Equality by lot

“What is the best government?” Can we answer this empirically? What kind of research could be conducted to answer these questions? I ask the readers of this blog to generate some ideas.

Potential Research Areas

Randomly Controlled Trials

Small-scale trials of democracy are indeed possible, and it might be possible to measure outcomes. Randomly Controlled Trials (RCT) are the typical gold standard, where a random selection of participants get a treatment and the rest are assigned as the control group.

For example labor unions or worker owned cooperatives might be persuaded with financial incentives to become test subjects. However, small scale trials might not be able to simulate the incentive structures of large democracies which may be subject to phenomenon such as “rational ignorance” and the reduction of the “expected value” of a participant’s vote.

On the other hand, perhaps small organizations *are* able to simulate these incentives. For example the much reviled American Homeowner’s Association, allegedly run democratically at smaller scales yet also hated by its participants. Because many do not participate, this leads to the self-selection of busy-bodies who annoy the nonparticipants. Could sortition be a solution to HOA hatred?

Testing can also be performed on for-profit oligarchic institutions such as Employee Stock Ownership Programs or shareholder corporations. If sortition is able to avoid rational ignorance, they may work in both democratically and oligarchically organized firms.

Lab Experimentation

Another possibility is laboratory experimentation on psychology students. Are there any open questions on human psychology and economics that can either corroborate or refute claims made by sortitionistas?

The advantage of these lab experiments is that they may be vastly cheaper than a field experiment. A lab experiment also may be needed to drum up interest for larger studies, though potentially the current deliberation work by for example James Fishkin may be sufficient.

Game Theory, Mechanism Design, and Social Choice Theory

Other further research areas to explore is to go back to theoretical economics and to perhaps explore how “boundedly rational agents” would behave in a sortition-based system. As far as I understand the academic landscape, this falls into the scope of “game theory”, “mechanism design”, and “social choice theory”. An advantage of this economics approach is that we could align the incentives in a proposed system we design to maximize the chances that “doing good” is rewarded whilst “doing bad” is punished.

Computational Simulation

In engineering and science, computer simulation is a common design tool. Simulations are used to simulate climate change, car crashes, orbital mechanics, engineering design, and a variety of phenomenon. Could simulation be used for politics?

Even if computer simulation may be vastly more inaccurate than the real world, it is vastly cheaper to do a simulation, than invest hundreds of millions of dollars in real life design iterations. A typical simulation procedure is called “Monte Carlo Simulation”. In Monte Carlo, highly uncertain parameters of the model are randomized, and the simulation is run thousands of millions of times. When comparing two designs (for example Election versus sortition), we can then evaluate at which conditions one design is better than another. Moreover, whereas a randomly controlled trial can test one treatment at a time, a computer simulation can test thousands of different treatments. Accurately modeling human behavior is obviously an enormous challenge, yet there may be profound benefits to doing so. Could simulation for example settle questions sortitionistas have on which sortition system is the best?

In the “election voting method” community, advocates such as Jameson Quinn believe that simulation at least may be sufficiently accurate to make relative recommendations of the superiority of one system over another. For example, Quinn’s simulations were used in support of a voting method called STAR voting.

A Call for Recommendations

Clearly much of the world continues to be skeptical of sortition, including in elite academic institutions. Can we change this? Can we generate clear, incontrovertible evidence that sortition is a superior form of government? Can we create an academic consensus that sortition is worth the trouble? Can we generate this evidence at the low prices funders might be willing to pay?

Filed under: Sortition |