Imaginary Numbers

Yesterday we came across a Reuters report on the rather pedestrian, to put it mildly, third quarter GDP report. Reported “real” GDP was greatly helped by the GDP deflator once again declining sharply, coming in a full percentage point lower than in the second quarter (1.2% instead of 2.2%). Perceptive readers have probably noticed this pattern already: weak GDP numbers always tend to coincide with outsized declines in the deflator.

A basic problem in this context is that “price inflation” in terms of a general price level cannot possible be measured; or let us rather say, such a measurement simply makes no sense, neither mathematically nor logically.

PCE chain-type price index (2009=100), annual rate of change

The main problem is that there is no fixed yardstick which can be used for measurement, as the value of money depends on supply and demand just as other goods do. In other words, one continually changing magnitude is measured with the help of another continually changing magnitude. This is basically nonsense. Moreover, there is an array of heterogeneous goods and services with an array of prices. Adding up cars, potatoes, movie tickets, rents, and so forth, making an “average” out of the result and calling it the “general level of prices” is absurd.

Just as various new calculation methods that have been adopted over the years have resulted in CPI and similar indicators (PCE is used for deflating GDP) becoming ever lower on average, the same methods have resulted in ever higher real GDP numbers. Needless to say, both effects have the side effect of tending to flatter the government’s economic policies and the Fed’s monetary interventions. Surely that is just a coincidence (cough cough).

The effects on GDP are such that it is fair to say that we are dealing with a complete fantasy number. Whatever it reflects, it is not reality and its connection with “economic growth” seems almost coincidental. This is beside the fact that GDP fails to account for the vast bulk of the economy’s production structure, as we have often discussed in these pages. If one wants to know about the actual distribution of economic activity, one needs to consult gross output data.

One effect of this is that many people erroneously assume that consumption is the biggest part of economic activity. It should be clear purely from a common sense perspective that this cannot be true: if we were to continually consume more than we produce, we would soon be living from hand to mouth.

“Unreal” GDP, calculated with great precision by the government’s statistics minions.

We have provided an extensive discussion of the pure fantasy numbers that make up GDP on occasion of the last revision of the calculation (even more fantasy numbers were included, which once again ended up flattering…oh well, you know the drill). We strongly recommend that readers take the time to check said article out in its entirety: “The Mirage of Economic Growth”. Here was just want to quote a few passages discussing especially glaring examples of GDP components that essentially consist of completely imaginary numbers:

“There are for one thing so-called ‘imputations’. These represent the ‘imputed value’ of services consumers get for free even though they apparently shouldn’t. For instance, if checking services are offered for free by banks to customers opening a current account with them, then the government adds the supposed value of these free services to GDP. Note here that no money has actually changed hands, but what is added to GDP are in fact money terms.

Another area in which imaginary numbers play an ever bigger role is the ‘hedonic indexing’ applied to all sorts of goods, something that has an especially large effect on all things to do with information technology. Here is an example from the second quarter of 2003 illustrating the effect (we have chosen this time period randomly, mainly because we happen to have the exact data at our fingertips. It should be pointed out though that the error in this data series compounds over time). In Q2 of 2003, actual spending on computers increased by $6.3 billion, from $76.3 billion to $82.6 billion.

If simply ‘every monetary transaction’ were added to GDP, then this is the number that would have been added, and thereafter it would have been massaged by the ‘deflator’. If not for hedonic indexing, that is. Before we tell you, try to guess how big an increase in spending on computers the government actually added to GDP in this instance. Was it 20% larger? 30%? Maybe even 50%? Hold on to your hat.

The number added by government to GDP instead of the $6.3 billion in actual additional spending was $38.2 billion. In other words, almost $32 billion in completely imaginary money that no-one ever spent or received, with the total number  used by the government amounting to more than 6 times the actual spending growth was used for the calculation of ‘real GDP’. It should probably be renamed ‘unreal GDP’.

And this is actually just a small slice of what is wrong with GDP. As we always stress, GDP is easily the by far most useless aggregate economic statistic that exists.

Why Calculate It At All?

Given that GDP is such a dubious statistic (the same is true of many other economic statistics as well actually), why even bother calculating it? Well, who is doing the calculation is actually already a strong hint. In an unhampered free market economy, it seems highly unlikely that a lot of time and effort would be wasted on compiling such statistics. Academic researchers (primarily historians) may have some interest in “macro” type data, and if this demand were strong enough to make the endeavor economically viable, someone would no doubt provide them.

Other than that, there may also be some private demand for certain industry-specific data, but even that demand would be highly dependent on whether fractional reserve banking and the associated credit expansions would actually be tolerated in such a hypothetical unhampered free market economy (as we have previously pointed out, there is a strong argument to be made that the practice of fractional reserve banking represents a flagrant violation of property rights).

For instance, a recent article at Wolf Street discusses the year-to-date collapse in the US load-to-truck ratio (another one of those recent “everything is awesome” economic statistics). This kind of data would undoubtedly be interesting to trucking companies and diesel engine makers. However, the extreme cyclicality of the ratio is solely a result of the business cycle (the companies in this business are undoubtedly fully aware of any purely seasonal effects). The business cycle in turn has monetary causes, i.e., it is a result of credit expansion. As Mises has pointed out, all attempts to assign non-monetary causes to the business cycle fall flat on even the most cursory examination.

The US load-to-truck ratio, via Wolf Street.

The only reason to compile assorted macro-economic statistics is for a) propaganda purposes and b) more importantly, to provide a justification for government meddling with the economy. Often they describe the atrocious results of previous interventions, which are then used as the justification for even more interventions. The Federal Reserve operates exclusively on this principle for instance. It first mucks the economy up, and the results are then presented as the reason for doing it all over again, only on an even grander scale.

This is something Sir John Cowperthwaite, the former governor of Hong Kong, was well aware of. When a UK delegation of bureaucrats and politicians came to visit him, they asked whether he could provide them with Hong Kong’s unemployment data. Cowperthwaite essentially told them (we are paraphrasing): “Sorry, we don’t collect such data. They would merely tempt government to meddle with the economy”. Amen.

Sir John James Cowperthwaite, the former governor of Hong Kong. When he started out in this job, the place was a sleepy fishing village largely made of wooden huts. Today it is one of the most free and richest economies in the world, something Sir John’s refusal to meddle with Hong Kong’s economy certainly had a lot to do with.

Photo via Wikimedia Commons

Undetected Ironies

Let us briefly get back to the above mentioned Reuters article on the Q3 GDP report. It contains a few interesting/funny details.

“U.S. economic growth braked sharply in the third quarter as businesses cut back on restocking warehouses to work off an inventory glut, but solid domestic demand could encourage the Federal Reserve to raise interest rates in December.

Gross domestic product increased at a 1.5 percent annual rate after expanding at a 3.9 percent clip in the second quarter, the Commerce Department said on Thursday.

The inventory drag, however, is likely to be temporary and economists expect growth to pick up in the fourth quarter given strong domestic fundamentals.”

Several paragraphs later, we are informed that inventories actually didn’t decline, except in selected sectors; overall, they merely grew at a slower pace than previously:

Businesses accumulated $56.8 billion worth of inventory in the third quarter, the smallest since the first quarter of 2014 and down sharply from $113.5 billion in the April-June period. There were declines in manufacturing, wholesale and retail inventories.

This was described as “healthy purge” by one observer! We believe the actual “purge” is actually still ahead. The article of course couldn’t fail to neglect mentioning the “consumption is everything” fallacy:

“The blow from inventories was, however, blunted by bullish consumers, who are getting a tailwind from cheaper gasoline and firming housing and labor markets. Consumer spending, which accounts for more than two-thirds of U.S. economic activity, grew at a 3.2 percent rate after expanding at a 3.6 percent pace in the second quarter. A measure of private domestic demand, which excludes trade, inventories and government spending, rose at a sturdy 3.2 percent pace.

“The consumer remains the main engine of economic growth. We expect this dynamic to remain in place,” said Jesse Hurwitz, an economist at Barclays in New York.

Sure enough, in GDP accounting, consumption is the largest component. However, this is (luckily) far from the economic reality. Naturally, it is not possible to consume oneself to prosperity. The ability to consume more is the result of growing prosperity, not its cause. But this is the kind of deranged economic reasoning that is par for the course for today: let’s put the cart before the horse!

Lastly, Reuters’ interview partners appear not have noticed the irony of the following statement:

“A separate report from the Labor Department showed new applications for unemployment benefits last week hovering near levels last seen in late 1973.”

Does anyone still remember what happened right after “late 1973”? We recall that what was then the worst recession and bear market since the Great Depression started more or less immediately after these seemingly uplifting data were recorded. As we have pointed out previously, extreme lows in unemployment claims are historically a contrary indicator (see also this chart). It seems that was true in 1973 as well.

Click on image to enlarge

S&P 500, weekly, during the 1970s. Maybe the “lowest unemployment claims since late 1973” aren’t such good news after all … 

Finally, it is mentioned that “trade was not a drag” on GDP this time. The idea that trade surpluses and trade deficits add to, resp. detract from economic growth is a mercantilist fallacy. Again, it is true in terms of GDP accounting, but it has been known for some time already that the balance of trade is not a yardstick of a nation’s prosperity. This is yet another part of this statistic that doesn’t make much sense.

“With the dollar strengthening, export growth decelerated in the third quarter. The drag was, however, offset by a slowdown in imports, especially automobiles, leaving trade’s impact on growth neutral.”

US citizens should of course be glad that their currency is strengthening for a change at the moment. Contrary to the poor rubes in Europe and Japan, who are increasingly impoverished by their central bankers having declared open season on the currencies they issue, they are actually gaining a little purchasing power. This means they must produce less to obtain the same amount of goods from abroad as before. Only the morons running modern central banks can possibly believe that this is a bad thing.

Conclusion

GDP remains as useless a statistic as ever. The attempts by assorted observers to make sense of these data are almost comical in their futility. Similar to many other macro-economic data series, there may be some usefulness to GDP data in terms of allowing historical comparisons, but even the historical trends have to be taken with a grain of salt given the many changes that have been implemented over the years in the calculation methods of various adjustment factors. Moreover, as Mish has pointed out, the initial guesstimate of the number is going to be revised several times, which makes it even more meaningless.

In a truly free market economy, there would be little point in bothering with such data beyond their possible historical interest. One must never lose sight of the fact that today’s widespread data obsession is in fact mostly driven by the fact that an ever larger share of the economy is subject to government interference and various forms of central planning.

Charts by: St. Louis Federal Rserve Research, DAT/Wolf Street, Chartstore