Re: Confidence: A Basic Politics Puzzle

Robin Hanson (hanson@hss.caltech.edu)
Fri, 21 Feb 1997 13:27:53 -0800 (PST)


I've looked over all the posts on this topic, and I don't have time to
respond to each one in detail. But overall I don't buy these theories:

HIGH STAKES -- Maybe the max/min possible outcomes are extreme, but
the expected stakes aren't out of line with other areas where people
act much less confident.

NOT VERIFIABLE -- There may be differences in fields in the ease of
verifying claims, but politics is comparable with many other areas
where people act much less confident: software/features you need for
your computer, how to drive safely, psychotherapy, preventing cancer.
Many of these areas have pretty high stakes, and poor ability to
verify expert claims, yet people admit their ignorance and defer to
others often.

CONFIDENCE PERSUADES -- Yes, confidence can help persuade sometimes.
But it can also make you look really stupid other times, and the issue
here it to explain a differnce: Why *more* confidence in politics.

These theories I consider sorta plausible:

SPECIFIC IGNORNANCE -- A standard "left" view is that politics is
mostly about values, and that a standard "right" view is that most
people are pretty stupid. These two views are wrong, but we can sorta
understand how people who hold them might be very condident and slow
to realize their error. But political confidence has been around for
quite some time; can people really be that slow?

SOCIAL SPORT -- We're in a signaling game where political knowledge is
taken as a sign of intelligence and social connectedness. So we try
to appear knowledgeable, even if this knowledge isn't otherwise very
useful. Hence the focus on persuasion over self-use, which encourages
confidence. Group leaders who might be in a position to influence
what sort of knowledge is used in this signaling game believe it is
good for their groups to use politics; it helps your interest group
get more benefits vs. other groups, and they suppose "democratic
participation is good for democracy".

A related idea is that people are trying to signal their membership in
particular social groups distinguished by political "beliefs" (really
verbal banners). Typically, signaling that you are between two groups
isn't such a good strategy.

Some specific comments:

Kathryn Aegis writes:
>Conflict theory states that the items that are less important to each
>side lend themselves to multiple avenues of negotiation, whereas
>parties tend to become entrenched on the bigger, or underlying issues
>involving face, power, authority, or influence.

This would argue for a differential confidence among political issues,
but not for a difference between politics and say psychotherapy.

Mark Crosby writes:
>Computer salespeople are rarely programmers! It is the marketing
>types that make promises such as "We can write software to do X in Y
>months". The programmers usually hear about this promise afterwards
>and can only roll their eyes and do their best to fullfill the hype.
>
>Perhaps we have a similar situation in the socio-economic policy
>marketplace where the chest-thumping politicians promise "We can
>create laws or programs to do X in Y months", and the socio-economic
>scientists can only roll their eyes and set out to gather as much data
>as they can from the latest experiment.

Yes, there is this similarity. So the puzzle is the difference: Most
people easily admit they don't know what computer stuff to buy, and
they take lots of advice from computer salesfolks.

Robin D. Hanson hanson@hss.caltech.edu http://hss.caltech.edu/~hanson/