The Circular Flow and National Income AccountingThe Circular Flow
|Until now attention
has been concentrated on how individual markets work, the decision-making
processes by which individual consumers (or households), firms, owners
of productive factors, and public officials go about the business of maximising
their best interests. Now the focus shifts from this microeconomic perspective
to a much broader one. From this point on the subject matter relates to
the functioning of the economy as a whole, a field of study known as macroeconomics.
Macroeconomics deals with a number of large totals or aggregates which are used to conceptualise and measure key components of the economy. The most fundamental of these is the total output of goods and services, conventionally referred to as the national income. (Official data in most countries is now actually reported on a "domestic" rather than a "national" basis. The distinction, which is unimportant for most purposes, relates to the treatment of investment income received from non-residents and paid to non-residents. "Domestic income" is that produced within a country by all producers operating there, whether foreign or not. "National income" is that produced only by "nationals" of that country, whether they are producing it there or elsewhere.)
There is nothing inconsistent in referring to total output as income. Although what is earned as income can be measured separately from what is produced, the two aggregates are necessarily the same in amount. Before going on to see why, note that in either case such large totals can be expressed only in terms of money, not physical products as such. It is impractical to try to measure output or income in real, physical terms, simply because it is impossible to sum apples and oranges or any of the millions of goods and services which are produced and received as income in a modern economy. Instead, physical quantities must be converted to a common measure and the measure used for this purpose is the national unit of account, the dollar, pound, or other currency.
The value of total output or income in an economy during some accounting period, usually a year or quarter of a year, is a significant statistic. It is generally used as an indicator of the economy’s performance. Because a larger output or income is equated with a rise in the economic well-being of a country’s population, a higher output or income is considered desirable and a lower one undesirable. The economy’s overall performance is tracked by the changing value of the total output or income statistic. Similarly, comparisons of relative well-being among different countries are based on these statistics and a host of political and social as well as economic implications flow from their behaviour over time.
A modern economy can be simply modelled in the aggregate by thinking of it as comprising two key sectors, households which consume produced goods and services and which supply labour and other productive services to firms, which use the labour and other productive services supplied by households to produce the goods and services the households consume. Households supply the services of productive factors (land, labour, capital, etc.) and the firms convert these inputs into produced goods and services which return to the households. Owners of firms are, of course, also part of the household sector where they function in their other capacity as consumers of goods and services.
The real flows of productive services and produced outputs have corresponding flows of money payments associated with them. Firms pay out wages and salaries in return for labour services, rents to owners of land and other natural resource inputs, and interest and profits to suppliers of capital and entrepreneurial inputs. Householders consequently have money income with which to pay for the produced goods and services that flow to them from firms. Thus, there are money flows corresponding to the real flows, but they move, of course, in the opposite direction.
Because the flows of payments for produced goods and services and payments for factor inputs are continuous, aggregate income/output in this simple model could be measured at any point, metering the flow anywhere in the circuit. If measured in terms of spending on produced goods and services, it would be natural to call this a measure of total spending or total expenditure. If measured in terms of outlays made for the services of productive factor inputs, it would be total income (from the point of view of the owners of those factor inputs). Obviously the two totals would have to be the same.
This is a greatly simplified model. One thing missing is the possibility of saving. If households do not spend all their income on produced goods and services, but hold some of it back as savings, every time income flows into the household sector the flow of payments made to producers will diminish. In terms of a mechanical (or hydraulic) model, there would be "leakage" of income/spending from the system and the volume of the flow would diminish—the level of national income would fall. But if there are savings, there could also be new investment. If businesses borrowed income saved by households and used it to finance the building of new plant or for other business purposes, it would be injected back into the income stream (in the form of payments to workers and other factor owners who supplied the necessary real inputs needed to produce the new capital). Banks and other financial intermediaries serve as the nexus through which savings are converted into investment spending and returned to the income stream.
If another complication, government, is added to the simple model, another potential for a leakage of income from the system is introduced. Governments impose taxes on households (and firms) and this results in a diversion of income from the private sector to government. This is a leakage similar to the savings leakage and it too has a corresponding potential for injecting such income back into the stream, this time in the form of government spending on produced goods and services.
Finally, most real world economies are not closed loops. Instead they are "open" to the rest of the world, with leakages from domestic income/expenditure flows in the form of payments made for goods and services produced by firms in other countries (imports) and injections of income back into the domestic flows as a result of sales of goods by domestic firms to consumers in other countries (exports). As already seen, there can also be important flows of savings and investment between one country and the rest of the world.
The important ideas
to understand at this point are that national income or expenditure can
be thought of as a continuous flow which can be measured in different ways
and that this simple process is complicated by the possibilities of leakages
and injections arising from private saving and investing; government taxation
and spending; and foreign trade and capital movements.
Each of us puts in what he has at one point of the circle of exchange and takes out what he wants at another.
—P.H. Wicksteed (1914)To Top
All the developed industrial economies today measure the volume of aggregate income, usually defined as Gross Domestic Product, in much the same way. (Here are links to some official sources.) Whatever their minor differences, all national accounting conventions follow the basic pattern identified in the preceding discussion of the circular flow of income and expenditure. There are always at least two main calculations, one which sums total expenditures on goods and services produced, the other of total income received as a result of producing those same goods and services. Because both are measures of the same thing they must, by definition, yield the same total.
Why two measures if the total must be the same? One reason is that two estimates provide a check on one another with respect to accuracy. Another is that the two measures break down into different components, some of which are more useful for certain purposes than others.
Measuring total output by the expenditure method involves breaking down total spending on all goods and services produced into four categories: spending done by consumers on goods and services (abbreviated simply to the letter C); spending done by businesses on capital goods (total investment spending, I); spending by governments on goods and services, G); and net exports (the total value of exports minus the total value of imports, X-M). Because all spending done in the country falls into one or other of these four categories, we can say that total expenditure is the sum of C+I+G+(X-M). We now examine each of these four main components of total spending.
Consumption spending is the total of all outlays made by households on final goods and services. In all countries it is by far the largest component of total spending. It covers spending on an enormous range of items, including durable goods like television sets and cars, non-durable goods like food and clothing, and personal services such as legal advice, hairdressing, and dental care. But it usually excludes spending on houses which is customarily (and arbitrarily) treated as investment expenditure. C also excludes purchases of second-hand goods that were produced in some earlier accounting period so as not to double count the value of such output.
Much of the spending done by governments in the developed countries today takes the form of simple transfers of income from taxpayers to those eligible for the wide range of income supplements available to assist the elderly, the sick and the unemployed, or as payments of interest to holders of the public debt. Such transfer payments do not represent spending on current production and consequently, like other transfers, are not counted in national income determination. What is counted is government spending on goods and services, many of which are bought by the government on behalf of the public and which are ultimately "consumed" by households: education, health care services, national defence, roads, water and sewage systems, postal services. Because so many of these goods and services are provided "free" or in other ways that bypass markets, it is difficult to determine their value in the same way that the value of the other items entering into C would be determined. Consequently, national income accountants value government spending on the basis of what the government pays for the goods and services it requires. (The attentive reader may find it interesting to recall the discussion of Niskanen’s bureaucracy .)
Another complication with government spending on goods and services is that such spending is often done on things like highways which are themselves capable of being used to assist in the production of other goods. Logically, such spending should be thought of as investment spending and included in the next category to be discussed. Some countries produce their accounts in such a form that government spending can be separated into two categories, current spending on goods and services, and investment spending, but if the main concern is to understand the causes of year-to-year cyclical fluctuations in the level of national income rather than the causes of its longer term growth (which may be strongly affected by the level of investment as opposed to current spending) it is convenient to stick with the traditional categories of spending which emphasise the different motivations driving the spending decisions of ordinary consumers, private investors and governments. Here, investment spending refers to private investment spending unless otherwise stated.
Total or gross investment spending may be divided into two main categories. The first is spending on capital goods—purchases of plant and equipment either to replace existing capacity that is wearing out or to increase capacity. This is often called fixed capital formation. A second type of investment spending is on inventories. Many businesses find it convenient or necessary to hold certain supplies of goods on hand, in which case investment in inventories may be considered voluntary. But business conditions are uncertain and so firms may also find themselves holding stocks because they miscalculated demand. In either case, firms are considered to be investing when they accumulate inventories. On the other hand, if their inventories decrease they are "disinvesting." Inventory investment is highly volatile, changing greatly in amount and composition from year to year.
Gross investment, then, is the total amount of (usually private) spending during the accounting period on capital goods (defined as structures, machinery and equipment, and inventories). Because capital by its nature consists of things that are used in the production of other goods and services, it is inevitable that it will wear out or "depreciate." Unless it is continually renewed, the stock of capital in the economy will gradually be depleted.
Handling depreciation is one of the more difficult parts of national income accounting. Again, the best treatment depends on what the data are meant to be used for. If the concern is with the long term growth of the economy, net investment (total investment during the accounting period minus depreciation) is the important concept because it measures the growth of the economy’s capital stock over time. But if the purpose is to understand short term, annual fluctuations in the level of total spending it is better to work with gross investment.
With the great expansion of world trade in recent decades, a significant part of total spending in most countries goes toward the purchase of goods produced abroad rather than domestically. As noted in discussing the circular flow, such outlays represent spending which leaks from the domestic economy to the rest of the world and is consequently treated as a negative entry in measures of total domestic spending. But it is offset to a greater or lesser degree by the spending of non-residents on goods produced and exported to international markets. It is often convenient, therefore, to take domestic spending on imports and foreign spending on exports as a combined value, usually called net exports, a value which may be positive or negative in any accounting period depending on which component, exports or imports, is larger.
Summing these four expenditure components, C+I+G+(X-M), gives a single figure, the total amount of spending done in the economy during the accounting period. It should be possible to arrive at exactly the same figure by summing all income received in the economy during the accounting period.
As seen in discussing the circular flow, what the firms producing the national output see as costs of production, owners of productive factors see as income. Factor costs and factor incomes are consequently the same thing viewed from different perspectives.
Quantitatively, by far the most important and certainly the simplest factor costs to measure are the payments made by employers for labour services. These payments are usually reported in the official statistics under a heading such as "Wages, salaries, and supplementary labour income," with the latter term referring to employee benefits such as pensions, workers’ compensation benefits, and employer contributions to unemployment insurance funds or other worker social security schemes. Most other factor payments, however, are much more difficult to track. Consider a farming operation. How should any net income derived from farming be classified? Some of it must be a return to the services of land the farmer is using—in which case it is rent. Some must be a return to the farmer’s own input of labour, which is wages. Some might be considered a return to setting up and operating the business, hence profit. But how are these to be separated? Because of such problems, the national accounts typically use definitions of factor payments which owe more to convenience than to the logic of factor classification: net income of farm operators, corporation profits, net income of unincorporated business, and interest and other investment income. Summing all these items yields the total amount received during the accounting period by the owners of productive factors. But if this figure for factor costs or income is compared with the total arrived at by the expenditure method, it falls considerably short of the amount expected.
The reason for this is that not all the costs of producing the total output have been counted so far. The prices paid for goods and services must be high enough to cover the costs of factors of production used to produce them. But producers also have other costs which they expect to be covered by the prices paid for the goods and services they produce. Governments impose certain kinds of taxes on producers which are passed on to consumers in the form of higher prices. They may also, however, pay some producers subsidies to enable them to lower the selling prices of their output. Deducting the latter from the former gives the net addition of indirect taxes to total costs of production. The second additional item which has to be included is one already encountered on the other side of the accounts—the cost of capital worn out during the accounting period. Firms must include in the market prices of their products some allowance for depreciation of capital, a capital consumption allowance. Although these items, indirect taxes less subsidies, and the capital consumption allowance, are small relative to the amount of factor payments, adding them in should bring the total expenses incurred by firms in producing the national output up to the amount arrived at by measuring total spending to purchase the national output.
Given the complexity of the measures involved and the reporting problems which may be encountered it would be surprising to get an exact match. Logically, however, the results must be the same, so any measurement errors are dealt with simply by dividing the difference in two, adding half to the low figure and subtracting half from the high figure to achieve the required equality in the result.
A third method is
available for estimating the total output of the economy and it is called
the "value added method" because it simply sums the net value of the output
produced by all the firms in the economy. This is conceptually simple,
but in practice complex because of the need to avoid double counting. There
is much interaction among firms in a modern economy. Many produce goods
which are sold not to final users as consumer goods, but to other firms.
Consider a firm producing power supply devices for computers. It buys components
from suppliers, assembles them, and sells the finished product to another
firm which incorporates it into a computer. If the value of the power supplies
was measured when they were produced and again as part of the price of
the finished computer, total output would obviously be exaggerated. Dealing
with this requires that the value of each firm’s output be reduced by the
amount of all payments made by that firm to obtain inputs. This involves
considerable work, but the resulting data are often very useful because
they yield a breakdown of national output on an industry-by-industry basis.
Like dreams, statistics are a form of wish fulfillment.
-Jean BaudrillardTo Top
Adjusting National Income Data to Allow for Price Changes
One difficulty with
using money values to express national accounting magnitudes is that the
value of money may change over time. If there is a general rise in all
prices, or a fall in all prices, the monetary unit either decreases or
increases in value. Trying to measure distance with a ruler that shrank
or expanded significantly between measurements would obviously be a frustrating
and not very useful activity. Inflation, defined as a general rise in the
price level, or deflation, a general fall in the price level, are common
enough to make it necessary to adjust national income data to remove the
effect of changes in the purchasing power of the dollar or other monetary
unit being used to measure the value of total output. This is done by developing
indexes which show how the prices of the goods and services produced in
any one year have changed relative to the prices of those goods and services
in some other year. Setting up these indexes of prices is not difficult
in principle, although it can be an expensive, time-consuming task in practice.
Consider a simple example in which only a single commodity is the subject
of interest, men’s shoes.In the following table the price of men’s shoes
in each year is compared with the price prevailing in one particular year.
The base year in the example is year 2, although it could have been any
one of the five years. The price in any particular year is then divided
by the price in the base year to get the ratio of prices shown in the third
column. Because these ratios are usually expressed as percentages, they
are then multiplied by 100 to obtain the price index numbers shown in the
These index numbers
can now be used to adjust the data on the value of men’s shoes produced
in each year, thereby eliminating the effect of price changes from the
series. Suppose the following production information is available.
The current dollar values shown in the second column turn out to be quite misleading as an indicator of the real changes in output. Because prices were lower in Year 1 than in the base year (Year 2), the output in Year 1 was understated, whereas, because prices in Years 3, 4, and 5 were higher than in the base year, the current dollar production values overstated the volume of output. The conversion to (Year 2) constant dollars in the third column was done by dividing the current dollar values of output for each year by the relevant index number (expressed as a percentage).
Men’s shoes are only one of thousands of commodities which are included in the total national income and, in practice, it is not feasible to develop price indexes for each item in this way. Instead, price indexes are built up for groups of commodities which are often defined in terms of who buys them. For example, a commonly used index measures changes in the amounts households spend on a selected bundle of goods and services. One of the problems with this kind of index is that it is very costly to determine which goods should be included in such a bundle. Surveys must be made of household buying habits to determine which goods households are buying in significant quantities and the relative importance of various goods in typical household budgets. Because of this, years may elapse between redefinitions of the goods which are included in the index which makes the information the index provides of dubious value toward the end of the redefinition cycle.
When constructing large price indexes for adjusting national income data, most statistical agencies build up a general index from a large number of specific commodity group indexes, so that changes in expenditure patterns within the component groups will not seriously affect the outcome. This composite index is known as a gross domestic product deflator. It can be used to convert any current dollar value of gross domestic product to a constant dollar basis using the relation:
GDP X 100 = GDP Deflator
If a number looks interesting it is probably wrong.
One of the principal developers of the national income accounting methods used today in most of the world was the Russian-born American economist, Simon Kuznets. While still very young Kuznets became head of the central statistical office in Ukraine, but emigrated to the US in the early 1920s. In 1970 his work in developing the system of national income accounting, which was so important in making macro-economic theory applicable to the real world, earned him the Nobel Prize in Economics. Kuznets was long associated with the US Department of Commerce which maintained the official US national accounts data. In the early 1940s, however, he became dissatisfied with the work the Department was doing, in particular its reluctance to expand the methodology so as to include such things as the value of unpaid household work in the estimates.
Since then there has been much interest in making the national accounts into a more comprehensive measure of welfare. It is widely recognised that in their current form the accounts provide only a very limited measure of changes in a country’s standard of living or "well-being". The decisions as to what to include and what to exclude from GDP are highly arbitrary. The production of military goods is typically included, although many people may doubt that they contribute to welfare. On the other hand, the production of illegal substances, such as certain drugs, is excluded, although some people would see them as increasing well-being. Many exclusions are based on practical problems of measurement. As a general rule, want-satisfying things that do not pass through markets (i.e., are not bought and sold) are excluded. Probably the largest group of these items relate to household production: food preparation, cleaning, clothing care, home maintenance and repair, gardening, pet care, shopping, providing physical care for children, tutoring children. A recent Canadian study has estimated that the labour utilised in providing these functions in households has a dollar value equivalent to between 30 and 40 per cent of GDP. The variation is due to two different methods being used to arrive at the estimate, the one basing it on the opportunity cost of such labour and the other on what it would have been necessary for householders to pay someone to do it for them. Comparing recent estimates with older ones showed that the amount of household work as a percentage of Canadian GDP has been declining, perhaps because of the great increase in participation in the labour force by married women.
If such estimates can be made, why is household work not included in official GDP? It cannot be simply because the product is not marketed, for there are already non-marketed elements incorporated in some official accounts. (In Canada and the US, for example, the value of occupant-owned housing is estimated and included in GDP.) One reason given by the authorities is that the techniques for making such estimates are still rudimentary and may not be capable of yielding systematic and consistent data. One of the basic difficulties lies in defining household production with any precision. If a person goes out to a movie the entertainment provided is included in GDP. What if the same person stays home and watches a home video? Should that be included? This soon leads to the question of how "doing nothing" should be treated. Leisure is conventionally thought of as a "good". It might be consistent with this to propose that if a person does nothing this is a component of total well-being that should be included in GDP. But what if the person is unemployed?
This kind of problem has plagued all those who have tried to develop a superior, all-in-one measure of human well-being. Especially in the 1960s and early 1970s there was a great deal of interest in devising a comprehensive measure of economic welfare (MEW). In the United States two mainstream economists, James Tobin and William Nordhaus, both at Yale University, pushed the available methodology to its limits and came up with the conclusion that it was impossible to devise a fully satisfactory measure of this type. The index they did manage to compute showed, for example, that during the Great Depression of the 1930s, total well-being in the US rose during the worst part of the decade, in the years 1929 to 1935! The reason was that they had quite sensibly included leisure as a positive factor in determining welfare. Massive unemployment generated a great increase in this component of the index.
Such dilemmas have not, however, led everyone to give up on the idea of developing an index to measure well-being. The latest effort has been sponsored by the United Nations which now publishes a remarkably ambitious Human Development Index (HDI). The HDI combines estimates of personal purchasing power, based on official national accounting income price level data, with indices measuring life expectancy and literacy. The conceptual basis upon which the HDI rests was set out in the United Nations’ Human Development Report of 1990. In it, human development is defined as "a process of enlarging people’s choices." In principle, these choices can be infinite and change over time. But at all levels of development, the three essential ones are for people to lead a long and healthy life, to acquire knowledge and to have access to resources needed for a decent standard of living. If these choices are not available, many other opportunities remain inaccessible."
Most of the criticism from academic reviewers of the HDI focuses on the issue of weighting. The HDI gives equal weight to its three main components, longevity, literacy and purchasing power. Would you agree that this is appropriate? How much additional income would you trade for an additional year of life? Is it better to die old and illiterate or young and well-read? These questions can never be answered with any degree of certainty, as Tobin and Nordhaus showed more than thirty years ago, and many economists would prefer to work with the relatively simple measures provided by the national accounts despite their inadequacies of coverage. Nevertheless, income is not everything, and the HDI yields a ranking of the goodness of life in different countries which is different than that provided by a simple GDP per capita ordering. Thus, Canada has usually placed as the highest ranking country in terms of HDI , despite having lower per capita income levels than several other countries. The United States, despite being near the top of the ranking in terms of income, typically comes in well behind Canada, Japan, Switzerland and Sweden, mainly because of lower life expectancy. (More on the HDI.)