Sense of Agency and Decision Making

The intricate relationship between our subjective sense of control over our actions and the resulting outcomes, known as agency, is fundamental to human psychology and decision-making processes. Recent theoretical advancements, particularly the Agency Disinhibits the Dopamine System (ADDS) theory, propose a compelling neurobiological framework for understanding how fluctuations in agency impact brain function, specifically the dopamine system. This article delves into the ADDS theory, extending its implications to various facets of decision-making under uncertainty, offering novel predictions and reviewing supporting empirical evidence.

At its core, agency is the perception that one is the author of one’s actions and their consequences. Extensive research underscores its critical role in psychological well-being and a wide spectrum of human behaviors. While the psychological underpinnings of agency have been extensively studied, a growing body of work is now exploring its neural basis. The ADDS theory, as put forth by Ashby et al. (2024), posits that increases in agency lead to the disinhibition of the dopamine (DA) system, consequently increasing the number of tonically active dopamine neurons in the ventral tegmental area (VTA). This theory not only elucidates a specific neural network mediating these effects but also successfully accounts for a variety of existing neuroscience and behavioral findings. This article aims to build upon ADDS by deriving novel predictions regarding how the sense of agency influences key components of decision-making.

The ADDS theory posits that enhanced agency leads to an increased number of tonically active dopamine neurons in the VTA. This mechanism is thought to involve a neural network where increased agency amplifies excitatory inputs to the VTA, thereby releasing dopamine neurons from tonic inhibition. This process, in turn, elevates tonic dopamine levels in target brain regions like the frontal cortex and broadens the pool of dopamine neurons available to respond to environmental stimuli. Conversely, a diminished sense of agency is predicted to lead to increased inhibition of VTA dopamine neurons, reducing their tonic activity.

This neurobiological model yields several crucial predictions regarding decision-making:

• Increased Value of Positive Outcomes: ADDS predicts that heightened agency should amplify the subjective value, or utility, assigned to positive outcomes. This effect is expected to be more pronounced for gains than for losses.
• Increased Risk-Taking: The asymmetric impact on the utility of gains versus losses is hypothesized to lead to an increased willingness to engage in risky behaviors.
• Reduced Temporal Discounting: Generally, ADDS suggests that an increased sense of agency should lead individuals to discount future rewards less heavily. However, an important caveat is introduced: in situations where endogenous dopamine levels are already exceptionally high, increased agency might paradoxically lead to greater temporal discounting, manifesting as impulsivity.

These predictions extend beyond theoretical postulations, finding resonance with a substantial body of existing empirical research, as well as suggesting avenues for future experimental inquiry.

The Neural Underpinnings of Agency and Dopamine

The ADDS theory hinges on the intricate interplay between the brain’s assessment of agency and the dopamine system. The neural network proposed by ADDS, detailed in Figure 1 of the original publication, suggests that outputs from brain regions responsible for estimating agency gate dopamine release in the VTA via projections through the ventral subiculum of the hippocampus (vSub). This pathway influences the tonic firing rate of dopamine neurons, which are crucial for regulating various cognitive and behavioral processes.

Research indicates that the VTA contains distinct subpopulations of dopamine neurons with differential response properties. While some neurons are primarily driven by reward prediction errors (RPEs) – the difference between expected and actual outcomes – others are more closely linked to action selection and motor control. ADDS focuses on the VTA’s mesolimbic and mesocortical pathways, which heavily influence executive functions and value-based decision-making.

A key aspect of the ADDS model is its explanation of how agency influences dopamine release. When agency is high, the vSub excites the nucleus accumbens, which in turn inhibits the ventral pallidum. This disinhibition of the ventral pallidum leads to a reduction in tonic inhibition of VTA dopamine neurons. Consequently, more dopamine neurons become tonically active, increasing baseline dopamine levels in target areas and enhancing the system’s responsiveness to salient events. Conversely, a low sense of agency leads to increased tonic inhibition, reducing the number of tonically active dopamine neurons.

This mechanism leads to two fundamental predictions:

1. Increased Tonic Dopamine Levels: Higher agency leads to increased tonic dopamine levels in VTA target regions, such as the frontal cortex.
2. Amplified Dopamine Response to Outcomes: Increased agency enhances the pool of dopamine neurons capable of phasic firing. This means that better-than-expected outcomes will elicit a stronger positive dopamine burst, while worse-than-expected outcomes will result in a more pronounced decrease in dopamine activity below baseline.

Computational modeling of the ADDS network has successfully replicated empirical findings from neuroscience studies, further validating the proposed neural architecture. For instance, the model accurately predicted the differential effects of activating the vSub and the pedunculopontine tegmental nucleus on VTA dopamine neuron activity, lending credibility to the theory’s core mechanisms.

Decision Making Under Uncertainty: A New Perspective

Decision-making, particularly under conditions of uncertainty, is a cornerstone of human cognition. Two primary theoretical frameworks exist: statistical decision theory, which focuses on uncertainty about the state of the world, and decision theory, prevalent in economics and psychology, which addresses uncertainty about outcomes resulting from actions. ADDS significantly contributes to the latter by providing a neurobiologically grounded explanation for how agency shapes these choices.

Traditional decision theories, such as subjective-expected utility theory and prospect theory, posit that individuals make choices based on the perceived utility or value of potential outcomes, influenced by factors like objective value, subjective probability, risk, and temporal delays. ADDS offers a new lens by suggesting that the sense of agency directly modulates these factors.

One significant implication of ADDS is its prediction that increased agency enhances motivation. This heightened drive is likely to lead to a greater engagement in decision-making behaviors. Supporting this, studies have shown that individuals are more likely to engage in betting behaviors when they have direct control over the dice roll compared to when an external agent controls it, suggesting a link between agency and the willingness to seek potential gains.

Furthermore, ADDS predicts that enhanced agency should improve executive functions, leading to more rational and optimal decisions. While direct tests of this prediction are still emerging, research has indicated a reciprocal relationship: more optimal decisions are associated with a higher sense of agency. This bidirectional influence highlights the profound impact of perceived control on cognitive processes underlying decision-making.

The Amplified Value of Gains and the Nuances of Losses

The ADDS theory predicts a significant asymmetry in how agency affects the valuation of outcomes. Specifically, it posits that increased agency substantially amplifies the utility of positive outcomes while having a comparatively minor effect on the utility of negative outcomes. This prediction stems from the understanding that dopamine plays a more prominent role in processing rewards and positive predictions than in processing losses or negative predictions.

Empirical evidence supporting this prediction comes from studies on the "choice-induced preference shift." These studies consistently demonstrate that the act of choosing an option, which inherently enhances agency, leads to an increased valuation of that chosen option. This effect is particularly robust for positive outcomes. For instance, individuals report greater happiness and higher valuations for items they actively selected compared to items they were simply assigned, and this effect has been shown to be partly mediated by dopamine levels.

In contrast, the impact of active choice on the valuation of losses is significantly less pronounced. Experiments comparing the effects of active choice on gains versus losses have found that while choice enhances the perceived value of gains, it does not similarly alter the perceived disutility of losses. This dissociation between gains and losses, and the differential influence of agency, aligns with ADDS’s prediction that the dopamine system, central to agency, is more strongly engaged by positive valence outcomes.

Increased Risk-Taking Behavior

The asymmetric impact of agency on the valuation of gains and losses directly translates into a prediction of increased risk-taking behavior. When the potential rewards of successful gambles are perceived as more valuable due to heightened agency, individuals are more likely to accept gambles that carry a higher risk of loss. This is because the increased utility of potential gains outweighs the comparatively less affected disutility of potential losses.

Studies utilizing the Balloon Analog Risk Task (BART) have provided direct evidence for this prediction. In BART, participants inflate a virtual balloon, earning rewards for each pump, but risking all accumulated rewards if the balloon pops. Manipulating agency, either by varying the delay between action and outcome or by altering the degree of control over the pumping mechanism, has consistently shown that higher agency leads to greater risk-taking. Participants with a stronger sense of control inflate the balloon more, accepting a higher probability of loss for the chance of greater rewards.

Further indirect support comes from the extensive literature linking dopamine levels to risk-taking. Pharmacological interventions that increase dopamine neurotransmission, such as L-DOPA administration in Parkinson’s disease patients, are often associated with an increased propensity for risky behaviors, including gambling. These findings collectively suggest that dopamine, modulated by agency, plays a critical role in driving risk-seeking tendencies.

Subjective Probability and Temporal Discounting

The ADDS theory also offers insights into how agency might influence estimations of subjective probability and the valuation of future rewards over time. While research on the direct impact of agency on subjective probability estimates is less developed, studies suggest that dopamine plays a role in probability biases, such as overestimating small probabilities and underestimating large ones. ADDS predicts that alterations in agency, and consequently dopamine levels, could modulate these biases. However, the precise nature of this modulation requires further investigation.

A more robust prediction of ADDS concerns temporal discounting – the tendency to devalue future rewards compared to immediate ones. Research has established that dopamine neuron activity exhibits hyperbolic temporal discounting, mirroring human choice behavior. This has led to the hypothesis that the dopamine system underlies temporal discounting.

ADDS extends this by predicting that increased agency should generally reduce temporal discounting, encouraging individuals to wait for larger, delayed rewards. This is supported by studies where participants with a greater sense of control over certain aspects of a task, even if trivial, displayed less temporal discounting. Similarly, interventions that are thought to enhance agency, such as episodic future thinking, have been shown to decrease temporal discounting.

However, ADDS introduces a critical nuance: the effect of agency on temporal discounting may be non-monotonic and dependent on baseline dopamine levels. In individuals with already high endogenous dopamine, increased agency could potentially exacerbate temporal discounting, leading to impulsive behavior. This hypothesis is partly supported by observations in Parkinson’s disease patients undergoing dopamine replacement therapy, where an increase in impulsivity and temporal discounting has been noted, potentially due to dopamine overdose in specific brain regions.

Implications and Future Directions

The ADDS theory offers a unifying framework that connects the sense of agency to fundamental aspects of decision-making through its influence on the dopamine system. The predictions derived from ADDS – amplified valuation of gains, increased risk-taking, and generally reduced temporal discounting – have significant implications for understanding a wide range of human behaviors, from everyday choices to the development of addictive behaviors and responses to various neuropsychiatric disorders.

The adaptive value of these predictions is also compelling. In environments where control is high, a willingness to take calculated risks and prioritize future rewards becomes more advantageous. Conversely, in low-agency environments, a risk-averse strategy and preference for immediate gratification may be more prudent.

While considerable empirical evidence supports the core tenets of ADDS and its extensions to decision-making, rigorous testing of these predictions requires further research. Future studies should aim to:

• Directly manipulate agency and measure its impact on the utility of gains and losses, as well as risk-taking and temporal discounting, using a wider array of experimental paradigms.
• Investigate the specific neural mechanisms through which agency influences subjective probability estimates.
• Clarify the non-monotonic relationship between dopamine levels, agency, and temporal discounting, particularly in populations with varying endogenous dopamine levels.
• Explore the differential effects of agency on dopamine neurons in the VTA versus the substantia nigra pars compacta (SNpc) and their respective roles in mediating behavior.

By bridging the gap between the subjective experience of control and the neurobiological underpinnings of decision-making, ADDS provides a powerful new framework for understanding human behavior and offers a promising avenue for future research. The continued exploration of agency’s role in decision-making promises to deepen our understanding of motivation, reward processing, and the complex interplay of factors that guide our choices.