Future Of Work
June 20, 2019
June 20, 2019
The cost of individuals coordinating is too high. This is the conclusion found by Ronald Coase when he set out decades ago to "discover why a firm emerges at all." What he found was that:
"a firm will tend to expand until the costs of organizing an extra transaction within the firm become equal to the costs of carrying out the same transaction by means of an exchange on the open market or the costs of organizing in another firm." (1)
The firm exists because it is cheaper (and easier) to consolidate necessary functions within a company than it is to find external providers of those functions on the market.
In a world where communication is slow and limited - as it was in 1937 when Coase conducted his analysis - it makes sense why external coordination was expensive. It makes sense why firms kept core competencies internal to avoid those transactions costs. Work required a unity of time and space.
The context that lead to this observation has since changed.
As companies grow, the cost of the marginal transaction must be less than completing that transaction in the open market. For any given transaction, does a company want that competency to be in house or outsourced to external providers? How hard is it to do that transaction internally vs. externally? The more difficult it is to transact externally, the more likely it will be brought in house.
Communication is no longer slow and limited, it is fast and unlimited. The internet and digital platforms reduce the cost of coordination. It is feasible to operate beyond the firm, to orchestrate many transactions by individual contributors, and it has become profitable to do so. (2) While there is still friction organizing people in different places and in different times, it is now possible; it is no longer necessary to be in the same place at the same time. Remote work is a testament to this.
The rise of the internet and open communication has inverted the nature of the firm. It makes increasing sense to outsource everything to the open market except for core competency.
It is not unreasonable to envision a world where hierarchical bureaucratic corporations give way to a networked web of external providers with focal points of core competencies. Firm responsibilities will be itemized. The optimal size of the firm is constantly shrinking.
Information technology makes it easier to transact externally; it diminishes the friction.
Cost of coordination has fallen because of the increased efficacy of:
Payment Infrastructure - it is easier to pay people and settle transactions
Reputation Systems - there now exists increased trust between unknown parties and rich information signaling
Search - it is easier to find services & talent and connect with them on demand
Communication Channels - it is easier to work remotely and manage distributed workforces
While the firm emerged as the right tool to service the Industrial Revolution, the Digital Revolution calls for a different tool - the nature of work is changing once again.
Not only is the cost of external coordination falling. Simultaneously, what used to require millions of dollars in startup capital, a multitude of employees, and years to reach market scale can now be accomplished with thousands of dollars, a handful of employees, and a shortened time horizon.
Technological leverage drives this change and amplifies the capabilities of individuals. People can now build and do things they couldn’t before by building on top of existing infrastructure and using technological tools for leverage.
As the nature of the firm changes, so too does the nature of the job.
Recall that as recently as 1900, the US was predominantly an agrarian society with the vast majority of the population working in agriculture and ~50% of the workforce being self-employed. (3) The Industrial Revolution codified the idea of having masses of individuals working on problems with strict schedules, times to show up, and rigid hierarchies. Again, jobs required a unity of time and space. Today, less than 2% of US workers are focused on agriculture. While many of these jobs were rendered obsolete, vast more opportunities were unlocked by the very innovation that rendered those jobs obsolete; this historically is the cycle of innovation and technology.
'9 to 5' made sense in that world. It was hard to coordinate externally and communication capabilities were limited. To be an effective firm and minimize the cost of coordination, everyone had to be in the same place at the same time. The pendulum swung from independent workers to structured employer-employee relationships as a means to manage manufacturing activities. Implicit in these traditional jobs is hierarchy of employment and wide scopes of responsibilities.
The pendulum is beginning to swing back the other way. With the proliferation of information technology (lower costs to connect, communicate, cooperate, coordinate), a new kind of work - one that values autonomy and output over time and input - is here. The narrowing scope of firm core competency coincides with a transition from individual jobs to tasks. Tasks are, by definition, more narrow in scope than the jobs they were encompassed by.
It marks a transition from paying for time to paying for outcome.
It is similar to artisan work. You pay a locksmith for fixing your lock, not for the time it takes her to do so. The compensation is tied to quality of output. It is the same for the independent gig earners who will be compensated by specific assignment, contract, task, or volume of sales - in direct contrast to salaried full time workers paid for their time spent working.
This is why the development of differentiated skills / offerings is important.
"As friction goes away, we will consider far fewer variables in driving decision making. Abundance of cheap infrastructure and unified technology stacks allows for very specific product offerings and proliferation of consumer niches. The broader the technology layer, the more narrow preferences can be supported" (h/t Alex Danco)
In the transition from generalized jobs to specific tasks, differentiated expertise (that is relatively less vulnerable to automation - more on this later) will be valuable. Individuals can hop from task to task. It bears little resemblance to a traditional 9-5 job with a single employer; lifetime employment becomes a relic of the past.
It is the itemization of the firm - the individual becomes the firm; Henry Ford's assembly line in the digital age.
The Gig Economy is the manifestation of this change.
The nature of the firm is changing. The nature of the job is changing. If firms are increasingly individual, and if jobs are increasingly tasks, how does coordination work? In the terminology of Coase, how are transactions costs minimized?
Firms shrink by focusing solely on core competency and outsourcing everything else. 'Everything else' becomes core competency for a growing number of other firms specializing and shaving little pieces off of formerly Goliath businesses.
Enter Uber and all "Uber for X" derivatives. What is Uber though? or Lyft? or Deliveroo, Taskrabbit, Fixerr, Etsy, Dawanda, Freelancer, Thumbtack, Upwork, etc.?
They are digital platforms that power the sharing economy on the consumer side, and the gig economy on the work side.
As a concept, they are not new though; there is a historical precedent for these platforms. These Gig Economy startups have really just automated what most traditional staffing agencies have done for decades.
Uber is a staffing firm. Lyft is a staffing firm. All these digital Gig Economy platforms are just temp staffing agencies leveraging technology to match workers with jobs, on demand.
It aligns with the individualization of the firm and the narrowing of tasks. The core competency of Gig Economy platforms is on demand coordination. They are the glue of the decentralized web of individuals making up the Gig Economy. A specialized company can now plug into the on demand labor force as required in much the same way technology companies call on other's APIs. There exists a core competency and an outsourcing of all else.
If outsourcing is the emergent behavior, the question becomes whether it is outsourced to humans or outsourced to AI (automated)?
The pernicious implication of narrow work (tasks) is that labor becomes a commodity - a race to the bottom for workers and a further shift in the power dynamic to capital from labor.
Amazon Mechanical Turk is the most blatant example of this - coined "platform capitalism" and "digital sweatshops." (4) However, as tasks become increasingly narrow, they become more susceptible to narrow AI. So even though narrow labor may be a race to the bottom for human workers, it is that very labor which is most likely to be automated away anyway.
Human desire to automate tasks — that are either prohibitively difficult for humans, or so expensive and repetitively boring that there is no justification to risk human fallibility nor waste human labor on them — is ancient.
Ancient as far back as Greek Mythology. Hephaestus' creations manifest themselves as automation:
Argus is the modern equivalent of unmanned surveillance systems
Tripodes de Automatoi were labor saving machines and autonomous vehicles
Golden Maidens were autonomous intelligent assistants
Phaeacian Fleet mirrors modern GPS infrastructure and the impending rise of autonomous transportation
Talos is the autonomous military industrial robot
Pygmalion, the autonomous sexbot
Mythological science fiction has proven prescient in its prediction of automation. Wasn’t the promise of technology to free humans to focus on other things?
As Aristotle noted,
"If every tool could perform its own work in anticipation of the need, like...the tripods of Hephaestus...then craftsmen would have no need of servants and masters would have no need of slaves”
Amazon Mechanical Turk and similar platforms represent the certain tasks that remain difficult for AI, but not for long. In many cases today, if you’re doing your job well as a knowledge worker, you’re automating yourself out of the job (alternatively, you're able to do far more than you used to be able to do — these are net productivity gains for workers and so again, firms may not need to hire the quantity of workers that they used to).
The first wave of globalization was to go to cheaper labor sources and replace expensive local talent by task. The second wave is to automation. While many view this existentially as a threat to jobs, it is also the very opportunity to bring these outsourced competencies back in house for markets that will be of greater demand in the future — energy consumption, storage, transportation, manufacturing, convenience.
If the firm exists to minimize transactions costs, what are the implications of blockchain?
Consider this - the most valuable "firm" in recent history has no board, has no formal leadership, has no employees, raised no money, and let anyone aware of its existence to invest.
To date, Bitcoin is worth over $160 Billion USD. (5) Is Bitcoin a firm? In a traditional denotation, maybe not. But it's emblematic of what is increasingly possible. DAOs — Decentralized Autonomous Organizations— are a growing phenomena under the blockchain umbrella. DAOs are as their name suggests, decentralized autonomous organizations that incorporate collaborative decision-making and bottom-up coordination from inception. There is no managerial overhead.
Trust is the underlying innovation of blockchain. Blockchain allows for building trust in trustless (digital) environments.
Trust used to require personally knowing someone. Society had personal dependence. Colloquially, in a small village, everyone you used to depend on you knew personally. When firms emerged as the coordination cost minimization tool, you began working with and depending on people you did not know. The firm lessened the degree of trust you used to have placed in individuals.
Today — in our market capitalist economy — there is far more impersonal dependence and less personal dependence. The Gig Economy and reputation systems are early building blocks of establishing impersonal trust. Blockchain allows for a marketplace where no one knows each other, everyone else is assumed adversarial, yet trust can be built regardless.
When trust can be built in digital environments, when remote communication is information rich, when distributed workers can effectively connect, the cost of coordination continues to fall. Whether DAOs represent the future is anyone's guess, but the reality is cost of coordination will continue to fall and this has profound implications for the future of the firm, future of jobs, and future of work.
Coase, Ronald Harry. "The nature of the firm." economica 4.16 (1937): 386-405.
Manyika, James, et al. "Independent work: Choice, necessity, and the gig economy." McKinsey Global Institute 2016 (2016)
Meda, Dominique. "The future of work: the meaning and value of work in Europe." (2017).
Cardon, D.; Casilli, A. 2015. Qu'est-ce que le digital labor? (Bry-sur-Marne, INA).
Gods and Robots - Myths, Machines and Ancient Dreams of Technology, by Adrienne Mayor