Artificially Limiting the Blocksize to Create a “Fee Market” = Another Variety of Lifting the 21 Million Bitcoin Cap
Note: This article has been kindly translated into Chinese. (中文链接)
Bitcoin backers have grown accustomed to thinking of Bitcoin as “inflation-proof” because of the iron-clad issuance schedule for its monetary tokens, culminating in the famously final total of 21,000,000 BTC sometime in the next century. Libertarians and other sound money advocates immediately recognize why Bitcoin is revolutionary, once they are able to grasp the nature of money as a ledger and how that makes Bitcoin truly as good as gold.
Yet the real epiphany hits when the former goldbug sees the script get flipped: Gold itself is in fact just a low-tech “sneakernet” ledger where unfakeable, scarce pieces of metal are traded as proof of ledger holdings instead of cryptographic signatures. Bitcoin simply upgrades this process with more reliable scarcity into faster, lighter, hideable, deniable, teleportable, unforgeable money. Libertarians hail it as a true sea change in personal liberty and the very fabric of society with the potential to disintegrate existing power structures at the seams.
“You’ve got to get a piece of this landgrab,” they say. “You’ve got to secure your portion of the ledger early on, before everyone jumps in.”
The key word here is scarcity.
But why? What benefit does scarcity in the form of a 21-million coin cap really confer? Of course, the answer goes, if a significant number of bitcoins could be issued contrary to the original schedule, over and above 21 million, its sound money property would be in jeopardy. It would be no better than the Federal Reserve with its willy-nilly inflationary policies.
Yet here one detail has been skipped over. Issued to whom? As everyone knows, the new Bitcoins are issued to the miners — thereby diluting the value of all bitcoin holders’ bitcoins slightly — as a reward for their services to the network. Said another way, bitcoin holders who have “secured their percentage of the ledger in this epic landgrab” forfeit a small part of that percentage to the miners every time a block is mined.
For instance, the Winkelvoss twins came out as owning 1% of the entire Bitcoin ledger as of early 2013. At the time, this figure represented about 100,000 BTC, because there were a little over 10,000,000 BTC issued. Now there are 16,000,000 BTC outstanding, meaning their stake in the current ledger has fallen to a little over 0.6%, a 40% loss in the proportion of the total ledger they own — a 40% shrinkage of their total land ownership of the slowly accreting Bitcoin island, as the traditional analogy goes.
That 40% loss wasn’t just “inflated away.” It went to the miners. Thus it is equally valid to view the situation as follows:
The Bitcoin island didn’t really grow; there isn’t really any inflation so to speak. There is just a miner tax. The island is the same size, but the Winkelvoss twins lost 40% of their property size to the miners. This is of course no problem, as they knew all this going in! However, I hope the reader can see that this is just as accurate a conception as the more familiar one. Why view it this way, though? Because it is more revealing as to what is actually going on.
Seen this clearer way, the block reward (now 12.5 BTC per block) is indeed a miners’ subsidy for their services. Since it was known from the beginning, this is all perfectly in keeping with the predictable issuance schedule of sound money. A certain known percentage of your stake in the Bitcoin island will go to the miners on a known schedule.
The other fact that was known from the beginning was that miners are allowed to charge fees to include transactions in a block. This is an additional yielding of landmass to the miners as is understood by all. It was always understood that this was necessary to incentivize miners to include transactions, as well as to scale mining subsidy as fee-paying transaction volume grew, since higher volume of high-priority transactions would mean Bitcoin had become a bigger deal economically and therefore needed even more protection from attack.
In the early days, critics were quick to point out that the miners could charge whatever fees they wanted to. Some socialists who don’t understand how economics work even contended the miners would charge exorbitant fees, like $500 per transaction, and people would be forced to pay. Was that not essentially an infinitely expandable additional miners’ subsidy?
“What meaning did the 21M cap have if miners could assess arbitrarily high fees on your ‘digital gold’ and you have to just stand there and take it?” they naively asked.
Fortunately most Bitcoiners are familiar with the counterintuitive aspects of free-market economics: unfettered competition among miners results in a bidding war. Even if some miner were to charge $500, other miners would undercut each other in a continual price war, the lowest bidder taking the entire pot of accumulated transaction fees. In this way, transaction fees reach the market clearing rate, where the fee rate on a given transaction approaches the marginal cost to a miner of including it in the next block. This isn’t a perfect system in terms allocating system expenses, but it certainly works to prevent price gouging. Again, Econ 101: any miner trying to charge even a little bit more for the exact same service will just be undercut by another miner and lose that potential income.
Therefore, in the final analysis, Bitcoin holders could rest assured that their savings — their percentage of the total Bitcoin island landmass — would be a predictable percentage at a predictable time, and furthermore that the tolls exacted when they move their money will be reasonable as they will be determined solely by free market competition. Perfectly predictable holdings and an un-gameable free market in fees that ensured fees were tied to actual miner costs. This was sound money incarnate.
Enter the blocksize cap.
The hard-coded cap on the maximum allowable size of blocks in the Bitcoin network stands at 1MB, and as everyone following the blocksize debate knows, this number had no discernable effect on fees per transaction until recently. You can be sure this fact isn’t itself a point of controversy precisely because one school has continually lamented it and the other school has continually lauded it. However, as is also a matter of record (and basic economics), rejoiced at by the former school and lamented by the latter, the fees per transaction have risen markedly since the transaction volume has started filling up the 1MB blocks.
No longer is the price per transaction determined by the free market, but instead by a hard-coded limit that intervenes in the market. This results in more money being paid to the miners for their services per transaction than otherwise would have.
Here I must make a gentle note to non-economists: this doesn’t mean more total money went to the miners, a common error — for the same reason that a 5-star steak restaurant doesn’t make more total money than McDonald’s. There are less total transactions than there would otherwise have been, as people start to economize, which is again an intended effect of the school of thought that celebrates this “fee market” (note that a free fee market already existed, hence this one should be distinguished by calling it something like an “artificial fee market” or a “developer-regulated fee market,” much like the many government-regulated markets out there in a standard mixed economy). This isn’t any kind boon for the miners, any more than it would be a boon to McDonald’s if the government came in and told them they have to triple all their prices.
But even if the reader is unfamiliar with economics to the point that there seems a way for this to somehow be more profitable for the miners, exorbitantly over-free-market-rate fees nevertheless violate the original sound-money aspects of Bitcoin equally as much as raising the 21 million coin cap. Many will object here: losing your purchasing power only when you spend it is different than losing it when you are just holding it. Yes, it is different, and that difference does create a change in user behavior; they will tend to consolidate transactions more, spend less often, etc.
However, these differences in incentivized behavior must never be allowed to obscure the fact that there is no principled distinction, as far as sound money is concerned, between (1) unsoundness by arbitrary additional payment to miners via extra block rewards (>21M) and (2) unsoundness by arbitrary additional per-tx payment to miners via fees that are many times higher than the free-market clearing rate. Either way, users are leached of a great deal of purchasing power they invested believing they would not be leached of. This undermines the entire point of holding BTC in the first place.
At this point this may seem like splitting hairs. Hey, fees aren’t that high yet. Even the block reward schedule moves around a tiny bit. Plus miners might be getting a bit of a free ride in the past, undercharging for transaction transmission. All of these points are true.
Nevertheless, this is exactly the same — in terms of destroying Bitcoin’s digital gold properties — as saying, “Relax, we’re only raising the 21M cap a little bit. What’s the big deal?”
So far the fees have risen quite a lot, and the stated goal of the school of small blocks is to have them rise much higher. They reason that something like the Lightning Network will work out to enable microtransactions — but not on Bitcoin.
This is not what investors signed up for. Instead store of value is undermined with the promise that it will be just the thing that will save store of value, as if investors need to be saved from themselves and their own bad decision-making in thinking Bitcoin fees could actually be left to the free market. Now if the Lightning Network can ever get those pesky decentralized routing problems sorted out, maybe this would be an argument to entertain — for those willing to compromise sound money.
Whenever you hear that a “fee market” needs to be incentivized by some artificial means, understand that this is equivalent to a call to raise the 21 million coin cap, in the capacity of being a means of defeating the security of Bitcoin as a store of value.
“By hook or by crook, we will eat out Bitcoin’s substance and render it a poor store of wealth,” the bankers say.
In Princes of the Yen, Richard Werner describes a conniving trick the Bank of Japan pulled to inflate the yen massively with no one noticing. While everyone was focused on the headline interest rate, the central bank gradually adjusted a little-known mechanism called “window guidance,” which was merely an informal “suggestion” pointing to certain quotas for loans banks should be issuing. This guidance was actually followed religiously, and ended up inflating a massive real estate bubble. It turned out to be equivalent to slashing interest rates in terms of sound money destruction. Here again the sticklers would have objected that it is not equivalent in its exact effects — this window guidance incentivized real estate speculation specifically — yet again this cannot in any way obscure the central effect that sound money was allowed to be decimated, by hook or by crook, indeed.
In Bitcoin as well, every mechanism must be watched carefully. The headline 21M coin cap is what everyone sees. Are the miners mining extra block rewards? Is anyone calling for lifting the cap? Nope. We’re safe.
No we’re not. One school of thought is asking the rest of Bitcoin to lift the 21M cap, in sound-money effect, by increasing per-transaction payments made to miners artificially, far beyond the original investment proposition that attracted so many billions of dollars into Bitcoin, that inspired legions of precious metals fans, that lit up the world with its defiant stance of unshakable monetary solidity.
I don’t wish to imply ulterior motives. While bankers may want to destroy Bitcoin, this school seems to have good motives, just a poor understanding of economic principles. After all, bankers stand to benefit from Bitcoin’s demise, but no one benefits from much higher transaction fees over the free market rate. Remember the triple-price Big Mac? No one benefits from that either. McDonald’s loses most of its customers, and the remaining customers must pay much more. It’s a lose-lose situation.
And lest some say that the Blockstream company, which I assume is still alive and kicking, stands to gain from this through its off-chain projects, no, this kind of policy is unequivocally destructive for all businesses in the space. One would think that the superiority of the free market could be left unheralded in a community built around free-market money, but it seems not. All I can say is that if this were Blockstream’s plan, there is no need to be concerned as such a plan can only fail. You don’t turn someone to stone before you try to squeeze blood from them. Having fees many multiples above the free-market clearing rate is indeed analogous to turning Bitcoin to stone.
Not because it would be unusable. Far from it. But because — and here apologies again go to the less economically literate if this explanation is too brief — Bitcoin is not competing only against fiat money and gold. Bitcoin is competing against the most powerful, fastest, most optimized, most cost-efficient version of itself that can be created.
That means Bitcoin of today competes against every altcoin, albeit with a massive advantage in network effect. However, massively overpriced fees due to developer-knows-best economic intervention are just the trick to open a great big gash in that network effect advantage.
More relevant than the meager competition from altcoins thus far is that Bitcoin as it currently stands “competes” (in a win-win fashion for holders) against every other possible superior version of Bitcoin, i.e., every hard fork that may be proposed. How difficult is it to compete against a version of Bitcoin where the fees are astronomically higher than the free-market clearing rate? Pretty easy, both because users and businesses will prefer the great savings and because investors will prefer a fee market that is free rather than frankensteined with sound-money-breaking artificial caps and limits handed down by development teams that can threaten to cease support of Bitcoin if the market interventions are not enforced. By hook or by crook…
It bears saying one more time: greatly overpriced transaction fees are just as store-of-value-wrecking as lifting the 21M coin cap. No one who supports artificial economic intervention into Bitcoin fees has any grounds on which to scoff at Keynesians who think Fed-controlled inflation is a good thing.
Clinging to semantics about “inflation” won’t save your purchasing power. Waving the banner of “scarcity” will not prevent those scarce coins from being sucked from your wallet every time you need to actually reap the real-world benefits of your hodling.
A final objection will be raised here.
What if we actually need a blocksize cap to save Bitcoin from centralization?
Indeed. The careful reader will notice I didn’t say Bitcoin doesn’t need any kind of limit on the size of its blocks. I merely said an arbitrary limit interferes with market clearing. Miners cannot be allowed to force everyone to accept giant blocks if the network can’t support them, though — right? There are many ways in which the blocksize could be capped voluntarily by miners, probably way above the level where it would induce a regulated fee market. Miners are after all the ones whose business depends on getting blocks propagated properly and are (or soon will be, as the ecosystem matures) far more qualified than developers to be setting this number.
Again by the same principles of competition and coordination, there “should” be no way to abuse this…but what do I know? I am not the one to be determining whether Bitcoin has a blocksize cap, what it is, or how it is enforced. Please get this centralist thinking out of your head. No person or group is qualified to dictate consensus settings, even should we dare allow one to, and I find it in bad taste for developers to cajole people into adopting such settings — as has been the practice heretofore — by bundling them up in a package with the rest of their code offerings in a “love it or leave it” gesture that forces users to choose their controversial consensus settings and their dev teams as a tightly bound set. Gratitude only goes so far. This is like the landlord who demands to be part of every important decision, such as who you date, because they are letting you stay for free.
This is why I like Bitcoin Unlimited. The dev team is all from the big block school, but they don’t try to force you into any controversial consensus settings just because you prefer their team or their implementation. Bitcoin Unlimited is the “no strings attached” implementation. In that way, they have opened up a market process among miners, which is a major step. Whereas — for better or worse — Bitcoin Core has so far been the “clingy ex-girlfriend” implementation.
Whichever team or philosophy you like best, I hope everyone in Bitcoin — being a free-marketeer — realizes that only the market can effectively aggregate and distill the knowledge of diverse persons across the community, anonymous and well-known, by putting their money where their mouth is. This process of opening the Bitcoin ledger to market decision-making has a name in the world of blockchains. It is called a “hard fork.” A controversial hard fork, to be precise, because without controversy there is no need to bring the matter to the market — in such mundane cases it is merely a protocol upgrade.
Besides some technical precautions that need to be taken in advance, the only reason not to present a controversial matter to the market for judgment is in the hope that the controversy isn’t so galling to your opponents that they won’t eventually roll over and accept a soft fork that accomplishes something similar, which shifts the hard-fork orchestration problem over to them. Indeed, if the change is mild enough and the market is immature enough for the process of market inquisition to still entail a lot of friction and uncertainty, this soft-fork poking can work for a while. However, don’t expect the market to react well to such a rude snub.
