|
Price as a Basis of Total Output
A response to the dilemma set out in Part I readily suggests itself. Bales of wool processed and exported, and motor cars manufactured and repaired, etc., clearly cannot be combined into a measure of total output; but surely the amounts of money expended on these – and, indeed, all – goods and services can be readily summed to form such an estimate. Upon reflection, however, this solution is far more apparent than real. To calculate the revenue generated by the buying-and-selling of a specific good or service, we multiply the good’s price by the quantity of goods exchanged. But what is a price? As discussed in Value Investing and the (Mis)Measurement of Results, it is the rate of exchange between two goods. Price is the amount of one good that can be exchanged for an amount of a second good, and is expressed as a ratio of the first to the second. Where money exists, price will be the amount of money divided by the amount of the good or service being exchanged. A price is established in a transaction involving two particular goods, two parties (i.e., one buyer and one seller) and a specific place and point in time. Accordingly, a change in any of these elements of the transaction – goods, buyer and seller, place and time – is likely to change the goods’ rate of exchange. Accordingly, and as participants in currency, commodity and equity markets know only too well, goods and services can remain constant but their prices are subject to constant and sometimes violent change. These changes occur for many reasons and very often for unknown reasons. On these grounds, plus the related point that revenue and output are not synonyms, prices provide a tenuous base for the estimation of total output and hence of productivity.
Our original mathematical problem also refuses stubbornly to disappear. To see this, suppose that two transactions are conducted. In the first, a Dell Inspiron 2100 Notebook computer is exchanged for $A2,949; and in the second, a Dell Dimension 4100 Desktop PC is traded for $A1,249. Assuming for simplicity’s sake that these transactions comprise the totality of computer-related exchanges, in order to estimate total computer output (as opposed to revenue) we must add these two ratios. Even with respect to “very similar” goods, however, and even with just two transactions, it is conceptually meaningless to add $2,949/1 Dell Inspiron to $1249/1 Dell Dimension: we lack the common denominator required in order to do so. Simply to relabel each denominator as a “computer” is to blur essential information and to ignore the underlying problem. The reason a buyer is prepared to exchange more than twice the amount of money for the Inspiron is surely because he regards it as different from and more valuable than the Dimension. Our problem is no more difficult but more apparent when we attempt to combine the average price and quantity of “different” goods – which we must do if by this procedure we are to estimate total output and hence productivity. In this context it is useful to note that commodity markets quote ratios of money to fungible goods, i.e., $US per barrel of West Texas Intermediate crude oil, $A per kilo of 19 micron wool, etc. It makes no sense to combine these ratios and no attempt is made to do so. Similarly, it makes no sense to calculate an average of the exchange rates $A/£, $A/$US, etc. More generally, Murray Rothbard concluded in his magnum opus Man, Economy, and State that “any concept [that] involves adding or multiplying quantities of completely different units of goods, such as butter, hats, sugar, etc., and is … meaningless and illegitimate. Even pounds of sugar and pounds of butter cannot be added together, because they are two different goods and their valuation is completely different.” Our inability to compute total output and hence productivity is seemingly no closer to resolution.
A Fixed Weight Price Index as a Basis of Total Output
Another response to this dilemma, the use of a “fixed weight price index has been suggested. Such an index, according to its proponents, establishes changes in the overall purchasing power of money and on this basis infers changes in output. If the volume of monetary transactions increases by 5% and the purchasing power of money falls by 2.5% then one infers that real outlays have grown by 2.5%. Consider a simple example suggested by Frank Shostak in his excellent article The Labour Productivity Myth (which served as the starting point for this analysis). In Period 1 Tom buys 100 hamburgers at a price of $2 each and 5 shirts at $20 each. Assuming that these comprise his total outlay for the period, we have $2*100 + $20*5 = $300. Note that the “weight” of hamburgers in Tom’s total outlay is 0.67 (i.e., two-thirds of his expenditure consists in hamburgers) and that the shirts’ weight is 0.33. In Period 2 the price of hamburgers rises from $2 to $3 (an increase of 50%) and the price of shirts rises to $25 (an increase of 25%). Assuming that Tom’s total consumption and taste for burgers and shirts remains unchanged, then we might be tempted to apply unchanged weights to his consumption. If so then we infer that his purchasing power has fallen 42% (i.e., 50%*0.67 + 25%*0.33 = 0.42).
If we assume further that Tom is an “average” consumer then we might also be tempted to infer that from Period 1 to Period 2 the average purchasing power of money fell by 41.7%. If we also observe that total spending increased from $100 million in Period 1 to $140 million in Period 2 (i.e. by 40%), then, by applying the inference that the purchasing power of money fell by 42%, we infer that “real” spending stood at $99 million in Period 2 and therefore that it fell 1% between the two periods. Every 5-10 years government statisticians undertake extensive surveys in order to establish the “typical” consumer’s pattern of expenditure. The obtained weights serve as bases for inferences about (changes in) average prices and hence in the purchasing power of money. Once changes in the purchasing power of money are established one can estimate changes in total real output – and productivity.
So What’s Wrong With That?
This procedure is questionable on several grounds. Consider the assumption that weights remain constant over a prolonged period of time. This assumption invites us to accept not only that that individuals’ preferences are immutable over extended periods, not only that there exists a composite “average person but also that this person’s behaviour represents that of many other people. Rothbard, however, redirects our attention from the fictional composite to the concrete individual and reminds us that “there are only individual buyers, and each buyer has bought a different proportion and type of goods. If one person purchases a TV set, and another goes to the movies, each activity is the result of different value scales, and each has different effects on the various commodities. There is no ‘average person’ who goes partly to the movies and buys part of a TV set. There is therefore no ‘average housewife’ buying some given proportion of a totality of goods. Goods are not bought in their totality against money, but only by individuals in individual transactions, and therefore there can be no scientific method of combining them.”
If Rothbard is correct then the fixed weight price index has no basis in reality, the overall purchasing power of money cannot be established and hence we possess no valid and reliable means to estimate overall productivity. If Rothbard is correct, in other words, then our original dilemma remains unresolved. In his view similar points apply to variable weight price indexes: “all sorts of index numbers have been spawned in a vain attempt to surmount these difficulties: quantity weights have been chosen that vary for each year covered; arithmetical, geometrical, and harmonic averages have been taken at variable and fixed weights; ‘ideal’ formulas have been explored – all with no realization of the futility of these endeavours. No such index number, no attempt to separate and measure prices and quantities, can be valid.”
Measuring Productivity: Intractable Difficulties and Perfidious Consequences
Are Australians becoming more productive? Our first conclusion is iconoclastic: alas, we do not and cannot know the answer. This is because the concept of total labour productivity (as opposed to the productivity of an individual worker and the purchasing power of money in a single transaction), when stripped to its most fundamental elements, is fraught with a host of seemingly intractable difficulties. If the methods which underlie productivity statistics may be questioned, so too may the data which generate them. It is not a case of lies, damned lies and statistics; rather, official statistics must necessarily be estimates; and government agencies (whose staffs we have no reason to doubt are diligent and dedicated) revise these estimates as new, better and more complete information comes to hand. Accordingly, despite the best efforts of their compilers the validity and reliability of these statistics should not be exaggerated. Alas, they are usually received and interpreted as writ. As Oskar Morganstern wrote a generation ago in his classic work On the Accuracy of Economic Observations: “textbooks … show little if any evidence of the awareness of [government statistics’] difficulties and limitations. In Great Britain, as in the U.S. and elsewhere, statistics are still being interpreted as if their accuracy compared favourably with the measurement of the speed of light.”
In some instances economic statistics are revised repeatedly over several years, and the cumulative effect of these revisions can differ drastically from the initial estimate. Market participants who act upon initial estimates may thus make egregious mistakes. In the U.S., for example, according to The Australian Financial Review (3 August 2001), “a revision of numbers over the past 3 ½ years has cast doubt on the so-called ‘miracle economy’ so warmly embraced by investors and has highlighted the significant pressure on companies in the world’s largest share market.” An economist quoted in the article stated that “the rewriting of US economic data highlights a deterioration of fundamentals over the past few years beyond that of most expectations. The revisions were stunning for corporate America, showing that the recent past wasn’t as rosy as we had thought … The economic boom wasn’t as steamy, productivity growth was much slower and profit margins were squeezed much harder.” The article concluded that “the fiddling of statistics has undermined the concept of the ‘miracle economy’ and, given the flow of weak economic data, US investors may be in for more troubled times ahead.”
Premature, incomplete or erroneous official figures can also cause government policy-makers to err grievously. In this context Frank Shostak supplies another insight. He notes that as a rule the greater the credit created by central and commercial banks the larger the subsequent volume of economic activity. In his words, this “means that the rate of growth of what governments call ‘total real output’ will closely mirror rises in money supply. Essentially, the more money pumped the greater the ‘total output’ will be. But this is not real production but only a statistical illusion.” It follows that estimates of productivity in countries whose central and commercial banks have in recent years created vast amounts of credit not backed by savings, such as Australia, have likely and unintentionally been misleadingly optimistic. Hence a second stark conclusion: assuming (if not heralding) a “New Era” and cheered by the vast majority of businessmen, economists and market participants, interest rates have been forced far too low for far too long; and to the extent that the ebb-and-flow of the business cycle has been exacerbated then the “bust” stage of the process outlined in Interest Rates, Corporate Debt and the Business Cycle has by no means run its course.
 
Designed & maintained by
Artist Web Design
©1999-2008 All Rights Rese |
