正念冥想不会让人更偏向当下
Mindfulness meditation does not make people more present-biased
正念冥想不会让人更偏向当下。正念在临床和企业场景广泛应用,有人担忧关注当下会加重现时偏向;该担忧未获支持。
Breathing in, I calm body and mind. Breathing out, I smile. Dwelling in the present moment I know this is the only moment.
—Thich Nhat Hanh.
Introduction
Mindfulness is no longer confined to yoga studios and meditation retreats; it has become mainstream in boardrooms and executive training programs worldwide. Many Fortune 500 companies—including industry giants like Google, Apple, Nike, and Goldman Sachs—have adopted mindfulness initiatives aimed at sharpening leaders’ attention, reducing stress, and improving decision-making1,2. Yet, despite its widespread popularity and enthusiastic endorsements by millions of professionals, mindfulness practice has recently come under scrutiny for potential unintended consequences. Some scholars and commentators have raised concerns that mindfulness, by enhancing focus on the present moment, might unintentionally encourage present-biased decision-making—leading people to prioritize immediate gratification at the expense of future benefits3,4,5,6. Given mindfulness’s rapid and extensive integration into corporate cultures, rigorously investigating its impact on intertemporal choice is important for both organizational effectiveness and individual well-being.
General effects of mindfulness and potential concerns for intertemporal decision-making
Mindfulness, defined as “the awareness that arises through paying attention, on purpose, in the present moment, non-judgmentally”7, has positive effects that have been well-established across many domains. Mindfulness interventions have been frequently applied in clinical contexts—among groups diagnosed with, or considered at risk for, health-related conditions—to address anxiety, personality disorders, substance abuse, stress, chronic pain, and other conditions8. Several reviews have summarized this literature and generally support the clinical effectiveness of mindfulness-based interventions (see9,10,11,12,13,14,15,16,17, with continued support from recent research (see16,18,19,20,21,22. Moreover, the positive effects of mindfulness are not limited to clinical contexts. Brain imaging studies have found a positive relationship between mindfulness and neural functions related to well-being23,24,25. Other studies have linked mindfulness to greater life satisfaction through reduced deviations from a balanced time perspective21,26, and still others have found positive effects on immune system parameters27,28,29.
Research on the effects of mindfulness on decision-making is much more limited, but studies have also found positive effects in important domains, including ethical decisions30,31, cooperative and pro-social decisions32,33, and the sunk-cost bias34. Mindfulness was also associated with enhanced self-regulation, reduced aversion to waiting35, and lower risk-aversion36. Overall, although mindfulness received some criticism37,38,39, the overarching conclusion is that it has desirable effects on mental health, well-being and decision-making.
While mindfulness may improve ethical, pro-social, and risk-related decisions people make for their present self, could it hurt the intertemporal choices they make for their future self? Mindfulness practices are commonly designed to focus attention on the present moment, reducing attention to concerns related to the past or future. This present-focused attention may have benefits for well-being, but it could also lead people to discount future outcomes more heavily and become more present-biased. This concern is partly grounded in foundational research on the cognitive underpinnings of intertemporal choice. Past studies have shown that episodic future thinking—the mental simulation of future events—significantly reduces delay discounting, making individuals more patient and future-oriented in their choices88. Relatedly, Smallwood, Ruby, and Singer92 found that mind-wandering, which frequently involves future-oriented thought, is associated with reduced delay discounting, suggesting that the capacity to mentally disengage from the present and consider future goals supports more patient decision-making. Mindfulness, by design, trains individuals to redirect attention away from such future-oriented thought and toward present-moment experience. If future thinking is a key mechanism through which people sustain the subjective value of delayed rewards80, then a practice that systematically reduces future-oriented cognition could, in principle, weaken the mental representation of delayed outcomes and thereby strengthen present bias.
A large literature in economics, psychology, and neuroscience has shown that when people make choices that involve intertemporal trade-offs—such as the ones related to saving, climate change mitigation, healthy eating, etc.—they tend to hyperbolically discount the value of future rewards and display a preference for immediate gratification (see40,41. These tendencies have been linked to several detrimental behaviors, including lower academic performance42, insufficient saving43, unhealthy eating and obesity44,45, and substance and alcohol use disorders46. They have also been linked to important life outcomes such as divorce47 and suicide48. If mindfulness were to increase present bias by focusing people more strongly on the present, it could have unintended consequences for decisions involving future costs and benefits.
Some existing research has linked mindfulness with patterns of substance abuse related to impulsivity49. Specifically, neurocognitive models of addiction suggest that attentional biases and dysregulated reward processing are key mechanisms driving impulsive substance use, and that mindfulness-based interventions target these same attentional and appraisal processes90. At the same time, impulsivity-related substance use is robustly associated with steep delay discounting—the tendency to prefer smaller, immediate rewards over larger, delayed ones—as shown by meta-analytic evidence across multiple substance types89. This body of work connects the attentional and reward-valuation processes engaged by mindfulness to the same present-oriented decision tendencies that characterize intertemporal choice, lending further plausibility to the concern that mindfulness could affect temporal discounting. Given the increased prevalence of mindfulness across Western societies, it is important to examine whether this practice affects intertemporal decision-making. Therefore, in the present paper, we aim to systematically study how mindfulness meditation affects intertemporal choice.
Empirical evidence on the effects of mindfulness on intertemporal choice
Several studies have examined the effects of mindfulness or related treatments on intertemporal decision-making in clinical and general populations—with very mixed results for both. At the clinical population level, several studies concluded that mindfulness can successfully reduce delay discounting by improving self-regulation, tolerance, and present-moment acceptance (see19,20,36,50,51,52. However, other studies found null effects53,54, and others argued that mindfulness can even increase discount rates depending on the population under study (see55. For instance, Rasmussen et al.55 found that a mindful eating treatment increased the discounting of rewards for food-insecure women. These findings, which run contrary to the results of previous studies, are interpreted as mindfulness inducing them into a ‘survival mode’ by increasing their focus on present scarcity.
At the general population level, several studies found that mindfulness can reduce discount rates35,50,56,57,58,59,60,61, but that this effect can be contingent upon framing35 and whether the type of reward is primary (e.g. food) or monetary51. These contingent effects are consistent with other studies that found that delay discounting is decreased when time is framed in terms of calendar dates rather than days62,63; and studies that found that primary rewards are discounted more steeply than monetary rewards40,64. Other studies found no significant effects of mindfulness or related treatments on intertemporal choices18,51,65. Finally, some studies suggested that mindfulness induces higher discount rates when making financial decisions55.
Beyond influencing decision outcomes, mindfulness may fundamentally alter the decision-making process itself by enhancing individuals’ metacognitive awareness of their choices and internal decision signals. Recent experimental evidence supports this proposition: Lachaud et al.91 demonstrated that experienced mindfulness practitioners exhibit superior detection of subtle external influences on their decisions compared to non-meditators, suggesting that mindfulness training enhances introspective access and attenuates choice blindness—a bias reflecting diminished metacognitive quality in decision-making. This finding positions the present investigation of mindfulness and intertemporal choice within a broader theoretical framework documenting mindfulness effects on decision processes that extend beyond temporal discounting per se.
Limitations of previous research
Research on the effects of mindfulness on intertemporal choice is still underexplored, and existing studies are limited in at least four ways.
First, most studies focus on very specific clinical populations and decision-making contexts. For instance, Hendrickson and Rasmussen50,51 studied the effects of mindfulness to reduce impulsive food choices to curb obesity. Similarly, Rasmussen et al.55 focused on food-insecure women, while Shead et al.54 focused on regular gamblers, and Yao et al.52 on gaming addiction. The relatively narrow focus of these studies—while important—makes it difficult to draw meaningful conclusions for the hundreds of millions of non-clinical mindfulness practitioners worldwide who make everyday decisions about how to invest their time and money (see19,20,36,50,51,52,53,54,55.
Second, the handful of studies that did examine the relationship between mindfulness and inter-temporal choice in the general population mainly did so using observational data. For instance, Duchêne et al.65, Marcowski et al.36, and Murphy & McKillop20 did not induce mindfulness but estimated statistical relationships based on self-reported traits, using questionnaires like the Five Facet Mindfulness Questionnaire. This makes it impossible to establish causality, especially given socially desirable responding and the conceptual overlap between trait-mindfulness and inter-temporal choice questionnaires.
Third, regardless of the focus and methodological approach of the study, mindfulness manipulations or traits have often been mixed with other factors and interventions (see36,52,53,59,60,61. For instance, Yao et al.52 analyzed the impact of a mindfulness intervention on decision impulsiveness to curb Internet gaming disorder. However, the authors used a combination of classic therapy and mindfulness meditation as a treatment, which did not allow for the isolation of the effects of mindfulness. Similarly, Morrison et al.59 and Morrison et al.53 used acceptance-based training sessions as a treatment to investigate the effects of mindfulness on impulsivity. Although acceptance-based therapy is rooted in the philosophy that underpins the mindfulness movement, in the acceptance-based therapy used in the study, no formal mindfulness training was given to treated participants. Scholten et al.60 also mixed mindfulness and acceptance-based training as a treatment, whereas Wang et al.61 used a broad Buddhist practice intervention that includes some mindfulness but also other elements like chanting. Finally, observational research by Marcowski et al.36 used a composite of trait mindfulness and psychological flexibility scores to create behavioral profiles and relate them to discounting.
Fourth, existing studies often exclusively used very short mindfulness treatments and focused on measuring the immediate effects of mindfulness on intertemporal decision-making (see35,54,56,57,58. For instance, Bazley et al.56, Dixon et al.57, and Errmann35 used mindfulness video or audio inductions of just about 5 min, casting doubt on the conceptual validity of the manipulation and providing no insights into the medium- and long-term effects of mindfulness. To the best of our knowledge, only Alem et al.18 has investigated the effects of a fully-fledged mindfulness program (conducted online for four weeks) on intertemporal choices—finding some patterns suggestive of reduced time discounting but no significant results.
We present four studies—three lab studies and one longitudinal field study—that address the aforementioned gaps and improve upon the existing literature in five main ways. Studies 1, 2, and 3 were laboratory studies that manipulated mindfulness using short-term induction paradigms. In these studies, we compared the effects of mindfulness meditation against specific control conditions: a passive control group (where participants simply performed the decision tasks) and/or an active ‘mind-wandering’ control group (where participants listened to an audio clip instructing participants to let their mind wander). This allowed us to isolate the specific effects of mindfulness from general attention or mind-wandering effects.
First, expanding on previous studies on the topic, our research combines both manipulation- and training-based approaches. This is important because we can compare the effects of momentary interventions that change the current state of the participants to the consequences of a longer program that may reshape behavior at a deeper level.
Second, we also combine our manipulation- and training-based approaches with the measurement of participants’ trait mindfulness levels and experience in mindfulness practice. This allows us to analyze if either personal dispositions or experience over longer periods of time are associated with any particular effects.
Third, our research uses three different tasks to measure intertemporal choice behaviors. In Studies 1 and 4, we used a reduced number of classic choices between smaller amounts of money sooner and larger amounts of money later, designed to measure both intertemporal preferences and time inconsistency. In Study 3, we implemented a more sophisticated design involving more choices that allows us to estimate model parameters related to time discounting and to preferences for immediate gratification. In Study 2, we used hypothetical real-world scenarios to analyze intertemporal decisions in more realistic settings. This combination of tasks provides a relatively comprehensive analysis of inter-temporal decision making.
Fourth, in our research, we use a mixture of hypothetical and real choices (i.e., participants knew that in certain tasks, they could actually obtain the outcome that they chose at the proposed time). This provides a test of the effects of mindfulness on intertemporal choice across different incentive systems.
Finally, Study 4 moved beyond the laboratory to test the ecological validity of these findings in a real-world setting. We partnered with a provider of the Mindfulness-Based Stress Reduction (MBSR) program to conduct a longitudinal field study. In this study, we used a within-subjects design, measuring participants’ intertemporal choices before (pre-treatment) and after (post-treatment) completing the intensive 8-week course. This allowed us to assess whether longer-term, intensive mindfulness training—rather than just a brief induction—altered intertemporal preferences within the same individuals. This improves on the most robust mindfulness manipulation used until now in the domain of intertemporal choice (see18, which lasted for four weeks and was conducted online.
The central concern we test is whether a brief mindfulness induction increases impatience and present bias, shifting intertemporal preferences toward smaller, sooner rewards. Because mindfulness could in principle move intertemporal choice in either direction, we evaluate this concern with two-sided tests throughout. It is tested in each of the three induction studies (Studies 1–3) using the main dependent variable of each study: proportion of patient choices (Study 1), proportion of money allocated to longer-term saving options and proportion choosing the waiting alternative (Study 2), and the structurally estimated discount factor δ, present-bias parameter β, and prevalence of present-biased preferences (Study 3). In Study 4 (field study), the test is the pre–post comparison of patient choices following an 8-week MBSR course. All remaining analyses—including decompositions by item type, time-consistency analyses, mixed-effects models and session-clustered robustness checks, covariate-adjusted regressions, and interactions with trait mindfulness—are considered exploratory. This distinction is maintained throughout the Results sections.
Study 1
Method
We assigned 323 participants—in two different waves—(219 women, 104 men; mean age = 21 years, range = 18–51 years; any meditation experience = 41.5%) to mindfulness, mind-wandering, and control conditions, randomizing per session. Participants were students and local residents from the Pompeu Fabra University Behavioral and Experimental Sciences Laboratory participant pool who responded to an advertisement offering 8€ for participation. Each participant sat in a semi-private cubicle within a laboratory. Our mindfulness and mind-wandering induction procedures drew on established methods7,34,66,67. All methods were carried out in accordance with relevant guidelines and regulations. Informed consent was obtained from all participants involved in the studies. A sensitivity analysis indicated 80% power (α = 0.05) to detect a small-to-medium effect of Cohen’s d = 0.38; full per-comparison calculations are reported in Supplementary Materials, Section S1.1.
The study was conducted in a computer laboratory across five data-collection days (June 14 and 15, July 1, and October 10 and 14, 2016), with multiple sessions per day. Each session consisted of a group of participants who completed the study simultaneously. All participants within a given session received the same experimental induction (mindfulness meditation, mind-wandering, or control audio). The order of conditions across sessions within each day was pseudorandomized: we systematically varied which treatment was administered in each time slot using a balanced rotation, ensuring that each condition appeared at different times of the day across the data-collection period. This design yielded 18 sessions in total (6 per treatment condition), with session sizes ranging from 3 to 28 participants (Mdn = 17).
Participants listened to a 15-min audio-recorded induction created specifically for this research by a professional mindfulness-meditation instructor. Participants were led through a focused-breathing meditation exercise that instructed them to focus on the physical sensations of breath entering and leaving their body and repeatedly reminded them to focus on their experience of breathing. The content of the mind-wandering induction repeatedly instructed participants to think of whatever came to mind. Although this type of induction has commonly been used as a control condition in prior mindfulness studies34,66,67 because it replicates a waking, baseline mental state68, research has shown that such “forced” mind-wandering can be aversive (e.g69.,. Therefore, going beyond prior studies, we also included a true control condition in which participants were not subjected to any manipulation, ensuring that observed mindfulness effects were not merely driven by negative reactions to the mind-wandering condition. Thus, the study employed a three-group design comprising the mindfulness treatment, the mind-wandering active control, and the passive control.
Then, participants in the first of the two waves of the study, in which we had a total of 119 participants (78 women, 41 men; mean age = 22 years; age range = 18–51), had to complete three manipulation check items in order to advance in the study—something we did not include in the second wave and in the rest of the studies since we demonstrated the effectiveness of our manipulations in this first wave. The three items are included in the instructions in our online repository. The first two items were designed to test to which extent participants had been focused on the present moment or on the physical sensations in their bodies. The third item was designed to test to which extent participants had been mind-wandering. We measured the three items with a 5-point Likert scale.
Finally, all participants made four hypothetical choices, presented in Table 1, between two assets. Asset A always offered a smaller amount of money (200€) sooner in time (either now or in 12 weeks). Asset B always gave a larger amount of money (either 220€ or 250€) at a later time (either in 4 or in 16 weeks). In two of the choices, Asset A offered amounts in the present moment, while in the other two choices, both Asset A and B offered delayed monetary rewards. Thus, participants made choices between smaller immediate rewards and larger later rewards or between smaller later rewards and larger even later rewards. Additionally, by offering the same monetary amounts with the same temporal distance between them at different moments in time, we can compare whether participants behaved in a time-consistent manner. Time inconsistency is one of the central phenomena in research on intertemporal decision making, and it typically takes the form of choosing the smaller sooner amount when it is close to the present (as in choices 1 and 3) and switching to the larger later reward when the same options are delayed into the future (as in choices 2 and 4) (see41. The study also included other items, mostly related to risk aversion, ambiguity aversion, and loss aversion, not relevant to the current study.
The studies reported in the paper were conducted as part of the project “Harnessing Collective Wisdom to Improve Individual Decision-Making in Complex Social Environments” (ID: PSI2013-41909-P), which was approved and funded by the Spanish Ministry of Economy and Competitiveness.
Results and discussion
When reporting the results, we supplement frequentist tests with standardized effect sizes (Cohen’s d for continuous outcomes, Cohen’s h for proportions), 95% confidence intervals, and BIC-based Bayes factors (BF₀₁83, where BF₀₁ > 3 indicates moderate, > 10 strong, and > 30 very strong evidence for the null85,86. Categorical outcomes are compared across conditions using Pearson chi-square tests of independence; for all two-group (2 × 2) comparisons we apply Yates’s continuity correction and report the statistic as χ²(1, N), where N is the combined size of the two groups compared. Because the four binary intertemporal choices are clustered within participants, all analyses are conducted at the participant level. For the continuous outcome (proportion of patient choices), we report one-way ANOVAs and pairwise t-tests on per-participant proportions, together with OLS regressions using cluster-robust standard errors at the session level (18 clusters; these standard errors allow observations within the same session to be correlated). For the binary time-consistency outcome, we use logistic regression with session-clustered standard errors. As a robustness check, we also estimate a mixed-effects logistic regression (a generalized linear mixed model) with a random intercept for each participant (full specification in Supplementary Materials, Section S1.3). The primary hypothesis-driven test is the omnibus ANOVA and pairwise comparisons on the participant-level proportion of patient choices; all other analyses (item-type decompositions, time consistency, mixed-effects models, session-clustered regressions, and session-level aggregations) are exploratory.
Manipulation check
Both our mindfulness and mind-wandering manipulations were effective. Participants in the mindfulness condition reported being significantly more focused on the present moment (mean = 3.64) than participants in the mind-wandering condition (mean = 2.61; t = 4.87, p <.001) and the control condition (mean = 2.98; t = 3.00, p =.004). There was no significant difference between the control and mind-wandering conditions (t = 1.41, p =.162). Likewise, participants in the mindfulness condition reported being significantly more focused on their body sensations (mean = 3.73) than participants in the mind-wandering condition (mean = 1.95; t = 8.81, p <.001) and control condition (mean = 2.30; t = 6.35, p <.001). Again, there was no significant difference between the control and mind-wandering conditions (t = 1.36, p =.177). Finally, participants in the mind-wandering condition reported more mind-wandering (mean = 3.47) than participants in the mindfulness condition (mean = 2.82; t = 2.76, p =.007) and control condition (mean = 2.83; t = 2.47, p =.015). There was no significant difference between the mindfulness and control conditions (t = − 0.05, p =.960). Regarding effect sizes, the mindfulness manipulation produced medium-to-large effects on the targeted constructs: mindfulness vs. mind-wandering on present-moment focus, d = 0.67, 95% CI [0.39, 0.95]; mindfulness vs. control, d = 0.40, 95% CI [0.14, 0.67]. For body-sensation focus: mindfulness vs. mind-wandering, d = 1.21, 95% CI [0.92, 1.50]; mindfulness vs. control, d = 0.86, 95% CI [0.58, 1.13]. For mind-wandering: mind-wandering vs. mindfulness, d = 0.38, 95% CI [0.11, 0.65]; mind-wandering vs. control, d = 0.34, 95% CI [0.07, 0.61].
Time Preference [hypothesis-driven analysis]
As depicted in Fig. 1, most participants preferred more money later over less money sooner, regardless of the experimental condition. Throughout, we refer to choices of the larger, later reward interchangeably as patient, delayed, or distant choices. Participants’ proportion of patient (delayed) choices across the four intertemporal tasks did not differ significantly between conditions: mindfulness (M = 0.713, SD = 0.340), control (M = 0.723, SD = 0.329), and mind-wandering (M = 0.731, SD = 0.324), F(2, 320) = 0.07, p =.930, η² < 0.001. A nonparametric Kruskal-Wallis test confirmed this null result, H(2) = 0.04, p =.979. Pairwise comparisons revealed no significant differences: mindfulness vs. mind-wandering, t(210) = − 0.38, p =.706, d = − 0.05, 95% CI [− 0.32, 0.22], BF₀₁ = 13.55; mindfulness vs. control, t(218) = − 0.21, p =.831, d = − 0.03, 95% CI [− 0.29, 0.24], BF₀₁ = 14.50; mind-wandering vs. control, t(212) = 0.17, p =.865, d = + 0.02, 95% CI [− 0.25, 0.29], BF₀₁ = 14.42. All Bayes factors exceeded 10, providing strong evidence for the null hypothesis that mindfulness induction did not affect intertemporal choice.
Proportion of distant choices per condition. This figure shows the proportion of distant choices combining the four intertemporal decisions as a function of the condition. Larger values indicate a greater proportion of distant choices. Error bars indicate the 95% confidence interval.
Present vs. future options [exploratory]
A similar pattern emerged when separately analyzing items that include and do not include a present option (Fig. 2). When analyzing choices involving an immediate reward (choices 1 and 3), no treatment differences emerged: mindfulness (M = 0.688, SD = 0.396), control (M = 0.703, SD = 0.372), mind-wandering (M = 0.709, SD = 0.361), F(2, 320) = 0.09, p =.918, η² < 0.001. All pairwise d values were trivially small (all |d| < 0.06, all BF₀₁ > 13). For tasks without an immediate option (tasks 2 and 4): mindfulness (M = 0.739, SD = 0.376), control (M = 0.743, SD = 0.375), mind-wandering (M = 0.752, SD = 0.376), F(2, 320) = 0.04, p =.963. All pairwise |d| < 0.04, all BF₀₁ > 14. In every case, effect sizes were negligible and Bayes Factors provided strong evidence for the null.
Proportion of distant choices per condition for choices between immediate vs. future rewards (Panel A) and near vs. distant future rewards (Panel B). Error bars indicate the 95% confidence interval.
Time consistency [exploratory]
In Study 1, participants could potentially be time inconsistent two times, one comparing choices 1 and 2 and another comparing choices 3 and 4. To analyze if there were differences between conditions in this respect, we gave participants a time-consistency score of 1 if their choices were fully time-consistent and zero otherwise. The proportion of participants exhibiting fully consistent preferences, depicted in Fig. 3, did not differ across conditions, as assessed by a Pearson chi-square test of independence: 67.0% in the mindfulness group, 72.1% in the control group, and 65.0% in the mind-wandering group, χ²(2, N = 323) = 1.31, p =.519. Pairwise comparisons yielded no significant differences: mindfulness vs. mind-wandering, χ²(1, N = 212) = 0.09, p =.768, h = + 0.04, 95% CI [− 0.23, 0.31], BF₀₁ = 13.94; mindfulness vs. control, χ²(1, N = 220) = 0.68, p =.411, h = − 0.11, 95% CI [− 0.38, 0.15], BF₀₁ = 10.58; mind-wandering vs. control, χ²(1, N = 214) = 1.23, p =.268, h = − 0.15, 95% CI [− 0.42, 0.12], BF₀₁ = 7.93. All Bayes factors indicated moderate to strong evidence for the null. Logistic regression with session-clustered standard errors confirmed the absence of treatment effects on consistency: mindfulness vs. control, b = − 0.24, SE = 0.20, z = − 1.21, p =.225, OR = 0.79; mind-wandering vs. control, b = − 0.33, SE = 0.20, z = − 1.64, p =.102, OR = 0.72.
Proportion of time consistent participants per condition. This figure shows the proportion of time consistent participants. We define a participant as time consistent if they chose the same asset selections in the first and second choices and in the third and fourth choices. Large numbers indicate a greater proportion of time-consistent choices. Error bars indicate the 95% confidence interval.
Robustness
Three robustness analyses confirmed the null: a mixed-effects logistic regression with a random intercept for participant (BF₀₁ = 17.58–17.93), an OLS regression with session-clustered standard errors (with and without age and gender covariates), and a session-level ANOVA on the 18 session means. Full results, including the regression table, are reported in Supplementary Materials, Section S1.3.
Summary of effect sizes and Bayesian evidence
Across all outcomes and comparisons in Study 1, standardized effect sizes were consistently negligible: Cohen’s d ranged in absolute value from 0.01 to 0.06 for continuous outcomes, and Cohen’s h from 0.04 to 0.15 for binary proportions. Every 95% confidence interval included zero. BIC-based Bayes Factors83 uniformly favoured the null hypothesis: BF₀₁ values ranged from 7.93 (moderate evidence) to 14.50 (strong evidence) for participant-level pairwise comparisons, with the majority exceeding the conventional threshold of 10 for strong support of H₀85,86. Mixed-effects models accounting for within-participant clustering yielded even stronger Bayesian evidence (BF₀₁ = 17.58–17.93). These results establish that the absence of mindfulness effects on intertemporal choice in Study 1 reflects a genuine null finding rather than statistical insensitivity. The full pairwise table of effect sizes, 95% CIs, and Bayes factors is reported in Supplementary Materials, Table S1.1.
Overall, Study 1 shows that a 15-min mindfulness meditation did not increase the preference for short-term gratification using a typical task to study intertemporal decision making. This is true both when considering differences in choices involving all types of delays and when breaking down decisions between immediate vs. future alternatives and near vs. distant future alternatives. Moreover, mindfulness did not change the consistency of people’s time preferences. Finally, all these null results were robust to including participants’ age and gender as control variables (see Tables S1.2 and S1.3 in the Supplementary Materials, located in our online repository).
Study 2
Method
To investigate the effects of mindfulness beyond the classical intertemporal tasks found in the decision-making literature, we decided to use additional tasks inspired by real-world scenarios. In Study 2, our procedures were generally the same as in Study 1. We assigned 140 participants (83 women, 57 men; mean age = 21 years; age range = 18–37; any meditation experience = 39.3%) to a mindfulness, a mind-wandering, or a control condition, randomized per session. A sensitivity analysis indicated 80% power (α = 0.05) to detect a Cohen’s d of 0.58–0.60; full calculations are reported in Supplementary Materials, Section S2.1. Participants in the mindfulness and mind-wandering conditions completed the same mindfulness or mind-wandering induction procedures as in Study 1 and all procedures were conducted in accordance with the same standards described for Study 1.
Next, all participants completed three hypothetical resource allocation scenarios. Participants were first asked to imagine that they had won 10,000€ and to indicate how much of this money they would allocate to immediate hedonic spending and how much they would save in a checking account for a rainy day. Then participants were told to imagine that they earned 2,000€ net per month (a decent salary at the time for our Spanish undergraduate participants) and indicate how they would allocate this monthly income between saving in a checking account, pension plan, rent and home expenses, leisure, food, and other expenses (all summing to 2,000€). Finally, participants were asked to imagine a scenario where they needed to buy a car but had only half the necessary funds. They were presented with two options: borrow the remaining amount from a bank and pay interest, or wait two years until they had saved enough, avoiding any interest payments (see the instructions in the online repository for full details).
These three tasks capture in different ways the trade-off between smaller short-term rewards and larger longer-term ones. In the first task, allocating more money to savings indicates a stronger preference for the long-term rewards, which is typically associated with less time discounting. In the second task, the options that represent longer-term rewards are saving and investing in a pension plan. In the third one, the longer-term option is waiting to avoid paying interest. These tasks were not directly incentivized; participants received a fixed payment for participation. The study also included other items, mostly related to risk aversion, ambiguity aversion, and loss aversion, not relevant to the current study. Psychometric properties of the different measures are reported in Supplementary Materials, Section S2.2.
Results and discussion
We follow the same analytical approach as in Study 1. Because Study 2 includes only 6 sessions (2 per condition), cluster-robust standard errors are imprecise84; we therefore rely on participant-level tests as the primary inferential approach, with session-level robustness checks reported in Supplementary Materials, Section S2.6. The primary hypothesis-driven tests are pairwise comparisons of (a) the proportion of money allocated to longer-term saving across Scenarios 1 and 2, and (b) the proportion of participants choosing the waiting alternative in Scenario 3. Full pairwise and sensitivity calculations are reported in Supplementary Materials, Section S2.3.
Saving behavior [hypothesis-driven]
Before examining treatment effects, we verified that the saving-allocation tasks were sensitive to meaningful variation. The within-person comparison between Scenarios 1 and 2 yielded a very large effect (paired t(138) = 20.09, p <.001, dz = 1.70), confirming that the allocation tasks discriminate between different saving contexts. A full sensitivity analysis (per-scenario means and SDs, the paired t-test, and an explanation of how this large within-person effect rules out low task sensitivity as an explanation for the null effect of mindfulness) is reported in Supplementary Materials, Section S2.4. The proportion of money allocated to longer-term saving options across Scenarios 1 and 2 depicted in Fig. 4 (N = 139; one participant excluded due to missing allocation data) differed marginally across conditions: mindfulness (M = 0.669, SD = 0.175, n = 41), control (M = 0.705, SD = 0.146, n = 47), and mind-wandering (M = 0.621, SD = 0.180, n = 51), F(2, 136) = 3.07, p =.050, η² = 0.043. A Kruskal-Wallis test confirmed this pattern, H(2) = 7.90, p =.019. Pairwise comparisons revealed that participants in the mind-wandering condition allocated a significantly lower proportion to longer-term options than participants in the control condition, t(96) = − 2.51, p =.014, d = − 0.51, 95% CI [− 0.91, − 0.11], BF₀₁ = 0.42. Although the effect size is moderate, the Bayesian evidence favouring an effect of mind-wandering on saving is only anecdotal (BF₁₀ ≈ 2.4) by the thresholds adopted here85,86. By contrast, the mindfulness condition did not differ significantly from the control condition, t(86) = − 1.03, p =.306, d = − 0.22, 95% CI [− 0.64, 0.20], BF₀₁ = 5.52, or from the mind-wandering condition, t(90) = 1.30, p =.199, d = + 0.27, 95% CI [− 0.14, 0.69], BF₀₁ = 4.15. OLS regression with heteroskedasticity-robust (HC1) standard errors confirmed the mind-wandering vs. control difference: mind-wandering b = − 0.084, SE = 0.033, t = − 2.53, p =.011; mindfulness b = − 0.035, SE = 0.035, t = − 1.02, p =.308. The full pairwise table of effect sizes, 95% CIs, Bayes factors, and HC1-robust regression coefficients is reported in Supplementary Materials, Table S2.1.
Proportion of money allocated to longer-term options per condition in tasks 1 and 2. This figure shows the average proportion of money allocated to the longer-term options as a function of the condition. Larger values indicate greater allocations to the long term. Error bars indicate the 95% confidence interval.
Saving behavior: Scenario 3 (waiting choice) [hypothesis-driven]
The proportion of participants choosing to save in Scenario 3, depicted in Fig. 5, did not differ across conditions: 82.9% in the mindfulness group (n = 41), 80.9% in the control group (n = 47), and 76.9% in the mind-wandering group (n = 52), χ²(2, N = 140) = 0.55, p =.760. Pairwise comparisons revealed no significant differences: mindfulness vs. mind-wandering, χ²(1, N = 93) = 0.21, p =.650, h = + 0.15, 95% CI [− 0.26, 0.56], BF₀₁ = 8.70; mindfulness vs. control, χ²(1, N = 88) = 0.00, p >.999, h = + 0.05, 95% CI [− 0.37, 0.47], BF₀₁ = 9.38; mind-wandering vs. control, χ²(1, N = 99) = 0.05, p =.817, h = − 0.10, 95% CI [− 0.49, 0.30], BF₀₁ = 9.69. All Bayes factors indicate moderate to strong evidence for the null hypothesis. A session-clustered logistic regression confirmed the null: mindfulness b = 0.14, SE = 0.42, z = 0.33, p =.740, OR = 1.15; mind-wandering b = − 0.24, SE = 0.52, z = − 0.45, p =.652, OR = 0.79.
Proportion of participants who chose the waiting alternative per condition in task 3. This figure shows the proportion of waiting decisions per condition in task 3. Large numbers indicate more waiting. Error bars indicate the 95% confidence interval.
Robustness: Session-level analysis
A session-level robustness analysis on the 6 session means (2 per condition) confirmed the null for both outcomes; full results are reported in Supplementary Materials, Section S2.6.
Summary of effect sizes
For the primary mindfulness comparison, effect sizes were small to negligible across all outcomes: mindfulness vs. control d = − 0.22 (proportion saved) and h = + 0.05 (Scenario 3), with Bayes factors of 5.52 and 9.38, respectively. The mind-wandering vs. control comparison yielded a moderate effect on proportion saved (d = − 0.51, BF₀₁ = 0.42), but no effect on the Scenario 3 saving choice (h = − 0.10, BF₀₁ = 9.69). Importantly, the saving-allocation tasks demonstrated strong structural sensitivity (within-person scenario effect dz = 1.70) and could detect a genuine between-group effect (MW vs. control d = − 0.51), confirming that the null mindfulness results reflect an absence of effect rather than task insensitivity.
The results of Study 2 replicate the findings from Study 1 using more naturalistic scenarios: A 15-min mindfulness meditation did not increase people’s propensity to pursue immediate gratification over longer-term rewards. These results hold true even when controlling for age, gender, and previous meditation practice (see Table S2.2 in the Supplementary Materials). However, Studies 1 and 2 relied on hypothetical decisions, which may not fully reflect actual decision-making behavior77. In Study 3, we address this limitation by examining whether mindfulness affects intertemporal choices in consequential, incentive-compatible contexts. Going beyond basic choice scenarios, we used a more comprehensive measure of intertemporal choice, allowing us to separately measure how people choose between delayed rewards and how strongly they favor immediate rewards over delayed ones (present bias).
Study 3
Method
We assigned 389 participants (263 women, 126 men; mean age = 21 years; age range = 18–51; any meditation experience = 41.6%) to mindfulness (n = 121), mind-wandering (n = 137), and control (n = 131) conditions, randomizing per session. Participants in the mindfulness and mind-wandering conditions completed the same mindfulness or mind-wandering induction procedure as in Studies 1 and 2. A sensitivity analysis indicated 80% power (α = 0.05) to detect a Cohen’s d of 0.35; full calculations are reported in Supplementary Materials, Section S3.1. Again, participants in the control condition were not subjected to any procedure and all procedures were conducted in accordance with the same standards described for Study 1.
Next, all participants made 42 choices as in McClure et al.70 between receiving smaller cash amounts (between 5€ and 34€) earlier (immediately, two weeks from the day of the study, or four weeks from the day of the study) and larger cash amounts (between 7€ and 43€) later (two, four, or six weeks, respectively, from the day of the study). The full list of the 42 choices is available in the instructions in our online repository. Participants made choices between smaller immediate rewards and larger later rewards, or between smaller later rewards and larger even later rewards. We incentivized participants to express their true preferences by randomly selecting 1 out of every 25 participants to realize one of their choices, paying the preferred alternative for a randomly selected choice pair. The payment was made at the chosen moment and for the chosen amount using electronically sent Amazon.com gift certificates. The study also included other items, mostly related to risk aversion, ambiguity aversion, and loss aversion, not relevant to the current study.
This task allows us to estimate the parameters of a model that distinguishes between two types of processes, as captured by the following quasi-hyperbolic discounting function71,72:
$$D(t)=\begin{cases}1 & \text{if } t=0\\ \beta\delta^{t} & \text{if } t>0\end{cases}$$
.
One of the parameters of the function (δ) captures time discounting when choosing between delayed rewards, assuming exponential discounting. The other parameter (β) captures what is known as present bias: the additional discounting occurring between rewards in the present and any delayed rewards. This bias occurs when 0 < β < 1, and it is also associated with time-inconsistent behavior.
The 42-item battery showed excellent internal consistency (Cronbach’s α = 0.962). Full psychometric properties are reported in Supplementary Materials, Section S3.2.
Results and discussion
We follow the analytical approach used in Studies 1 and 2. Treatment was assigned at the session level (18 sessions, 6 per condition), and a sensitivity power analysis indicated 80% power (α = 0.05) to detect effects of d ≥ 0.34–0.35 across the three pairwise condition comparisons. Unlike Studies 1 and 4, Study 3 is not affected by the inflated-N concern, since the dependent variables—the long-run discount factor δ, the present-bias parameter β, and the present-bias classification (β < 1)—are person-level structural estimates derived from 42 choices rather than single choice observations. These three outcomes are the primary hypothesis-driven tests; session-clustered regressions and session-level aggregated ANOVAs serve as exploratory robustness checks.
To test whether the parameters differ between the three conditions, we estimated them by fitting each participant’s choices to the quasi-hyperbolic discounting function using maximum-likelihood estimation, constraining β and δ to be between 0 and 1. This approach generated two parameters per participant, allowing us to compare the distribution of β and δ across conditions.
Present-bias prevalence [hypothesis-driven]
To begin with, Fig. 6 presents the proportion of participants who showed present bias (by displaying a β lower than 1) in the different conditions. β was statistically equally likely to be lower than 1 among mindfulness participants (74%), mind-wandering participants (77%), and control participants (82%). Pairwise comparisons revealed no significant differences in present-bias prevalence: mindfulness vs. control, χ²(1, N = 252) = 1.96, p =.162, h = − 0.20, 95% CI [− 0.44, 0.05], BF₀₁ = 5.97; mindfulness vs. mind-wandering, χ²(1, N = 258) = 0.32, p =.571, h = − 0.09, 95% CI [− 0.33, 0.16], BF₀₁ = 13.68; mind-wandering vs. control, χ²(1, N = 268) = 0.52, p =.471, h = − 0.11, 95% CI [− 0.35, 0.13], BF₀₁ = 12.62. All Bayes factors exceed 5.9, providing moderate to strong evidence for the null. The full pairwise table of effect sizes, 95% CIs, and Bayes factors for δ, β, and present-bias prevalence is reported in Supplementary Materials, Table S3.1.
Proportion of present-biased participants per condition. This figure shows the proportion of individuals displaying present bias (β parameters lower than 1) as a function of the condition. Error bars indicate the 95% confidence interval.
Discount and present-bias parameters [hypothesis-driven]
Figure 7 shows the mean values for both parameters (β and δ) in the three conditions. When we compare conditions focusing on discounting in choosing between delayed rewards (δ), we observe that mindful participants (mean δ = 0.310, median δ = 0.154) discounted similarly to mind-wandering participants (mean δ = 0.245, median δ = 0.104; W = 9238, p =.113). Because the distributions of β and δ are markedly skewed (see Figs. 7 and 8), violating the normality assumption underlying the parametric t-test, we compared conditions using the distribution-free Mann–Whitney U test, which makes no assumption about the shape of the distribution. The results show there were also no differences in terms of δ between mindful participants and control participants (mean δ = 0.256, median δ = 0.126; W = 8757, p =.151) or between mind-wandering and control participants (W = 9045, p =.912). The Bayes factors for the mindfulness comparisons (BF₀₁ = 4.62 and 7.14) provide moderate null evidence, while the MW vs. control comparison (BF₀₁ = 15.74) yields strong null evidence.
Average individual discount factor (δ) (left graph) and present bias (β) (right graph) as a function of condition. Large numbers indicate more patience. Error bars indicate the 95% confidence interval.
Focusing on the degree of present bias (β), we see that mindful participants (mean β = 0.972, median β = 0.989) did not significantly differ from mind-wandering participants (mean β = 0.951, median β = 0.987; W = 8629, p =.567), or control participants (mean β = 0.951, median β = 0.993; W = 8282, p =.536). The level of present bias in mind-wandering (mean β = 0.951, median β = 0.987) and control participants did not significantly differ either (W = 8581, p >.999). The Bayes factors for the mindfulness comparisons (BF₀₁ = 2.91 and 3.00) represent borderline anecdotal-to-moderate null evidence, while the MW vs. control comparison yields strong null evidence.
To further test whether the distributions of individual parameters of participants in the three conditions differed, we conducted a series of Kolmogorov-Smirnov tests. As depicted in Fig. 8, the distribution of the δ parameters did not differ between the mindfulness and control conditions (D(252) = 0.11; p =.359), the mindfulness and the mind-wandering conditions (D(258) = 0.11; p =.367), or between the control and mind-wandering conditions (D(268) = 0.07; p =.900).
Histogram of δ per condition. This figure shows the histogram plot of the discount factor (δ), plotted as a function of the condition.
Similarly, as depicted in Fig. 9, Kolmogorov-Smirnov tests for the distribution of the β parameters showed no difference between the mindfulness and control conditions (D(252) = 0.09; p =.664), the mindfulness and the mind-wandering conditions (D(258) = 0.06; p =.956), or between the control and mind-wandering conditions (D(268) = 0.10; p =.476).
Histogram of β per condition. This figure shows the histogram plot of the present bias parameter (β), plotted as a function of the condition.
Taken together, these results suggest that while we observe a small average trend in the individual parameter estimations, whereby mindfulness slightly decreased the desire to get something sooner, we did not observe significant differences between the three experimental conditions in our tests. Session-clustered regressions (OLS for δ and β; logistic for present-bias prevalence) and session-level ANOVAs on the 18 session means confirmed the null. Full results, including the regression table, are reported in Supplementary Materials, Section S3.4 (Table S3.2).
Summary of effect sizes and Bayesian evidence
For the primary mindfulness comparison, effect sizes were small across all three dependent variables: d = 0.16 (δ), d = 0.23 (β), and h = − 0.20 (present bias prevalence), relative to the control group. The 95% CIs for all comparisons include zero. Bayes factors ranged from 3.00 to 7.14 for the mindfulness comparisons, providing anecdotal to moderate evidence for the null hypothesis.
In sum, Study 3 replicates the null effect of mindfulness on intertemporal choice using incentivized decisions and a more sophisticated estimation approach. One possibility, however, is that our short-term mindfulness induction may not have been powerful enough to fundamentally shift underlying decision-making patterns. To address this potential limitation, Study 4 moves beyond brief experimental interventions, exploring whether a longer, structured mindfulness training—the classic eight-week Mindfulness-Based Stress Reduction (MBSR) program—can meaningfully influence intertemporal choices.
Study 4
Method
We partnered with the largest provider of the Mindfulness-Based Stress Reduction (MBSR) program in Spain to conduct a field study testing the effects of mindfulness on intertemporal choices. MBSR is a standardized 8-week therapeutic intervention developed by Jon Kabat-Zinn7 that teaches mindfulness through yoga and meditation, and it is the most popular standardized mindfulness program worldwide73. The intervention was facilitated by the director of Instituto EsMindfulness, a Doctor in Psychology with more than 20 years of experience and a certified MBSR instructor by the Center for Mindfulness in Medicine, Health Care and Society, University of Massachusetts Medical School, USA. The program meets the Standards of Practice of the Center for Mindfulness, recognized by the Spanish Association of Mindfulness Instructors (http://www.mbsr-instructores.org/). The intervention is based on exercises to cultivate awareness of internal present-moment experiences with an accepting, non-judgmental stance, along with stress management techniques and other pleasurable activities focused on enhancing adherence to the program. It consists of 8 weeks of group classes with 15–20 participants, one full day, and daily home practice of 45 min. Sessions include didactic presentations, formal meditation practices, and group discussion, during which mindfulness meditation skills are taught to help participants discover relationships between mindful practice and an increased ability to deal effectively with stress. Home practice consists of 45 min of daily formal practice (e.g., sitting and walking meditation, body scan, and yoga stretching) and informal practice (e.g., mindfulness of daily activities and the 3-minute breathing space). Participants received MP3s/CDs of formal meditations for home practice, along with a book and a notebook with relevant readings and space to register their training.
Because random assignment to a separate control group was not ethically or practically feasible within this fee-based field partnership, we used a longitudinal within-subjects design, measuring participants’ intertemporal choices at two time points: Pre-treatment, during the initial orientation session before the course began, and Post-treatment, during the final session of the 8-week program—specifically, after a guided meditation midway through the two-hour session. This design allows us to detect changes in decision-making relative to each participant’s baseline. Participation was voluntary and involved no direct monetary compensation; participants who completed both assessments received a mindfulness book authored by the center’s director as a token of appreciation.
A total of 57 participants completed both the pre-treatment and post-treatment waves (the attrition in the second wave was 35%). Two of them did not provide complete data on the MAAS and meditation-practice measures. Analyses involving these covariates—the manipulation check and the covariate-adjusted regressions—are therefore based on the 55 participants with complete data (36 women, 19 men; mean age = 43 years; age range = 23–69; any meditation experience = 52.7%; months of meditation experience of participants with any experience mean = 24 months, median = 6 months), whereas the unadjusted pre–post comparisons of intertemporal choice use all 57 respondents. A sensitivity analysis indicated 80% power (α = 0.05) to detect a within-subjects effect of Cohen’s dz = 0.38; full calculations are reported in Supplementary Materials, Section S4.1. All methods were carried out in accordance with the relevant guidelines and regulations. Informed consent was obtained from all participants involved in the studies.
For practical reasons, we used the short 4-item measure of intertemporal choice from Study 1 in a booklet format (see the online repository). We also elicited trait mindfulness using the Mindful Attention Awareness Scale (MAAS), developed by Brown and Ryan87. To verify the effectiveness of the intervention (manipulation check), participants completed the MAAS at both the pre-test (Orientation Session) and the post-test (Final Session) to assess changes in trait mindfulness over the 8-week period. The MAAS is a 15-item self-report questionnaire designed to assess an individual’s dispositional tendency to attend to and be aware of present-moment experiences in daily life. Respondents rate items on a 6-point Likert scale ranging from 1 (“almost always”) to 6 (“almost never”), with higher average scores indicating greater trait mindfulness. The MAAS has been widely validated and is commonly used in psychological and behavioral research to quantify individual differences in mindful attention and awareness. In addition, we asked participants two questions related to their prior mindfulness meditation experience: whether they had ever practiced mindfulness before the program, and for how many months they had practiced mindfulness. The study also included other items, mostly related to risk aversion, ambiguity aversion, and loss aversion, not relevant to the current study.
The 4-item task showed good internal consistency (Cronbach’s α = 0.876 and 0.889 across the two waves) and structural sensitivity (within-person item-discrimination dz = 0.50 and 0.60), with test–retest reliability r =.699 (p <.001). Full psychometric properties are reported in Supplementary Materials, Section S4.2 (Table S4.1).
Results and discussion
We follow the same analytical approach as in the previous studies, with one adaptation for the within-subjects design. Because each participant made 4 binary choices per wave, analyzing choices at the item level would inflate the effective sample size relative to the actual N = 57 participants. We therefore aggregate choices to the participant level (proportion of patient choices per person) and report a supplementary mixed-effects logistic regression estimates with a random intercept for participant as a robustness check. Effect sizes for the paired pre–post comparison are reported as Cohen’s dz. A sensitivity power analysis indicated 80% power (α = 0.05) to detect an effect of dz ≥ 0.38.
Manipulation check (MAAS pre vs. post)
As a manipulation check, we compared participants’ MAAS scores (computed as the mean of the 15 items) at the pre-test (Orientation Session, before the MBSR course began) and the post-test (Final Session). MAAS scores increased substantially from pre-treatment (M = 3.63, SD = 0.84) to post-treatment (M = 4.39, SD = 0.68); the mean within-person change was ΔM = + 0.76 (SD of the change = 0.82). A paired t-test was highly significant: t(54) = 6.84, p <.001, with a large within-subjects effect size, Cohen’s dz = 0.92, 95% CI [0.61, 1.24]. A Wilcoxon signed-rank test confirmed this pattern (W = 104, p <.001), and the BIC-based Bayes factor provided overwhelming evidence in favour of the alternative hypothesis (BF₁₀ = 3.8 × 10⁶; because this is the only confirmatory test in the paper, we report it in the direction of the alternative hypothesis, equivalently BF₀₁ = 2.6 × 10⁻⁷). The 8-week MBSR program therefore produced a robust within-subjects increase in trait mindfulness, validating the manipulation. The substantive intertemporal-choice analyses below are interpretable against an effective intervention.
Intertemporal choices [hypothesis-driven]
The proportion of patient (larger-later) choices depicted in Fig. 10 did not change significantly from pre-training (M = 0.640, SD = 0.406) to post-training (M = 0.671, SD = 0.401): paired t(56) = 0.74, p =.463, dz = 0.098, 95% CI [− 0.16, 0.36], BF₀₁ = 5.75. A Wilcoxon signed-rank test confirmed the null, W = 110, p =.574. Of the 57 participants, 12 increased their proportion of patient choices after the course, 10 decreased, and 35 showed no change. The Bayes factor provides moderate evidence for the null hypothesis (dz = 0.098, well below the minimum detectable effect (MDE) of dz = 0.38). A mixed-effects logistic regression on the item-level choices, with a random intercept for participant, also confirmed the null (OR = 1.35, p =.310; population-averaged d = 0.07, 95% CI [− 0.12, 0.27]). Item-level McNemar tests likewise revealed no significant pre-post shifts for any individual item (all p >.220). Full mixed-model results, including specifications with item fixed effects, are reported in the Supplementary Materials, Section S4.3, Table S4.2.
Proportion of distant choices per wave. This figure shows the proportion of distant choices combining the four intertemporal decisions as a function of the wave. Larger values indicate a greater proportion of distant choices. Error bars indicate the 95% confidence interval.
Table 2 reports the results from three OLS regression models further examining the impact of mindfulness training and trait mindfulness on the number of distant intertemporal choices participants made before and after the eight-week MBSR program. The dependent variable is the number of distant (i.e., delayed reward) options selected among the four binary decisions in a given wave (0–4), stacked across the pre-treatment and post-treatment waves, yielding 110 participant-wave observations. The primary independent variables include a post-treatment indicator (Posti), a trait mindfulness score (MAASi), and their interaction. Additional covariates include our prior meditation experience variables, age and gender. Standard errors are clustered at the participant level and are robust to heteroskedasticity.
In the model in column 1, which includes no controls or fixed effects, the coefficient for Posti was not statistically significant (b = 0.880, SE = 1.471, p =.552), suggesting that the average number of distant choices did not increase following the mindfulness course. Similarly, neither the main effect of trait mindfulness (b = 0.044, SE = 0.309, p =.887) nor its interaction with the post-treatment indicator (b = − 0.162, SE = 0.350, p =.643) were statistically significant, indicating no meaningful association between mindfulness traits and intertemporal choice, nor evidence of moderation by trait mindfulness. The model explains very little of the variance in the dependent variable (R² = 0.006). Bayesian analysis of the regression coefficient for the post-treatment indicator similarly favored the null hypothesis. The Bayes Factor derived from the t-statistic for the treatment effect (t = 0.60) was BF₀₁ = 6.88, providing moderate evidence that the intervention did not produce changes in the number of distant choices.
The model in column 2 adds individual-level covariates, including two indicators for prior mindfulness practice, age, and gender. The post-treatment coefficient remained statistically non-significant (b = 0.925, SE = 1.409, p =.514), as did the interaction term (Posti × MAASi; b = − 0.161, SE = 0.330, p =.628). Notably, the number of months of prior meditation practice emerged as a positive and statistically significant predictor of distant choices (b = 0.0038, SE = 0.0018, p =.036), suggesting that individuals with longer-standing mindfulness experience are more inclined toward delayed gratification, regardless of the formal course. Other covariates, including gender (b = 0.558, SE = 0.407, p =.176) and age (b = 0.0045, SE = 0.0207, p =.828), were not significantly associated with the outcome. The inclusion of these controls slightly improved model fit (R² = 0.048).
Column 3 introduces participant fixed effects to account for unobserved, time-invariant heterogeneity at the individual level, while omitting the additional controls. In this within-subject specification, the post-treatment coefficient is essentially unchanged and remains non-significant (b = 0.919, SE = 1.858, p =.623). The MAAS score shows a larger, though still statistically insignificant, association with distant choices (b = 0.376, SE = 0.326, p =.253), and the interaction term remains insignificant (b = − 0.229, SE = 0.434, p =.601). The fixed effects specification substantially increases explanatory power, with the model explaining 87% of the variance in the outcome (R² = 0.87), consistent with the inclusion of subject-level intercepts. Clustered OLS regressions (participant-level cluster-robust standard errors) confirmed that the pre-post change in patient choices did not vary as a function of dispositional mindfulness (MAAS mean score): post × MAAS b = − 0.162, SE = 0.350, p =.643. Including demographic controls (age, gender, self-reported mindfulness practice) did not alter this conclusion: post × MAAS b = − 0.161, SE = 0.330, p =.626.
Summary of effect sizes and Bayesian evidence
The 8-week MBSR course did not significantly alter intertemporal choices. The pre-post effect was trivially small (dz = 0.098), with moderate Bayesian evidence for the null (BF₀₁ = 5.75). The participant-level analysis, mixed-effects logistic regression, and item-level McNemar tests all converge on the same conclusion. Participant-level paired tests and item-level mixed-effects model results—including subject- and item-fixed-effects specifications, an item-type decomposition, and item-level McNemar tests—are reported in Supplementary Materials, Tables S4.2–S4.5.
Overall, across all specifications, there is no evidence that participation in the mindfulness program influenced intertemporal choice behavior. Trait mindfulness as measured by the MAAS did not predict distant choices nor interact meaningfully with the intervention. The only statistically significant finding is the positive relationship between months of prior meditation practice and patient decision-making, once the covariates are introduced in the second regression. This may suggest that accumulated meditation practice over longer periods is associated with greater patience, although this cross-sectional association is correlational and cannot be attributed to the intervention itself.
General Discussion
There has been growing concern both inside and outside academia about potential deleterious effects of mindfulness for decision-makers. Our research provides the most comprehensive investigation into the relationship between mindfulness and intertemporal choice conducted to date. Across four studies—including non-incentivized and incentivized decisions, hypothetical real-life scenarios, and a longitudinal field intervention—we found no evidence that mindfulness makes people more present-biased or impatient. While some minor differences in the opposite direction emerged in Study 3 (see also the regressions reported in Supplementary Materials, Section S3.4), they did not generalize across studies, outcomes, or methods; if anything, the fact that the only directional signal we observed pointed toward greater patience runs counter to the present-bias concern rather than supporting it. Importantly, even a structured, eight-week Mindfulness-Based Stress Reduction (MBSR) program did not produce significant changes in intertemporal decision-making. Bayesian analyses further strengthen these conclusions.
Mini Meta-Analysis of Induction Studies (Studies 1–3)
To provide a quantitative synthesis of the relationship between mindfulness and intertemporal-choice across the three induction studies, we conducted a mini meta-analysis pooling the standardized effect sizes (Cohen’s d or h) from all hypothesis-driven outcomes (k = 6). Using inverse-variance weighting, the fixed-effect pooled estimate was d = 0.020, 95% CI [−0.096, 0.136], z = 0.35, p =.730. The random-effects estimate (DerSimonian–Laird) was virtually identical: d = 0.014, 95% CI [−0.141, 0.169], z = 0.18, p =.858. Heterogeneity was low to moderate, Cochran’s Q(5) = 8.42, p =.134, I² = 40.6%, τ² = 0.0149, suggesting that the individual effect sizes are broadly consistent across studies and outcome measures.
A BIC-based meta-analytic Bayes Factor yielded BF₀₁ = 31.98 for the all-outcomes model, providing very strong evidence that the pooled mindfulness effect on intertemporal choice is indistinguishable from zero. When restricting to one primary outcome per study (k = 3), the pooled effect remained negligible: fixed-effect d = 0.025, 95% CI [−0.142, 0.192], p =.772, BF₀₁ = 22.69. The product of independent Bayes Factors across the three studies was BF₀₁ = 571.5, indicating overwhelming cumulative evidence for the null.
A parallel meta-analysis of the mindfulness vs. mind-wandering contrast yielded a similarly null pooled estimate (d = 0.116, BF₀₁ = 9.26). The full study-by-study effect-size table and forest plot for the primary mindfulness-vs-control analysis are reported in Supplementary Materials, Section S5.1 (Table S5.1 and Figure S5.1).
Practical and theoretical implications
These findings suggest that the extensively documented well-being benefits associated with cultivating awareness of the present moment do not appear to compromise people’s ability to make patient decisions regarding the future. This is particularly reassuring given the widespread adoption of mindfulness-based interventions in corporate and educational settings (see1,74, where the balance between immediate and delayed gratification is essential to success (see, for example75,76.
Additionally, our findings contribute to the broader theoretical understanding of intertemporal decision-making by suggesting that temporal attention and temporal valuation might be more separable than previously assumed. Specifically, our research implies that enhancing present-focused attention through mindfulness practices does not necessarily influence how individuals value immediate versus delayed rewards. This dissociation aligns with neuroscientific evidence highlighting separate neural networks for attentional control (e.g., the dorsal frontoparietal network78,79 and valuation processes (e.g., the ventromedial prefrontal cortex and ventral striatum80,81. Recent research further emphasizes flexible interactions among these networks rather than uniform responses to attentional manipulation alone82.
Interpreting the null results
It is important to consider what our null results do and do not establish. A failure to reject the null hypothesis does not, by itself, demonstrate the absence of an effect. However, several features of the present research strengthen the evidential value of our null findings. First, the consistency of the null across four independent studies—spanning different induction methods (brief audio vs. 8-week MBSR), different tasks (binary choices, real-world scenarios, structurally estimated parameters), and different incentive structures (hypothetical and consequential)—substantially reduces the probability that a genuine effect was missed by chance in any single study. Second, the mini meta-analysis of the three induction studies yields a pooled effect indistinguishable from zero (d = 0.020, 95% CI [− 0.096, 0.136]) with very strong Bayesian evidence for the null (BF₀₁ = 31.98), placing a tight upper bound on any potential mindfulness effect. Third, the tasks demonstrated adequate psychometric properties (Cronbach’s α ranging from 0.759 to 0.962 across studies; per-study reliabilities are reported in the Supplementary Materials, Sections S1.1, S2.2, S3.2, and S4.2) and structural sensitivity—detecting within-person present-bias effects, delay-length effects, and even a between-group mind-wandering effect in Study 2—confirming that the measurement instruments were capable of detecting effects of the magnitude typically reported in related literatures.
Boundary conditions
Nonetheless, our findings are bounded by several important factors. The brief-induction studies (Studies 1–3) tested the immediate effects of a single 15-minute session. It is possible that repeated mindfulness practice could produce cumulative shifts in intertemporal preferences that our design would not capture. Although Study 4 addressed this concern with an 8-week MBSR program, its sample (N = 57) was smaller and statistically powered to detect only medium-to-large effects (MDE dz ≥ 0.38). A longer training with a larger sample might reveal subtler shifts. Moreover, our participants were drawn from a non-clinical, predominantly young-adult, university-based population; the effects of mindfulness on intertemporal choice may differ among clinical populations (e.g., individuals with substance-use disorders or ADHD) for whom both mindfulness interventions and impulsive decision-making are of particular clinical relevance46,47,48,49. Similarly, our intertemporal tasks involved monetary outcomes. Mindfulness might differentially affect intertemporal decisions in domains such as health behaviors, dietary choices, or environmental sustainability, where emotional and visceral factors play a larger role.
Why might mindfulness leave intertemporal choice unchanged?
As noted above, the neural networks supporting attentional control and temporal valuation are largely dissociable, and our results are consistent with the view that enhancing present-moment awareness does not spill over into how people weigh immediate versus delayed rewards. A second possibility is that mindfulness produces competing effects that cancel out: increased attention to present rewards may be offset by increased self-regulation and reduced impulsivity, yielding no net change in observed choice behavior. A third possibility, which we cannot rule out, is that the effect exists but is too small to be detected even with our combined sample (N = 852 across Studies 1–3). However, the meta-analytic 95% confidence interval ([− 0.096, 0.136]) bounds any such effect to a negligible magnitude, well below the d = 0.20 conventionally considered small. If a mindfulness effect on intertemporal choice exists at all in non-clinical populations, it is unlikely to be practically meaningful.
Implications and future directions
Our null results also carry implications for how researchers interpret and report mindfulness effects on economic decision-making more broadly. The existing literature has been marked by small samples, heterogeneous tasks, and a potential publication bias favoring significant effects107. Our multi-study package, with hypothesis-driven analyses, effect size analyses, Bayes Factors, and a meta-analytic synthesis, illustrates the value of treating null results as informative and publishable. Future research should seek to replicate our findings in clinical populations, with longer and more intensive training protocols, and using neuroimaging to test whether mindfulness training modifies the neural correlates of temporal discounting even when behavioral choice remains unchanged. Furthermore, ecological momentary assessment designs could capture whether state mindfulness fluctuations within individuals predict day-to-day variation in intertemporal preferences, a level of analysis not accessible through the between-subjects induction designs employed here.
Nonetheless, these results should be interpreted with care. First, although we controlled for basic demographic variables (age, gender) and previous meditation experience, we did not collect data on participants’ socioeconomic status (SES) or living standards—factors often associated with intertemporal preferences. However, because conditions were randomly assigned at the session level, such individual differences should still be distributed across the mindfulness, mind-wandering, and control conditions, although with relatively few sessions per condition this balance is less assured than under individual-level randomization; we therefore report session-clustered standard errors throughout. Moreover, the predominance of university students in Studies 1–3 provides a relatively homogeneous sample regarding current financial context, minimizing the likelihood that SES disparities acted as a confound.
Second, individual differences in baseline impulsivity represent an important potential moderator that warrants future investigation. Impulsivity is a multidimensional construct, encompassing attentional, motor, and non-planning components93, each of which may interact differently with mindfulness interventions. Notably, trait mindfulness and dysfunctional impulsivity are negatively correlated, whereas functional impulsivity—the tendency to act quickly when such action is adaptive—is positively associated with mindfulness94. This distinction suggests that the relationship between mindfulness and impulsive behavior is not unidimensional. Moreover, an eight-week MBSR program failed to reduce impulsivity on either the BIS-11 or a go/no-go task relative to active and waitlist controls95, paralleling our own null findings for intertemporal choice after MBSR. Nevertheless, individuals with high baseline impulsivity might respond differently to mindfulness inductions than those with low impulsivity, especially if temporal discounting is partly driven by failures of inhibitory control96. Some evidence from brief mindfulness training suggests that acceptance-based components can momentarily reduce cognitive and motor impulsivity, but not impulsive decision-making as captured by delay discounting tasks97. If the effect of mindfulness on intertemporal choice operates through inhibitory control rather than through altered time perception, then baseline individual differences in self-control capacity may be a crucial boundary condition that future research should systematically examine.
Cultural context constitutes a third important moderator to consider. Cross-cultural research has documented meaningful differences in temporal discounting across societies, with neuroimaging evidence suggesting that these differences originate in reward-related brain regions (e.g., the ventral striatum) rather than in executive control systems98. One framework proposes that analytic versus holistic attentional styles—which differ systematically between Western and Eastern cultures—influence the degree of discounting and time inconsistency in intertemporal choice99. At the same time, a large-scale study across 61 countries (N = 13,629) found that the fundamental structure of temporal discounting is globally consistent, with economic inequality and financial instability—rather than cultural orientation per se—driving population-level variation in choice patterns100. Our samples were drawn from Spain and were relatively homogeneous in terms of cultural background. Although the use of random assignment in Studies 1–3 mitigates concerns about cultural confounding within our design, the broader question of whether mindfulness affects intertemporal preferences differently across cultures—particularly in societies with markedly different temporal orientations, individualism-collectivism dimensions, or economic environments—remains open. Replication across culturally diverse populations would therefore be valuable.
Fourth, the specific type of mindfulness practice employed may matter. Mindfulness meditation is commonly decomposed into two core components: focused attention (FA), which involves sustained concentration on a chosen object such as the breath, and open monitoring (OM), which entails non-reactive, moment-to-moment awareness of ongoing experience without a fixed attentional anchor101. These two practices engage partially distinct neurocognitive processes. FA meditation strengthens top-down attentional control and reduces mind-wandering, whereas OM meditation enhances non-judgmental monitoring of internal and external stimuli and decreases habitual reactivity102,103. A dismantling trial comparing FA, OM, and their combination in MBCT found that FA training had the greatest impact on emotional regulation, while OM training produced more limited benefits restricted to subjective outcomes104. One might hypothesize that OM meditation, with its emphasis on present-moment immersion, could plausibly heighten present-bias more than FA meditation, which cultivates executive control over attention. Our mindfulness inductions in Studies 1–3 used guided exercises combining breath-focused attention with broader awareness of thoughts and sensations—an approach typical of standard MBSR-style practice—and the MBSR program in Study 4 similarly integrated both FA and OM components. However, we did not systematically vary or isolate these components. Future dismantling studies that separately test FA and OM interventions on intertemporal choice tasks would be informative for determining whether any component of mindfulness practice selectively alters temporal valuation.
Taken together, socioeconomic status, baseline impulsivity, cultural factors, and the specific meditation component employed represent four plausible moderating variables that could shape the relationship between mindfulness and intertemporal choice. Our null findings were robust across multiple operationalizations and study designs, but they were obtained in culturally homogeneous, primarily student samples using combined FA/OM inductions. Investigating these moderators in future research would help establish not only whether mindfulness affects temporal preferences under specific conditions, but also illuminate the cognitive and affective mechanisms through which any such influence might operate.
Finally, we acknowledge specific limitations regarding our field study (Study 4). First, due to logistical and ethical constraints associated with partnering with a fee-based MBSR course, we utilized a single-arm, within-subjects design rather than a randomized controlled trial. While this design controls for time-invariant individual heterogeneity, it cannot rule out non-specific effects such as history or maturation, though the stability of preferences in the adult population makes spontaneous changes over eight weeks unlikely105,106. Second, the study experienced a 35% attrition rate—common in longitudinal field research but a constraint on statistical power. However, sensitivity analyses indicated that the covariate-complete subsample (N = 55) remained sufficiently powered to detect medium within-subject effects (dz = 0.38), and the Bayesian analysis of the pre/post change in patient choices yielded moderate evidence for the null hypothesis (BF₀₁ = 5.75; see Supplementary Materials, Section S4.3, Table S4.2). Thus, while Study 4 should be interpreted as a complementary ecological audit rather than a definitive trial, its results converge with the rigorous laboratory findings to suggest that even intensive mindfulness training does not systematically alter intertemporal preferences. Third, there was an asymmetry in the immediate context of the assessments. The pre-test was administered during an introductory orientation session (prior to the start of the course), whereas the post-test was administered at the conclusion of the final class, potentially following a period of practice. Consequently, the post-test measures could theoretically reflect a combination of cumulative training effects and immediate ‘state’ effects. However, it is worth noting that even with the potential additive influence of an immediate state induction at the post-test, no significant shift in intertemporal choice was observed. This suggests that neither the long-term acquisition of mindfulness skills nor the immediate physiological state of meditation was sufficient to alter time preferences in this context. Fourth, the pre- and post-test measures of intertemporal choice were identical. While keeping the tasks constant was necessary to calculate within-subject changes in discounting parameters, this design raises the possibility of test-retest or recognition effects. Although the eight-week interval between sessions likely reduced the precision of specific memory for prior choices, we cannot rule out the possibility that participants strove for consistency with their past selves, potentially dampening the observation of treatment effects.
Overall, our research consistently demonstrates that mindfulness does not inherently promote present-biased choices. In this sense, our findings provide no evidence that mindfulness practices undermine long-term decision-making in the populations we studied.
Data and code availability
The analyses for all our studies were conducted using Stata and R. The data, code, experimental instructions, and supplementary materials and analyses can be found in the following repository: https://doi.org/10.17605/OSF.IO/SWG79.
References
Gelles, D. Mindful work: How meditation is changing business from the inside out (Houghton Mifflin Harcourt, 2015).
Good, D. J. et al. Contemplating mindfulness at work: An integrative review. J. Manage. 42 (1), 114–142 (2016).
Hafenbrack, A. C. & Vohs, K. D. Mindfulness meditation impairs task motivation but not performance. Organ. Behav. Hum. Decis. Process. 147, 1–15 (2018).
Schlosser, M., Demnitz-King, H., Whitfield, T., Wirth, M. & Marchant, N. L. Repetitive negative thinking is associated with subjective cognitive decline in older adults: a cross-sectional study. BMC Psychiatry. 20, 1–10 (2020).
Grant, A. Can we end the meditation madness? [Internet]. The New York Times; Oct 10 [cited 2025 May 26]. (2015). Available from: https://www.nytimes.com/2015/10/10/opinion/can-we-end-the-meditation-madness.html
Whippman, R. Actually, let’s not be in the moment [Internet]. The New York Times; Nov 26 [cited 2025 May 26]. (2016). Available from: https://www.nytimes.com/2016/11/26/opinion/sunday/actually-lets-not-be-in-the-moment.html
Kabat-Zinn, J. Full catastrophe living: using the wisdom of your body and mind to face stress, pain, and illness (Delacorte, 1990).
Sauer, S. et al. Assessment of mindfulness: Review on state of the art. Mindfulness 4 (1), 3–17 (2013).
Bowen, S. et al. Mindfulness meditation and substance use in an incarcerated population. Psychol. Addict. Behav. 20 (3), 343–347 (2006).
Burke, C. A. Mindfulness-based approaches with children and adolescents: A preliminary review of current research in an emergent field. J. Child. Fam Stud. 19 (2), 133–144 (2010).
Chiesa, A., Calati, R. & Serretti, A. Does mindfulness training improve cognitive abilities? A systematic review of neuropsychological findings. Clin. Psychol. Rev. 31 (3), 449–464 (2011).
Creswell, J. D. Mindfulness interventions. Annu. Rev. Psychol. 68, 491–516 (2017).
Fjorback, L. O., Arendt, M., Ornbøl, E., Fink, P. & Walach, H. Mindfulness-based stress reduction and mindfulness-based cognitive therapy: a systematic review of randomized controlled trials. Acta Psychiatr Scand. 124 (2), 102–119 (2011).
Hofmann, S. G., Sawyer, A. T., Witt, A. A. & Oh, D. The effect of mindfulness-based therapy on anxiety and depression: A meta-analytic review. J. Consult Clin. Psychol. 78 (2), 169–183 (2010).
Mars, T. S. & Abbey, H. Mindfulness meditation practise as a healthcare intervention: A systematic review. Int. J. Osteopath. Med. 13 (2), 56–66 (2010).
Sharma, M. K., Bhargav, H., Kumar, A., Digambhar, V. & Mani, T. A. Mindfulness-based interventions: Potentials for management of Internet gaming disorder. Int. J. Yoga. 14 (3), 244–247. https://doi.org/10.4103/ijoy.IJOY_46_21 (2021 Sep–Dec).
Walach, H., Ferrari, M. L. G., Sauer, S. & Kohls, N. Mind-body practices in integrative medicine. Religions 3 (1), 50–81 (2012).
Alem, Y., Behrendt, H., Belot, M. & Bíró, A. Mind training, stress and behaviour—A randomised experiment. PLoS One. 16 (11), e0258172. https://doi.org/10.1371/journal.pone.0258172 (2021).
Ashe, M. L., Newman, M. G. & Wilson, S. J. Delay discounting and the use of mindful attention versus distraction in the treatment of drug addiction: A conceptual review. J. Exp. Anal. Behav. 103 (1), 234–248. https://doi.org/10.1002/jeab.122 (2015).
Murphy, C. & MacKillop, J. Living in the here and now: Interrelationships between impulsivity, mindfulness, and alcohol misuse. Psychopharmacol. (Berl). 219 (2), 527–536. https://doi.org/10.1007/s00213-011-2573-0 (2012).
Rönnlund, M. et al. Mindfulness promotes a more balanced time perspective: Correlational and intervention-based evidence. Mindfulness 10 (7), 1579–1591. https://doi.org/10.1007/s12671-019-01113-x (2019).
Westbrook, C. et al. Mindful attention reduces neural and self-reported cue-induced craving in smokers. Soc. Cogn. Affect. Neurosci. 8 (1), 73–84. https://doi.org/10.1093/scan/nsr076 (2013).
Farb, N. A. S. et al. Attending to the present: mindfulness meditation reveals distinct neural modes of self-reference. Soc. Cogn. Affect. Neurosci. 2 (4), 313–322 (2007).
Tang, Y. Y. et al. Central and autonomic nervous system interaction is altered by short-term meditation. Proc. Natl. Acad. Sci. U S A. 106 (22), 8865–8870 (2009).
Tang, Y. Y. et al. Short-term meditation training improves attention and self-regulation. Proc. Natl. Acad. Sci. U S A. 104 (43), 17152–17156 (2007).
Stolarski, M., Vowinckel, J., Jankowski, K. S. & Zajenkowski, M. Mind the balance, be contented: Balanced time perspective mediates the relationship between mindfulness and life satisfaction. Pers. Individ Dif. 93, 27–31. https://doi.org/10.1016/j.paid.2015.09.039 (2016).
Carlson, L. E., Speca, M., Patel, K. D. & Goodey, E. Mindfulness-based stress reduction in relation to quality of life, mood, symptoms of stress and levels of cortisol, dehydroepiandrosterone sulfate (DHEAS) and melatonin in breast and prostate cancer outpatients. Psychoneuroendocrinology 29 (4), 448–474 (2004).
Davidson, R. J. et al. Alterations in brain and immune function produced by mindfulness meditation. Psychosom. Med. 65(4), 564–570 (2003 Jul–Aug).
Witek-Janusek, L. et al. Effect of mindfulness based stress reduction on immune function, quality of life and coping in women newly diagnosed with early stage breast cancer. Brain Behav. Immun. 22 (6), 969–981 (2008).
Ruedy, N. E. & Schweitzer, M. E. In the moment: The effect of mindfulness on ethical decision making. J. Bus. Ethics. 95 (Suppl 1), 73–87. https://doi.org/10.1007/s10551-011-0796-y (2010).
Shapiro, S. L., Jazaieri, H. & Goldin, P. R. Mindfulness-based stress reduction effects on moral reasoning and decision making. J. Posit. Psychol. 7 (6), 504–515. https://doi.org/10.1080/17439760.2012.723732 (2012).
Hafenbrack, A. C. et al. Helping people by being in the present: Mindfulness increases prosocial behavior. Organ. Behav. Hum. Decis. Process. 159, 21–38. https://doi.org/10.1016/j.obhdp.2019.08.005 (2020).
Kirk, U. et al. Mindfulness training increases cooperative decision making in economic exchanges: Evidence from fMRI. Neuroimage 138, 274–283. https://doi.org/10.1016/j.neuroimage.2016.05.075 (2016).
Hafenbrack, A. C., Kinias, Z. & Barsade, S. G. Debiasing the mind through meditation: Mindfulness and the sunk-cost bias. Psychol. Sci. 25 (2), 369–376 (2014).
Errmann, A. C. Re-positioning the present moment: Effects of mindfulness on consumer behaviour [dissertation]. Auckland: University of Auckland; (2022).
Marcowski, P., Białaszek, W., Dudek, J. & Ostaszewski, P. Higher behavioral profile of mindfulness and psychological flexibility is related to reduced impulsivity in smokers, and reduced risk aversion regardless of smoking status. Pol. Psychol. Bull. 48 (4), 445–455. https://doi.org/10.1515/ppb-2017-0051 (2017).
Bazzano, M. After mindfulness: new perspectives on psychology and meditation (Palgrave Macmillan UK, 2014).
Foster, D. Is mindfulness making us ill? The Guardian [Internet]. Jan 23 [cited 2020 May 27]. (2016). Available from: https://www.theguardian.com/lifeandstyle/2016/jan/23/is-mindfulness-making-us-ill
Shonin, E., Van Gordon, W. & Griffiths, M. Mindfulness in psychology: A breath of fresh air? Psychologist 28, 28–31 (2015).
Cohen, J., Ericson, K. M., Laibson, D. & White, J. M. Measuring time preferences. J. Econ. Lit. 58 (2), 299–347. https://doi.org/10.1257/jel.20191074 (2020).
Frederick, S., Loewenstein, G. & O’Donoghue, T. Time discounting and time preference: A critical review. J. Econ. Lit. 40 (2), 351–401 (2002).
Mischel, W., Shoda, Y. & Peake, P. K. The nature of adolescent competencies predicted by preschool delay of gratification. J. Pers. Soc. Psychol. 54 (4), 687–696. https://doi.org/10.1037/0022-3514.54.4.687 (1988).
Epper, T. et al. Time discounting and wealth inequality. Am. Econ. Rev. 110 (4), 1177–1205. https://doi.org/10.1257/aer.20181096 (2020).
Barlow, P., Reeves, A., McKee, M., Galea, G. & Stuckler, D. Unhealthy diets, obesity and time discounting: A systematic literature review and network analysis. Obes. Rev. 17, 810–819. https://doi.org/10.1111/obr.12431 (2016).
Komlos, J., Smith, P. K. & Bogin, B. Obesity and the rate of time preference: Is there a connection? J. Biosoc Sci. 36 (2), 209–219. https://doi.org/10.1017/S0021932003006205 (2004).
Li, C. R. & Sinha, R. Inhibitory control and emotional stress regulation: Neuroimaging evidence for frontal limbic dysfunction in psycho-stimulant addiction. Neurosci. Biobehav Rev. 32, 581–597 (2008).
De Paola, M. & Gioia, F. Does patience matter in marriage stability? Some evidence from Italy. Rev. Econ. Househ. 15 (2), 549–577 (2017).
Wang, L. et al. Associations between impulsivity, aggression, and suicide in Chinese college students. BMC Public. Health. 14, 551 (2014).
Karyadi, K. A., VanderVeen, J. D. & Cyders, M. A. A meta-analysis of the relationship between trait mindfulness and substance use behaviors. Drug Alcohol Depend. 143, 1–10. https://doi.org/10.1016/j.drugalcdep.2014.07.014 (2014).
Hendrickson, K. L. & Rasmussen, E. B. Effects of mindful eating training on delay and probability discounting for food and money in obese and healthy-weight individuals. Behav. Res. Ther. 51 (7), 399–409 (2013).
Hendrickson, K. L. & Rasmussen, E. B. Mindful eating reduces impulsive food choice in adolescents and adults. Health Psychol. 36 (3), 226–235 (2017).
Yao, Y. W. et al. Combined reality therapy and mindfulness meditation decrease intertemporal decisional impulsivity in young adults with internet gaming disorder. Comput. Hum. Behav. 68, 210–216 (2017).
Morrison, K. L. et al. Effects of Acceptance and Commitment Therapy on impulsive decision-making. Behav. Modif. 44 (4), 598–621. https://doi.org/10.1177/0145445519833041 (2020).
Shead, N. W., Champod, A. S. & MacDonald, A. Effect of a brief meditation intervention on gambling cravings and rates of delay discounting. Int. J. Ment Health Addict. 18 (4), 1247–1263. https://doi.org/10.1007/s11469-019-00133-x (2020).
Rasmussen, E. B., Rodriguez, L. R. & Pemberton, S. Acute and enduring effects of mindful eating on delay and probability discounting for food and money in food-insecure women. Mindfulness (N Y). 13 (3), 712–729. https://doi.org/10.1007/s12671-022-01828-4 (2022).
Bazley, W., Cuculiza, C. & Pisciotta, K. Auditory cues, temporal attention, and intertemporal choices [Internet]. Rochester (NY): SSRN; 2023 [cited 2025 Apr 11]. Available from: https://ssrn.com/abstract=3921871
Dixon, M. R. et al. The effect of brief mindfulness training on momentary impulsivity. J. Contextual Behav. Sci. 11, 15–20. https://doi.org/10.1016/j.jcbs.2018.11.003 (2019).
Errmann, A. All the time in the world: Effects of mindfulness on intertemporal decisions. In: Proceedings of the Global Fashion Management Conference; ; Seoul, South Korea. pp. 283–91. (2021).
Morrison, K. L., Madden, G. J., Odum, A. L., Friedel, J. E. & Twohig, M. P. Altering impulsive decision making with an acceptance-based procedure. Behav. Ther. 45 (5), 630–639 (2014).
Scholten, H. et al. Behavioral trainings and manipulations to reduce delay discounting: A systematic review. Psychon Bull. Rev. 26 (6), 1803–1849 (2019).
Wang, Z. et al. Experiencing Buddhist practices promotes nonbelievers’ preference for future rewards in intertemporal choices. Eur. J. Soc. Psychol. 53 (1), 230–244. https://doi.org/10.1002/ejsp.2900 (2023).
LeBoeuf, R. A. Discount rates for time versus dates: The sensitivity of discounting to time-interval description. J. Mark. Res. 43 (1), 59–72. https://doi.org/10.1509/jmkr.43.1.59 (2006).
Read, D., Frederick, S., Orsel, B. & Rahman, J. Four score and seven years from now: The date/delay effect in temporal discounting. Manag Sci. 51 (9), 1326–1335. https://doi.org/10.1287/mnsc.1050.0412 (2005).
Odum, A. L. et al. Delay discounting of different outcomes: Review and theory. J. Exp. Anal. Behav. 113 (3), 657–679. https://doi.org/10.1002/jeab.589 (2020).
Duchêne, S., Guillon, M. & Rafaï, I. Association between mindfulness and risk and time preferences. J. Econ. Sci. Assoc. 10 (2), 199–212. https://doi.org/10.1007/s40881-024-00169-3 (2025).
Arch, J. J. & Craske, M. G. Mechanisms of mindfulness: Emotion regulation following a focused breathing induction. Behav. Res. Ther. 44 (12), 1849–1858 (2006).
Kiken, L. G. & Shook, N. J. Looking up: Mindfulness increases positive judgments and reduces negativity bias. Soc. Psychol. Personal Sci. 2 (4), 425–431 (2011).
Mason, M. F. et al. Wandering minds: The default network and stimulus-independent thought. Science 315 (5810), 393–395 (2007).
Wilson, T. D. et al. Just think: The challenges of the disengaged mind. Science [Internet]. 2014 Jul 4 [cited 2025 Apr 27];345(6192):75–7. Available from: https://www.science.org/doi/https://doi.org/10.1126/science.1250830
McClure, S. M., Laibson, D. I., Loewenstein, G. & Cohen, J. D. Separate neural systems value immediate and delayed monetary rewards. Science 306 (5695), 503–507 (2004).
Laibson, D. Golden eggs and hyperbolic discounting. Q. J. Econ. 112 (2), 443–477 (1997).
O’Donoghue, T. & Rabin, M. Doing it now or later. Am. Econ. Rev. 89 (1), 103–124 (1999).
Rosenkranz, M. A., Dunne, J. D. & Davidson, R. J. The next generation of mindfulness-based intervention research: what have we learned and where are we headed? Curr. Opin. Psychol. 28, 179–183. https://doi.org/10.1016/j.copsyc.2018.12.022 (2019). Epub 2019 Jan 4. PMID: 30739006; PMCID: PMC6609495.
Sheinman, N. & Russo–Netzer, P. Mindfulness in education: Insights towards an integrative paradigm. In The Palgrave Handbook of Positive Education (eds Kern, M. L. & Wehmeyer, M. L.) 609–642 (Palgrave Macmillan, 2021). https://doi.org/10.1007/978-3-030-64537-3_24.
Gschwandtner, A., Jewell, S. & Kambhampati, U. S. Lifestyle and life satisfaction: The role of delayed gratification. J. Happiness Stud. 23, 1043–1072. https://doi.org/10.1007/s10902-021-00440-y (2022).
Shoda, Y., Mischel, W. & Peake, P. K. Predicting adolescent cognitive and self-regulatory competencies from preschool delay of gratification: Identifying diagnostic conditions. Dev. Psychol. 26 (6), 978–986. https://doi.org/10.1037/0012-1649.26.6.978 (1990).
Galizzi, M. M. & Navarro-Martinez, D. On the external validity of social preference games: A systematic lab-field study. Manag Sci. 65 (3), 976–1002. https://doi.org/10.1287/mnsc.2017.2908 (2019).
Corbetta, M. & Shulman, G. L. Control of goal-directed and stimulus-driven attention in the brain. Nat Rev Neurosci. ;3(3):201 – 15. (2002). https://doi.org/10.1038/nrn755. PMID: 11994752.
Buschman, T. J. & Kastner, S. From Behavior to Neural Dynamics: An Integrated Theory of Attention. Neuron 88 (1), 127–144. https://doi.org/10.1016/j.neuron.2015.09.017 (2015).
Kable, J. W. & Glimcher, P. W. The neural correlates of subjective value during intertemporal choice. Nat. Neurosci. 10 (12), 1625–1633. https://doi.org/10.1038/nn2007 (2007).
Varma, M. M., Zhen, S. & Yu, R. ‘Not all discounts are created equal: Regional activity and brain networks in temporal and effort discounting’, NeuroImage, 280, p. 120363. (2023). https://doi.org/10.1016/j.neuroimage.2023.120363
Wang, Q. et al. Neural representations of the amount and the delay time of reward in intertemporal decision making. Hum. Brain Mapp. 42, 3450–3469. https://doi.org/10.1002/hbm.25445 (2021).
Wagenmakers, E-J. A practical solution to the pervasive problems of p values. Psychon Bull. Rev. 14 (5), 779–804 (2007).
Cameron, A. C. & Miller, D. L. A practitioner’s guide to cluster-robust inference. J. Hum. Resour. 50 (2), 317–372 (2015).
Jeffreys, H. Theory of probability 3rd edn (Oxford University Press, 1961).
Kass, R. E. & Raftery, A. E. Bayes factors. J. Am. Stat. Assoc. 90 (430), 773–795 (1995).
Brown, K. W. & Ryan, R. M. The benefits of being present: mindfulness and its role in psychological well-being. J. Pers. Soc. Psychol. 84 (4), 822–848 (2003).
Peters, J. & Büchel, C. Episodic future thinking reduces reward delay discounting through an enhancement of prefrontal-mediotemporal interactions. Neuron 66 (1), 138–148 (2010).
Amlung, M., Vedelago, L., Acker, J., Balodis, I. & MacKillop, J. Steep delay discounting and addictive behavior: a meta-analysis of continuous associations. Addiction 112 (1), 51–62 (2017).
Garland, E. L., Froeliger, B. & Howard, M. O. Mindfulness training targets neurocognitive mechanisms of addiction at the attention-appraisal-emotion interface. Front. Psychiatry. 4, 173 (2014).
Lachaud, C. M., Renault, R., Bhatt, M., Bherer, L. & Bhatt, S. Mindfulness meditation and choice blindness: an experimental study. Mindfulness 13 (8), 2020–2030 (2022).
Smallwood, J., Ruby, F. J. M. & Singer, T. Letting go of the present: mind-wandering is associated with reduced delay discounting. Conscious. Cogn. 22 (1), 1–7 (2013).
Patton, J. H., Stanford, M. S. & Barratt, E. S. Factor structure of the Barratt impulsiveness scale. J. Clin. Psychol. 51 (6), 768–774 (1995).
Yigitçetin, R., Stoyanova, S. & Prenda, J. The relationship between mindfulness and impulsivity: the role of meditation. Mindfulness 15, 1649–1661 (2024).
Korponay, C. et al. The effect of mindfulness meditation on impulsivity and its neurobiological correlates in healthy adults. Sci. Rep. 9, 11963 (2019).
Peters, J. & Büchel, C. The neural mechanisms of inter-temporal decision-making: understanding variability. Trends Cogn. Sci. 15 (5), 227–239 (2011).
Qu, Y., Li, M. & Xiao, W. Brief online training in mindfulness meditation momentarily reduces some impulsive behaviors: a randomized controlled trial. Mindfulness 16, 42 (2025).
Kim, B., Sung, Y. S. & McClure, S. M. The neural basis of cultural differences in delay discounting. Philos. Trans. R Soc. Lond. B Biol. Sci. 367 (1589), 650–659 (2012).
Takahashi, T. et al. Cultural neuroeconomics of intertemporal choice. Neuro Endocrinol. Lett. 30 (2), 185–191 (2009).
Ruggeri, K. et al. The globalizability of temporal discounting. Nat. Hum. Behav. 6 (10), 1386–1397 (2022).
Lutz, A., Slagter, H. A., Dunne, J. D. & Davidson, R. J. Attention regulation and monitoring in meditation. Trends Cogn. Sci. 12 (4), 163–169 (2008).
Lippelt, D. P., Hommel, B. & Colzato, L. S. Focused attention, open monitoring and loving kindness meditation: effects on attention, conflict monitoring, and creativity – a review. Front. Psychol. 5, 1083 (2014).
Ooishi, Y., Fujino, M., Inoue, V., Nomura, M. & Kitagawa, N. Differential effects of focused attention and open monitoring meditation on autonomic cardiac modulation and cortisol secretion. Front. Physiol. 12, 675899 (2021).
Brown, K. W. et al. Comparing impacts of meditation training in focused attention, open monitoring, and mindfulness-based cognitive therapy on emotion reactivity and regulation: neural and subjective evidence from a dismantling study. Psychophysiology 59 (11), e14111 (2022).
Gelino, B. W., Schlitzer, R. D., Reed, D. D. & Strickland, J. C. A systematic review and meta-analysis of test–retest reliability and stability of delay and probability discounting. J. Exp. Anal. Behav. 121 (3), 358–372 (2024).
Strickland, J. C. et al. Temporal reliability and stability of delay discounting: a 2-year repeated assessments study of the Monetary Choice Questionnaire. Exp. Clin. Psychopharmacol. 31 (5), 902–907 (2023).
Van Dam, N. T. et al. Mind the hype: a critical evaluation and prescriptive agenda for research on mindfulness and meditation. Perspect. Psychol. Sci. 13 (1), 36–61 (2018).
Acknowledgements
We are thankful to Guillermo Mobayed Caballé for his assistance in reviewing and evaluating empirical studies referenced in this paper. We also thank Ferran Mestanza for producing the induction audios and for guidance on mindfulness practice and philosophy.
Funding
We gratefully acknowledge the generous support of the Janssen family, whose funding made this research possible. We also thank the Spanish Ministry of Science and Innovation (PID2019-105249GB-I00, PID2022-137908NB-I00), the BBVA Foundation (Fundacion BBVA-EI-2019-D.Navarro), the Ramon Areces Foundation (Fundacion Ramon Areces 2019- Navarro), and ICREA (ICREA Academia 2024-Daniel Navarro) for their financial support. This paper was partially funded by PID2023-152707NA-I00 granted by MCIU / AEI / 10.13039/501100011033 / FEDER, UE.
Author information
Authors and Affiliations
Department of Finance, IE Business School, Madrid, Spain
Josep Gisbert
Department of Economics and Business, Pompeu Fabra University, Barcelona, Spain
Daniel Navarro-Martinez
Instituto EsMindfulness ES, Barcelona, Spain
Andrés Martín-Asuero
Department of People Management and Organization, ESADE, Sant Cugat, Spain
Jordi Quoidbach
IE University, Madrid, Spain
Josep Gisbert
Barcelona School of Economics, Barcelona, Spain
Daniel Navarro-Martinez
Barcelona School of Management, Barcelona, Spain
Daniel Navarro-Martinez
Authors
- Josep Gisbert
- Daniel Navarro-Martinez
- Andrés Martín-Asuero
- Jordi Quoidbach
Contributions
J.G. (Josep Gisbert) conceived the study, designed the methodology, developed the software and materials, performed the formal analyses, and wrote the original draft. D.N.-M. (Daniel Navarro‑Martínez) conceived the study, designed the methodology, developed the software, performed the formal analyses, supervised the research, and reviewed and edited the manuscript. A.M.-A. (Andrés Martín‑Asuero) contributed to the methodology, provided resources, and reviewed and edited the manuscript. J.Q. (Jordi Quoidbach) contributed to the conceptualization and methodology, supervised the research, and reviewed and edited the manuscript. All authors reviewed and approved the final manuscript.
Corresponding author
Correspondence to Josep Gisbert.
Ethics declarations
Competing interests
Andrés Martín-Asuero is the director of Instituto EsMindfulness, the organization that delivered the Mindfulness-Based Stress Reduction (MBSR) program evaluated in Study 4. Josep Gisbert, Daniel Navarro-Martinez, and Jordi Quoidbach declare no competing interests.
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
About this article
Cite this article
Gisbert, J., Navarro-Martinez, D., Martín-Asuero, A. et al. Mindfulness meditation does not make people more present-biased. Sci Rep 16, 32477 (2026). https://doi.org/10.1038/s41598-026-64198-0
Received: 17 October 2025
Accepted: 23 July 2026
Published: 15 October 2026
Version of record: 15 October 2026
DOI: https://doi.org/10.1038/s41598-026-64198-0
Keywords
来源:Nature 健康护理主题 · nature.com