And now I shall be off scarfing smoked salmon and fresh apple pie seasoned with free market tears, or
Why a collection of confused retards does not amount to a free market.

Thursday, 10 October, Year 5 d.Tr. | Author: Mircea Popescu

Note that this isn't the longest title of a Trilema article.


The free market is the most efficient solution to a specified class of problems. It is not a universal panacea, it is not a means to violate basic thermodynamics, it's not a procedure to bring forth milk out of stone and cheese from the very moon.i

For the benefit of those who seriously aim to improve their understanding of this vaunted concept past naked, naive fetishism and into some sort of rationally usable system, there are two major limits on the applicability of the free market :

I. The free market requires information symmetry. The cannonical form this problem is captured in could be the observation that "an exchange in which Joe trades sight unseen a satchel in which he might have put some gold for a sack in which Moe might have put some potatoes is not free nor can it be part of a market."

A very common example to easily understand how this works in practice is the used car market. For those looking to buy it is expensive to establish exactly the wear&tear levels of any one used car. This creates an information asymmetry : the seller knows much better than the buyer what the item is worth. Consequently, the buyer hedges his bets, and in order to make sure he is not taken advantage of too badly he is unwilling to pay more than what he perceives to be "the average price".

The problem is that the seller is also in a position to know this "average price"ii and so he has an array of attacks at his disposal. In the car market what he does is he avoids putting on the market any car which is objectively worth more than the "average price" cutoff. What this does over time, of course, is lower the average value of cars on sale, which in turn lowers the "average price" which in the end ensures that nothing but lemons will ever be available to buy. Left untreated the situation can devolve to absolutely comedic extremes.

In other cases the seller favours other attacks, such as for instance manipulations of the perceived "average price". In the case of the alleged "domain name market" the sellers of fundamentally worthless items (domain names are under no circumstances worth more than what they cost to register) enter into highly publicised self-trades aimed to create a convenient if unsupported impression among the general buyer population that the "average price" is significantly above the value of the items on sale. Other Internet scams, such as the stuff Burnside or Nefarioiii were peddling use a very similar approach to prop the value of insubstantial "assets".

II. The free market requires completeness. The free market fails to work - and spectacularly so - when it is applied over partial fields, in which actors are able to hide externalities.

For instance, take the case of the girl that baked my apple pie depicted above. If she is allowed to not clean up the kitchen after herself, she will be able to bake pies at a lower per-pie cost than the other girls, which do also clean up afterwards as part of baking the pie. The net result of this will be a very dirty and also increasingly dirty kitchen tableiv.

This problem becomes extremely serious when coupled with welfare or entitlement systems. For instance, in a country in which health care is offered free to all citizens, a fast food place which sells very cheaply extremely bad food secure in the knowledge that the public budget will treat for "no cost" all the people it makes sick is in fact running more of a tax fraud business than a legitimate eatery business.

To understand each other : it's absolutely true that free markets are very good at allocating resources. It is however the case that free markets need some items to work their magic, they don't just operate by nominal invocation, like the Holy Spirit. One major such element is information symmetry. Another such element is containment, or the absence of hidden externalities. Absent either of these, it can not be the case that whatever is going on qualifies as a free market, even if it appears that way to the naive observer.

  1. Notably, a century or so ago Romania's national poet was complaining of the same disease brought by a different fetish :

    Da-i soseaua rea, încât ti se frânge caru-n drum? Libertate, egalitate si fraternitate si toate vor merge bine. Dar se înmultesc datoriile publice? Libertate, egalitate si fraternitate da oamenilor, si s-or plati. Da-i scoala rea, da' nu stiu profesorii carte, da' taranul saraceste, dar breslele dau înapoi, dar nu se face grâu, da'-i boala de vite?... Libertate, egalitate si fraternitate, si toate or merge bine ca prin minune.

    which in English would read,

    So the road's broken and unmaintained, so your cart breaks down ? Freedom, equality and brotherhood and everything will be fine. But is the public debt growing ? Freedom, equality and brotherhood so that it may be paid. But the school's bad, the teachers can't read, the peasants are growing poorer, the artisans are disappearing, the crops falter, the livestock sick ? Freedom, equality and brotherhood and everything will miraculously improve.

    The impact of the French revolution is quite visible in the form the disease was taking in the XIXth century, but its substance is the same, unchanged. []

  2. The reason for the quotes is that averages can only be calculated over collections of identical items. There's no average gender of the human race, nor is there an average fruit. []
  3. Aka James McCarthey []
  4. Speaking of which, if you're young, fetching and comfortable being topless I'm looking for more pie bakers, apply below. []
Category: 3 ani experienta
Comments feed : RSS 2.0. Leave your own comment below, or send a trackback.

8 Responses

  1. Mateusz Wywial`s avatar
    Mateusz Wywial 
    Thursday, 10 October 2013

    I suppose your first point is much weaker than one could expect. For an exchange to be mutually beneficial, it is only necessary that the economic actors know that they do not have perfect information, which is much weaker assumption than symmetric information.

    Still, asymmetric information can ruin a market, but this is an empirical and not a theoretical question. Consider two numerical examples:

    A) (from Bryan Caplan:

    Suppose that the value of a used car to a current owner is uniformly distributed on the interval from $0 to $100. (That is a fancy way to say that every dollar value from 0 to 100 is equally likely). These cars would however be worth three times as much to someone else. The catch: Current owners know the true worth of their car; buyers only know the average value of cars on the market.
    What happens? If the buyers bid $50, then every car worth $50 or less gets sold. The average value to the original owners: $25. The average value to the new owners: $75. Profit: $25. Competition forces the buyers to bid more. So they bid up to $75. Average value to original owners: $37.50. Average value to new owners: $112.5. Profit: $37.50. Competition ultimately pushes the price up to $150. Average value to original owners: $50. Average value to new owners: $150. Profit: $0.

    B) same argument (again from Caplan), This time, the car is only twice as valuable to the buyer than to the seller. The buyers initially bid 50, and anyone whose car is worth between 0 and 50 sells. The average car sold, therefore, is worth 25*1.5=37.5 to the buyer. The buyer has to pay 50 to for an average payout of 37.5. In equilibrium, the market price falls to 0, and the whole market disappears.

    Obviously, you could further manipulate the numbers and distributions, but it is clear that the result is very much dependent on the numeric parameters. Usually, the very existence of a market implies that it somehow overcame the information problem.

    Finally, even in case B there should emerge a meta-market for institutional setup of a market - it's a market solution of a market failure. The mighty MPEx is a clear example of a winner in such market for market. Other market solutions include guarantees included by the car sellers, insurances and so forth.

    Your second point, concerning completeness of a market, seems much stronger. Yet, even in this case, ruling out the market solution should me empirical and not a priori, as pointed by Ronald Coase in his famous Problem of Social Cost. Not being applicable to the example of fast-food w/ free healthcare, it may be quite applicable to the dirty kitchen case. Lacking costless property law execution or these being poorly defined, the most efficient baker girl can negotiate table cleaning with the dirty baker girl and the outcome can still turn out efficient.

  2. Mircea Popescu`s avatar
    Mircea Popescu 
    Thursday, 10 October 2013

    @Mateusz Wywial We are not discussing the same thing here. In my model the car has an objective value attached to it, which would be its "true value". In your model the car has no objective value attached to it, and instead it is represented by context-values. Thus it ends up having one value in the buyer's context (presumably some sort of utility-value) and another in the seller's (presumably some sort of "what can it be sold for" value). Thus in my perspective the market operates to discover the "true value" of the car, within arbitrary precision, whereas in your perspective the market operates to negotiate translation between contexts.

    I am not proposing either approach is wrong and inadequate in principle, but it must be recognised that they being so fundamentally different, it shouldn't come as a suprise that the implications of theory formulated in one contradict, at least on the surface, the implications of theories formulated in another. For a common example of this problem, consider that both civil and common law systems generally work to keep people unrobbed, unraped, undefrauded, fully taxed to within an inch of the maximum they can bear and so on. Nevertheless, the exact mechanisms one system uses to work towards this goal are often enough anathema in the other. More generally speaking, while simple parts of two homologuous systems may be readily recognised as equivalent, the homomorphisms between more complex parts may be extremely difficult to follow. It often works out that the start and end point are simple and the middle complex, and therefore...

    Of particular interest in this vein is the following example of petitio principii :

    Usually, the very existence of a market implies that it somehow overcame the information problem.

    You certainly must be aware that you are bringing the "existence is proof of existence" argument in a discussion of the very inexistence. That's the thesis here, that a broken market is not in fact a market, even if a number of people labour under the delusion that they are participating in one.

    To be perfectly clear, my argument as presented in the article is neither intended nor in fact logically capable of carrying some sort of government-equivalent structure that'd "fix" the market. It's certainly not a matter of "ruling out the market" in any sense. Consider for illumination of this point the meta-problem of finding the truth in science [which is very much homologuous to the problem of price discovery in economics]. It can well be argued that the only method of finding the truth is through free discussion among the proponents of various theories as to that truth. While one may make the observation that two groups speaking completely different languages are not in fact engaging in a debate as to the merits of their respective theories, irrespective of whether they think they do, that is not to be construed as a general admonition to go consult the village shaman / holy text y / the pope so as to directly obtain the truth.

  3. I have no idea what you two are smoking, but the A example made no sense whatsoever. "Profit: $25, Profit: $37.50, Profit: 0" whaa? Did either of you actually read that?!

  4. Mircea Popescu`s avatar
    Mircea Popescu 
    Thursday, 10 October 2013

    Yes, yes. It's something like this : suppose a tribe of Indians meets a bunch of Dutchmen. Suppose these unrelated tribes that never met before use a common currency, for the same value and in the same way (this is patently nonsensical, even if the currency is a bullion metal, but hey, we're in the area of academic economics, so this is par for the course). Suppose they are trading an item, be that item "cars".

    For the Indians, the utility-value of cars is normally distributed in the [1, 100] interval. For the Dutch, the utility-value of cars is normally distributed in the [200, 400] interval. (This was actually the case historically, with various spices, tobacco and whatnot. Pomodori are ethimologically reflective of the situation.)

    In this highly contrived nonsensical context, should the buyers cartel (as they normally would in similar, real circumstances) and offer 50 per car, they will acquire all the cars in the [1, 49] interval, or roughly half the cars. This again is patent nonsense, as the Indian who sees his friend sell his 13 worth car for a 300% profit is not likely to sell his 49 car for a 2% profit, and this is also how the Dutch find out about how to value cars in the Indian system, but hey.

    The average profit of the Indian in this theoretical example would be 25 per sold car. The average profit of the Dutchman would be 250 per bought car. Together they realise a profit of 275, which is unevenly split.

    Admitting no further cars are produced as a result of this, and the demand on the Dutch side is not satisfyied, the next round will result in as many cars being sold, for a price of 100 each, yielding the Indians the same 25 per sold car in profit, but the Dutch merely 200, for a total of 225, which is somewhat more evenly split this time.

    At this point it's game over and the Dutch have all the fishies (which is not exactly unlike the result of "free market" interactions between the actual Indians and the actual Europeans in the continental US, as we can find today in the field). The Dutch continue to happily trade their cars among themselves for something in the [200, 400] range and that's that. The price has increased towards average values and the profit distribution has pushed towards balance, slightly, somewhat, for as long as the Indians still had any cars at least. After which it was a 100% Dutch affair and good luck Daisy.

  5. Mateusz Wywial`s avatar
    Mateusz Wywial 
    Thursday, 10 October 2013

    We are not discussing the same thing here. In my model the car has an objective value attached to it, which would be its “true value”. In your model the car has no objective value attached to it, and instead it is represented by context-values. Thus it ends up having one value in the buyer’s context (presumably some sort of utility-value) and another in the seller’s (presumably some sort of “what can it be sold for” value). Thus in my perspective the market operates to discover the “true value” of the car, within arbitrary precision, whereas in your perspective the market operates to negotiate translation between contexts.

    While I agree that when discussing a more complex problem, this difference would be very important, in this case objective-value approach can be translated to one presented by me quite easily. The value for buyers should be exactly the same as value for sellers. Then, obviously, the market will be broken. Still, I believe that my question remains valid: not every market is the one for goods with objective value (marginal value for marginal buyers seems much more frequent in reality) and thus to determine if asymmetric information is harmful one must inspect that particular market and not invoke the theory.

    As to the alleged petitio principii, it might be a result of inadequate clarification from my part. My reasoning was as follows:
    The theory points that upon asymmetric information, it will be either relatively irrelevant (as in my case A) or destructive (case B). If market exists, then it is probably case A. Unless participants are deluded and do not know that they do not possess perfect knowledge, or equilibrium is not yet reached in case B.
    I of course agree that market participants' delusion wreaks havoc on both the market and participants themselves, I noted it explicitly:

    For an exchange to be mutually beneficial, it is only necessary that the economic actors know that they do not have perfect information, which is much weaker assumption than symmetric information.

  6. Mircea Popescu`s avatar
    Mircea Popescu 
    Thursday, 10 October 2013

    I can easily grant that one can't simply barge in and declare any random market dysfunctional on the grounds that he's identified some real (or oft imagined) asymmetry.

    The problem with the "economic agents know they don't know" statement is that it requires possibly one of the most difficult mental skills for the human race. For some reason this is just... hard!

  7. It
    It's still not me 
    Monday, 8 May 2017

    Why am I back to year 5, do you really serve up old stale bread on your home page address? Otay though, all information is asymettrix. When all information is distributed evenly, there is no more downhill, and we have perfect smoothness. I for one, still do not believe in entropy, though I do love my Sister-in-law, Shannon R., but that is the other kind. Okay, how much does it cost to buy an FuckGoats? I really can't go out of my way to buy an FPGA, I really don't even like Xilinx, but JTAG rox. So, I didn't see the price on yer site, just the cost.

    I also ran into the hitlery porn shit, that is rude man. No wonder people say what they do. I still like you though, even though our bacon number is 2. You know Tim May, Anton, Lizard, Jamey Donnald? WoT!!!

  1. [...] curse of modernism, as an intellectual attempt, is precisely this "fake it, maybe you make it" thing. If the little Corsican who wasn't an Emperor could pretend like he were an emperor [...]

Add your cents! »
    If this is your first comment, it will wait to be approved. This usually takes a few hours. Subsequent comments are not delayed.