This is the third and final article in a series about blockchain-enabled “smart contracts” and their ability to address retail fraud. In this article, we continue our case study on the online platform OpenBazaar and its smart-contract approach to mitigating fraud, known as multisignature escrow. We evaluate its potential efficacy in reducing fraud and discuss potential ways to improve the system. For further background, please read the first and second articles in the series.
Efficacy of Multisignature Escrow to Prevent Fraud
The three systems to prevent fraud that we discussed in part 2, legal enforcement, private enforcement and smart-contract-enabled multisignature escrow, seek to prevent fraud rather than just catching and reversing it. They do so by lowering expected payoffs from fraud. OpenBazaar and decentralized blockchain-enabled retail platforms in general will only succeed if they are effective enough in preventing the types of fraud mostly likely to occur. This article discusses the potential efficacy of the multisignature-escrow system to prevent several types of retail fraud. The commitment device provided by the system can allow both buyers and sellers to accurately signal their intention not to commit fraud, but only if the underlying moderation process is reliable and accurate.
Fraud by buyers against sellers
LexisNexis identifies the two most common types of fraud committed by buyers as “friendly fraud” (where the buyer denies receiving the item) and fraudulent requests for returns and refunds. Together, they account for 56 percent of fraud committed against large online merchants.
The commitment device enabled by smart contracts reduces buyers’ incentive to commit such fraud. Buyers commit to losing their funds should the seller and a third party moderator agree with each other. By agreeing to this smart contract in advance, the buyer signals they do not intend to commit fraud.
This system only works if there is a low probability that the buyer could also defraud the moderator. Suppose a buyer purchases an item, fraudulently claims the item never arrived or failed to meet promised standards, and does not release the funds from escrow to the seller. The moderator in this case may have trouble determining with complete confidence which party is truthful. Reputation mechanisms could help with this dilemma, as could sellers’ sending items via carriers that provide confirmation of delivery.
Fraud by Sellers Against Buyers
Protecting buyers against unscrupulous sellers is a problem that the multisignature-escrow system must address. In traditional e-commerce, buyers rely either on the reputation of a large retailer such as Amazon or a large centralized platform such as eBay or Etsy. Reputation scoring for sellers is likewise important on OpenBazaar, and will be discussed further below.
Under multisignature escrow, the seller commits to not receiving payment until either the buyer or a moderator acknowledges that contract terms have been met. Once again, the value of the commitment is only as good as the underlying moderation system. Shipping via carriers who document delivery might provide a predictable way to make the system work at least most of the time. Issues of product quality, however, would provide room for subjectivity on the part of the moderator that could make the process less reliable and accurate.
Decentralized platforms such as OpenBazaar, especially when using a cryptocurrency as payment, deviate radically from traditional online retail when it comes to identity theft. Traditional online retail places the cost burden of identity theft (e.g., stealing a credit card and using it to buy a product online) on sellers (via chargebacks) and intermediaries such as credit card companies. Victims get refunds for the purchases they did not actually make.
If an identity thief could access the information necessary to use an individual’s Bitcoin wallet and made a purchase on OpenBazaar, the victim would have almost no recourse while the seller would not be harmed at all. The seller and fraudulent buyer would sign off on the transactions, and the funds would be irreversibly released to the seller. The seller would likely never know identity theft had occurred. This risk of identity theft is not borne specifically by OpenBazaar buyers but by anyone owning Bitcoin.
According to LexisNexis, identity theft accounts for 23 percent of fraud committed against large e-commerce merchants. It is generally thought to be harder to hack into a Bitcoin wallet than fraudulently obtain credit card information. But those who commit e-commerce fraud are adept at developing new techniques when large enough rewards are at stake. In OpenBazaar’s system using Bitcoin, sellers would be completely shielded from the impact of identity theft. But if such transactions became risky enough for consumers, it would threaten the viability of Bitcoin-using platforms as a whole.
Scope for Improvement
While OpenBazaar’s multisignature-escrow approach to fraud is promising for many reasons, we have also identified several challenges, including scalability, credibility of the moderation system, and buyer exposure to identity theft. This section discusses additions or changes to the current system that might address some of these issues.
Selection of Moderators
A multisignature-escrow system only works if the moderation process is reliable and accurate. In part 2 of this series, we noted several challenges to OpenBazaar’s system, such as scalability and misalignment of third party incentives. Some centralization on the part of moderators, while preserving the market-based approach, might make the system more scalable and predictable.
Firms offering moderation services could compete alongside individual moderators. Having firms that specialize in the service could routinize some procedures, making the system more scalable, and create brands with the incentive to build reputation. These firms already exist in the wider Bitcoin ecosystem, though it may only be worthwhile to compete on OpenBazaar if that platform grows.
On a platform such as OpenBazaar that allows a high degree of anonymity, reputation scoring is important to instill mutual confidence between buyers and sellers. OpenBazaar’s system for seller reputation asks buyers at the close of a transaction to rate sellers on a scale of 1 to 5 on four dimensions, including quality and delivery, and averages them into an overall feedback score. The system gives sellers some incentive to build a brand without interacting with the same buyer multiple times. It is more difficult to judge newer sellers with a shorter history of transactions, though buyers can presumably observe the number of transactions and take that uncertainty into account.
Ex ante Prevention of Fraud
Due to the irreversibility of transactions, Bitcoin shifts the burden of identity theft from sellers to buyers. Could a data-driven ex ante prevention system help protect Bitcoin holders from unauthorized purchases made on OpenBazaar? Implementing such a system on a fully decentralized platform would be very difficult. First, due to the relative anonymity of Bitcoin purchases, there is not a store of data that could match the credit card purchases used by PayPal. Second, there is no central entity with the incentive to implement such a system. A third party firm might come along and offer those services for a fee, but this would likely require changes to OpenBazaar’s code. Finally, since the risk is borne by Bitcoin users in general, even OpenBazaar buyers and sellers would lack a large incentive to provide such a system. Security of wallets will be an essential component in the success of Bitcoin as a whole.
A Promising System
Smart contracts coupled with blockchain technology have the important feature of not relying on any central authority, whether public or private, to enforce the execution of terms. They avoid such reliance by allowing the parties when entering the agreement to relinquish control of the terms of the contract being executed. Blockchain technology features both irreversibility of transactions and a high degree of anonymity, removing any credible threat of going to court, and thus making the pre-commitment fully binding.
Smart contracts look good in theory but must succeed in preventing real-world fraud. The multisignature-escrow approach employed by OpenBazaar will provide an early test of the efficacy of smart contracts. The platform is small and anonymous, meaning data on fraud prevention are not available, but anecdotal evidence will be informative as it grows. We have evaluated this method of preventing fraud by comparing it to other, more commonly used methods and looking at the incentives of potential fraudsters. Signaling one’s intentions to deal honestly by pre-committing to the smart contract will work if and only if the third party moderation system that solves disputes is reliable and accurate. We have discussed both merits of and concerns with OpenBazaar’s highly decentralized approach to moderator selection, and it will be interesting to see how it performs in practice.
Many blockchain and cryptocurrency enthusiasts tout decentralization as an unambiguous benefit of these new technologies. But decentralization typically involves trade-offs. The benefits of OpenBazaar, including lack of fees, lack of censorship and lack of reliance on a central authority, will only justify moving to the platform if the multisignature-escrow system adequately prevents market participants from being the victims of fraud. This outcome will greatly inform observers as to whether retail is an efficient use of blockchain technology.