About 1.5 years ago Digiconomist featured a blog on how a simple concept called “economies of scale” played an important role in explaining why Bitcoin’s Proof of Work algorithm incentivizes hashrate centralization. By restraining Bitcoin’s block size limit, Core developers are about to unleash the same concept on the actual use of Bitcoin rather than just its mining algorithm. Strangely, this has largely been overlooked or ignored by the Bitcoin community, even though the potential consequences are hard to overstate. Bitcoin developer Gavin Andresen appears to be the first to actually give this some thought, and wrote the following on Reddit (February 23 2016):
Smaller blocks can be a competitive advantage for larger companies, because they can:
- a) Contract with big miners or pools for guaranteed block space or “preferred” inclusion of some number of their transactions.
- b) Arrange once-a-day (or -hour/-week/-month) settlement with other large companies with whom their customers do a lot of transactions. E.g. Coinbase and BitPay might agree that any payment from a Coinbase customer to a BitPay merchant settles immediately off-chain, and is combined with all the other such payments into a single on-chain settling transaction.
That work (which is engineering and business development and lawyers) only makes sense if the savings and improved customer experience outweigh the costs.
Big, well-funded companies like Coinbase can afford it. Little companies may get squeezed out, and individuals will have an increasingly strong incentive to use a service like Coinbase’s centrally managed wallet– it will be more reliable and cost less.
If Brian was running a really big, old company then maybe he would be all for keeping the block size low to drive more business his way. But he’s running a start-up company in Silicon Valley, and he (and his investors) understand that growing the entire Bitcoin infrastructure and ecosystem is the only rational strategy. Spending any time even thinking about putting smaller competitors out of business is a waste of time when you’re a start-up in a brand-new market (or a start-up disrupting a huge potential market).
To examine Gavin’s remarks with the help of the concept of economies of scale, let’s first think back on its definition: “the cost advantages that enterprises obtain due to size, output, or scale of operation, with cost per unit of output generally decreasing with increasing scale as fixed costs are spread out over more units of output.” In short, efficiency gains from operating on a larger scale.
Economies of Scale Explained
In order for economies of scale to play an important role in a Bitcoin transaction fee market, it would thus require a possibility to save costs by operating on a larger scale. Interestingly, this possibility already exists within Bitcoin’s current fee market. It is mostly applied by intermediaries such as exchange, that mostly settle transactions between the users of their platforms “off chain”. E.g. if one Coinbase users “sends” Bitcoins to another Coinbases user, Coinbase can simply update both users’ account balances. No Bitcoins are actually moved, and therefore no fees are paid for this “transaction”. Coinbase happens to be a very honest company, as it doesn’t charge its users for this type of transactions either. In fact, Coinbase even keeps the transaction fees low for users sending money (to external users) that require an actual Bitcoin transaction. What Coinbase could do is charging a transaction fee to every single user regardless of whether they have a Coinbase account or not.
Suppose Alice has a Coinbase account and wants to send Bitcoins to Bob. Bob also has a Coinbase account, so Alice can send Bitcoins to Bob without paying any fees. Bob subsequently wants to send money to Carol, who doesn’t have a Coinbase account. Bob will pay a minor transaction fee (at a market rate) of $0.20 worth of Bitcoin. If Coinbase decides to change its policy, and always charges $0.10 worth of Bitcoin per transaction regardless of whether it is internal or external, both Alice and Bob will pay this amount. It might not be an advantage to Alice (who was paying nothing), but both Alice and Bob will be paying a transaction fee that is lower than the market rate of $0.20 worth of Bitcoin. Had Alice used an “on chain” transaction to send money to Bob, with Bob subsequently using an “on chain” transaction to send money to Carol, then they would have had to pay a total of $0.40 worth of Bitcoin in transaction fees at the market rate.
In the previous example, the average required costs per users would decrease if another one is added. Suppose Dave also has a Coinbase account and wants to send money to Alice, this could be processed as yet another internal “off chain” transaction. The only transaction that needs to be processed “on chain” remains the one between Bob and Carol requiring $0.20 worth of Bitcoin. Since there are now two internal transactions for which Coinbase can charge a fee as well, Alice, Bob and Dave would only have to pay less than 7 cents worth of Bitcoin each to cover all of Coinbase’s actual costs. In general, the more users Coinbase has (increasing scale), the less transactions it has to process “on chain”, lowering the average costs of internal/external transactions combined. We have thus established that there indeed exists a possibility to save costs by operating on a larger scale, hence economies of scale apply to a transaction fee market.
We can now get back to Gavin’s remark that “smaller blocks can be a competitive advantage for larger companies”. This statement should at least be amended to “smaller blocks can be more of a competitive advantage for larger companies”, as we’ve already established that larger companies already have an advantage due to the very existence of cost inefficiencies in a fee market. Even though we’ve already seen the signs of a fee market taking shape, the impact of Bitcoin’s current limit on overall fees has been too little for economies of scale to make much of a difference. A second key ingredient is therefore required for Gavin’s (amended) argument to hold. The benefit of operating on a larger scale simply needs to increase to a point where it is large enough to influence user choices.
Predicting the latter point is a futile exercise, as it can be anywhere from $0.50 to $5 or more worth of Bitcoin per transaction. To tackle this, it is only required to argue that we’ll inevitably reach a tipping point as a result of limited block size, which is a lot easier to do. It comes down to basic supply and demand curves, and Bitcoiners all too well know what a limited supply can do to prices. Smaller (limited) blocks mean that the free market fees for using block space will increase as block size becomes more scarce. As fees are rising, it increases the potential benefit of large companies that have large numbers of internal “off chain” transactions on which they can charge money as well (but process at zero costs). The mechanism explained in the example above can be used to lower the average costs per “on chain” transaction, and the biggest benefit will go to the company with the largest user base.
Economic theory thus supports Gavin’s (amended) statement that “smaller blocks can be [more of] a competitive advantage for larger companies”. The consequences of this could potentially be disastrous, as it implies that Bitcoin usage will centralize in the hands of a limited number of large companies (plus those users willing to spend a small fortune on transaction fees).
Those thinking that this is a little farfetched may have missed the fact that at least exchange BTCC is already gearing up for war with the release of its BlockPriority service, a service that exploits economies of scale, albeit with subtle differences from the example discussed because it is a free service. As more users join BTCC, the average transaction costs per user will still drop due to an increase in the number of possibilities for internal “off chain” transactions as the total number of users increases.
In the meanwhile, Blockstream is developing its own “off chain” solution for increasing fees called the Lightning Network. Even though this is a sidechain, it is very comparable to an exchange or wallet service processing transactions internally and having to settle only a limited number on the Bitcoin blockchain. Unfortunately, it could turn out to be a poor competitor against large companies. With the Lightning Network, users remain in control of their own funds, but it also reduces liquidity because funds will have to be locked in for a certain amount of time. More importantly, users already need an account at an exchange in order for them to get a hold of Bitcoin. The moment they open up an account at an exchange like BTCC, the platform can already offer a cheap way to transact Bitcoins. This is what gives large companies an advantages in terms of convenience.
Economies of scale is a powerful economic concept that has already proven itself in Bitcoin’s hashrate distribution. By creating scarcity in Bitcoin’s block space Core developers are now about to let the same concept get a hold on the usage of Bitcoin. Or, to put it differently, the usage of Bitcoin is at risk of becoming more controlled in the hands of intermediaries that the protocol once sought to eliminate.