Nouriel Roubini and David Backus

Chapter 1: Monitoring Macroeconomic Performance

Growth and Business Cycles

The classical question of economic growth is why some countries are richer and/or grow faster than others. (The two are clearly related, since countries that grow faster will eventually be richer.) Some examples are given in Figure 2, which graphs per capita GDP for three countries over the postwar period. [All are measured in 1980 US dollars.] This figure differs from the previous one, since I've expressed output in per capita (per person) terms by dividing GDP by population. This produces a more meaningful comparison between countries, since countries with more people don't automatically have higher numbers.

Figure 2 illustrates a number of differences among three countries: Japan, Argentina, and the US. Perhaps the most obvious feature is that the US is the richest country: by this measure in 1985, it was 30 percent richer than Japan and almost three times as rich as Argentina. These are averages so they ignore a lot of differences at the individual level, but they give you some idea of where these nations stand economically. The comparison with Argentina gives us an idea of the enormous differences between rich and poor countries. In fact, Argentineans are relatively well off, roughly five times better off than an average person in India. But the truly remarkable country is Japan. In 1913 Argentina was about 3 times richer than Japan, now it's the opposite. Japan's remarkable performance has lasted, thus far, for over a century. Argentina, on the other hand, has gone from one of the richest countries in the world at the turn of the century to an average Latin American country economically that experienced a severe economic and financial crisis in 2001.
 

Figure 3 does the same thing for the US, China, and Korea, where again we see sharp differences between countries. China used to be one of  the poorest of these countries but for the last 20 years China has been among the most rapidly growing countries in the world. Combined with China's enormous population, some estimates suggest that China is now the world's third largest market.

These comparisons are so striking I find it hard to leave them, but let's turn our attention to the other aspect of macroeconomics, business cycles. From a business point of view these short-term movements in the economy are of more immediate concern. You may want to know, for example, whether the economy will be in better shape when you finish your degree or whether your airline stock is going to be worth anything in 12 months (airlines are notoriously sensitive to recessions). You get a much better picture of the short-term fluctuations in Figure 4, where we graph annual growth rates of US GDP.

By annual growth rate, I mean the "year-on-year'' growth rate in quarterly data,

(GDP_t - GDP_t-4) /GDP_t-4

where GDP_t is GDP in quarter t (for example the third quarter of 2005) and GDP_t-4 is GDP four quarters before (for example the third quarter of 2004). Viewed from this perspective, the short-term movements seem a lot bigger than they did in Figure 1. For the postwar period as a whole the average growth rate of 3.3 percent per year is swamped by the year-to-year variations. [Statistically, we could say that the mean of 3.3 percent per year is only slightly larger than the standard deviation of 3.0 percent. A plus or minus two standard deviation interval is thus (-2.7,9.3). If you find this mysterious, review your statistics notes.] The nine downward spikes, all of which touch or pass the axis, are the nine postwar recessions, defined most simply as two consecutive quarters of declining GDP. The National Bureau of Economic Research, the de facto arbiter of business cycles in the US, has decided that the troughs (the bottom point) of these recessions occurred in November 1949, May 1954, April 1958, February 1961, November 1970, March 1975, July 1980, November 1982, April 1992 and November 2001.

Note that in Figure 4 the growth rate of GDP is defined as year-on-year growth rate of quarterly GDP. Note that there is an alternative way to define the growth rate of the economy: this is the way the growth rate of GDP is usually reported by the US Government and the press. It consists of measuring the growth rate of GDP in a particular quarter relative to the previous quarter and annualize such quarterly rate of growth by multiplying by four. Accordingly, the quarterly growth rate of the economy at an annual rate (AR) is:

4 x [(GDP_t - GDP_t-1) /GDP_t-1 ]

Figure 4' shows the growth rate of GDP according to this alternative measure. As a comparison of figures 4 and 4' shows, the second way of expressing the growth rate of the economy implies a greater volatility of output growth as quarterly changes in the rates of growth are amplified when measured at annualized rates. As the annualized quarterly growth rate gives a better measure of the very recent performance of the economy, this is the measure usually reported in the press and most closely analyzed in the business and financial sector. However, the year-on-year definition gives a better measure of the growth rate of the economy over a longer period, i.e. how the economy has actually grown over the last 4 quarters. A similar distinction between year-on-year growth rate and annualized quarterly growth rate holds for the other macroeconomic variables. To create quick charts of macro variables using these alternative definitions, you can use the Economic Chart Dispenser available on the Web. Tables with the most recent GDP data is available from the Bureau of Economic Analysis at the Department of Commerce. For more information on specific macroeconomic variables see the course homepage on the Hyptertext Glossary of Business Cycle Indicators.

One question you might ask is why the economy experiences such large short-term fluctuations. We'll return to this later in the course. For now let me just say that recessions happen: business cycles have been a property of all economies for as long as we've had data and, despite what politicians tell us, they show no sign of going away. You can see signs of cycles in other countries in Figure 5. In Figure 5 I report growth rates of real GNP (total, not per capita) in Germany and Japan, where we see that they, too, have had substantial fluctuations, despite their higher average growth rates. For Japan, though, there would be only recessions between World War II and 1990 if we defined a recession, as is typically done in the US, as negative growth. Note, however, that in the 1990s, Japan experience a period of protracted economic stagnation. The average growth rate per year was close to zero between 1992 and 1995. Growth recovered in 1996 but such recovery fizzled in 1997 when the economy went again into a slump. The weak economic performance of Japan in the 1990s and 1997 in particular contributed to exacerbate the 1997 economic crisis in East Asia: as Japan is a leading export market for many East Asian countries, the stagnation of growth in Japan in this decade led to a reduction (since 1995) in the export growth rate of many East Asian countries.

Gross Domestic Product

Our first goal is a measure of overall production, which we will refer to as Gross Domestic Product, or GDP. Gross National Product, or GNP, is closely related. Both are measures of the total production of goods and services of the US economy for a particular time period---say, the year 2004  or the first quarter of 2005 (January through March). We will discuss below the difference between these two measures.

We can think of total production in the US as the sum of production by all the individual firms, but there's a subtlety here that we can illustrate with a simple example. Consider a firm that assembles PCs from parts made in Taiwan. Its only other expenses are labor. Let's say that the firm's income statement looks something like this:

   Sales revenue           40,000,000


   Expenses                26,000,000

     Wages                 20,000,000
     Cost of Parts          6,000,000

   Net Income              14,000,000

Another way to compute value-added is to sum payments to labor and capital. In this case we add 20m paid to workers to 14m profit that goes to owners of the firm---capital. That gives us factor payments of 34m, the same number we found above using a different method, factor being a term used by economists to mean inputs

The term value-added has the connotation that the prices that underlie the firm's income statement reflect economic value in some deeper sense. When we compared the GDP's of three countries earlier we presumed that the country with the larger per capita GDP was richer in some useful sense. But suppose they produce different goods. Suppose country A produces 10 billion apples and country B produces 10 billion bananas. Which is richer? We generally assume that if apples are worth more than bananas then country A is richer. The idea is that market prices tell us which is more valuable, apples or bananas. The same thing underlies our measurement of value-added. Suppose, to make this concrete, that the 40m sales of our fictitious company was 20,000 PCs at $2,000 each. Our presumption is that the market price of $2,000 reflects economic value and we use it as part of our calculation of GDP. In some cases this isn't so easy. In, say, North Korea (or until recently, China), prices do not generally reflect market forces, so it's not easy to calculate economic values. There are also some subtle issues in market economies about how to value nonmarket activities like government spending, housework, pollution, and so on.
    I promised a little while ago to mention the difference between GDP and GNP. GDP is, to me, the more natural concept. It measures total value-added produced by firms operating in the US. GNP, on the other hand, measures value-added generated by factor inputs, capital and labor, owned by Americans. This is slightly different because there are foreign factors (labor and capital) producing in the US and American factors producing abroad. Here's a concrete example. An American working in London for Goldman Sachs would count in US GNP but not US GDP. She would also count in British GDP, since she's working there.
    To clarify the distinction between GDP and GDP take the following example. Suppose that the firm we considered before is partly owned by Japanese owners. Let us also assume that some of the workers in the firm are Japanese managers temporarily working in the U.S. Then:

   Sales revenue                  40,000,000

   Expenses                       26,000,000

     Wages                        20,000,000
        Paid to US workers        18,000,000
        Paid to Japanese managers   2,000,000
     Cost of Parts                 6,000,000

   Net Income                     14,000,000
        Paid to American owners    9,000,000
        Paid to Japanese owners    5,000,000

GDP = 34m = 40m - 6m = 20m + 14m

GNP = GDP - 2m - 5m = 27m = 18m + 9m

GNP = GDP - factors payments to foreigners (dividends, interest, rent to foreign residents owning assets in the US and wages of foreign residents working in the US) + factor payments from abroad to US residents (dividends, interest, rent to US residents owning assets abroad and wages of Americans working abroad).
    The difference between GDP and GNP is not very large in the U.S but can be very large for countries such as Mexico that have a large amount of foreign debt on which they pay interest to foreigners and countries such as Ireland where a large fraction of the factories are owned by foreign multinationals that receive profits and royalties on their Irish operations.

Examples (1987 data):

             GDP  +  Net Factor Income(+)    =  GNP    % difference
                     Payments (-) Abroad               between the two

   US        4540       4                       4544     0.08
   Mexico     192       -9                       183     -4.9
   Ireland    19.9     -1.9                       18      -10

NFA = Net Foreign Assets = Assets owned by Americans abroad - Liabilities of Americans towards foreigners = US Foreign Assets - US Foreign Debt

Assets (and liabilities) include stocks, bonds, loans from banks and other sources, real estate, firm ownership and so on.

If NFA > 0, the country is a creditor country.

If NFA < 0, the country is a debtor country.

If we define with i the:

i = average interest rate (rate of return) on net foreign assets (foreign assets - foreign liabilities)

i NFA = Net factor income from abroad = interest rate times net foreign assets.

Then the GNP is :

GNP = GDP + i NFA = GDP + Net factor income from abroad

Given the above identity, it is easy to see that GNP will be greater (smaller) than GDP if the country is a net creditor (net debtor).

Some examples of the national accounts at work:

1. GDP at factor cost. You'll note in the PC example that we could calculate value-added in two equivalent ways. We can take sales and subtract costs of raw materials: 40m - 6m = 34m. Or we could add up the profits and payments to labor: 20m + 14m = 34m. Double-entry bookkeeping always allows you multiple ways of deriving any number. Both of these methods are used in constructing the national accounts in the US. When there are capital costs these are counted, too, as part of value-added and GDP (next section).

2. Government services. Here there is no figure analogous to sales (unless you think of taxes this way). In the national accounts, value-added is generally computed by adding together payments to labor and (sometimes) capital. For example, payments to Commerce Department employees count as value-added in government services.

3. Imported oil. Suppose that the US economy continues to produce the same quantities of output at the same prices after an increase in the price of oil. The value of this output is, by assumption, the same after oil prices rise, but with more of this value going to oil producers a smaller share is left for domestic factors, capital and labor. The price increase thus leads to a decline in value-added. [Think of the PC assembler: if the cost of parts rises to 8m, what happens to value-added if other costs and revenues stay the same?]

4. Underground economy. By practical necessity only market activity is measured. The old example, not especially relevant these days, is that maids count in GDP but housewives do not. There's some question about the entire underground economy, which by its nature is hard to monitor and does not show up in GDP or GNP. In a curious example, economists recently estimated that Italy had a GDP as large as the UK once they included an estimate of its underground economy.

5. Clean air. There is no market transaction for clean air and pollution, so this aspect of our quality of life is not incorporated in GDP. GDP is not, then, a catchall measure of our well-being. What does show up in GDP is expenditures on pollution control equipment. [Perhaps the EPA's plan to allow firms to trade pollution rights in open markets will change this.]

Accounting Identities

The first is to think of value-added as payments to labor and capital. The point is that sales revenue shows up as income to someone. Intermediate goods are income to the firm that makes them, wages are income to workers, and profits are income to the people who own the firm. As a result, we can think of GDP as measuring either income or output: the two numbers are the same thing.

Let's go back to our PC assembler to see this in action, adding a few things to make it more realistic.

   Sales revenue             40,000,000


   Expenses                  32,000,000

     Wages                   20,000,000
     Cost of Parts            6,000,000
     Interest                 2,000,000
     Depreciation of capital  4,000,000

   Net Income                 8,000,000

Net Domestic Product = GDP - Depreciation = 34m - 4m = 30m

The reason we tend to stick with GDP is that economic depreciation (as opposed to what shows up on financial statements and tax returns) is difficult to measure.

The national income and product accounts do this at the aggregate level, with a couple added complications. The numbers in 1994 looked like this (in billions of dollars):

1. National Income              5,495.1

2. Compensation of employees    4,008.3

3. Proprietor's income            450.9

4. Corporate Profits              526.2

5. Rents                          116.6

6. Net Interest                   392.8

Our second look at GDP comes from the perspective of purchases of final goods: who buys them (consumers, firms, governments, or foreigners). The most common decomposition of this sort is

GDP = consumer expenditures + investment + government purchases of goods and services + net exports,

or, in a more compact notation,

GDP = C + I + G + NX.

Net exports is simply exports (X)  minus imports (M) or NX = X - M.  Net exports are also referred to as the trade balance. Consumption is expenditures on consumer goods by households. Investment in this course will always mean accumulation of physical capital: purchases of new buildings and machines, plant and equipment in the language of national income accountants (a close relative of the beloved PPE of financial accounting). It also includes accumulation of inventories (that is, the change in stocks of inventories). Government consumption here consists of purchases of goods and services (mainly wages) and does not include government outlays for social security, unemployment insurance, or interest on the debt. We think of these, instead, as transfers, since no goods or services are involved. We'll see more of this when we look at the government deficit. U.S. data on the various components of GDP are contained in Tables published in the Economic Report of the President. The data for 1994 are as follows:

                                                          % Share of GDP
GDP                                    6931.4                 100%
Consumption                            4698.7                67.8%
        Durable Goods                          580.9
        Non-Durable Goods                     1429.7
        Services                              2688.1
Gross Private Domestic Investment      1014.4                14.6%
        Non Residential                        667.2
        Residential                            287.7
        Change in Bus. Inventories              59.5
Government Consumption                 1314.7                18.9%
Net Exports of Goods and Services       -96.4                -1.3%
        Exports                                722.0                10.5%
        Imports                                818.4                11.8%

.............................................
Net Factor Incomes from abroad           -9.0

GNP                                    6922.4

Given the definition of net exports as X-M, we can also rewrite the national income identity as:

GDP + M = C + I + G + X

The left hand side of the expression represents the total supply of goods available in the country; such a supply is the sum domestic supply (GDP or domestically produced goods) and foreign supply of goods (imports). The right hand side says that the total supply of goods is purchased either by private consumers (C), firms for investment purposes (I), the government for its own public consumption (G) or foreign agents in the form of exports (X).

The Current Account

We will now define a very important concept,  the current account of the balance of payments, that is quite related to the trade balance (net exports, NX).

Given the definition of GNP, we also get:

GNP_t = GDP_t + i_t NFA_t = C_t + I_t + G_t + (NX_t + i_t NFA_t ) =

= C_t + I_t + G_t + CA_t

where:

CA_t = NX_t + i_t NFA_t

Current Account = Trade Balance + Net Factor Income from abroad

The subscript t refers to a period t variable. If we take data ar a yearly frequency, GNP_t would be GNP in year t, say 1997. The difference between the trade balance and the CA can be very large if a country is a large creditor or debtor.

Example: Brazil in 1986.

NX = + $ 8.3b
CA = - $ 5.3b
i  NFA = -$ 13.6b

In this example, Brazil had in 1986 a large current account deficit in spite of a trade surplus. In fact, Brazil was a heavy foreign debtor, having borrowed a lot in the 1970s and 1980s. By 1986 the total foreign debt of Brazil was above $100b and the net foreign interest payments on that debt (and profit repatriations of foreign firms owning assets in Brazil) equaled $13.6b.

As the table below shows, in Asia large current account deficits (as a share of the country GDP) were prevalent in the 1990s. They resulted from very large trade deficits (NX<0) and, in some countries, large interest payments on foreign debt  (i NFA <0) ; such large current account imblances eventually led to the currency and debt crisis of 1997.

                                Current Account Balance (% of GDP)
                        1990    1991    1992    1993    1994    1995    1996

Korea               -1.24   -3.16   -1.70   -0.16   -1.45     -1.91     -4.89
Indonesia         -4.40   -4.40   -2.46   -0.82   -1.54    -4.25     -3.41
Malaysia          -2.27   -9.08   -4.06  -10.11  -11.51  -13.45    -5.99
Philippines        -6.30   -2.46   -3.17   -6.69   -3.74    -5.06     -5.86
Singapore          9.45   12.36   12.38    8.48   18.12  17.93    16.26
Thailand           -8.74   -8.61   -6.28   -6.50   -7.16    -9.00     -9.18
Hong Kong        8.40    6.58    5.26    8.14    1.98    -2.21      0.58
China                 3.02    3.07    1.09   -2.17    1.17     1.02      -0.34
 

To understand better why a country may be running a current account deficit or surplus, one should notice that the current account is the difference between what a country produces (GNP) and what the country spends (total consumption plus investment). In fact:

CA = GNP - (C + G + I)

where GNP is income and (C +G +I) is domestic spending for consumption and investment purposes (formally called "absorption"). If a country produces more than it spends, the excess of goods produced over those bought at home for consumption and investment purposes must be on net exported to the rest of the world (a positive external balance). So, if GNP > Absorption, the external balance is positive or, equivalently, the current account is in surplus. Viceversa, if If a country produces less than it spends, the excess demand of goods for consumption and investment purposes over income/production  must be on net imported from the rest of the world (a negative external balance). So, if GNP < Absorption, the external balance is negative or, equivalently, the current account is in deficit.

Another way to understand the current account is to see that it is the difference between national savings and national investment. In fact, as for an individual, we can define savings as the difference between income and spending for consumption purposes. If I consume more (less) than my income my savings are negative (positive). In the case of a country consumption is made both by the private (C) and public sector (G). So, by definition, national savings are equal to:

S = GNP - C - G

Substituting this definition of savings in the expression for the current account, we get:

CA = S - I
 

To see why the current account is equal to the difference between savings and investment, consider the similarity of a country  with an individual. For simplicity, suppose initially that the investment of the individual is zero and that G=0. If an individual consumes (C) more than his/her income (GNP), the savings (S=GNP-C) of the individual will be negative (S<0). Since the individual investment is zero, the current account of the individual will be equal to his/her savings (CA=S<0). So, an individual with negative savings has a deficit in its current account. In a similar way, if I=0, a country running a current account deficit is consuming (including both public and private consumption) more than it is producing as CA = S = GNP-C-G.

Consider now how positive investment (I>0) changes things. Take again the case of an individual who has now positive savings (S=GNP-C >0). Suppose now that the individual makes real investments; for example, he/she may buy a new home (residential investment). Suppose that the investment in the new home is greater than the savings of the individual (I > S) as it is usually the case. In this case the current account of the individual is in deficit as CA = S - I <0. Since the income of the individual (GNP) is less than his/her total spending (for consumption and investment), the individual current account is in deficit, or the individual's savings are below the individual's investment. The same story holds for a country.  If a country invests more than its saves,  the country is producing an amount of output/income (GNP) that smaller than the total spending on goods for consumption and investment purposes (C+G+I). Therefore, the excess of spending (absorption) over income or, equivalently, the excess of investment over savings implies that the country is running a current account deficit.
 
 

Insight in the Asian economic crisis: Why current account deficits lead to the accumulation of a large stock of foreign debt.

It is very  important to understand that if a country runs a current account deficit (CA<0), as it is the case in many developing countries such as those currently in crisis in Asia , this means that the country is borrowing from the rest of the world and its foreign debt will increase over time. Thus, flows (items on income and cash flow statements) translate into changes in stocks (balance sheet items, like household wealth, the stock of capital, government debt, and net foreign debt).

To understand this important point, we need to be more specific about the distinction between stocks and flows. A stock is measured at a particular point in time such as the stock of capital at the end of 1997. A flow instead represents the change in the stock over a particular period of time: for example net investment in capital in the year 1997 is equal to the difference between the stock of capital between the end of 1997 and the end of 1996. So, if we define with K the stock of physical capital, this stock is related to the flow of net investment (I - depreciation) by:

K_t+1 = K_t+ I_t - Depreciation_t

or:

Stock of K at time t+1 = Stock of K at time t + (Net Investment in new capital in period t)

Then, the flow of new investment is equal to the change in the stock of capital

I_t - Depreciation_t = K_t+1 - K_t

Note that macroeconomists typically measure K at replacement cost rather than book value.

Similarly, the current account in the year 1997 is equal to the difference in the stock of net foreign assets of the country between the end of 1997 and the end of 1996. A current account surplus results in an increase in the net foreign assets of a country while a current account deficit results in a decrease of these assets or, if the country is already a net debtor, it results in an increase in the net foreign debt of the country.

    To understand why a current account deficit leads to an increase in the stock of foreign debt of a country, consider the similarity of a country  with the budget constraint of an individual. For simplicity, suppose initially that the investment of the individual is zero (I=0). If an individual consumes (C) more than his/her income (GNP), the savings (S=GNP-C) of the individual will be negative (S<0). Since the individual investment is zero, the current account of the individual will be equal to his/her savings (CA=S<0). So, an individual with negative savings has a deficit in its current account. If the individual has an initial positive wealth (NFA=(Assets-Liabilities)>0), then these negative savings (current account deficit)  will lead to a fall of his/her net wealth (assets minus liabilities) as he/she will run down his/her assets or, for given gross assets, he/she will borrow to pay for the excess of the consumption over income. In either case (regardless whether gross assets are run down or new gross borrowing are made) his/her net wealth will fall as personal assets fall and/or personal debt goes up.  If savings are negative year after year, at some point net assets will fall to zero and the individual will become a net debtor (assets-liabilities < 0). In this case negative savings will lead over time to a growing net debt of the individual.
    In a similar way, if I=0, a country running a current account deficit is consuming (including both public and private consumption) more than it is producing as CA = S = GNP - C- G.  Therefore, to finance such a deficit the country needs to run down its assets and/or borrow to pay for the excess of consumption (C+G) over income/output (GNP). In either case (regardless whether gross assets are run down or new gross foreign borrowing are made) the country's net foreign wealth (NFA = Foreign Assets - Foreign Liabilities)  will fall as foreign assets fall and/or foreign debt goes up. If the country is initially a net creditor (NFA>0), over time current account deficits will lead the country to become a net debtor (NFA<0) as net assets fall and eventually become negative; to finance the deficit, each year the country will borrow from the rest of the world an amount of funds that is equal to the excess of income over consumption. So the new borrowing (the increase in foreign debt)  is equal each year to the current account deficit. So, if a country is already a net debtor, further current account deficits will lead this country to increase its stock of net foreign debt.
    Consider now how investment changes things. Take again the case of an individual who has now positive savings (S=GNP-C >0). Suppose now that the individual makes real investments; for example, he may buy a new home (residential investment). Suppose that the investment in the new home is greater than the savings of the individual (I > S) as it is usually the case. In this case the current account of the individual is in deficit as CA = S - I <0.  To finance the excess of his/her investment over savings, the individual can do two things: either run down his/her financial assets (if there are enough assets to be run down) and/or borrow to finance the new investment. In either case, the excess of I over S leads to a reduction of the net assets (assets-liabilities) of the individual. If such current account deficits occur over time net assets will fall to zero and the individual will become a net debtor; the increase in stock of debt will be each year equal to the current account deficit.
    The same holds for a country that has a current account deficit. If a country invests more than its saves, it has to borrow from the rest of the world to finance this deficit. In fact, a CA deficit means that the country is producing an amount of output/income (GNP) that falls short of the total spending on the goods of the country ( the sum of consumption and investment):

CA = GNP - C - G - I

To finance the excess of investment over savings, the country can do two things: either run down its financial foreign assets (if there are enough foreign assets to be run down) and/or borrow from the rest of the world to finance the new investment. In either case, the excess of I over S leads to a reduction of the net foreign assets (foreign assets - foreign liabilities) of the individual. If such current account deficits continue year after year net foreign assets will fall to zero and the country will become a net debtor; in each year the increase in stock of foreign debt will be equal to the current account deficit. More formally, the change in the net foreign asset of a country (a change in stocks) will therefore be equal to the current account (a flow) or:

NFA_t+1 - NFA_t = CA_t

If CA>0 net foreign assets will increase (or net foreign debt will become smaller if the country was starting with net foreign debt, NFA<0); if CA<0 net foreign assets will decrease (or net foreign debt will become bigger if the country was starting with net foreign debt, NFA<0). In each period net foreign borrowing will be equal to the current account deficit (or net accumulation of foreign assets will be equal to the current account surplus).

Another way to see that the previous equation holds is to notice that  the net foreign assets at the beginning of next period (t+1) must be equal to those in period t plus total national income (GNP) minus the part of national income that is consumed (C and G) or invested (I):

NFA_t+1 = NFA_t + GDP_t + i_t  NFA_t - C_t - G_t - I_t = NFA_t + CA_t

Therefore:

NFA_t+1 = NFA_t + CA_t = NFA_t + NX_t + i_t NFA_t

We refer to NFA_t as the initial balance and NFA_t+1 as the ending balance.

The above discussion clarifies why some countries have a very large stock of foreign debt: like in the case of an individual, if  you consume and invest more than you produce (earn income) year after year, you must borrow over time to finance this current account deficit (excess of consumption and investment over income or excess of investment over savings). Therefore, your individual's or country's net foreign debt must increase over time. So countries with a large stock of foreign debt have had in the past large current account deficits that have led to an accumulation of this debt. This is very important to understand what happened in Asia in 1997. During the 1990s, all the Asian "crisis countries" run very large and increasing current account deficits as their national income (GNP) was below their domestic absorption (C+G+I) (or as their investment rates I were above their savings rates); this led to a large accumulation of foreign debt that eventually became unsustainable.

What Causes Current Account Deficits? Are Such Deficits Bad?

Now that we have understood the meaning of the current account and how it relates to the foreign debt of the country, we want to analyze in more detail the link between the current account, private savings and government budget deficits. This will help us to understand whether current account deficits are caused by budget deficits (the "twin deficits" hypothesis).

We take our earlier national income account  identity (GNP = C + I + G + CA) and do a little algebra to get:

(GNP_t -T_t -C_t ) = I_t + (G_t -T_t ) + CA_t ,

where

GNP_t - T_t - C_t = S_t^p= Private Savings

and T_t are taxes collected by the government (TX_t ) net of transfer payments (TR_t ) and interest payments on the public debt (i_t Debt_t ). So:

T_t = TX_t - TR _t - i_t Debt_t .

T is intended to measure all revenues and expenses of the government not included in G, so G-T is the government deficit, NIPA version, a close relative of the number bandied about in the business press. It's only a relative because (i) the press generally focuses only on the federal government and (ii) the Administration and Congress typically have more imaginative measures of the deficit. Note the sign convention: unlike what you generally do in accounting, a deficit is a positive value of G-T. Continuing with the identity: GNP-T measures the amount of income households have on hand once we take into account things like taxes paid to the government, social security payments, and interest on the government debt. GNP-T-C is thus the amount of income households do not spend on goods and services, namely private saving S^p. Conversely, we can define public (government savings) S^g as the difference between government revenues and spending. So:

Def_t = (G_t - T_t ) = G_t - TX_t + TR_t + i_t Debt_t = - S_t^g

or

S_t^g = - Def_t = T_t - G_t

Thus we can write the identity

S_t^p = I_t+ Def_t + CA_t (1)

where Def = G-T is the government deficit as measured by the NIPA. This connects private saving, investment, the government deficit (negative public savings) and the trade balance. Sometimes we combine S and Def, as in

S_t = S_t^p - Def_t = S_t^p+ S_t^g = I_t + CA_t

or

S_t = I_t + CA_t (2)

that implies our earlier definition of the current account:

CA_t = S_t - I_t(3)

where S is a comprehensive measure of national savings, the sum of private and public savings or, if the government is running a deficit, it is total savings net of government dissavings.

The first identity (1), which is based on flows of goods, suggests our earlier interpretation of how current accounts lead to a change in the stock of assets. Private savings, under this interpretation, are a source of new financial capital, since saving leads to purchases of assets. Savers can purchase either corporate securities (which finance new investment by firms in plant and equipment, I), government securities (which go to finance the government deficit, Def), or foreign securities (which finance a current account surplus if CA is positive); the latter purchase of foreign assets leads to an accumulation of net foreign assets. If the CA is negative, this means that private savings are not enough to finance both investment and the budget deficit; therefore foreign savings (borrowing from the rest of the world in the form of an accumulation of foreign debt) is required to finance the excess demand of funds by firms (for investment) or government (for deficit financing purposes) relative to the quantity of private savings . This also tells us, for example, that the government and private industry may be competitors in capital markets for the pool of private savings: if the government takes more, there is less to support private investment. The second identity expresses national savings (S) as equal to national investment (I) plus the current account (CA). The third identity expresses the current account (CA) as the difference between national savings (S) and national investment (I).

There are a couple of connections here that get one thinking about the operation of the economy. One is the connection between the government deficit (Def = G-T) and the current account deficit (-CA ). A government deficit must be matched by some combination of higher saving, lower investment, or a trade deficit. To the extent it's the latter, a large government deficit will be associated with a large trade (current account) deficit. One of the questions we want to keep in mind for the future is whether the trade deficit is largely the result of the government deficit, rather than more fundamental problems with US competitiveness. Another issue is the relation between saving and growth. Two of the things we know are (i) countries that save a lot are also countries that invest a lot and (ii) countries that invest a lot grow faster. We'll return to (ii) in a week or two. For now, let me say simply it's not clear what the direction of causality here: whether higher investment leads countries to grow faster, or countries that grow fast for other reasons (technology?) invest a lot. It's clear, though, that growth and investment are closely related in the data. As for (i), I've computed ratios of S, I, and CA to real GNP (defined with the variable Y) for a number of major countries, and reported them in Table 1. The definition of saving is here total national savings

S = Y - C - G

We then have the identity S = I + CA . You see in Table 1 that the US saves and invests much less, as a fraction of national output, than most other developed countries. Japan, on the other hand, saves and invests substantially more. You might plot the growth rates vs saving and investment rates to see how they are related.

Finally, note that, given our definition of budget deficits, and our previous discussion of how flows lead to changes in stocks, we can show that a government deficit results in an increase in the stock of government debt or:

Debt_t+1 = Debt_t+ G_t - T_t = Debt_t+ (G_t + TR_t  - TX_t) + i_t Debt_t

We refer to Debt_t as the beginning balance and Debt_t+1 as the ending balance.

Another detail. You might be asking yourself (if not, don't) why all taxes are paid by households: what about the corporate income tax? The answer is that firms are owned (for the most part) by households and we are consolidating their books. We attribute to households all the before-tax profits of firms (in value added). We then have them pay the firms' taxes. This is equivalent to just giving them after-tax profits in the first place. The only fudging arises with firms not owned by Americans. In the real accounts the rest of the world (i.e., foreigners) can own some US firms, pay taxes, collect interest on US government debt, and so on, which would complicate the international part of the accounts. For most of this course we'll ignore that to make things simpler. Life is complicated enough as it is.
 

Are Current Account Deficits Good or Bad? Are Large Deficits Sustainable?

The recent experience in Asia shows that large current account deficits led to an accumulation of foreign debt that eventualy became unsustainable and led to a currency crisis. This leads to the following question: is it a bad idea to run a current account deficit? The answer is actually quite complex because running a current account deficit may me a good or bad, sustainable or not sustainable, depending on the cause of the current account deficit.

To specify a definition of sustainability, consider a situation where current macroeconomic conditions continue (i.e. there are no exogenous shocks) and that there are no changes in macroeconomic policy.  In this instance the current account deficit can be argued to be sustainable as long as no external sector crisis occurs.  An external sector crisis could come in the form of an exchange rate crisis or a foreign debt crisis.  An exchange rate crisis could be a panic that leads to the rapid depreciation of the currency or a run on the central bank’s foreign exchange reserves.  A debt crisis could be the inability to obtain further international financing or to meet repayments or an actual default on debt obligations.  A sustainable current account deficit is one that can be maintained without any of these crises occurring.  Of course, sustainability can only be judged after the fact, but we will be examining the characteristics of the economy that are indicative of crises occurring.

If we rewrite our definition of the current account, we can see that there are three main causes of current account deficits:

CA_t = S_t^p - I_t - Def_t

A current account deficit may be caused by:

1. An increase in national investment

2. A fall in national savings; specifically:

    2a. A fall in private savings and/or
    2b. An increase in budget deficits (a fall in public savings).

We want to show that a current account deficit may be bad or good depending on its source.

1. A boom in domestic investment.
We consider first the case where the current account deficit is caused by a boom in investment. In this case running a current account deficit is a good idea and the accumulation of foreign debt associated with the deficits should not be viewed with concern. To see why, notice that a country is like a firm. Suppose that a firm has identified good profitable investment projects but that the savings of the firm (i.e. the firm's retained earnings) are below the value of profitable investment projects. Then, it makes sense for the firm to go to capital markets external to the firm and borrow funds equal to the difference between the value of the new investment projects and the firm's savings (retained earnings). This firm borrowing can take various forms: it could borrow funds from banks; it could issue corporate bonds or it could issue new equity that is purchased by agents in the economy. Such borrowing by the firms is optimal as long as the financed investment projects are profitable (i.e. as long as the return on the investment is as high as the cost of borrowed funds). In fact,  over time, the earnings generated by the capital created by the new investment will be sufficient to pay back the principal and interest on the borrowed funds.
    Now, note that a country is like a firm as in a country thousands of firms make individual investment decisions. Suppose that the country experiences an investment boom. The reasons for such investment boom can be several: new natural resources are found in the country (oil, minerals); technological progress leads to new products that can be profitably developed and produced; structural economic reforms (like trade liberalization or capital market liberalization) or macroeconomic stabilization policies (such as a reduction in inflation, a cut in budget deficits and reduction in distortionary taxes on income and capital) lead to expectation of high future economic growth and high profitability of new investments.
In all these cases, the country will have an investment boom that has to be financed with some savings. If the national savings of the country (the sum of private and public savings) are not sufficient to finance all new profitable investment projects, then it is optimal for the country (like it was for a firm)  to run a current account deficit, i.e. rely on foreign savings to finance the excess of investment over national savings. Such a current account deficit will imply the accumulation of new foreign debt, i.e. a capital inflow as foreign funds will be borrowed to finance domestic investment. The forms of such a capital inflow are similar to those of a firm. First, the country (or better the country's firms) could directly borrow from foreign banks; second, the domestic firms could borrow from domestic banks but these in turn borrow from foreign banks; third, the firm could issue new bonds that are bought by foreign investors; fourth, the firm can issue new equity that is purchased by foreign investors.  Finally, if the new investment is originally made by a foreign firm that has decided to build a new plant in the domestic economy, the flow of foreign capital that finances this investment project is called Foreign Direct Investment (FDI). In all these cases, a current account deficit (CA= S-I <0) is financed by some form of foreign saving (foreign capital). And, as in the  case of a domestic firm, it is optimal for the country to borrow funds from the rest of the world and accumulate foreign debt as long as the new investment projects are profitable. Over time, the goods produced by the new capital will lead to increased country exports that will generate the trade and current account surpluses that are necessary to eventually repay the foreign debt and interest on it.
    So, in general a persistent current account deficit and foreign debt accumulation generated by a boom in investment should not be considered with too much concern and it might actually increase the rate of growth of an economy where domestic savings are not sufficient to finance all profitable investment projects. There are however several caveats to be made to this argument.
    First, borrowing form the rest of the world to finance investment that produces new goods is especially good if the new investments are in the traded sector of the economy (i.e. the sectors of the economy that produce goods that can be sold in foreign markets). In fact, at some point in time the foreign debt has to be repaid back and, for a country, the only way to pay back foreign debt it to run at some point trade and current account surpluses. If the new investments are instead in the non-traded sector of the economy (such as commercial and residential investment), they create goods (housing services)  that cannot be sold abroad. So, in this case the long run ability of the country to repay its debts through trade surpluses may be limited and this can create a problem. For example, many Asian countries in the 1990s were running large and increasing current account deficits that were financing new and excessive investments in the non-traded real estate sector (residential and commercial building). Such investments went bust in 1996-97 because of a glut of real estate and the collapse of the real estate asset price bubble that lead to a rapid fall in the price of land and real estate values; then,  the firms and individuals that had borrowed foreign funds (and/or the banks that had borrowed the foreign funds and in turn lent these funds to domestic firms and households) to finance real estate investments went all into a financial crisis. They had borrowed too much in foreign currency to finance investments that had a low or negative returns. Moreover, the exchange rate depreciation associated with this crisis made things worse as the value in domestic currency of funds borrowed in foreign currencies (Dollars, Yen, Marks) increased enormously once the currencies depreciated rapidly. This real increase in the burden of foreign debt caused a financial crisis for the banks, firms and individuals heavily exposed in non-traded sectors (such as real estate) and led to widespread bankruptcies. So the first caveat is that is is dangerous to run a current account deficit to finance excessive investments in non-traded sectors of the economy.
    The second caveat is relevant both for traded sector firms and non-traded sector firms. Every firm knows that it is optimal to borrow funds to finance investments only as long as the return on these investments are at least as high as the cost of the borrowed funds; otherwise, a firm that borrowed too much and invested in bad projects will eventually experience losses, a financial crisis and potentially go bankrupt if most investments turn out to be bad. The story of the Asian crisis is in part one of a current account deficit and foreign debt accumulation caused by a boom of investment that turned out to be excessive. In Asia, there were too many investments (both in traded and non-traded sectors) that turned out to be not very profitable.
    How can one rationally explain such overinvestment in wrong projects? Why did the firms make such investments and borrow the funds? Why did the domestic banks lend them the funds and did not monitor the quality of the investments? To see understand this we need to introduce some politics and the behavior of governments. Many governments in Asia were trying to maximize the rate of economic growth; since growth and the production of goods requires a lot of labor and capital, a necessary condition for high economic growth is a very high rate of national investment. It appears that many governments in the region were pursuing economic growth targets that were excessive. Governments gave incentives (such as subsidies) to firms to invest too much and incentives to the domestic banks to borrow too much from abroad to finance dubious investment projects by the firms.
    Banks, in turn,  borrowed too much from abroad for many reasons, mostly related to the implicit promise of a government bail-out in case things went wrong: first, their risk capital was usually small and owners of banks risked relatively little if the banks went bankrupt; second, several banks were public or controlled indirectly by the government that was directing credit to politically favored firms, sectors and investment projects; third, depositors of the banks were offered implicit or explicit deposit insurance and therefore did not monitor the lending decisions of banks; fourth, the banks themselves were given implicit guarantees of a government bail-out if their financial conditions went sour because of excessive foreign borrowing; fifth, international banks (Japanese, American and European ones) lent vast sums of money to the domestic banks of the Asian countries because they knew that governments would bail-out the domestic banks if things went wrong. The outcome of all this was twofold: first, banks borrowed too much from abroad and lent too much to domestic firms; second, because of all these implicit public guarantees of bail-out, the interest rate at which domestic banks could borrow abroad and lend at home was low (relative to the riskiness of the projects being financed) so that domestic firms invested too much in projects that were marginal if not outright not profitable. Once these investment projects turned out not to be profitable, the firms (and the banks that lent them large sum) found themselves with a huge amount of foreign debt (mostly in foreign currencies) that could not be repaid. The exchange rate crisis that ensued made things only worse as the currency depreciation dramatically increased real burden in domestic currencies of the debt that was denominated in foreign currencies.

2. A current account deficit caused by a fall in national savings: a fall in private savings  or an increase in budget deficits (a fall in public savings).

Apart form the previous case of an investment boom, a current account deficits may also be caused by a fall in national savings. A current account imbalance caused by a fall in the national savings rates can be due to either a fall in private savings or in public savings (higher budget deficits). A fall in national savings caused by lower public savings (higher budget deficit) is potentially more dangerous than a fall in private savings. The reason for this is that a fall in private savings is more likely to be a transitory phenomenon while structural public sector deficits are often hard to get rid of.  The private savings rate will recover when future income increases occur. On the other hand, large and persistent structural budget deficits may result in an unsustainable build-up of foreign debt. For example, in the late 1970s many developing countries were running very large budget deficits to finance large and growing government spending; to finance these deficits, the governments borrowed heavily in the world capital markets (either directly from international banks or indirectly by issuing bonds purchased by foreign investors). In this case, the large and growing budget deficits led to large current account deficits and the accumulation of a very large stock of foreign debt. By 1982, the size of this public foreign debt was so large (often close to or above 100% of GDP) that many governments began having difficulties in repaying interest and/or principal on their foreign liabilities; therefore, a severe Debt Crisis emerged in  the 1980s with many countries risking default on their foreign debt and having to negotiate a rescheduling of their foreign liabilities. So the lesson is that running current account deficits and borrowing from abroad to finance budget deficits is a dangerous game that will eventually lead to a debt crisis. Unlike firms that borrow to finance investment projects that will be eventually self-financing (as they generate trade surpluses that will be used to repay the original foreign debt), fiscal deficits are rarely self-financing, especially if such deficits are chronic, the result of excessive spending and structural lack of tax revenues.

Unlike the case of a current account caused by a fall in public savings (a larger budget deficit), a current account caused by a fall in private savings is usually considered with less concern. A fall in private savings rate may be transitory and occur when expectations of higher future GDP growth result in an increase in current consumption above current income. For example, an MBA student in school will usually have zero or close to zero income in the two years he/she is in school. Since consumption is positive while in school (you got to eat and cloth to live!), the student has negative savings (S=GNP-C < 0 as GNP=0 and C>0) and a current account deficit. [Note also that the student  is borrowing money not only to finance its negative savings but also to finance its MBA tuition: this is an Investment in human capital that will eventually lead to higher income; so it is also optimal to borrow to finance that tuition investment]. In this case, negative savings lead to a current account deficit and accumulation of personal debt; however, this borrowing is optimal since the student is consuming today not on the basis of his/her current low income but on the basis of its permanent income that is high because of the expected higher income after school. So, this transitory fall in savings and accumulation of debt is optimal since the higher income after school will be above consumption and lead to the repayment of the debt incurred while in school. The same happens for a country: an economic reform or stabilization may lead to a consumption boom (especially purchases of durable goods) even if current incomes have not increased yet so much because households in the economy expect high future incomes because of the expectations of future high economic growth. In this case, current consumption (C) goes up a lot today while income (GNP) grows only over time; this consumption boom leads to a fall in private savings; the ensuing current account deficit is financed (at the aggregate country level) through an inflow of capital from abroad. This accumulation of foreign debt is not worrisome as long as future income growth is realized and individuals are able to repay their debts (foreign liabilities).

Needless to say, many episodes of unsustainable current account deficits do not fit the patterns described. For example,  the deterioration of the current account balance in the years preceding the 1994 Mexican peso crisis was largely due to a fall in private savings. In the Mexican episode, the boom in private consumption and the sharp fall in private savings rates was fueled by the combined forces of overly optimistic expectations about future growth and permanent income increase together with the loosening of liquidity constraints on consumption deriving from the liberalization of domestic capital markets. Under such conditions, the fall in private savings rates led to a rapid and eventually unsustainable current account deterioration. Moreover, while the 1980s foreign debt crisis was caused by very large budget deficits, more recent episodes of debt crisis do not seem to have their source in a fiscal imbalance. For example,  the 1990-94 Mexican episode and the 1997 Asian crises occurred in spite of the fact that the fiscal balances were in surplus; the large and increasing current account deficits and foreign debt accumulation were caused by the private sector behavior, a fall in private savings and an increase in investment. This suggests that current account deficits that are driven by structurally low and falling private sector saving rates may be a matter of concern even if they are the results of the "optimal" consumption and savings decisions of private agents. This is especially the case when the private consumption boom, like in Asia in the 1990s, is in part the consequence of an excessively rapid liberalization of domestic financial markets that gives access to credit to households that were previously borrowing-constrained.

Whether a large current account deficit is sustainable or not also depends on a number of other macroeconomic factors: 1. the country's growth rate; 2. the composition of the current account deficit; 3. the degree of openess of the economy (as measured by the ratio of exports to GDP); 4. the size of the current account deficit (relative to GDP).

1. Large current account deficits may be more sustainable if economic growth is higher. High GDP growth tends to lead to higher investment rates as expected profitability  increases.  At the same time, high growth might lead to higher expected future income and (as noted above) transitory declines in private savings rates. Generally, higher growth rates are related to more sustainability of the current account deficit because, everything else equal, higher growth will lead to a smaller increase in the foreign debt to GDP ratio and make the country more able to service its external debt. However,, many episodes of unsustainable current account deficits do not fit the patterns described.  In particular, the examples of Chile in 1979-81, Mexico in 1977-81 and the Asian countries in 1997 come to mind.  In all these instances the average real GDP growth rate in the years preceding the crisis was above 7%: what happened was that excessively optimistic expectations that the high economic growth would persist for the long-term led to an excessive investment boom and a boom in private consumption (a fall in private savings) that resulted in current account deficits and growth of foreign debt; the latter eventually became unsustainable and caused a currency and debt crisis (as in Asia in 1997-98).

2. The composition of the current account balance which is approximately  equal to the sum of the trade balance and the net factor income from abroad will affect the sustainability of any given imbalance.  A current account imbalance may be less sustainable if it is derived from a large trade deficit rather than a large negative net factor income from abroad component. In fact, for a given current account deficit, large and persistent trade deficits may indicate structural competitiveness problems while large and negative net foreign factor incomes may be the historical remnant of foreign debt incurred in the past.

3. Since a country's ability to service its external debt in the future depends on its ability to generate foreign currency receipts, the size of its exports as a share of GDP (the country's openness) is another important indicator of sustainability.

4. Most episodes of unsustainable current account imbalances that have led to a crisis have occurred when the current account deficit was large relative to GDP.  Lawrence Summers, the U.S. deputy Treasury secretary, wrote in The Economist on the anniversary of the Mexican financial crisis (Dec. 23, 1995-Jan. 5, 1996, pp. 46-48)  “that close attention should be paid to any current-account deficit in excess of 5% of GDP, particularly if it is financed in a way that could lead to rapid reversals.”  By this standard, many of the Asian economies provided ample source for concern in the 1990s as they had very large and increasing deficits, well above the 5% red flag.

The above analysis suggest that there is not anything inherently good or bad about a current account deficit. Like and individual or a firm that borrows funds, a country may be borrowing funds from the rest of the world for good or bad reasons. So a current account deficit and the ensuing accumulation of foreign debt may be good, sustainable and lead to higher long-run growth or may be eventually unsustainable and lead to a currency and debt crisis depending on what drives the current account deficit. We will return to the discussion of current account and foreign debt sustainability in Chapter 3.

Prices and Real Quantities

As a result, a great deal of effort goes into measuring "real'' (as opposed to "money'' or "nominal'') GDP and related quantities and constructing indexes of "average'' dollar prices. For GDP we would generally like to compare quantities of output produced in different periods, so that an increase in GDP means we are producing more of something.

How to measure correctly the real value of GDP and the correct level of the inflation rate is a difficult issue. Until the end of 1995, the U.S. followed a "fixed-weight" approach to the measurement of real GDP but has since moved to a "chain-weight" method. This move was. however, somewhat controversial and object of a serious debate. For what concerns the inflation rate, we can measure it by using the price deflator series derived from the calculation of real and nominal GDP or we can measure it by calculating the CPI (Consumer Price Index) inflation rate. Recently, however, it has been argues that the CPI inflation rate overstates the true inflation rate. In December 1996, the Boskin Commission appointed by the Senate Finance Committee, reached the conclusion that the CPI overstates the annual inflation rate by 1% to 2% per year. To understand these recent debates on the correct measurement of GDP and inflation, we need to consider in more detail these issue. In particular, we need to start by understanding why the US switched from a fixed-weight to a chain-weight method to measure real GDP and why the CPI inflation rate might be overestimated. Let us start with the fixed-weight GDP measure.

Suppose, for example, we want to compare GDP in 1993 to GDP in 1992. The (fixed-weight) measures of nominal and real GDP using 1987 as the base year (the method used until the end of 1995) were:
 

	Nominal GDP	Real GDP
1987	4539.9	4539.9
1992	6020.2	4979.3
1993	6343.3	5134.5

The growth rate of nominal GDP in 1993 was:

5.3% = 100 x (6343.3 - 6020.2)/6020.2

But how much of that reflects a decline in the value of the dollar? What we might do is measure the 1992 and 1993 quantities and value them at the same prices to get a "constant'' price comparison. The NIPA, for example, used to measure everything in 1987 prices; 1987 is referred to as the base year. This was a "fixed-weight" method since it implied measuring quantities of goods in different years at the prices prevailing in a base year. Using this method, GDP in 1987 prices was 4979.3 in 1992 and 5134.5 in 1993, implying a grow rate of real GDP of

3.1% = 100 x (5134.5 - 4979.3)/4979.3

Thus it appears that 2.2 percent (5.3% - 3.1%) of the growth in current dollar GDP was simply a general increase in dollar prices of goods.

This general increase in prices is implicit in the real and nominal measures of GDP. One measure of the average price is the ratio of GDP in current prices to GDP in 1987 prices. We call this measure of prices the GDP implicit price deflator:

GDP Price Deflator = GDP in current prices (Nominal GDP) / GDP in base year prices (Real GDP)

Nominal GDP (NY) = Real GDP (Y) x GDP deflator (P)

Or:

NY_t = Y_t x P_t

where the subscript refers to the year t value of the the corresponding variable. We typically report this price deflator as an index, with 1987 = 100. The index was
 

1987	100
1992	120.9 = 100 x 6020.2/4979.3
1993	123.5 = 100 x 6343.3/5134.5

for an inflation rate of 2.2 percent (= 126.3/121.3 -1).

Here, we are defining the inflation rate p as the % rate of change of the price level (the GDP deflator) between period t-1 and period t, or:

p_t = (P_t - P_t-1)/P_t-1 = inflation rate in year t.

More formally, the rate of growth of nominal GDP (ny_t) is equal to the rate of growth of real GDP (y_t ) plus the rate of inflation. In fact:

(ny)_t = (NY_t - NY_t-1)/NY_t-1 = (NY_t / NY_t-1) -1 = (Y_t x P_t) / (Y_t-1 x P_t-1) - 1 =

(Y_t / Y_t-1) x (P_t / P_t-1) - 1

Therefore:

ny = ( 1 + y) x (1 + p) - 1 = y + p + yp (*)

Since yp is a small number, the expression (*) is approximately equal to:

ny_t = y_t + p_t

Or:

(NY_t - NY_t-1)/NY_t-1 = (Y_t - Y_t-1)/Y_t-1 + (P_t - P_t-1)/P_t-1.

Figure 6 shows the levels of nominal and real GDP for the U.S. economy; note that since the base year for the comparison is 1992, nominal and real GDP are equal to each other in that year as the deflator is equal to 1 by choice of the base period. Figure 7 presents a graph of the rate of growth of nominal and real GDP for the U.S. economy. As inflation is positive, nominal GDP growth is above real GDP growth.

This is simply one example of a price measure. There are also price deflators for components of GDP: consumption, investment, government spending, exports and imports. The most common measure of price movements, though, has nothing to do with the national income accounts.

The Consumer Price Index measures the dollar price of a "fixed basket'' of goods rather than the constant price of a changing basket of goods used to compute the "fixed-weight" GDP and its nominal price deflator.

The idea is to calculate the price of a constant list of goods at different points in time. Eg, consider 5 gallons of gas, one haircut, 2 pounds of chicken, 3 bottles of soda, and so on. The Bureau of Labor Statistics at the Department of Labor sends people to stores every month to collect prices of the various goods, and then computes prices of various "baskets.'' The Consumer Price Index (CPI)  is the total price of all of these goods at different dates, normalized to equal 100 at some date. Same idea, really, as the Dow Jones Industrial Average or the S&P 500. The CPI takes its basket of goods from the typical spending patterns of an American family.

The conceptual problem for both price indexes---the fixed-weight GDP deflator and the fixed basket CPI deflator ---is that it's not clear how to measure the purchasing power of the dollar when the dollar prices of different goods are changing at different rates. Conversely, it's not clear how to combine quantities of different goods when their relative prices are changing. As usual, this is easier to see with an example.

Example (made-up numbers).

Our economy produces two goods, fish and and chips (computer chips, not potato ones). At date 1 we produce ten fish and and ten chips. Fish cost 0.25 cents and chips 50 cents. At date 2 the price of fish has risen to 50 cents and of chips to 75 cents and the quantities have changed to 8 and 12.
 

	Price of Chips	Quantity of Chips	Price of Fish	Quantity of Fish
Date 1	0.5	10	0.25	10
Date 2	0.75	12	0.50	8

Note that the two prices have not gone up by the same amount: fish inflation is 100 percent but chip inflation is 50 percent. Another way to say the same thing is that the relative price of chips to fish has fallen from 2 (=.50/.25) to 1.5 (=.75/.50). What is the change in the price level?

Example continued (fixed-weight GDP deflator and fixed-weight real GDP). We construct GDP at both dates in current prices and in date 1 prices.

Date 1 Nominal GDP = $7.50 (= .50x10 + .25x10)

At date 2

Date 2 Nominal GDP = 13.00 (= .75x12 + .50x8).

In date 1 prices ("real'') GDP is:

Date 1 prices ("real'') GDP = 8.00 (= .50x12 + .25x8).

The GDP deflator (the ratio of current price GDP to GDP in base year prices, here date 1) rises from 1.0 in the base year date 1 to 1.625 (= 13/8) at date 2, an inflation rate of 62.5 percent.

The the, real GDP growth measured with fixed weights is:

6.66% = 100 x (8-7.5)/7.5

In fact, since we know from (*) above that::

(1 + ny) = ( 1 + y) x (1 + p)

real growth y is:

y = [(1 + ny) / (1+p)] -1 = [(1 + 0.733)/(1 + 0.625)] -1 = 0.066

Consider now what happens to our measure of real GDP growth when we use a "fixed-basket" based measure of inflation (the CPI index).

Example continued ("fixed-basket" CPI deflator and real GDP). The consumer price index uses quantities in a base year to compute the costs of the same basket of goods at 2 different dates. Let's say here that the basket of goods is 10 fish and 10 chips (the same composition as GDP). Then:

CPI at date 1 = 7.50 (= .50x10 + .25x10).

CPI at date 2 = 12.50 (= .75x10 + .50x10).

The implied CPI inflation rate is 66.6 (= 100 x (12.50-7.50)/7.50) percent.

Note the difference between the two indexes: the CPI uses date 1 quantities while the GDP deflator uses date 2 quantities to compute the date 2 price index. (Check out the CPI Calculation Machine at the Minneapolis Fed home page to get, say, the price of a cup of coffee in 1963).

Since nominal GDP growth is again 73.3% and the fixed-basket (CPI based) measure of inflation is 66.6%, now the fixed basket measure of real GDP is 4% rather than the higher 6.66% obtained by using the fixed-weight method. In fact:

y = [(1 + ny) / (1+p)] -1 = [(1 + 0.733)/(1 + 0.666)] -1 = 0.04

How can we compute directly the real GDP growth if we use the CPI deflator ? Simple: compute real GDP in the second period by taking period 2 as the base year (rather than period 1 as in the fixed-weight method). Then:

Period 2 Real GDP using date 2 as the base year: 13 = 0.75x12+0.5x8

Period 1 Real GDP using date 2 as the base year: 12.5 =0.75x10+0.5x10

Implied Real (fixed-basket) GDP growth using period 2 as base year: 4% =(100 x (13-12.5)/12.5)

You see that, depending on which deflator we use, our estimate of real GDP growth will be different (6.66% versus 4%).

So which method is better ?

The point is this: there is no unique or best way to separate relative price movements from general movements in the price level, even in theory. This problem involves some subtle issues about price measurement, like what quantities to use, date 1 or date 2. How much difference does this make in practice? Some, but in high inflation periods, especially, the movements in different prices indexes are similar. You can see this from the graphs of the CPI and GDP deflator in levels and rates of change (Figure 8 and Figure 9).

Note also that, the fixed-weight method used by the US until 1995 had the disadvantage that it was giving too much weight in the calculation of real GDP to the good whose relative price had fallen over time (in this example chips). Because of this bias, the value of the real output of chips was overestimated and led to an overestimation (6.66%) of the value of the growth rate of the economy.

To see this issue in more detail consider the following example:
 

	Price of Chips	Quantity of Chips	Price of Fish	Quantity of Fish
Date 1	1	10	1	10
Date 2	0.5	20	2	5

In this example:

Date 1 Nominal GDP = 20 (= 1x10+1x10)

Date 2 Nominal GDP = 20 (= 0.5x20+2x5)

Note intuitively that, in this example, real GDP has not changed in period 2 relative to period 1. In fact the share of the 2 goods in nominal output is 50% and the quantity produced of one good (chips) doubled while the quantity of the other was cut by half.

So, what happens when we estimate real growth of GDP using the fixed-weight and CPI methods ?

Fixed-weight approach:

Date 2 Real GDP (in date 1 prices) = $ 25 (= 1x20 + 1x5)

Real 'fixed weight' GDP growth: 25% = (25/20)-1

GDP deflator inflation: -20%

Nominal GDP growth = 0 = (20-20/20) = (1 - 0.2)(1 + 0.25) -1

CPI (fixed basket) approach:

CPI inflation: 25% = [( 0.5x10 + 2x10) / 20] -1

Period 1 Real GDP using date 2 as the base year: 25

Period 2 Real GDP using date 2 as the base year: 20

Real GDP growth using date 2 as the base year: -20%

Nominal GDP growth = 0 = (20-20/20) = (1 + 0.25)(1 - 0.20) -1

The problem is that in fixed-weight approach, too much weight is given to production of the good (chip) whose price has fallen over time. If we use a fixed-weight method, the output level and growth rate is biased upward (we get an estimate of 25% real growth) because we are overestimating the value of the output of the good whose price has fallen.

It is like computing the real output of a PC computers in 1997 by taking the 1987 price of an equivalent machine (approximately $6,000) as the base for valuing the real value added of a PC that is priced only at $2,000 today. It does not make sense to value the quantity of computers produced today at prices that were prevailing 10 years ago. So, the fixed-weight method led to an overestimation of the value added of the computer industry.

When the U.S. relied on the fixed-weight method, it was giving too much weight in the calculation of real GDP to the good whose relative price had fallen over time (in this example chips and in reality computers, semiconductors and other high tech sectors of the economy). Because of this bias, the value of the real output of chips was overestimated and led to an overestimation of the growth rate of the economy. This issue became serious over the 1980's as the price of computers was falling in absolute and relative terms while the fixed-weight, by using the high prices of computers prevailing in the base year, was leading to an overestimate of the real GDP created by computers. In order to eliminate such a bias, the Department of Commerce switched at the end of 1995 to a chain-weight method of measuring real GDP. The chain weight method is a combination of the fixed-weight method and the fixed-basket method. Real GDP is estimated twice, first using the previous year prices as the base (fixed-weight) and the second time using the current year prices as the base and the previous year quantities to compute real GDP in the previous year. Then, a (geometric) average of the two is taken. Using this method:

Growth rate of chained GDP = [(1 + 0.25)(1- 0.2)-1]/2 = 0

i.e. the growth rate of chained GDP is equal to zero that is the sensible economic answer since real output in the example above had not changed in a substantial sense.

There are however several potential problems also with the chain-weight method:

1. Quality changes are not correctly measured (examples: computers, light) leading to under-estimate of the product of industries where such quality changes occur.

2. Major productivity growth in the service industries (ATM's, telecommunications, quality of health care) not measured by standard GDP measures.

First, an important issue in computing price indexes is how they deal with quality change and new goods. One of the facts of life in growing economies is that the goods change: candy bars change size, PCs have ever-increasing capabilities, and some goods simply didn't exist in the base period. Candy bars are the easiest: we simply regard a five oz bar as half a ten oz bar. But what about PCs? If a 286 sells for $2000 and a 386 for $4000, has there been inflation or is the 386 machine twice as good as the 286? It's even more difficult if the commodity has no counterpart in the base period. How do we include VCRs in the calculation when they didn't exist, for all practical purposes, prior to the 1980s? For this reason, some people think that price indexes and real GDP do not adequately reflect quality improvements---that real GDP is growing faster than we think because quality is constantly improving. That's especially true now of new high-tech capital goods.

Second, related issues show up in services. Many authors (including the Fed Chairman Alan Greenspan) have argued that major productivity growth in the service industries are not measured by standard GDP measures. Moreover, there are other subtle measurement issues: if the price of one hour of a lawyer's time goes up, does this represent an improvement in quality or just a rise in the price?

Critics of the switch from fixed-weights to chain-weights have argued that, while the fixed-weight method overestimated the contribution of computers to real GDP, the chain-weight method fixes one problem but does nothing to address the two issues above; that, on net, leads to an underestimation of real GDP. So the new measure might overall tend to underestimate GDP and its growth rate.

At the same time, a number of authors have argued that the use of the CPI inflation rate also tends to overestimate the true level of inflation rate in the US economy because of a number of biases. In December 1996, the Boskin Commission appointed by the Senate Finance Committee, reached the conclusion that the CPI overstates the true inflation rate by 1% to 2% per year. Note that, if inflation is overestimated, then our measure of real GDP growth is underestimated as well as more of the growth of nominal GDP is imputed to an increase in prices than to an increase in quantities produced. A wide debate on the CPI  has followed the publication of the Boskin Commission recommendations. Fed Chairman Alan Greenspan has expressed his views on this debate in a testimony in Congress  in January 1997 and a recent speech in November 1997.

For more discussion on these issues see the home page on the debate on whether output and CPI inflation are mismeasured.

Summary

1. GDP measures the total value of production at market prices, the sum of value-added by every production unit in the economy.
2. Identities.
• Output (GDP) = Income (payments to labor and capital, gross of depreciation).
• GDP = C + I + G + NX .
• GNP = GDP + i NFA = C + I + G + (NX + i NFA) .
• GNP = C + I + G + CA .
• S = I + CA .
3. Price indexes measure the purchasing power of money. Issues: relative price vs general price level movements, quality change and new goods.
4. Bottom line: economic indicators clearly contain useful information, but like accounting statements they must be interpreted with care. [Data are like sausages: if you like them, you shouldn't think too much about what goes into them.]

Further Readings

On paper, there are two good (by which I mean informative and readable) books on the uses, sources, and meaning of economic data: Norman Frumkin's Guide to Economic Indicators (Armonk: Sharpe, 1994, 2nd edition) or Tracking America's Economy (Armonk: Sharpe, 1992). Not needed for this course, but if you ever have to look something up it's a good place to start. A slightly more technical introduction to macroeconomic data is available from the Richmond Fed: Macroeconomic Data: A User's Guide, edited by Roy Webb. Both of these cover the US, but in many cases the methods are similar to those used in other countries (especially for national income and product accounts, for which there is a United Nations standard).

Further Web Links and Readings

The course home page on the controversy about whether output and CPI inflation are mismeasured is a useful source of materials on the chain-weight measure of GDP, on the results of the Boskin Commission and the debate on these results. The debates on the chain-weight system of measuring GDP and the biases in measuring the inflation rate are also related to the question of whether we are correctly measuring productivity growth and whether there has been a resurgence of productivity growth in the 1990s after the dismal productivity experience in the 1973-1990 period (see also Chapter 4). On this debate and the related issue of the productivity slowdown see the homepages on the controversies Productivity Growth, Its Slowdown in the 1973-90 period and its resurgence in the 1990s: Truth or a Statistical Fluke? and the New Economy.

Entries are percentages, averages of quarterly data over the period 1970:1 to 1989:4. Data are from the OECD's Quarterly National Accounts, seasonally adjusted, except US, from Citibase. Variables are: Y = GNP or GDP; S = Y-C-G, where C is consumption and G is government purchases of goods and services; I = gross fixed capital formation. All variables are measured in current prices. Numbers may not sum to zero because of rounding, and because my measure of investment does not include the change in business inventories.

      Country               S/Y     I/Y        CA/Y    Y Growth

      Australia             24.1     24.6       1.1     3.33
      Austria               26.6     25.2       0.1     2.95
      Canada                23.7     22.1       1.2     2.82
      France                23.3     22.2       0.2     2.83
      Germany               25.1     21.4       3.1     2.51
      Italy                 22.8     22.7      -0.1     3.06
      Japan                 33.6     31.2       1.5     4.49
      United Kingdom        18.2     18.2       0.0     2.38
      United States         16.0     15.5       0.1     2.77

Figure 1. US Real GDP

FIGURE 2. Per Capita GDP: International Comparisons

.

.

Figure 3. Per Capita GDP: International Comparison 2

FIGURE 4

.

.

.

.

.

.

.

.

.

.

.

.

FIGURE 4'

Figure 5. GNP Growth in Germany and Japan

.

Figure 6

Nominal and Real GDP

Figure 7

Nominal and Real Growth Rate of GDP

.

.

.

.

.

.

Figure 8. CPI Level

and its Percentage Annual Rate of Change

Figure 9. GNP Deflator Index

and its Percentage Annual Rate of Change