A Journal for Western Man

 

A Review of Robert Murphy's

Chaos Theory: Two Essays on Market Anarchy

Dr. Charles N. Steele

Issue XCI- February 25, 2007

-----------------------------------

Principal Index

-----------------------------------

Old Superstructure

-----------------------------------

Old Master Index

-----------------------------------

Contributors

-----------------------------------

The Rational Business Journal

-----------------------------------

Forum

-----------------------------------

Yahoo! Group

-----------------------------------

Gallery of Rational Art

-----------------------------------

Online Store

-----------------------------------

Henry Ford Award

-----------------------------------

Johannes Gutenberg Award

-----------------------------------

CMFF: Fight Death

-----------------------------------

Eden against the Colossus

-----------------------------------

A Rational Cosmology

-----------------------------------

Links

-----------------------------------

Mr. Stolyarov's Articles on Helium.com

-----------------------------------

Submit/Contact

-----------------------------------

Statement of Policy

-----------------------------------

 

Review of Chaos Theory: Two Essays on Market Anarchy, by Robert Murphy. RJ Communications, 2002, 58 pages.  

 

Anarchism is one of the extreme ends of the spectrum of answers to the question “what should be the role of the state?”  “There is no role,” answers anarchism.  Totalitarianism, the other extreme, replies “everything!”  If the objective is to develop a sustainable society that generates well-being for its members, there is no case to be made for the for totalitarian prescription.  Totalitarian systems inherently incorporate a number of fatal flaws: inability to overcome the calculation/knowledge problem is one, and extreme vulnerability to “public choice” type problems is another.  In totalitarian system, planners have no way of determining what choices would be sustainable and “maximize welfare,” nor, given their total power, is there any way of constraining them to seek this objective in the first place.

 

Debates over the anarchist prescription are fought over different grounds.  Since the state entails an important set of constraints that arguably are important for maintaining a harmonious social order, one important challenge to anarchism is to explain how order might be generated and maintained in the absence of fundamental institutions (e.g. a legal system with police and court procedures) to constrain improper, opportunistic behavior that turns social relations from positive sum to zero sum.  Left anarchism (or utopian anarchism) essentially holds that individuals’ anti-social tendencies are a product of bad institutions; abolish these institutions, and human nature will change (e.g. William Godwin).  Or, alternatively, change individuals’ consciousness, say through education, and the institutions that comprise the state will become superfluous.  There was probably never any reason to believe the “perfectibility of man” argument, but certainly research in neuropsychology refutes it, and studies of primate societies show that murder, theft, and abusive power relationships can be pervasive even in societies never polluted by the “unnatural” institutions of the state.

 

The “right” or anarcho-capitalist alternative does not argue for the perfectibility of man.  Instead, it argues that free market contractual institutions can supplant the state.  The rejoinder is “in the absence of a state to enforce contracts, how can the market, or at least anything more than the simplest forms of market exchange, exist?”  Hence at least as much anarcho-capitalist ink is poured out in arguing for the feasibility of the system, as opposed to its desirability.  Robert Murphy’s Chaos Theory explicitly deals almost exclusively with this feasibility issue, and proposes what I think is a somewhat original solution to the question.  Unfortunately, his solution is fatally flawed in several dimensions.  As I’ll show, the flaws are not minor; they are fundamental.  The version of anarcho-capitalism given in Chaos Theory is internally inconsistent, and very much at odds with what we know of how the world really works.  As such, this work completely fails to explain how anarcho-capitalism could serve as a replacement for the state.

 

My review proceeds in ten numbered sections, each addressing a separate issue.  Before proceeding, I suppose a disclaimer is in order.  I give a strong “two thumbs down” to Chaos Theory, and argue that it makes egregious errors on simple and basic economic issues.  I don’t want to be misunderstood as attacking anarcho-capitalism, or libertarianism, or Dr. Murphy himself.  In regards to this last, I hasten to point out that the professorship I currently occupy was previously held  by Dr. Murphy.  While I don’t know Dr. Murphy, his students speak highly of him, and do so convincingly, which is good enough for me.  This review should be taken for exactly what it is, and nothing more: a critique of what I think is a very bad argument for anarcho-capitalism.

 

Section 1: Contractual Basis of Society

 

Dr. Murphy begins by proposing a society that is conducted entirely on the basis of contract.  (Call this society MA, Murphy Anarchism, for ease of exposition.)  Fair enough; this is an idea that seems inherent to anarcho-capitalist thought, although it is not exclusively anarcho-capitalist, since a state may have a contractual basis as well.  But under MA, all human relations should have a contractual basis, and not simply an implicit one.  In MA, an explicit formal contractual agreement covers every human interaction, at least between adults.  MA private contracts cover every contingency, since they replace “government” law.  These purely voluntary contracts assign liability for the results of every possible action that individuals might take.  For example (one of Murphy’s own examples), when a moviegoer buys a cinema ticket, the ticket is a contract specifying every possible event, including what happens if the moviegoer commits a crime while attending…and what is and isn’t a crime is included in the contract as well.  This needn’t all be spelled out in fine print on the ticket, which might simply read something like “this contract is governed by standard contract provisions A.”  But it is all spelled out in “standard contract provisions A.”

 

Why this web of contracts, covering everything, and not instead some universal law, such as the “non-aggression principle?”  Murphy doesn’t explicitly say, but I gather that his reason is to avoid the possibility that any member of society be coerced into accepting something, even if it is the non-aggression principle.  I think this is something original, since, if it could be implemented, every single member of society would have agreed to all of the rules under which s/he subsequently lives.  However, Murphy’s proposal fails for two important reasons.

 

First, even a passing familiarity with the economics of contracting reveals that this objective is literally impossible.  Contracts are always incomplete, because contracting is costly.  Murphy posits a complete set of private contracts to replace government law.  But in the real world, such a set of contracts would be prohibitively costly (i.e., impossible) to develop and enforce.  This is different from the question of whether a new contract would have to be drawn from scratch for each transaction.  Murphy correctly points out that standard contracts could be used (as on the movie ticket referring to “standard contract A”), but even standard contracts will be incomplete.  So, since both private law and government law must be imperfect and incomplete, this raises a fundamental question that Murphy entirely ignores: might not general laws, established by government, in some cases and spheres, entail more and better coverage, and at lower costs, than individual contracting?  If the government defines, say, murder, assault, rape, and robbery, and simply prohibits them, this saves on the costs of individuals having to do it themselves, or hiring firms to do it piecemeal.  And use of a “reasonable man standard” or some other method of applying the law to unforeseen contingencies can overcome the necessary incompleteness of contracts.  It’s hardly far-fetched to think that a one-size-fits-all contract on these issues, established and imposed by government, might easily generate lower transaction costs.  (In fact, this is probably a major reason why we don’t see stateless societies.)  Hence Murphy’s claim that his MA system is more efficient than a system with a state is not established… and is extremely doubtful (see point 10 below).

 

Second, what if some individual simply refuses to sign such contracts?  Since s/he has never agreed to anything, is s/he scot-free to do as s/he wishes, or do we have to apply some standard, e.g., the non-aggression principle, to tell us what we can and can’t do to curb this individual?  If the former, Murphy’s system is indeed chaos, and if the latter, then we need to explicitly enshrine the non-aggression principle as the law of the land and drop the fiction that only those laws apply to an individual that the individual has explicitly agreed to.

 

Section 2: The Incentive to Seek Better Contracts

 

Suppose, for the sake of argument, that my above objections do not hold, and the MA world is established: all human relations are governed by voluntary contracts.  Murphy argues that individuals will have an incentive to seek out better contractual arrangements and avoid worse ones, and hence lead to an improvement in law.  In fact, even in the real world it’s impossibly costly for individuals to do this, and in the real world people sign all sorts of contracts without having the faintest idea what they entail.  If I want to buy a movie ticket from a theatre I don’t usually visit, is it really likely that I want to spend time researching provisions of standard contracts B or C that, to my surprise, are part of the ticket?  Murphy no doubt holds multiple credit cards from different firms: I challenge him to tell me the extent to which each contract provides for private arbitration of disputes, what arbitration companies are involved, in what states, what the arbitration rules of those companies are, the applicable laws of the states, what the companies’ reputations are, and all the other things he says people will have incentive to learn.  In the MA world, there’s even more scope for these things to differ, and the costs to an individual of acquiring familiarity potentially skyrocket.  My point isn’t that markets can’t work in contract selection, but rather that Murphy ignores contracting and information costs as limits to market transactions.  This simply sidesteps the very real possibility that government law might sometimes be the low-cost solution, and enhance both efficiency and freedom.

 

Section 3: Contracts and Incentives to Behave Properly

 

MA contracts work by imposing liability for actions.  If each person is fully responsible for all consequences of her/his actions, then s/he will have the right incentives for civilized behavior.  But strictly speaking, this argument is false.  Suicide bombers, for example, aren’t generally deterred by financial liability for their actions.  Common criminals tend to have very short time horizons, and hence tend to behave more impulsively.  These are the people we need to deter; most people are already well-behaved.  It’s true that the world would be better if everyone bore more responsibility for their actions, and that at the margin people would behave better; it’s not true that this would substitute for force.  Imposing liability on criminals will hardly eliminate crimes such as murder.  Some murders will still occur.  But under MA, the murder victim’s sole remedy is that his estate can demand $y millions in restitution from the killer – small comfort if the killer is broke, in which case s/he’s off scot-free.

 

Section 4: The Insurance Solution

 

Murphy counters the last objection in section 3 by arguing that since few could afford to pay the $y millions, everyone will carry an insurance policy that will cover the restitution, should they commit a murder or other crime.  Anyone who didn’t carry such a policy would be ostracized.  This solution is utterly unworkable.  Never mind the idea of having to show your proof of insurance before a store owner lets you enter her china shop (Murphy’s example), or at the start of a date, or eating at a restaurant, or listening to (or delivering) a university lecture.  While these all seem bizarre to us, perhaps they are not fundamental objections; I suppose we could all get used to this annoyance.  Sometimes I suspect it might even be a good idea.  But much more importantly, no sane entrepreneur would offer the sort of insurance Murphy endorses – insurance against the consequences of one’s own intentional behavior.  This is the best (worst) example of moral hazard in insurance imaginable.  If a person decided s/he wanted to commit a murder, all s/he’d need would be a cheap insurance contract.  Murphy assures us this would be inexpensive stuff, since the odds of any member of society committing a murder are small.  For a mere $10, one could kill with impunity.  No, not impunity; one’s future insurance costs would go up.  “Take that, Charles Manson!  No more cheap insurance for you!  And you have to buy insurance, or they won’t let you into McDonald’s!” 

 

Furthermore, Murphy appears not to realize that this insurance destroys the liability he originally proposes as a shaper of incentives.  Again, if an insurance company issues a policy that indemnifies me against consequences of my house burning down, it had better make certain that the contract excludes indemnities if I intentionally burn my own house down.  Otherwise it has created incentives for me to burn my house!

 

What Murphy is proposing really is not insurance, but bonding.  Insurance covers unavoidable risk, not the obvious consequences of one’s own chosen actions.  However, in some limited circumstances, bonding can function in the way that Murphy proposes.  In this context, a bond is a sort of hostage one offers to be forfeited in the event of bad behavior.  Familiar examples are the bail bond, or the bond posted by a bonded locksmith.  For these bonds to work, they must be substantial, so as to signal that the individual has strong incentive to behave properly.  Hence, bonding does not solve the problem at all; I would need to post a bond approaching $y million dollars to “insure” myself against committing murder.

 

(An aside: somewhat ironically, Murphy laments that criminals on parole often commit crimes, and faults ivory tower intellectuals for this.  But many crimes are also committed by accused criminals out on bond; he doesn’t realize that his own armchair system would generate the same phenomenon – pay the insurance company enough to make it worth their while, and they’ll let you go.  Maybe he’d argue insurance companies would require higher bonds than government does, but using Murphy’s own argument, they might well not.  See point 9 below.)

 

These four points are sufficient to show that MA is fatally flawed.  1. The web of contracts is impossible.  2. Even if they were possible, the details would be so complex that no individual would know what s/he he had agreed to.  3. Liability by itself is insufficient deterrence for criminal behavior.  4. The insurance scheme is utterly unworkable.  This means the core of MA is completely unworkable as well.

 

However, there are further points that are somewhat unique to MA.  While I find these interesting, they also are badly flawed.

 

Section 5: Insurance Company Incentives and Pre-Empting Criminal Behavior

 

Murphy repeatedly argues that criminals (one of his examples is an ax murderer) will carry personal insurance to pay for the consequences of their crimes.  And their insurance carriers will have an interest in keeping these clients isolated (in prison) since otherwise the clients will commit further mayhem and the insurers will have to pay more indemnities.  Hence criminals will be kept off the streets in MA.  None of this makes any sense at all.  Why would the ax murderer carry this insurance?  And who would want to insure him?  As I’ve already argued, this insurance makes no sense, but suppose for the sake of argument it existed.  If you were an insurer and had insured someone and so  must pay when he commits a crime, and he proved to be an ax murderer, what do you do?  Pay the indemnity and cancel his policy!  If not, at the very least, his premiums would skyrocket to cover the costs of his incarceration, and he’d not be free to go to work, so how would he pay for this?  The argument is absurd, all the more so when Murphy claims that in such prisons the guards would have to treat prisoners well, or else the prisoners would change their insurance to a different company that keeps nicer prisons.  “Hello, GEICO?  I’m an ax murderer, and I need insurance against the consequences of my homicidal tendencies.  And I’m very disappointed with my current insurer, State Farm.”

 

Section 6: Big Brother is Watching

 

Under MA, the ubiquitous insurance agencies must be collecting enormous amounts of information on their clients – almost everything about their behavior, including the number and kinds of firearms they own (Murphy is quite explicit on this).  Even the (U.S.) federal government doesn’t currently collect this information!  Has Murphy not heard of identity theft or other privacy concerns?  Murphy would likely respond that insurance companies can’t collect information that individuals don’t voluntarily relinquish – but that’s not the point.  The point is twofold: insurance data collection would be very costly to insurers, and also very costly (including risk) to the would-be insured… and hence there’d be much less insurance than Murphy suggests…presuming that the insurance could work in the first place, which it cannot.

 

The thing that I found surprising in contemplating Murphy’s arguments is that it brought home to me, an anarcho-capitalist, just how difficult is the problem that government solves: it’s very possible that a one-size-fits-all set of rules enforced by a monopoly (government) might be a lower-cost, more feasible method of providing defensive services.  The question of how to properly constrain government remains, of course.

 

All of the above is from Murphy’s first essay.  The second essay addresses the problem of how defensive services, primarily against foreign invasion, might be provided under MA.  The following points are from both essays, and I place them together as they all deal with, in part, claims about the nature of government provision of services.

 

Section 7: Would Insurers Man the Trenches?

 

Murphy argues that in the event of a foreign invasion, insurance companies would have incentive to organize defense so as to avoid paying indemnities to its insured in the event that the invader harms them.  This is a very different kind of insurance than he proposed in the first essay, and is, in fact, real insurance – people purchase insurance to be protected from risks, in this case from foreign attack.  Exactly why people would tend to buy such insurance is unclear, for the standard reason that they could free-ride on one another (or on the defense the insurance companies will provide).  Free-riding on public goods is a fundamental problem for proponents of a stateless system to address, but Murphy essentially just assumes that people will purchase such insurance, and asserts that the public goods nature of defense is exaggerated.

 

If we grant, for the sake of argument, that the free-rider problem is somehow solved, the question arises: why would it be cost-effective for insurers to fight against an invader?  War is expensive, and if the insurers win, they’ll still have to pay some indemnities if the invaders inflict damage.  But if they lose, it’s questionable whether they’ll pay any indemnities at all.  If the invaders are the equivalent of the Nazis (Murphy’s example), when they take over, the old contracts will presumably not be enforced.  Perhaps the insurers will find their assets seized by the conquerors.  Insurers would thus have an incentive to simply abandon the MA society once it is attacked, taking the pooled funds with them.  While insurers might fight, the profit motive is a weak hook to hang one’s defense on, since the more profitable thing for the insurer could easily be to flee.

 

For this reason, it is better to trust one’s defense to an organization motivated by something more than pecuniary profit, whether this be love of justice, or homeland, or pride, or duty, etc. – to depend on individuals who, for whatever reason, are willing to fight at all costs, to their own deaths if need be.  If I get to choose who defends me, I’d rather have a bunch of crazy Texans in an Alamo or a General McAuliffe at Bastogne, rather than someone who checks with the accountants to see whether firing the next shot at the enemy or at me would bring more profit.  While this doesn’t show the superiority of the state to an stateless society, it casts tremendous doubt on Murphy’s contention that private companies, motivated purely by profit, would provide better defense than alternative institutions.

 

Section 8: Would Market-Based Defense Beat a Centrally Planned Offense?

 

Murphy goes well beyond the point that private firms would provide better defense.  He also argues they’d conduct defense on the basis of market transactions.  Here he again loses all sight of transaction costs, in this case with respect to their influence on organizational structure. As an example, Murphy argues that at Stalingrad the Red Army had a harder time of it than necessary because its actions were centrally planned.  They’d have defeated the Germans more easily if they’d organized their defense on a market basis.  Private insurers, interested in avoiding paying indemnities to clients victimized by the Nazis, would have offered bounties for the destruction of German targets.  Profit-seeking entrepreneurs would have competed to earn these bounties.  Some would have organized armored units, others infantry units, and still others mortar batteries (Murphy’s examples) and then they’d have competed with each other to try to find better ways to defeat the Germans and earn bounties.  In this way, the market, not a central planner, would decide how to fight the war. 

 

I hardly know what to say about this bizarre contention.  It ignores the obvious economies of scale involved in conventional military conflict, and the obvious benefits of large numbers of actions being coordinated ex ante.  Having a central commander directing a single large organization can be appropriate in such situations.  Any account of battle reveals the importance of such top-down coordination.  As Patton put it, a mediocre plan, executed immediately and boldly, beats a lot of dithering that eventually generates an optimal plan.  And this competition among small uncoordinated military units, guided only by the “invisible hand” of profit, sounds like a lot of dithering.  But Murphy isn’t even true to his own argument: why should entrepreneurs organize and centrally direct their small units?  This is just central planning on a smaller scale.  To be consistent, Murphy should argue that each soldier be a completely independent private contractor, ready to act on her/his individual initiative and own knowledge of particular circumstances of time and place.  Hierarchical military organization is always central planning, and markets always outperform central planning, so a collection of unorganized independent mercenaries will always do better than the same group would if they had a single commanding officer directing and coordinating their actions…right?  Clearly not.  For certain purposes, hierarchical control outperforms individual market transactions.  If it weren’t so, there’d be no reason for firms to exist.

 

The problem with Murphy’s analysis is that he’s taken a valid point with respect to an economy and assumed it’s a universal rule: market transactions beat central planning.  But it is simply not so.  Imagine if, in a huddle (American football), the quarterback were to not call a play (central planning) but rather offer substantial cash bonuses to the player who finds the ”best” way to beat the opposing team on the next play.  This would not unleash entrepreneurship, nor get the invisible hand to working on the problem.  It would just be a mess…chaos, if you will.

 

Section 9: Anarcho-Capitalism: Safer and  More Efficient?

 

Murphy argues that MA would do a better job at ensuring product safety than does government.  Consider airlines, for example.  Murphy points out that insurance companies would force client airlines to pursue safety measures so long as the expected savings in avoided indemnities exceed the costs of the safety measures.  This correctly recognizes there could be such a thing as too much safety, but he seems to miss this implication.  He instead says that the FAA has no incentive to reduce crashes and actually benefits from them.  This is false, and if were it were true, the FAA would act to maximize crashes, something we do not see.  It is very possible that the FAA actually imposes too much safety from an economic standpoint. (Economic analyses of FAA regulations for aircraft de-icing have sometimes concluded this.  FAA may be over-cautious, since they don’t bear the costs of too much de-icing but do get and earful from angry families and their Congressional representatives in the event of a fatal crash.). 

 

The point isn’t that Murphy’s system wouldn’t work in providing safety, but rather, that it doesn’t do what he thinks it does.  He clearly doesn’t understand implications of his own system, and doesn’t understand how government functions, either.  A real world example that Murphy cites elsewhere is that the FDA is too careful to permit new drugs to enter the market, for safety reasons.  This is likely correct, but undercuts Murphy’s own claims about insurance companies. 

 

It’s important to emphasize that Murphy’s analysis really doesn’t sensibly characterize government at all, which in his writing is simply an evil, incompetent, wasteful monolith.  This picture of government is completely misleading, as another real-world example illustrates.  On identical, side-by-side forest tracts in Montana, the State of Montana consistently earns positive net revenues by issuing logging leases to private companies, while the federal government consistently loses enormous sums on its similar leases to private firms.  The post-logging reclamation on the State lands is widely agreed (by the State foresters, federal foresters, and environmentalists) to be superior to that on federal lands.  And the leases on State lands are as profitable to the logging firms as those on federal lands.  So why the crucial difference?  It has to do with the way that the respective government agencies are set up – the rules under which the state bureaucrats work are very different from those governing the feds.  Such fundamental differences are necessarily ignored in a simplistic analysis in which “the state” is just an evil, stupid, monolithic entity.  In fact, “the state” is always a collection of people, operating under very complex rules and constraints; different sets of constraints generate states that operate very differently.  Murphy contrasts government at its worst with a market system with costless transacting, and concludes that the market system is superior.  But this comparison is completely irrelevant for the real world.  Among real-world alternatives, which would give better results, more freedom and “efficiency”: working to improve imperfect government so that it is properly constrained and agencies operate more like those managing Montana’s School Trust Lands, or working to establish imperfect free-market alternatives?  I think there are numerous reasons to favor the market alternative, but Murphy’s analysis is completely useless for tackling these serious questions.

 

Section 10: The Doubtfulness of Superior Efficiency

 

Murphy claims that MA would be more efficient than statism, and also that if everything in the world, or within a society, were held by a single owner, that owner would charge rental rates to maximize his rents.  But this means that any state, or government officials, could convert their existing system to an anarcho-capitalist system with the officials as rent-maximizing owners, and make themselves wealthier.  Murphy says “states” have non-pecuniary motives (which, as I have argued in section 8, may actually be an advantage of the state over profit-seeking firms for some purposes).  But surely some officials  somewhere (e.g., infamous kleptocrat Mobutu Sese Seko of Zaire) must care primarily about money.  But Mobutu didn’t establish anarcho-capitalism in Zaire, and nowhere do we seem to see an anarcho-capitalist system emerging.

 

In sum, Murphy’s presentation of the workability and superiority of his MA alternative falls completely flat.  It’s not that anarcho-capitalism has been refuted, but Murphy’s version of it makes no sense, and neither does his view of the alternative, the state.  In short, I learned nothing from the second essay.

 

Conclusion

 

Dr. Murphy seems very sincere, and respectful of his readers regardless of their viewpoints.  (This can’t be said of J. Sapienza, the author of the Introduction, who raises a reasonable question that might be asked about anarcho-capitalism, and then responds to the hypothetical questioner with rather rude invective.  This is too bad, as Murphy himself keeps a high tone.)  Murphy seems genuinely interested in liberty.  Hence, it is unfortunate to have to be so negative.  But this work is the worst discussion I’ve ever read of how an anarcho-capitalist society could function.  The book shows not even a faint understanding of the economics of contracts and economics of insurance, even though contracts and insurance are the fundamental elements of Murphy’s system.  Transaction costs and the economics of organization are entirely ignored.  No attention is paid to the incentives facing insurers, who are central to his system.  The system described is internally inconsistent, and at odds with the facts of reality as well. This is quite unfortunate, because it leaves the reader with the impression that libertarianism and anarcho-capitalism are patently silly, which they really are not. There are serious economic cases for anarcho-capitalism (e.g., David Friedman) but this isn’t one of them.

Charles N. Steele is Assistant Professor of Economics at Hillsdale College.  He has previously taught economics at the graduate and undergraduate levels in China, Russia, and Ukraine.  He also has extensive professional experience as a private consultant in the insurance industry on problems of design and evaluation of insurance programs.

This TRA feature has been edited in accordance with TRA’s Statement of Policy.

Click here to return to TRA's Issue XCI Index.

Learn about Mr. Stolyarov's novel, Eden against the Colossus, here.

Read Mr. Stolyarov's new comprehensive treatise, A Rational Cosmology, explicating such terms as the universe, matter, space, time, sound, light, life, consciousness, and volition, at http://www.geocities.com/rational_argumentator/rc.html.