Enforcing the Consequent

Posted on June 8, 2020 by Jack Kelly
Tags: coronavirus, logic

To successfully do anything at all, you must have some idea of how cause and effect works in the world around you. I’m getting very worried by a pattern I’ve noticed, where large organisations are actively disconnecting themselves from reality because they cannot honestly achieve their goals. This cannot work, but it’s worse than that: the now-blind organisations cannot notice it not working, and risk becoming stuck in their delusions.

This phenomenon feels like a special case of Goodhart’s Law (“when a measure becomes a target, it ceases to be a good measure”) crossed with the fallacy of Affirming the Consequent (“All men are mortal. Socrates is mortal. Therefore, all men are Socrates.”). The pattern, which I’m calling enforcing the consequent, looks like this:

Boosting B so aggressively often means turning off feedback mechanisms which would signal that more B is a bad idea. I’m seeing this happen in really large organisations and systems, and that scares me: big systems can make big mistakes. Examples after the jump.

Higher Education

A university education used to be a near-certain ticket to a good job, and employers could be confident that degree-qualified applicants were of high quality. Giving degrees to more people doesn’t magically make them better or more productive, especially if you have to lower standards to do it. Universities are forced to steadily grow at the best of times (so that there’s enough new professorships to motivate the postdocs (so that there’s enough postdoc places to motivate the PhD students (so that…))), and a rapid expansion of undergraduate places means even more rushed growth.

Sustaining this growth is a funding challenge, and universities have addressed this by taking on a students-as-customers mindset, adding more international student places, expanding undergraduate offerings, and lowering standards (“More than half of staff surveyed did not believe their university offered a better level of education than it did five years ago” - and this report is from 2008!).

This is enforcing the consequent. Having decided that we want more of the sorts of people that used to graduate university, we crank up the number of graduates, close our eyes, and hope that the people who come out are still as good. Degree printer go brrr.

This has predictable effects on student outcomes: declining full-time employment rates for recent graduates, increasing underemployment for graduates and reduced graduate starting salaries. As the value of a university degree conveys less information about the person holding it, employers (including Apple, Google, Netflix, and the big four consultancies) move towards other ways of assessing candidates.

WHO

A larger, more recent, and possibly more dangerous example, and the reason this post is also tagged coronavirus: Youtube’s CEO has indicated they’ll remove anything that contradicts WHO guidelines around SARS-CoV-2 response (BBC, Business Insider). Other tech companies have stated similar intentions. Let me remind you that this is the same WHO that dragged its feet on acknowledging human-to-human transmission, on declaring a pandemic, and on saying that COVID-19 is worse than the flu.

If the World Health Organisation fails to secure world health, it is perfectly reasonable that people on the ground lose trust in it. If the WHO has become less credible than people pushing snake oil, then it has failed.

There’s no point having an organisation like the WHO unless it is trusted on matters of world health. The right way to do this, the long way, would be for the WHO to apologise, correct the processes by which it offered so much dangerously wrong information, and begin consistently offering correct advice about world health. Instead, the response has been to enforce the consequent:

Let’s all hope like hell that the WHO magically does better next time, because it’ll be the most trusted voice left in a very empty room.

Why it Scares Me

The common thread connecting these examples is the intentional decision to disconnect from an error-correcting mechanism. But once you’ve done that, how do you know when you’re going the wrong way?

Lowering standards to pass more students blinds a university’s internal feedback mechanism, and the university can carry on issuing degrees for a long time before reality catches up. The external feedback loop is much longer; it comes from employers and graduate admissions offices judging the quality of a university’s graduates. Students first have to complete their degrees, attempt to enter the workforce, and then the outside world needs to start noticing that graduates aren’t as consistent as they used to be. Only then can any pressure to improve graduate quality begin finding its way back.

Large organisations like the WHO digest information and speak truth slowly, if they ever get around to speaking it at all. This is disastrous for an organisation that must act quickly in response to health emergencies. It was contrarian voices who first got their heads around this crisis, and cutting them out of the conversation will get a lot of people killed when the next pandemic rolls around.

In both cases, the policy changes amount to putting on the blinders and crossing your fingers, and the consequences of these changes will not appear until years or decades later. That’s a long time for errors to accumulate, and those errors will do a tremendous amount of damage before any error-correction can kick in.

Exercise: look at the news media through this lens, and then try to sleep soundly. Not only is the whole industry a fractal of people enforcing consequents to further their narratives (recent example: downplaying SARS-CoV-2 was enforcing the consequent), the news media is supposed to be an important part of how the public gathers information and makes sense of the world. This information flow has been compromised; how many systems in our society now run on damaged feedback loops?

Previous Post
All Posts | RSS | Atom
Copyright © 2020 Jack Kelly
Site generated by Hakyll (source)