A Better Framework for Token Design: Internalizing Externalities
By Alejandro Ballesteros, Venture Investor, Decasonic
For the past couple of months I have been thinking a lot about token economics, from working with our portfolio companies on designing tokens for sustainable platform growth, to attending talks, listening to podcasts, and studying prior case studies. When I initially started diving into the subject I felt that I was drinking from a fire hydrant, it became clear that many people have vastly different opinions on the role of tokens, and I could never find a gold standard framework for optimal token design or even understanding the role of tokens in an ecosystem. This essay is my attempt to succinctly explain the insights I’ve gained along the way.
The “Supply and Demand” Framework
I started by reading Nat Eliason’s Tokenomics 101-104 series. In Tokenomics 101, Nat outlines that Tokenomics all comes down to supply and demand. This is in some ways a truism, given that the price of any good, service, or asset in a market always boils down to supply and demand, but nevertheless I explored this lens.
Analyzing Token Supply
Nat argues that the main question you are trying to answer when studying a token’s supply is:
Based on supply alone, should I expect this token to hold or increase its value? Or will that value be inflated away?
In other words, is the token’s supply fixed, decreasing, or increasing? I agree with Nat that this is a useful question to ask, but many times people use the terms inflationary or deflationary to describe these supply side dynamics. I think these terms cause a lot more confusion than they are worth.
Inflation and deflation are technical terms that are typically used to describe money supply in an economy. In the Milton Friedman Monetarist lens, the basis of a currency being deflationary or inflationary revolves around whether the price of goods in that economy denoted in that economy’s native currency is increasing or decreasing. If in the US, we engage in monetary policy that results in an expansion of the money supply, we would expect the price of goods as denoted in US dollars to increase because each dollar’s value is diluted. On the other hand if we engage in monetary policy that contracts the money supply we would expect the price of goods as denoted in US dollars to decrease as every dollar is worth more of that good. It’s important to note that this rough logic assumes that the amount of goods and services produced remains constant and the money supply is the only variable that is changing.
This is not necessarily a useful parallel for a crypto economic ecosystem for many reasons. For one, purchasing power of the native currency relative to “goods and services” is not something we care about in all circumstances. When we think about Ethereum’s economics, the only native good or service that we reliably purchase is blockspace via gas fees. Gas fees are determined by supply and demand for blockspace on Ethereum. If we assume users denote gas fees in ETH, supply and demand for block space stays constant, and more ETH is issued such that outstanding supply increases, each ETH should theoretically be able to purchase less gas units than before and the cost of Gas as denoted in ETH (or GWEI) should increase. This doesn’t tell us whether the dollar value of ETH should change in any direction, however.
If we take the last example, but assume people are paying for gas and denoting gas fees in dollars (which I think is true, as many users get paid in USD and you can buy many more goods and services in USD than in ETH globally), the situation changes entirely. Except for the denomination of gas fees in USD, let’s make the same assumptions as before: demand and supply for Ethereum blockspace is constant and more ETH is issued such that outstanding supply increases. Since the cost to execute a transaction in dollars should stay the same (purchasing power of the dollar has not changed), the price of gas units in ETH increases, but the value of ETH in dollars decreases.
So using supply to understand more about the economics of this ecosystem comes down to what our assumptions are and what we are trying to measure. Do we care about:
1). Value of ETH in USD assuming constant supply and demand for ETH Blockspace
2). Cost of Gas in ETH, assuming constant supply and demand for ETH Blockspace
3). Cost of Gas in USD, assuming constant supply and demand for ETH Blockspace
I believe most people analyze supply to understand (1) but the use of the technical terms inflationary and deflationary actually convey that we are interested in understanding (2), or the impact on the price of goods in the native currency. Therefore, the terms inflationary and deflationary just add a lot more confusion than they are worth.
Furthermore, say we wanted to understand the difference between (1) and (3). We first have to answer the question of what Ethereum is: is it a security or a currency? If Ethereum is a currency and consumers are just using it to pay for gas, (1) and (3) are measuring very related things. In this situation, the value or dollar exchange rate of ETH (1) is always a function of supply and demand for Ethereum blockspace (cost of gas in USD (3)) and outstanding Ethereum supply (assuming constant US dollar money supply). If however, Ethereum is viewed by market participants as a security or speculative investment, (1) measures the perceived net present value of that investment in dollars, and (3) measures the cost of gas in USD, and they become very different things. Therefore, the assumptions we make about the perception of Ethereum by market participants also impact our possible understanding of the economics.
You can quickly see how using terms like “inflationary and deflationary” to understand the exchange rate (or dollar value) of the native token involves making tons of different assumptions about other variables like how market participants perceive the token, the supply and demand for blockspace or other native goods and services purchased with it, etc. For this reason I much prefer to ask: “is the supply of this token decreasing, constant, or increasing?”, with the understanding that this gives us limited information on whether the price of goods and services purchased with this token decreases or increases.
We don’t perfectly understand ahead of time how market participants will view or interact with a token. As a result, analyzing supply in a vacuum doesn’t tell you all that much. We are still forced to make large sweeping theoretical assumptions that may not hold up to reality when using supply dynamics to understand the future value of a token.
Analyzing Token Demand
Ok, so what if we add demand to the picture? Many people boil down demand to all sorts of complicated factors: game theory, memes, expectation of future price, access to content, etc. This can all be encapsulated by asking one simple question:
“Why should anyone buy your token?”
The correct answer to this question should be “differentiated utility”. In other words, users should buy the token because if they do, they will gain the ability to do something they really want to do and currently cannot via any other mechanism in a sustainable way.
But once again, this brings about all sorts of questions like “what do users want to do?” and “can they achieve what they want to do in other ways that are not buying my token?”. It is possible to analyze tokenomics through distilling down different elements of supply and demand, but we risk making faulty assumptions, not understanding human behavior, and missing critical variables. Analyzing token economics through “supply and demand” generally turns out to be a truism that actually does not provide any useful context on the role of a token in an ecosystem.
A Better Framework: Internalizing Externalities
When I heard Jump Crypto’s Nihar Shah speaking about Tokenomics, it became clear there was a much more intuitive and useful framework for understanding the role of tokens. Nihar’s view is that the tokens should provide common good, incentivize value creation, and implement non-restrictive staking. Building upon, his ideas, I believe there is an even clearer framework for designing durable tokens that encapsulates all of this: decreasing negative externalities and increasing positive externalities.
I argue that in trying to design a durable token, it becomes much clearer to use a framework that has existed for hundreds of years in economics: internalizing externalities. It may come as a shock but if we view fungible tokens as ways of internalizing externalities, Tax Credits (EITCs, AOTCs, etc.) and Carbon Offset Credits are examples of tokens that have existed for quite some time.
These credits have existed for a long time in order to incentivize citizens to produce more positive externalities and less negative externalities. An externality is when an individual or entity produces a socially-suboptimal amount of an activity or good as a result of not bearing all of the costs or benefits of that activity wholly. If the cost of doing something is concentrated and the benefits are diffuse or vice-versa this activity is likely prone to externalities.
If I smoke in a public setting, I am benefitting from all of the benefits of smoking but am not bearing the cost of any of the second hand smoke I am producing (only the first hand), if I am a factory owner and decide to dump harmful waste into a lake, I am not bearing the entire cost of this pollution’s harm on people or the environment but I am receiving all of the benefits.
There can also be positive externalities: If I produce and maintain open source software I am bearing all of the costs associated with this but am receiving a small portion of the benefits. Similarly, if I am an early adopter of costly electric vehicles I am not receiving all of the benefits to the environment and other citizens, but I am bearing all of the cost. For these reasons, the socially optimal amount of goods and services with diffuse benefits and concentrated costs or vice-versa are rarely produced.
https://en.wikipedia.org/wiki/Externality
In particular, the example of governments creating subsidies for early adopters of electric vehicle ownership is particularly salient. Because producing electric vehicles is subject to “economies of scale”, meaning that unit costs decrease as more are produced and users gain more utility the more total users adopt them (more charging and servicing infrastructure), there is a massive cold start problem with electric vehicle adoption, and less of them get adopted than is socially optimal. This is a big parallel to growing a decentralized protocol or platform. Oftentimes initial users receive little utility because the network is not very large, but expend large sums of energy and time costs to participate. The only way to get to the socially optimal level of production and increase the size of the network is to reward these early adopters, and therefore, internalize the externality they face.
In the example of a negative externality like those of emitting dangerous amounts of CO2 into the atmosphere, governments have tackled this through carbon offset credits. These credits force high-emitting companies to internalize this negative externality by having to pay low emitters for their credits to be able to emit more CO2. In the case of a protocol or platform, tokens can be used to disincentivize antisocial behavior like spamming, creating bots, or stealing. If users of networks were forced to stake tokens that at any point could be programmatically custodied by the platform for antisocial behavior, there would be a lot less of it.
We have gone wrong by believing that tokens are a protocol’s entire value proposition, that supply and demand (and therefore human behavior) are entirely predictable, and that outstanding supply of tokens is the most important thing to consider when investing. Tokens have a very specific use case in protocols: they are useful for coordinating large remote user bases by internalizing externalities. If a token does this correctly and its value does not increase steeply to the right, it is still a successful token. It might be the case that a token does accrue value over time, but that is for investors to decide.