Yesterday I gave a guest lecture in WIT’s management department on the economics of blockchains (here are the slides; here is the code used for the original figures and cryptography examples). Putting the lecture together was great fun, and the students enjoyed it, despite the 8am start time and lousy weather. One student even posted a picture to Reddit. True to form, many comments are equal parts hilarious and off-topic. Yet several commenters described an unmet demand for lectures on stuff like blockchains at their universities. While blockchain discussions are more common in computer science and engineering circles, economics has a lot to contribute as well, particularly to the question: when should you use a blockchain?

I framed the lecture around the problem of incomplete contracts. These are agreements between people (“I will buy your used car”; “I will hire you for this job”; “Let’s agree to make a trade in the future”) that rely on information not readily available. One side of the agreement may have private information they want to hide (turns out there used to be raccoons living in the trunk). Or maybe somebody cannot credibly reveal what they know (how does the good student reveal their quality to the employer and get the right wage?). Or maybe there is information nobody can access (e.g. who knows what tomorrow brings?). Contigencies abound.

The lack of perfect information in many economic settings means most contracts are incomplete in some way. Amazing, really, that we trade at all, given all the things that go wrong. Trust can be very expensive.

This is why I began the lecture tens of thousands of years ago in the Pleistocene/Holocene transition. There is a literature that asks why humanity gave up hunter-gathering and took up farming, when our first attempts at domesticating grains made it harder to “make a living” (produce enough calories to survive). Bowles and Choi (2013) tell a fascinating story in which farming takes over not because it is a superior technology, but because it co-evolves with the institution of private property. Hunter-gatherer bands shared nearly everything. Sounds nice. But when nearly everything is communal property, trade, especially with anyone outside your band, is nigh impossible. Contracts boil down to agreements to transfer private property from one person to another, and trading private property, from millet and grain to securities and houses, is what makes us rich (alongside many other things, of course). Early man had no idea his future self might be scrolling through Reddit on the bus to work. Yet in this light, farming likely took over because people realized the benefits of increasing the scope of economic activity. The rest is (pre) history.

Let them eat crypto

The key idea behind blockchains is that they too can increase the amount and/or type of contracts we enter, largely by reducing the cost of verification. Most markets today rely on some intermediary to approve the deals you and I enter (e.g. Amazon, Ebay, your local bank, the central bank). Transactions are tracked in a ledger, and the role of the intermediary is that of the bookkeeper, making sure there is no funny business. Otherwise, I might be tempted to go in and cook the books (“No, you didn’t pay me, see?”).

You often hear that a blockchain is a digital ledger. It is much more. By bundling transactions into “blocks” that are “chained” using cryptography, the ledger can be made public, and transactions can be verified by anyone in the market, so long as you provide the right incentive (hence “mining”, the reward for contributing your computing power to maintain the ledger).¹

Why get rid of the intermediary? The intermediary in any market is the monopolist on verification, and monopolists often have incentives out of line with everyone else. Specifically, the intermediary has the power to censor transactions and charge fees above and beyond the value it adds to the market. But perhaps most important, once more, is the role of information. To verify transactions, the intermediary needs a lot of background about everyone in the market – more, in fact, than if contracts were complete. Bad stuff can happen. So much private information (credit cards, social security numbers) stored in one place invites theft. What else can the intermediary do with this information? Maybe sell it, or allow third-parties (e.g. app developers) to access it.

Back, then, to the original question. Blockchains are good at tracking stuff and storing information in an immutable ledger that is very robust to fraud. So, when do you use a blockchain? In markets where the scope of economic activity is dramatically reduced due to the incompleteness of contracts. Specifically, when a market has large verification costs, and intermediaries charge large fees for their service.

Not so fast

Tradeoffs cannot be ignored. In the lecture, I focused on the Bitcoin blockchain, which for all its innovation, is terribly slow (many orders of magnitude slower at processing transactions than, say, VISA) and relies on wasteful computation to deter fraud, consuming massive amounts of electricity in the process. Ironically, economies of scale in mining blocks turned an idyllic decentralized market into a market of a few major players with a competitive fringe. The Bitcoin blockchain is essentially a mechanism for reaching decentralized consensus (people “agree” on the ledger by accepting a mined block and working on the next one). Yet developers struggle to agree how to scale up. Then there are questions of regulation.

Not being an expert, I can’t say what will happen to Bitcoin.² My take away is that Bitcoin is a great proof-of-concept. At the same time, Bitcoin also shows that there is no one blockchain to rule them all. How you design and implement a blockchain depends on things like transaction size, privacy, security and so on. The Bitcoin blockchain is “permissionless” (anyone can view the ledger and mine blocks), but some recent blockchains, like the one used by the Australian Securities Exchange to track financial trade, is “permissioned”, meaning a central authority has the exclusive right to access and edit the ledger. The advantage of a permissioned blockchain is that it reduces information requirements (the ledger does not need to be a public broadcast). The disadvantage is that the ledger no longer benefits from the reduced cost of networking (i.e. having a bunch of miners toiling away at verification).

The world is your blockchain

Given these tradeoffs, there are great conversations to be had about the uses of blockchains. Digital goods can stored in blockchains, but so can attributes of a digital good, like property rights to a song, allowing more seamless trade and enforcement of intellectual property. You can also think about markets that could be created out of thin air using blockchains, since they have the potential to handle a wide variety of so-called “smart contracts” that bundle the terms of trade and their execution and enforcement all in one.³ Many goods like computing power (an attribute of a computer) are more efficiently allocated through trade but difficult to contract on. (Imagine trying to orchestrate a trade where someone rents your CPU for a night by borrowing the entire computer!) Etherium was created to allow individuals to trade computing power in exchange for a cryptocrurrency (Ether).

Far out ideas abound. For instance, many students were late to the lecture because they had to drive through notorious Boston traffic. Imagine a world in which self-driving cars could write smart contracts with each other. If you were late getting out the door, you could see if somebody else was willing to sell their spot on the road (e.g. agree to leave later than you). If you strike a deal, the smart contract immediately transfers payment to other driver, whose car is then “locked up” until the agreed upon time.

Crazier ideas dust down from space when you think about how blockchains can mitigate fraud and reach consensus. What if blockchains stored identities, allowing you to sell your privacy at your discretion? What if special counsels used blockchains to audit information and manage investigations? What if voting was held on blockchains? Tradeoffs abound, though many center around the question: how will society trade privacy for transparency? Questions can also be asked about the future of monetary policy. Should the central bank adopt a digital currency? Or should the central bank be stripped of its monopoly power in domestic monies?

Students are right: blockchain discussions should happen in more classes, especially economics. For the interested, there are many great economics papers that study blockchains. The ones I used for the lecture are listed on the last slide. A great start is “Some simple economics of the blockhain”.