Ch. 11: Summary
Aaron Swartz
We can think of institutions as equilibrium outcomes of some
underlying game. Change can happen within groups, where a whole group
tips from one institution to another, or between groups, where a group
with a different institution outperforms and replaces other groups.
Institutional change often involves six factors: (1) conflicting
interests, leading some subgroup to have defection as its best option,
(2) chance or exogenous developments, which make the old institutions
no longer a best response, (3) collective action, so that some
coordinated subgroup can try to tip the group to a new institution,
(4) inefficient institutions, where everyone is better off by
switching, (5) punctuated equilibrium, where unraveling of some
institutions undermine the stability of others leading to large-scale
change, and (6) coevolutionary processes, where preferences lead to
new institutions and new institutions lead to new preferences.
Our preferences do not come out of nowhere, they're caused by genetic
inheritance and cultural learning. Advertising and indoctrination
presumably shifts preferences as well, but studies have found that
advertising is less important that the influence of one's friends.
(uncited) "Preferences are like accents; we can try to acquire them --
learning to love Prokofiev and snails, or adopting an 'upper class
accent' -- but for the most part we are only dimly aware of how we
acquired them." So we'll model preference shifts like linguists have
modeled language shifts. "Our basic language system is not acquired
from school teachers or from radio announcers, but from friends and
competitors: those who we admire and those who we have to be good
enough to beat." (Labov 1983:23)
For the most part, we've looked at how people make decisions so as to
benefit themselves, but humans are very conformist creatures and
they'll often do things just because everyone around them is. But
institutions also privilege people who take the popular side of
things, giving a structural advantage to conformists. When conformism
is modeled, we find that adopting lower-payoff institutions is an
evolutionarily stable strategy (explaining inefficient institutions).
Furthermore, conformism creates tipping point effects where everyone
adopts one strategy or another instead of some mix of strategies
(explaining punctuated equilibria). If we imagine that some people are
learners and others are imitators, we see that there's a free-riding
problem. Imitators copy the learners, providing an increase in social
benefit that the learners don't get compensated for. Imitators can
insist on only pairing with imitators, but this has the same effects
as copyrights or patents: it reduces social welfare.
Now examining real cases of institutional preference change is tricky
because people's preferences are often developed during adolescence
and institutions rarely change. But a cool example is when markets
enter into nonmarket societies: witchcraft is often seen to follow.
When they started exporting cocoa from Ghana, theinflux of money
undermined the previous communal property regime and witchdoctors
"sprang up to adjudicate disputes on the expanding boundary of the
crop. Similar episodes occurred in olivia with the advent of tin
mining, in Colombia with the spread of sugar cultivation, and in
seventeenth-century Salem Village (massachusetts) with the growth of
commerce along the road running north from Boston."
One can also do experiments. Doing the ultimatum game around the
world, we found that the more the society had increasing returns to
scale and market-like systems, the more likely its members would
reject low offers. Why? It's also been found that childrearing
strategies are highly correlated with economic structure. It seems
likely that in societies which would benefit from cooperative
institutions, there is preference change to be more cooperative.
How did humans transition from collectivist hunter-gatherers to
market-based farmers? Imagine there are three types of people: sharers
(who share the spoils with their partner), grabbers (who try to take
the spoils from their partner), and punishers (who act like sharers
unless they meet a grabber, in which case they call all the other
punishers in the population to punish the grabber). This game has two
equilibria: the Hobbesian equilibrium of mostly grabbers and some
sharers (a handful of punishers aren't much improvement over the
sharers) and the Rousseauian equilibrium of sharers and punishers.
Interestingly, while the punishers are good at fending off the
grabbers, they don't do any better than the sharers and so tend to get
replaced by them over time, allowing a later influx of grabbers and
thus to the Hobbesian equilibrium. The price of civilization is
eternal vigilance, I suppose.
So why is real civilization much more like Rousseau? Conformism may
push societies toward the all-punisher Rousseauian equilibrium, making
it more stable. If everyone's a punisher, even weak conformism will
keep them that way because the benefits to becoming a sharer are near
zero and the punishment for becoming a grabber is severe and cheap to
administer. And if a few grabbers ever happen to appear by random
chance, the punishers will benefit at the expense of the sharers,
providing another impetus to stay vigilant.
But another interesting thing humans do is they punish the
non-punishers. With this form of enforced conformity (even if the
punishment is very small), the all-punisher dynamic becomes much more
stable -- after all, all it needs to do is prevent drift toward
becoming a sharer, which doesn't have any compensatory benefits. And
if there are increasing returns to scale, all the more reason to be
cooperative.
Getting back to agriculture, moving to a world in which people might
have clear property rights allows for a Bourgeois strategy for this
game: grab if you own something and share if you don't. Bourgeois can
completely take over the Hobbesian equilibrium, replacing everyone
with Bourgeois, which is even more stable than Hobbes. This was
impossible in the hunter-gatherer world, because there was no way to
clearly delineate property rights in unhunted game. Further,
agriculture allowed for personal stockpiling which reduced the need
for communal sharing to smooth out consumption. And, of course, it
increased productivity, allowing for growth.
Let's be really crazy and try and do some detailed simulations of
early human history. (The simulations are so complicated that we can't
simply solve them with equations like we usually do, we have to really
simulate things this time.) It's made up of people on a plane that
wraps around at the edges (which, if you think about it, is a torus),
who play the Sharer-Grabber-Punisher game with each other, then talk
to a neighbor and see what strategy they've pursued and how much
they've made, occasionally go to war with each other (the richest
society usually wins and the losing society learns from the winning
one and has to hand over some winnings), and so on.
With (1) intergroup conflicts, (2) second-order punishment (enforced
conformity for punishers), and (3) conformism, you get Rousseau.
Without them, you get Hobbes. With just 2 and 3, you get more grabbers
and less sharers (3 favors the grabbers, while 2 hurts the sharers).
With 1 and 2, it moves back and forth between Rousseau and Hobbes.
Intergroup conflict between punishers and sharers tends to favor
sharers, causing a downward spiral to Hobbes. With just 1, the
population is evenly divided. If you replace Grabbers with Bourgeois
(who they strictly dominate), it depends on how clear the property
rights are: Bourgeois take over when property rights are clear more
than half the time.
[In other words, the important thing about Bourgeois is that it allows
for a decentralized mechanism of sharing-enforcement (i.e.
Punishment). Punisher requires the whole society to gang up on
violators of its norms, but under the Bourgeois strategy people can
settle disputes for themselves and it's in their own individual best
interest to do so. This insight is at the core of the
free-markets-are-natural case. --AS]
In short, just as Marx said, the shift in institutions was caused by a
shift in technology. And we've seen this happen more recently as well:
barbed wire let people fence in portions of the American Southwest,
New England water mills transformed riparian rights, the Plains Native
Americans were transformed by the introduction of horses. And future
changes in technology may lead to future transformations.
See you next time.