Lately, I’ve been writing a lot of rambles about a subject that’s been on my mind without necessarily coming to some kind of big, obvious concluding point. I like writing these kinds of things and I’m not certain I deeply care whether anyone else likes to read them, so my incentives are pretty clear and I’m going to write another one! This is one is about theories of the firm

First, read Yanis Varoufakis’ post on this subject, specifically on how it applies to his employer, Valve.

Now, Valve is a little bit of an odd case – it’s likely more homogenous than most firms in a couple key respects, namely age, experience, salary, and job description, and therefore inherently requires less internal management than most firms. But this all still begs the question – why do firms exist at all? Surprisingly, it’s still a question that doesn’t have a clear or even an accepted received-wisdom answer from economics, especially surprising since a) firms are really important parts of the economy and b) they are really weird, at least from a neoclassical perspective. Because, basically, firms are vast marketless spaces in the market. Essentially, in a truly complete and full market, every individual would be an independent contractor, signing small, finite, discrete contracts for each sensible unit of production they create, and the economy would be a multifarious network of one-on-one contracts. It would look like this:

also known as the timeline of primer

Where the black dots are individuals and the red lines are market relationships.

Instead, we have this:

also known as the andromeda strain

Where black dots are individuals and red lines are market relationships but the blue circles represent firms. So most people are in firms, and firms mostly do business with other firms, and inside of the firms there are no market relationships. Basically, individuals sign up for long term relationships with a firm at a fixed wage or something close to it and inside the firm there is hierarchical command-and-control. Certainly there are market-like incentives – good performance can lead to bonuses and promotion – but that’s different from a fundamentally market relationship where the only relationship between two equals is a contract, as opposed to having an employer and an employed.

Now, economists have come up with many theories which Varoufakis excellently summarizes, and each of these theories definitely has a part of the story. Adam Smith talks about the efficiency of labor division, which (other than Stephen Marglin) nobody I know of has thoroughly disputed, and which seems like its some part of the process. But that doesn’t explain firms, it explains specialization. Marx posited firms as basically vehicles for tacit collusion among the capitalist class to ensure uniform wages below the true average value of labor. There’s some truth to that, but I think if the only advantages to firm formation were to the capitalist class then the model would have proven less persistent than it has. Joseph Schumpeter thought that certain kinds of monopoly were necessary to create the economies of scale, both supply-side and demand-side, that financed major innovation. This is probably also true, but doesn’t quite get to the heart of the question because there are plenty of firms that are not part of the Schumpeterian innovation process that simply trudge along at a steady state. Ronald Coase’s transaction costs theory which is obviously correct and also is the only one whose theory so far explains a major corollary of the question – not merely why are there firms, but why do firms form the way they do?

There are two other established theories I’d like to discuss. One is Frank Knight’s classic distinction between risk and uncertainty and how the latter is the wellspring of profit. While not directly a theory of the firm, it is related to the theory of the firm (at least the modern firm) because a key aspect of the firm is limited liability. This is worth breaking down into two parts:

1)     Where does the liability go? It goes to the corporation itself, not its managers and employees. Obviously if someone breaks the law they break the law but no individual employee of the firm is responsible for restituting losses of the firm if it is unprofitable or fails.

2)     How much is the liability? The liability is limited to the assets of the corporation, not its owners.

So, a quick example – I own a firm that owes the bank $1 million. I have five employees, and unfortunately they are stupid and make bad decisions and the firm loses money and is forced to close. The five employees lose their job but keep all their income to date and property; the bank seizes all the assets of the corporation but not my assets.

So what does this have to do with uncertainty? Well, let’s say I have an idea that I think could change the world and/or make me fabulously rich. But I’m not certain that it will. In fact, there is some pretty large uncertainty involved here, which is probably why nobody else has put this idea into action. But thanks to limited liability, if I decide to take the big plunge and start a business to pursue this fabulous idea, I might lose all the (likely substantial) money I put into the business, but I won’t lose my house and my car and my shirt unless I explicitly agree to name those things as collateral. This is good, because it enables risk-taking (or should it be uncertainty-taking?) beyond what a “purely free market” might provide but the thing about uncertainty-taking is that a small number of successes can produce wealth that vastly outstrips the sum of all the failures.*

Another theory is the Nobel Memorial Prize-winning work of Oliver Williamson, who actually wrote about a network of effects, including some proto-behavioral effects like bounded rationality and opportunism, that lead to hierarchical firms emerging, but one in particular that I want to discuss is asset specificity and small-numbers bargaining.

Let’s say the government wants to build a road from City A to City B, and in between there are a thousand small lots each owned by a separate individual. This is a tricky situation. The issue is (and let’s just say for the sake of argument that this is through some narrow valley so you can’t just go around the lots) is that the government absolutely needs each and every lot to build the road. If I want to start a café I don’t need to buy all the coffee beans in the world, just a couple tons. This is oft cited as a clear justification for a government power like eminent domain – to deal with what isn’t one market with a thousand participants but a thousand markets with a sole monopolist.

But private individuals can’t use eminent domain! And fortunately threatening to injure or kill people is illegal and frowned-upon besides. So if a private person, say, that makes a relatively obscure widget needs a relatively obscure component, you’re going to find yourself in monopolist-monopsonist bargaining which as economists will tell you ends up taking a long fracking time and surprisingly just failing altogether. So if you don’t make just novelty birthday cards (which requires commodities like card stock and ink) but birthday cards that sing “Happy Birthday To You” when you open them then you need the computer chip that goes in the card that plays “Happy Birthday To You,” and maybe that’s a small market so there’s really only one dude who makes that, and so you’re the only person buying them and they’re the only person making them and you both have something over the other guy but you also need their thing and this is actually for all intents a purposes a hostage negotiation of sorts.

So instead you basically form a firm (“Irritating and Expensive Cards, Inc”) that internalizes and therefore negates that negotiation, you collaborate to make as many chips and then as many cards as possible and split the profits.

The thing about this logic is that it also brings surprising benefits. Say, for example, the price of card stock jumps up. If there’s no firm then maybe the cardmaker can’t make ends meet and both of you go out of business; if there is a firm you maybe find a way to cut costs elsewhere (hey look China makes cheaper chip components!) but even if there isn’t you could both stay afloat at a narrower margin for at least a little while. There’s security in a firm.

There are also network effects. This gets us to Sears. Mina Kimes wrote a phenomenal and fascinating story about how idiot Master-of-the-Universe Eddie Lampert basically attempted to atomize Sears – he should have read about the theory of the firm! Each individual element of Sears is an undiversified and volatile investment, but Sears on the whole is a diverse portfolio that creates security and stability. More importantly, there is branding and internal unity of purpose and harmony that is undermined (maybe we’ll talk about that later).

Theories of the firm can explain why firms might arise but can’t necessarily explain why firms settle, at least for some time, at the size and diversity they do. Here’s an odd case – Terminix. Terminix is a company that, superficially, eradicates pests, especially termites. But in truth Terminix is an insurance company – once Terminix certifies your home as termite-free, you can sign a contract with them that obligates them to not merely eradicate termites should they emerge but also to pay for all damage those termites would cause. My question: why do existing homeowner insurance companies not do what Terminix does? What distinguishes termites from the other long list of things that insurance companies will protect you from? Why hasn’t State Farm or AIG purchased Terminix? Why doesn’t Terminix merely exterminate as a contractor for home insurance companies who offer termite protection?

The answer is likely branding and inertia. The former is difficult to quantify, difficult to accumulate, extremely valuable, but also easy to squander. The latter has something to do with Coase’s transaction costs theory – that the Terminix/homeowner’s insurance division could emerge for arbitrary reasons, but once emerged, it could persist simply because there is not enough benefit inherent in undoing the status quo. Private equity essentially exists to find sufficiently large “inefficiencies” in the status quo that it is profitable to invest the time and resources to undo them (inefficiencies, in this case, often meaning non-legally-binding trust networks), but certainly there must be some non-negligible number of cases where the status quo persists because even if it is less inefficient than some hypothetically superior alternative the cost of transition is less than the net benefit (or at the very least the capturable benefit for the transactor).

Things like this make you wonder what, exactly, binds the now Washington Post-less Washington Post Company. My guess is just that, in general, to phrase it econometrically the best predictor of t is t-1; that is, intertia is a large force, and the finitude of human experience alone means that most of the time most things persist over the short-to-medium term. This force is worth incorporating into one’s model of the world – one should not overly privilege the status quo in any case, but certainly not when someone argues that free markets inherently lead to just or efficient outcomes.


*It’s worth noting here that this is why universal health care is a really good idea.