Trustworthy corporations

An exotic model for corporate oversight.
April 3, 2018 (policy, economics)

This is an outline for how to construct a trustworthy corporation, in certain circumstances. Some parts need more fleshing out to be usable.

Why does it matter? In the world we live in, large corporations produce a huge amount of value, and do many many useful things.

There is a big downside though - large corporations aren’t trustworthy. They have a drive to be useful and efficient and innovative, but also a drive to be underhanded and manipulative and monopolistic, and both come from the same place - seeking for profit in a market economy.

So, I’d like to find a way to keep that profit-based motivation for the good things - the efficiency and innovation - but eliminate it for the negative behaviors, to somehow arrange for what we would consider greed to not pay off.

I don’t have a general solution - for reasons of my own, I’m only going to consider cases where there’s a simple, primary purpose to the corporation. For example, we could summarize Intel as “they build us CPUs”, or we could summarize Comcast as “they provide us with internet access”.

The basic structure is to define an acceptable level of service and price. For example, for an internet provider that might be “100Mbps for $50/month”, indexed respectively to networking tech advancements and to general inflation.

Then, rather than the traditional strategies of

- some government agency enforces this
- hope the corporation plays nice
- count on competition to keep them in line

which all have serious problems which you’re probably familar with and that I won’t go into here, we apply a novel enforcement mechanism.

Every year, the corporation creates a new legal entity, a small spin-off that we’ll call “Enforcer”. Then, the corporation signs a very one-sided contract with Enforcer. The contract says that if the corporation doesn’t “play nice” (defining this with the appropriate rigor and level of abstraction is the biggest open question here, but I think it’s often possible), Enforcer can sue the corporation for 3x its profits that year. Also Enforcer gets access to the corporation’s internal communications and so forth during the suit, so if something shady is going on it can definitely find out.

Then Enforcer gets auctioned off to the highest bidder, proceeds to charity or launched into the sun.

And that’s basically it! So why does it work?

It works because it changes the incentive curve for the corporation. It can still make a nice little profit by doing what it’s good at and charging a reasonable profit, and that’s fine. The problem before was that it had the option of making even more profit by being underhanded - but that’s no longer the case, because Enforcer is in the way. Now the choice is “be nice and make money OR be not nice and lose everything” - and I’m pretty ok with trusting faceless corporations to make that decision for themselves.

Let’s dig into the Enforcer a little more, because this is pretty strange. Instead of making the corporation answerable to the government, or to its customers, we make it responsible to some arbitrary third party - whoever feels like buying up Enforcer at auction. Who would that be? Well, probably your favorite Wall Street boogeyman. Some other rich, profit-seeking corporation, with lots of resources and highly-paid lawyers, is going to buy Enforcer, and they’re going to be super aggressive about rooting out bad behavior - because there’s a ton of money in it.

So we’ve basically constructed a customized regulator, which is profit-driven to regulate effectively, and is bootstrapped on top of the court system as a source of relatively even-handed arbitration that’s subject to neither regulatory capture nor populist mob mentalities.

It’s a little unfortunate that in the event of some transgression, it won’t be the wronged parties who get the payment money, but rather whoever bought Enforcer. I can live with that though, because it’s better to arrange for transgressions to be few and far between, than to arrange for them to happen all the time and try to make up for it with reparations.

To summarize

- this is immune to regulatory capture
- it causes the corporation to want to be nice, not to want to be only as greedy as it can get away with
- this doesn't require any laws passed or government support, aside from a court system

And there’s one last big question - why would the corporation ever agree to this? This is a topic I may get deeper into another time, but it boils down to the expectation of more engagement as a result of promising, in a very concrete way, to play nice.

To see prior art of this dynamic playing out, look at free software, for example Linux. Open-source software is very similarly a one-sided promise to play nice (effectively the creators give you a license to use it without asking for anything in return). The payback is that people are much more willing to use and contribute to these projects.

So it would probably be adopted by a new entrant to some market that can use the I-play-nice guarantee to attract engagement, rather than being retrofitted onto some corporation that is already dominant.