# Corruption potential and efficiency are anti-correlated Corruption is a [[Nebulous]] term, but it has a sense of someone going outside the official way of doing something in a way that benefits both themselves and an external third party. Corruption undoubtably has a negative connotation. It also has a connotation of waste and things slowing down. I want to challenge the latter connotation and bring up some uncomfortable questions about the former. Corruption is all about *process* and failing to follow legible processes. It’s easy and common to assume that corruption always leads to bad outcomes because when we talk about it we implicitly assume that the process leads to good outcomes. However, when you step back and think about all the processes you’ve encountered in your life, it’s clear that this is a terrible assumption. Many processes suck and lead to bad outcomes. So we can actually divide corruption into inefficient corruption that has bad outcomes and efficient corruption that has good outcomes. The classic cases of corruption fall into the first category - an official taking a bribe to ignore a violation of health policies or to give the contract to a less-qualified contractor. However, what would happen if someone took a bribe to get a beneficial drug through FDA trials faster? What is an official getting their extremely competent but on-paper unqualified friend a job? These cases are the latter category of corruption. [[Railroads were unprofitable at first]] and there is a strong argument that the subsidies and land that allowed them to expand all over the US were driven by corruption. Railroad expansion would have been much slower otherwise. [[Governing is all about new tradeoffs]], and at some point there may be a tradeoff between fairness, efficiency-generating corruption, and efficiency-killing corruption. [[China’s Corrupt Meritocracy]] argues that the Chinese government has sacrificed fairness in favor of efficiency-generating corruption that also pushes out efficiency-killing corruption. The danger with this tradeoff is that a system that enables efficiency-generating corruption can easily flip to inefficiency-generating corruption. The way that the Chinese system handles that is that all of the corruption is potential grounds for dismissal. As long as you’re deemed to be doing a good job, no problem. Do a bad job (flip from efficiency-generating to efficiency-killing) and there’s tons of ground for dismissal and ridicule. The upshot may be that corruption is an unavoidable effect of a relatively unbuffered governing system. [[The ARPA Model is an unbuffered system]]. The tradeoff between corruption and efficiency suggests that the other tradeoff to consider is the one between scope and potential downside. ### Related * [[Legible games enable people outside of an institution to control it]] [Web URL for this note](http://notes.benjaminreinhardt.com/Corruption+potential+and+efficiency+are+anti-correlated) [Comment on this note](http://via.hypothes.is/http://notes.benjaminreinhardt.com/Corruption+potential+and+efficiency+are+anti-correlated)