Part 2 of this series asked whether Bitcoin could be adequately funded merely through the spontaneous charity of network users. Now we look at the potential of inflation and demurrage to sustain the network.
Introduction
The funding proposals discussed previously, voluntary fees, assurance contracts, and altruism, were all essentially different ways of arguing for non-action: just leave the protocol alone and someone (fee-payers, major self-interested parties, altruists) will take care of the rest. If we have established that non-action is not a workable solution to the funding problem, then inflation is perhaps the “lightest” solution next in line. After all, right now Bitcoin’s security is already being solely funded through inflation. Inflation as of 2015 is approximately 10% p/a, but slated to taper off over the coming decades, dwindling to a negligible trickle before, some time around 2140, disappearing altogether.
What if the rules were changed, and the inflationary era prolonged indefinitely? The funding problem could be solved in arguably the most economical way possible, simply by making the coinbase transaction continue at a certain set level forever. This would also avoid the difficulties of constructing a system to collect small payments from every single transactor on the network. Instead, all users would be taxed proportional to their investment in Bitcoin as a currency. The more bitcoin you own, the more you contribute. Sounds fair enough, right?
Politically impossible in the mainchain
The simplest argument against this proposal is its total political infeasibility. The guarantee that inflation will end and that ultimately one bitcoin will always be worth 1/21 millionth of all bitcoins in existence, forever, is one of the major attractions of Bitcoin, one of the principal reasons for its success thus far. Many Bitcoin enthusiasts see this aspect of the currency as offering a sanctuary from a world ridden with nation states reliant upon perpetual, oft wanton debasement to, among other sins of government, hide budgetary incompetence, fulfill otherwise-impossible electoral promise, and defend the oppression and exploitation of populaces.
Anyone proposing such a drastic change would struggle (to put it lightly) to garner support from the core dev team, and so would inevitably be forced to release their change as a non-official fork of the source code. Miners, fearing correctly that the fork, if implemented, would cause massive damage to confidence in Bitcoin, would almost certainly shun it. So it is for all intents and purposes impossible that such changes will ever come to constitute the mainchain of Bitcoin, and that Bitcoin the currency itself will ever be directly inflated above 21 million units.
Because of its emotional dimensions, grounded in a long history of currency abuse by nation states, this objection is essentially irrespective of whether perpetual inflation is in itself a sound funding proposal or not —it is simply a political non-starter in the mainchain. But if we ignore the issue of political practicality, and assumed that it was somehow possible to implement, would it even be a good idea in the first instance?
Inflation on its own merits
At its simplest, perpetual inflation would essentially shift the burden of network sponsorship from those who are actively transacting in the present to those who are passively holding bitcoin. So the longer someone leaves a UTXO dormant, the more they will pay from its value to fund the security of new transactions. Is this fair? One might argue that all holders of bitcoin require a secure network for the currency to retain its value, and therefore should pay for its maintenance. But this argument is contrary to how network security actually works.
Fundamentally, network security is most important when transactions are newest. As transactions age, they accumulate more confirmations, and with confirmations the cost of reversing transactions scales exponentially. Therefore even in a weak and poorly-secured network a very old transaction can still be trustworthy, because the accumulated cost of reversing it has become extremely high. But the the more recently you have received funds on the Bitcoin network, and the faster you want to be able to trust that those funds are not double-spent, the more you care about security.
Inflationary funding however would tax all holders of bitcoin, irrespective of the age of their transactions or the number of transactions they are currently processing. In so doing it would disconnect the principal consumers of security from the principal funders — a recipe for an unsatisfying, unpopular payment system, and the opposite of what we should hope to achieve. While it is perhaps true that all holders of UTXOs benefit in perpetuity from a secure, stable, usable network, if Bitcoin succeeds it will nevertheless be used by everyone, everyday. It is thus most sensible and fair to design a funding system that assesses contributions on active use of the network. For all these reasons inflation would likely be a poor solution even if it was politically feasible.
A sidenote on demurrage
Demurrage is an alternative to inflation where the total number of currency units stays static, but the value of UTXOs are gradually diminished to fund the block reward. Demurrage is really just inflation wearing a different hat: whether a currency inflates or is demurred, as long as the rates are equal, the overall loss of value to each holder is the same. The only advantage of demurrage might be psychological, in that the total number of bitcoin would stay the same, and thus the costs of recalculating prices inherent in an inflationary regime would be avoided. In contrast however, its first major disadvantage is greatly increased technical complexity. Compared to the simplicity of merely prolonging the coinbase, a system would have to be devised to recalculate the contents of UTXOs when they are consumed as inputs, or something similarly complex, requiring in turn significant code changes in many pieces of the Bitcoin ecosystem. Furthermore demurrage is perhaps even less politically feasible than inflation, because demurrage implies a direct intrusion into the holdings of currency owners.
Conclusion
As noted, many users of Bitcoin are attracted by its strong contrast to the current roster of unpredictable, elite-controlled inflationary currencies. A change to the slated inflation schema would thus be seen as opening the door to further dilutions of the Bitcoin vision, and would likely critically damage confidence in the currency. Therefore, without inventing a pointless altcoin, an inflationary funding model would only be possible as an experiment on a sidechain, once that technology becomes available. The sidechain could be designed with an initial conversion bitcoin rate that reduced as more of the sidechain token was produced. That said, even an inflationary sidechain would probably fail. Such a sidechain would be chronically prone to exploitation, with users moving into the sidechain to process high numbers of transactions and moving out, consuming the good of network security without contributing by maintaining their funds in the chain, a classic example of a failure of a common good.
