse OFC different cell groups analogous to those previously found in the primate OFC. More recently we recorded from OFC using 2-photon calcium imaging. We found that the representation of decision variables is longitudinally stable (Zhang, Livi, et al, bioRxiv). Furthermore, decision variables are differentially represented across cortical layers — neurons representing the decision inputs (offer value) populate primarily layers 2/3, while neurons representing the decision output (choice outcome, chosen value) populate primarily layer 5 (Livi, Zhang, et al, in preparation). Taken together with the anatomical organization of cortical layers, this result directly supports the notion of a decision circuit.
Current directions
Ongoing research proceeds in several directions:
- Dissecting the decision circuit. Using our mouse model, we continue to dissect the organization of the neural decision circuit. In particular, we are working to assess whether different groups of cells identified in relation to choice behavior are primarily excitatory or inhibitory (and what kind of inhibitory). Other experiments focus on long-range connectivity between cell groups and other brain regions. Going forward, we will combine 2-photon microscopy and optogenetic stimulation to assess the connectivity of different cell groups.
- Generalizing current notions. Current notions rely almost exclusively on studies in which two options were offered simultaneously. However, many real-life choices entail three or more options. Other choices are — at least prima facie — decisions to accept of reject a particular opportunity. We are interested in assessing whether notions emerged from previous studies generalize to more complex situations. These experiments rely on neurophysiology in monkeys.
- Role of other brain regions. In addition to OFC, we previously examined neuronal activity in lateral prefrontal cortex, anterior cingulate cortex, gustatory cortex, and the amygdala. One key question concerns the construction of subjective values. Going forward we plan to extend our investigation to other frontal regions and to visual cortex.
