What exactly are free agents free from? In sports, a free agent is someone who is free of allegiances to a particular team. In knight-errantry, a freelancer is somebody who is free of allegiances to a particular feudal estate or monarch.
So what are free agents in the gig economy free from?
In the gig economy, freedom is primarily freedom from being managed. It’s a freedom that can seem like a curse to those who either enjoy being managed, or are too inexperienced to have learned adequate self-management behaviors. But like it or not, this is the freedom you have in the gig economy, and there is an art to thriving under this freedom you must learn, or it turns into a burden.
This freedom exists not because we free agents brave the wild open economy and carve out sweet manager-free territories for ourselves, but because there are rapidly growing regimes of valuable work (knowledge intensive ones in particular) where traditional people management as a function simply fails, and where the typical organizational response is increasingly to just eliminate it in those regimes. Such elimination moves often create gig economy roles as a side-effect.
The regime of effectively unmanaged work is increasing in size in the modern economy. I don’t have numbers to back this up, but my anecdotal experiences suggest managerial failures are a significant driver of growth in the gig economy.
Bad managers — represented by fictional archetypes such as the pointy-haired boss in Dilbert, Bill Lumberg of Office Space, and Michael Scott of The Office — are a symptom rather than the cause. Their existence points increasingly points to the growing untenability of people management as a function rather than individuals failing to be good managers.
Does this mean good management does not or cannot exist under modern conditions? Of course not. But it is playing a shrinking role.
The free agent economy in some ways represents a civilizational bet on a radical idea: what if, instead of trying to FIX managerial failure, we try to do without it altogether?
For students of management and organizations, the free-agent economy is in some ways the control group of people management theories. If your theory of people management cannot deliver better outcomes than the essentially unmanaged work processes of the free-agent economy, it is not a contender.
The existence of pointy-haired bosses, Bill Lumbergs, and Michael Scotts in the world is a symptom of the failure of management as a paradigm for coordinating work. It is no accident that many of these familiar contemporary “bad manager” archetypes are from fictional software companies.
Are Free Agents Managed?
I have had about 5 years experience being managed in a traditional sense, and 9 years in the gig economy where even if I stretch the definition, I cannot say I have been managed at all, in any sense I recognize.
As a free agent, you might have a client with expectations, and you might deal with someone in a “vendor management” function when it comes to paperwork and contracting logistics, but you are generally not managed the way employees are. The exception is subcontracting. If you’re managed at all, there’s a good chance another free agent — a prime contractor — is doing it. There’s also a good chance they’re doing it badly, but that’s a story for another day.
As a free agent you might also be free of other things, such as a 9-5 schedule, certain kinds of paperwork and training burdens, and so on, but those freedoms are generally not as robust. On a big gig with significant coordination needs with client employees, you might end up on a 9-5 schedule anyway. Your paperwork burdens as a small business owner or contractor might end up being greater than those of employees, depending on the client. You might have to put more effort into training yourself and acquiring certifications that open doors to gigs than employees do.
But the one robust freedom is freedom from being managed. So it is important to understand what it is to be managed, either well or poorly, and what it means to do without.
What Managers Do (In Theory)
One of the very interesting things I’m learning as a result of writing this newsletter is that increasingly, young people are directly entering the gig economy as their first foray into the workforce, and since the gig economy doesn’t really have managers (unless you count algorithms), they acquire no experience of being managed, have no idea what managers do, and no idea how to do it for themselves when necessary. So the freedom from management turns into a curse because they don’t know what they don’t know. Because they’ve never seen it. From a yearning distance, “being managed” can even start to seem like a blessing, something to aspire to if you haven’t experienced the reality of it.
So let’s review what managers actually do.
Managers do many things in organizations, but traditionally, the core of what they are supposed to do is manage the risks of individuals failing. Individuals can fail in many ways:
-
Do the wrong thing (misdirect effort)
-
Do the thing wrong (make mistakes)
-
Cut corners and do poor work out of laziness
-
Work too slowly, creating delays
-
Game incentives and work to minimal standards
-
Act maliciously due to unresolved resentments
-
Act unreliably due to personal life issues
-
Lie or cheat in reporting on work
-
Fail to resolve conflict with other employees
-
Become unable to work due to illness
-
Fail due to lack the right resources to succeed
-
Fail due to essential tools or systems failing
Managers exist because organizations need to function despite this vast potential for individual failure. Historically management arose as a function out of the need to mitigate the risks of such failure. Many of the other things managers traditionally do, such as set targets, supervise training, fight for budgets, and relay information, arise out of the primary job of managing the failure potential.
The cynical assumption that employees will fail in various ways unless actively managed is often known as Theory X, while the opposed idealistic theory, known as Theory Y, popularized in the 60s by Douglas McGregor, holds that left to themselves, employees will generally do the right thing. Under Theory Y, the core of what a manager supposedly does is promote the growth and well-being of employees, rather than manage the risks of their failures.
Both theories are empirical on the surface, and could in principle be tested. You could research a company to see if employees are succeeding or failing by default, and whether failure mitigation or growth promotion adds more value. But in practice, Theory X and Theory Y tend to be workplace ideologies adopted as untested values, rather than being selected as the more accurate description of the specific workplace.
Theory Y was popular for a couple of decades, but with deregulation and increasing competitive pressure in the 80s, Theory X enjoyed a renaissance.
The early phases of digital transformation of businesses in the 90s and 00s saw the increasing strengthening of Theory X tools (Big Brother at work basically) — worker surveillance tools, self-documenting, high-transparency workflows, open-plan offices, and an increasing emphasis on “collaborative” cultures (which is usually code for reliance on a culture of peer surveillance).
Theory Y tools also emerged — goal setting tools, “performance management” tools, “employee engagement” tools, and so on. But despite cosmetic Theory Y layers, most workplaces today, including knowledge-work-intensive workplaces, have a default Theory X culture.
In theory managers manage failures. So what happens when managers themselves fail?
When Managers Fail
Today, at the tail end of the neoliberal, globalized era, we can say that to a first approximation, Theory X (with a cosmetic veneer of Theory Y) is used to manage low-wage, low-skill, highly interchangeable and precarious employees, while Theory Y is used to manage a small subset of high-wage, high-skill, hard-to-replace secure employees.
Theory X is used to manage workers destined to have their jobs eliminated or dumbed down maximally through automation, while Theory Y is used to manage the shrinking number of workers who are irreplaceable by automation and very expensive to replace with other humans.
So why free agents?
Free agents exist because whether they are serving in a Theory X or Theory Y role, sometimes managers fail and turn into net liabilities. This is most likely when the work being done has a strong element of a principal-agent problem, due to knowledge or skill gaps (the manager doesn’t know or cannot do things the employee does). This can happen due to either specialization of roles, or situational differences when manager and employee are not collocated (as in remote work, or more recently, work-from-home conditions under Covid19; anecdotal evidence suggests managers are really having a hard time being effective under WFM conditions).
Either way, the ability of the manager to either address failures, or promote success, is sharply limited by lack of knowledge (contextual or specialized), lack of skills, or both.
So you have this matrix:
If the existence of managers isn’t helping failing employees become successful, or successful employees grow, what’s the point of them?
By presenting ways to do away with managers altogether, this is the question software tools allow you to ask, and often the answer is “there is no point.” So software eats managerial roles.
The resulting elimination of human managerial functions also drives the structural evolution that refactors the job itself as a gig economy job.
When the response to failing management is to do away with it, the free agent economy grows.
Managerial Failure = Gig Economy Growth
There are two potential outcomes of managerial failure: below the API gigs and above the API gigs.
In the below-the-API case, you can replace managerial roles with automated software layers that provide the necessary coordination function, dispense with the other softer functions, and rely on opt-in market dynamics and economic incentives to shape the managed function. This is how you get under-the-API gig economy: rideshare drivers and the like.
Note how this layer works. The failure modes are not “managed” for the most part. Absenteeism and slacking are not motivation problems to be “managed” for example, requiring either penalties imposed or a pep talk from a manager. Rideshare drivers simply work when they want to and are paid accordingly. If they do a poor job, bad ratings kick in and eliminate them. On the success side, if they do very well, the algorithm itself rewards them, along with better tipping dynamics.
Sweet deal, huh? Except of course that any job that is legible enough to be refactored this way is also very likely ripe for automation.
In the over-the-API case, you get the kind of free agency we mostly talk about in this newsletter: indie consulting, high-skill contracting and so forth. In this case too, there is no management. If you fail, you simply lose the gig. Management is reduced to hiring/firing free agents and dealing with them as vendors/suppliers of work rather than employees. If you do well, you get more inbound via referrals, and can either make more money by working more, or by raising your rates.
To succeed in either case, you have to learn the art of being unmanaged.
Managing Yourself
In the under-the-API case, being unmanaged is easy. You learn to play the algorithm’s game as best you can for your kind of work. As a rideshare driver, maybe you only work during surge pricing. Maybe you loiter in areas where you get many quick turn-around short trips rather than rare long trips with long empty backhaul legs. If the platform changes the algorithm, you change your behavior and figure out a new optimum or quit the system.
Managing yourself under the API simply means setting goals and then working the algorithms to hit them sustainably with the least effort possible. For example, I saw some research I can’t find now that showed that many rideshare drivers simply work till they hit a target daily amount, and then punch out, regardless of whether it’s a high demand period when they could make a lot more money in a short period. Where a managerial solution would try to motivate the drivers to work more to meet the demand, the algorithmic solution is to raise prices and move to a different market equilibrium. Customers change their expectations and intentions too. Those unwilling to pay surge prices simply wait out the surge or find alternative modes of transportation. As a result, the system, while not a one-to-one substitute for an equivalent “managed” service, works well enough to sustain itself.
In the above-the-API case, managing yourself is much more complex. The two obvious things you can do are simply supplying the Theory X and Theory Y functions yourself to the extent you need them. You can monitor your own failures and try to learn from them. You can watch for where you’re doing well and double down there by giving yourself pep talks.
But the biggest aspect of managing yourself above the API is to get to an oversubscribed state where you can pick and choose what gigs to take, and where you have the genuine freedom to walk away from gigs that aren’t working out. The better part of managing yourself is simply working on the right sorts of gigs, and saying no to the wrong sorts.
In other words, instead of managing yourself, you seek out work where you don’t need as much management.
The Unmanaged Future of Collaborative Work
As the free agent economy grows, and takes on more complex functions, requiring increasing coordination among free agents, the art of being unmanaged will evolve.
Some of the experiments we are doing over at the Yak Collective involve researching precisely this evolution.
How do you run team projects without a traditional project manager?
How can a multi-role project get staffed and completed with much less cat-herding by a manager, or even no cat-herding?
Can you create rideshare like algorithmic platforms at the scale of a single small project?
These are problems that we are just starting to figure out. In every case, there is a temptation to simply reproduce the traditional solution: “manage” the problem with a “manager.” But it is the solutions that reduce or eliminate the need for management, while producing equivalent or better output that will be the interesting and exciting ones.
In most cases, these won’t be 1:1 substitutes for managed project outputs. When you work in unmanaged ways, you tend to do different kinds of work as well. The nature of the economy itself changes. It creates wealth and meets needs in different ways than one with managers. You get more things that look like crowd sourcing, and fewer things that look like curated, “managed” outcomes. Things aren’t managed or left unmanaged via unexamined defaults or due to the inertia of past practices.
Instead, management becomes yet another design variable in business models.