« Less Wrong: Progress Report | Main | Prestige Matters More For Smarts »

April 26, 2009


I'll start with the obvious: The institution of scientific academia and scientific journals is often used to track, experimentally, the results of imposing particular institutions - albeit deliberately controlled studies are rare.

Patri Friedman's Seasteading seems to be getting at this point.

1. What institutions are especially good as meta-institutions?

The market-like institution of institutional competition between jurisdictions.

Institutional competition deals especially well with the limitations of human knowledge and cognitive capabilities. It acts as a discovery procedure for better and more suitable institutions and weeds out unworkable ones.

Also, it can serve to constrain the power of the necessary but imperfect institution of the state, because it gives citizens alternatives to choose from (possibilities to invest, live or produce under different legal settings)

Lastly, it serves especially well as a motivating device for politicians, who choose institutions in their country, because they don'T want to loose investment or economic activity to other jurisdictions.

Note that institutional competition is, like the market, a process framed by rules and not "just something that happens" or plain anarchy. Like the market, the process must be framed by appropriate institutions to yield results in the common interest of the participants. These would be negative abstract rules that forbid certain undesirable actions in the process of institutional competition.

2. What institutions should we use to evaluate meta-institutions?

To evaluate their working properties, probably economic and social science.

However, due to the limitations of the human mind and the complexity of social processes, decentralized experiments with institutions in the process of instituional competition yield the most reliable knowledge about their actual working properties under real circumstances.

To evaluate their "goodness" or legitimacy, movements of people or movements of resources towards other jurisdictions. For this, people need not be informed about the working properties of the specific institutions, and need decide only by observing outcomes of different jurisdictions. The observed outcomes of individual market-like choice between institutions are a much better indicator on the question whether an institution corresponds to the preferences of the people, than the guesstimates of economic scientist about these preferences.

Eliezer, yes, academia is the institution now that most often explicitly and verbally evaluates other institutions. But is it the best, or just the loudest?

TGGP, how do I use seasteading to evaluate other institutions, e.g., academic peer review.

bbb, I expected to hear such a party line; I was hoping to hear better arguments.

We may have different opinions on what constitutes a "good" argument.

I thought you meant arguments concerning institutions that had some potential to remedy the worst problems of our existing intitutions of collective choice, and that are relevant because there is still much room for improvement.

RH: ".. how do I use seasteading to evaluate other institutions, e.g., academic peer review."

Mike Gibson responds at Seasteading's official blog. From his response,

Two very difficult questions to answer with respect to governance. Robust transnational mobility with low transaction costs would serve as a good meta-institution. Alas, these are but inchoate longings. As for question two, Hanson’s preferred meta-institution, the prediction market, excels at aggregating information and delivering judgments. Are there any prediction market contracts out on the feasibility of seasteading or competitive government? I can only find one market on Ephemerisle.

In democracies, voting is the overriding meta-institution. All other institutions exist only on the tolerance of the voters. This is why we don't have Idea Futures markets yet - voters are not convinced they would be a good idea.

General weaknesses of voting have been extensively analyzed. Specifically as a meta-institution, voting seems to be especially bad at evaluating hypothetical institutions. Once an institution exists and people have experience with it, voting is a more reliable guide to whether that institution's effects are positive or negative. If people find markets produce too much inequality or instability, they will limit markets. To the degree that academia is seen as producing social value, it will be funded and supported. Voting seems to provide good feedback for existing institutions.

In terms of evaluating itself, voters are curiously eager to put limits on the voting process. They vote to support constitutions whose main purpose is to prevent future voters from enacting certain kinds of policies. They vote to enact complicated systems of indirect representation rather than relying on the wisdom of the electorate more directly. They vote for term limits and constraints on future spending decisions, often seeking to make it difficult to reverse these votes.

In a few historical cases, voters have even more or less knowingly voted in leaders who will shut down democracy and take on autocratic powers. However these cases are the exception, and for the most part voters are eager to retain the fundamentals of democracy.

The biggest problem I see with voting as a meta institution is the difficulty of getting voters to seriously consider the potential and problems of new institutions. However it's possible that this conservatism is wise, that voters have learned that most proposals don't work out as well as the proponents claim, and that voters are justified in their reluctance to endorse new institutions and in erecting substantial barriers to acceptance.

Hal, I'm less convinced about how well voters evaluate existing institutions, but I agree it is curious how much voters say to restrict themselves.

Zac, I still don't see how transnational mobility could evaluate peer review. Are we to presume that nations which gain more members have better forms of peer review?

The comments to this entry are closed.

Less Wrong (sister site)

May 2009

Sun Mon Tue Wed Thu Fri Sat
          1 2
3 4 5 6 7 8 9
10 11 12 13 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30