CONCLUSION: Governing economies, with markets, states, and communities
Aaron Swartz
Chicago communities with low turnover are much more effective at
enforcing community norms against crime. Japanese fishing cooperatives
with synchronized schedules allow their members to pool resources and
share income. In general, strong communities with multifaceted
interactions are able to sustain other-regarding behaviors, instead of
just relying on people's self-interest.
Such cooperative societies were the topic of philosophers from
Aristotle through Burke. But starting with Machiavelli and Hobbes,
philosophers began treating self-interest as primary. Mandville
suggested self-interest could be harnessed for good and Adam Smith
made that the basis of his economics. Good rules thus replaced good
people as the basis for a good society.
The debate soon turned to a question of socialism versus laissez
faire. (Options in between were only rarely discussed in theory,
although they commonly existed in practice.) Mises and Hayek argued
rational planning required prices; Lange, Barone, and Lerner countered
that prices are implicit in any optimizing problem and that planning
could adopt all the information markets could and more. By the 1940s,
it was clear the socialists had one. Even their fierce opponents, like
Schumpeter and Pareto, had conceded (Schumpeter: "there is nothing
wrong with the pure theory of socialism"; Pareto: "pure economics
[can't decide] between the organization of society based on private
property and a socialist organization").
So why did socialism lose in the end? The debate had assumed the
Walrasian paradigm, but Hayek's "The Uses of Information in Society"
corrected this error and pointed out that information acquisition had
costs. Decentralized markets made efficient use of dispersed
information. Now markets are rarely as ideal as Hayek suggested, but
what information is available, markets are able to use. Hayek's paper
was a landmark, much like Mandeville's fable.
But eliminating the assumption of perfect information countered the
argument of both sides, leading Coase to study why capitalist
economies were shot dominated by miniature planned societies: firms.
Neoclassical economics could not understand why socialism failed, nor
why our system of capitalism succeeded. Let us try to do a little
better.
First, let's distinguish the main features of the standard (Walrasian)
model and the new (evolutionary) model we've tried to develop in this
book. We admit that contracts are often incomplete and markets
frequently noncompetitive. We admit positive feedback effects in
technology and social interactions. Instead of using perfect
information to predict the future, individuals update imperfectly by
looking at their local experience of the past. Instead of a unique
stable equilibrium based on stationary actions, we admit many
equilibria in which long-term averages result from varying behavior
among individuals. (Thus around 60 people may show up at the bar each
Thursday, but they are not always the same 60 people.) Statics replace
dynamics, chance becomes an essential driver of evolution instead of
simply a component of risk, instead of being self-contained the
economy co-evolves with preferences and institutions, people care
about others as well as themselves, and about process as well as
outcome. Wealth is a binding constraint on actors and society is
important in addition to simply individuals.
Further, if we are to think about policy improvements to our system,
we must model policymakers along with everything else. Public-spirited
officials do not simply enforce a policy they believe to be optimal,
but instead policy is the aggregate result of the mutual best
responses of multiple officials with different preferences.
One final note. "Extensive empirical evidence suggests that by the
standard of well-being, people are bad choosers." In lousy moods, we
take irreversible actions that deeply hurt ourselves. We are myopic,
bad at predicting our future preferences, frequently regretful,
inattentive to what actually makes us happy, inconsistent, and bad
with probability. This is true of all of us, from Harvard professors
to NYC cabbies.
What to do about this is a difficult question. Should we intervene to
help people suffering from addiction? Should we account for loss
aversion when tallying costs and benefits? (This would provide a
powerful boost for the status quo!) These are questions for another
day.
Returning to our story, the neoclassical economists not only agreed
that market socialism and market capitalism were equivalent, they
formally proved it as a necessary result of their assumptions of
complete contracts and perfect information. So to get at the real
issues, we must examine the violations of these assumptions.
The optimal system depends on the situation. In a case where there are
strong increasing returns, both individual production and market
competition are infeasible, because the positive feedback loops will
lead to one giant producer.
Institutions influence the distribution of information, power,
interest, and preferences. They may address coordination failures by
(1) giving people more control over their work and its products and
thus reducing interdependence, (2) reducing conflicts of interest over
noncontractible results, or (3) reducing the importance of private
information.
Markets make it difficult for people to collude and thus fix prices.
They reward people who honestly publish their costs of production
through sale prices set at marginal cost. (Pricing is thus an n-person
prisoners' dilemma.) They force people to publish some of their
preferences through purchase prices. (In other systems, consumers have
the incentive to overstate their needs and producers to understate
their production capacity.) When effort is aligned with reward,
markets provide a decentralized and difficult to cheat mechanism for
rewarding performance.
But this alignment is rare, usually market discipline is mistargeted.
Most people do not get paid more on days that they work harder. A
plant closing punishes hundreds of fine workers but rarely affects the
manager actually responsible. And even when effort and reward are
aligned, environmental externalities (like pollution) cause random
other people to get punished.
States, by contrast, can compel people to cooperate when doing so is
more efficient (as in undesirable prisoners' dilemmas, in contrast to
the desirable ones helpfully created by markets). States are needed to
define, assign, and enforce property rights; to provide public goods;
to regulate environmental effects; to regulate natural monopolies; to
provide some forms of insurance in the face of adverse selection
effects; and to stabilize the macroeconomy.
When multiple equilibria exist, states can help by forcing the system
into a more socially desirable equilibrium. A paper by Basu and Van
shows how a one-time ban on child labor can move a society out of a
poverty trap equilibrium and make children and their families better
off.
We've seen other examples in previous chapters (the sleepyhead in 6,
the redistribution of wealth in 9). Protecting collective bargaining
can address some of the power issues discussed in 10.
The state's flaws include its lack of access to the private
information of its citizens' and its citizens' lack of access to the
private information of the state. There is also the problem that there
is no perfect voting system to aggregate individual preferences and
most voting systems have difficulty representing preference intensity.
Government action can also lead to some groups receiving unjustified
windfalls, giving them extra incentive and resources to engage in the
political process to protect them (rent-seeking).
The problem is that the state works by being universal, but this very
universality makes it difficult to subject to competition. And the
imperfection of voting systems makes it important that there are
nonelectoral ways to influence decisions, which is what opens the
state up to interest group influence. Of course it is still possible
to subject the state to competition in many areas and designing it to
maximize accountability.
Community-based institutions were traditionally thought of as wildly
inefficient. The common lands, it was believed, were wasted until they
were enclosed for private property. Recent economic historians have
shown the opposite was frequently the case. "The open field system
used local information and peer pressure to foster innovation and
solve [] allocational problems..." Unlike the farmers of Palanpur (ch.
1), in England "three fieldmen were chosen on the first of each year
to establish the dates when [crops] would be planted, when animals
would graze and to enforce the maintenance provisions."
In Germany, community-based governance allowed for more efficient
credit contracts while networks of ethnic motel owners mobilize credit
today. Online, communities develop open source software. Communities
can collect private information that other organizations can't, and
provide enforcement through social norms. And they can harness
powerful human incentives: trust, solidarity, reciprocity, reputation,
personal pride, respect, vengeance, and retribution.
Communities fail when their preference for internal trade blinds them
to the gains they could reap from the outside world. Their tendency to
be homogenous makes it harder for them to reap the benefits of
diversity. Neither of these are insurmountable, but they are more
difficult than in impersonal state bureaucracies or markets.
Communities have trouble where different members have wildly different
costs of exiting. But markets have pretty much the same problem: when
inequality is high, the preferences of the wealthy swamp the purchase
desires of the poor.
But most institutions are a mix of market, state, and community. The
lobster fishermen of Maine developed gangs who used force to ensure
only members could set traps in their territory. More recently, the
state has formalized these gangs and required they elect leaders
democratically and intervene to prevent inter-gang violence. But this
balanced system means only six state officers are required to enforce
environmental regulations against 6800 lobstermen. State rules are
easier to enforce because of the gangs and the gangs are more
effective because they know they'll be backed up by the state.
By contrast, the Himalayan forests in Uttar Pradesh, India were once
regulated by village panchayats who levied fines against outsiders who
removed forest products. But instead of supporting this system,
British colonial administrators claimed the forest for themselves,
resulting in "incendiary protests that destroyed large strands of
pine." In retreat, they awarded the forest to the traditional
villagers as a group, which undermined the careful village system of
local regulation. The courts upheld the right of one village to steal
from another.
In nearby Palanpur we find another example "institutional crowding
out": the wider labor market has reduced the value of local
reputation, encouraging people to break their lending contracts. In
Haiffa, fining parents who picked their kids up late from daycare (ch.
3) seems to have crowded out community incentives. Experiments have
shown similar results with wage-effort questions (offering people high
wages leads them to work harder than a pay-for-performance system). A
study of Costa Rica businessmen found people were most trustworthy
when the principal was permitted to fine the agent for untrustworthy
behavior, but declined to. This seemed to set an example of trust
which was then reciprocated.
Since Mandeville we have focused on how to harness people's selfish
desires for the collective good. But these examples show there is much
more than selfishness to be harnessed and that more efficient motives
may be crowded out by systems designed to appeal only to selfishness.
"A constitution for knaves may produce more knaves."
The phrase "good fences make good neighbors" is often thought to be a
defense of clear property rights, but Frost's point was rather
different:
> He only says, "Good fences make good neighbors."
> Spring is the mischief in me, and I wonder
> If I could put a notion in his head:
> "Why do they make good neighbors? Isn't it
> Where there are cows? But here there are no cows.
> Before I built a wall I'd ask to know
> What I was walling in or walling out,
> And to whom I was like to give offence.
> Something there is that doesn't love a wall,
> That wants it down!" ...
> He will not go behind his father's saying,
> And he likes having thought of it so well
> He says again, "Good fences make good neighbors.
Perfect contracts keep getting harder, perfect information keeps
getting rarer. Generalized increasing returns, zero marginal costs,
information that wants to be free are all taking us away from markets
and back toward the mobile foraging band of prehistory. Pursuing good
ideas is more like hunting big game than working in a factory, with
new drugs as the antelope and open source biohackers as the foraging
band. We may come to find that, when fences fail, we are left
depending on good neighbors.