Following up on what I wrote last night, I think it is time to look at a problem that makes all efforts at scientific economic management impossible. That is the fact that, outside of a freely set market price, we have no mechanism to accurately measure economic phenomena.

Let us start by looking at some simple concepts, the supply and demand curves. These are probably the most basic elements in an economic system. As any econ 101 student can tell you, the curves tell us what amount of a product will be demanded at a given price and what amount will be supplied. Now, under some systems, such as full communism, the supply curve may have less importance, but at least the demand side needs to be known, in order that we not over produce a product that is no longer wanted. Even if communism did away with money exchanges entirely, the managers would still want to know the relative demand for good, so that they would not flood the market with boots when there was an unfulfilled demand for lawn mowers. So, while they would not need a demand curve precisely, they would need some cognate which measured relative demand for all goods.

The problems with calculating a demand curve are many.

First, the demand curve is made up of the aggregate demand of every consumer. Since we cannot easily poll every consumer about how many of a given good he would buy at each price increment, we have to content ourselves with approximations. These would likely take two forms, either a survey of a representative sample, or a curve generated using various historical data. However, as those two methods simply exacerbate inherent problems, I will instead imagine that we do have the capacity to instantly poll every member int he market, and show that this perfect knowledge is still of little use. From that we can then see how our approximations simply make the problems even worse.

Perhaps if we look at only one user the problem will be more obvious. Suppose that the poll asks how many potatoes you would buy at various prices. You respond honestly and they add you to the calculation. A day later they poll again, and this time rice is cheap, so your numbers have changed and are lower. Another day they call again, and you are hungry and the numbers are higher. Next poll you are craving potatoes and the numbers again change. And finally they call after you satisfied your craving and you can't stand the thought of eating another potato, and the numbers drop.

Now, what this shows is how inconstant the numbers are. From second to second your preferences change. Now multiply that by every buyer int he market and you will see how impossible it is to generate a demand curve. And even if we could, it would be valid for only the split second it was generated, the next second making it invalid. (This is why historical aggregates cannot be sued to generate curves, as they show differing conditions as well as differing prices, and thus are essentially comparing apples and oranges.)

The other problem is one of interaction. While economists gloss over this by adding "ceteris paribus" to all their statements, meaning all other things remain equal, the truth is that a change in the economy ripples out and changes all sort of factors. A drop in wheat prices causes a drop in meat prices which causes an increase in demand for grills, which causes a rise int he price of steel, which increases the price of cars. It is probably a minute change, but when they all add up, these curves all interact in unforeseen ways. Rather than discrete sets of supply and demand curves, the economy is a massive set of countless simultaneous equations.

Nor is it any better on the supply side, as the supply side is simply the combination of a number of demand curves for the components of the good, including raw materials and labor, as well as an allowance for a reasonable return and capital depreciation. Actually even those last two are demand curve, the demand curve for investment funds and the demand curve for the capital goods being replaced. So, if demand curves are in constant flux, then so are supply curves. Of course, long term supply and labor contracts do tend to flatten these out a bit more on the supply side, but not that much. With the possibility of new firms entering and old firms dropping out of a market, as well as retooling for alternate production, even the supply side is more fluid than one would imagine.

I am sure some will now ask, "if predicting supply and demand curves is impossible, then how do businesses function?" But that is a simple answer. Businesses do not need a precise knowledge of supply and demand to function. Unlike those planning scientific management of the whole economy, businessmen can work with approximations. They can also adjust prices slowly to measure the effects on profits, and can also use the knowledge of their competitor's prices to set their prices. And, more important than any of those things, the market itself forces them to set their prices close to the optimum. If they charge too much, they see sales go to competitors, if they charge too little they see inventory sell out while profits are low. In either case, their low profits will draw investors to their competitors, giving them a sure sign that they need to change their prices. With no knowledge at all of the supply and demand curves they can set something close to the optimum. This is not an option available to those planning an entire economy.

Some will doubtless claim that my statements are true, but only apply to microeconomics. They will claim that neo-Keynesian macroeconomic measures are accurate and easily generated. But that is completely untrue.

Just ask an expert what the total money supply of the US is at a given moment and you will see the problem. Not only are there dozens of definitions of various money supply measures*, each with a group of partisans arguing that it is the one true measure, but even those figures are themselves often difficult to determine. For example, is a line of credit an expansion of the money supply, even when it has not bee used? Or is it part of the money supply only when it has been spent? Or should we ignore lines of credit entirely? If we are counting demand deposits, do we also count money market funds with checking privileges which mimic demand deposits while acting also as investment vehicles? Does money still in existence but slated for destruction count toward the money supply? Are debts owed by the government to itself considered real debts? Likewise, are fictional transactions within corporate divisions undertaken for accounting purposes only real debts, or not?

While economists pretend that such questions are easily resolved, in reality, any such figures are at best an approximation. And even then, they are bad approximations. As with the supply and demand curves, as soon as a money supply figure contains anything other than physical cash reserves, it becomes quite time sensitive, as there are any number of very short term transactions carried out between financial institutions as well as other organizations for a number of reasons. For example, overnight money borrowed to cover reserve requirements can, on paper, increase money supplies, which then shrink again in the morning, yet this is normally not reflected in the figures for money supply.

But perhaps financial institutions and money supply are a bit vague, let us look at something always reported with great precision, the trade deficit. Surely that figure must be certain, and easily calculated, right?

No.

Let us just take one example. Toyota Motors sells a number of autos to the US, creating a deficit of, say , $10 million. Toytoa then takes that $10 million and buys a parcel of land. What is the trade deficit?

Most of the time, those who promote the idea of a trade deficit do not consider "capital purchases" as trade, so they would say there was still a $10 million trade deficit. But why? Toyota spent $10 million in the US, why does this purchase not count? Because it is a "capital investment"? What if they bought the land to provide employee housing for their US plants? Does that not make it an end use, and thus a "retail" purchase? And so should the balance of trade be even? Similarly, why would we consider the balance of trade to be even if Japan bought $10 million of machine tools and took them to Japan, but if they spent $10 million building a factory in Kansas, we still consider there is a $10 million trade deficit?**

But a lot of the problems with balance of trade issues is ideological. Let us look at a figure which has problems that are not ideological but practical. That figure is the gross domestic product. The most obvious problem of the GDP is that it is a monetary figure that is generated over a longish time span. During periods of low inflation this is a minimal issue, but during periods of serious inflation, or int he case of some nations hyperinflation, this can seriously distort figures, as the dollars used to value items may be worth much more than the dollars used to value items in December. But even if we were to work in a system with an inflation of 0%, there would still be problems.

The problem is that GDP does not truly measure what a country produces, but what is purchased. (Which is why inflation has such an impact on the numbers.) Since it measures essentially cash spent, it generates a tangible number, but one of rather dubious utility. If we wanted to increase our GDP drastically, all we would have to do is charge one another for each line of conversation. I would pay you $1, you would pay me $1, and so on. By increasing the rate of "services" rendered, we could massively inflated the GDP without actually increasing our aggregate wealth one bit. On the other hand, we could also have a massive, wealthy economy, but operate purely through barter or gold only contracts, and have a GDP of $0. Which means that the GDP, while of obvious interest to tax collectors, is of little true utility in determining the economic health of a nation. A nation caught in the death spiral of hyperinflation could still post a respectable GDP, while a relatively underdeveloped nation could have a GDP which would severely understate the true amount of economic activity.

Now, the shortcomings of these aggregates, and even of the supply and demand curve, would be no problem were they used as originally intended, as illustrations for teaching and perhaps as broad generalizations to help with rough comparisons. The problem came when people saw the numbers attached to these statistics and assumed that numbers were the same as precision. Thus was econometrics born, and the horribly dangerous myth that it was possible to scientifically manage the economy. But numbers can be attached to anything, the presence of digits is not an indication of precision.

When we cannot even generate the most fundamental numbers with any reliability, what hope is there for scientific management of the economy? It is pipe dream. Yet one that persists, continuing to do harm. Hopefully one day someone will point out to those who believe in this theory that human behavior and preference is not the same as the laws of physics and that one cannot manage humans scientifically, no matter what any theory might say.

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

* Though M0, M1, M2, &M3 are currently the most widely used, at various times there have been a number of additional measures and variations on the current measures. Especially during the monetary chaos of the 1970's and the peculiar financial transactions of the 1990's, the number of such measures proliferated to try to create a better representation of the economy.

** Many of the idiocies of the trade deficit are hold overs from the silly mercantilism of the 16th century, which was also popular in the early 20th century when Keynesianism was born. Protectionists also tend to use these figures to argue for more impediments to trade. As both groups have an interest in limiting imports, the theories behind the figures tend to skew them towards always showing a deficit.

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

Note: All that I wrote here is dealt with at great length in Ludwig von Mises' Human Action. He also explains quite extensively how the impossibility of economic calculation without a market makes communism impossible in his work Socialism. And I enjoy mentioning both of those works, as this is one of the few times I can claim that my description is more concise. (Though his still reads better.)