By : Benjamin Edelman, Associate Professor Harvard Business School
Payment gurus are buzzing with excitement about Bitcoin. A decentralized peer-to-peer payment system with no one to impose arbitrary rules as well as a cutting-edge technical underpinning—it’s no wonder Bitcoin gets praise from sources as diverse as venture capitalist Marc Andreessen and former treasury secretary Larry Summers.
But will consumers actually want to use Bitcoin? I’m less sure.
At its start, Bitcoin attracted buyers drawn to its black-market sellers. For example, Silk Road sold weapons, drugs, and even murder for hire. Those services don’t accept payment through mainstream systems like credit cards, so their customers were willing to learn the ins and outs of Bitcoin. These distinctive sellers attracted initial consumers, but they were largely unlawful and unsustainable. (Indeed, the FBI shut down Silk Road in 2013.) Illegal sales won’t drive Bitcoin’s growth.
In its most ambitious form, Bitcoin could seek to join the ranks of standard consumer payment systems—letting consumers buy things from Main Street merchants. But Bitcoin is poorly positioned to compete in that market. Experience from other payment systems is instructive.
Usually, consumers pay the same bottom-line price no matter what payment mechanism they choose. Cash, credit, and (perhaps) Bitcoin are all the same price. Savvy consumers choose a payment mechanism based on benefits, seeking the best rebates or points. This market structure has predictable incentives: I choose a Visa Signature Preferred card with 2.2% cash back not because it’s the cheapest to merchants (it’s not) but because it’s the best for me. (It’s hard to find a card with a larger rebate.) Indeed, to a merchant, my Signature Preferred is surely the worst of Visa’s offerings because it carries the highest interchange fees (charged to credit card processors, and in turn to merchants) of any Visa card. But a consumer has no reason to consider or care about those costs to merchants.
How does Bitcoin fit in? Suppose I wanted to buy shoes at Overstock that cost $100. If I pay by credit card, the receipt says $100, but my card’s rebate means I actually only pay $97.80. If I wanted to pay with Bitcoin instead, I’d need to open a Bitcoin wallet and pay $101 to Coinbase to get $100 of Bitcoins (at the current exchange rate). (The extra dollar covers a 1% fee to Coinbase.) If I hurry straight to Overstock, my Bitcoins should still be worth $100. (They’re as likely to go up as down in the time I have to wait.) But notice: The transaction ends up costing me $101 by Bitcoin, versus $97.80 by credit card. I might try it once as an experiment. But I have every incentive to stick with my credit card going forward.
To spur consumer adoption of Bitcoin, merchants should offer discounts for consumers who use it. Suppose Overstock’s credit card processor charges 2.9%, a relatively standard fee. Is there a discount that makes Bitcoin preferable to credit cards both for me and for Overstock? It turns out that there is not. If Overstock reduces its price to me by 2.9% when I pay by Bitcoin, I still have to pay 1% to get the Bitcoins, which means I pay $98.10 to get the shoes. That’s still more than the $97.80 I would pay by using my credit card.
Even if I already have some Bitcoins, paying with Bitcoin still isn’t a no-brainer. Suppose I have $100 of Bitcoin. I could use that $100 to pay Overstock. Or, I could cash that out and get $99. Notice: With $99, I could pay my $97.80 credit card bill and still have $1.20 leftover. So even if I already have enough Bitcoin, I’d rather pay by credit card.
The crux of the problem is price coherence—a market structure wherein buyers pay the same price whether buying directly or via an intermediary. Consider: paying cash versus credit card, reaching a merchant by typing in its domain name versus clicking an ad on Google, or booking on an airline’s site versus an online travel agent like Expedia. In this market structure, intermediaries don’t compete to offer lower prices to merchants. Rather, they compete to offer buyers as much benefit as possible. To tempt me away from my 2.2% Visa card, MasterCard needs to offer a 2.3% alternative. To keep me searching at Google (rather than switching to Bing), Google pays engineers to build ever-cleverer search functions. The same is true of other affected markets. Of course merchants and advertisers then face ever-higher fees—the predictable effect of this market structure. (For details, see my working paper with Julian Wright: Price Coherence and Adverse Intermediation.)
Bitcoin doesn’t play this game. There’s a certain appeal to rejecting the escalating benefits and ever-higher fees to merchants. But consumers have little reason to forego these benefits—all the more so if Bitcoin then rejects the important consumer benefits (consolidated billing, 30 to 60 day float, and dispute resolution rights) that accompany mainstream payment models.
Perhaps Bitcoin shouldn’t be chasing Main Street merchants after all. Summers sees opportunity in the most inefficient parts of the financial system, so we might especially look to markets that are badly served by credit cards and other payment incumbents. I’m struck by the market for international payments—sending money to friends and family in other countries—where fees were historically quite high. Would Bitcoin work well there? Consider Bitcoin’s unpredictable exchange rates, technical complexity, and lack of accountability if something goes wrong. These features are a poor fit for customers of modest means who may not speak English natively and who are in no position to gamble on uncertain technology. Meanwhile, Western Union will send $1000 to dozens of countries for just $8. Running the same transaction through Bitcoin and Coinbase would cost $20 (1% to send, 1% to receive). Bitcoin fees need to drop threefold to be cheaper than Western Union, even putting aside service differences.
Recent surveys reveal consumer interest in Bitcoin not as a means of paying, but as a store of value that could gain future appreciation. But Bitcoins have no intrinsic value. These consumers might as well invest in tulips.
Bitcoin’s technical architecture—the distributed transaction register—is a clever innovation. It’s a fine addition to the Internet, and surely it will find some appropriate uses. Ex-Googler Mike Hearn says Bitcoin’s block chain algorithm could help run a more efficient taxi dispatch system or help an online review service recommend restaurants. I suspect those benefits could be provided just as well—and much more simply—through other designs. But if the future of Bitcoin shifts from payments to data storage and data processing, Bitcoin’s role will be correspondingly less revolutionary. As to the prospect of regular folks using Bitcoin payments day in and day out: Don’t hold your breath.