How well do estimates of GDP and National Income measure economic performance?  These accounts, though well-suited for some purposes, have limitations that fall into four broad categories: (a) inaccuracy, (b) incompleteness, (c) misclassification, and (d) ambiguity.

 

Inaccuracy

            There is a tendency toward specious accuracy, a pretense that things have been counted more precisely than they can be. ... The classic case is, of course, the story in which a man, asked about the age of a river, states that it is 3,000,021 years old. Asked how he could give such accurate information, the answer was that twenty-one years ago the river's age was given as three million years.

Oskar Morgenstern

"Qui Numerare Incipit Errore Incipit"

“He who begins to count beings to err.”

 

            Just how accurate must GDP accounts be to be useful to planners and forecasters? Estimation errors are clearly biased. For example, many people systematically understate the incomes reported on their tax forms. Statisticians commonly assume that measurement errors are offsetting, Underreporting biases estimates of GDP downwards, especially for countries in which tax avoidance (legal) and tax evasion (illegal) are significant.

 

            Another problem is that most accounts are reported to the exact dollar. For example, per capita income for Utah was reported as $12,893. Would it be just as useful (and less misleading) to learn that per capita income in Utah was roughly $13,000?

 

Incompleteness: Nonmarket Transactions

 

            GDP is the value of all production so it should cover all outputs. National accounts focus on market transactions because dollar sales figures are reasonably available, but not all productive activity is marketed. Some nonmarket activities are included in GDP because crude estimates of their values are available (e.g., estimated rents for owner-occupied housing and estimates for food consumed by its producers). But the values of homemakers' services are excluded from GDP.  So, the GDP is increased when you hire someone to clean your home but not when a stay-at-home parent or self-employed consultant does it herself.  This holds true for other do-it-yourself services, such as homemade haircuts, or household repairs.

 

            Only crude estimates are available for unrecorded activities such as cash or barter transactions, so the government excluded these estimates until 1986. Illegal activities have traditionally been treated as socially unproductive, so they are excluded from GDP. But should legalization of gambling or marijuana lead to growth in measured GDP even if people's behavior doesn't change? Under current accounting practices, it would. Such problems mean that comparisons of GDP over time or among countries must be tempered by recognition that GDP accounts are affected by the relative importance of do-it-yourself production and barter, and by differences in laws and regulations.

 

Misclassification

 

            Some GDP items seem misclassified. For example, individual spending on education is treated as personal consumption. Football games, parties, and some frivolous courses may qualify as consumption activities. But time and effort spent studying that will increase your future productivity should be classified as investment, not consumption. (Should your body's deterioration as you age be considered "depreciation"?) Government investments in flood control, research and development, transportation networks, and so on are treated similarly. These investments are reported as government spending instead of investment. Classification in this way tends to understate the extent to which present consumption is diverted to activities that enhance our future productivity, output, growth, and well-being.

 

Ambiguity: Government Output

 

The government seldom directly charges prices for its services to final users, so the value of government output is somewhat ambiguous. Critics who view most government spending as waste might advocate discounting its purchases to more accurately measure the value of total GDP. On the other hand, people who view government as a bargain might think that GDP would be more accurate if government spending were assigned a premium. Consequently, National Income accountants use the best estimate available—out-of-pocket government spending—primarily, wages paid to government workers.

 

Ambiguity: International Comparisons

 

The cliché, "You can't compare apples and oranges," is often echoed in debate. Economists view such tasks as easier than this adage suggests—money prices provide a common denominator. But such comparisons get tougher when different types of money are used to do the pricing.

 

The Exchange Rate Approach

 

            Currencies (different types of money) of most countries are traded in international financial markets 24 hours each day at exchange rates driven primarily by international supplies and demands of these currencies.

An exchange rate is the relative price of one currency in terms of another.

 

            Exchange rates are quoted in the financial sections of most newspapers.  An exchange rate for the German deutsche mark (DM) of 2.00 per dollar, for example, would mean that each DM is worth $0.50, and that $1 will buy 2DM.

           

            The easiest approach for comparing GDP between countries is to use exchange rates for conversion to a common denominator. Thus, if German per capita GDP is 20,000DM and the exchange rate is 2DM per dollar, then 20,000DM = $10,000 per capita income in Germany, when converted into dollars. The exchange rate approach has been used for decades, primarily because of its computational convenience. Unfortunately, this approach often yields very misleading results.

 

            If the international economy and all domestic economies were perfectly competitive, exchange rates would adjust so that the price of any single good or resource or service would be identical everywhere in the world. In other words, if transactions costs were zero, price deviations for any given good would be zero, a concept is known as the law of one price.

 

            The Law of One Price assumes that transaction costs are zero and concludes that only one price can exist at any given moment for a homogeneous item. Absent transactions costs, a haircut in Bombay would cost the same, after converting Indian rupees for U.S. dollars, as the same haircut in Baltimore. The same would be true of computer discs or heads of lettuce. The real cost of living would be the same everywhere, after adjusting for such amenities as weather and scenic beauty.

 

            In the real world, however, transactions costs are not zero, markets are imperfectly competitive, governments intervene to set exchange rates artificially, and quotas and tariffs hinder free flows of goods and resources across international borders. Consequently, international exchange rates for currencies often fail to reflect the relative costs of living between countries.

 

The Purchasing Power Parity (PPP) Approach

 

Growing recognition of flaws in the exchange rate approach have lead researchers to adjust GDPs according to relative costs of living. If such basic goods as food, clothing, and shelter (housing) cost much more (after exchange rate conversion) in one country (say, Japan) than in another (say, Brazil), then the well-being of the Brazilians relative to the Japanese is understated by the exchange rate approach, while Japanese well-being is relatively overstated.

 

The purchasing power parity (PPP) approach adjusts exchange rate conversions for differences in costs of living before comparing GDPs between countries.

Table 6 illustrates the shifts in relative magnitudes that occur when per capita GDPs are adjusted by the PPP approach. Chinese per capita GDP, for example, is much greater after adjustments for purchasing power parity than if exchange rates alone are used for conversion.

 

Table 6  Exchange Rates vs. Purchasing Power Parity Adjustments to GDPs in Developing Countries
1992	$GDP per head, market exchange rates	$ GDP per head, purchasing–power parity	Total GDP, $B of purchasing- power parity
China	370	2,460	2,870
India	275	1,255	1,105
Brazil	2,525	4,940	770
Mexico	3,700	6,590	590
Indonesia	650	2,770	510
South Korea	6,790	8,635	380
Thailand	1,780	5,580	320
Pakistan	400	2,075	240
Argentina	6,870	5,930	190
Nigeria	275	1,560	190
Egypt	655	3,350	180
Philippines	820	2,400	155
Malaysia	2,980	7,110	130
Disparate per capita GDPs between developed and developing economies tend to be far more pronounced when exchange rates are used for conversions than if purchasing power parity is taken into account. These differences exist because costs of living in less developed countries are far less than the exchange rates of their currencies would suggest; i.e., exchange rates for currencies from less developed nations tend to be undervalued.
Sources: World Economic Outlook 1993, International Monetary Fund, EIU; World Bank; OECD, as reported in "Chinese Puzzles, The Economist, May 15, 1993, p. 83.

 

            Conversion according to purchasing power parity is still in its infancy, largely because data to measure consumer price indices are very crude, at best, in many developing nations. The importance of refining such statistical measures is attested to by the fact that, using the PPP technique, many analysts now forecast the possibility that, if its recent 10+ percent annual growth rates continue, the Chinese economy, now second in the world (as shown in Figure 5), may surpass that of the United States to become the world's largest early next century. (A 10 percent growth rate, compounded, doubles an economy in a bit more than 7 years.) Because the exchange rate approach reflects an artificially low exchange rate for Chinese currency relative to other currencies, such forecasts are contradicted when using the traditional exchange rate approach—the current Chinese GDP is much less when adjusted by exchange rates than when adjustments are made under the PPP approach.

Figure 5   The World's Largest Economies in 1992

MISSING!

             (trillions of dollars of GDP)

 

The GDPs of many developing countries are closer to those of developed nations when purchasing power parity is used to adjust exchange rates than when exchange rates are not adjusted.

 

Source:  World Economic Outlook, International Monetary Fund, 1993, as summarized in "Chinese Puzzles," The Economist, May 15, 1993, p. 83 

 

ED:  1 column figure—We will adjust colors and try to update.

 

            GDP accounts were created to measure economic performance. These accounts may be misleading when used to weigh well-being in various countries—especially if the relative amounts of self-production, barter, or the underground economy vary much among countries, or if their relative costs of living are out of synch with the relative values of their respective exchange rates. Nuances in GDP accounting and the qualifiers created by the problems we've described cause some critics to assert that these aggregate data are worthless. It may be that parts of our measures of GDP and some of its relatives are a bit like the following example.

 

            Imagine that all your classmates were transported back in time to around 1803. Each is assigned by President Thomas Jefferson to travel to various parts of North America and then to return to Washington, D.C. with estimates of distances between points. To standardize measurements, everyone is to pace the distances. People walking to New York or Boston would cross-check each other, ensuring reasonable estimates for such short distances. But  stride lengths differ, some people might wander in circles, and others might guess at the distance while riding in wagons. Still others might not even go to faraway destinations but would fill in travel vouchers as if they had. The figures turned in to President Jefferson might resemble parts of GDP accounting—far from perfect, but still better than no data at all.