Cryptoasset Psychology: Applying Insights from Behavioral Economics to Bitcoin
Note: This article is a word-for-word reproduction of a short 2015 Term Paper I wrote for a class at the University of Chicago taught by Boaz Keysar called “The Psychology of Decision Making.”
This piece uses the lenses of behavioral economics and behavioral psychology to explore roadblocks (as I perceived them in 2015) for bitcoin’s adoption as a payments mechanism, although it’s easy to extend the concepts to other cryptoassets and cryptocurrencies. At the time, I wasn’t aware of the Digital Gold or Store of Value hypotheses so I don’t consider cryptoassets in that context here. The References section is great if you’re interested in digging into seminal behavioral economics papers.
Written by Johnny Antos, originally submitted December 15, 2015.
Biases in Bitcoin Adoption and Implications for the Future Digital Currency
Table of Contents
- The Issue
- The Phenomena
- Conclusions and Extensions
- Full Paper
What is Bitcoin? About 49% of Americans have never heard of Bitcoin and of the 51% who have heard of it, only 3% have actually used it (2014 CSBS Consumer Survey). In short, Bitcoin is a peer-to-peer digital currency that requires no central intermediary. It has gained more traction than any other past digital currency due to the fact that no central, trusted intermediary is required for a transaction. When Alice sends Bitcoin to Julia, it is immediately verified through a decentralized network (not a bank, as in the past) and recorded in a public ledger (Popper, 2015, p. 358). With investors pouring money into Bitcoin-based companies at a faster rate than during the early days of the Internet, there has been a buzzing interest in Bitcoin and its underlying technology amongst elite tech circles throughout the world. However, even if Bitcoin is one of the greatest human technological achievements since the Internet and the growth of Information Technology, widespread adoption by consumers and businesses (meaning people using Bitcoin in transactions on a regular basis) has been slow and has only changed marginally over Bitcoin’s 7-year lifespan (Popper, 2015, p. 305).
Current research in digital currencies focuses on economic incentive structures and the underlying technology while ignoring many of the psychology of decision-making biases that present the steepest hurdle to adoption. Treating digital currencies as a classic network good (where each consumer’s utility from using a product increases exponentially as more consumers use it; think Facebook) suggests that, given enough time, widespread adoption is likely; however, analysis using only traditional utility theory as a basis masks the full story (Popper, 2015, p. 113). Kahneman and Tversky’s (1974) availability bias is a steep barrier to Bitcoin adoption due to the general public’s vivid memories of news stories tying Bitcoin to illegal activity, including the fall of the Silk Road, an Amazon for illicit substances, as well as other Bitcoin exchange security breaches. Complexity bias (Kruger, 1999) combines with loss aversion and status quo bias (Kahneman & Tversky, 1984) to be a potent influence on adoption; if people are fine using credit cards or cash, why should they switch to a new payment system that they perceive is extremely complex and too difficult for them to use effectively? Lastly, it can be hypothesized that using digital currency may entirely alter the conception of value. This may have negative effects on risk preferences, gambling, and spending tendencies.
In October 2014, the FBI conducted a sting operation to arrest Ross William Ulbricht, the secretive creator of the Silk Road, an online marketplace for peer-to-peer illicit drug sales. A large part of the Silk Road’s success was that buyers and sellers interacted using Bitcoin, which was widely believed to be a secure and anonymous method of payment as opposed to credit cards (Bearman & Hanuka, The Untold Story of the Silk Road). After the arrest, major media outlets widely reported that there had been an online drug marketplace operating, and emphasized the key to their business model had been a nebulous digital currency called Bitcoin. This was the first time that many people in the general public heard the term “Bitcoin”, and in the reporting, it was almost inseparable from the illicit activity that Ross had employed it for (Popper, 2015, p. 245).
Tversky and Kahneman (1974) demonstrate that an availability bias sometimes exists when people evaluate the probability of a certain event or the frequency of a class. For example, as opposed to evaluating the success of a new business venture based on a balanced analysis of costs and benefits and incorporating financial projections and multiple interviews with industry insiders, someone may mentally evaluate the business’ success based on imagining the many difficulties that the business could encounter. Tversky and Kahneman also demonstrate that often people employ an anchoring and adjustment heuristic; that is, people set some initial anchor (perhaps based on prior knowledge or the phrasing of the situation) and then adjust away from the anchor to get to an estimate that they feel is correct. The researchers showed that when people were asked to estimate the percentage of African countries in the United Nations, the initial anchor (which was set by the random spin of a wheel) had a significant effect on participants’ adjustment process. Specifically, participants adjusted insufficiently from the initial arbitrary anchor (Tversky & Kahneman, 1974).
In short, because the high-profile arrest of Ross Ulbricht and Bitcoin’s ties to the illegal Silk Road permeates in the general public’s mind, it is the most available information that people rely upon when evaluating or discussing Bitcoin. When a prominent entrepreneur went to Silicon Valley in late 2014 to raise money for a new Bitcoin company, a key concern was that, “for those who had heard of [Bitcoin], the first question was always about whether it was anything more than a token for online drug dealers.” (Popper, 2015, p. 158) When forming opinions about Bitcoin, instead of taking into account the thousands of completely legal Bitcoin transactions that occur successfully every day between law-abiding individuals, people (and even sophisticated Silicon Valley investors) overweight the Silk Road incident. Put another way, they anchor in “Bitcoin is only used for illegal purposes” and there is very little adjustment towards “Bitcoin has many innovative uses that are perfectly legal” because the general public almost never hears about the many completely legal $10 Bitcoin transactions between legitimate users.
This availability bias persists even within the tech-savvy community. There have been several large-scale hacking incidents, in which malicious groups of programmers attack Bitcoin exchanges and are able to steal millions of dollars in Bitcoin (Popper, 2015, p. 309). Programmers began to anchor in “Bitcoin is not secure” and fail to adequately adjust to “Bitcoin is mostly secure”. When asked about the importance of various factors in whether they would use Bitcoin, about 20% of people say they are extremely worried about the safety of Bitcoin, and 75% are extremely doubtful about its security (2014 CSBS Consumer Survey). However, much has changed in Bitcoin since the days of frequent hacks, and the truth is that the underlying Bitcoin technology is extremely secure (Popper, 2015, p. 342).
Bitcoin is extremely complex, even for computer science geeks. Bitcoin runs on the Blockchain, which is an entirely new protocol, similar to how the Internet runs on the IP protocol. It is nearly impossible for an expert to fully explain exactly how the Bitcoin system works if the person has no background knowledge, and even then, it is challenging (Popper, 2015, p. xi-xiii). Svenson (1981) demonstrated that an overconfidence effect is prevalent when people are asked to rate their own skill relative to that of a peer group. For example, 70–80% of people believe themselves to be in the safer half of all drivers, even though this cannot be. Kruger (1999) demonstrated that as opposed to an overconfidence effect, where individuals tend to be overconfident in their own abilities relative to that of a peer group a “below-average” effect also exists in specific scenarios. Namely, Kruger showed that when a person perceives that their absolute skills are high, the overconfidence effect exists; however, in domains where people perceive their absolute skills to be low, there is an underconfidence effect, or complexity bias. For example, most people consider their level of absolute skill in juggling to be low; thus, people tend to estimate that they are below-average compared to their peer group in the domain of juggling (Kruger, 1999).
Only 3% of people have ever used Bitcoin, and even if all of these people understand Bitcoin perfectly (which is an aggressive assumption), that means that 97% of people have no idea as to how the digital currency system works (2014 CSBS Consumer Survey). Even people who are deeply involved with digital currencies have shown significant ability to misunderstand how the technology works. For example, the vast majority of people have the perception that because an illegal online marketplace like the Silk Road would use Bitcoin, it must be anonymous and untraceable (Bearman & Hanuka, The Untold Story of the Silk Road). However, this was shown to be stunningly false when the two FBI agents who arrested Ross Ulbricht were found to have laundered about $1M in Bitcoin into their own pockets during the investigation, because they believed it to be anonymous (Popper, 2015, p. 249. Thus, if even the elite, technology-focused FBI agents showed a significant misunderstanding of how Bitcoin works, the general public’s understanding of Bitcoin is close to zero. This likely plays a significant role for the 25% of people who had never heard of Bitcoin, but who say that they will never use it (2014 CSBS Consumer Survey). Complexity bias and an underconfidence manifests itself in the domain of digital currencies and Bitcoin, as people are likely to drastically underestimate their ability to use Bitcoin because they perceive their absolute skills in the domain of the digital world as low. However, people also perceived that there was high complexity in the shift from cash payments to credit card payments and then to online payments (Popper, 2015, xi).
Any large-scale shift to Bitcoin would require consumers and businesses to make a choice between adopting the new system or sticking with the old status quo. Tversky and Kahneman (1984) demonstrated that loss aversion is a robust quality, in that people strongly prefer avoiding losses to pursuing gains. Similarly, Kahneman, D., Knetsch, J., & Thaler, R. (1990) showed that an endowment effect exists, in that people attribute value to a particular object simply because it is already in their possession. When combined, these produce status quo bias (Kahneman, D., Knetsch, J., & Thaler, R., 1991) which is a strong emotional preference for the current state of affairs and an aversion to a shift away from the current state. A large-scale movement towards Bitcoin would likely necessitate significant erosion in credit card usage, and thus, people may view the loss of their current payment systems as looming larger than potential efficiency gains from adopting Bitcoin.
It is fairly well-understood that these biases can be significant barriers to consumers adopting a new product or system such as Bitcoin (Gourville, 2006); however, they are especially prevalent when the subject is digital currencies due to the fact that people view digital currencies as inherently riskier than traditional payment systems (Popper, 2015, p. 216), and thus, risk-aversion plays a large role. Also, the exact benefits of Bitcoin are not well-understood by the public, and thus, when people consider the tradeoff between their current payment systems and Bitcoin, they have almost zero idea of the costs or benefits of Bitcoin. Without any direct, imaginable or vivid benefits from digital currencies, loss aversion plays a large role, and if people operate using reason-based choice (Shafir, Simonson & Tversky, 1993), then there is almost no compelling reason that could help overcome loss aversion to justify a switch to Bitcoin.
Conclusions and Extensions
Clearly, there are many psychology of decision-making biases that have created significant difficulties for people who want Bitcoin to become a commonly used currency. However, there may be issues due to the increased psychological distance of using Bitcoin as opposed to cash or even credit cards. Prelec and Loewenstein (1998) showed that “cash decoupling” occurs, which means that credit cards have different associative networks than cash. Many researchers, including Feinberg (1986) have shown that a credit card premium exists, that is, people have a higher propensity to spend when using credit card as opposed to cash. Moreover, Raghubir and Srivastava (2008) show that people treat payment methods differently based on their respective physical resemblances to cash [i.e. a gift card is closer to cash and induces “pain of payment” (Prelec & Loewenstein, 1998), whereas a credit card is less like cash].
Thus, if the world were to move completely to digital money with no physical basis, perhaps the “pain of payment” is gone completely. People would tend to spend more using Bitcoin for this reason, ceteris paribus, as they fail to feel the pain of the cost and instead sense only the benefit from the gain. Chatterjee and Rose (2011) suggest that more empirical research should be done on the online spending, and specifically the effects of more disconnection of the consumer from the payment mechanism (e.g. if your credit card information is stored so you don’t have to enter it). Raghubir and Srivastava (2002) demonstrated that when using foreign currencies during travel, people anchor in the nominal price and fail to adequately adjust depending on whether the currency is above or below the U.S. Dollar. They find that there is underspending when the local currency is a multiple of the U.S. Dollar (e.g. 4 Malaysian ringgits = 1 USD) and overspending when it is a fraction (e.g., .4 Bahraini dinar = 1 USD). Using this framework, people would drastically overspend using Bitcoin, as 0.0025 Bitcoin = 1 USD.
Perhaps a useful paradigm to examine something like how risk preferences change with digital currencies would be to look at how people behave when asked to take gambles using something that they don’t conceive as their own money. Perhaps an effect similar to Thaler’s “house money effect” (1990), where when gambling, people who had prior profits felt they were playing with “house money”, and as a result, their tolerance for taking gambles and higher risks increased significantly. Intuition says that when using Bitcoin, risk tolerance should be even higher than when using cash, as Bitcoin lacks a physical presence and due to the exchange rate being around $300 per 1 Bitcoin, it is unlikely that, say, betting 0.3 Bitcoins (equivalent to about $90) has the exact same mental conception as wagering $90 directly. Similar to how people may overspend with Bitcoin due to insufficient adjustment after anchoring in the nominal value, further empirical study could examine whether the anchoring and adjustment mechanism leads to excessive risk taking with Bitcoin.
Furthermore, evidence from prior psychology research on gambling may provide insights on how risk preferences are affected by the choice of units used to frame prices or bets. Loba et al. (2001) demonstrated that pathological gamblers’ willingness to gamble changes based on whether the bets are displayed in cash or credits; namely, gamblers are able to quit sooner when cash amounts are displayed, as opposed to credits. Lapuz and Griffiths (2010) examined Texas Hold’em players to show that people gambled significantly more when using chips as opposed to real cash. Although it’s dangerous to extrapolate these results directly to a people’s risk preferences and spending patterns in a Bitcoin economy, the results do suggest that there could be significant negative effects as a result of increased risk tolerance when value is decoupled from cash. Namely, combined with the idea that the “pain of payment” (Prelec & Loewenstein, 1998) is severely reduced even when using credit cards, it seems that a Bitcoin economy could be significantly different, psychologically, as opposed to a cash economy. This may have broad implications if risk tolerance increases significantly when using Bitcoin as opposed to cash or credit, which seems reasonable. To conduct this study, researchers could examine the few Bitcoin online poker markets that exist and compare whether player’s risk tolerance differs significantly when compared to using tokens or cash. However, Griffiths (1999) concludes that it’s unlikely internet gambling is “doubly addictive” when compared to regular gambling; instead, the internet may simply offer a convenient medium to exercise the vice. Thus, perhaps risk preferences and spending habits with Bitcoin wouldn’t be drastically different than with credit cards, although risk tolerance likely increases.
In terms of general future research, the intersection between the psychology of decision making and digital currencies is currently wide open. Many intriguing studies could be done, including replicating past studies such as those found in Tversky & Kahneman (1984) that describe the nonlinearity of decision weights. Thaler’s (1999) extensive work on mental accounting could be extended to examine whether people demonstrate similar conceptions of value when faced with choices that are framed in digital currencies like Bitcoin.
Blockchain Advisory Group (BAG) provides technical and principled strategic advisory services to high-quality crypto core teams and traditional private and public corporate management teams.
2014 CSBS Consumer Survey: Consumer Attitudes on Bitcoin and Other Virtual Currencies. (2014). Retrieved November 29, 2015, from https://www.csbs.org/regulatory/ep/Documents/Final Report-VirtualCurrencySurvey2014.pdf
Bearman, J., & Hanuka, T. (2015, May 23). The Untold Story of Silk Road. Retrieved November 29, 2015, from http://www.wired.com/2015/05/silk-road
Chatterjee, P., & Rose, R. (2012). Do Payment Mechanisms Change the Way Consumers Perceive Products? Journal of Consumer Research, 38, 1129–1139. Retrieved November 29, 2015, from http://jcr.oxfordjournals.org/content/jcr/38/6/1129.full.pdf
Feinberg, Richard A. (1986), “Credit Cards as Spending Facili- tating Stimuli: A Conditioning Interpretation,” Journal of Consumer Research, 13 (December), 348–56.
Gourville, J. (2006, June 1). Eager Sellers and Stony Buyers: Understanding the Psychology of New-Product Adoption. Retrieved November 29, 2015, from https://hbr.org/2006/06/eager-sellers-and-stony-buyers-understanding-the-psychology-of-new-product-adoption
Griffiths, M. D. (1999). Internet addiction: Internet fuels other addictions. Student British Medical Journal, 7, 428–429.
Kahneman, D., & Tversky, A. (1984). Choices, values, and frames. American Psychologist, 39(4), 341–350. Retrieved November 29, 2015, from PsycArticles.
Kahneman, Daniel, Jack L. Knetsch, and Richard H. Thaler. 1991. “Anomalies: The Endowment Effect, Loss Aversion, and Status Quo Bias.” Journal of Economic Perspectives, 5(1): 193–206.
Kahneman, D., Knetsch, J., & Thaler, R. (1990). Experimental Tests of the Endowment Effect and the Coase Theorem. Journal of Political Economy, 98(6), 1325–1348. Retrieved November 29, 2015, from JSTOR.
Kruger, J. (1999). Lake Wobegon be gone! The “below-average effect” and the egocentric nature of comparative ability judgments. Journal of Personality and Social Psychology, 77, 221–232.
Lapuz, J., & Griffiths, M. (2010). The role of chips in poker gambling: An empirical pilot study. Gambling Research, 22(1), 34–39.
Loba, P., Stewart, S.H., Klein R.M., & Blackburn, J.R. (2001). Manipulations of the features of standard video lottery terminal (VLT) games: Effects in pathological and non-pathological gamblers. Journal of Gambling Studies, 17, 297–320.
Popper, N. (2015). Digital gold: Bitcoin and the inside story of the misfits and millionaires trying to reinvent money. New York, New York: Harper.
Prelec, Drazen and George Loewenstein (1998), “The Red and the Black: Mental Accounting of Savings and Debt,” Marketing Science, 17 (Winter), 4–28.
Raghubir, P., & Srivastava, J. (2002). Effect of Face Value on Product Valuation in Foreign Currencies. Journal of Consumer Research, 335–347.
Raghubir, Priya and Joydeep Srivastava (2008), “Monopoly Money: The Effect of Payment Coupling and Form on Spend- ing Behavior,” Journal of Experimental Psychology: Applied, 14 (3), 213–25.
Shafir, E., Simonson, I., & Tversky, A. (1993). Reason-based choice. Cognition, 49, 11–36.
Svenson, O. (1981). Are we less risky and more skillful than our fellow drivers? Acta Psychologica, 47, 143–151.
Thaler, R. (1999). Mental accounting matters. Journal of Behavioral Decision Making, 12, 183–206.
Thaler, R., & Johnson, E. (1990). Gambling with the House Money and Trying to Break Even: The Effects of Prior Outcomes on Risky Choice. Management Science, 36(6), 643–660. Retrieved November 29, 2015, from http://faculty.chicagobooth.edu/Richard.Thaler/research/pdf/Gambling with the House Money and Trying to Break Even The Effects of Prior Outcomes in Risky Choice.pdf
Tversky, A., & Kahneman, D. (1974). Judgment under uncertainty: Heuristics and biases. Science, 185(4157), 1124–1131.
If you’d like to download the full PDF, please see this link.