Our results demonstrate that willingness to delay gratification depends on social trust. Whether contemplating a single interaction with one individual or multiple interactions with different individuals, people are less willing to wait for rewards with individuals they see as less trustworthy, when there is reason to doubt that an individual would actually deliver the delayed reward in the future. Directly experiencing the unreliability of an individual was unnecessary; here, impressions of trustworthiness from vignettes and faces produced powerful effects. This work complements prior correlational work, which suggested a link between trust and delaying gratification but did not establish causality (Mischel, 1961
; Harris and Madden, 2002
), and prior experimental work, which suggested that trust could influence delaying gratification but did not manipulate trust independent of rewards that can influence self-control (Mahrer, 1956
; Kidd et al., 2012
). Our studies provide the first experimental manipulation of trust while avoiding manipulations of reward, and thus critically demonstrate a causal role for social trust in delaying gratification, independent of other factors that can influence self-control.
Our findings add to a growing literature emphasizing the role of social factors in cognitive processes (Sanfey, 2007
; Bernier et al., 2011
; Meyer et al., 2012
), and indicate the need to revise prominent theories of delay of gratification. Most theories focus on the role of cognitive control, basic valuation, and prospective mechanisms (Peters and Büchel, 2011
). The role of social factors, while raised early on in this domain (Mischel, 1961
), has been largely overlooked in subsequent theorizing and testing. For example, social factors go unmentioned in the burgeoning literature on the neural mechanisms supporting delay of gratification (Wittmann and Paulus, 2008
; Luhmann, 2009
; Peters and Büchel, 2011
). However, our results demonstrate that delaying gratification does not occur in a social vacuum.
Our findings are also relevant to the study of social trust. For example, while existing studies have focused on the consequences of trust for immediate processing and behavior (McCabe et al., 2001
; Delgado et al., 2005
; Kosfeld et al., 2005
; Van den Bos et al., 2011
), the present work demonstrates how a lack of trust may also negatively impact planning for the future. In addition, in both experiments, we find non-linear effects of trust manipulations, like those observed in studies testing other effects of trust. For example, individuals invest less in the Trust Game with partners judged to be untrustworthy, but invest similarly with neutral and trustworthy partners (Fareri et al., 2012
). In our experiments, participants were less willing to delay gratification with untrustworthy partners, and showed less differentiation between neutral and trustworthy partners. Such findings indicate a “threshold” for social trust, where decreases in trust below a certain threshold influence behavior more than increases in trust above that threshold, which seems worthy of further investigation.
Because our studies used trust vignettes and faces and hypothetical choices about whether to delay gratification, future work should build on these findings to examine the ability to actually wait for delayed gratification, with real people and rewards. Hypothetical and real rewards elicit similar patterns of temporal discounting behavior (Madden et al., 2003
) and associated neural activity (Bickel et al., 2009
), but social factors such as trust may matter more in situations involving an actual person rather than a vignette, or actual rewards rather than hypothetical ones. Similarly, hypothetical choices about delaying gratification and actually delaying gratification are correlated (Johnson and Bickel, 2002
; Duckworth and Kern, 2011
), but the influence of social factors may be more apparent when a hypothetical choice is first made rather than when a delayed choice must continue to be abided by (as in traditional delay of gratification paradigms), which may require additional processes, such as inhibitory control. Future studies could address such possibilities by comparing the influences of social trust on delaying gratification with vignettes vs. an experimenter (e.g., who behaves in a trustworthy or untrustworthy manner), with choices for immediate or delayed rewards vs. actually waiting for a delayed reward, and with hypothetical vs. real rewards (e.g., using primary rewards such as food, or randomly selecting and honoring one response from the intertemporal choice task; Reynolds et al., 2003
Social factors suggest intriguing alternative interpretations of prior findings on delay of gratification, and suggest new directions for intervention. For example, the struggles of certain populations, such as addicts, criminals, and youth, might reflect their reduced ability to trust that rewards will be delivered as promised. Such variations in trust might reflect experience (e.g., children have little control over whether parents will provide a promised toy) and predisposition (e.g., with genetic variations predicting trust; Krueger et al., 2012
). Children show little change in their ability to delay gratification across the 2–5 years age range (Beck et al., 2011
), despite dramatic improvements in self-control, indicating that other factors must be at work. The fact that delay of gratification at 4-years predicts successful outcomes years or decades later (Casey et al., 2011
; Shoda et al., 1990
) might reflect the importance of delaying gratification in other processes, or the importance of individual differences in trust from an early age (e.g., Kidd et al., 2012
). From this perspective, emphasizing social trustworthiness might be important in interventions for delaying gratification, not just for increasing accuracy of information collected from individuals with deficits (as emphasized for some interventions, e.g., with juvenile delinquents), but for improving behavior. Testing such possibilities for the role of social trust, and investigating how social and other factors interact, may greatly advance our understanding of the fundamental ability to delay gratification.