Preliminary findings on cryptocurrency trading among regular gamblers: Anew risk for problem gambling? PDF

Preview

Summary

Addictive Behaviors 92 (2019) 136–140 Contents lists available at ScienceDirect Addictive Behaviors journal homepage: www.elsevier.com/locate/addictbeh Short Communication Preliminary findings on cryptocurrency trading among regular gamblers: A T new risk for problem gambling? ⁎ Devin J. Mills , Lia ...

Description

Accelerat ing t he world's research.

Preliminary findings on cryptocurrency trading among regular gamblers: Anew risk for problem gambling? Lia Nower Addictive Behaviors

Cite this paper

Downloaded from Academia.edu 

Get the citation in MLA, APA, or Chicago styles

Related papers

Download a PDF Pack of t he best relat ed papers 

GBQ VALIDAT ION 1 An Assessment of t he Validit y of t he Gamblers Belief Quest ionnaire Kahlil Philander, Sally Gainsbury Prevalence and Correlat es of Problem Gambling in a Represent at ive Sample of Norwegian 17-Year-Ol… Alex Blaszczynski Work in Progress- Crypt onomics: Invest ment behaviour in t he crypt ocurrency market Mohamad Saleh

Addictive Behaviors 92 (2019) 136–140

Contents lists available at ScienceDirect

Addictive Behaviors journal homepage: www.elsevier.com/locate/addictbeh

Short Communication

Preliminary findings on cryptocurrency trading among regular gamblers: A new risk for problem gambling?

T

Devin J. Mills , Lia Nower ⁎

Rutgers University, Center for Gambling Studies, 390 George Street, Room 706, New Brunswick, NJ 08901, United States

H I GH L IG H T S

> 50% of regular gamblers have traded cryptocurrencies in the past year • Trading is linked to high-risk stock trading • Trading cryptocurrencies cryptocurrencies is associated with risk for problem gambling, depression and anxiety • Further directions for cryptocurrency research are provided •

A R T I C LE I N FO

A B S T R A C T

Keywords: Gambling Cryptocurrency Addiction Mental Health Stock Market

Cryptocurrencies are emerging digital currencies that allow anonymity in accessing various risk-taking activities through the Internet (e.g., drugs, gambling). However, given conceptual links to high-risk stocks, the present study sought to assess the association between trading cryptocurrencies and problem gambling. Data was collected through a cross-sectional online survey. Advertisement for the survey was posted on Amazon's Mechanical Turk. Participants were adults who had gambled at least monthly in the past year (N = 876; 58.33% male; M = 33.74 years, SD = 9.73). Participants completed the Problem Gambling Severity Index, Patient Health Questionnaire (2-item version), and Generalized Anxiety Disorder scale (2-item version). Trading cryptocurrencies is strongly associated with problem gambling severity (r = 0.53, p < .001). Results from a linear regression with backwards elimination revealed that sports betting, daily fantasy sports, high-risk stock trading, and problem gambling severity contribute to trading cryptocurrencies more frequently in the past year, whereas gambling in on-land casinos contributed to less cryptocurrency trading. Finally, trading cryptocurrencies overlapped strongly with trading high-risk stocks. Moreover, gamblers who engaged in both forms of trading reported greater problem gambling and depression and anxiety symptoms relative to those trading either cryptocurrencies or high-risk stocks, but not both. The present results suggest that trading cryptocurrencies may be appealing to gamblers that are exhibiting greater problem gambling severity. Future research should begin to include cryptocurrency trading in screening, assessment, and treatment protocols, particularly with regular gamblers.

1. Introduction Cryptocurrency or “crypto” is digital money, created and transferred by computers in limited supply that is unrestricted by any bank or country. Unlike cash, it is held and traded with an anonymous user ID, providing anonymity and freedom from asset disclosure. Underlying crypto is blockchain technology, which ensures security and data integrity and hides the value of the currency and the identity of the holder unless they have the digital key (Singhal, Dhameja, & Panda, 2018). A limited but growing body of research is focused on understanding how

⁎

cryptos are used to hide individuals' online gambling activity and/or their access to illicit drugs through the “Dark Net” (Gainsbury & Blaszczynski, 2017; Orsolini, Papanti, Corkery, & Schifano, 2017). Yet, a further concern that has not been given any attention in research is whether or not trading cryptos is an appealing activity among gamblers, and if crypto trading is associated with increased risk for problem gambling. Users generally see cryptos as an investment versus a currency (Gao, Clark, & Lindqvist, 2015), despite the intense volatility of crypto markets. For example, Bitcoin lost > 70% of its value over nine months in

Corresponding author. E-mail addresses: [email protected] (D.J. Mills), [email protected] (L. Nower).

https://doi.org/10.1016/j.addbeh.2019.01.005 Received 11 October 2018; Received in revised form 7 January 2019; Accepted 7 January 2019 Available online 08 January 2019 0306-4603/ © 2019 Published by Elsevier Ltd.

Addictive Behaviors 92 (2019) 136–140

D.J. Mills, L. Nower

2018 (CoinMarketCap, 2018b, 2018a). Furthermore, crypto markets have increased the access to trading options, which significantly increases the risk and volatility associated with this activity (Wilmoth, 2018). In this way, trading cryptos is analogous to trading high-risk stocks (e.g. margins and options), which are highly variable, frequently traded rather than held, and preferred by those with a high tolerance for risk. Conceptually, there are many similarities between trading cryptos and high-risk stocks that stems, in large part, from the volatility that generally discourages investors with a lower tolerance for risk (see review by Arthur, Williams, & Delfabbro, 2016). Although stock trading is typically conceptualized as investment toward future gain, several studies have identified stock trading patterns (i.e. day-trading) and investment choices (e.g. margins, options) that mimic problem gambling behavior (Arthur, Delfabbro, & Williams, 2015; Grall-Bronnec et al., 2017; Youn, Choi, Kim, & Choi, 2016); analogous to the “rush” experienced by action-oriented gamblers, high-risk stock traders buy and sell highly variably stocks for the highs and lows of the volatility alone rather than to nurture a growing investment over time. Likewise, Griffiths (2018) has identified the craving and chasing associated with crypto trading — “crypto-addiction” — which may be appealing to problem gamblers. Additionally, highly frequent crypto trading may be predicted by issues with depression and anxiety, as observed in many of the case studies of excessive stock traders that were recently reviewed by Grall-Bronnec et al. (2017). The present study represents a first step in exploring crypto trading among regular gamblers (i.e., those who gamble at least every month). Crypto trading is hypothesized to be positively associated with problem gambling, and given conceptual the overlaps, positively associated with high-risk stock trading. The present study explores the psychological harms associated with crypto trading frequency. It is hypothesized that crypto trading will be positively associated with depression and anxiety symptoms. Finally, the present study explores which factors (e.g., demographics, gambling frequency, problem gambling, depression) are most strongly associated with crypto trading frequency among regular gamblers.

Table 1 Differences in crypto trading frequency across demographics and thresholds for risk of problem gambling, depression, and anxiety. Frequency of trading cryptocurrency N

Total Sample

%

876

SD

M

1.71

Gender (t(854.96) = 8.24, p < .001, d = 0.55) Male 511 58.33% 2.15 Female 365 41.67% 1.10 Race (F(3, 872) = 8.25, p < .001, ηp2 = 0.03) White or Caucasian 362 41.32% 1.52a Hispanic 205 23.40% 2.31b Black or African 197 22.49% 1.58a American Asian and Other 112 12.79% 1.47a

Lower

Upper

2

1.58

1.84

2.07 1.71

1.97 0.92

2.33 1.27

1.91 2.13 1.95

1.32 2.01 1.31

1.71 2.6 1.86

1.89

1.12

1.83

1.61

1.95

0.83

1.7

1.45

1.92

1.87 1.99

1.08 1.55

1.61 1.92

2.06

1.71

2.23

ηp2 = 0.10) 2.05 2.03 1.71 0.87 2.05 2.08

2.54 1.2 2.64

Marital Status (F(2, 873) = 2.01, p = .135, ηp2 = 0.01) Married, Living 538 61.42% 1.78 2.03 with a partner 67 7.65% 1.27 1.79 Separated, Divorced, Widowed Single (Never been 271 30.94% 1.68 1.97 married) Income (F(2, 873) = 5.38, p = .005, ηp2 = 0.01) Less than $30,000 195 22.26% 1.34a $30,000 to 439 50.11% 1.73a,b $69,999 $70,000 or more 242 27.63% 1.97b Gambling Medium (F(2, 873) = 50.44, p < .001, Primarily Online 249 28.42% 2.29a Primarily On-Land 417 47.60% 1.04b Equally Online and 210 23.97% 2.36a On-Land Problem Gambling Severity Non-Problem 78 (Score 0) Low-Risk (Score 1 139 to 2) Moderate Risk 246 (Score 3 to 7) High Risk (Score 8 413 or more)

2. Methods 2.1. Participants and procedure The university ethics board approved the present study prior to data collection. A request for “frequent gamblers” over the age of 18 was posted on Amazon's Mechanical Turk (MTurk) with an incentive of $0.75. MTurk qualifications were used to only recruit those residing in the U.S. and having obtained a 95% approval rate. MTurk is reliable and valid approach for recruiting highly engaged gamblers and for conducting social science research more generally (Buhrmester, Kwang, & Gosling, 2011; Kim & Hodgins, 2017). Online recruitment was utilized to access participants who are more likely to be familiar with interactive technologies and crypto currency. Participants were provided complete information regarding the study prior to beginning the survey. Although participants were not made aware of the cutoff for gambling frequency, those that did not participate in at least one gambling activity monthly were not eligible for participation in the study. Additionally, as part of the consent form, each participant was informed that their surveys would be reviewed before payment would be approved in accordance with MTurk guidelines. A total of 1124 completed the survey, however, 116 failed one or both of the attention items, and were thus excluded. Participants were also excluded for completing the entire survey suspiciously fast (i.e., in less than five minutes; n = 63) and for submitting only extreme responses (n = 42). Finally, of the remaining 903 participants, 27 (2.90%) did not submit responses to one or more items, and were thus excluded. The final sample of was 876 regular gamblers; nearly half

95% CI

(PGSI) (F(3, 872) = 72.21, p < .001, ηp2 = 0.20) 8.90% 0.32a 0.83 0.13 0.51 15.87% 28.08% 47.15%

0.96a,b

1.63

0.69

1.24

1.03

b

1.67

0.82

1.24

2.63

c

2.02

2.44

2.83

Depression (PHQ-2) (t(451.64) = 8.35, p < .001, d = 0.63) No Risk (Score 605 69.06% 1.33 1.81 1.18 < 3) No Risk (Score 3 or 271 30.94% 2.57 2.13 2.31 more) Anxiety (GAD-2) (t(522.46) = 6.34, p < .001, d = 0.46) No Risk (Score 586 66.89% 1.41 1.87 < 3) Risk (Score 3 or 290 33.11% 2.33 2.09 more)

1.47 2.82

1.25

1.56

2.09

2.57

For variables with more than two groups, means with different superscripts indicate a significant post-hoc test (p < .05).

gamble only at land-based venues, with about a fourth gambling only online and a fourth gambling both online and in land-based venues. Sample characteristics are presented in Table 1. 2.2. Measures The present study includes data from the demographics, gambling involvement, problem gambling, and mental health sections of a large online survey. Participants rated their level of engagement in gamblingrelated activities including crypto and high-risk stock trading, ranging from “Not at all” (0) to “4 or more times per week” (6). Next, participants completed the 9-item Problem Gambling Severity Index (PGSI), a 137

Addictive Behaviors 92 (2019) 136–140

D.J. Mills, L. Nower

cryptos, high-risk stocks, and both cryptos and high-risk stocks in the past year. > 63% of gamblers (n = 552) traded either cryptos (n = 77; 59.74% male), high-risk stocks (n = 87; 63.21% male), or both in the past year (n = 388; 71.39% male). Although differences in gender were noted, results from a Chi square test indicated that these differences did not reach significance (p = .07). Additionally, these groups did not differ in terms of participants' race (p = .75) or marital status (p = .14). However, there were differences when comparing income levels (χ2(4) = 10.16, p = .04). Specifically, for those who solely traded cryptos, only 19.48% were high-earners (i.e., annual income of $70,000 or more) relative to 36.78% of those who solely traded high-risk stocks. No significant differences in incomes were noted between those who traded both cryptos and high-risk stocks. There were significant differences among these three groups in terms of problem gambling (F(2,549) = 47.20, p < .001, ηp2 = 0.15), depression symptoms (F(2,549) = 8.42, p < .001, ηp2 = 0.03), and anxiety symptoms (F(2,549) = 3.95, p = .02, ηp2 = 0.01). Specifically, those who traded both cryptos and high-risk stocks reported high levels of problem gambling severity (M = 12.26, SD = 6.65), depression (M = 2.31, SD = 1.88), and anxiety (M = 2.33, SD = 1.85) compared to those having traded only cryptos or high-risk stocks (problem gambling: Mcryptos = 7.01, SD = 5.24; Mstocks = 6.26, SD = 5.18; depression: Mcryptos = 1.56, SD = 1.59; Mstocks = 1.69, SD = 1.66; anxiety: Mcryptos = 1.81, SD = 1.82; Mstocks = 1.90, SD = 1.67). The differences between those solely trading either cryptos or high-risk stocks did not reach significance.

widely used measure for assessing problem gambling in the general population (Ferris & Wynne, 2001). Finally, anxiety and depression symptoms were assessed using two-item versions of the Patient Health Questionnaire (PHQ-2) and Generalized Anxiety Disorder scale (GAD-2) (Donker, van Straten, Marks, & Cuijpers, 2011; Löwe, Kroenke, & Gräfe, 2005). The PGSI, PHQ-2, and GAD-2 offer cutoffs for establishing risk of gambling severity, depression, and anxiety, which are presented in Table 1. These classifications are used only for a descriptive purpose, as the composite scores will be used in the primary analyses. 2.3. Statistical analyses Student's t-tests and one-way analyses of variance (ANOVAs) were conducted to compare crypto trading frequency across various demographics and thresholds for risk of problem gambling, depression, and anxiety. Bivariate correlations were computed to assess the relation between crypto trading frequency and age, frequency of engaging in gambling activities, problem gambling, and depression and anxiety symptoms. Finally, a stepwise regression with backwards elimination was used to identify predictors of crypto trading frequency. This type of regression was selected given both the quantity of variables considered as possibly explaining crypto trading frequency and the exploratory nature of the present study. 3. Result 3.1. Bivariate associations

4. Discussion Of the total sample, 53.08% (n = 465) traded cryptocurrency with at least some frequency in the past year. There were no significant differences for crypto trading frequency by marital status; however, there were differences by gender, race, and income, with males, Hispanics, and higher earners reporting greater trading frequency relative to their respective counterparts (see Table 1). Additionally, those at moderate- and high-risk for problem gambling or at risk for either depression or anxiety reported more frequent crypto trading in the past year relative to their counterparts who are either not at risk for problem gambling, depression, or anxiety. Finally, bivariate correlations show that crypto trading frequency was negatively correlated with age (r = −0.26, p < .001), but was positively correlated with problem gambling (r = 0.53, p < .001) and symptoms of depression (r = 0.29, p < .001) and anxiety (r = 0.20, p < .001). Crypto trading frequency was also positively correlated with frequency of engaging in all other gambling activities ranging from purchasing scratch tickets (r = 0.17, p < .001) to games of skill (r = 0.42, p < .001) to high-risk stock trading (r = 0.67, p < .001).

The present study is the first to explore the prevalence of crypto trading among regular gamblers and its association with problem gambling as well as symptoms of depression and anxiety. More than half of the regular gamblers reported trading cryptos in the past year. Demographically, greater crypto trading was reported by males, Hispanics, higher earners, and/or online gamblers relative to their direct counterparts. Crypto trading was associated with greater engagement in all other gambling activities as well as with problem gambling, depression, and anxiety symptoms. However, within a linear model, depression and anxiety were not significant predictors of crypto trading frequency suggesting that crypto trading by itself may not be associated with psychological harm. Greater crypto trading frequency was associated with increased sports betting, daily fantasy sports playing, highrisk stock trading, and greater problem gambling severity, and a preference for online gambling. Collectively, these preliminary findings suggest that crypto trading may appeal to those who gamble at least monthly as a new, largely unregulated risk-taking activity that offers a rush similar to other gambling-related activities. The present findings, albeit preliminary, do suggest that crypto trading could be conceived as an extension to high-risk day trading activity; especially do to the high overlap between crypto and high risk stock traders with > 75% of high-risk stock traders also trading cryptos. Additionally, crypto and high risk stock traders also shared similar demographics and gambling preferences. For instance, previous studies have found that high-risk stock traders tend to be male, White, married, and wealthy as well as more frequently bet on sports, play poker, or engage in games of skills relative to non-high-risk stock traders (see Arthur et al., 2015; Arthur & Delfabbro, 2017). The present findings suggest that crypto traders largely match these demographics expect in the case of race, where Hispanics were found to trade cryptos more frequently. Moreover, gamblers trading either cryptos or high-risk stocks, but not both, did not differ significantly in their reports of problem gambling or symptoms of depression or anxiety. However, gamblers trading both cryptos and high-risk stocks reported higher problem gambling and greater depression and anxiety symptoms than to gamblers trading either cryptos or high-risk stocks. Thus, crypto and high-risk stock trading appears have an additive effect on the

3.2. Multivariate analyses Participant demographics, level of engagement in gambling activities, and composite scores on the PGSI, PHQ-2, and GAD-2 were included as predictors within a stepwise linear regression with backwards elimination predicting crypto trading frequency. The resulting model accounted for 53.73% variance of crypto trading frequency in the past year (F(3, 872) = 8.25, p < .001). As shown in Table 2, more frequent crypto trading was predicted by greater engagement in sports betting, daily fantasy sports, high-risk stock trading, and increased problem gambling severity. Additionally, preference for gambling in land-based casinos was a negative predictor indicating that those gambling equally online and in on-land casinos and those that gamble exclusively online tend to trade cryptos more ...