Table of contents
- Social science data revolutions
- DAOs as constitutional laboratories
- Experimenting with Optimism’s airdrops
- Our first study: Optimism
- Broader implications for studying governance
In the second century BC, the Greek historian Polybius wrote, “What chiefly attracts and chiefly benefits students of history is just this — the study of causes and the consequent power of choosing what is best in each case. Now the chief cause of success or the reverse in all matters is the form of a state’s constitution; for springing from this, as from a fountain-head, all designs and plans of action not only originate, but reach their consummation.”
What Polybius was getting at is this: A central goal for those who study political economy is designing optimal political institutions. Interest in the question goes beyond political economy, of course: Understanding how to design durable governance structures, allocate decision-making power, and effectively coordinate various actors span business, law, policy, and beyond.
We’ve been working on the problem for 2200 years, give or take, and the answers are still subject to much study and debate. And while political scientists and political economists have developed many important theories about which constitutional changes lead to which outcomes, the issue is that these theories — however convincing — are hard to confirm. We can only test them when we find historical conditions that match both the changes our theories consider and can collect the necessary data. Despite these challenges, we’ve made remarkable progress, developing deep bodies of research around questions like the differences between presidential and parliamentary systems and the impact of term limits, term length, politician compensation, and many other design features on democratic accountability and representation.
We could learn more about the effects of constitutional designs on political outcomes if we could intervene and change institutional designs more frequently and more deliberately, observing the consequences in actual communities and then iterating based on them — that is, if we could experiment. For obvious reasons, no government or polity wants us to experimentally intervene, even with the best intentions.
But there is another option, recently developed and ripe for exactly this kind of experimentation: web3 governance.
Online platforms in web3 can allow us to design, test, and redesign political systems for online communities, organizations that will be an increasingly important part of the world in the future. web3 platforms are unique in their focus on writing and committing to explicit, written constitutions, and they provide open data on their political structures and outcomes. These are ideal circumstances for carefully designed experiments in governance with many participants (sometimes millions of voters), clear outcomes, and useful data.
This opportunity has parallels to other fields of social science. In the study of economic markets, for example, the growth of online platforms created an explosion in the ability for economists to test how consumers respond to different incentives in practice, and to design different auctions and other market mechanisms. Today, most online platforms employ large teams of economists to work on these issues, and the top economics journals are full of articles that use data from online platforms to understand economic behavior.
The study of human social networks experienced a similar transformation. The original “small world” experiments of the 1960s involved a few hundred people mailing letters. More recently, social media platforms like Facebook, Twitter, Instagram, and LinkedIn have become massive repositories of information on the complex web of human interactions. Researchers are now able to explore a wide range of sociological and behavioral phenomena, with real-time updates to high-dimensional datasets containing demographic and geographic information. They can intervene in these systems via experiments and learn how things work and why, doing experiments that would be impossible — or at least deeply impracticable — in the real world.
A core idea in web3 is that platforms should be decentralized, which means that no central actor is in charge of the platform. In an effort to remain decentralized, most web3 projects are governed by a body called a “decentralized autonomous organization,” or a DAO. People who hold the digital tokens issued by the DAO (or purchased in the secondary market) make and debate proposals and cast votes to make decisions collectively. These decisions can be high stakes, with a number of DAOs controlling large treasuries or being responsible for shared projects that people care deeply about — including issues like how to allocate shared resources, what changes to make to the project’s code base, what values to set key parameters to, or what business partnerships the project should enter into.
Early in its life, a web3 project typically specifies which decisions the DAO makes and how. This “constitution” is written into the project’s code, and the results of the votes automatically execute upon completion.
None of this implies that DAOs today are full-fledged democracies. Participation in DAO voting is often low, and voting power can be unevenly distributed. In some systems, for instance, wealthy token holders can sway the vote. Nevertheless, DAOs offer many instances of high-stakes voting, with meaningful electoral competition over important issues. For instance, the decentralized financial exchange Uniswsap held a closely contested vote to determine whether the protocol would start charging fees on certain transactions — an issue with major business ramifications for the protocol, and which was decided by a voting margin of roughly three percentage points.
The structure of DAOs provides three valuable characteristics for people studying the design of constitutions.
In cases where the DAO’s decisions matter to voters, these three features allow for a tight linkage between theory and practice usually unavailable in the study of constitutions.
DAOs struggle to engender robust political participation. But participation matters. Getting enough people to participate is important for a number of reasons. Broad, decentralized participation is core to the web3 ethos, and projects that fail to stimulate enough participation are vulnerable to “governance attacks” (akin to hostile takeovers in corporate governance), in which an adversarial actor purchases a sufficient number of tokens to take over the system.
Conveniently, a rich literature in political science has studied the determinants of political participation and experiments with different ways of giving people incentives to turn out and vote — ranging from social pressure and social rewards to legal requirements and even direct payments in the form of cash or lotto tickets.
Online governance provides a potentially broader design space for incentives that tap into other ways of motivating voters that have not yet been tried in the real world and therefore haven’t been studied empirically. For example, what if the rewards for voting are not purely monetary but instead give participants a stake in the future of the community? What if voters are rewarded over time for sustained participation and not just in a single shot?
Our recent paper offers one example of how DAOs can be useful laboratories for studying political institutions. Using public data from a web3 startup called Optimism, we analyzed a large-scale experiment that created a new incentive system for encouraging democratic participation. In the experiment, Optimism deployed more than 28 million dollars worth of digital tokens to users in order to reward democratic participation in the startup’s voting process. The clever design of Optimism’s retroactive airdrops allowed us to study these questions. We leveraged a natural experiment in Optimism’s airdrop 2 to empirically study the links between incentives and political participation at a scale and granularity not previously possible — an illustration of how web3 allows us to tightly connect theory and practice.
As we’ll explain, our paper offers insights into how to stimulate participation in online democracies. And, more generally, it shows how we can begin to experiment with governance: iterating between designing new political institutions, testing their impact, and adjusting their design by applying techniques learned from political economy and political science to the online world.
Optimism is a Layer 2 scaling solution built on Ethereum and powered by optimistic rollups. In addition to making Ethereum transactions more cost- and time-efficient, Optimism has pioneered community efforts to incentivize prosocial behavior in line with the broader Ethereum / web3 ethos. For example, Optimism allocates rewards in the form of OP tokens that convey irrevocable voting rights, and Optimism has promised an ongoing sequence of airdrop rewards by earmarking 19% of its budget for sequential airdrop rewards aimed at incentivizing prosocial behavior.
Here’s how airdrop 2 worked: On Feb 9, 2023, 11.7 million OP tokens (worth ~28 million USD) were airdropped directly to more than 300,000 unique Optimism addresses. The eligibility criteria were unknown to participants beforehand, announced the same day the rewards were distributed. This reward function consisted of
Optimism’s airdrop 2 aimed to incentivize prosocial behavior by promising future rewards and giving people a greater stake in the community through governance power, but users can sell their token rewards on the open market at any time. It’s possible the intervention could have no effect on the project. In fact, this trend of dumping and exiting immediately after receiving an airdrop has been documented in other projects. With this in mind, we explored two effects.
First, we find that there is a noticeable increase in new daily delegations immediately after the announcement of the airdrop reward on February 9, 2023. Specifically, this is an increase in new wallets that didn’t previously delegate but began to delegate post-airdrop.
Second, we find that on average, wallets that receive larger airdrop rewards subsequently delegate at higher rates — especially to other addresses (as opposed to self-delegation). To account for potential confounding — where, for instance, higher rewards simply go to addresses already more inclined to participate — we controlled for a wallet’s prior delegation behavior and examined the residual variation in rewards that comes from gas usage and bonus categories.
To check whether this was simply “delegation farming” in anticipation of future similar airdrops, we also examined the effects of the airdrop reward on subsequent voting. We do see an increase in voting as well as delegations, suggesting a positive relationship between reward size and governance participation more broadly.
From this analysis, we conclude that incentives for democratic participation can have a meaningful impact, at least temporarily, under the following conditions:
Practically, these lessons might be incorporated by tech platforms seeking to encourage higher levels of user input in decision-making processes. Additional design considerations that projects might adopt from this case of Optimism include: Specifically rewarding past governance behavior; designing a quasi-random, retroactive eligibility criteria which makes this reward more difficult to game; and cultivating a broader ecosystem emphasis and demonstrated commitment to promoting public goods.
These findings also offer support for existing theories suggesting that long-term economic stakes — like homeownership or universal basic income (UBI) payment programs — increase turnout in the physical world. It is also useful for political scientists to study how giving people a durable stake in their local community might increase participation, for example by encouraging participation in financial markets.
Our analysis of Optimism’s airdrop 2 showcases the potential for using web3 to study deep governance questions and test existing theories empirically. This isn’t just theoretical, and designing incentive systems for encouraging democratic participation is just one example. In other ongoing work, we are examining DAOs that switched from using direct democracy to indirect democracy. And there are many other opportunities to study reforms to political institutions in this environment. Some examples include:
Some of these lessons may generalize to voting systems in the physical world — including both democratic bodies as well as corporate governance — while others may not. But as more of human social, economic, cultural, and political life moves online, understanding how to design the governance of online platforms and communities will become increasingly important. Indeed, recent interest in the governance of AI platforms — including announced efforts by Meta, OpenAI, and Anthropic — suggest the growing importance of this field. There is much these new efforts can learn from the experiments already ongoing in web3. As we continue to work on additional projects on governance in the web3 space, we hope that other social scientists will join us.
***
Read our roadmap for how web3 projects can enable experiments with governance. And if you’d like to see a discussion of our earlier thoughts on the project, you can watch our presentation at the a16z crypto summer research seminar series last year.
Andy Hall is the Davies Family Professor of Political Economy at the Stanford Graduate School of Business, as well as a Professor of Political Science. Hall is a consultant to the a16z crypto research team and to the AR Family team at Meta Platforms, Inc.
Eliza Oak is a PhD candidate in political science at Yale University. She studies the politics of emerging tech and society with a focus on democratic online governance. She was a research intern at a16z crypto during the summer of 2023 and continues to work on governance research and experiments with Optimism, an a16z portfolio company.
***
The views expressed here are those of the individual AH Capital Management, L.L.C. (“a16z”) personnel quoted and are not the views of a16z or its affiliates. Certain information contained in here has been obtained from third-party sources, including from portfolio companies of funds managed by a16z. While taken from sources believed to be reliable, a16z has not independently verified such information and makes no representations about the enduring accuracy of the information or its appropriateness for a given situation.
This content is provided for informational purposes only, and should not be relied upon as legal, business, investment, or tax advice. You should consult your own advisers as to those matters. References to any securities, digital assets, tokens, and/or cryptocurrencies are for illustrative purposes only and do not constitute a recommendation to invest in any such instrument nor do such references constitute an offer to provide investment advisory services. Furthermore, this content is not directed at nor intended for use by any investors or prospective investors, and may not under any circumstances be relied upon when making a decision to invest in any fund managed by a16z. (An offering to invest in an a16z fund will be made only by the private placement memorandum, subscription agreement, and other relevant documentation of any such fund and should be read in their entirety.) Any investments or portfolio companies mentioned, referred to, or described are not representative of all investments in vehicles managed by a16z, and there can be no assurance that the investments will be profitable or that other investments made in the future will have similar characteristics or results. A list of investments made by funds managed by Andreessen Horowitz (excluding investments for which the issuer has not provided permission for a16z to disclose publicly as well as unannounced investments in publicly traded digital assets) is available at https://a16z.com/investments/.
Charts and graphs provided within are for informational purposes solely and should not be relied upon when making any investment decision. Past performance is not indicative of future results. The content speaks only as of the date indicated. Any projections, estimates, forecasts, targets, prospects, and/or opinions expressed in these materials are subject to change without notice and may differ or be contrary to opinions expressed by others. Please see https://a16z.com/disclosures for additional important information.