Real GDP, Technological Progress, and

the Price of Computers

FOCUS
A tough problem in computing real GDP is how to deal with new computers this year have a speed of 3 GHz compared
changes in quality of existing goods. One of the most dif- to a speed of 2 GHz for new computers last year. And sup-
ficult cases is computers. It would clearly be absurd to as- pose the dollar price of new computers this year is the
sume that a personal computer in 2010 is the same good as same as the dollar price of new computers last year. Then
a personal computer produced in 1981 (the year in which economists in charge of computing the adjusted price of
the IBM PC was introduced): The same amount of money computers will conclude that new computers are in fact
can clearly buy much more computing in 2010 than it could 10% cheaper than last year.
in 1981. But how much more? Does a 2010 computer pro- This approach, which treats goods as providing a col-
vide 10 times, 100 times, or 1,000 times the computing serv- lection of characteristics—for computers, speed, mem-
ices of a 1981 computer? How should we take into account ory, and so on—each with an implicit price, is called
the improvements in internal speed, the size of the random hedonic pricing (“hedone” means “pleasure” in Greek).
access memory (RAM) or of the hard disk, the fact that com- It is used by the Department of Commerce—which con-
puters can access the Internet, and so on? structs real GDP—to estimate changes in the price of
The approach used by economists to adjust for these complex and fast changing goods, such as automobiles
improvements is to look at the market for computers and and computers. Using this approach, the Department of
how it values computers with different characteristics in Commerce estimates that, for a given price, the quality of
a given year. Example: Suppose the evidence from prices new computers has increased on average by 18% a year
of different models on the market shows that people are since 1981. Put another way, a typical personal computer
willing to pay 10% more for a computer with a speed of 3 in 2010 delivers 1.1829 ⴝ 121 times the computing serv-
GHz (3,000 megahertz) rather than 2 GHz. (The first edi- ices a typical personal computer delivered in 1981.
tion of this book, published in 1996, compared two com- Not only do computers deliver more services, they have
puters, with speeds of 50 and 16 megaherz, respectively. become cheaper as well: Their dollar price has declined by
This change is a good indication of technological progress. about 10% a year since 1981. Putting this together with the
A further indication of technological progress is that, for information in the previous paragraph, this implies that
the past few years, progress has not been made by increas- their quality–adjusted price has fallen at an average rate
ing the speed of processors, but rather by using multicore of 18% ⴙ 10% ⴝ 28% per year. Put another way, a dol-
processors. We shall leave this aspect aside here, but peo- lar spent on a computer today buys 1.2829 ⴝ 1,285 times
ple in charge of national income accounts cannot; they more computing services than a dollar spent on a computer
have to take this change into account as well.) Suppose in 1981.

is no official definition of what constitutes a recession, but the convention is to refer to

a “recession” if the economy goes through at least two consecutive quarters of negative
growth. Although GDP growth was positive for 2001 as a whole, it was negative during
each of the first three quarters of 2001; thus 2001 qualifies as a (mild) recession.

2-2 The Unemployment Rate

Because it is a measure of aggregate activity, GDP is obviously the most important
macroeconomic variable. But two other variables, unemployment and inflation, tell us
about other important aspects of how an economy is performing. This section focuses
on the unemployment rate.
We start with two definitions: Employment is the number of people who have a
job. Unemployment is the number of people who do not have a job but are looking for
one. The labor force is the sum of employment and unemployment:
L = N + U
labor force = employment + unemployment

Chapter 2 A Tour of the Book 25

 Universal Uclick. Reprinted with permission. All rights
Non Sequitur © 2006 Wiley Ink, Inc. Distributed by

reserved.
The unemployment rate is the ratio of the number of people who are unemployed
to the number of people in the labor force:
U
u =
L
unemployment rate = unemployment>labor force
Constructing the unemployment rate is less obvious than you might have thought.
The cartoon above not withstanding, determining whether somebody is employed
is straightforward. Determining whether somebody is unemployed is harder. Recall
from the definition that, to be classified as unemployed, a person must meet two
conditions: that he or she does not have a job, and he or she is looking for one; this
second condition is harder to assess.
Until the 1940s in the United States, and until more recently in most other
countries, the only available source of data on unemployment was the number of
people registered at unemployment offices, and so only those workers who were
registered in unemployment offices were counted as unemployed. This system
led to a poor measure of unemployment. How many of those looking for jobs ac-
tually registered at the unemployment office varied both across countries and
across time. Those who had no incentive to register—for example, those who had
exhausted their unemployment benefits—were unlikely to take the time to come to
the unemployment office, so they were not counted. Countries with less generous
benefit systems were likely to have fewer unemployed registering, and therefore
smaller measured unemployment rates.
Today, most rich countries rely on large surveys of households to compute the
unemployment rate. In the United States, this survey is called the Current Population
Survey (CPS). It relies on interviews of 50,000 households every month. The survey
classifies a person as employed if he or she has a job at the time of the interview; it clas-
sifies a person as unemployed if he or she does not have a job and has been looking for
a job in the last four weeks. Most other countries use a similar definition of unemploy-
ment. In the United States, estimates based on the CPS show that, during 2010, an aver-
age of 139.0 million people were employed, and 14.8 million people were unemployed,
so the unemployment rate was 14.8>1139.0 + 14.8 2 = 9.6%.
Note that only those looking for a job are counted as unemployed; those who do
not have a job and are not looking for one are counted as not in the labor force. When
unemployment is high, some of the unemployed give up looking for a job and therefore
are no longer counted as unemployed. These people are known as discouraged workers.
Take an extreme example: If all workers without a job gave up looking for one, the

26 Introduction The Core

 10 Figure 2-3
U.S. unemployment rate,
1960–2010
9
Since 1960, the U.S. unemploy-
ment rate has fluctuated between
3 and 10%, going down during
8
expansions, and going up during
recessions. The effect of the crisis
is highly visible, with the unem-
7 ployment rate reaching close to
Percent

10%, the highest such rate since

the 1980s.

6 Source: Series UNRATE: Federal

Reserve Economic Data (FRED) http://
research.stlouisfed.org/fred2/

3
1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010

unemployment rate would equal zero. This would make the unemployment rate a very
poor indicator of what is happening in the labor market. This example is too extreme; in At the start of economic re-
practice, when the economy slows down, we typically observe both an increase in un- form in Eastern Europe in the
early 1990s, unemployment
employment and an increase in the number of people who drop out of the labor force. increased dramatically. But
Equivalently, a higher unemployment rate is typically associated with a lower participa- equally dramatic was the fall
tion rate, defined as the ratio of the labor force to the total population of working age.  in the participation rate. In
Figure 2-3 shows the evolution of unemployment in the United States since 1970. Poland in 1990, 70% of the
Since 1960, the U.S. unemployment rate has fluctuated between 3 and 10%, going up decrease in employment was
reflected in early retirements—
during recessions and down during expansions. Again, you can see the effect of the by people dropping out of the
crisis, with the unemployment rate reaching a peak at nearly 10% in 2010, the highest labor force rather than becom-
such rate since the 1980s. ing unemployed.

Why Do Economists Care about Unemployment?

Economists care about unemployment for two reasons. First, they care about
unemployment because of its direct effect on the welfare of the unemployed. Although
unemployment benefits are more generous today than they were during the Great
Depression, unemployment is still often associated with financial and psychologi-
cal suffering. How much suffering depends on the nature of the unemployment. One
image of unemployment is that of a stagnant pool, of people remaining unemployed
for long periods of time. In normal times, in the United States, this image is not right:
Every month, many people become unemployed, and many of the unemployed find
jobs. When unemployment increases, however, as is the case now, the image becomes
more accurate. Not only are more people unemployed, but also many of them are un-
employed for a long time. For example, the mean duration of unemployment, which
was 9 weeks on average during 2000–2007, increased to 33 weeks in 2010. In short,
when the unemployment increases, not only does unemployment become both more
widespread, but it also becomes more painful.

Chapter 2 A Tour of the Book 27

 Did Spain Have a 24% Unemployment
Rate in 1994?
FOCUS
In 1994, the official unemployment rate in Spain reached declaring it to the social security administration—accounted
24%. (It then decreased steadily, reaching a low of 8% in for between 10 and 15% of employment. But it was composed
2007, only to increase dramatically again since the begin- mostly of people who already had a job and were taking a sec-
ning of the crisis. It now exceeds 20% and is still increasing. ond or even a third job. The best estimate from the survey was
Thus, many of the issues in this Focus box are becoming that only about 15% of the unemployed were in fact working.
relevant again.) This was roughly the same unemployment This implied that the unemployment rate, which was officially
rate as in the United States in 1933, the worst year of the 21% at the time, was in fact closer to 18%, still a very high
Great Depression. Yet Spain in 1994 looked nothing like the number. In short, the Spanish underground economy was sig-
United States in 1933: There were few homeless, and most nificant, but it just was not the case that most of the Spanish
cities looked prosperous. Can we really believe that nearly unemployed work in the underground economy.
one–fifth of the Spanish labor force was looking for work? How did the unemployed survive? Did they survive be-
To answer this question, we must first examine how the cause unemployment benefits were unusually generous in
Spanish unemployment number is put together. Like the Spain? No. Except for very generous unemployment ben-
CPS in the United States, unemployment is measured using efits in two regions, Andalusia and Extremadura—which,
a large survey of 60,000 households. People are classified as not surprisingly, had even higher unemployment than the
unemployed if they indicate that they are not working but rest of the country—unemployment benefits were roughly
are seeking work. in line with unemployment benefits in other OECD coun-
Can we be sure that people tell the truth? No. Although tries. Benefits were typically 70% of the wage for the first six
there is no obvious incentive to lie—answers to the survey months, and 60% thereafter. They were given for a period of
are confidential and are not used to determine whether 4 to 24 months, depending on how long people had worked
people are eligible for unemployment benefits—those before becoming unemployed. The 30% of the unemployed
who are working in the underground economy may prefer who had been unemployed for more than two years did not
to play it safe and report that they are unemployed instead. receive unemployment benefits.
The size of the underground economy—the part of eco- So how did they survive? A key to the answer lies with
nomic activity that is not measured in official statistics, either the Spanish family structure. The unemployment rate was
because the activity is illegal or because firms and workers highest among the young: In 1994, it was close to 50% for
would rather not report it and thus not pay taxes—is an old those between 16 and 19, and around 40% for those be-
issue in Spain. And because of that, we actually know more tween 20 and 24. The young typically stay at home until
about the underground economy in Spain than in many other their late 20s, and have increasingly done so as unemploy-
countries: In 1985, the Spanish government tried to find out ment increased. Looking at households rather than at
more and organized a detailed survey of 60,000 individuals. individuals, the proportion of households where nobody
To try to elicit the truth from those interviewed, the question- was employed was less than 10% in 1994; the proportion
naire asked interviewees for an extremely precise account of of households that received neither wage income nor un-
the use of their time, making it more difficult to misreport. employment benefits was around 3%. In short, the family
The answers were interesting. The underground economy structure, and transfers from the rest of the family, were the
in Spain—defined as the number of people working without factors that allowed many of the unemployed to survive.

Second, economists also care about the unemployment rate because it provides
a signal that the economy may not be using some of its resources efficiently. Many
workers who want to work do not find jobs; the economy is not utilizing its human
It is probably because of resources efficiently. From this viewpoint, can very low unemployment also be a prob-

statements like this that eco- lem? The answer is yes. Like an engine running at too high a speed, an economy in
nomics is known as the “dis- which unemployment is very low may be overutilizing its resources and run into labor
mal science.”
shortages. How low is “too low”? This is a difficult question, a question we will take up
at more length later in the book. The question came up in 2000 in the United States. At
the end of 2000, some economists worried that the unemployment rate, 4% at the time,
was indeed too low. So, while they did not advocate triggering a recession, they favored
lower (but positive) output growth for some time, so as to allow the unemployment
rate to increase to a somewhat higher level. It turned out that they got more than they
had asked for: a recession rather than a slowdown.

28 Introduction The Core

2-3 The Inflation Rate
Deflation is rare, but it hap-
Inflation is a sustained rise in the general level of prices—the price level. The infla-pens. Japan has had defla-
tion rate is the rate at which the price level increases. (Symmetrically, deflation is tion, off and on, since the
a sustained decline in the price level. It corresponds to a negative inflation rate).  late 1990s. The United States
The practical issue is how to define the price level so the inflation rate can be e x p e r i e n c e d d e f l a t i o n i n
measured. Macroeconomists typically look at two measures of the price level, at two the 1930s during the Great
Depression.
price indexes: the GDP deflator and the Consumer Price Index.

The GDP Deflator

We saw earlier how increases in nominal GDP can come either from an increase in real
GDP, or from an increase in prices. Put another way, if we see nominal GDP increase
faster than real GDP, the difference must come from an increase in prices.
This remark motivates the definition of the GDP deflator. The GDP deflator in year
t, Pt , is defined as the ratio of nominal GDP to real GDP in year t: Index numbers are often set
Nominal GDPt $Yt equal to 100 (in the base year)
Pt = = rather than to 1. If you look at
Real GDPt Yt the Economic Report of the
President (see Chapter 1) you
Note that, in the year in which, by construction, real GDP is equal to nominal GDP will see that the GDP deflator,
(2005 at this point in the United States), this definition implies that the price level is reported in Table B-3, is equal
equal to 1. This is worth emphasizing: The GDP deflator is called an index number. Its  to 100 for 2005 (the base year),
level is chosen arbitrarily—here it is equal to 1 in 2005—and has no economic interpre- 103.2 in 2006, and so on.
tation. But its rate of change, 1 Pt - Pt - 1 2 >Pt - 1 (which we shall denote by pt in the rest
of the book), has a clear economic interpretation: It gives the rate at which the general Compute the GDP deflator and
level of prices increases over time—the rate of inflation.  the associated rate of inflation
One advantage to defining the price level as the GDP deflator is that it implies a from 2004 to 2005 and from
simple relation among nominal GDP, real GDP, and the GDP deflator. To see this, reor- 2005 to 2006 in our car exam-
ple in Section 2-1, when real
ganize the previous equation to get: GDP is constructed using the
2005 price of cars as the com-
$Yt = Pt Yt
mon price.
Nominal GDP is equal to the GDP deflator times real GDP. Or, putting it in terms of
rates of change: The rate of growth of nominal GDP is equal to the rate of inflation plus  For a refresher, see Appendix 2,
the rate of growth of real GDP. Proposition 7.

The Consumer Price Index

The GDP deflator gives the average price of output—the final goods produced in the
economy. But consumers care about the average price of consumption—the goods
they consume. The two prices need not be the same: The set of goods produced in the
economy is not the same as the set of goods purchased by consumers, for two reasons:
■ Some of the goods in GDP are sold not to consumers but to firms (machine tools,
for example), to the government, or to foreigners.
■ Some of the goods bought by consumers are not produced domestically but are
imported from abroad.
To measure the average price of consumption, or, equivalently, the cost of living, Do not confuse the CPI with the
PPI, or producer price index,
macroeconomists look at another index, the Consumer Price Index, or CPI. The CPI has which is an index of prices of
been in existence in the United States since 1917 and is published monthly (in contrast, domestically produced goods
numbers for GDP and the GDP deflator are only constructed and published quarterly).  in manufacturing, mining, ag-
The CPI gives the cost in dollars of a specific list of goods and services over riculture, fishing, forestry, and
time. The list, which is based on a detailed study of consumer spending, attempts to electric utility industries.

Chapter 2 A Tour of the Book 29

Figure 2-4 14

Inflation rate, using the 12 CPI

Inflation Rate (percent per year)

CPI and the GDP deflator,
1960–2010 10

The inflation rates, com- 8

puted using either the CPI or
the GDP deflator, are largely 6
similar.
4
Source: Calculated using series
GDPDEF, CPI-AUSCL Federal Re- 2
serve Economic Data (FRED) http:// GDP
research.stlouisfed.org/fred2/ deflator
0

–2
1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010

represent the consumption basket of a typical urban consumer and is updated roughly
only once every 10 years.
Each month, Bureau of Labor Statistics (BLS) employees visit stores to find out
what has happened to the price of the goods on the list; prices are collected in 87 cit-
ies, from about 23,000 retail stores, car dealerships, gas stations, hospitals, and so on.
These prices are then used to construct the Consumer Price Index.
Like the GDP deflator (the price level associated with aggregate output, GDP), the CPI
is an index. It is set equal to 100 in the period chosen as the base period and so its level has
no particular significance. The current base period is 1982 to 1984, so the average for the
Do not ask why such a strange
 period 1982 to 1984 is equal to 100. In 2010, the CPI was 222.8; thus, it cost more than twice
base period was chosen. as much in dollars to purchase the same consumption basket than in 1982–1984.
Nobody seems to remember. You may wonder how the rate of inflation differs depending on whether the GDP
deflator or the CPI is used to measure it. The answer is given in Figure 2-4, which plots
the two inflation rates since 1960 for the United States. The figure yields two conclusions:
■ The CPI and the GDP deflator move together most of the time. In most years, the
two inflation rates differ by less than 1%.
■ But there are clear exceptions. In 1979 and 1980, the increase in the CPI was signifi-
cantly larger than the increase in the GDP deflator. The reason is not hard to find. Re-
call that the GDP deflator is the price of goods produced in the United States, whereas
the CPI is the price of goods consumed in the United States. That means when the
price of imported goods increases relative to the price of goods produced in the United
You may wonder why the effect
States, the CPI increases faster than the GDP deflator. This is precisely what happened
of the increases in the price of in 1979 and 1980. The price of oil doubled. And although the United States is a pro-

oil since 1999 is much less vis- ducer of oil, it produces much less than it consumes: It was and still is a major oil im-
ible in the figure. The answer: porter. The result was a large increase in the CPI compared to the GDP deflator.
The increases have taken place
more slowly over time, and In what follows, we shall typically assume that the two indexes move together so we
other factors have worked in do not need to distinguish between them. we shall simply talk about the price level and
the opposite direction. denote it by Pt, without indicating whether we have the CPI or the GDP deflator in mind.

Why Do Economists Care about Inflation?

If a higher inflation rate meant just a faster but proportional increase in all prices and
wages—a case called pure inflation—inflation would be only a minor inconvenience,
as relative prices would be unaffected.

30 Introduction The Core

 Take, for example, the workers’ real wage—the wage measured in terms of goods
rather than in dollars. In an economy with 10% more inflation, prices would increase
by 10% more a year. But wages would also increase by 10% more a year, so real wages
would be unaffected by inflation. Inflation would not be entirely irrelevant; people
would have to keep track of the increase in prices and wages when making decisions.
But this would be a small burden, hardly justifying making control of the inflation rate
one of the major goals of macroeconomic policy.
So why do economists care about inflation? Precisely because there is no such
thing as pure inflation:
■ During periods of inflation, not all prices and wages rise proportionately. Because they
don’t, inflation affects income distribution. For example, retirees in many countries re-
ceive payments that do not keep up with the price level, so they lose in relation to other
groups when inflation is high. This is not the case in the United States, where Social
Security benefits automatically rise with the CPI, protecting retirees from inflation. But
during the very high inflation that took place in Russia in the 1990s, retirement pen-
sions did not keep up with inflation, and many retirees were pushed to near starvation.
■ Inflation leads to other distortions. Variations in relative prices also lead to more
uncertainty, making it harder for firms to make decisions about the future, such This is known as bracket creep.
as investment decisions. Some prices, which are fixed by law or by regulation, lag In the United States, the tax
behind the others, leading to changes in relative prices. Taxation interacts with brackets are adjusted automat-
ically for inflation: If inflation is
inflation to create more distortions. If tax brackets are not adjusted for inflation, 5%, all tax brackets also go up
for example, people move into higher and higher tax brackets as their nominal in- by 5%—in other words, there is
come increases, even if their real income remains the same.  no bracket creep.

If inflation is so bad, does this imply that deflation (negative inflation) is good?  Newspapers sometimes con-
The answer is no. First, high deflation (a large negative rate of inflation) would create fuse deflation and recession.
many of the same problems as high inflation, from distortions to increased uncertainty. Sec- They may happen together but
they are not the same. Deflation
ond, as we shall see later in the book, even a low rate of deflation limits the ability of monetary is a decrease in the price level.
policy to affect output. So what is the “best” rate of inflation? Most macroeconomists believe A recession is a decrease in
that the best rate of inflation is a low and stable rate of inflation, somewhere between 1 and real output.
4%. We shall look at the pros and cons of different rates of inflation later in the book.

2-4 Output, Unemployment, and the Inflation

Rate: Okun’s Law and the Phillips Curve
We have looked separately at the three main dimensions of aggregate economic activ-
ity: output growth, the unemployment rate, and the inflation rate. Clearly they are not
independent, and much of this book will be spent looking at the relations among them
in detail. But it is useful to have a first look now.

Okun’s Law
Intuition suggests that if output growth is high, unemployment will decrease, and this is
indeed true. This relation was first examined by American economist Arthur Okun and
for this reason has become known as Okun’s law. Figure 2-5 plots the change in the un-  Arthur Okun was an adviser
employment rate on the vertical axis against the rate of growth of output on the horizon- t o P re s i d e n t K e n n e d y i n
tal axis for the United States since 1960. It also draws the line that best fits the cloud of the 1960s. Okun’s law is, of
course, not a law, but an em-
points in the figure. Looking at the figure and the line suggests two conclusions: pirical regularity.
■ The line is downward sloping and fits the cloud of points quite well. Put in eco-
nomic terms: There is a tight relation between the two variables: Higher output

Chapter 2 A Tour of the Book 31

Figure 2-5 4

Changes in the

Change in the unemployment rate

3
unemployment rate versus
output growth in the
United States, 1960–2010 2

(percentage points)
Output growth that is higher
than usual is associated with 1
a reduction in the unemploy-
ment rate; output growth that 0
is lower than usual is associ-
ated with an increase in the
–1
unemployment rate.

Source: See Figures 2-2 and 2-3. –2

–3
–4 –2 0 2 4 6 8
Output growth (percent)


Such a graph, plotting one growth leads to a decrease in unemployment. The slope of the line is -0.4. This
variable against another, is
implies that, on average, an increase in the growth rate of 1% decreases the unem-
called a scatterplot. The line
is called a regression line. For ployment rate by roughly -0.4%. This is why unemployment goes up in recessions
more on regressions, see Ap- and down in expansions. This relation has a simple but important implication: The
pendix 3. key to decreasing unemployment is a high enough rate of growth.
■ This vertical line crosses the horizontal axis at the point where output growth is
roughly equal to 3%. In economic terms: It takes a growth rate of about 3% to keep
unemployment constant. This is for two reasons. The first is that population, and
thus the labor force, increases over time, so employment must grow over time just to
keep the unemployment rate constant. The second is that output per worker is also
increasing with time, which implies that output growth is higher than employment
growth. Suppose, for example, that the labor force grows at 1% and that output per
worker grows at 2%. Then output growth must be equal to 3% 11% + 2%2 just to
keep the unemployment rate constant.

The Phillips Curve

Okun’s law implies that, with strong enough growth, one can decrease the unemploy-
ment rate to very low levels. But intuition suggests that, when unemployment becomes
very low, the economy is likely to overheat, and that this will lead to upward pressure
on inflation. And, to a large extent, this is true. This relation was first explored in 1958
by a New Zealand economist, A. W. Phillips, and has become known as the Phillips
It should probably be known
 curve. Phillips plotted the rate of inflation against the unemployment rate. Since then,
as the Phillips relation, but it is the Phillips curve has been redefined as a relation between the change in the rate of
too late to change that. inflation and the unemployment rate. Figure 2-6 plots the change in the inflation rate
(measured using the CPI) on the vertical axis against the unemployment rate on the
horizontal axis, together with the line that fits the cloud of points best, for the United
States since 1960. Looking at the figure again suggests two conclusions:
■ The line is downward sloping, although the fit is not as tight as it was for Okun’s
law: Higher unemployment leads, on average, to a decrease in inflation; lower
unemployment leads to an increase in inflation. But this is only true on average.
Sometimes, high unemployment is associated with an increase in inflation.

32 Introduction The Core

 6 Figure 2-6

Change in inflation rate (percentage points)

Changes in the
4 inflation rate versus the
unemployment rate in the
United States, 1960–2010
2
A low unemployment rate
leads to an increase in the in-
0 flation rate, a high unemploy-
ment rate to a decrease in the
inflation rate.
–2
Source: See Figures 2-3 and 2-4.
–4

–6
3 4 5 6 7 8 9 10
Unemployment (percent)

■ The line crosses the horizontal axis at the point where the unemployment rate is
roughly equal to 6%. In economic terms: When unemployment has been below 6%,
inflation has typically increased, suggesting that the economy was overheating, oper-
ating above its potential. When unemployment has been above 6%, inflation has typi-
cally decreased, suggesting that the economy was operating below potential. But, again
here, the relation is not tight enough that the unemployment rate at which the econ-
omy overheats can be pinned down very precisely. This explains why some economists
believe that we should try to maintain a lower unemployment rate, say 4 or 5%, and oth-
ers believe that it may be dangerous, leading to overheating and increasing inflation.
Clearly, a successful economy is an economy that combines high output growth, low
unemployment, and low inflation. Can all these objectives be achieved simultane-
ously? Is low unemployment compatible with low and stable inflation? Do policy mak-
ers have the tools to sustain growth, to achieve low unemployment while maintaining
low inflation? These are the questions we shall take up as we go through the book. The
next two sections give you the road map.

2-5 The Short Run, the Medium Run,

the Long Run
What determines the level of aggregate output in an economy?
■ Reading newspapers suggests a first answer: Movements in output come from move-
ments in the demand for goods. You probably have read news stories that begin like
this: “Production and sales of automobiles were higher last month due to a surge in
consumer confidence, which drove consumers to showrooms in record numbers.” Sto-
ries like these highlight the role demand plays in determining aggregate output; they
point to factors that affect demand, ranging from consumer confidence to interest rates.
■ But, surely, no amount of Indian consumers rushing to Indian showrooms can
increase India’s output to the level of output in the United States. This suggests
a second answer: What matters when it comes to aggregate output is the sup-
ply side—how much the economy can produce. How much can be produced de-
pends on how advanced the technology of the country is, how much capital it is
using, and the size and the skills of its labor force. These factors—not consumer
confidence—are the fundamental determinants of a country’s level of output.

Chapter 2 A Tour of the Book 33