The Pervasive Role of Priors: Part Two
Continuing the thread from my previous post, available here, on how prior beliefs and ways of viewing the world affect one’s conclusions:
One person who commented on my prior post noted that people have priors as to how much justification is needed for government action. I completely agree with this. I think that in many cases, that prior is importantly related to, and to some extent caused by, another prior, about what sort of people are in government.
I was brought up thinking that people who went into government were generally “good” – that they went into government to do what they thought of as the right thing, and that in many cases, what they thought was the right thing was pretty good. I recall watching the Alan Alda movie The Seduction of Joe Tynan, which I remember as being about a “good” politician completely compromising his principles to retain power, and feeling both sad and startled.
In a sense, my prior has stayed with me, even though I know of many-probably thousands at this point- examples to the contrary. I recall reading fairly recently (over at Marginal Revolution) about a government decision to keep the speeding limit on a stretch of highway too low. The low speed limit causes accidents, but yields lots of revenue from speeding tickets, presumably the motivation for keeping the limit as it is. My reaction was a bit of shock and dismay.
Many people have quite a different prior: that people in government mainly want power (whether or not they started out that way). Maybe they (also) want to achieve something, and maybe it’s something they think is “good.” (But what they think of as “good” may not be good to many if not most correct-thinking people.) Yet maybe they just want goodies for themselves or their friends. And, whether they are well- or badly-intentioned, they won’t be subject to the rigors of the marketplace, and hence may be wasteful and inefficient.
Clearly, there is evidence for both priors (and of course others as well). It seems unlikely that either prior can be proven right. It also seems unlikely that a person holding one of these priors could be convinced of the other one, especially by the person most likely to try to convince him, the person holding the other one.
That I reacted with shock and dismay tells me that some part of my prior is still intact. I apparently organize my worldview by viewing these sorts of examples as exceptional. This sounds like a confirmation bias, but that’s terminology (the word ‘bias’) I want to resist here.
Because the ultimate truth here isn’t knowable, the principles by which I’ll take in new information should, I think, be dictated in part by just plain usefulness, which is completely consistent with having a prior but understanding that there are some exceptions. I can have the prior I described above but be open to being convinced that in the case of not changing the speed limit, government acted badly.
Here is terminology I want to embrace: cognitive miser. Having a prior subject to revision but trying hard to get the revision process right strikes me as the guiding principle — in this case, there’s no strong reason why that should require getting rid of the prior. I think this may be true more than it typically acknowledged: while jettisoning your prior may sometimes be the right thing to do, there is no reason to not accord your prior some deference.
I of course think the dynamic I describe happens on the other side(s) too. Indeed, I think the facts of the sticky speed limit case I linked to above could be recounted a bit differently, in a way that wouldn’t serve the government-as-corrupt narrative quite so well (see the report linked to in the post).
My bottom line is this. A view that government regulation needs to be very strongly justified may come in part from a view about what sort of people are in government. Arguments about the merits of a particular regulation, or regulation in general, may not get anywhere because the person favoring regulation has a more government-favoring prior than the person against the regulation.
Neither may know about the other’s prior. And whether or not they know, the priors will resist being changed. A person who thinks government mostly screws up or just helps itself will be far more wary about anything government proposes to do than will a person who has some faith that government is ‘good.’