The impact of agency on time and risk preferences

The concept of personal agency, the extent of an individual's control over their life, has been a central theme in philosophy, political science, and economics. Philosophers like Aristotle and Rawls, and economists ranging from Marx to Friedman, have highlighted its importance. Agency is defined relative to an individual's choice set; a larger set implies greater agency. The value of agency is understood through two main lenses: instrumental value (a larger choice set maximizes utility) and intrinsic value (people value agency itself). Existing research shows that individuals react negatively to diminished agency, seeking to punish those responsible and experiencing improved well-being when agency is restored, even without financial benefits. This paper explores a third dimension of agency: its impact on shaping preferences, specifically patience and risk tolerance. The hypothesis is that greater agency increases both risk tolerance and patience, leading to a preference for larger, later rewards. A key aspect of the research is investigating whether the mere knowledge of having agency, even without exercising it, influences behavior. This has significant implications for decision-making involving time and risk tradeoffs, such as saving, education, and health choices. The theoretical framework posits that greater agency reduces perceived risk, increasing risk tolerance. Since decisions involving time inherently involve risk (the further into the future, the greater the uncertainty), increased risk tolerance should lead to increased patience and preference for larger, delayed rewards. Initial evidence from the 2010–2014 World Values Survey (N = 86,272) shows a positive correlation between reported agency and savings, suggesting a link between agency and patience. This correlation, while not implying causality, motivates two experiments designed to establish causal effects.

Prior research extensively links resource scarcity with increased risk aversion, impatience, and a preference for immediate rewards. This behavior is often attributed to the experience of scarcity itself. However, scarcity frequently co-occurs with a lack of agency; individuals facing scarcity often lack control over their situation. This makes it difficult to isolate whether the scarcity itself or the lack of agency drives the observed behavioral changes. This study aims to disentangle these effects, focusing on the role of agency as a mediator in the relationship between adverse states and impatience. Studies examining the effects of poverty alleviation programs, particularly those using unconditional cash transfers, demonstrate positive outcomes like improved schooling and health. These programs enhance agency by providing recipients with greater control over resource allocation. This research aims to shed light on the potential role of this increased agency in shaping more forward-looking behavior, beyond the immediate instrumental and intrinsic benefits.

The study employs two experiments to test the causal relationship between agency and preferences. **Study 1** investigated the impact of agency on patience and risk tolerance under conditions of resource scarcity. Participants (N=211) were randomly assigned to three groups: (1) Scarcity-Agency (Agency), (2) Scarcity-No Agency (No Agency), and (3) Control (no scarcity). The scarcity manipulation involved time constraints for a cognitive task. Both scarcity conditions had equal scarcity levels, but only the Agency group had the option to alleviate scarcity by gaining extra time; however, this option incurred a cost to discourage its use. Patience was measured by choices between smaller, sooner and larger, later financial rewards (token allocation). Risk tolerance was assessed through binary choices between risky and safe gambles. Manipulation checks confirmed the effectiveness of scarcity and agency manipulations. The key comparison was between the Agency and No Agency groups, both experiencing the same scarcity but differing in agency. Study 1b, a manipulation check, examined demand for agency, emotional states, and trust in the experimenter to ensure that the manipulation affected agency and not other constructs. **Study 2** generalized the findings to other adverse states. Participants (N=113) were asked to complete an anagram task while exposed to aversive noise. The Agency group could remove the noise at a cost, while the No Agency group couldn't. Patience was again measured through choices between sooner and later rewards. Similarly, Study 2b included a manipulation check on agency using a measure of demand for agency. In both studies, statistical analyses (OLS regression, Mann-Whitney U test) were used to compare the groups, focusing on the effects of agency on patience and risk tolerance. In Study 1, mediation analysis examined the role of risk tolerance in the relationship between agency and patience. Lee bounds were used to address a potential self-selection bias stemming from participants not exercising their agency. Hyperbolic discounting analysis in Study 2 provided another measure of patience using data from time-preference choices.

Study 1 revealed that participants with agency over resource scarcity were significantly more patient than those without agency, even though the majority in the Agency group didn't exercise their option. Their patience levels were comparable to the Control group (no scarcity). This suggests that the mere perception of agency is sufficient to increase patience. The analysis showed that the relationship between agency and patience is partially mediated by risk tolerance; greater agency led to greater risk tolerance, which in turn led to greater patience. Study 2 replicated these findings using an aversive noise manipulation; the Agency group (capable of removing the noise at a cost) showed significantly greater patience than the No Agency group, despite equal exposure to noise. Both parametric and nonparametric analyses confirmed these findings across both studies. Manipulation checks consistently validated the effectiveness of both scarcity and agency manipulations in the studies.

These findings significantly contribute to the literature on adverse states and decision-making. The studies demonstrate a previously under-recognized pathway through which resource scarcity might affect behavior: the reduction of agency. Providing individuals with agency, even if not exercised, can effectively mitigate the negative impacts of scarcity on patience and risk tolerance. This highlights the importance of considering agency in policies aimed at alleviating poverty and other adverse conditions. The results challenge existing theories of poverty that focus solely on increasing access and opportunities or assume immutable preferences among those in scarcity. Instead, the findings emphasize the importance of designing programs that enhance agency, potentially fostering more forward-looking behavior and maximizing the impact of interventions. The finding that the positive effects of agency hold even when it is not exercised points to the potential for cost-effective interventions that cultivate patience and risk tolerance.

This research demonstrates the significant impact of personal agency on time and risk preferences, even in the absence of actual agency exercise. The findings highlight the potential for designing poverty alleviation programs that focus on increasing agency rather than simply increasing resources or opportunities. Future research could explore the optimal levels of agency provision, the generalizability of these effects across different cultures and contexts, and the long-term impacts of agency-based interventions on various economic and social outcomes. Additional research could also examine different types of agency and their relative effects on preferences and behaviors.

One potential limitation of Study 1 is the possibility that the negative impact of the adverse state, and hence the agency manipulation, could have been greater for participants with lower cognitive abilities. Although random assignment ensures the validity of the conclusions about the effects of agency on behavior on average, this possibility cannot be entirely ruled out. Another limitation is the use of hypothetical scenarios in some measurements; future studies could incorporate real monetary incentives to further validate the findings. The samples used in the studies might not be fully representative of the broader population, limiting the generalizability of the findings to some extent.