PLEASE NOTE: These basic lecture notes on Hall and Taylor (HT) Chapter 3 are required for Econ 302. If possible, students should read these notes prior to attending class lectures on HT chapter 3. A more advanced detailed version of these notes is also available on this web site; the advanced notes are recommended but not required.
The consumer price index (CPI) measures the annual cost of living of a typical U.S. urban family. More precisely, the CPI is the current market value of a basket of goods and services deemed to be representative of the type and amount of goods and services purchased by a typical U.S. urban family.
The relative amounts of goods and services included in the basket are determined by a survey of urban consumer buying habits conducted approximately once every ten years. The levels of goods and services included in the basket are normalized so that the value of the basket is $100 when valued in prices for some conventionally determined base year. The BLS updates the CPI each month by using current prices to re-compute the market value of the CPI basket of goods.
For example, suppose in 1995 that a typical U.S. urban family ate 150 pounds of chicken and 50 pounds of beef, and that the price of chicken in 1995 was $2 per lb and the price of beef in 1995 was $3 per lb. Suppose 1987 is the base year, and the price of chicken in 1987 was $1.00 per lb and the price of beef in 1987 was $2.00 per lb.
The CPI basket for 1995 is then determined by the following two conditions: It should contain 3 times more chicken than beef (because 150/50 = 3), implying that c = 3b; and the total value of the CPI basket measured in 1987 prices must be $100, implying that $1xc + $2xb = $100. The amounts c and b of chicken and beef to be included in the CPI basket for 1995 can then be found by solving these two equations for the two unknowns c and b, which gives c=60 lbs and b=20 lbs. The CPI for 1995 is then given by the 1995 market value of a basket containing 60 lbs of chicken and 20 lbs of beef, which is $2x60 + $3x20 = $180. The fact that the CPI for 1995 is greater than the base year value of $100 is interpreted as an indication that the cost of living for a typical U.S. urban family has increased since the base year.
In the U.S., the CPI is the most widely used measure of the purchasing power of the dollar. For example, wage contracts which contain cost-of-living-adjustment (COLA) provisions to protect workers against price increases require wages to increase in proportion to increases in the CPI. Social security benefits are also now indexed to the CPI or to a wage index, whichever is lower. The CPI is strongly affected by energy and food costs, which can be extremely volatile. Omitting these more volatile elements from the CPI results in a more stable measure of inflation, referred to as the core rate of inflation, and one commonly sees this core rate of inflation reported in addition to or even instead of the CPI. See, for example, the ERP (Chapter 2, pages 53 and 68-69).
An important difficulty with the CPI is that the basket of goods used to represent typical purchases of an urban family is not updated very often and can become outdated. In particular, the relative amounts of goods in the basket are not varied in response to price changes even though consumer purchases may be highly elastic with respect to price changes.
For example, after the 1974 oil price shock, U.S. consumers greatly reduced their use of energy in response to increased energy prices. However, the CPI basket in use from 1973 through 1986 was based on a 1973 survey in which energy usage was high, hence it did not reflect this relative drop in energy use. Consequently, the CPI surged upward in these years in response to increased energy prices, greatly exaggerating the actual impact of inflation on a typical U.S. urban family.
Recall from HT2 that the GNP implicit price deflator P(T) for any period T is defined to be nominal GDP for period T divided by real GDP for period T. Often the rate of inflation INF(T,T+1) from T to T+1 is measured as the percentage rate of change in P(T) from T to T+1, as follows:
P(T+1) - P(T)
INF(T,T+1) = --------------- .
P(T)
For concreteness, we will use INF(T,T+1) as our measure of inflation
from T to T+1 throughout all remaining lectures.
By construction, P(T) reflects prices for all newly produced final goods and services in the economy in period T. Consequently, INF(T,T+1) is a fairly broad measure of inflation as experienced by all purchasers of final goods and services.
Alternatively, economists sometimes measure inflation as the rate of change in the CPI rather than the rate of change in the GNP implicit price deflator. See, for example, the ERP (Chapter 2). The CPI measure of inflation more narrowly focuses on inflation as experienced by households.
-- SEE HT FIGURE 3-1:CONSUMER PRICE INFLATION (USING CPI)--
In order to discuss carefully the appropriate policy responses to unemployment, it is important to understand how unemployment is actually measured in practice. Information on employment in the U.S. comes from two surveys.
The Current Population Survey undertaken in a particular month M counts a person as unemployed for month M if that person did not work either full-time or part-time during the survey week for month M but did look for work during the previous four weeks. Given this definition, one can then define practical measures for the labor force, the unemployment rate, and the labor force participation rate as follows:
Labor Force for month M = Number of legal-aged persons who were
(U.S. definition) either working or unemployed during month
M according to the Current Population Survey.
Number of Unemployed Workers During Month M
Unemployment Rate = ---------------------------------------------
for month M Labor Force During Month M
Labor Force During Month M
Labor Force Participation = ------------------------------------
Rate for Month M Working Age Population During Month M
As sensible and straightforward as these measures may seem, they can result in counter-intuitive observations. For example, at the start of an economic expansion, the unemployment rate often rises. How can this be?
By definition, any person who currently is not working and who has stopped seeking work for whatever reason is not counted as part of the labor force. At the start of an economic expansion, many people without jobs who previously had been too discouraged to seek work now begin to seek work and hence are counted as new additions to the labor force. Thus, although the number of employed workers is increasing in absolute terms, the labor force may be increasing at an even faster pace, which results in an increase in the rate of unemployment.
To define potential GDP, we first need to clarify what is meant by full employment. We will start with a purely verbal (and somewhat vague) characterization for full employment. Later this will be interpreted in more formal quantitative terms as the level of employment that obtains when the labor market is in equilibrium in the sense that labor demand equals labor supply.
Given any time period T, full employment for period T -- denoted by N*(T) -- is defined to be the amount of employment that would be realized, given existing incentives such as fringe benefits and unemployment insurance, if wages were fully flexible and responsive to demand and supply pressures.
Note, in particular, that N*(T) need not be the maximum amount of employment that the economy is capable of supplying during period T, because some people may choose to continue seeking work (to remain unemployed) rather than to accept the jobs that are currently available to them. The latter type of unemployment is sometimes referred to as voluntary frictional unemployment.
The unemployment rate U*(T) that prevails when employment is equal to full employment N*(T), and which therefore corresponds to voluntary frictional unemployment, is often called the natural rate of unemployment. Currently in the U.S. the natural rate of unemployment is calculated to be about 6 percent. For reasons that will be clarified in later lectures, the natural rate of unemployment is sometimes also referred to as the "sustainable rate of unemployment" or the "non-accelerating inflation rate of unemployment" (NAIRU); see the ERP (Chapter 2, Box 2-4).
The level of real GDP that would result in period T if employment were equal to full employment N*(T), denoted by Y*(T), is called the potential GDP for period T. The percentage departure of actual real GDP Y(T) from potential GDP Y*(T) is called the GDP gap for period T. That is,
GDP GAP FOR PERIOD T:
Y(T) - Y*(T)
-------------- ,
Y*(T)
where
Y(T) = real GDP for period T ;
Y*(T) = potential GDP for period T .
Empirically, employment is procyclical in the sense that employment moves in step with the business cycle. More precisely, employment is high when real GDP is high relative to potential GDP and low when real GDP is low relative to potential GDP.
-- SEE HT FIGURE 3-2: GROWTH AND FLUCTUATIONS OF HOURS WORKED --
Given that employment is procyclical, it follows that unemployment is countercyclical -- that is, unemployment is high when GDP is low relative to potential GDP (i.e., during times of economic recession) and low when GDP is high relative to potential GDP (i.e., during times of economic expansion).
-- SEE HT FIGURE 3-4: THE UNEMPLOYMENT RATE --
The empirically observed countercyclical nature of the unemployment rate is captured in Okun's Law, expressed as follows: For each percentage point by which the unemployment rate is above the natural unemployment rate, real GDP is 3 percent below potential GDP.
ALGEBRAIC FORM OF OKUN'S LAW:
Y(T) - Y*(T)
---------------- = - 3 x [U(T) - U*(T)] ,
Y*(T)
| |
| |
__ GDP gap __
in period T
where
Y(T) = real GDP for period T ;
Y*(T) = potential GDP for period T ;
U(T) = rate of unemployment for period T ;
U*(T) = natural rate of unemployment for period T .
-- SEE HT FIGURE 3-5: OKUN'S LAW --
E. THE MEASUREMENT OF WAGES AND LABOR PRODUCTIVITY
The nominal wage rate for the U.S. is measured as total labor earnings (dollar value of wages plus fringe benefits) divided by total hours worked. The real wage for the U.S. is then measured as the nominal wage divided by some measure of the general price level, generally either the CPI or the GNP implicit price deflator P. In remaining lectures, the following notation will be used:
W(T) = nominal wage during time T
Total labor earnings during T
= ------------------------------------
Total number of hours worked during T
= Average hourly wage during T.
w(T) = real wage during time T
W(T) W(T)
= ------- OR ------ .
P(T) CPI(T)
As noted in the ERP (Chapter 2, p. 60), the stagnation in real wages in the U.S. since the early 1970's is a major current policy concern.
-- SEE HT FIGURE 3-6: GROWTH IN REAL WAGE (MEASURED AS W/CPI) --
Productivity is the amount of output produced per unit of input. The most popular measure of productivity, put out by the Bureau of Labor Statistics (BLS), is as follows:
Labor Productivity = Average output per hour of labor during T
During Period T
= Real GDP originating in the business sector
during T divided by the total hours of work
in the business sector during T
Labor productivity growth in the U.S. has been disappointing since 1973, in the sense that it has been below the normal trendline growth rate established during the period 1946-1973. As noted in the ERP (Chapter 2, pages 57-59), this slowdown in labor productivity is an ongoing puzzle and a major policy concern.
-- SEE HT FIGURE 3-7: SLOWDOWN IN LABOR PRODUCTIVITY GROWTH --