One oversimplified but potentially useful way to think about corporate regulation is to prevent it from being profitable to cause harm.
Of course, this can be very challenging to achieve, and one can debate what “harm” means, how broadly it should be construed, and what to do in cases where there is substantial uncertainty about how to make harm unprofitable. But insofar as it’s unprofitable to cause harm, and companies behave as profit maximizers, companies are basically forces for good.
How to make companies into engines of good: make harm unprofitable. If a machine is designed to do X, it shouldn’t come as a big surprise if it approximately does X. However, we want to make sure that X rules out things that we really don’t want the machine to achieve. If it doesn’t, we might end up hating the machine.
In other words, if we think of companies as an optimizing system that (to some degree of approximation) attempts to maximize profit for shareholders, then if the set of very profitable options includes doing harm, we can predict that companies will sometimes take those harmful options.
Making it unprofitable to do harm is different than, say, punishing bad things companies do in proportion to how bad a crime seems intuitively. It ideally takes into account how much punishment is necessary to offset the average profit that could be made from violating that rule (otherwise, if the profit potential is large enough, even once the potential punishment is accounted for, we can predict the rule will be violated by many companies).
This piece was first written on July 20, 2017, and first appeared on my website on June 10, 2025.
Comments