Saturday, May 26, 2007


A recent discussion has inspired me to articulate my own design for better government, long simmering in the back of my head and in various scribbles. My focus is on the "action end" of government where disputes are resolved and laws enforced, namely what in modern democratic systems we call the judicial and executive branches. Indeed except for the Articles of Incorporation, which is the constitution of Juristopia, its laws evolve mostly by the common law method of precedent. Subsidiary organizations are free to pass bylaws if they wish, as long as they do not contravene Juristopia's own Articles of Incorporation or disobey an Order of the Extraordinary Court. There is very narrow legislative role for the Board of Franchisors which will be stated below. The main democratic element is juries, which play an important role in limiting the exercise of governmental powers.

To simplify discussions of political power and rebellion, let's posit a system of magic Rings of Power, a la Tolkien. However, there is no master One Ring. (Tolkien surely had the correct answer for such a totalitarian power: it must be destroyed at all costs). Instead these are Rings of Discontempt that operate as follows:

(1) There are N Rings, where N is odd, say 9. Any Ring in the possession of an individual who possesses another Ring has no power (that includes the first Ring possessed -- the second renders the first impotent -- so that there is a strong disincentive to accumulate Rings).

(2) To operate the ring, a quorum of holders of the Rings, called the Court of Extraordinary Justices, touch their Rings together and prouounce the unique name of an individual to be punished for Contempt of Court, and pronounce a punishment.

(3) The quorum of joined Rings can administer the following magical punishments to the named person at any range:
(a) imprisonment in a magical jail for any specified number of time, requiring the aid of no human jailers
(b) instant and painless death

(4) The Rings confer on their wearers complete invulnerability to any kind of violence.

There is no other magic of any sort in Juristopia.

Like the Catholic College of Cardinals, the Extraordinary Justices choose their own successors. Board of Franchsisor members also choose their own successors, but have no Rings. Combined with the magical Rings, self-succession allows the Extraordinary Court to avoid the two main sources of political bias and pressure from other branches of government they now face: from their initial appointments by other branches, and by the ultimate ability of an executive branch to ignore a court order that it despises. An executive officer who refuses to execute a Court order is punished with Contempt of Court; the Rings insure that no rebellion against such punishment is possible.

The Articles of Incorporation state that the Court of Extraordinary Justices exercises only a very narrow jurisdiction, called the Extraordinary Jurisdiction. All Ordinary Jurisdiction, and all armed forces and police powers besides the Rings, are auctioned off to the highest bidder as franchises.
Franchises are property rights to exercise certain narrowly defined governmental powers, subject to limitations on the monopolistic fees or taxes they may charge, determined by juries selected by lottery from the feepayers or taxpayers.

A number of general kinds of franchises are specified by the Articles: public nuisance suppression, fraud suppression, armed forces, police forces, and common law courts are the five main categories of franchises, but there is a vast variety of possible jurisdictional "boundaries" for franchises within each category, and ownership of franchises may be bundled within or across boundaries in a dizzying variety of ways. The franchise market can thus assemble bundles that best persuade juries of feepayers to maximize fees. Each franchise collects its own monopoly fees from all persons within its territory; no franchise may charge fees that are more than what a jury considers "reasonable", including a reasonable profit. The general franchises stated in the Articles can be subdivided and narrowed, but not expanded, by the Board of Franchisors, which can pass bylaws that specify the subject matter and personal jurisdiction and territorial boundaries of each franchise. Some franchises are executive (similar to government agencies), and some are judicial (courts). The Board, like the Court, selects its own successors.

The Articles of Incorporation state that Juristopia Corporation itself does not have shareholders and may not earn a profit. Its only officers are members of the Board and Court, and the salaries of these are limited to salaries deemed "reasonable" by juries chosen by lottery from among Juristopia' taxpayers, as are its overall budget and the fees (taxes) it may charge within the territory of Juristopia. Its officers and employees may legally only engage in the business of Extraordinary Jurisdiction (Court) and writing initial franchise deeds (Board) -- they are are forbidden from engaging in ordinary business or Ordinary Jurisdiction.

All Justices and Board members must be eunuchs (if male) or artificially or naturally postmenstrual and sterile (if female), may own no amount or kind of property that a jury deems "unreasonable", must dispose of any properties that will create an imminent conflict of interest, and must recuse themselves from any legislation or cases that create a conflict of interest. They may not reproduce themselves in any fashion. The purpose of all this is to minimize potential biases, conflicts of interests, and motivations to use the Rings to expand power. (A digital protocol would be ideal if human judgment were not required to render verdicts and define new franchises).

The Articles state that all the jurisdictions and armed powers shall be auctioned off to the highest bidder as franchises. There would, for example, be Army, Navy, and Air Forces franchises empowered to charge a reasonable fee to the entire island for their services. There would be some public nuisance suppression franchises: for example there could be an air pollution franchise, allowing the franchise owner to regulate all pollution in Juristopia, similar to the U.S. EPA today. These franchises might be further unbundled into the particulates franchise, the sulfur dioxide franchise, the carbon dioxide franchise, etc. -- or alternatively the car tailpipe franchise, the electric plant emissions franchise, etc. These franchises would be allowed to auction off tradeable emission rights, or altenatively to charge emission fees. To the extent franchises are monopolies, or coercive, or both, the fees they can charge are limited to what are "reasonable" fees as determined by juries.

The Articles restrict the Extraordinary Court's jurisdiction to appeals in certain narrow (extraordinary) areas: choice of law and forum clauses, cases involving the property rights of franchises, and cases involving violations of a Bill of Rights, which would also be part of of the Articles. The Bill of Rights states a number of individual procedural rights, such as right to trial by juries, the right of juries chosen by a lottery of taxpayers to decide whether any taxes or franchise fees are reasonable, limits on search, seizure, and arrest, and so on that must be respected by any entity exercising a judicial or police power.

All Ordinary Jurisdictions are auctioned off to the highest bidder and can be freely traded. Among the Articles of Incorporation would be a strong requirement that Choice of Forum and Choice of Law clauses in contracts be enforced in a way much broader than current U.S. practice (e.g. allowing all torts of any kind between the two parties to be included in the scope of choice, overriding any franchise jurisdictions).

In the expanded form I propose, these will allow pairs of persons (individual or corporate) to opt out of the franchise that would otherwise have jurisdiction over future cases between them. This introduces a substantial degree of competition between franchises, and one might be able to assemble an AC-style system of insurance/protection companies from such clauses.

Under the Articles there are no Chief Justices, Presidents, or other lead executive officers that have special constitutional powers. Any Commander-in-Chief of the Armed Forces is optional and would be chosen by the various armed frachises themselves, as they may agree. The Court and Board are free to elect nominal leaders for administrative purposes if they wish.

The biggest open problem in Juristopia is how to stop the Extraordinary Court from cleverly interpreting the Articles of Incorporation to usurp the Ordinary Jurisdictions that properly belong to the franchises. Or from just using its control of the Rings to blatantly violate the Articles and enforce martial law and slavery for its own grand benefit and everbody else's pauperization, once enough people have settled on the island with high exit costs. A possible solution might include giving Extraordinary Juries, selected by lottery for each decision, their own set of Contempt of Jury Rings to enforce jury restrictions on the Court and Board.
Publish Post
I have also neglected powers to make and enforce treaties with other governments. Since such treaties typical assume external sovereignty, the very low degree of internal sovereignty in Juristopia, which is otherwise a very good feature, may pose an interesting problem for treaties.

(N.B.: I crudely described a previous version of Juristopia here).

UPDATE: spelling change.


Mencius Moldbug said...


You're solving a different problem than me, but this does not render our solutions incomparable, as my One Ring could certainly be used to enforce your system of Nine - just as in Tolkien :-)

A working solution to the problem, of course, would require no rings at all. Real force of some sort must be substituted for the rings. This awaits further discussion, and presumably it is not much harder for your design than for mine.

Of course I admire the multiple layers of jurisdiction (franchises). I am not sure the word "reasonable" is really a load-bearing material, so to speak. There is much to be said for juries chosen by lottery, however, although for such a method I think 12 is too small a number (having served on a couple of juries myself).

Juritopia has to be considered an attempt at solving the ancient ipsos custodes problem. As you say, it depends - not unlike its US relative - on a supreme authority which guards all other watchmen. (Although today's US answer might best be expressed as "The New York Times will watch the watchmen.")

As such it falls into the general class of one-authority solutions. That is, it assumes one benign and disinterested authority. It does not rely, like Montesquieu, on a balance of competing authorities. I think this is wise, because it is fairly clear that balance-of-power designs don't work - an attacker just takes over all the competing authorities, and you end up with some kind of a Polygon.

The main vulnerability of all such systems is that they tend to be captured by movements which are based on (post forthcoming) what I call adaptive misconceptions.

An adaptive misconception is a belief about reality that is false, but allows its believers to conspire maliciously in a way that would be benign cooperation if it were true.

For example, the belief that planned economies are more efficient than "chaotic" laissez-faire, at one time widely believed, was an adaptive misconception that enabled Communism to take over half the world.

Communism is best seen, I think, as a branch of the Progressive movement, which is best seen as a sect of Christianity. (See, eg, Rothbard's "Power and the Idealists.") It appears to me that your system would be vulnerable to this sort of attack, which I think would manifest itself in Extraordinary Justices appointing successors who were progressively more Progressive.

Daniel A. Nagy said...

1. What prevents the collectors of monopoly fees to bribe those who would impose limits on them?
2. What is the benefit of granting monopoly for judicial services? Why not let people decide unto whose ruling they are willing to submit?
3. Suppose, nine people with all the right qualifications, sharing your vision find these nine rings (which immediately have power over all humans) and some additional treasure trove that lets them survive long enough to get used to the fact that they are not subject to starvation and other problems associated with the lack of shelter and nutrients because of the rings' protection against violence. Is that enough for Juritopia to emerge and take over the world? Could you give us a sketch of the process?

Anonymous said...


"Reasonable" is basically just legal-speak for "let the jury decide the issue."

Having a jury cap the fees or taxes charged by a franchise comes straight from real monarcho-franchise law. It's a similar idea as requiring tax and budget legislation to come from Commons or Congress, except repackaged -- each different administrative function has its own "Congress." Two things prevent a jury from extorting rule-making power in exchange for greater fees (this is how Parliament achieved its overweaning power) -- (1) it is temporary and chosen by lottery, so this year's jurors have no incentive to empower next year's jurors, and (2) even if they tried, the scope of rule-making they can influence is limited by the property deed of their particular franchise.

There is a different fee structure and jury for franchise, and each franchise corresponds to a small subsection of one of today's administrative agencies -- one for auto tailpipe emissions, another for smokestacks, etc.

Actually the overall structure is rather Montesiquean: Board (legislative), Extraordinary Court (judicial), and the executive dispersed among the franchises. I've strengthened the ultimate authority of the judicial branch in severe disputes, but I argue that due to the much narrowed jurisdiction of the Court and the dispersed nature of the executive, they are still in fair balance with my weakened legislative branch (which may only draft the property deeds for new franchises, which is actually not all that different from what Congress does today when it sets up an administrative agency -- but I've severely seperated legislative power from the tax power).

I suspect the "Byzantine agreement" work in computer science can be used to prove that no possible form of government or any other configuration of police powers can be made immune from your adaptive misconceptions problem. And Godel probably showed that you can't set things up to make all substantial misconceptions locally maladaptive. But I'd love to be proven wrong.


(1) Fee or tax collection is handled in the same way as collection of any other debt -- there is no tax collection agency, witholding is voluntary just as with today's "autopay" options from your checking account, and if you don't pay on time it goes on your credit record and you could be sued. There will probably be commercial aggregators to prevent citizens from receiving hundreds bills from the hundreds of franchises.

Choosing by lottery makes it too hard for supporters to bribe politicians before they're elected, as often occurs in modern governments. Athens had a similar system for electing public officials, and Venice had an ingenious system of phased vote-lottery-vote-lottery-etc. system that resembled the confusion/diffusion phases of a cipher. The result was to combine majority rule with the bribe-reducing unpredictability of a lottery election.

Another aspect is that juries only set fees -- the franchises make the actual regulations within the scope of their deeds. So bribing juries is not the route to effect regulations, although bribing franchises might be. If juries set the fees high enough and the Board drafts the franchise deed carefully, I am hopeful franchises can be disincentivized from taking bribes. Also a crucial part of every deed -- we can make this an Article -- is that franchise and their employees may take their revenue from jury-approved fees and in no other way. Their bank accounts are closely monitored.

(2) There must be default jurisdictions for the same reason there is a background to every contract of default rules and a state of property. We can't expect everybody to necessarily contract with everybody else (even indirectly, i.e. a clause "We choose the same forum and law as the other cuustomers of X") and contracts are often sadly incomplete (Godel strikes again). So there have to be default terms and that includes default jurisdictions. If it turns out that the franchises oppress their default subjects and it becomes easy to switch via indirect choice clauses, AC might well emerge. (If most people have switched out juries will tend to defund the franchises). It's hard to imagine crafting mass agreements for large minorities to opt out certain rules, such as global pollution rules, but it would be fascinating exercise to figure out how to draft such contracts (a hard task that most AC proponents sweep under the rug). There's not much AFAIK in the Articles to discourage such an AC outcome if it is viable, but the franchise system is at least necessary as a transition and as a fallback in areas where AC doesn't work, for example (I suspect) global pollution like carbon dioxide.

(3) I'll have to punt -- I haven't thought much about how to start out on Mencius' island. To the extent I think of transitions it is from actual systems -- constitutional amendments and the like.

Anonymous said...

BTW, one flaw I suspect might occur is that each jury, not constrained by having to fit its franchise's budget into an overall budget, and influenced by it's franchise's PR to believe that its franchise solves The Most Urgent and Important Problem on Our Planet -- each jury will tend to overfund its particular franchise, since it represents only a small fraction of what the juror will pay in taxes.

Suggestions for solving this problem? I have some vague notion that there will be a "defense" at a "trial" arguing to the jury that the franchise's fees should be lowered, but I'm not sure who would be motivated to fund such a defense -- perhaps some of the companies regulated by the franchise. I don't want getting fee raises to be as easy for a franchise as getting a grand jury to indict the proverbial ham sandwich.

Mencius Moldbug said...

The idea of drawing any kind of analogy between algorithmic CS (such as Byzantine agreement) and politics terrifies me. An election is not a heapsort. The thing that happens when one of your replicated servers goes down is not democracy.

I am much more comfortable with fuzzy examples from what one might call the artistic wing of CS. For example, I'd argue that there is an analogy between the fact that the US has one system responsible for "national security" and another for "homeland security," and the fact that OS X has something like 27 systems responsible for interprocess communication, although it is descended from Unix which had precisely one.

Anyway, as an attacker working against this system, the juries are definitely the weakest point and I would concentrate on them, working up to the judges.

What you need is a feedback loop wherein some concept of "social justice," which is of course not the same thing as formalist justice, can attract supporters to itself by promising them power, money or both.

This would presumably be run through the newspapers, or other points of information control. Call it the extended educational system. I notice that you have not specified an information management architecture in this design. Since all benevolent systems of government depend ultimately on public opinion, one will create itself, and it is likely to attack the system. And win.

So we'll use this to generate a progressive mood in the population, a feeling that, perhaps due to some vague historical wrong, juries and judges should favor some groups of litigants over others. Progressives will take over the educational system, which is necessarily responsible for training both judges and juries.

By my count this is game over. Your judicial system will be the new legislative branch of Juristopia, which will in practice be run by its information system, just as in the Fourth Republic. Franchises will evolve into a stuffy, orthodox irrelevancy, and the result will be unlimited administrative government, as today.

Anonymous said...


I expect most of the PR that currently goes into the press to voters will instead be lots of evidence presented to the jury by the defense (the franchise).

BTW, I had some doubts about this before, but I now realize this is a real trial, or at least a strong legal fiction thereof. Any taxpayer has standing to sue to reduce or eliminate his taxes, and any other taxpayer can join -- it will generally become a class action lawsuit. Then the jury decides on the proper level of taxes for this agency.

Mencius Moldbug said...


I think you may be underestimating the power of information systems to organize collective action.

There's a reason our newspapers, schools, universities, etc, etc, tend to be dominated by collectivists. It is not a coincidence. It is a vicious cycle, a feedback loop, in which misguided or unprincipled collectivists, because they can capture the power of the State, regularly defeat their libertarian adversaries, whose principles force them to abjure it.

Juridical safeguards were supposed to keep this from happening in the US. They didn't. I think it is unwise to devise a replacement structure without at least some theory that explains this rather dramatic engineering failure.

Anonymous said...


Your argument against my franchises is also an argument against your own proposal. There is nothing preventing my franchises from making profits, issuing shares, and paying dividends, and I expect they often will. They might also be sole proprietorships.

The big differences are that I have two mechanisms that can put a cap on our for-profit coercers -- juries and constitutionally enforced legal competition (a contractual opt-out for subgroups among themselves in many areas of the law). You have none.

Your proposal is far more dangerous in terms of information systems, as the same organization that profits from high taxes also can legally control the press and education. My franchises are very restricted in their scope and can do no such thing. Nonterritorial decentralization -- i.e. unbundled property rights in legal subject matter -- is a very powerful idea, and that it solves your information system problem, whereas no other proposal has ever solved it, is a very good example of that power.

Mencius Moldbug said...


Sure, but the fact that your franchises are restricted creates a combination of conflict and uncertainty - jury results are anything but certain. The result is that people will scheme to evade the restrictions and return taxation to its Laffer maximum.

Since I already set taxation at the Laffer maximum, I have no need for any such restrictions. The whole system is at neutral pressure, so to speak. So there is no need to construct cunning conspiracies.

Now, that said, I think we agree on the merits of decentralization. I am just more concerned about providing economic motivations, as opposed to traditional motivations (respect for law is a wonderful tradition, but a tradition nonetheless), for making the resulting system stable in the long run.

Anonymous said...


I don't know what you mean by "people will scheme." Can you be specific? Who specifically will so scheme, and why?

The separations of powers and checks and balances incentivize people to respect the law -- as opposed to arbitrary power which incentivizes the power holder to disrespect it. I'm certainly not relying primarily on traditional respect for the law -- rather I am setting up a structure where maintaining and even reviving traditional respect for the law is the best strategy for all involved.

You finally admit that your government would set taxes at the Laffer maximum. Quite a libertarian outcome, to maximize government revenues! BTW, I'd bet that the vast majority of economists would say today, if you asked them, that most U.S. taxes are set well below the Laffer maximum revenue point. 20th century democracy is rather tax-heavy compared to some earlier eras (e.g. the 19th century), but far worse, including what you now admit you are proposing, is quite possible. The Laffer maximum has been reached only at the most outrageous of tax rates, e.g. the 70-90% tax rates at the top brackets in the U.S. during the middle of the last century. The Laffer maximum is about the same as what slave cost his slave owner in ongoing costs, i.e. about 30-40% of his output.

Mencius Moldbug said...

Intellectuals will scheme, because intellectuals (like everyone) like power, and are also very good at inventing pretexts that make them out to be concerned for the benefit of humanity, rather than scheming for power. See, again, the Progressive Movement and its present-day successors.

I'm surprised by your claim - I think most economists today would say reducing marginal top rates (note: marginal, not flat) was a Laffer optimization.

The reason the Laffer maximum involves low taxes is simple: growth is exponential. As long as you have high time preference, as monarchies certainly did, anything that reduced the exponent was a killer.

This is why monarchical taxation was much lower than democratic taxation. To say nothing of the various non-revenue-producing regulatory deadweights that democratic states have created.

Once again, the proprietor of a nation-sized patch of land profits from his tenants' prosperity. Interests are aligned. I am really not sure why this concept is so difficult.

I am a libertarian. But I think of libertarianism as an outcome, not as an intention. And I notice that the product of libertarian intentions has been a gigantic state that Louis XIV could only have dreamed of. I therefore suspect that there is some kind of fundamental engineering error in libertarian designs.

It might help me understand your argument if you gave your interpretation of why the American design, originally quite libertarian (Rothbard's Conceived in Liberty is a wonderful read) failed so utterly and disastrously.

Anonymous said...

Why the American design failed? Well, for starters let's get some perspective. Failed compared to what? The American design failed more slowly and to less an extent then most of the rest of the world. Only your little trading post countries, the liberty of which is well explained by their rare economic niche of attracting global traders, and thus needing to maintain low exit costs, have managed to beat the U.S. in liberty during the 20th century.

As for the general explanation of why the Constitution couldn't better constrain the growth of power: high exit costs. Once the American frontier filled up, and every patch of territory became a sovereign State ruled over by a vast national government, there was no place for the oppressed left to flee to. Social networks coalesced, government became less distant, and exit costs grew. The conditions so favorable to liberty in the American colonies Smith described so well (you really do need to read Chapter 7 of Smith, who was there, instead of a narrow ideologue like Rothbard who wasn't) -- cheap land and distance -- have both largely disappeared. It's no t much of a coincidence that the Jingo and Reform Burgess complained of sprang up in America during the same decades that last of the lower 48 were achieving Statehood.

Therer are some secondary explanations too, such as the rising efficiency of civil bureaucracy due to mass literacy, and the rise of modern corporate accounting which made the income tax feasible to implement.

At the same time the U.S. imorted ideologies, like Progressivism (which is basically just a sanitized compromise between communism and the free market), from places that went down the tubes farther and faster than the U.S., like Europe. And they went down farther in faster in large part because of their Justinian Code tradition of totalitarianism, which America has been busy importing during the 20th century, but at a rate well behind most of the rest of the world (and we've even managed to export a few counter-memes, like modern markets, freedom of speech, and republican controls).

I am very glad indeed that we had the U.S. Constitution with its checks and balances during the 20th century rather than the kind totalitarian system, derived from the Caeasars and Kaisars and Csars, that led to the vast domestic oppressions and genocides of fascism and communism. Extremely glad. I cannot imagine any sane American who lived through the middle of the 20th century who would now look back and say they would have prefered to live under Hitler or Mussolini or the many communist regimes. Would you? The U.S. has many choice targets for your withering critique, and I as well as you would love to make our politics better, but you should seriously put your critiques and notions of political failure into some perspective.

Anonymous said...

That should be "the kind _of_ totalitarian system", although you can read it without the "of" for irony if you like... :-)