<aside> đźš© Flagship Game: None yet.



With this chapter, we’re interested in redesigning the basic systems of society—systems like voting, markets, organizational hierarchies, social networks, academic science, or the court system.

One way to learn about these systems is to trace the boundaries of where they are used. Voting, for instance, isn’t used to decide everything inside a country or an organization. In a modern democracy, decisions involve a mix of law (both legislative and case law), administrative bureaucratic technocracy, court systems for adjusting laws, and elected representatives with powers defined by laws.

It would be disastrous to choose any one of these mechanisms to make every decision. Voting, for instance, is awful at long-term planning and at reconciling competing concerns—and both court systems and technocratic bureaucracies address these problems in different ways. You can imagine if popular votes were used to decide court cases, with no notion of precedent or law, and no written decision by a judge. Or if cities had no planning bureaucracies, and the direction of each road was subject to a popular vote. As it stands, voting is good for certain collective choices, while courts and bureaucracies are good for others.

It's similar with markets. If they were the only kind of economic decision-making needed then we’d have no firms—everyone would be an individual contractor. Or firms would make all their decisions using internal markets. The domain of markets also gets limited by long-term contracts and treaties, by regulation, by unions and households, and by class-action lawsuits. Noticing the length of contracts and the sizes of firms is one way to see where a non-market basis for decision making works better. Supply-chain contracts tend to be long-term and research-heavy firms tend to be large. This suggests that making and coordinating long-term plans and flows is something that markets, like voting, don’t do well.

Three Lenses

We'll practice looking at an institution through three lenses — as a decision-making device, a deal-making device and culture-shaping device. These lenses come from a variety of fields, including political theory, economics, game theory, mechanism design, social choice, game design, and sociology. Using them, non-experts can learn to make smart evaluations of institutional ideas quickly.

This is a simplification of these fields, but I hope you'll agree that it’s surprisingly powerful. I’ll give a quick tour.


Lens 1: Decision-making

If an institution seems to make poor decisions — in the sense that it justifies projects with a logic that doesn’t really work — it will be deemed illegitimate. For instance, markets support projects on the basis of purchases. The idea is that the customer expresses their desires and interests through purchases, and these desires justify a project as good for the customer. The market uses purchase information to decide which projects go forward.

But when markets using this rule decide that the sale of highly addictive drugs is a justified project, the decision-making process of markets begins to be seen as illegitimate (at least in this area). Decisions about addictive substances are then moved away from the domain of markets and into the domain of law, which uses a different decision-making procedure. (And of course, the law may be deemed illegitimate for other reasons — leading us to look for a third institution to make a decision, like courts or class action.)

Lens 2: Deal-making

Every institution must also harness the interests of participants towards a beneficial group outcome. That is to say institutions must also make deals. But often the participants are insufficiently motivated, or they are motivated to act in a way that undermines the beneficial outcome — for instance through conspiracy, or fraud, or manipulation. In these cases, the institution fails to make good deals.

Voting fails at deal-making when the voters have no incentive to show up or to inform themselves, or when political deals are made behind the scenes. Political parties have some benefits, but they are also a kind of conspiracy which can rig elections (for instance, by exercising a form of monopoly power, through hacks like gerrymandering, or by manipulating media narratives). These perverse incentives can make voting with political parties ineffective as deal-making devices.

Lens 3: Culture-shaping

Finally, every institution requires a certain culture, and it may or may not support the culture it requires. Academic science depends on specific values — like rigor and curiosity — or it doesn’t go well. But the mechanisms of science (like publications and tenure) may in certain cases undermine the very values that science depends on (for instance, if the pressure to publish leads to a decline in rigor). Science will then undermine itself, and the “values” of science will become performances. Something similar can happen with the values of democracy (like public debate and open-mindedness) or with those of markets (like fair dealing).

Summing up

To summarize, institutions must be robust against three kinds of failure. First, if they produce unjustifiable outcomes, they fail at decision-making, and lose legitimacy. Second, if their incentives are gamed, they fail at deal-making, and end up rewarding fraud, defection, or bad actors. Finally, if they undermine the values they depend on, they fail at culture-shaping, and participation becomes meaningless, performative, or careerist.