Decentralized Autonomous Organizations and the Atomization of the Firm
The optimal size of the firm is shrinking.
The optimal size of the firm is shrinking.
To understand why it’s happening and why it’s important, let’s start by talking about firms. Very simply, a firm organizes people and has them make something to sell it for profit. To discern why firms exist in the first place we need to clarify 2 things: the costs of organizing people to make something, and the costs of actually making it.
The Costs of Organization
In his 1937 paper, The Nature of the Firm, the famed economist Ronald Coase predicted that the optimal size of a firm in any given industry would decrease as transaction costs in that industry decreased. He defined transaction costs as the costs incurred in the process of making a transaction. Transaction costs might include:
Finding a party to exchange with
Negotiating a sales agreement
Enforcing a sales agreement
Coase experienced the second industrial revolution first-hand, a monumental clustering event that bled into the early 20th century. The rise of mass manufacturing and production line technology consolidated labor, requiring huge numbers of people to band together to actually make industry work.
According to Coase, the only reason for this clustering, rather than numbers of individuals simply freelancing with each other, was to keep transaction costs down. For example, if each part of a car was made by a different individual, the effects of those transaction costs would be multiplied: you’d need to find thousands of parties, drafts thousands of sales agreements and enforce all of them after finding someone else to put your car together. Doing everything under one roof kept transaction costs to a minimum.
Seeing this in action, Coase predicted falling transaction costs would work to shrink the firm. Simply speaking, if the cost of finding, negotiating and enforcing a sale was lessened across an industry, there’d be no reason to build a factory and organize people inside it in the first place. Thus, the costs of centralization would outweigh the costs of decentralization.
Though unforseen by him, the advent of the internet has only brought Coase’s Theorem to the extreme, as information and our ability to transact with one another is now literally millions-fold cheaper. Before we see this in action though, we need to address the other half of what justifies a firm to begin with: making things.
The Costs of Production
Traditionally, a firm’s marginal cost of production is modeled as a convex, smile-shaped curve, with the optimal size of a firm at the bottom of that curve. As scale increases, marginal costs decline, and economies of scale are realized. This is when a business benefits from its size— for example, it’s much cheaper and more efficient to serve 100 people in a restaurant than just one. The marginal cost of serving the 101st person is negligible.
After a certain point, though, the costs of consolidation outweigh the benefits of scale. For example, feasibly building and operating a restaurant that serves 1,000,000 people doesn’t make any commercial sense.
Here’s where it gets interesting: the internet brings our marginal cost of production to near zero. After you make a video, there’s no difference in cost to you whether it’s distributed to 1 person via email or millions of people on YouTube— you can build once and distribute infinitely. As the internet eliminates more and more of our marginal costs, the optimal size of the firm begins to shrink. Suddenly, we can serve many more restaurant-goers at a much lower cost.
The Atomization of the Firm
As the industrial age became the information age, we started replacing labor specialization with software specialization. Today, it’s become commonplace for startups to bundle and unbundle businesses— to shave off pieces from large companies and turn them into massive markets. What once looked like a small “vacation rentals” section on Craigslist is now AirBnB, a 100 billion dollar company.
This trend isn’t immediately apparent. Over the last 2 decades, the biggest companies in the world have grown substantially— WalMart, the US’ largest employer, doubled its workforce from 1.1 to 2.2 million people. The average size of companies with over 500 employees also increased. As a result, most Americans now work for a company with more than 500 people, which wasn’t the case in 2000.
Crucially though, the average number of employees at a firm is decreasing, having fallen during each year of the first decade of the 2000s and continues to do so. A firm with less than 500 employees accounts for 99.7% of all businesses in the US. In 2000, the average size of a firm with less than 500 employees was around 19. Today that number is 10.
Crucially though, the optimal size of the firm has shrunk dramatically; the labor market just hasn’t caught up yet. GDP adjusted, the impact of startup companies with small headcounts has skyrocketed over the last 2 decades. Comprehensively, this data tells us one of two things. Either the very largest companies are consolidating labor, or firms are suddenly able to accomplish more with less people.
I would argue the latter.
If we simply assume that software decreases the marginal costs of a firm and that firms will become more technologically dependent over time, we find that software will inevitably atomize the firm, eventually including even the largest.
Companies of One
Naval Ravikant extrapolates on this concept, imagining a decentralized end-scenario where the only work that hasn’t yet been automated is fundamentally human creative work and every individual is a freelancer:
“...What I think we're going to see is whether it's 10, 20, 50, 100 years from now, high quality work will be available [on-demand]. We're not talking about driving an Uber, we're talking about super high quality work available in a gig fashion, where you'll wake up in the morning, your phone will buzz and you'll have five different jobs from people who you’ve worked with in the past or have been referred to you...So the information revolution, by making it easier to communicate, connect and cooperate, is allowing us to go back to working for ourselves.”
Whether or not we agree with his hypothesis, the underlying change is undeniable: As the optimal size of the firm shrinks, everything trends toward decentralization.
The Spectrum of Governance
To better understand how organizations of the future might govern themselves, we need to understand where decentralization lies today.
Suppose we visualize this as a spectrum, with absolute dictatorship at one end and absolute decentralization at the other. In these bounds, we’ll let the legacy firm lie at 1/2, representative democracy lie at 3/4 and Bitcoin lie near 1. As we move from 0 to 1, self-governance becomes more and more important for the functioning of each structure.
An Analysis of the Legacy Firm
We’ll focus on the legacy firm, our chosen tool for organizing ourselves and making things. Take Apple, Nike, or even McDonald’s for example— the legacy firm is deeply ingrained in the regulatory fabric and culture of western society, and that embeddedness in our systems of thought doesn’t look to be changing any time soon.
That structure is also undoubtedly effective. People’s ability to organize themselves and be compensated for working has turned capitalism, despite its flaws, into one of the most positive engines of growth in history. As Peter Diamandis outlines in The Future is Faster Than You Think, over the last 100 years, income has increased 300%, lifespan has increased 250%, and food is 13 times cheaper, made possible in large part by the legacy firm.
For all those benefits, it’s important to keep in mind that what worked in the second industrial revolution at the height of 1870s may not be best suited for the height of the 2020s. 3 main problems arise:
Firstly, governance is concentrated at the top. All formal decisions are typically made between shareholders, a board, and executive management. Let’s picture an iceberg: all parties below the waterline— the majority—employees and customers, are simply instruments of labor and capital input that keep the firm afloat. With this in mind, we find that concentrated governance causes a misalignment of incentives: shareholders, the board and executive management are incentivized to maximize shareholder value, while those below the waterline are incentivized to build the best product or service. These are often very different objectives. We try and compensate for this with bonuses and commissions for employees or marketing campaigns for customers, but they come at a cost and the underlying misalignment remains.
Secondly, governance is difficult to access if you aren’t already at the top. If you aren’t, you have two options: climb the corporate ladder (spend time) or buy equity (spend money). The problem with climbing the corporate ladder is straightforward: you can only choose one ladder. The problem with buying equity is a bit more complex. For starters, only half of Americans actually own any public stock. In addition, over 85% of U.S. firms with 500 or more employees aren’t publicly traded. The minimum requirement for someone to become an accredited investor and access this 85+% is having a net worth of at least $1M, which only 8% of Americans can claim.
Lastly, we arrive at a classic problem of economics, the Moral Hazard. A Moral Hazard arises when one party becomes much more risk-loving because they won’t personally incur the associated costs if those risks are realized. If you’re insured, for instance, you might take on more risk knowing you won’t be ruined if things go wrong. We’ve seen this misalignment of risk and cost play out plenty of times: the 2008 subprime mortgage crisis, the BP Horizon Water spill, the Facebook Cambridge Analytica scandal, and many more. Louis Grx summarizes this well:
“[These] cases of mismanagement show that some companies have become too important in people’s lives and for society as a whole to be handled by current governance systems. [They] are the result of too much power in the hands of biased individuals, and corporate governance as well as hierarchical organizations failure to keep everyone aligned and accountable. A few attempts like Corporate Social Responsibility (CSR), i.e. companies implementing self-control over potential hams they may cause, has remained useless over the years.”
As it turns out, the internet has an exacerbating effect on this. As we’re able to work and transact with each other across the globe unimpeded by borders, the negative effects of concentrated governance are magnified.
Borderlessness
For all intents and purposes, the internet is stateless. If you’re determined enough, national borders become irrelevant, even in the face of most media censors through use of tools like VPNs and mix networks. For better or worse, this homogenizes culture, but also has incredibly powerful implications on how we govern ourselves.
In The Sovereign Individual, James Dale Davidson and William Rees-Mogg argue information is so unstoppable that the internet presents an existential threat to government itself. Let’s take online work for example:
Online work reduces the attractiveness of physical land, because it’s becoming less and less necessary for effective business. As a prime example, see the death of the American mall.
Online work has also increased regulatory arbitrage. States and countries are becoming more competitive with each other for tax revenue, being forced to undercut each other since individuals can work with organizations across the globe.
Online work has allowed individuals to earn in stronger currencies than that of the country they live in and work from.
In the short run, all of this puts pressure on currencies with undesirable monetary policies: No one wants to earn in the Venezuelan Bolívar. In the long run though, this puts pressure on any currency that doesn’t have a totally transparent monetary policy. Fundamentally, this is the argument for Bitcoin: a transparent, non-inflationary and universally verifiable asset might be the greatest store of value available. Of course, this assumes no regulatory intervention, which isn’t realistic, at least in the near future.
Though the internet has been mainstream for 30 years, we’re only now seeing the implications of Davidson and Rees-Mogg’s argument come to life. COVID has sped up remote work’s adoption into the economy, and countries are now beginning to struggle legislating, adopting or abandoning cryptocurrencies. The world is trending more rapidly toward decentralization, and at a global scale, countries are approaching a fork in the road: Embracing further decentralization erodes legacy governance, but abandoning it presents the real possibility of being left behind for good.
Enter the DAO
Assuming an inevitable trend toward decentralization, what solutions do we have at our disposal that we can leverage to continue aligning ourselves and working toward common goals?
Enter the Decentralized Autonomous Organization (DAO for short), a new, internet-native organizational and governance paradigm, enforced over a blockchain. Very simply, as Cooper Turley puts it, “DAOs are internet communities with a shared cap table and bank account.”
A DAO is decentralized because it gives independent governance shares to everyone involved. Any member of a DAO can vote on proposals or delegate their share of governance to another member at will.
A DAO is autonomous because its core functionality is written into code. For instance, if a DAO’s treasury receives money, its allocation across any verticals of that DAO are automatically executed by smart contracts.
What makes a DAO tick is trustless incentive alignment. Contributors of a DAO, unlike employees of a company, don’t need to trust each other or know their collaborators’ identities. Instead, individuals are aligned financially by the DAO’s token, which they’re rewarded in. Above all else, this ensures they have skin in the game. As the DAO grows, its token appreciates.
In practice, this often looks like a Discord server with access gated by a utility token and governance allocated to those token holders. Decisions are voted on collectively, recorded through 3rd party tools like Snapshot and executed by smart contracts. Yet even though DAOs have seen upwards of $15B in total value locked at the end of 2021, they look a lot more like direct democracy than a traditional corporation.
To break this down, we can examine the organizational benefits of a DAO. Foundation gives a great 4-point breakdown in their aptly titled piece, Everything you need to know about DAOs, which I’ll quote here:
“Transparency: A DAO provides users with a clear set of rules and regulations. Members in a DAO can see the code that governs the network, and all the transactions that take place on the blockchain.
Efficiency: DAOs are a great way to collaborate globally, as the technology makes them truly borderless. The rules of participation are also explicit, and once a decision is made based on the DAO's framework, contracts are automatically executed.
Autonomy: Because DAOs are self-governed by their communities, they can operate without an overarching authority figure, empowering everyone involved.
Anonymity: Members can invest in a DAO anonymously, which often means they can be more flexible and experimental with their investments.”
While it’s a great exercise to compare DAOs to the legacy firm, it’s important to keep in mind that DAOs aren’t totally analogous to businesses. Instead, we should simply be treating them as a new and frictionless framework under which we can efficiently collaborate with each other from anywhere in the world.
The endgame here harkens back to Naval Ravikant’s vision of a world in which we all freelance and in might more immediately resemble what Scott Belsky refers to as “polygamous careers” in 10 Forecasts For The Near Future Of Tech:
“The next generation of talent entering the workforce will overwhelmingly opt for what I’ve come to call “polygamous careers.” The desire to generate income and feel fulfilled from multiple projects will increase retention (you don’t leave a job if your “other interests” are being fulfilled elsewhere), increase workplace productivity (no more face time…people will be busier and more efficient), and help many projects and companies engage top talent that would otherwise be out of reach. One’s profession will be a portfolio of projects, whether you’re a designer, engineer, sales person, or investor. The idea of “exclusivity” in an offer letter will be laughable faster than we think.”
DAO Examples
Though a nascent organizational structure, DAOs have already seen many use-cases, especially over the last year. Broadly speaking, we can break these down into 4 categories, largely through the lens of curation:
Capital Curation: Individuals pool funds via a DAO and collectively vote on investments or grants to be made with that capital. They can do this for the sake of social capital, financial returns, or the collection of assets (usually NFTs). Here, a DAO curates capital.
People Curation: Individuals create token-gated clubs via a DAO for the sake of socializing or allocating talent to promising projects. Here, a DAO curates people.
Media Curation: Individuals cut out middlemen between artists, writers and content creators and their audience via a DAO. Here, a DAO curates media.
Protocols and Tools: Just as today’s tech giants came to power by lowering the technical barrier for individuals’ participation, Protocol DAOs have grown by creating tools to aid the issuance, management and exchange of tokens, placing power in the hands of community.
Some interesting highlights over the last year:
ConstitutionDAO raised $41M from 17,437 contributors to buy an original copy of the US Constitution at auction.
KrauseHouseDAO was formed to buy and operate an NBA team as a DAO.
PleasrDAO was founded to collectively buy and own “culturally significant” NFTs, leading to a series of groundbreaking bids and eyebrow-raising valuations.
CityDAO, an experiment in decentralized asset ownership, bought a parcel of land in Wyoming. This came to fruition following a bill the state passed last April recognizing DAOs as legal entities.
Future Use Cases
As such an early experiment in collective governance, it’s difficult to know what the future holds for DAOs. The simple answer might be a convergence of what constitutes a company and what constitutes a piece of software.
Much like software allows for rapid experimentation of systems, DAOs will allow for rapid experimentation of governance structures. And much like software or digital networks, DAOs might follow a pattern of interdependency through meta-governance. Stefen Deleveaux and Numa Oliveira liken this phenomenon to nature in Tapping Into DAO Ecology:
“One of the most observed features of planet Earth is its ability to generate resilience, and it does this largely by fostering a web of interdependence between organisms. In the same manner, a DAO with an ownership network diversified enough could discover and resolve unforeseen challenges, and could create an auspicious environment for innovation and coordination...
But what exactly will these interactions mean for the communities themselves? As a science, ecology is the study of the connections and relationships between various communities of organisms and their habitats. In the same way, we need to study and understand how the various DAO communities can and will interact with each other, and how this could result in a truly interconnected ecosystem. This method of analysis, or “DAO ecology”, would actually be quite useful for understanding the health of the ecosystem as a whole.”
As DAOs themselves serve as participants in governance for each other, we find a natural progression of interdependency one might see in software coupling, APIs, or indeed, in nature.
Potential Issues
As with any nascent technology, there comes a myriad of potential issues. Linda Xie highlights 4 great points in her piece on DAOs:
Firstly, DAOs are oftentimes still legal grey area in much of the world. This seems to be changing quickly, as we can observe in the previously mentioned Wyoming DAO Bill passed in April of 2021. By nature, any crypto-enabled technology is susceptible to regulation, but if trends continue, the future looks promising.
Secondly, there’s the question of effectively coordinating individuals toward a common goal. Financial alignment via tokens provides a practical solution, but in the face of too many moving parts, it can be difficult to make meaningful progress in one direction. A solution here likely lies in defining the mission of a DAO narrowly to avoid distraction on the periphery.
Thirdly, the question of individual apathy becomes important. Just as only 60% of US citizens actually find themselves voting every 4 years, not all members of a DAO will find themselves caring or having the time to vote on a given proposal. Many DAOs have worked to combat this by recreating representative democracy. In practice they allow individuals to lend their governance power at will to delegates, who make their position on various topics clear.
Lastly, we have the problem of signal versus noise internally. As a side-effect of being ultra-accessible, we find DAOs can run into the problem of low-quality or non-mission aligned discourse, especially if barriers to entry are low. On the contrary, since DAOs function through financial alignment, raising the barrier to entry can also mean making a DAO only accessible to more wealthy individuals. Current solutions include community-reviewed applications and crowdfunded grants to those who can’t afford the token.
In Conclusion
As we bring ourselves back to the present and take a few steps back, we get a sense of the bigger picture. COVID’s acceleration of remote work’s adoption has made us realize the future is likely decentralized. Software’s atomization of the firm, coupled with trustless incentive alignment enabled through cryptonetworks means it might soon become more fruitful to work as an individual than with a traditional, legacy firm.
DAOs, or their more matured descendants, present a great vehicle by which this might take place. If 2021 taught us anything, it’s that change can come about much more quickly than we think. Let’s not be surprised if it speeds up.
Thanks for reading,
Alex
Sources:
10 Forecasts For The Near Future Of Tech by Scott Belsky
A Beginner's Guide to DAOs by Linda Xie
A DAO Defined: The Big Picture by Aragon
A few limits of traditional organizations and their governance systems by Louis Grx
An Introduction to DAOs by Pet3rpan
Community DAOs by Patrick Rivera
Cooperation Economy by Packy McCormack
DAO Landscape by Cooper Turley
DAO Nation by Clay Robins
DAOs: a new organizational and governance paradigm by Louis Grx
Everything you Need to Know About DAOs by Foundation
How the Second Industrial Revolution Changed Americans' Lives by Eric Niiler
Platforms vs Verticals and the Next Great Unbundling by Jeff Jordan and D'Arcy Coolican
Tapping into DAO Ecology by Stefen Deleveaux and Numa Oliveira
The Amazing Shrinking Firm by Francis Kim
The billion-dollar, one-person startup by Roy Bahat
The corporation is dead, long live the corporation! by Roy Bahat
The Future is Faster Than You Think by Peter Diamandis and Steven Kotler
The Nature of the Firm by Abhishek Joshi
The Nature of the Firm by Ronald Coase
The Sovereign Individual by James Dale Davidson and William Rees-Mogg
Transaction Costs in the New Economy by Alan Patrick
World After Capital by Albert Wenger