Delay discounting: Trait variable?
Highlights
► Delay discounting is associated with a variety of maladaptive behaviors. ► The current analyses shows that discounting for one outcome is related to discounting for other outcomes. ► These and other data suggest delay discounting may be a personality trait.
Introduction
Impulsivity is a multifaceted concept that includes relative insensitivity to delayed outcomes (see e.g., de Wit, 2008). Delay discounting refers to the decrease in the present value of an outcome when its receipt is delayed (Mazur, 1987). Across a variety of species and outcome types, the form of the function relating delay to value is hyperbolic: small delays to the receipt of the outcome have a proportionally greater impact on value than do longer delays (e.g., Rachlin et al., 1991, Richards et al., 1997). Steep hyperbolic discounting of value by delay can lead to one form of impulsivity: choice of a smaller sooner outcome over a larger later outcome (Logue, 1988).
Delay discounting is of growing interest because of its relation to a number of socially important problems. For example, cigarette smokers show greater discounting of delayed monetary outcomes than do matched non-smokers (e.g., Bickel et al., 1999). Furthermore, steep discounting is predictive of the initiation of regular smoking in adolescents (e.g., Audrain-McGovern et al., 2009). Cigarette smokers with shallower discount functions are also more likely to achieve abstinence from cigarettes (e.g., MacKillop and Kahler, 2009).
Substantial evidence indicates that delay discounting is sensitive to both state and trait influences (see Odum and Baumann, 2010, for review). A state variable influences behavior over a relatively short time frame, whereas a trait variable is a relatively stable pre-existing characteristic an individual brings to a situation. State variables shown to affect the degree of discounting include the type of outcome (e.g., Madden et al., 1997, Odum and Rainaud, 2003), the magnitude of an outcome (e.g., Green et al., 1997), and the context in which a choice is made (e.g., Dixon et al., 2006).
A number of trait variables have been shown to affect the degree to which an outcome is discounted as well. For example, steeper discounting is associated with more fatalism (Johnson et al., 2010), less agreeableness (Miller et al., 2008) and less empathy (Kirby et al., 1999) as measured by personality scales. Similarly, certain psychiatric conditions, such as disinhibitory behavior problems and antisocial personality disorders (e.g., Bobova et al., 2009, Crean et al., 2000, Petry, 2002), as well as pathological gambling and drug abuse, are associated with elevated levels of delay discounting (see Petry and Madden, 2010, Yi et al., 2010 for review). It is interesting, however, that self-report measures of impulsivity have weak and often inconsistent relations to the degree of delay discounting across studies (e.g., Bobova et al., 2009, Janis and Nock, 2009, Perales et al., 2009, Smith and Hantula, 2008; see de Wit et al., 2007 for review). These inconsistencies may be due to the fact that these self-report scales measure multiple facets of impulsivity, including the tendency to act without thinking and the inability to withhold a pre-potent response, rather than aspects more specifically related to sensitivity to delayed outcomes.
Although delay discounting is clearly related to some personality and individual characteristics, there has been little formal consideration of whether delay discounting itself might be considered a trait (see de Wit, 2008, Kirby, 2009, Reimers et al., 2009). A reasonably common definition of a personality trait is ‘a relatively enduring pattern of thoughts, feelings, and behaviors that reflects the tendency to respond in certain ways under certain circumstances’ (Roberts, 2009). What evidence is there that the degree of discounting in which a person engages might be an overt component of such a trait?
First, consider the initial portion of the definition of a personality trait: that delay discounting is relatively enduring. This question may be answered in part by data concerning test-retest reliability. Test-retest reliability may be measured with the same form of the test on two occasions, or with one form of the test on one occasion and an alternate form of the test on another occasion.
Same form test-retest reliability is good for delay discounting as measured by a variety of techniques and over test-retest intervals ranging from weeks to a year. For example, Simpson and Vuchinich (2000) found evidence for strong test-retest reliability (r = .91) with a test-retest interval of 1 week using a choice task in which the delayed amount was $1000 and the outcomes were not actually delivered to the undergraduate participants (called a hypothetical money choice task). Test-retest reliability remains good up to intervals of one year (r = .71; Kirby, 2009). Several other studies have yielded similar conclusions regarding test-retest reliability of the degree of delay discounting using intermediate test-retest intervals and different monetary amounts, populations, and versions of the delay-discounting task (Baker et al., 2003, Beck and Triplett, 2009, Black and Rosen, 2011, Ohmura et al., 2006, Takahashi et al., 2007). Further investigation is warranted to determine whether discounting is stable over longer time frames.
Alternate form test-retest reliability is also good for delay discounting as assessed by a number of different versions of the task. For example, Johnson and Bickel (2002) found a robust relation (r = .83) between the degree of discounting on the hypothetical money choice task and a ‘real rewards’ version of the task in which one randomly selected outcome is delivered. Smith and Hantula (2008) reported good overall correspondence between different methods of task administration (paper and pencil vs. computer based) and task types (binary choice vs. fill in the blank; r = .75). Robles et al. (2009) obtained a moderate relation between the degree of discounting as determined by different sequences of presentation of the choices between the immediate and delayed outcomes (ascending vs. descending outcome amounts; ρ = .44). The degree of delay discounting was strongly correlated between versions of the task that used fixed versus titrating amount sequences of presentation of the choices between the immediate and delayed outcomes (r = .81; Rodzon et al., 2011). Other studies have also reported good correlations between the degree of discounting obtained using alternate methods (e.g., Epstein et al., 2003, Kowal et al., 2007). Thus, there is evidence that delay discounting is reasonably stable over modest time frames and with different assessment techniques, providing at least a tentative positive answer to the first criterion for a trait given above, that it be ‘relatively enduring’.
What evidence is there for the second part of the definition of a trait, that it ‘reflects the tendency to respond in certain ways under certain circumstances?’ (This requirement does not mean that people must behave identically in all situations, but that their behavior be ‘meaningfully consistent’; Roberts, 2009.) There are multiple ways to answer this question. One possible direction, which I shall pursue in this paper, is to examine the consistency of delay discounting across different types of outcomes.
Prior research has extensively shown that money is discounted less steeply than other types of outcomes. For example, Madden et al. (1997) found that people with opioid dependence discounted money less than they did an amount of heroin that was equated for monetary value. The finding that people discount their drug of abuse more steeply than money has been widely replicated with a variety of drug classes, including alcohol, cocaine, marijuana, and nicotine (e.g., Bickel et al., 1999, Coffey et al., 2003, Johnson et al., 2010, Petry, 2001). Although initially researchers speculated that the differences in discounting for drugs and money could be related to addiction specifically, Odum and Rainaud (2003) found that people without problematic drinking also discount alcohol more steeply than money. Furthermore, a variety of commodities are discounted more steeply than money, including an assortment of food and non-alcoholic beverages (e.g., Estle et al., 2007, Odum and Baumann, 2007, Odum and Rainaud, 2003, Odum et al., 2006), as well as music, CDs, and DVDs (Charlton and Fantino, 2008).
Examining discounting of different commodities opens the possibility to examine the second part of the definition of a personality trait, the ‘tendency to respond in certain ways in certain circumstances’. People who show relatively steep discounting of one outcome (e.g., money) should also show relatively steep discounting of another outcome (e.g., food) if they indeed behave similarly with respect to different outcomes. In the words of Green and Myerson (2010), “If impatience is a trait, [impatient] individuals would be expected to consistently discount delayed outcomes more than do other individuals, regardless of what the outcome is…” (p. 68). Only a few studies to date, however, have investigated this issue. Charlton and Fantino (2008) found moderate to strong correlations, all positive, between discounting among various forms of entertainment (books, CDs, DVDs), food, and money. Johnson et al. (2010) reported a strong positive correlation (r = .72) between the degree of discounting for hypothetical marijuana and money in current marijuana users. In the present paper, I determine the generality of these findings by analyzing a number of archival data sets to determine whether there is tendency for people, across a variety of populations and outcome types, to discount different commodities in a related manner.
Section snippets
Studies
The data sets were drawn from all prior studies that included discounting for more than one commodity that my colleagues and I have published. Table 1 summarizes the studies in terms of characteristics and number of participants and outcomes types and amounts. These studies include college students, community members screened for psychological disorders, and people with drug abuse problems. The outcomes used were food, alcohol, cigarettes, and/or heroin. The outcomes values ranged from amounts
Results
Fig. 1 shows the mean AUC values across participants for studies that examined discounting of money and one other outcome. In all cases, consumable commodities had lower AUCs, showing they were discounted more steeply, than money. Cigarette smokers (top panel) discounted cigarettes more than money (t22 = 3.21; p = 0.004), and people with opioid dependence (second panel) discounted heroin more so than money (t31 = 4.93; p < 0.0001). College students discounted $100 worth of food (third panel: t51 = 5.24; p
Discussion
This examination of archival data investigated AUC values, indicative of the steepness of discounting by delay, across different commodities. In all cases, the AUC values for consumable commodities were significantly lower than values for money, indicating money was discounted less steeply. This finding with AUC replicates conclusions conducted in some of the original studies using a different measure of discounting (the derived parameter k, from the hyperbolic model; Mazur, 1987) as well as
Acknowledgement
I would like to thank Tim Shahan, Greg Madden, and my laboratory group for helpful discussion.
References (97)
- et al.
Pathological gambling severity is associated with impulsivity in a delay discounting procedure
Behav. Processes
(2003) - et al.
Effects of clomipramine on self-control choice in Lewis and Fischer 344 rats
Pharmacol. Biochem. Behav.
(2005) - et al.
Does delay discounting play an etiological role in smoking or is it a consequence of smoking?
Drug Alcohol Depend.
(2009) - et al.
Epigenetics and the biological basis of gene × environment interactions
J. Am. Acad. Child Adolesc. Psychiat.
(2010) - et al.
Remember the future: working memory training decreases delay discounting among stimulant addicts
Biol. Psychiat.
(2011) - et al.
A money management-based substance use treatment increases valuation of future rewards
Addict. Behav.
(2011) - et al.
Commodity specific rates of temporal discounting: does metabolic function underlie differences in rates of discounting?
Behav. Processes
(2008) - et al.
Immediate pleasures and future consequences: A neuropsychological study of binge eating and obesity
Appetite
(2010) - et al.
IQ and nonplanning impulsivity are independently associated with delay discounting in middle-aged adults
Pers. Individ. Dif.
(2007) - et al.
Are self-injurers impulsive? Results from two behavioral laboratory studies
Psychiat. Res.
(2009)
A comparison of two algorithms in computerized temporal discounting procedures
Behav. Processes
Delay discounting of real and hypothetical rewards III: steady-state assessments, forced-choice trials, and all real rewards
Behav. Processes
Delayed reward discounting predicts treatment response for heavy drinkers receiving smoking cessation treatment
Drug Alcohol Depend.
Cigarette smokers show steeper discounting of both food and cigarettes than money
Drug Alcohol Depend.
Discounting of delayed alcohol, food, and money
Behav. Processes
Discounting of delayed hypothetical money and food: effects of amount
Behav. Processes
Needle sharing in opioid-dependent outpatients: psychological processes underlying risk
Drug Alcohol Depend.
Development of hot and cool executive function during the transition to adolescence
J. Exp. Child Psychol.
Associations between a one-shot delay discounting measure and age, income, education and real-world impulsive behavior
Person. Ind. Diff.
Delay discounting by the children of smokers and nonsmokers
Drug Alcohol Depend.
Within-subject differences in degree of delay discounting as a function of order of presentation of hypothetical cash rewards
Behav. Processes
Within-subject comparison of degree of delay discounting using titrating and fixed sequence procedures
Behav. Processes
Back to the future: personality and assessment and personality development
J. Res. Pers.
Good things come to those who wait: attenuated discounting of delayed rewards in aged Fischer 344 rats
Neurobiol. Aging
Neuroanatomical correlates of personality in the elderly
Neuroimage
Heritability of delay discounting in adolescence: a longitudinal twin study
Behav. Genet.
Delay discounting in current and never-before cigarette smokers: similarities and differences across commodity, sign, and magnitude
J. Abnorm. Psychol.
Test-retest reliability of a group-administered paper-pencil measure of delay discounting
Exp. Clin. Psychopharmacol.
Impulsivity and cigarette smoking: delay discounting in current, never, and ex-smokers
Psychopharmacology
Disinhibitory psychopathology and delay discounting in alcohol dependence: personality and cognitive correlates
Exp. Clin. Psychopharmacol.
Animal personality
Curr. Biol.
Comparative neuroscience of stimulant-induced memory dysfunction: role for neurogenesis in the adult hippocampus
Behav. Pharmacol.
Functional brain mapping of extraversion and neuroticism: learning from individual differences in emotion processing
J. Pers.
Impulsivity and rapid discounting of delayed hypothetical rewards in cocaine-dependent individuals
Exp. Clin. Psychopharmacol.
Reward discounting as a measure of impulsive behavior in a psychiatric outpatient population
Exp. Clin. Psychopharmacol.
Impulsivity as a determinant and consequence of drug use: a review of underlying processes
Addict. Biol.
Self-control in children with autism: response allocation during delays to reinforcement
J. Appl. Behav. Anal.
Preference for progressive delays and concurrent physical therapy exercise in an adult with acquired brain injury
J. Appl. Behav. Anal.
Using a self-control training procedure to increase appropriate behavior
J. Appl. Behav. Anal.
Teaching self-control to small groups of dually diagnosed adults
J. Appl. Behav. Anal.
Contextual control of delayed discounting by pathological gamblers
J. Appl. Behav. Anal.
The effects of choice on self-control
J. Appl. Behav. Anal.
Cross-cultural comparisons of discounting delayed and probabilistic rewards
Psychol. Rec.
Examining impulsivity as an endophenotype using a behavioral approach: a DRD2 Taql A and DRD4 48-bp VNTR association study
Behav. Brain Funct.
Comparison between two measures of delay discounting in smokers
Exp. Clin. Psychopharmacol.
Discounting of monetary and directly consumable rewards
Psychol. Sci.
Facilitating tolerance of delayed reinforcement during functional communication training
Behav. Modif.
Discounting of delayed rewards: a life-span comparison
Psychol. Sci.
Cited by (297)
Delay discounting as a behavioral phenotype associated with social rank in female and male cynomolgus monkeys: Correlation with kappa opioid receptor availability
2023, Pharmacology Biochemistry and BehaviorDifferential methylation in CD44 and SEC23A is associated with time preference in older individuals
2023, Economics and Human BiologyDevelopmental trajectories of delay discounting from childhood to young adulthood: longitudinal associations and test-retest reliability
2022, Cognitive PsychologyCitation Excerpt :This finding is consistent with the latter interpretation that fitting participants’ DD behavior to a theoretical model introduces systematic error that reduces the reliabilities. Thus, though there are limitations to the use of AUC, we contend that it may better model the tendency of individuals to discount future rewards across a variety of outcomes, and that this general tendency better accounts for response consistency that is typical of personality traits (Odum, 2011b; Roberts, 2009) because it is free from errors that arise due to constraining DD behavior to a particular theoretical model. Cognitive impulsivity, as measured by the delay discounting task, may index broad trait domains such as Conscientiousness (McCrae & Costa, 2003) and Constraint (Tellegen & Waller, 2008) as well as narrower traits of disinhibition or impulsivity (Moreira & Barbosa, 2019).
Using rodent data to elucidate dopaminergic mechanisms of ADHD: Implications for human personality
2024, Personality NeuroscienceDevelopment of a cortical delay discounting assay: A potential biomarker of externalizing disorders
2023, Psychological MedicineA systematic review and meta-analysis of test–retest reliability and stability of delay and probability discounting
2024, Journal of the Experimental Analysis of Behavior