Tuesday, 19 May 2015

How GDP Metrics Distort Our View of the Economy

Everyone thinks that GDP measures all of a country’s economic activity. But everyone who thinks that would be wrong …
Guest post by Christopher Casey


GDP purports to measure economic activity while largely divorcing itself from the quality, profitability, depth, breadth, improvement, advancement, and rationalization of goods and services provided.

For example, even if a ship — built at great expense — cruised without passengers, fished without success, or ferried without cargo; it nevertheless contributed to GDP. Profitable for investors or stranded in the sand; it added to GDP. Plying the seas or rusting into an orange honeycomb shell; the nation’s GDP grew.1

Stated alternatively, GDP fails to accurately assess the value of goods and services provided or estimate a society’s standard of living. It is a ruler with irregular hash marks and a clock with erratic ticks.

As proof, observe this absurdity: in 1990, Soviet GDP equalled half of US GDP, according to the 1991 CIA Factbook. No one visiting the Soviet Union in 1990 would believe their economy came close to 50 percent of the quality and quantity of the goods and services produced in America. GDP-defined production may have been strong, but laying roads to nowhere, smelting unusable steel, and baking barely edible breads stretches the definition of “production.” And this describes the goods which were actually produced. There is no accounting for the opportunity cost of forfeited essential goods and services.

How can this be? Why does GDP poorly reflect economic size and vitality? The blame largely resides with three fallacious concepts embedded within GDP “measurements”:

    1. intermediate goods (e.g., steel) must be eliminated to avoid “double counting”;
    2. government expenditures consist of viable economic activities; and
    3. imports should be netted against exports.

Let’s look at each in turn.

1. The Overstatement of Consumption

GDP stands for Gross Domestic Production. It might more accurately be called Net Domestic Consumption.

Which transactions should be included within GDP? Since most products consist of other products, the architects of GDP attempted to avoid “double counting” transactions by largely including only “final” goods and services produced – the ones you see on your shelves. By their methods, the production of a car is counted (as an increase in inventory), but the metal, rubber, and plastic purchased in its creation is not. But the rules behind what makes a transaction “final” are arbitrary. The logic could just as easily justify including the sale of an automobile to a consumer and disregarding its previous production.

Intermediate goods are produced too, metal, rubber, plastic, but GDP’s gurus just don’t measure them. They’re already included, they say, in their measurements of “final” goods. But they’re not. They represent economic activity that is simply unmeasured.

In addition, any “final” transaction during a given time period does not necessarily include in it the intermediate goods that are produced in that same time period: metal, rubber, and plastic purchased today will likely be for a different car produced or sold in a different (future) time period.

Regardless as to the arbitrary nature of determining final sales and notwithstanding the problem of temporally matching intermediate goods with their associated final sales, the exclusion of certain “intermediate” transactions simply excludes massive volumes of economic activity. Thus, GDP understates the economy as a whole while grossly overstating its consumption component relative to business investment.

Thus, it allows ignoramuses to say that NZers rebuilding after an earthquake and selling houses to each other are the foundations for a “rock star” economy!

A better measure of overall production was created in 2014 when the US Commerce Department began publishing Gross Output which does incorporate intermediate transactions. Using GDP, ignorant commentators  commonly cite the fantasy statistic that consumption accounts for 70 percent of all economic activity. Using Gross Output, we see that consumption  represents a mere 40 percent.

2. The Treatment of Government Expenditures as Productive

If the GDP figure purports to measures production, then why does it include government spending?

If GDP purports to measure economic activity which benefits society, the inclusion of government expenditures is dubious – especially because of the temptation for governments to goose the figures with extra spending they can’t afford (we’re looking at you, every finance minister in the western world).

If all government spending is included no matter the outcome, then GDP “produced” in the Soviet Union is no different than GDP “produced” by any government — the difference is but one of scale.

In this sense, reducing GDP would actually increase economic growth. Because, frankly, all government spending is not production but consumption. For as George Reisman notes:

In the same sense as a housewife, the government is not a producer but a consumer, who is dependent upon producers.  All of its physical production, like hers, is in the last analysis a consumptive production. It is a production which cannot replace the means with which it began ...  a production which leaves the government poorer by the amount of funds it has expended.  In order to continue the activity, resort must be had to an external source of funds -- in the government's case, the taxpayers or the printing press.

And all government spending is to some degree malinvestment, for as Murray Rothbard noted:

Spending only measures value of output in the private economy because that spending is voluntary for services rendered. In government, the situation is entirely different ... its spending has no necessary relation to the services that it might be providing to the private sector. There is no way, in fact, to gauge [the monetary value] these services.

Without voluntary action, the price system is impotent; without true price discovery, true monetary value cannot be ascertained. True price discovery is impossible with government spending.  This does not mean all goods and services provided by government would cease to exist; rather, some production (e.g., hospitals, schools, roads, etc.) would revert to the private sector (and then nurses and teachers might be paid what they deserve).

When measured (and produced) this way, the true government contribution to GDP may be positive but overstated (it currently approximates 20 percent of US GDP) – it also offers a flawed basis for a modern form voodoo economics that claims government deficits actually encourage saving. A more accurate depiction of economic activity would reduce if not eliminate altogether both voodoo economics and the contribution of government expenditures.

Or perhaps, as Murray Rothbard argued, the higher of government receipts or expenditures should actually be deducted from GDP since “all government spending is a clear depredation upon, rather than an addition” to the economy.

3. The Problems of Subtracting Imports from Exports

Way to ignore the benefits of trade, guys …

As Robert Murphy has noted several times, the netting of imports against exports in determining GDP seriously understates the contribution of trade to overall economic activity. To wit, an economy which exports $1 and imports $1 will have the same GDP contribution (zero) as one which exports $100 billion and imports $100 billion. Obviously, the latter economy would be far worse off with the sudden cessation of trade.

A fixture of GDP is the mercantilist mentality of treating exports positively and imports negatively. Why are exports additive to GDP while imports are deductive? If the goal of GDP is to measure the goods and services provided to people within a geographic region, then imports are the real benefit— not exports. Exports are but payment for imports.

The problem and confusion arises because the GDP calculation unrealistically excludes other forms of payment: it should make a difference if imports are funded with increasing debt levels or if funds are accumulated from previous years of compensated exports. If China converted over $1 trillion in US debt instruments into imports of American goods and services, then its people benefit today; but under GDP accounting, the negative impact of imports would offset greater consumption and/or government spending (the increase in GDP was previously realised in the years during which exports created a trade surplus).

Conclusion: GDP Advances the Failed Keynesian Agenda

Yes,  GDP does advance a failed agenda …

Simon Kuznets (1901–1985) revolutionised econometrics and standardized measurements of GDP, with his research culminating in his 1941 book, National Income and Its Composition, 1919–1938. While not a Keynesian per se, since central planning requires economic statistics the nature and timing of his research fuelled the Keynesian revolution. As Murray Rothbard noted:

Statistics are the eyes and ears of the bureaucrat, the politician, the socialistic reformer. Only by statistics can they know, or at least have any idea about, what is going on in the economy. Only by statistics can they find out ... who “needs” what throughout the economy, and how much government money should be channelled in what directions.

Further, the overstatement of consumption perfectly matches the Keynesian agenda, which happily dismisses the savings necessary for the production of the intermediate goods the GDP measurement system ignores.

The fact is, GDP’s faulty theoretical underpinnings and politically-motivated acceptance distort the performance and nature of an economy while failing to satisfactorily estimate a society’s standard of living. In fact, Kuznets himself partially understood this. In his very first report to the US Congress in 1934, Kuznets said “the welfare of a nation [can] scarcely be inferred from a measure of national income.”

Yet the blind usage of GDP persists. That its permanence and persistence only serves the Keynesian policies of greater consumer spending, increased government expenditures, and larger exports through currency debasement should not be considered coincidental. Unfortunately, the resulting economic stagnation, debt accumulation, and price inflation are as inevitable as they are predictable.



  • 1.Starting in December of 1991, the Bureau of Economic Analysis (BEA) of the U.S. Department of Commerce emphasised gross domestic product (GDP) over that of gross national product (GNP) as a measurement of production within the U.S. The difference between GNP and GDP lies in the treatment of income from foreign sources: GNP measures the value of goods and services produced by U.S. nationals, while GDP measures the value of goods and services produced within the boundaries of the U.S., regardless as to the nationality of ownership. For purposes of this article, the differences between each measurement are unimportant and therefore “GDP” is utilised synonymously with GNP.

1 comment:

Sarah Wilshaw-Sparkes said...

Another shortcoming of GDP is that it ignores unpaid labour ie it only measures paid "market" transactions. The idea that NZ would become more productive if our unpaid volunteers and homemakers could be enticed into paid work flows directly from this flaw. Those who think unpaid labour has no value to the economy should try plugging the gap left by a homemaker who is sick/dead/in full time paid employment.

How Reisman can equate a housewife with the government and suggest she is a consumer dependent on producers is beyond me. When the 'main breadwinner' comes home to a hot meal, clean clothes and a full tube of toothpaste I suggest said breadwinner is a consumer of what the housewife has produced.

Kuznets, one of the architects of GDP, was well aware, as you note near the end of the post, that the measure was a specific tool for a specific use that necessarily excluded a range of valuable activities like housecare, and he worried, seemingly rightly, that it would be misapplied.