58
©2014 Pearson Education, Inc.
Chapter 4
Demand
Chapter Outline
4.1 Deriving Demand Functions
System of Demand Functions
Graphical Interpretation
Application: Quitting Smoking
4.2 Effects of an Increase in Income
How Income Changes Shift Demand Curves
Solved Problem 4.1
Consumer Theory and Income Elasticities
Income Elasticity
Income-Consumption Curves and Income Elasticities
Application: Fast-Food Engel Curves
Some Goods Must Be Normal
Weighted Income Elasticities
Solved Problem 4.2
4.3 Effects of a Price Increase
Income and Substitution Effects with a Normal Good
Solved Problem 4.3
Application: Shipping the Good Stuff Away
Income and Substitution Effects with an Inferior Good
Solved Problem 4.4
Compensated Demand Curve
Solved Problem 4.5
Slutsky Equation
4.4 Cost-of-Living Adjustment
Inflation Indexes
Effects of Inflation Adjustments
CPI Adjustment
True Cost-of-Living Adjustment
Size of the CPI Substitution Bias
Application: Fixing the CPI Substitution Bias
Chapter 4 Demand 59
©2014 Pearson Education, Inc.
4.5 Revealed Preference
Recovering Preferences
Substitution Effect
Teaching Tips
This chapter contains a great deal of important material, and requires several classes to cover effectively.
The material is not intuitively difficult, but students need to be clear about these concepts in order to have
the substitution and income effect material make sense. You may want to spread the presentation of the
substitution and income effects over more than one period because students will benefit from having some
time to process the first run through, as well as refer back to the book. This tends to be the most
conceptually difficult part of the consumer theory section of the course; if students master substitution and
income effects, they are likely to do well in the rest of the course.
When presenting the substitution and income effects, try to set up the presentation such that the class can
take good quality notes on the graphs. Unfortunately, it is often difficult for students to master these graphs
without practice, and thus their notes may be poor. To minimize this problem, in addition to reminding
them to bring in a protractor and colored pencils with which to take notes, do the following: The first time
demonstrate the separation of the total effect into the two component effects, replicate an example that is in
the text (such as the live music and music tracks example in Figure 4.5). Tell the class that you are doing
this, but ask them to take notes as they normally would rather than just watch me and look at the book. This
way, if they make errors in note taking, they can refer to the text to see the correct graph. The other thing
that can be helpful is to supply them with pre-drawn indifference curves, as this is where most of the
trouble occurs. For example, if they draw their curves much differently than yours when you are
demonstrating the separation of income and substitution effects for an inferior good, theirs may not turn
out to be inferior. Another possibility is to give the class coordinates of points that make up the needed
indifference curves. Working through mathematical solutions also helps reinforce the graphs.
Perhaps the most important aspect of the Slutsky equation is that the class sees the relevance of separating
out the effects of the price change. The examples in the text, especially the extended example of the CPI
substitution bias, can work to motivate understanding. You could also show the separation for a price
increase in the context of a sin tax, but not until you are comfortable that separations for price decreases
are well understood.
Additional Applications
Rats Treat Quinine Solution as a Giffen Good1
Economists have long sought empirical confirmation of the Giffen good phenomenon—the occurrence
of a negative income effect so large that it overwhelms the substitution effect, creating a positively sloped
demand curve. Battalio, Kagel, and Kogut (1991) used a novel experimental procedure to demonstrate
that, within a certain income range, the Giffen phenomenon can occur at the individual level.
1Raymond Battalio, John H. Kagel, and Carl A. Kogut, “Experimental Confirmation of the Existence of a Giffen Good,”
American Economic Review, 81(4) September 1991:961–70.
60 Perloff • Microeconomics: Theory and Applications with Calculus, Third Edition
©2014 Pearson Education, Inc.
The authors began by providing six rats with the opportunity to consume liquid in the form of root beer, a
quinine solution (0.1 gram per liter), and water. The rats strongly preferred root beer to water and water to
quinine solution. Given these preferences, root beer was chosen as the normal good and quinine was intended
to serve as the Giffen good. In the experiment, the rats could “purchase” each liquid by pressing on a wall-
mounted lever (one for root beer, one for quinine) in their cage. Income was controlled by fixing the number
of presses that would produce liquid. The rats could allocate their total presses across the two levers in
whatever proportion they chose.
The authors first searched for an income range in which quinine was strongly inferior (a necessary condition
for producing the Giffen phenomenon). They then imposed income–constant price changes at these income
levels to demonstrate the Giffen effect. The quinine price was changed by altering the amount received per
lever press. As the price went up, less quinine was received per press.
Of the six rats tested, three produced a Giffen response, and three did not. Not surprisingly, the Giffen-
response–producing rats were the same three that had treated quinine as a strongly inferior good in the
income range that was tested. The other three rats, for whom quinine was not strongly inferior, did not
exhibit the Giffen phenomenon.
The authors concluded that the Giffen good is observed rarely for two reasons. First, it is difficult empirically
to generate (or to observe in human experience) the “initial conditions of strong inferiority that the theory
calls for.” Further, when the responses were averaged across rats to produce a market effect, the effect was
not significantly different from zero. Thus although some individuals did exhibit the Giffen phenomenon
in a specific income range, the heterogeneity of preferences and income levels across individuals makes
the observation of the Giffen phenomenon at the market level extremely unlikely.
1. The authors were able to demonstrate this result only at very low income levels. Why would this be so?
2. Does this experiment reinforce or weaken the theory of choice?
Discussion Questions
1. What kind of experiment could a firm conduct to determine the demand curve it faces?
2. Suppose the government wants to discourage consumption of some good (such as cigarettes or
liquor). How effective will specific taxes and lump sum taxes (a pure reduction in income) be in
reducing consumption? What type of information do you need to answer this question?
3. How can the firm use the information contained in an Engel curve and government forecasts of
income to predict future demand?
4. Are you aware of any Giffen goods? What types of goods might these be?
5. Think of several goods that, at your current income levels, you would consider normal and others that
you would consider inferior. Try to determine the defining characteristics of normal and inferior
goods by evaluating your list. What do the two groups have in common? How are they different from
each other?
6. Would you consider different goods normal and inferior at different income levels?
7. How are your preferences revealed by your buying choices?