Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2024 Nov 1.
Published in final edited form as: Pharmacol Biochem Behav. 2023 Sep 16;232:173638. doi: 10.1016/j.pbb.2023.173638

The endowment effect and temporal discounting of drug and non-drug commodities

Sean D Regnier 1, Mark J Rzeszutek 5, Justin C Strickland 6, Thomas P Shellenberg 1, William W Stoops 1,2,3,4
PMCID: PMC10947334  NIHMSID: NIHMS1935310  PMID: 37717822

Abstract

Objectives:

Despite a rich history of behavioral economic research on substance use there remains a need for further exploration of behavioral mechanisms that may underlie the etiology or persistence of substance use disorder. The purpose of this study was to measure the association between delay discounting and the endowment effect in people who smoke cigarettes, use cocaine, and controls, using online crowdsourcing.

Methods:

Participants were categorized to a cocaine group (n = 36), cigarette group (n = 48), or control group (n = 47) based on recent reported drug use. Based on group, participants completed up to three delay discounting tasks (i.e., money, cigarettes and cocaine), an endowment effect task for multiple commodities, and other questionnaires.

Results:

Participants in the cocaine and cigarette group demonstrated an increased rate in discounting for money compared to controls. Participants in the cocaine group had a less pronounced endowment effect for beer, compared to controls, as suggested by willingness to accept less to sell beer. A significant negative association was found between endowment ratios for non-drug commodities and delay discounting for cigarettes, but not monetary or cocaine delay discounting, indicating an inconsistent relationship between the two measures.

Conclusions:

These results support prior research demonstrating a relationship between cocaine and cigarette use and delay discounting and extend that work by measuring the association between delay discounting and the endowment effect. Future research should include both loss aversion and endowment effect tasks and compare their relationship with delay discounting among people that use drugs.

Keywords: Behavioral Economics, Endowment Effect, Delay Discounting, Loss Aversion, Substance Use Disorder

Introduction

Over half a million Americans die each year due to drug-related causes (CDC, 2020; 2022) warranting further investigation of the individual factors related to substance use to inform evidence-based treatment and policy. Considerable progress in addiction science can be attributed to the application of behavioral economics in the field (i.e., microeconomic principles relevant to individual decision making; Bickel et al., 2014; Koffarnus & Kaplan, 2018; Strickland & Lacy, 2020). Despite a rich history of behavioral economic research on substance use disorder (Koffarnus & Kaplan, 2018) there remains a need for further exploration of behavioral economic principles that may inform our understanding of drug use patterns.

There are several decision-making biases demonstrated to predict substance use disorder risk. One variable that is especially relevant to substance use is delay discounting: the decline in reinforcing value of a stimulus as a function of its delay to receipt. A substantial body of evidence suggests people who use drugs have higher discounting rates (i.e., devalue a stimulus by delay at a faster rate) than those who do not (Amlung et al., 2017). This has been demonstrated with varied tobacco products (Bickel et al., 1999; DeHart et al., 2020), heroin (Kirby & Petry, 2004), alcohol (Petry, 2001), and cocaine (Garcia-Rodriguez et al., 2013; Heil et al., 2006). Additionally, people with a higher intention to quit smoking typically have lower rates of discounting compared to those who do not (Athamneh et al., 2017). Taken together, delay discounting rates appear to be predictive of substance use frequency and severity. These findings are consistent with a theoretical framework that posits people with a preference for smaller, sooner outcomes may discount the long-term health benefits or adverse consequences of abstinence and drug use, respectively, to a greater degree than those who do not.

Another decision-making bias that appears to be related to substance use is loss aversion. Loss aversion refers to the propensity for potential gains to be outweighed by equivalent losses (Kahneman et al., 1991). In a laboratory setting, loss aversion is typically measured using a decision-making task in which participants accept or reject gambles with a 50/50 chance (coinflip) of gaining or losing money, with gain and loss amounts varying between trials (Thrailkill et al., 2022b; Tom et al., 2007). In this arrangement, lower ratios of gambles accepted to rejected indicate greater loss aversion.

Recently, delay discounting and loss aversion have been shown to be independent factors associated with drug use in people who smoke cigarettes, drink alcohol, and use other drugs. These individuals had higher rates of delay discounting and lower loss aversion compared to non-drug-using groups (Thrailkill et al., 2022a, 2022b). However, in one study, although loss aversion and delay discounting were independently associated with drug use, the authors found an interaction between loss aversion and delay discounting (Thrailkill et al., 2022b): participants who smoked cigarettes and displayed co-occurring high levels of delay discounting had an increased risk of other substance use if they also had low loss aversion.

A decision-making bias believed to be related to loss aversion, and potentially substance use, is the endowment effect. The endowment effect can be described as the observed increase in the subjective valuation of a good after it has been acquired (Kahneman et al., 1990; Morewedge & Giblin, 2015). The endowment effect has been observed for dozens of commodities, including movie tickets, chocolate, soda, time, and intellectual property (Horowitz & McConnell, 2002). One way the endowment effect is measured is using the valuation paradigm. In the valuation paradigm, half of the participants (sellers) are randomly assigned to receive a good and provided an opportunity to sell it back to the experimenter. The other half of participants (buyers) do not receive anything but are provided an opportunity to buy it from the experimenter. In several studies, buyers are willing to pay (WTP) significantly less than sellers are typically willing to accept (WTA), described as the WTA-WTP gap (Morewedge & Giblin, 2015). This measure offers an opportunity to calculate a WTA-WTP ratio (Novemsky & Kahneman, 2005) which can also be called an endowment effect coefficient. Loss aversion is believed to explain the endowment effect through the loss in utility associated with selling a good (i.e., WTA) being greater than the gain associated with acquiring it (i.e., WTP; Ericson & Fuster, 2014; Thaler, 1980).

Although delay discounting and loss aversion appear to be factors related to substance use (Thrailkill et al., 2022a, 2022b), the association between delay discounting and the endowment effect has yet to be explored and extended to drugs like cocaine. Therefore, the purpose of this study was to measure the association of delay discounting and the endowment effect in people who smoke cigarettes, use cocaine, and controls, using an online sample.

Methods

Participants and Recruitment

Amazon Mechanical Turk (MTurk) was used for subject recruitment. MTurk is an online labor market where “requesters” (researchers or organizations requesting information) can post jobs for “workers” (the employees). In this study, the work requested is called a Human intelligence task (“HITs”), which can vary from answering surveys, to more involved tasks such as transcribing audio and conducting web searches. In this case participants answered a survey. The study survey was a HIT on the MTurk website, visible to participants that met initial inclusion criteria. Participants were recruited in April 2016. To be eligible to participate in the study participants must have 1) self-reported being at least 18 years old; 2) been located in the United States; 3) had a Human Intelligence Task (HIT) approval rate of at least 95%; and 4) had at least 50 approved HITs (Robinson et al., 2019). Having a 95% approval rating helps increase the likelihood that participants are filling out their responses truthfully and not randomly.

General Procedure

The informed consent process was completed on the first page of the survey. Those who agreed to the terms of the informed consent completed a 19-question screener regarding their age, sex, and use of cocaine, cigarettes, alcohol, and marijuana. Their responses to the screener questions dictated the flow of the survey and their group assignment (see below). Data from onehundred thirty-one (n = 131) participants were included in this analysis. Participants were categorized as a cocaine group, a cigarette group, or a control group based on screener responses. Participants who reported using cocaine at least once in the past three months, and at least six times during their lifetime during the screener were categorized as the cocaine group (n = 36). These criteria align with prior research (i.e., Garcia-Rodriguez et al., 2013) and allow for a broad range of cocaine use between participants, rather than focusing specifically on those with cocaine use disorder. To be categorized in the cigarette group (n = 48) participants must have 1) reported smoking at least one cigarette per day in the past three months; and 2) report not using prescription or elicit opioids or cocaine at least once in the past three months, or at least six times during their lifetime. Participants who did not report any opioid, cigarette, or cocaine use in the past three months and reported using less than six times during their lifetime during the screener were categorized as the control group (n = 47). All participants completed at least one Delay Discounting Task, an Endowment Effect Task, and other demographics-related questionnaires. Data on cocaine abuse and dependence was collected using the DSM-IV.

Study Tasks

Delay Discounting Task

Five-trial adjusting delay discounting tasks (i.e., “minute discounting”; Koffarnus & Bickel, 2014) were completed by all participants. In these tasks, participants made sequential, hypothetical choices between a delayed reinforcer of a predetermined value (Cocaine = 4 grams; Cigarettes = 4 packs; Money = $1000), and half of the value available immediately (e.g., “Would you rather receive 2 grams of cocaine now or wait to receive 4 grams of cocaine in 3 weeks?”). These discounting tasks measured the delay that decreases subjective value of the delayed reinforcer by 50% (i.e., ED50; Yoon & Higgins, 2008) All participants in the cocaine group completed the cocaine and money delay discounting tasks. To complete the cigarette discounting task participants in the cocaine group must have also reported smoking at least one cigarette in the past three months and regularly smoking at least one cigarette per day. Participants in the cigarette group completed the delay discounting tasks for cigarettes and money, while participants in the control group completed the money discounting task, only.

Endowment Effect Task

All participants completed the endowment effect task. In this task (Matthews et al., 2016), participants were presented a valuation paradigm in which they were given two series of questions and asked the price they would pay (series 1) and accept (i.e., sell; series 2) eight commodities (3-pound bag of jellybeans, 40 deluxe Belgian chocolates, a leather-bound notebook, one year subscription to Scientific American, a Bissell bagless vacuum cleaner, a 14-inch gemstone globe, a Samsung Galaxy Gear smartwatch, and a TomTom navigation device) after being given a hypothetical scenario. These varieties of commodities were selected to produce range of prices. Hypothetical scenarios produce similar results compared to real-life scenarios (Johnson & Bickel, 2002). In the Willing to Pay (WTP) scenario participants were asked to state the maximum amount they would be willing to pay ($.50 to $300; using a slider scale) for each item in a hypothetical auction (i.e., their bid). In the Willing to Accept (WTA) scenario participants were asked to select the minimum amount they would be willing to sell ($.50 to $300) the item to a potential buyer for after they won the item as a prize. This methodology was also extended to WTA/WTP for beer (6-pack; $.50 to $300) cigarettes (1 pack; $.50 to $300), and cocaine (1 gram; $.50 to $300). WTA/WTP for cigarettes and cocaine were only presented to those who indicated smoking cigarettes or using cocaine.

Other Questionnaires

The last block of the Qualtrics survey contained sixteen demographics questions (e.g., age, weight, race) and the Fagerstrom Test for Nicotine Dependence (FTND) (Heatherton et al., 1991). Participants also answered general questions about their frequency of drug use.

Data Analysis

All data analyses were conducted using R 4.2.0 (R Core Team, 2022). The package lme4 (Bates et al., 2015) was used for linear mixed-effects models. The car package (Fox & Weisberg, 2019) was used for type-II Wald χ2 omnibus tests and the emmeans package (Lenth, 2020) was used for post-hoc comparisons. Use of a linear mixed-effects model corrects for fixed effects based on missing data (e.g., control group not completing cigarette and cocaine questions) and emmeans allows for appropriate post-hoc comparisons based on group while controlling for multiple tests. This also means that interactions reported are based on conditions for which interactions were possible (e.g., money and cigarette discounting for cigarette and cocaine groups). The psych package (Revelle, 2020) was used for Spearman rank correlations. An approximation of Cohen’s d was calculated by using the pooled standard deviation from the mixed-effects model (Lenth, 2020). Because data from both the delay discounting and endowment effect tasks were log-transformed prior to analyses these approximated effect sizes are in proportional units. See supplemental document for data not reported in the manuscript.

Delay Discounting

The primary dependent measure of the delay discounting task was the delay discounting rate (i.e., k) for money, cigarettes, and cocaine (if applicable; labeled “condition”). Values of k were acquired by calculating the inverse of the obtained ED50 (Koffarnus & Bickel, 2014). All k values were log10 transformed prior to analyses. A linear mixed-effect model was used to compare k values between condition (i.e., money, cigarette, and cocaine) and group (i.e., cocaine, cigarettes, control). Data were analyzed as group and condition (and their interaction) as fixed effects, while participant was set as the random effect.

Endowment Effect

The endowment effect was defined as the ratio of how much a participant was willing to accept (WTA) and willing to pay (WTP) for the same item (i.e., WTA/WTP ratio; Knetsch & Sinden, 1984; Morewedge & Giblin, 2015). Coefficients were calculated by dividing the participant’s WTA by their WTP for each commodity. Participants were said to have a more pronounced endowment effect if their WTA/WTP coefficient was greater than 1. To reduce the total number of WTA/WTP for analysis purposes, a participant’s median value (to help control for outliers) across the eight WTA/WTP ratios was used for non-drug commodities, with beer, cigarettes, and cocaine analyzed separately. A linear mixed-effect model was used to compare WTA/WTP ratios between commodities (i.e., median non-drug, beer, cigarette, and cocaine) and group (i.e., cocaine, cigarettes, control). Data were analyzed as group and condition (and their interaction) as fixed effects, while participant was set as the random effect. WTA/WTP ratios were log-scaled prior to analysis to approximate normality.

Delay Discounting x Endowment Effect

The relationship between delay discounting and endowment effect was calculated by obtaining Spearman’s rank order correlations for each group based on the tasks completed. For the control group, this consisted of median non-drug and beer WTA/WTP ratios and monetary discounting. For the cigarette group, this also included cigarette WTA/WTP ratios and cigarette discounting. For the cocaine group, this also consisted of cocaine WTA/WTP ratios and cocaine discounting. Obtained Spearman’s coefficients were considered statistically significant at p < 0.05, with a Holm-Bonferroni correction for multiple comparisons.

Other Questionnaires

Demographic variables are presented as a rate (e.g., drinks per week, binge days per month [4 or more for women, 5 or more for men in a two-hour period] cigarettes per day) and were analyzed using a multivariable general linear model with a Bonferroni correction for multiple comparisons. Variables presented as counts (e.g., sex, race, # unemployed) were analyzed using a χ2 Test of Independence across groups (control, cigarettes, cocaine).

Results

Demographics

See Table 1 for an overview of participant demographics and a summary of the other questionnaires. No significant relationships between age, sex, employment, and race were found between groups. Participants in the cigarette group reported consuming a significantly greater number of drinks per week than controls (p = .027). Participants in the cocaine group reported a significantly greater number of drinks per week and binge days per month than those in the cigarette (p = .034; <.001, respectively) and control groups (p <.001; <.001, respectively).

Table 1.

Demographic and drug use for each of the three groups (columns). Sex, employment, cocaine abuse, and cocaine dependence (DSM-IV) are presented as percent of participants that met criteria. Race is presented by count (% of group). All other variables are presented as mean (SD). Asterisks (*) indicate a statistically significant mean difference versus controls was revealed by the multivariate general linear model (with a Bonferroni correction for multiple comparisons). No significant relationships between group and age or group and sex were observed.

Control (n = 47) Cigarette (n = 48) Cocaine (n = 36)
Age 35.8 (10.37) 37.0 (11.04) 32.4 (8.38)
Female 57.45% 52.08% 47.22%
Unemployed 23.4% 14.6% 25.0%
Race
 White 43 (91.49%) 41 (85.42%) 29 (80.56%)
 Black 2 (4.26%) 4 (8.33%) 4 (11.11%)
 Hispanic 1 (2.13%) 3 (6.25%) 2 (5.56%)
 East/Southeast Asian 0 0 1 (2.78%)
 Other 1 (2.13%) 0 0
Alcohol
Drinks/Week 2.11 (4.45) 7.96 (12.92)* 19.22 (19.93)*
Binge Days per Month 1.07 (2.89) 3.21 (6.59) 9.38 (9.18)*
Cigarettes
Cigarettes/Day - 15.58 (4.86) 11.42 (12.89)
FTND - 4.50 (1.81) 4.19 (2.73)
Cocaine
Past Month Cocaine (Days) - - 3.31 (4.15)
Lifetime Cocaine - - 122.56 (234.29)
Cocaine Abuse - - 27.78%
Cocaine Dependence - - 50.00%
Open in a new tab

Delay Discounting

Delay discounting by group can be found in the top panel of Figure 1. There was a significant effect of condition (p < .001, χ2(2) = 118.723) and substance use group p < .001, χ2(2) = 15.192), but no significant interaction between the two (p = .411, χ2(1) = .675). Post-hoc comparisons indicated a significant increase in monetary delay discounting for participants in the cigarette group compared to controls (d = 0.757, p = .027, t(187) = 2.596), and the cocaine group compared to controls (d = 1.22, p < .001, t(187) = 3.882). Furthermore, the cigarette group discounted money less than cigarettes (d = −1.809, p < .001, t(111) = −8.861), while the cocaine group discounted money less than cigarettes (d = −1.536, p < .001, t(116) = −5.832) and cocaine (d = −1.148, p < .001, t(111) = −4.872). See discussion for more details on the interpretation of comparing discount rates across commodities due to magnitude size discrepancies. All additional comparisons for discounting can be found in the supplemental document, Tables S1 and S2.

Figure 1. Top Panel: Discounting by Group and Commodity. Bottom Panel: Median Endowment Ratios by Group.

Figure 1.

Open in a new tab

Top Panel: Boxplots of discounting based on control group (dark grey, left of triads), cigarette group (light grey, center of triads), and cocaine group (white, right of triads). Commodity being discounted is on the x-axis. Values of k are along the y-axis, which is log10 scaled. Bottom Panel: Boxplots of willingness to accept (WTA) and of willingness to pay (WTP) ratios for control group, cigarette group, and cocaine group. Commodities are on the x-axis, median non-drug is the median ratio for all non-drug commodities. Values of WTA/WTP ratios are along the y-axis, which is log10 scaled. Boxes indicate the middle 50% of the distribution, black bars indicate medians, white squares represent the means. *: p < .05. **: p < .001.

Endowment Effect

There was an overall significant effect of commodity (p = .003, χ2(3) = 13.894) and substance use group (p = .023, χ2(2) = 7.504), on WTA/WTP ratios. There was no significant interaction between the two (p = .842, χ2(3) = 0.831). Post-hoc comparisons identified significant differences between cigarette and non-drug median ratios for the cocaine group (d = −0.688, p = .041, t(243) = −2.658) and a significant difference for beer ratios between the cocaine group and control group (d = −0.715, p = .038, t(271) = −2.468). No other significant differences were identified between groups or conditions for WTA/WTP ratios. Boxplots of WTA/WTP ratios coefficients can be found in the bottom panel of Figure 1. All additional comparisons for WTA/WTP ratios can be found in the supplemental document, Tables S3 and S4.

Discounting and Endowment

Spearman correlation matrices for the cigarette group and cocaine group can be found in Tables 2 and 3, respectively. There was a significant positive relationship between non-drug median and beer WTA/WTP ratios (rs = .439) for the control group, but no relationship between WTA/WTP ratios and monetary discounting (rs = .05). For the cigarette group, there were significant positive relationships between non-drug medians, beer, and cigarette WTA/WTP ratios, a significant positive relationship between discounting for cigarettes and money, and significant negative relationships between cigarette discounting and non-drug and cigarette WTA/WTP ratios (see Table 2). For the cocaine group, there were significant positive relationships between non-drug, beer, and cigarette WTA/WTP ratios, but not with cocaine WTA/WTP. All discounting measures were positively related to each other, but monetary discounting was the only measure related to cocaine WTA/WTP (see Table 3). For scatterplots, histograms, and Pearson correlations for all endowment and discounting measures, see supplemental Figures S1, S2, and S3. To see Spearman correlations when all groups were aggregated, see supplemental Table S5.

Table 2.

Spearman rank correlations for the cigarette group between WTA/WTP ratios from median non-drug commodities and beer and cigarette questions as well as k values for discounting of money and cigarettes. *: p < .05. **: p < .001. P-values above the diagonal have been adjusted for multiple comparisons. Participants in the Cigarette group were not asked questions regarding cocaine.

WTA/WTP Ratios Discounting
Non-Drug Beer Cigarette Money k Cigarette k
Non-Drug .576** .327 −.203 −.377*
Beer .576** .537** −.036 −.193
Cigarette .327* .537** −.206 −.425*
Money k −.203 −.036 −.206 .522*
Cigarette k −.377* −.193 −.425* .522**
Open in a new tab

Table 3.

Spearman rank correlations for the cocaine group between WTA/WTP ratios from median non-drug commodities and beer, cigarette, and cocaine questions as well as k values for discounting of money, cigarettes, and cocaine. *: p < .05. **: p < .001. P-values above the diagonal have been adjusted for multiple comparisons.

WTA/WTP Ratios Discounting
Non-
Drug		 Beer Cigarette Cocaine Money
k 		Cigarette
k 		Cocaine
k
Non-Drug .550* .401 .255 .024 −.348 −.201
Beer .550** .542 .307 .063 .087 −.038
Cigarette .401* .542* .364 .127 .249 .034
Cocaine .255 .307 .364 .496 .259 .333
Money k .024 .063 .127 .496* .505 .647**
Cigarette k −.348 .087 .249 .259 .505* .458
Cocaine k −.201 −.038 .034 .333 .647** .458*
Open in a new tab

Discussion

The primary findings of this analysis were that people who use cocaine and people who use cigarettes had increased monetary discounting relative to a control group. There was a minimal difference in the WTA/WTP ratios (i.e., minimal evidence of differential endowment effect) between these groups: participants in the cocaine group had a less pronounced endowment effect for beer, compared to controls, but no other between-group differences were found. There was an inconsistent relationship between discounting and WTA/WTP ratios within and across groups. Monetary discounting was only positively related to WTA/WTP ratios for cocaine, whereas cigarette discounting was negatively related to non-drug and cigarette WTA/WTP ratios in the cigarette group, only.

The results of this study support prior research demonstrating that people who use cocaine (Garcia-Rodriguez et al., 2013; Heil et al., 2006) or cigarettes (Bickel et al., 1999; DeHart et al., 2020) typically discount reinforcers with increasing delay at higher rates compared to controls, and the supposition that high rates of delay discounting may be a predictor of substance use. Some notable differences exist between the present study and Garcia-Rodriguez et al. Although both studies required at least one use in the last three months, participants in Garcia-Rodriguez et al. must have met DSM-IV criteria for cocaine dependence. Additionally, while there was a difference in group-specific demographic information, participants in the current study may have had higher rates of cocaine use (3.31 use days in the last 30 days) than in Garcia-Rodriguez et al. (33.1-45.1 days since last use). Finally, Garcia-Rodriguez et al. utilized a lengthier adjusting amount discounting task, which took approximately 10 minutes to complete. The similar results obtained in the current study provide support for the five-trial adjusting delay discounting task and its use in participants who use cocaine. Although the overall findings were similar, Garcia-Rodriguez et al. also found that delay discounting was significantly greater for participants who used both cocaine and nicotine compared to nicotine only. In the present sample, participants in the Cigarette and Cocaine groups reported similar daily cigarette use and nicotine dependence, making drug-specific effects on discounting difficult to determine. Future research should control for the potential effects of polysubstance use by separating participants into groups specific to only cocaine, alcohol, and cigarettes.

The similar endowment effect outcomes observed between groups is inconsistent with a previous study (Strickland et al., 2017), which found a decrease in WTA/WTP coefficients for participants who used cocaine. However, this was relative to a population normative value rather than a control group. Additionally, participants in that study reported an average of 15.7 use days in the last month, compared to 3.31 in the present study. This suggests that differences in endowment may be observed in those with increased use frequency and severity. In the present study, cigarette and cocaine groups had generally lower WTA/WTP coefficients than controls, but these differences were not statistically significant. Decreases in loss aversion relative to controls have been previously reported in people who smoke cigarettes. Atlhough loss aversion and the endowment effect may be similar constructs, loss aversion does not explain the endowment effect in its entirety (Ericson & Fuster, 2014; Morewedge & Giblin, 2015). Although significant differences in alcohol intake were reported between participants in the cigarette and cocaine groups, it is difficult to compare these results to the loss aversion research with alcohol given the lack of purposeful sampling based on alcohol use (Strickland et al., 2017). Future research is needed to understand the individual relationship between cocaine, cigarettes, and alcohol, and the endowment effect.

This is the first study to assess the relationship between delay discounting and a direct measure of the endowment effect, limiting ability to compare to other literature. However, the limited relationship between the two may be best compared to the prior research on delay discounting and loss aversion. The endowment effect was originally developed as an explanation of loss aversion in choice paradigms without risk (Ericson & Fuster, 2014; Thaler, 1980), as loss aversion arrangements typically involve a gamble (Thrailkill et al., 2022b; Tom et al., 2007). This has resulted in a long history of loss aversion being the primary explanation of the endowment effect (Ericson & Fuster, 2014). Although loss aversion remains the leading explanatory theory of the endowment effect, there is an increasing body of evidence that it may not provide a complete account (Ericson & Fuster, 2014; Morewedge & Giblin, 2015). This may explain the lack of a statistically significant relationship between discounting and the endowment effect found here. This is because participants completing the endowment effect task in the present study were not making decisions under risk, which may suggest that risk is an important variable related to delay discounting. Future studies should include the addition of a loss aversion task to understand the relationship between the two and their connection to delay discounting.

There are several limitations to the present study that must be acknowledged. First, the WTP and WTA tasks included in this study were completed within-subject, which contrasts with prior research that typically includes two independent groups (Kahneman et al., 1990). The consequences of completing endowment effect tasks entirely within subject are unclear, but within-subject designs can produce order effects and larger effect sizes due to decreases in variability compared to between-subject. Second, drug use requirements were less stringent in the present study compared to others measuring similar constructs (Garcia-Rodriguez et al., 2013). Cannabis use was permitted in all three groups, and participants in the cocaine and cigarette group reported significantly more alcohol consumption than controls. This prevents the analysis of the independent relationship of specific drugs on delay discounting and the endowment effect, though these results may generalize to individuals who use multiple substances. Because different discounting commodities were not equal in magnitude (i.e., financial value) differences in k values across commodities must be interpreted with caution. It is likely that if cigarette and cocaine discounting conditions were presented as “$500 of cigarettes/cocaine now or $1000 of cigarettes/cocaine at some delay,” discount rates would decrease (e.g., Holt et al., 2016). However, rank ordered correlations detected that those who discounted money more steeply also were more likely to discount cigarettes and/or cocaine more steeply as would be expected (Odum et al. 2020) despite the differences in magnitude. Finally, data were collected in 2016, which restricts the generalization of these results to drug using populations in the current climate, especially following the COVID-19 pandemic.

Conclusion

The results of this study support prior research demonstrating a significant relationship between cocaine use and delay discounting and showed a weak relationship between discounting and the endowment effect. This is the first study to directly measure the association between these phenomena in a population of persons who use cocaine and cigarettes. The endowment effect and delay discounting do not appear to have a significant relationship in this study, which may be explained by lack of risk in the endowment effect task. Future research should include both loss aversion and endowment effect tasks and compare their relationship with delay discounting in a drug-using sample.

Supplementary Material

1

Highlights.

  • This study measured delay discounting and the endowment effect in an mTurk sample

  • People who reported drug use had increased delay discounting compared to controls

  • People who used cocaine had a less pronounced endowment effect for beer

  • Endowment ratios and delay discounting were generally unassociated

Grant Support

Sean Regnier’s time on this project was supported by grants from the National Institute on Drug Abuse (T32DA035200) of the National Institutes of Health. Mark J. Rzeszutek’s time was supported by the National Institute on Alcohol Abuse and Alcoholism of the National Institutes of Health Grant R01 AA026605 awarded to Mikhail N. Koffarnus. Thomas Shellenberg's time was supported by grants from the National Institute on Drug Abuse (R01DA045023, R01DA047368). This research was also funded by the University of Kentucky’s Department of Behavioral Science and National Science Foundation. The funding agencies had no role in study design, data collection or analysis, or preparation and submission of the manuscript. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health or National Science Foundation.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Amlung M, Vedelago L, Acker J, Balodis I, & MacKillop J (2017). Steep delay discounting and addictive behavior: a meta-analysis of continuous associations. Addiction, 112(1), 51–62. 10.1111/add.13535 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Athamneh LN, Stein JS, & Bickel WK (2017). Will delay discounting predict intention to quit smoking? Exp Clin Psychopharmacol, 25(4), 273–280. 10.1037/pha0000129 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bates D, Machler M, Bolker BM, & Walker SC (2015). Fitting Linear Mixed-Effects Models Using lme4. Journal of Statistical Software, 67(1), 1–48. https://doi.org/DOI 10.18637/jss.v067.i01 [DOI] [Google Scholar]
  4. Bickel WK, Johnson MW, Koffarnus MN, MacKillop J, & Murphy JG (2014). The Behavioral Economics of Substance Use Disorders: Reinforcement Pathologies and Their Repair. Annual Review of Clinical Psychology, Vol 10, 10, 641–677. 10.1146/annurev-clinpsy-032813-153724 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bickel WK, Odum AL, & Madden GJ (1999). Impulsivity and cigarette smoking: delay discounting in current, never, and ex-smokers. Psychopharmacology (Berl), 146(4), 447–454. 10.1007/pl00005490 [DOI] [PubMed] [Google Scholar]
  6. DeHart WB, Friedel JE, Berry M, Frye CCJ, Galizio A, & Odum AL (2020). Comparison of delay discounting of different outcomes in cigarette smokers, smokeless tobacco users, e-cigarette users, and non-tobacco users. J Exp Anal Behav, 114(2), 203–215. 10.1002/jeab.623 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Ericson KMM, & Fuster A (2014). The Endowment Effect. Annual Review of Economics, Vol 6, 6, 555–579. 10.1146/annurev-economics-080213-041320 [DOI] [Google Scholar]
  8. Fox J, & Weisberg A (2019). An R companion to applied regression (3rd ed.). https://doi.org/https://socialsciences.mcmaster.ca/jfox/Books/Companion/ [Google Scholar]
  9. Garcia-Rodriguez O, Secades-Villa R, Weidberg S, & Yoon JH (2013). A systematic assessment of delay discounting in relation to cocaine and nicotine dependence. Behav Processes, 99, 100–105. 10.1016/j.beproc.2013.07.007 [DOI] [PubMed] [Google Scholar]
  10. Heatherton TF, Kozlowski LT, Frecker RC, & Fagerstrom KO (1991). The Fagerstrom Test for Nicotine Dependence: a revision of the Fagerstrom Tolerance Questionnaire. Br J Addict, 86(9), 1119–1127. 10.1111/j.1360-0443.1991.tb01879.x [DOI] [PubMed] [Google Scholar]
  11. Heil SH, Johnson MW, Higgins ST, & Bickel WK (2006). Delay discounting in currently using and currently abstinent cocaine-dependent outpatients and non-drug-using matched controls. Addict Behav, 31(7), 1290–1294. 10.1016/j.addbeh.2005.09.005 [DOI] [PubMed] [Google Scholar]
  12. Holt DD, Glodowski K, Smits-Seemnan, & Tiry AM (2016). The domain effect in delay discounting: The roles of fungibility and perishability. Behavioral Processes, 131, 47–52. [DOI] [PubMed] [Google Scholar]
  13. Horowitz JK, & McConnell KE (2002). A review of WTA/WTP studies. Journal of Environmental Economics and Management, 44(3), 426–447. 10.1006/jeem.2001.1215 [DOI] [Google Scholar]
  14. Johnson MW, & Bickel WK (2002). Within-subject comparison of real and hypothetical money rewards in delay discounting. J Exp Anal Behav, 77(2), 129–146. 10.1901/jeab.2002.77-129 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Kahneman D, Knetsch JL, & Thaler RH (1990). Experimental Tests of the Endowment Effect and the Coase Theorem. Journal of Political Economy, 98(6), 1325–1348. https://doi.org/Doi 10.1086/261737 [DOI] [Google Scholar]
  16. Kahneman D, Knetsch JL, & Thaler RH (1991). Anomalies - the Endowment Effect, Loss Aversion, and Status-Quo Bias. Journal of Economic Perspectives, 5(1), 193–206. https://doi.org/DOI 10.1257/jep.5.1.193 [DOI] [Google Scholar]
  17. Kirby KN, & Petry NM (2004). Heroin and cocaine abusers have higher discount rates for delayed rewards than alcoholics or non-drug-using controls. Addiction, 99(4), 461–471. 10.1111/j.1360-0443.2003.00669.x [DOI] [PubMed] [Google Scholar]
  18. Knetsch JL, & Sinden JA (1984). Willingness to Pay and Compensation Demanded - Experimental-Evidence of an Unexpected Disparity in Measures of Value. Quarterly Journal of Economics, 99(3), 507–521. https://doi.org/Doi 10.2307/1885962 [DOI] [Google Scholar]
  19. Koffarnus MN, & Bickel WK (2014). A 5-trial adjusting delay discounting task: accurate discount rates in less than one minute. Exp Clin Psychopharmacol, 22(3), 222–228. 10.1037/a0035973 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Koffarnus MN, & Kaplan BA (2018). Clinical models of decision making in addiction. Pharmacol Biochem Behav, 164, 71–83. 10.1016/j.pbb.2017.08.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Lenth R. (2020). emmeans: Estimated marginal means, aka least-squares means (1.5.2-1). In [Google Scholar]
  22. Matthews WJ, Gheorghiu AI, & Callan MJ (2016). Why do we overestimate others' willingness to pay? Judgment and Decision Making, 11(1), 21–39. <Go to ISI>://WOS:000369237900003 [Google Scholar]
  23. Morewedge CK, & Giblin CE (2015). Explanations of the endowment effect: an integrative review. Trends in Cognitive Sciences, 19(6), 339–348. 10.1016/j.tics.2015.04.004 [DOI] [PubMed] [Google Scholar]
  24. Novemsky N, & Kahneman D (2005). The boundaries of loss aversion. Journal of Marketing Research, 42(2), 119–128. https://doi.org/DOI 10.1509/jmkr.42.2.119.62292 [DOI] [Google Scholar]
  25. Odum AL, Becker RJ, Haynes JM, Galizio A, Frye CCJ, Downey H, Friedel JE, & Perez DM (2020). Delay discounting of different outcomes: Review and theory. Journal of the Experimental Analysis of Behavior, 113(3), 657–679. 10.1002/jeab.589 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Petry NM (2001). Delay discounting of money and alcohol in actively using alcoholics, currently abstinent alcoholics, and controls. Psychopharmacology (Berl), 154(3), 243–250. 10.1007/s002130000638 [DOI] [PubMed] [Google Scholar]
  27. R Core Team. (2022). R: A language for statistical computing (4.2.0). . In R Foundation for Statistical Computing. [Google Scholar]
  28. Revelle W. (2020). psych: Procedures for Psychological, Psychometric, and Personality Research (2.0.8). https://doi.org/https://CRAN.R-project.org/package=psych [Google Scholar]
  29. Robinson J, Rosenzweig C, Moss AJ, & Litman L (2019). Tapped out or barely tapped? Recommendations for how to harness the vast and largely unused potential of the Mechanical Turk participant pool. PLoS One, 14(12), e0226394. 10.1371/journal.pone.0226394 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Strickland JC, Beckmann JS, Rush CR, & Stoops WW (2017). A pilot study of loss aversion for drug and non-drug commodities in cocaine users. Drug Alcohol Depend, 180, 223–226. 10.1016/j.drugalcdep.2017.08.020 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Strickland JC, & Lacy RT (2020). Behavioral economic demand as a unifying language for addiction science: Promoting collaboration and integration of animal and human models. Exp Clin Psychopharmacol, 28(4), 404–416. 10.1037/pha0000358 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Thaler R. (1980). Toward a Positive Theory of Consumer Choice. Journal of Economic Behavior & Organization, 1(1), 39–60. https://doi.org/Doi 10.1016/0167-2681(80)90051-7 [DOI] [Google Scholar]
  33. Thrailkill EA, DeSarno M, & Higgins ST (2022a). Intersections between environmental reward availability, loss aversion, and delay discounting as potential risk factors for cigarette smoking and other substance use. Prev Med, 107270. 10.1016/j.ypmed.2022.107270 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Thrailkill EA, DeSarno M, & Higgins ST (2022b). Loss aversion and risk for cigarette smoking and other substance use. Drug Alcohol Depend, 232, 109307. 10.1016/j.drugalcdep.2022.109307 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Tom SM, Fox CR, Trepel C, & Poldrack RA (2007). The neural basis of loss aversion in decision-making under risk. Science, 315(5811), 515–518. 10.1126/science.1134239 [DOI] [PubMed] [Google Scholar]
  36. Yoon JH, & Higgins ST (2008). Turning k on its head: comments on use of an ED50 in delay discounting research. Drug Alcohol Depend, 95(1-2), 169–172. 10.1016/j.drugalcdep.2007.12.011 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS1935310-supplement-1.docx (2.3MB, docx)

RESOURCES