Sunday, March 2, 2014

False Precision



NOTE: I am reproducing several articles from my old blog. The first, "The Best Argument Against the FairTax", should have been included among the other essays I reposted recently, but was not. The second, "The Problem of Pornography", is one I found while searching for the FairTax article, and as it seemed interesting, I decided to reproduce it here. "I  Knew I Was On To Something" is cited in the second essay, and is one of my favorite essays about a topic which troubles me greatly. In fact, I am surprised I did not reproduce it earlier. (The same quote it discusses is the subject of my "Stupid Quote of the Day" series on December 30, 2012. Since I plan to reproduce that whole series sometime this week, I am not copying that essay tonight.) "The Politics of Psychiatry" mentions the Potter Stewart quote examined in "The Problem of Pornography" (and is amusingly controversial to boot), and "Finding What You Are Looking For" is cited in that essay. "False Precision" provides a good analogy, showing the futility of "scientific management" of economics, a topic examined again in "The Perfect Model". "Be Careful When 'Sticking It' to 'Big Business'", though burdened with a horrifying number of needless scare quotes (thankfully, something I have given up in recent years), is an interesting predecessor to the argument I made best in "In Praise of Contracts". "Kelo, Home Schooling and Drug Laws - Inconsistent Theories of "Social Costs"" is an interesting examination of the harm done by pragmatic solutions, often proposed by moderates and even conservatives. "Trial by Parenthood" consists of my reflections on how having children divides us into two groups. "An Interesting Contrast" meanders here and there, covering everything from the definition of hypocrisy to the benefits of search engines, yet, somehow, it manages to amuse me. "Why Term limits Will Fail (And Should)" is one of my earlier examinations of the subject of term limits, a topic examined again in "Critique of a Congressional Reform", "Some Thoughts on Term Limits", "Damn the Torpedoes!", "Making the President Irrelevant", "Racketeering Through Legislation", "Power and Disorder", "Our Aristocracy" and "The Problem of Professional Politicians, or, The Impossibility of a True 'Ousider' Candidate". "The Cost of Big Government" takes an all too brief look at precisely that, the cost of big government. "Transparency, Corruption and Reform" is a frequently cited essay looking at the relationship between the size of government and the amount of corruption.  Finally, "A Perfect Quote" provides a very brief statement I found on IMDB, which perfectly summarizes the greatest problem with the political left.



I have written several times on the limits of "scientific" management of the economy. In one specific essay, I argued that econometrics, while providing the appearance of precision, often actually made things more confused.

One of the points I raised is that desires are constantly changing and cannot be properly quantified. The price system, as established by an absolutely free market does the best job of representing the relative valuations of the entire population, but even there there is some lag and some imprecision. However, should we try to substitute some other mechanism for the freely established price, or even to simply manipulate the market price, we will end up misunderstanding the true desires of the consumers and our economy will produce less satisfaction than one without such manipulation would.

The argument that the price system is the sole tool for understanding consumer desires is not my own, it was one of von Mises' greatest contributions to the understanding of socialism, making this argument explicit and showing how communist systems cannot function due to a lack of information about consumer desires. And int he years since von Mises made this argument, some socialists have tried to circumvent it by proposing methods for measuring desire separate from prices.

And to some that may sound valid. Rather than relying on the free market, many would argue, why not determine the relative desires of the consumers and then use those in calculation? Especially with our modern computing technology, surely something could be done to track the shifting desires and determine a scale of desires.

It sounds plausible, but it will not work. The problem is that valuations are ordinal, not cardinal. We can add numbers to desires to get a mock precision, a consumer could say "I value cheese 7 and strawberries 14", but does that mean he will always trade 2 units of cheese to get one unit of strawberries? No, it simply is adding mock precision to a very vague preference. We can say a>b>c>d when discussing preferences, but we cannot say 1a=3b=7c=12d*. Attempts to apply cardinal numbers to ordinal lists may appear to provide precision, but it is an illusion. Would you believe someone who says "I love you 7, but Mary 12"? Then why trust anyone who thinks they can numerically represent individual preferences for goods?

And I have now stumbled upon the perfect example. An attempt to place numbers to a subjective, ordinal list to turn it into a set of cardinal numbers. It is the Schmidt Sting Pain Index. The original idea is a valid one, to evaluate the relative damage done by various stings. However, rather than saying "X is more damaging than Y" the creator started to give them numeric values. And that is where it is useful for our example.

There is nothing wrong with providing numbers, so long as everyone involved recalls they are simply arbitrary choices, selected solely to keep an ordinal list in the correct sequence. At best they may be seen as a single individual's rough estimate of the relative strengths. However, once someone starts to take the numbers seriously, they do much more harm than good, by providing the appearance of precision to something arbitrary.

For example, bullet ant stings are rated 4.0, while sweat bees are rated 1.0. Does that mean that one would willingly undergo 3 sweat bee stings rather than a bullet ant sting? But they would rather be stung by a bullet ant than 5 sweat bees? Given that sweat bees cause almost no pain while bullet ants cause intense agony, I would think not. But the numbers give the illusion that such calculations can be made, and, more importantly, that they are valid conclusions.

 And that is the problem with plans to use some sort of numerical index of individual desires to replace the price mechanism. It uses arbitrary numbers as if they were precise.

Of course, some are asking why I am spending so much time on this, as it is only applicable to a few socialist theoreticians. However, it is not. These socialists may be the only ones who have explicitly formulated such a theory, but they are hardly the only ones who think they can somehow divine the "real" situation behind the prices.

Think of how many times you have recently heard that oil prices are not explained by "market forces", that only speculation could explain it? Have you ever stopped and asked yourself how anyone could know that

Now, obviously, the answer is, they don't, they are just politicians stumping for reelection and looking for someone to blame, but the question remains why people believe them. And the answer is, people think there really is some way we can tell what people "really" want, set some numbers to desires, and measure it against the current situation. In other words, the econometricians have convinced people, or rather bamboozled people, into believing there is a way to tell whether the price is "justified" or not.

The price simply is. We cannot know the fictional supply and demand curves which establish it, as they fluctuate with the second. We may be able to correlate the price with certain other events and argue that this or that event caused the price to rise, but the price itself is neither right nor wrong, it simply is. All the price tells us is, given the circumstances at time X, people who were willing to pay Y could get all they wanted of a given good. Beyond that, the price tells us nothing, and is neither right nor wrong. A price an no more be "wrong" than a child's name can be wrong. It simply is, it is a primary datum.

To try to find any more meaning than that in a price is akin to trying to decide how many stings from a sweat bee could be traded for one sting from a bullet ant. Simply worthless.

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

* Attempting to place numbers also obscures another problem. In terms of individual preference, a>b and a>c says nothing about the relationship between b and c. 

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

POSTSCRIPT


I have, from time to time, heard various stupid proposals about lowering oil prices defended with the argument, "who cares if they are right,a s long as they work". Thus I have seen people defend limiting speculation, even if the speculators are not themselves to blame for rising prices. For those who believe in such arguments, I would refer them to my older essay "The Shortcomings of Pragmatism". While stopping speculation may slow the rise of prices slightly, for the moment, it will have no impact on the final price, and it will do nothing but cause prise shocks to be more severe and sudden. It will also massively increase government intervention in the future markets, which will likely have nothing but unpleasant consequences.

Originally posted in Random Notes on 2008/07/28.

No comments:

Post a Comment