Economic Statistics: Good,
Bad, or Indifferent? by Carole E. Scott

Carole Scott is a Professor of Economics at the State University of West Georgia and Editor-in-Chief of B>Quest.

"For data users, confidence in official statistics is very much a matter of their confidence in the objectivity and professionalism of the producing agency."[McLenaghan, 43]

"If GNP and the investment rate did not exist, it would be necessary to invent them. Indeed, it has been necessary to invent them....Economists have repeatedly recognized the shortcomings of GNP as a measure of welfare and of the investment rate as the propellant of economic growth. But to little avail....

National accounts exist because national governments have believed them useful for governmental economic control; they reflect the expanding ambitions of governments....Many [economists] are on public payrolls, directly or indirectly, and their ambitions frequently find fulfillment in the business of national government. Consequently their receptivity to fiction is abnormally developed." [Rimmer, 13, 16]

"There are three kinds of lies: lies, damned lies, and statistics." - Benjamin Disraeli

Following the lead of other nations, a few years ago the United States replaced GNP (Gross National Product) with Gross Domestic Product (GDP) as the measure of the value of the nation's output of goods and services. The difference between them is that, while the latter measures all the output produced domestically (within the borders of the U.S.), the former does not, as it only measures the output of firms owned by Americans. However, GNP includes the output of American-owned firms located abroad, while GDP does not. Due to the fact that in recent years domestic savings have been inadequate to finance the desired level of domestic investment, GDP and GNP may in the future increasingly diverge because of a resulting rise in foreign ownership of firms located in the U.S.

GDP is the best known of the National Income and Product Accounts. GDP is a summary measure used to measure national well being in a quantified manner. The National Income and Product Accounts describe the relationship between income and product (output) and trace the principal economic flows among the major sectors of the economy. Many economists today complain that the federal government spends too little on collecting this and related data and, as a result, it is not adequate or accurate enough.

GDP and other macroeconomic statistics have a significant impact on governments' and businesses' plans for the future, and the press and the public look to them as a primary indicators of how well the country is doing. If, according to them, the country is not doing well during a presidential election year, many believe the incumbent president has little chance of being reelected, as "the assumption that the President manages the economy is the core of prevailing political wisdom, dinned into the public mind by a generation of pundits, a convention of discourse endlessly repeated but rarely examined. The fact is, presidential elections have become referenda on the business cycle, whose fortuitous turnings are personified in the President--thus the 'Bush recession' yields to the 'Clinton recovery'. This is not economics, it is anthropology--an exercise in collective magic." [Morris, 49]

Improving the quality of macroeconomic data is today a high priority issue throughout the world. Concern on the part of the International Monetary Fund (IMF) arising out of the international financial crisis of late 1994 through early 1995 led to the IMF pressing the nations of the world to improve their data dissemination practices because it believes this will reduce the potential for the spreading of a future financial crisis.

The increasingly globalized economy, evidenced by the major expansion of the international capital markets through the 1980's and the early 1990's and the rapid increase in the number of countries participating in those markets, has drawn attention to the potentially destabilizing influences of sudden and large changes in the direction of capital movements. This in turn has highlighted the importance of ensuring that market participants-- whose transactions take many forms and whose interests may well diverge--are able to make decisions on the basis of timely and good quality information on developments and prospects in individual countries. The growth of around-the-clock electronic trading in financial instruments among an increasing number of countries, both the established financial centers and the emerging market countries, has given prominence to the importance of dissemination of economic and financial data by countries and the means by which those data can be accessed. [McLenaghan, 39]

A major reason why the data provided by National Income accounting is considered to be so very important is that it is essential to those who want to employ Keynesian economic theory because, according to this theory, the government can and should micro-manage the economy through the use of fiscal and monetary policy. According to Keynesian theory, effective monetary and fiscal policy requires this data and that it be very reliable, timely, and describe a broad range of economic activity. Otherwise, government management of the economy would be like driving a car with no speedometer and a blacked-out windshield. (Fiscal policy is concerned with government spending and revenues, while monetary policy is concerned with the supply and cost of money and credit.)

Some, however, are skeptical of the validity of this theory. "After a period of eclipse during the Depression, the scientific pretensions of American economics and its sister social studies were powerfully reinforced by the sweeping triumph of Keynesianism. Ignoring Keynes's [sic] own warnings about the waywardness of real markets, American academics forged a rigidly mechanistic vision of the economic apparatus: pull this lever and investment rises, turn this flywheel and consumption goes up--all the pieces clicking smoothly into place like stainless-steel tumblers. Faith in deus in machina prompted John Kennedy's wildly unprescient declaration in 1962 that there were no ideological issues left to solve; the country faced only 'technological problems...administrative problems'." [Morris, 50]

The Difficulties Involved In Measuring The Nation's Output

It's easy for companies like Ford Motor Company to accurately measure its output of automobiles. Counting every one that comes off the assembly line is not difficult or expensive, either absolutely or relative to Ford's total costs. But how about counting everything that is produced in this country? This is neither easy nor inexpensive.

Ford can directly observe how many cars it produces. Other than going from door to door asking, which is very expensive even for a modest sample survey, how could you count the number of meals prepared during the year in the nation's homes?

Meals prepared in the home have no market price because they are not sold; so how can a monetary value be placed upon them? Because it is not feasible to reasonably value meals prepared in the home, unlike restaurant meals, meals prepared in the home are not included in GDP. Because a larger percent of meals are eaten outside the home today than used to be the case, some of the increase in GDP we have experienced in recent decades is fallacious.

If a family washes its clothes at home in its washing machine, GDP is unaffected. If, instead, it takes them to a laundromat, it is. If it gets around in a family-owned car, GDP is unaffected. If it rents the car it uses, it is. If it visits Disneyland, GDP is affected. If it visits a National Park, it isn't.

Measuring the nation's output by the number of units produced would not be meaningful, because 9 cars and one paper clip would add the same number of units, 10, to total output as 1 car and 9 paper clips. To deal with this problem, GDP is measured in dollars. That is, a car adds, say, $22,000 to the value of output, while the paper clip adds one cent.

In solving the problem of adding apples and oranges by measuring the output of each on the basis of how much was spent to buy them, another one is created, because the price level changes over time. Therefore, if output did not change in terms of the number of units produced or in its composition from one year to the next, it would probably have a different dollar value because the average price per unit of output changed. To deal with this, government economists estimate the price level in order to adjust GDP measured at the prices things actually sold for (actual GDP). GDP adjusted for changes in the price level is called real or constant-dollar GDP.

To compute its profit, Ford subtracts from its revenues the cost of the resources it uses up in producing its cars. GDP, however, is not reduced because gasoline produced from domestic oil reduces the amount of the nation's oil reserves. A hurricane may destroy a tremendous amount of property, but GDP is not reduced to reflect this loss. Instead, a hurricane increases GDP by the amount of money spent to repair the damage. An increase in crime reduces the quality of life in the country, but it will increase GDP by the amount it costs to apprehend, prosecute, and imprison criminals.

An electronic calculator is vastly superior to a slide rule; yet, despite the rise in the price level, many electronic calculators cost less than a good slide rule cost back when they were used. Therefore, even though it is far more useful, an electronic calculator produced today adds less to GDP than a slide rule did in the past.

Because people don't have to buy a car, and if they do, because there are other car manufacturers, they don't have to buy a Ford, what people pay for Fords doesn't exceed what they are worth to the buyer. The same isn't necessarily true of government-provided services because the public is forced to pay for them through taxes, regardless of the valuation the public places on government-provided services. (The ballot box is the only way the public can do anything about the cost of government-provided services exceeding their value to the public.)

While what Fords sell for and how many of them are sold determines how much what Ford produces adds to GDP, how much the government spends determines how much what it produces adds to GDP. So, the more money the government wastes, the higher GDP will be.

In measuring the output of services industries, there is a conceptual problem that does not exist with manufacturing. "Quantity indexes usually are not available--in fact, there often is no physical 'good' to count--and prices also are not observed directly. To derive 'real output,' therefore,.." what is done is that an extrapolation "using indirect measures of input costs as proxies for price changes" is employed. In some industries, "...such as banking and recreation, measures of inputs are used--often number of employees--to proxy for changes in the level of production." [Schmidt, 2] The accuracy of this method of measuring output has often been questioned, and some believe that it has produced a spurious conclusion that productivity growth in services has been relatively low.

Macroeconomic Statistics' Accuracy And Relevance Is Questioned

Economists and others have been complaining for years that GDP and the price indexes (Implicit GDP Deflator, Consumer Price Index, and Producers Price Index) that are used to adjust total spending (GDP) and its components, consumer and business spending, for changes in the price level from year to year are inadequate. The questions raised fall into three categories: (1) Is what is being measured being measured accurately and appropriately? (2) Is what is being measured what is supposed to be measured? (3) Is what is supposed to be measured what should be measured?

GDP measures in dollars the value of each year's output of final goods and services. GDP is measured in both current- and constant-value dollars. When measured in current dollars, GDP is valued at the number of dollars the goods and services whose value it measures sold for. It is measured in constant-value dollars when, using the Implicit GDP Deflator, the rise in its value since a base (reference) period due to inflation is subtracted. The more meaningful inflation-adjusted GDP is called real GDP. (By removing that part of an increase in the value of GDP due to a rise in the price level, any remaining increase is the result of an increase in the quantity and/or quality of goods and services produced. The best measure of the material well being of the public is per capita real GDP.)

GDP is computed in two different ways: (1) the sum of what is spent by U.S. consumers, businesses, and government and foreigners to buy what the U.S. produces, and (2) the income earned by Americans for producing the nation's output of goods and services. This income is either spent, saved, or paid to the government as taxes. The estimate of the value of GDP computed by adding up what is spent to buy it uses data collected by the Bureau of the Census and other organizations, while the estimate of its value based on the income earned by those producing it is based on the results of a number of income surveys. These two figures can differ. If they do, the difference has to be reconciled.

The Commerce Department's Bureau of Economic Analysis [BEA] is the U.S. government agency responsible for making quarterly estimates of the size of GDP and the related measures of economic performance that comprise the National Income and Product Accounts. Unfortunately, according to William Beach, former chief economist for the Spring Corporation, "...regular quarterly estimates of gross domestic product (GDP)...have been off by an average of 50 percent during the nearly 20-year period beginning in 1976." [Heritage Foundation, 1]

"The federal government," Beach complains, "is forcing businesses to make decisions about jobs and investment based on faulty estimates of the economy's strength. In this regard, the GDP forecasts [of it] do more harm than good." [Heritage Foundation, 1] He recommends the BEA delay for at least 60 days the release of this estimate so that, because additional data will then be available, it will be more accurate. (It should be emphasized that what is being forecast is an estimate of the size of GDP, as GDP is not something we can "count" like how many cattle go through a gate. In short, what the BEA does is forecast quarterly, on the basis of less than all the data it will ultimately have, what it then will estimate is the size of GDP.)

GDP is estimated from a sample of sales collected monthly. Quarterly estimates of GDP are made well before all this data has been collected. As a result, initial estimates of GDP sometimes differ significantly from later estimates. Even relatively small, subsequent revisions in these estimates as more data becomes available, can significantly alter the estimated annual rate of growth of GDP. According to later, revised estimates, twelve percent of the time from 1978 to 1991, the advance estimate of the annual rate at which GDP was growing or shrinking was either positive when it should have been negative or vice versa. The estimated annualized GDP growth rate for the first quarter of 1992 was only 2.0 percent. Subsequently it was revised to 4.7 percent. During the Reagan administration the BEA underestimated the growth rate of GDP 63 percent of the time. [Beach, 1]

Despite their lack of accuracy, advance estimates of the growth rate of GDP are "quickly seized upon by politicians of both parties either to support their own records or to attack those of their opponents. And unlike the publication of advance estimates, revisions of these quarterly data by the BEA hardly make the news. This process of correcting and finalizing GDP estimates can last for months--even years--following that initial release and can result in huge revisions which tell a completely different story about how well the economy performed." [Beach, 3]

In light of this, it is reasonable to ask why quarterly estimates are made and released to the public. Would a drug company be allowed to continuously publish equally inaccurate data about its drugs? No matter how unimportant its product or service, would any business be allowed to make available to the public figures so likely to be significantly inaccurate?

Are National Income Accounting Statistics Appropriate And Accurate?

"In view of the conceptual difficulties of defining and drawing welfare inferences from GNP, its statistical unreliability may seem secondary. Yet it is astonishing that the accuracy of these seldom questioned....[even though] estimates are subject to repeated revisions....[As Professor Oskar Morgenstern has observed, 'national income statistics are still being taken at their face value and interpreted as if their accuracy compared favourably [sic] with the measurement of light'." [Rimmer, 27]

According to a 1994 article in Business Week. "ln the midst of the greatest information explosion in history, the government is pumping out a stream of statistics that are nothing but myths and misinformation. Most of the surging information economy--including software, telecommunications, and entertainment--is poorly covered by the data. While figures such as a 5.9% unemployment rate or a 3% inflation rate seem to have a reassuring solidity, in fact the connection between the government's statistics and the reality is getting more tenuous every year. Such official measures as gross domestic product (GDP), producer price index (PPI), or capacity utilization say far more about what's happening in old-line manufacturing industries than they do about such leading-edge companies as Microsoft, Disney, MCI, Fidelity, and Intel." [Business Week, 110]

Alan Greenspan, chairman of the Board of Governors of the Federal Reserve System believes GDP is not being measured accurately. He has "...noted inflation measurement problems in various major industrial countries, including the U.S." [Dow Jones News Service, 1] Technological change, innovations within economies and the introduction of new types of services, he observes, are all variables that increase the complexity of measuring economic output, and "the concept of a unit of output is becoming increasingly difficult to craft....Clearly, if you cannot define the unit of output, you cannot define price. And even if you succeed, in an adequate proxy, for unit of output, unless it is substantially unchanged over a period of time, price change isn't defined."[Dow Jones News Service, 1] Because economic data is relied upon heavily in making forecasts of the future, and, therefore, observes Greenspan, poor data has serious consequences because expectations about inflation can "distort economic decision making." [Dow Jones News Services, 1].

How do you measure the output of companies that provide cable TV service? Hours viewers have their TVs on? Having access to cable service is a benefit and, although, two companies may provide the same number of hours of viewing, more homes may be served by one than the other. How can output be measured so as to also measure this, or should the number of homes served be the measure of output? How is a value to be placed on having access to 80 channels, rather than 20?

Does what is labeled in the National Income Accounts "investment" include what it should? Investment goods differ from consumer goods in that they are not directly consumed. Investment goods are things like tools, equipment, buildings, etc. The objective of those acquiring them is not direct consumption, but indirect consumption, that is, to use them to produce things which are directly consumed and more investment goods. (Investment goods means the same thing as capital goods, that is, real capital. The term "real" is used to distinguish it from capital in the sense of cash and securities.)

The productivity of the resources used to produce goods and services is measured by the value of the output produced by a resource relative to the cost of that resource. Thus, the productivity of workers is measured by the value of output relative to the cost of labor. Investment is evaluated on the basis of the rate of return earned on it, that is, profits to the amount invested in capital goods.

More and better capital makes it possible to produce more. In the former case, this is because there is more of it. In the latter case, this is due to the fact that it is more productive. We end a year with more capital than we began the year with only if the value of the capital used up during the year in producing consumer and capital goods falls short of the value of the capital produced during the year. We have only very rough estimates of the value of the capital used up during the year.

Computer hardware counts as investment; computer software does not unless its cost is included in the price of the hardware is used with. Homes count as investment, but durable goods purchased by consumers do not.

Modern technology has made it both very difficult to measure much of the nation's output and how productive its workers are. FedEx provides a good example of such a situation. It has been estimated that it costs FedEx 75 cents or more when a customer calls on the telephone to inquire about their shipment. Recently FedEx established a site on the Web that enables a customer to type in an airbill number and determine their shipment's status. It is estimated that FedEx can cover the capital expenditure required to establish this alternative method of handling customer inquiries in four months. [Kessler, 294]

The Impact Of Chain-Weighing

Many economists think that productivity growth is inaccurately measured because no satisfactory method has been developed to measure productivity gains in services. A recent change in how the BEA measures real GDP called chain-weighing caused its estimate of productivity growth to decline from what was considered a healthy rate to an anemic one. According to the revised figures, productivity gains in the '90s have been no better than in the '70s and '80s.

With the chain-weighing method the BEA has switched to, the price of an item in each and every year is compared, not to its price in a given earlier year, but with its average price over two years: the current year and the previous year. Another change the BEA has made is to take into account the fact that people switch to a substitute product when the price of the product they formerly bought goes up relative to the substitute's.

Before the BEA switched to chain-weighting, it utilized a single base year which it moved forward every five years. The base year was moved forward because, if it wasn't, there would be an ever larger number of goods and services produced in the current year that were not produced in the base year. Also, changes in the relative prices of various goods and services would probably be greater. Qualitative changes would also be greater. (There is an enormous difference between automobiles in 1929 and 1996. There is a significant difference even between 1986 and 1996 automobiles. So, while if automobiles had not changed, due simply to inflation, their price would have been greater in 1996 than in 1929 and 1986; some of the increase in their price was due to their higher quality.)

Prior to the switch to chain-weighing, real GDP in 1985 was initially determined by computing what that year's output would have sold for in 1982. Later the base year was moved forward, and it was recalculated using 1987 prices. Because while, say, in 1982 the ratio of the price of product X to product Y was 4, but by 1987 it had changed to 2, the ratio of real GDP in 1984 to 1985 would not be the same when the base year was 1987 as it was when the base year was 1985. Therefore, every time the base year was changed, the growth rate of real GDP from year to year changed. Chain-weighing eliminates this.

The "BEA has estimated that shifting the base year from 1982 to 1987 resulted in an average drop of 0.3 percentage point in GDP growth for 1982-88. This adjustment is not trivial; for instance, it is close to some economists' estimates of by how much the decline in U.S. saving--including the increase in the budget deficit--cut U.S. growth rates... in the 1980s." [Steindel, 2]

Proponents of the new system for deflating GDP note that,"the chain-weighted procedure will...freeze the historic record of growth. History will no longer be rewritten by an arbitrary switch of the base year--the growth rates for a year will instead remain fixed." [Steindel, 4] On the other hand, with this method "we can no longer measure precisely manufacturing [or any other sector's] share of real GDP." [Steindel, 5]

Computer prices "declined at an average annual rate of 17 percent during 1982 - 1987, while computer output increased at a 34-percent rate; as a result, computers caused significant revisions in the GDP estimates when the weights and base period were updated. For example, when BEA shifted the weights and base period from 1982 to 1987...computers contributed significantly to the downward revision of 0.2 percentage point in the annual growth rate of real GDP for 1977 - 90." [Landefeld, 34] But, while chain-weighing will preclude this in the future, the resulting elimination of fixed-price weights will prevent the sum of the values of its components equaling GDP.

1972 was once the base year. There were no personal computers produced in 1972. To deal with this fact in computing real GDP in later years when they did exist, the BEA looked at how the prices of similar products such as electric typewriters changed. So, if, say, the price of electric typewriters in 1972 were 60 percent of their 1980 price, personal computers produced in 1980 would be valued at 60 percent of the prices they actually sold for that year. Quality improvements in a product from 1972 to 1980 were dealt with by estimating their value.

Chain weighing produces a rather interesting phenomenon. To demonstrate this, for simplicity, it will be assumed that only two products, X and Y, are produced. First we will compute real GDP for Years 2 and 3 the old-fashioned way, letting Year 1 be the base year.



(1) Price (in dollars)

(2) Quantity

(1) X (2)













Assume X's price rises to $3 and Y's to $6, but output stays the same. Actual GDP will rise to $120. Real GDP will be $80 because the level of output hasn't changed, and Year 2 will be repriced using Year 1 prices.


Assume X's price rises to $4 and Y's to $10, but output stays the same. Actual GDP will rise to $180. Real GDP will be $80 because the level of output hasn't changed, and Year 3 will be repriced using Year 1 prices.


The average price of X in Years 1 and 2 is $2.50, while that of Y is $5.00 (X's=$2 plus $3 divided by 2 and Y's=$4 plus $6 divided by 2.) Using these prices to compute real GDP in Year 2 gives us a real GDP of $100. ($2.50 X 20 plus $5 X 10 = $50 plus $50 = $100.) Because the average price of X in Years 2 and 3 is $3.50, while that of Y is $8.00, real GDP in Year 3 is $150. So, even though what is produced hasn't changed at all, real GDP has!

Observe, too, that with the old system both X and Y each year comprised one-half of real GDP in each year. With the new system this is not true. In Year 3, the value of X = $3.50 X 20 = $70, while the value of Y = $8 X 10 = $80. This difference is the product of the fact that the difference in their relative prices in Year 1 and Year 3 had no effect on real GDP in Year 3 when Year 1 was the base period, but this did affect real GDP when Years 2 and 3 were the base period. This is one reason why chain-weighing was adopted.

GDP As Income - The Savings Problem

The United States' relatively low savings rate compared to the other industrialized nations has been a matter of concern for many years. Some have speculated that the low savings rate has not had a devastating impact on real capital formation only because a substantial amount of foreign savings have been invested in the United States.

There are a number of hypotheses as to its cause of the decline in the savings rate. Some, however, question whether savings is being measured properly. Saving, as economist Arthur B. Laffer has observed, is very important because "it is society's only way of accumulating capital--and thus of spurring technological inventions, discoveries and developments. Sooner of later an economy will have to come to a grinding halt if it is deprived of new capital. And productivity will stagnate without the technology found only in new capital. Therefore, the faster the amount of available capital increases and the more capital there is, the faster the economy will grow and the more able society will be able to solve its economic problems." [Laffer]

Laffer claims that the measuring of saving as being income less spending on consumption and taxes is not the proper way to measure saving. What is wrong with this is that if, say, a person spent his entire income this year, this person would add nothing to the nation's savings this year as measured by the BEA; yet his or her wealth may have increased during the year through a rise in the market value of his or her investments. On the other hand, someone who sank part of his or her income into a worthless investment would add to the nation's savings this year as measured by the BEA.

Is What Supposedly Is Measured What Is Measured?

GDP is supposed to measure the value of all final goods and services produced during the year. Final goods and services, rather than all goods and services, are measured because much of what is produced during the year, like batteries an automobile assembly plant installs in automobiles, is subsequently, in effect, resold, as part of another product; so to include these intermediateproducts would involve double counting: so only final goods like the automobile consumers buy for personal use are counted. (The value of the automobile to society is what is paid for it by the consumer; not this plus what the automobile assembly plant pays for the battery plus what the battery maker pays for the components it of the battery it produces, plus what the automobile assembly plant pays for the tires, plus what the tire manufacturer pays for the rubber it makes the tires from, and so forth.)

Things like used car sales during the year are also not counted. This is because they were included in the year in which they were produced.

It is obvious that GDP doesn't include all final goods. Household production is not included, because there is no feasible way to measure how much of it is going on. GDP also doesn't measure the output of individuals producing illegal goods, like cocaine, and illegal services, such as sports betting. Barter transactions are also not counted because of the difficulty in measuring and valuing them.

Even thought, they too, are not sold in the marketplace, the services of government workers are included in GDP. Substituting for the market prices used to value services provided by private sector is what it cost the government to provide these services. One of the problems with this is that, while adjusted for inflation, the price of a privately-provided service will not rise unless its value to the public has increased, as otherwise the public would not buy it; whereas the public is forced to pay, through taxes, for a government service whatever the government chooses to spend on providing it.

Another problem with the way government output is valued is that it involves double counting of the value of the government's output. What a private business pays for security guards is not included in GDP because their services are considered an input in the productive process, and counting what they are paid would involve double counting, since what the business sells its products for covers the costs incurred in producing them. What is paid policemen employed by a city government, however, is included in GDP; yet they provide society a like service needed to facilitate the productive process.

Yet another difficulty associated with the handling of the cost of government is the fact that the prices at which private sector output is sold covers the taxes the private sector pays. For example, the price of an automobile is higher than it otherwise would be because an automobile assembly plant pays property taxes. The price of an automobile is higher, too, than it otherwise would be because, due to income taxes, assembly line workers' gross pay is higher than their take-home pay. Therefore, when government spending increases, and this increase is tax financed, GDP rises both due to the increased spending by government and the resulting increase in the prices of privately-produced goods and services.

Is What Is Measured What Should Be Measured?

At the heart of the current controversy over how the CPI should be computed is whether it should measure the cost of a given set of consumer goods and services or whether it should measure the cost of living--the two differing because, over time, what the typical consumer purchases changes. (For example, before CD players existed, people gathered around the family piano and played and sang. When lumber was relatively cheaper, the floors and walls of the typical home were made of solid planks of wood.) There is a like controversy over what GDP should measure. Some argue that it should measure the totality of the public's well being and, therefore, pollution and the depletion of natural resources should be subtracted from GDP. Doing this would necessarily involve very subjective judgments, as there is no objective way to measure the value of these and the other suggested subtractions.

It has been proposed that a number of things not currently included in GDP have be measured and included. Critics, for example, point out that, while private businesses deduct from their income the depletion of natural resources that they own, such as oil bearing land, GDP is not reduced by the value of the natural resources used up in the process of producing goods and services. (Because the price a natural-resource using firm charges is higher than otherwise would be the case so it can cover its depletion allowance, the depletion of natural resources increases GDP.) On the other hand, some say that because a shorter work week increases people's well being, GDP should be increased when the amount of leisure time the average person enjoys is increased. In this way, GDP today would reflect the fact the work week has shrunk this century from 6-days, many of which were 12-hour days, to the 5-day week and 8-hour days. (A problem with this is the fact that leisure time also rises when the unemployment rate rises; so, presumably, an increase in the unemployment rate should reduce GDP.)

Other changes necessary if GDP is to measure over all well being as critics think it should, includes subtracting from GDP the costs pollution imposes on society and losses from the commission of crimes. The cost of the destruction brought about by such natural disasters as hurricanes, too, is something some critics of the current system advocate be subtracted.

If GDP is to accurately measure overall well being, some very difficult problems in addition to those already cited arise. For example, the internet is a substitute for the telephone, regular mail, radio, and television. How can we realistically compare the well being of people before the internet existed to their well being after it came into being? (During the lifetimes of many people living today we have gone from hand copying material in reference books in a library to Xeroxing the material to downloading it at home from the internet!)

Many economists believe that the CPI as it is currently computed overstates the actual rate of inflation of consumer goods, and they advocate the method of computing it be changed so as to eliminate this error. The substantial negative impact on many people's income of the proposed change has brought home to the American public how very important to them is the government's estimate of the rate at which the price level is changing. (The negative impact on income is the product of the fact that the new method produces a significantly lower estimate of the rate of inflation, and many people's incomes are tied to the CPI. Social security and other entitlements: guaranteed transfer payments--unearned income--provided by the government, are increased as the price level, and so is the pay of many workers in the private sector.)

Some wonder whether the method of measuring the CPI is being changed because it is flawed, or whether the method is being changed in order to balance the federal budget. A suspicious Wall Street Journal writer, for example, says that adjusting the prices paid by consumers "for quality, substitutability, 'satisfaction' and the like is tricky--unless you have a desired answer in mind ahead of time." [Lowenstein, C1] 

One of the chief reasons why the CPI is believed to overstate inflation is that it is thought that it is not adjusted to properly reflect changes over time in the quality of goods and services. If, for example, something's price has doubled, but it is four times better, its "true" price has actually declined by one-half! However, someone's whose income has been increased by a percent which has been computed by properly evaluating quality changes over time will find their real income has declined because some of yesterday's inferior goods and services are no longer on the market. An inferior medical procedure used in the past may be unavailable today, and the vastly superior one available today costs much more. (If it is still available, people will not want to use it.) The inferior car of the past is no longer produced, and today's superior one costs much more.


There are plenty of reasons to question whether what is being measured is what should be measured and whether what should be measured or what is actually being measured can be measured sufficiently accurately for many of its uses.



1. We need to know by how much the price level changes from year to year in order to determine by how much the level of output changes from year to year. Example: If what we spend on GDP this year exceeds what we spent on it last year by 21 percent, and the price level rose by 10 percent this year, we know that this year the quantity of goods and services we consumed increased by only 10 percent; not 21 percent. This is because, if we let GDP last year = P (average price per unit) times Q (the number of units of output produced), then, if both the price and output levels rise by 10 percent, this year's GDP will be 21 percent higher because 1.10P X 1.10Q = 1.21GDP.

The price level is measured by an index number whose value in what is called the base period is set at 100. So, if the prices in another year are, say, ten percent higher, that year's price level is 110. This number is used to deflate actual (current-dollar) GDP to real GDP. Example: If in the base period, actual GDP (the number of dollars actually spent on it) is $200 trillion, and it was $220 trillion in a later year when prices were ten percent higher, real (constant-dollar) GDP in that year is $200 trillion: $220 trillion divided by 1.10. If GDP in the later year was, instead, $440 trillion, real GDP in that year is $400 trillion.

Real GDP is actual (nominal) GDP adjusted for changes in the price level. Therefore, it measures only changes in the level and quality of output; not a combination of a change in the price level and the level of output as does actual (nominal) GDP.

The CPI measures the rate of change in the prices of consumer goods and services; while the GDP deflator measures the rate of change in the prices of everything (price level) we produce (both consumer and producer goods and services). The PPI measures the rate of change in the prices of producers' goods and services (things businesses buy: capital goods). All three are index numbers, Example: If the price level measured by one of these this year is, say, seven percent higher than it was in the base period, the index this year would be 107, and the average price per unit in the base period multiplied by 1.07 would give you the average price per unit this year.

The prices whose level the CPI measures are the prices of a specified group of consumer goods and services. The group of goods and services whose prices the CPI measures consists of those purchased by a typical urban family. This group is called the market basket. The CPI is estimated by the Bureau of Labor Statistics (BLS). The 1987 (a base period) market basket was based on interviews with nearly 10,000 American families from 1982 to 1984. It included 184 different goods and services.

The prices used in computing the CPI are gathered monthly by 250 "price collectors" employed by the BLS. The composition of the market basket changes over the years because what families buy changes over time, and periodic surveys by the BLS reveal this.

The CPI is actually a weighted average of several component price indexes: housing, transportation, food and beverages, etc. "Weight" refers to the relative importance of a given class of purchases. If, for example, one type of goods accounts for twice as large a share of each dollar spend by the average family, it will have twice the impact on the size of the CPI.

The CPI is a better measure of changes in the cost of living than is the Implicit GDP Deflator, because the latter measures the rate of changes in the prices of all final goods and services; not just those purchased by the average urban family.

The CPI is used to determine real income.

Real income equals = actual (nominal) income divided by the CPI divided by 100

The CPI can also be used to compute the purchasing power of the dollar.

The purchasing power of the dollar = $1 divided by the CPI divided by 100

2. As Expenditures, GDP can be viewed as:

(1) Consumption + Investment (business' capital goods spending) + Government + Exports;

that is:: what American consumers spend on (1) goods and services produced domestically plus what American (2) businesses spend on them, plus what our (3) government spends on them, plus (4) exports (what foreigners spend on them);

as Income, GDP can be viewed as:

(2) Consumption + Savings + Taxes + Imports

that is: what Americans do with their income is: (1) spend it on domestically-produced consumer goods and services plus what they don't spend, either because it is (2) saved or cannot be kept because it is paid to the government as (3) taxes, plus what they spend on (4) imports (goods and services produced abroad)..

Because what is spent on output is paid out as income:

(3) Consumption + Investment + Government + Exports = Consumption + Savings + Taxes + Imports

GDP is defined as:

(4) Consumption + Investment + Government + Net Exports.


(5) Net Exports = Exports - Imports

(Net Exports may be either a positive (exports exceed imports) or a negative number (imports exceed exports).

Equation (4) is derived by subtracting imports from both side of the above (3) identity. It means that GDP is the value of all domestically produced goods and services plus, if exports exceed imports, what is owed to us by foreigners or, if imports exceed exports, less what we owe foreigners..

More meaningful for measuring material well being than Gross Domestic Product is Net Domestic Product (NDP) or National Income. Some of the capital goods (tools, machinery, etc.) produced in earlier years and in stock at the start of the current year are used up during the year in the production of consumer goods and capital goods. While GDP includes all the capital goods produced during the year (gross investment), GDP less the value of capital goods used up during the year (depreciation) equals NDP; so NDP includes only net investment (the increase in our stock of capital). However, economists often use GDP instead of NDP because estimates of the wearing-out and obsolescence of real capital are clearly not highly reliable.

Analogous to gross investment is gross saving, and net saving is no more reliable than is net investment because its accuracy, too, is dependent upon the accuracy of the estimate of depreciation. (Depreciation as recorded by accountants does not closely reflect the actual wearing out of real capital because many business select the depreciation method they use so as to defer taxes, rather than accurately reflect the actual physical deterioration of their real capital.)

Works Cited

Beach, William and Gareth Davis, "Why The Second Quarter Economic Growth Figures Probably Will Be Wrong." The Heritage Foundation, FYI No. 114, July 29, 1996:,

Business Week, "The Real Truth About The Economy." November 7, 1994, pp. 110 - 118.

Dow Jones News Services, "Fed Chairman Acknowledges Difficulty Of Gauging Inflation." The Wall Street Journal (interactive edition), August 30, 1996,

Heritage Foundation (news release), "Take GDP Growth Numbers With A Grain Of Salt, Heritage Economist Says."

Kessler, Andrew J., "Software That Pays For Itself,."Fortune, October 21, 1996, p. 294.

Laffer, Arthur B., "Creating Wealth, Not Just 'Savings',." The Wall Street Journal, October 15, 1996.

Landefeld, J. Steven and Jack E. Triplett, "Preview Of The Comprehensive Revision Of The National Income And Product Accounts: BEA's New Featured Measures Of Output And Prices." Survey Of Current Business, July 1995, pp. 31 -38.

McLenaghan, John B., "Standards For The Dissemination Of Economic And Financial Statistics." Survey Of Current Business, October 1996, pp. 39 - 47.

Rimmer, Douglas, Macromancy, The Ideology Of "Development Economics". Hobard Paper 55, The Institute Of Economic Affairs (Great Britain), 1973.

Schmidt, Ronald H., "Services: A Future Of Low Productivity Growth?" Weekly Letter (Federal Reserve Bank of San Francisco), February 14, 1992, pp. 1 - 3.

Steindel, Charles, "Chain-weighting: The New Approach To Measuring GDP." Current Issues In Economics And Finance (Federal Reserve Bank Of New York), December 1995, pp. 1 - 6.



Beach, William and Gareth Davis, "Why The Second Quarter Economic Growth Figures Probably Will Be Wrong." The Heritage Foundation, FYI No. 114, July 29, 1996:

Boskin, Michael J., "Prisoners Of Faulty Statistics." December 5, 1996, The Wall Street Journal, p. A20.

Business Week, "The Real Truth About The Economy." November 7, 1994, pp. 110 - 118.

Cobb, Clifford, Ted Halstead, and Jonathan Rowe, "If The GDP Is Up, Why Is America Down?" The Atlantic Monthly, October 1995, pp. 59 - 78.

Dentzer, Susam, "The Growing Mysteries Of The GDP." U.S. News And World Report, October 30, 1995, p. 67.

Dentzer, Susan, "Warp 6 For The U.S. Economy." U.S. News And World Report, November 6, 1995, p. 77.

Dow Jones News Services, "Fed Chairman Acknowledges Difficulty Of Gauging Inflation." The Wall Street Journal (interactive edition), August 30, 1996,

Eisner, Robert, "What Counts And How To Count It." The Wall Street Journal, November 12, 1996.

Fleming, Martin, "The Statistics Corner: The GDP Revisions And Understanding Sluggish Productivity Growth." Business Economics, April 1996, pp. 62 - 65.

Goldstein, Henry, "Should We Fret About Our Low Net National Savings Rate?" Cato Journal, January 1990.

Hequet, Marc, "Poof! Productivity Vanishes." Training, December 1995, p. 16.

Heritage Foundation (news release), "Take GDP Growth Numbers With A Grain Of Salt, Heritage Economist Says."

Hershey, Robert D., Jr., The New York Times, "Concern Is Voiced Over The Quality Of Economic Data." November 24, 1996, pp. 1, 18.

Kessler, Andrew J., "Software That Pays For Itself." Fortune, October 21, 1996, p. 294.

Laffer, Arthur B., "Creating Wealth, Not Just 'Savings'." The Wall Street Journal, October 15, 1996.

Landefeld, J. Steven and Jack E. Triplett, "Preview Of The Comprehensive Revision Of The National Income And Product Accounts: BEA's New Featured Measures Of Output And Prices."Survey Of Current Business, July 1995, pp. 31 -38.

Lowenstein, Roger, "Betraying A Fondness For The Old CPI." The Wall Street Journal, December 5, 1996, p. C1.

McLenaghan, John B., "Standards For The Dissemination Of Economic And Financial Statistics." Survey Of Current Business, October 1996, pp. 39 - 47.

Morris, Charles R., "It's Not The Economy, Stupid." The Atlantic, July 1993, pp. 49 - 62.

Rimmer, Douglas, Macromancy, The Ideology Of "Development Economics". Hobard Paper 55, The Institute Of Economic Affairs (Great Britain), 1973.

Stahl, David, America's Community Banker, "The Savings Problem." April 1995, pp. 14-19.

Schmidt, Ronald H., "Services: A Future Of Low Productivity Growth?" Weekly Letter (Federal Reserve Bank of San Francisco), February 14, 1992, pp. 1 - 3.

Steindel, Charles, "Chain-weighting: The New Approach To Measuring GDP." Current Issues In Economics And Finance (Federal Reserve Bank Of New York), December 1995, pp. 1 - 6.

Survey of Current Business, "A Look At How BEA Presents The National Income And Product Accounts." May 1996, pp. 33 - 37.