Chapter 1

Consumer and Demand Theory

The goal of this chapter is to provide a foundation on elementary concepts of consumer and

demand theory. A basic understanding of these concepts is essential in understanding price

behavior. The chapter begins with a theoretical discussion of utility maximization, consumer

demand, and market demand. This is followed by a discussion of elasticity of demand and price

flexibility coefficients. The chapter concludes with an examination of empirical estimates of

demand elasticities and price flexibility coefficients.

Utility

A fundamental assumption of neoclassical economic theory is that people are rational

decision makers. Rational, in this context, means that people make decisions that optimize an

important objective or a set of objectives subject to various constraints such as budget and

technology restrictions. For example, owners of business strive to maximize profits or minimize

costs in planning their production decisions subject to constraints on the level of production

inputs they own or control. Consumers maximize their satisfaction when deciding on what to

consume subject to a budget constraint. Consumer satisfaction, or utility as economists

referred to it, is derived through consumption of goods and services. While almost everyone

would agree that firms and consumers do not behave exactly in this way, defenders of

neoclassical economic theory argue that most people behave close enough to this assumption to

make any derivation from this assumption to demand and supply theory a reasonable

approximation of reality. Additionally, theory based on the assumption of optimality or

1
rationality provides a roadmap for best practices for decision makers, or what we sometimes call

normative theory meaning what people ought to do to behave rationally.

In theory, there are several axioms that follow on how consumers rank preferences in

terms of deriving utility. For example, it is assumed that consumers gain more utility from

consuming more rather than less goods and service (i.e., non-satiation), and consumers can

rank their preferences in a way that satisfies certain logical conditions (see Capps and

Havlicek, 1987, for a detailed discussion of these conditions, which include the following logical

conditions: transitivity, symmetry, completeness, reflexivity, and completeness). Based on this

setup, the decision problem that consumers face is to choose a bundle of goods and services to

consume that maximizes their utility subject to a constraint that they cannot spend more than

their income level (i.e., a budget constraint). This can be expressed and solved mathematically as

the following constrained optimization problem for an individual consumer:

Max: U=U(x1,….xn)

s.t.: P1x1 +…..+ Pnxn < Y,

where U is the utility function, which depends upon consumption of all goods and services, xi

(i=1…n), Pi is the price of good i, and Y is the consumer’s income.

The optimal solution to this problem gives the consumption bundle that maximizes utility

for the consumer while satisfying the budget constraint. Based on utility maximization, a demand

function can be derived for each good and service for the consumer whose form depends

primarily upon the form of the utility function as well as the prices of xi and the consumer’s

income level. In reality, the maximization of utility is used as a conceptual construct to illustrate

consumer demand theory, but utility is not actually observed in the real world.

2
Consumer Demand

A consumer’s demand function gives the relationship between the price and quantity

demanded for any good or service, holding constant all other factors that impact demand.

Consumer demand is often depicted as a demand curve, which is a two-dimensional relationship

where the price of the good is plotted on one axis (e.g., the Y axis) and quantity demanded is

plotted on the other axis. One of the most famous laws in economics is the Law of Demand,

which states that price and quantity demanded are inversely related, e.g., as the price

decreases, quantity demanded increases and as the price increases, quantity demanded

decreases, holding constant all other demand factors. That is, demand curves have a

negative slope. The theoretical derivation of a demand curve from utility maximization can be

illustrated graphically with the use of indifference curves.

Figure 2.1 shows a map of hypothetical indifference curves for two goods: at-home food

vs. restaurant food. An indifference curve defines the various combinations of two goods that

gives the same level of utility to the consumer, e.g., in Figure 2.1, 8 units of at-home food vs.

10 units of restaurant food each give the same level of utility (U1) to the consumer. Further, as

indifference curves shift out from the origin (towards the northeast), the level of utility becomes

higher due to our assumption that more is preferred to less. Hence, in Figure 2.1, U3 is preferred

to U2 and U2 is preferred to U1. This notion of utility is called ordinal utility, which means the

consumer prefers U2 to U1, but the magnitude of the preference does not exist, e.g., we

cannot say that the consumer prefers U2 to U1 by a specific amount. Note it is assumed that

quantities of each can be consumed in fractional amounts (i.e., quantities are perfectly divisible).

Also, the convex shape of the curve (curve is bowed inward towards the origin) illustrates the

concept of diminishing marginal utility. This concept states that as more and more of one

3
good (e.g., at-home food) is consumed, the incremental gain in utility eventually diminishes,

which is reflected in the graph by the slope of the indifference curve, referred as the

marginal rate of substitution, changing for each point along the curve. For instance, consider

a point along U1 such as point P in Figure 2.1, which features a lot of restaurant food, but only a

little at-home food. In this case, in-order to get one more unit of at-home food along U1, the

consumer would have to give up a lot more than one unit of restaurant food due diminishing

marginal utility associated with a lot of restaurant food being consumed at point P. Put

differently, at point P, the marginal utility of at-home food is a lot higher than the marginal

utility of restaurant food. The exact opposite would be true of point L on U1 in Figure 2.1, which

features a lot of at-home food and little restaurant food being consumed.

<Figure 2.1>

To maximize utility, a consumer strives to reach the farthest (towards the right)

indifference curve possible, but is restricted by a budget constraint, i.e., a consumer cannot spend

more than their income level. For instance, if the consumer has $500 to spend each month on

food, and the price of at-home food is $5 per meal and the price of restaurant food is $10 per

meal, respectively, the consumer’s budget constraint (assuming the entire $500 is spent on food

during the month) is equal to the following equation:

5xh + 10xr = 500,

where xh is units of at-home food and xr is units of restaurant food consumed per month. This

budget line is depicted graphically in Figure 2.2 as the line labeled ZA. Note that the intercepts

on both axes can be derived by dividing the monthly income endowment for this consumer,

$500, by each price, i.e., xr axis is 500/5 = 100 and xr axis is 500/10 = 50. This line shows the

feasible set of at-home and restaurant food consumption for the consumer, which lies at or below

4
this line. The slope of ZA is determined by the price of restaurant food relative to at-home food,

and the distance of the line from the origin is based on the income endowment. The combination

of at-home food and restaurant food consumption that maximizes utility subject to staying within

the $500 budget can be found by the point of tangency of where the highest indifference curve

has just one point that touches the ZA budget line, which is Point B in Figure 2.2. In this

example, the utility maximizing solution is 50 units of at-home food and 25 units of restaurant

food [note that this solution perfectly exhausts all $500 = 5(50) + 10(25)].

<Figure 2.2>

What happens when the price of one of the goods changes? Suppose that the price of at-

home food decreases from $5 to $4? In this case, the budget line changes to:

4xh + 10xr = 500,

The new intercepts in Figure 2.2 are xh = 500/4 = 125 and xr axis remains at 500/10 = 50, and the

new budget line becomes ZC in Figure 2.2. Notice that now at-home food is less expensive

relative to restaurant food. With the new price of at-home food, the new point of tangency that

maximizes utility is at point K on U2. That is, the new optimum is 60 units of at-home food and

26 units of restaurant food (note that 4 (60) + 10 (26) = 500). With the decrease in the price of

at-home food, the consumer maximizes utility by purchasing 10 additional units of at-home food

and still has sufficient income left over to purchase one additional unit of restaurant food. In

some cases, the consumption of restaurant food might decrease due to a decrease in the price of

at-home food.1

1
More specifically, if the demand for one good, X, is “inelastic,” then a decrease in the price of X will result in an
increase in the consumption of the other good, Y. This is due to the fact that for inelastic goods such as X in this
case, the percentage increase in the quantity of X is less than the percentage decrease in the price of X, and hence
the total expenditure on X increases with the decrease in the price of X. On the other hand, if the demand for X is
price elastic, then a decrease in the price of X leads to a decrease in the consumption of Y. We discuss the concept
5
 By varying the price of one good, e.g., at-home food, and solving for the utility

maximizing solution, one can trace out the demand function for at-home food for the consumer.

The demand function for at-home food is depicted in the bottom part of Figure 2.2. In this case,

the demand function for at-home food can be expressed as:

Dh = f(Phú Pr, Y),

Where Dh is the demand function for at-home food, Ph is the price of at-home food, Pr is the price

of restaurant food, and Y is consumer income. Here, the notation ofú Pr, Y means that we are

holding constant Pr and Y and expressing how Dh and Ph vary together.

Income vs. Substitution Effects

A change in the price of one good relative to other goods has two effects on quantity.

The first is the income effect. The decrease in the price of at-home food in the example above

means that the consumer can now purchase more at-home food with the same level of income,

$500. For most goods, the income effect is generally positive, i.e., as the consumer has more

income, more of the good is consumed. However, there are exceptions to this that are called

inferior goods. An inferior good is inversely related with consumer income. Dry beans is an

example; as consumers income increase, consumption of dry beans decrease (Tomek and Kaiser,

2014).

The second is the substitution effect. That is, the decrease in the price of at-home food

means that it is now relatively less expensive than restaurant food, which results in more at-home

of elasticity of demand later in this chapter.

6
food to be purchased. The substitution effect is negatively related to the price change which is

consistent with the Law of Demand (Okrent and Alston 2011).

How much of the 10 additional units of consumption due to the at-home food price

decrease is due to the income vs. substitution effect? We can isolate the income vs. substitution

effects using indifference curves in Figure 2.3, which is the same as Figure 2.2 but is enlarged

<Figure 2.3>

to more clearly illustrate the two effects. In Figure 2.3, a new line (NM) is added that is parallel

to the new budget line ZC, but is pushed down to be just tangent at point Q along the original

indifference curve (U1) before the price decrease. This line represents the decrease in income

required to fully offset the income effect due to the fall in the price of at-home food. Now, with

the lower at-home food price and the lower income represented by line NM, point Q (along the

original indifference curve U1) represents a solution where the consumer is equally well off

compared with the lower price and higher real income. The move along indifference curve U1

from B to Q is attributable to the substitution effect of lowering the price of at-home food. On

the horizonal quantity of units of at home food axes of Figure 2.3, the substitution effect of the

at-home food price decrease is 4 units, while the income effect is 6 units.

There are two general types of demand functions. The first, which was derived above, is

called an “ordinary” or “Marshallian” (after the economist Alfred Marshall) demand curve. The

bottom part of Figure 2.2 is an example of such a demand curve. The second is called a

compensated or Hicksian (after the economist John Hicks) demand curve (Tomek and Kaiser,

2014. A compensated demand function shows the quantities a consumer would demand as prices

vary, but in addition adjusts the consumer’s income (either added or subtracted) so as to leave

the consumer’s level of utility the same. In other words, an ordinary demand function is derived

7
from the utility maximization subject to a budget constraint, while the compensated demand

function is derived by minimizing the consumer’s expenditures subject to the same level of

utility (Okrent and Alston 2011). Applied economists often estimate both types of demand

functions.

Do People Really Maximize Utility?

Probably most, if not all economists, would say the answer to this is no. Utility does not

even exist in reality so it would be difficult for even the most rational consumer to maximize it.

There are two general schools of thought on this question.

As previously mentioned, neoclassical economists would argue that even though

consumers do not maximize utility subject to a budget constraint, the basic tenants of utility

and demand theory are a reasonable approximation of reality that also yield very useful

normative theory.

On the other hand, other economists, particularly behavioral economists, have published

studies that have demonstrated violations in the basic assumptions of consumers being

rational (see Thaler, 2019, for a great discussion of this). Behavioral economics combines

psychology and economics and rather than looking at the normative behavior assumed by

neoclassical economists, try to explain how people actually behave in reality. This is referred as

positive theory.

In this book, both neoclassical and behavioral concepts are discussed, and I will leave it

to the reader to decide which camp they think is best.

Market Demand

8
 Market demand is the aggregation of all consumer demand in a given market. Like

consumer demand, market demand gives the relationship between the price and quantity

demanded for any good or service in the market, holding constant all other factors that impact

market-level demand. Since it is impossible to derive a market demand curve from utility

maximization, market demand curves are estimated by market-level data on prices, quantities,

aggregate market wide income, and other factors that influence demand.

Applied economists often estimate demand functions for food items. To do this, data on

prices and quantities are necessary. Three broad types of data are often used based on secondary

data sources from the government or private data venders. First, market-level data on total

volume and average prices over a time-period and given frequency, e.g., quarterly data

from 2000 through 2022. The frequency of such data can be annual, quarterly, monthly,

weekly, and so on. Second, cross-sectional data on either individuals or households can be

used to estimate a demand function. Cross-sectional data includes prices and quantities

purchased, but in addition include household demographic data. Finally, panel-data, which

combines cross-sectional and time-series data as observations. Panel-data are especially rich

since it offers data points over individuals as well as over time.

Market demand is often estimated with quantity demanded as the dependent

variable and price as one of the independent variables. This assumes that price is given or

exogenously determined, and therefore changes in price “cause” changes in quantity

demanded. However, in reality both price and quantity demanded may occur simultaneously and

depend upon each other. In this case, both price and quantity demanded are endogenous variables

and special statistical techniques are used to correct for simultaneity bias.

9
 Alternatively, causality may run in the opposite directions, i.e., changes in quantity

demanded leads to changes in price. Indeed, this may often be the case for agricultural products

that are harvested once per year, e.g., apples are harvested in the fall, and the size of the harvest

determines the price. In this case, applied economists often estimate a price inverse demand

function, where price is the dependent variable and quantity demanded is the independent

variable, which is consistent with changes in quantity demanded causing changes in price.

Examples of both regular and price inverse demand functions are given throughout in this book.

Changes in Quantity Demanded vs. Changes in Demand

For a demand curve, a movement along the curve in response to a price change is

referred as a change in quantity demanded. That is, a decrease in price results in an increase in

quantity demanded, while an increase in price results in a decrease in quantity demanded. But

there are other factors that drive demand besides the price of the good. These factors will be

referred to as demand determinants, demand shifters, or demand drivers, and they impact the

position of the demand curve. Hence, a change in a demand determinant will result in a

change in demand, which is a shifting inward or outward of the demand curve. An increase

in demand means that consumers are either willing to buy more of the good at the same price, or

are willing to buy the same amount at a higher price. A decrease in demand means the opposite.

Demand shifters include economic, demographic, psychological, new product development,

government policies, marketing, and other factors that impact the position of the demand curve.

In terms of economic factors, the utility maximization example above illustrated the two

prominent determinants. The first is the price of related goods. That is, the demand for a good

not only depends on its own-price, but also on the prices of other goods. For instance, the price

10
of Pepsi is a demand determinant for Coke. An increase in the price of Pepsi, holding all other

factors constant, should increase the demand for Coke since Coke is now less expensive relative

to its main substitute, Pepsi. Hence, the prices of substitute products are positively related with

the demand for the product, i.e., an increase in the price of a substitute product, Y, should

increase the demand for its substitute product, X. Alternatively, the prices of complimentary

products (i.e., products that are consumed together) are inversely related to the demand, e.g., an

increase in the price of milk should decrease the demand for breakfast cereal. These demand

shifters are discussed in detail later in this chapter as cross-price elasticities of demand.

The second main economic demand shifter is consumer income, as previously discussed

in the income effect presentation. For normal goods, an increase in income results in an increase

in demand. Hence, there is a positive relationship between a change in income and change in

demand for normal goods. For inferior goods, an increase in income decreases demand. Finally,

it is possible that some goods are not impacted at all by changes in income. These would likely

be very low-priced goods such as salt.

Demographic characteristics of the market can have significant impacts on the position of

the demand curve. Demographic characteristics include the population of the market as well as

the distribution of the population by such characteristics as age, gender, race, ethnicity,

education-level, religious affiliation, and other population attributes. Consider the following

three examples to illustrate the importance of demographics as demand determinant.

First, the level of the market population obviously has an important impact on demand. If

population grows by, say, 1% per year, demand will likely grow by that amount as well. That is,

increases in population normally leads to increases in demand. Indeed, over the long-run, the

change in population is probably the most important determinant of demand. When predicting

11
what the world demand for food will be over the next 100 years, it is insightful to look at world

population growth projections, which currently indicate that the world population will grow by a

little over 0.5% per year reaching 11 billion people by 2100 (United Nations, 2019,

https://www.un.org/en/global-issues/population).

Second, the distribution of market population by age can have an important impact on

demand. For example, the largest segment of beer consumers in a market is the cohort of the

population between 18 and 26 years of age. If Market A has a higher proportion of people in this

cohort than Market B, this will likely mean that beer demand in Market A will be higher than in

Market B, all other factors held constant.

Third, the racial composition of the market’s population has important implications

for demand. For example, a study of milk demand in the New York City market over a 30-year

time period found that two of the most important demand determinants were the proportion of

Hispanics and African Americans in the NYC market (Chung and Kaiser, 2000). African

Americans have a much larger incidence of lactose intolerance than Caucasians and therefore

drink a lot less milk per capita than Whites. On the other hand, Hispanics consume a much

higher amount of milk per capita than Whites. So, it follows that changes in milk consumption in

NYC depends significantly on how the proportion of African Americans and Hispanics in NYC

changes.

The point of these examples is that the level of population and population demographics

can be very important determinants of demand, and must be accounted for in cross sectional

and/or time series demand analysis.

Positive or negative information about a product can be an important demand shifter.

Examples of positive information shifters would be news that consumption of the product has

12
health or other benefits to the consumer. For example, new medical research shows that regularly

consuming cranberry juice helps protect against urinary tract infections. Or, drinking skim or

low-fat milk helps prevent osteoporosis, which elderly women may suffer from. Another

example of positive news would be advertising, which seeks to increase demand by portraying

the product in a positive light. In addition, other marketing activities such as channels marketing

or public relations typically increase the demand for a product. The opposite is true regarding

negative information. For example, new medical research shows that consuming whole milk and

other higher fat dairy products such as cheese and butter can lead to coronary disease. Such

information would likely have a negative impact on dairy demand. Or, a food safety incidence

that occurs such as an outbreak of e coli occurring with fresh fruit may cause a decrease in fruit

demand. We will explore the impact of positive and negative information in more detail later in

the book.

A related, but usually longer term, demand shifter is changes in consumer tastes and

preferences. Consumers’ preferences are likely influenced by advertising, experience, and

education (Tomek and Kaiser, 2014). A prominent example of this is per capita milk

consumption since 1975, which is shown in Figure 2.4. It is clear from this figure that the long-

term trend in milk consumption has been declining at a very steady pace; about 1% per year. Of

course, some of this trend is due to changes in other demand drivers such as changing

demographics, prices of compliments and substitutes for milk, introduction of new products

competing against milk, etc. But some of the negative trend in milk demand is due to a declining

consumer preferences for milk. This has been a major problem for the milk industry in the

United States.

<Figure 2.4>

13
 The introduction of new products to the market can be an important demand shifter. An

excellent example of this is the introduction bottled water and sports drinks in the beverage

market. Another example is the introduction of high fructose corn syrup, which essentially

replaced sugar as a sweetener in soda and other sweetened products. Other recent examples

include non-dairy milk beverages, high protein beverages and bars, hemp products, and Keurig

coffee just to name a few.

Finally, another important demand determinant is government policies, which can either

have a positive or negative effect on demand. One of the most prominent examples of positive

government policies that substantially increase food demand is the Supplemental Nutrition

Assistance Program (SNAP) also known as the Food Stamp Program. This program gives credits

to eligible households for purchasing food at grocery stores instead of using cash. Tax rebates for

purchasing electric cars or solar panels or geothermal heating/cooling are also examples of

government policies that increase demand. Government restrictions on smoking would be an

example of a negative demand determinant. The most extreme example would be the

government making consumption of a product illegal such as bans on illicit drugs. There are

many other government policies that impact the demand for a variety of products.

Derived Demand

Most likely, if you have taken a course in microeconomics, the concept of demand was

presented in the context of a final product at the retail level, e.g., a new car. We will refer to this

as primary demand, which is a function of the price that the consumer pays for the product,

which is usually at the retail-level.

14
 In the agricultural sector, the demand for most commodities is not from the final

consumer, but rather from a processor, wholesaler, or “first handler” who buys the farm

commodity, and then processes and sells it to retailers, who then sell the final product to the

consumer. In contrast to primary demand, derived demand is the demand that these

middlemen have for the raw farm commodity. For the wholesaler/processors, derived demand

for the agricultural commodity is analogous to the demand they have for other inputs used to

produce the final good. Derived demand has similar characteristics as primary demand, but some

of the factors that impact it differ. For example, consider the derived demand for corn in the

United States. The derived demand for corn has different and competing uses. In addition to the

demand for corn for feeding animals, corn is also used to make fuel (ethanol), sweeteners, and

exports to other countries. The derived demand for corn can therefore have different demand

determinants depending on what it is being used for. In terms of ethanol (which is blended with

gasoline), the derived demand for corn will also depend critically on the gasoline price, other

factors impacting that market, and government subsidies and policies pertaining to ethanol. Corn

that is exported to other countries will depend upon international factors as the exchange rate for

the U.S. dollar relative to other currencies, consumer income in other countries, population in

importing countries, the price of corn from competing country exporters, and government trade

policies.

Milk is another example of a commodity that has many uses, and hence the derived

demand for it will vary by use. Farm milk is used to make a variety of final products such as

beverage milk, cheese, butter, nonfat dry milk, ice cream, yogurt, and other dairy products.

Furthermore, milk has different components that have different uses. For instance, milk can be

divided into fat, which is used to make butter, and the remainder can be dried to produce nonfat

15
dry milk. Milk protein is particularly useful in making cheese. Similar to corn, therefore, the

derived demand for farm milk has different demand drivers depending on the end use of the

milk.

Tomek and Kaiser (2014) discuss two approaches to estimating derived demand. The first

approach estimates derived demand as a function of the farm price (e.g., milk) relative to the

price of the output it is used to produce (e.g., cheese). The second approach assumes derived

demand is basically the retail-level demand function minus all marketing costs from the post-

farmgate, which include processing, transporting, and retailing costs. This approach assumes that

the end-product is made in fixed proportion to the farm commodity, e.g., it takes about 10

pounds of farm milk to produce a pound of cheese. Under the second approach, derived demand

depends on the same demand determinants as primary demand as well as a marketing cost

variable. In agricultural economics, both primary and derived demand functions are important

concepts, and each will be included in some of the remaining chapters of this book.

Price Elasticity of Demand

While the Law of Demand states that quantity demanded and price are inversely related,

it does not say anything about the magnitude of this relationship. Economists borrow the physics

concept of an elasticity to measure the magnitude of the how sensitive quantity demanded is to a

price change. The elasticity of demand with respect to a product’s own-price is defined as the

percentage change in quantity demanded given a 1% change in price, holding all other demand

determinants constant. Mathematically, the elasticity of demand with respect to own-price is

equal to:

h = (DQd/DP)(P/Qd),

16
where D means change in, Qd is quantity demanded, and P is price. You can see from this

formula that the price elasticity of demand has two components: (1) the slope of the demand

curve, and (2) the point along the demand curve (P, Qd) where the elasticity is being evaluated.

Suppose you had two observations for price and quantity demanded, e.g., the annual

average price and total quantity demanded in 2021 and 2022:

Year Price Quantity demanded

2021 $10 100
2022 $9 125

Using these two annual observations as points along the demand curve, and evaluating the

elasticity for 2021 as the base point, the elasticity of demand with respect to price for this

example would be:

h = {(100-125)/(10-9)}(10/100) = -25(1/10) = -2.5.

Using 2022 instead of 2021 as the base point, the elasticity of demand with respect to price is:

h = {(125-100)/(9-10)}(9/125) = -25(0.072) = -1.8.

In this case, the elasticity of demand with respect to price ranges between -1.8 and -2.5. What

does this mean? Mathematically, it means that a 1% increase (decrease) in price, holding all

other demand determinants constant, results in a 1.8% or 2.5% decrease (increase) in quantity

demanded. In this case, consumers are relatively sensitive to changes in price because their

change in quantity demanded is larger than the 1% price change. The example above pertains to

a point estimate for the elasticity, since each estimate is defined for a unique point along the

demand curve, e.g., for 2021 or 2022. For most forms of demand functions, the value of the

elasticity will be unique for each point along the demand curve. For example, for a linear

demand function, all points on the curve represent unique elasticities with the midpoint being

17
equal to unitary, while points above that representing the elastic range, and points below it

represent the inelastic range.

A second type of elasticity formula is called an arc elasticity, which is equal to the

formula:

h = [(Q0 - Q1)/(Q0 + Q1)]/[(P0 - P1)/(P0 + P1] = [(Q0 - Q1))/(Q0 + Q1)][(P0 + P1)/(P0 - P1)].

In this formula, the subscripts o and 1 represent two different points on a demand curve. This

formula represents the average elasticity between two points on the demand curve. Applying the

arc formula to the example above, the arc elasticity can be computed as:

h = [(100-125)/(100+125)]/[(10-9)/(10+9)] = (-25/225)(19/1) = -2.11.

Note that the arc elasticity value lies in between the two-point estimates.

Students who have taken introductory Calculus will note that the slope of the demand

curve can be computed as the derivative of Qd with respect to the own-price. Hence, the formula

for the elasticity becomes:

h = (dQd/dP)(P/Qd).

For example, consider the following demand function:

Qd = 100 – 5P,

If we evaluate this curve at a combination of (Qd, P) of (75, 5), the point elasticity is:

h = -5(5/75) = -1/3.

In this case, demand is said to be in the inelastic range since 1/3 < 1.

There are three ranges for elasticities that measure the degree of price sensitivity.

Demand is called inelastic if the absolute value of the coefficient lies between zero and 1. In this

case, consumers are relatively insensitive to price changes since it means a 1% change in price

leads to less than a 1% change in quantity demanded. Demand is called elastic if the absolute
18
value of the coefficient is larger than 1. In this case, consumers are relative sensitive to price

changes since a 1% change in price leads to a greater than 1% change in quantity demanded.

Finally, if the coefficient is exactly -1, this is called unitary elastic since a 1% change in price

leads to exactly a % change in quantity demanded.

There are two limiting cases to the range in elasticity of demand. The first is h = 0. In this

case, demand is perfectly fixed represented by a vertical demand curve. This represents a

perfectly inelastic demand curve, and buyers will pay whatever price is set in the market. The

second is h = negative infinity. This means the demand curve is perfectly elastic and is

represented by a horizontal demand curve. We will see examples of both of these cases later on

in the book.

Because the price elasticity generally varies for each point along the demand curve, it is

not technically correct to say that the demand for a good is elastic or inelastic (Tomek and

Kaiser). Rather, we can say that demand is elastic or inelastic within a given range of prices.

So, for example, milk tends to be highly price inelastic based on retail prices ranging from $2.50

to $4.00 per gallon. Nevertheless, for the majority of retail milk prices, we often say that the

price of milk tends to be inelastic. Also, it is important to note that the elasticity of demand is

based on a very small (i.e., marginal or 1%) change in price. Consequently, it does not apply to

large price changes, e.g., if the price elasticity at a $2.00 price is -0.25, then we should not apply

it for a projection of a 45% price change to get 45(-0.25) = -11.25. The reason for this is with a

much larger price change (45% in this example), the price elasticity itself (-0.25 in this example)

will change.

What determines the elasticity of demand with respect to price? There are several factors

that determine this. The first is the number of substitutes or close substitutes that the product has.
19
Generally, the greater the number of substitutes or the closer the substitutes, the greater the

elasticity of demand. For example, a new drug with a patent has no substitutes due to the

exclusive monopoly the patent has given the firm, and therefore the demand for the new drug

will likely be highly inelastic. On the other hand, a drug (e.g., Bayer aspirin) that no longer has a

patent and has a lot of generic substitutes will likely be price elastic because consumers could

easily switch to the next best brand should Bayer raise the price.

Second, the price of the product in relation to consumers’ incomes is likely to influence

demand. The more expensive the product, and therefore the larger the proportion of consumers’

incomes spent on it, the greater the elasticity. New cars or houses are big ticket items, and we

should therefore expect the price elasticity of demand to be high. Consumers will spend a lot of

time shopping around for the best price on big ticket items, which means they have an elastic

demand. At the other extreme, products like salt that do not cost much are more price inelastic.

Relatedly, relatively wealthy people tend to have more inelastic demand than people with lower

incomes. This is one reason why estimates of price elasticities for food products in the United

States tend to be relatively more inelastic than price elasticities for food in poorer countries.

Third, whether the product is a necessity or a luxury item will influence the demand

elasticity. Product that are necessities (e.g., insulin for diabetics) tend to have an inelastic

demand, e.g., imagine a drug that can save your life, you are going to pay whatever the price is to

obtain it. On the other hand, products that are deemed luxury items (e.g., boats) tend to have

more elastic demands we can live without luxury items.

Fourth, goods that are staple and frequently purchased products tend to be more price

inelastic than non-staple products. Families with children regularly buy staple products such as

20
milk and breakfast cereal while paying less attention to their prices compared with infrequently

purchased products.

Fifth, a firm’s advertising can influence price elasticity. When a firm advertises or

conducts a general marketing campaign, the goal is generally twofold: (1) increase demand for

the firm’s product(s), and (2) make the demand for the product(s) more price inelastic. If a firm

is successful in making the demand for its product(s) more price inelastic, it can increase its

revenue by raising the price for the product (see discussion below on the relationship between

price elasticity and a firm’s total revenue).

Finally, demand is generally more price elastic the longer the time-period allowed for

consumers to adjust to a price change. For example, suppose the price of gasoline has increased

by $0.50 per gallon. Chances are that most consumers will not make too many adjustments in

their quantity demanded for gas in the first week of the price increase. However, if the price

increase lasts a long time, consumers will make adjustments to decrease their quantity demanded

for gas by using public transit more, walking instead of driving, and even buying more fuel-

efficient cars. The same is true of other goods; the longer the time-period for adjustment, the

higher the price elasticity of demand, holding all other factors constant.

The Relationship Between Price Elasticity of Demand and Total Revenue

Figure 2.5 illustrates the relationship between price, quantity demanded, elasticity, and

total revenue, which for any point along the demand curve is equal to price times quantity. For

this linear demand curve, the midpoint, labeled as point A, is where the price elasticity of

demand is unitary, i.e., -1.0. To understand the relationship between total revenue and elasticity

of demand, first consider all the points along the demand curve above the midpoint A. For this

21
range of the demand curve, all elasticity coefficients are elastic, meaning that a 1% change in

price has a larger than 1% change in quantity demanded. For any point along this range, a change

in price is inversely related to a change in total revenue. For example, a 1% decrease (increase)

in price will lead to a larger than 1% increase (decrease) in total revenue. In the case of a price

decrease within the elastic range of the demand curve, a 1% decrease in price will lead to a larger

than 1% increase in quantity demanded and the net result will be an increase in total revenue

(P*Qd). In the case of a price increase in this range, a 1% increase in price will lead to a larger

than 1% decrease in quantity demanded and, hence, total revenue will fall.

<Figure 2.5>

What about points along the demand curve below the midpoint, A? This range of the

demand curve represents inelastic points meaning that a 1% change in price will lead to a less

than 1% change in quantity demanded. For any point along this range, a change in price is

positively related to a change in total revenue. That is, an increase in price will increase total

revenue and a decrease in price will decrease total revenue. For example, by definition, a price

increase within the inelastic range means a 1% increase in price will result in a less than 1%

decrease in quantity demanded thereby increasing total revenue. The opposite is true for a price

decrease.

Figure 2.5 illustrates this, more generally, by showing demand in the top graph, total

revenue in the middle graph, and elasticity of demand in the bottom graph. The price that

maximizes total revenue corresponds to the midpoint A which has a unitary price elasticity of

demand. This relationship has major implications for agricultural policy. This can be seen by

looking at the derived demand for agricultural commodities in more detail.

22
 For most agricultural commodities, the derived demand is thought to be in the inelastic

range. As we just saw, this means that price and total revenue vary in a positive way. This fact is

the main justification for supply control policies in the agricultural sector, that impose quantity

restrictions on farmers in the hope of improving farm revenue. Supply control policy is the

foundation of Canadian agricultural policy and has been used to raise farm income in a lot of

countries, including the United States. Indeed, as early as 1915, future Secretary of Agriculture

and Vice President Henry A. Wallace wrote: “The Demand Laws...indicate to me that the

farming class as a whole is penalized by over-production and rewarded for under-production” (as

cited in Stigler 1962, and Tomek and Kaiser, 2014). In addition to raising total revenue,

restricting supply in this case also lowers production costs which further enhances farm

profitability.

Elasticities of Demand Determinants

The concept of an elasticity can and is applied to other demand determinants as well as

the own-price. For example, consider the function for a dependent variable y, as a general

function of the explanatory or independent variables, xi:

y = f(x1, x2, …, xn)

The general formula for the elasticity of y with respect to any variable, xi, is:

ey.xi = (Dy/Dxi)(xi/y), or using Calculus

ey.xi = (¶y/¶xi)(xi/y)

This general formula will be useful in discussion of all other elasticities of demand in this

chapter, and supply elasticities in the next chapter.

23
Cross-Price Elasticity. An important economic demand determinant previously discussed is the

price of related goods. A cross-price elasticity measures the percentage change in demand given

a 1% change in the price of a related good holding all other demand determinants and own-price

constant. Note that in this definition, we say a change in demand rather than a change in quantity

demanded since a cross-price is a demand shifter. Mathematically, a cross-price elasticity is

defined as:

hij = (DQdi/DPj)(Pi/Qdi), or, using Calculus

hij = (¶Qdi/¶Pj)(Ppi/Qdi).

where the i subscript represents demand for good i and the j subscript represents a price for a

related good j.

As previously discussed, there are three potential types of related goods and each can be

defined based on the substitution effect by the sign of hij. In general, a good is a substitute if hij

is positive, while a compliment if hij is negative. A related good is called independent if hij = 0.2

It is important to note that cross-price elasticities are normally not symmetric, i.e., hij

does not necessarily or even usually equal hji. For example, the cross-price elasticity of milk

with respect to the price of breakfast cereal is not usually equal to the cross-price elasticity of

breakfast cereal with respect to milk.

2 If we factor in the income effect, there may in rare cases be an exception to this. The income effect generally is
inversely related for cross-price elasticities, i.e., a decrease in pj increases demand for good i while an increase in pj
decreases demand for good i. If the income effect is larger than the substitution effect, there will be a net reduction
in the demand for good i when pj increases. Normally, consumers substitute good i for good j when the price of j
increases, but in this case the increase in pj reduces income, which adversely impacts consumption of both i and j.
Thus, the real income effect on consumption of i will be negative, while the substitution effect will be positive. If
the former exceeds the latter, the net effect may be negative even though the two products are substitutes (Tomek
and Kaiser, 2014).
24
Income Elasticity. An income elasticity measures the percentage change in demand given a 1%

change in income. It is a measure of how sensitive demand is to income changes, and it is

generally positive for most goods. Mathematically, the income elasticity is computed based on

the following formula:

hiy = (DQdi/DY)(Y/Qdi), or

hiy = (¶Qdi/¶Y)(Y/Qdi).

where Y is income.

For most food items, income elasticities decline as income levels increase. For example,

it is likely that the income elasticity for food for a millionaire is substantially smaller than for a

person living below the federal poverty line. A similar comparison would be true of comparing

general income elasticities between a rich and poor country. Tomek and Kaiser (2014) state that

the average income elasticity for food products in the United States is probably around 0.2, i.e., a

1% increase in U.S. income should result in about a 0.2% increase in the demand for food.

Other demand shifters. Computation of the elasticity of demand with respect to any other

demand determinant is done in exactly the same way as for income or cross-prices. For example,

advertising is a potential demand shifter, and to compute the demand elasticity for advertising,

we would use the following formula:

hAD = (DQdi/DAD)(AD/Qdi), or

hAD = (¶Qdi/¶AD)(AD/Qdi).

where AD is a measure of advertising effort such as expenditures by a firm on advertising their

product.

25
Total Elasticity

All definitions of elasticities discussed thus far assume that all other demand

determinants are held constant, which is also referred as the ceteris paribus (which is Latin for

all other things equal) assumption. For instance, an own-price elasticity of demand gives the

percentage change in quantity demanded given a 1% change in price, ceteris paribus. However,

in reality, it would be unusual for just one price to change ceteris paribus. If the own-price

changes, it is likely that the price of related goods will also change. We will therefore refer all

previously discussed elasticities as partial elasticities since everything else is being held

constant.

Focusing on substitute prices, there generally is a positive relationship between changes

in the own-price and changes in the price of its substitutes. Take pork demand for example. A

decrease in the price of pork will likely cause the demand for beef to decrease since pork and

beef are fairly-strong substitutes. Assuming that the beef supply is fixed in the short-run, a

decrease in demand for beef will cause the beef price to decrease. The decrease in the price of

beef will then cause the demand for pork to decrease since the beef price is a positive associated

demand shifter for pork demand. This process will continue until a new equilibrium is reached.

Thus, in the case where substitute product prices are allowed to change as well in response to an

own price decrease, the total response will actually be more price inelastic than the partial

elasticity based on the ceteris paribus assumption. This is due to the fact that the own price

decrease causes a decrease in the substitute price(s) which shifts the own product demand curve

inward leading to a lower quantity response. In essence, the own-price decreases causes two

countervailing effects: (1) an increase in quantity demanded and (2) a decrease in demand. This

process is illustrated graphically in Figure 2.6, where demand is initially D1, but then the process

26
begins with a price decrease in the own price, leading to a decrease in demand and price of the

substitute price, leading to a decrease in demand for the own product to D2, and subsequently to

D3, and so on.

<Figure 2.6>

The notion of a total demand response, or a total elasticity was first introduced by an

agricultural economist in 1958 (Buse 1958: 882). Buse devised a formula for total demand

response that is based on the own-price partial elasticity, cross-price elasticities, and a measure

of their price relationship, which is equal to the following for good i:

n
Ti = hi + S hijSji,
j=1

where Ti is the total or net response curve for good i, hii is the partial own price elasticity of

demand for good i, hij is the cross-price elasticity of good i with respect to substitute good j, and

Sji is the percentage change in the price of j given a one percent change in the price of i. This

measures the complete response of demand to a price change when substitute prices are “allowed

to act and interact as the market structure requires to reach a new equilibrium level” (Buse 1958:

882). Here it is assumed that the main demand determinant that is allowed to vary is the prices

of substitute goods, while other demand determinants are held constant. Looking more closely at

this formula one can see that Ti is smaller (more inelastic) in absolute value than hI since hij is

positive for substitutes and Sji is positive. Hence, for the equality in this formula to hold it

follows that Ti is smaller (more price inelastic) in absolute value than hi.

Suppose that the demand for beef has an own price elasticity of -0.3 and the cross-price

elasticity of beef demand with respect to the price of poultry and pork are 0.10 and 0.15,

respectively. Also, assume that the percentage change in the price of poultry given a one percent
27
change in the price of beef is 0.4 and the percentage change in the price of pork given a one

percent change in the price of beef is 0.5. With this information, we can compute the total

demand response for beef as:

Tbbeef = -0.3 + 0.10(0.4) + 0.15(0.5) = -0.185

In other words, the partial own price demand elasticity for beef is relatively more elastic

(-0.3) than its total elasticity (-0.185).

Price Flexibility Coefficients

As was mentioned previously in this chapter, sometimes applied economist use price

inverse demand functions, where price is the dependent variable and quantity demanded is one of

the explanatory variables. In this case, there is a similar concept to the elasticity of demand

called a price flexibility coefficient. This coefficient measures the percentage change in price

given a 1% change in the explanatory variables (including quantity demanded) holding

other demand (measured as price) determinants constant. For example, consider the

following inverse demand function:

P = D-1(Qd)

In this case, the price flexibility coefficient is equal to:

F = (DP/DQd)(Qd/P), or

F = (¶P/¶Qd)(Qd/P).

Inverse demand functions and their price flexibility coefficients are particularly relevant for

many agricultural commodities because of the biological nature of agricultural production. That

is, for many agricultural commodities, output is not instantaneously produced, but rather there is

a lag between when production decisions are made and when output is realized. Because of this,
28
current production is not influenced by the current price, but rather the expected price at the time

when production decisions were made. Because of this, supply, once it is harvested, is “pre-

determined” by a past price, and is fixed at harvest-time and cannot be changed. In order for

markets to clear, the current price is then determined by the fixed quantity of the pre-determined

supply. Hence, in this case, the direction of causality runs in reverse of a quantity dependent

demand function, i.e., price is determined by the pre-determined quantity of supply. The Law of

Demand still holds; higher predetermined supply results in lower prices while lower

predetermined supply leads to higher prices. Because of this, the own-price flexibility coefficient

is negative just like a price elasticity of demand.

The formula for a price flexibility coefficient appears to be the inverse of a demand

elasticity, i.e.,

F = 1/h.

However, this is not exactly true because one must also consider other price determinants in the

demand model. This can best be illustrated by examining a typical quantity demand function vs.

a typical price inverse demand function. Suppose we postulate that the quantity demand function

for beef is the following:

Qdb = f(Pb, Pc, Pp, Y),

where Qdb is quantity demanded of beef, Pb is the own-price of beef, Pc is the price of chicken, Pp

is the price of pork, and Y is consumer income. The price inverse counterpart to this would

typically be:

Pb = f(Qdb, Qc, Qp, Y),

where Qc is quantity demanded of chicken, Qp is quantity demanded of pork, and all other

variables are as previously defined. So, quantity dependent demand functions have the prices of

29
related goods as demand shifters, while price inverse demand functions use quantities instead.

As Tomek and Kaiser (2014) point out, because different variables are being held constant in the

two functions, the reciprocal of the flexibility need not equal the elasticity. Rather, the

following relationship holds between the price elasticity of demand and the price flexibility:

|h| > |1/F|.

Therefore, the own-price elasticity will generally be larger than the reciprocal of the flexibility.

The interpretation of the price flexibility coefficient is similar to that for regular demand

elasticities, but the relationship is the inverse. For instance, demand is considered to be inelastic

if the price flexibility coefficient, in absolute value, is larger than one. The price flexibility

coefficient, in absolute value, will be between zero and one for elastic demand. Of course, one

needs to keep in mind the above relationship between h and F for this general definition of

elastic vs. inelastic demand based on price flexibility.

We can estimate price flexibilities with respect to income and cross-price as well (indeed,

price flexibility coefficients can be computed for all demand (i.e., price) determinants. The

definition of a price flexibility coefficient with respect to income is the percentage change in

price given a 1% change in income holding all other demand determinants constant. For a normal

good, the sign of this coefficient is positive, while for an inferior good, the sign is negative.

Since the quantities demanded for related goods are usually used instead of their prices, cross-

price elasticities are not computed from price inverse demand functions. However, a related price

flexibility can be computed using the quantities instead of the prices which provides a measure

of the percentage change in the own-price given a 1% change in quantity demanded of a related

good, i.e.,

Fij = (DPi/DQj)(Qj/Pi).
30
where Fij is the price flexibility of good i with respect to the quantity demanded for good j,

holding all other demand determinants constant.

If Fij is negative, that means a 1% increase (decrease) in quantity demanded for good j

would result in a decrease (increase) in the price of good i. This indicates that goods i and j are

substitutes since the decrease (increase) in demand for one leads to an increase (decrease) in

price (the measure of demand) for the other product. To better understand this, one needs to note

that the direction of causality for price inverse demand functions is that predetermined quantities

determine price. If Fij is positive, that means a 1% increase (decrease) in the quantity demanded

for good j results in an increase (decrease) in the price of good i. In this case, the two products

are compliments since the increase (decrease) in demand for one product leads an increase in

price (the measure of demand) for the other product.

Empirical Elasticities

Applied economists have published many studies that estimate demand elasticities or

price flexibility coefficients for agricultural and food goods. Table 2.1 lists estimated own-price

and income (or expenditure, which is analogous to income) elasticities for various food products.

These measures were estimated using econometric models of market demand with various types

of data.

Tomek and Kaiser (2014) caution that these estimates are not the same as the true,

underlying, elasticity, since that is impossible to measure. There are errors in the data and the

models are not perfect due to missing variables and other problems, and hence it is usually useful

to compute confidence intervals for the point estimates. Many studies do this. Several studies

31
have focused on the potential problems of empirical price analysis (Alston and Chalfant 1991;

Davis 1997; Tomek 2000).

<Table 2.1>

It is clear from Table 2.1, that the demand for agricultural commodities tend to be both

price and income (or expenditure) inelastic. For instance, with the exception of bananas and

citrus, all price elasticities are less than one in absolute value. With the exception of coffee and

tea, all income or expenditure elasticities are below one as well. This inelastic nature of

agricultural and food demand in the United States is well known, and the majority of studies

have demonstrated this.

It should be pointed out, however, that this inelasticity is a feature of the broad categories

that are being estimated here, e.g., milk, beef, pork, etc. For more disaggregated demand studies,

the price elasticity generally increases. For example, if one was looking at a n individual product

produced by a firm within a competitive market, e.g., Brand X ice cream, it is very likely that the

price elasticity of demand for Brand X is in the elastic range due to the existence of substitutes.

The same is true as we disaggregate the commodity. Hence, if we were looking at numerous

different milk products such as whole milk, low-fat milk, skim milk, organic milk, rBST-free

milk, flavored milk, and so forth, the price elasticity would almost certainly more elastic than

viewing milk as an aggregate commodity.

Concluding Remarks

This chapter reviewed consumer and demand concepts. We began with a discussion of

maximization of utility subject to a budget constraint and the derivation of a consumer’s demand

function. While abstract, demand theory based on utility maximization provides useful

32
foundations for applied economists interested in estimating market demand functions for various

products.

We then discussed the two important effects of a price change on demand: (1)

substitution effect, and (2) income effect. Assuming the price of one product falls, the

substitution effect means consumers will increase their quantity demanded for the good since it is

now less expensive relative to other products they buy. The income effect, in this case, means

that people increase their demand for the product whose price has fallen since it is akin to an

increase in the real income of the consumer. Of course, the opposite effects will occur in case of

a price increase of a product.

Next, the distinction between a change in quantity demanded and a change in demand

was discussed. The former involves a movement along the demand curve in response to a change

in the product’s price, while the latter is a shifting of the demand curve due to a change in

another demand determinant such as income or cross-prices.

In agriculture, the demand for farmers’ commodities by processors is called derived

demand. Unlike primary demand by consumers for final products, derived demand for

agricultural commodities by processors is analogous to the demand they have for other inputs

used to produce the final good. Derived demand has similar characteristics as primary demand,

but we argued that some of the factors that impact it differ. For example, commodities like farm

milk have many uses and hence the derived demand for milk by a cheese processor will be

different from that of a butter manufacturer or beverage milk processor.

The very important measure of elasticity of demand was discussed in detail. We argued

that the price elasticity of demand for most food products tends to be relative inelastic. In a later

chapter, we will discuss how this inelasticity along with the inelasticity of agricultural supply

33
presents price instability issues in the agricultural sector. The related concept of total demand

response, which does not hold constant the price of substitute products, was then presented. And,

price flexibility coefficients, which are the analog of price elasticities for price inverse demand

functions, were examined. Finally, the chapter concluded with a brief discussion of empirical

estimates of food elasticities. Most estimates for food are highly price inelastic due to such

factors as food being a necessity, relatively small share of consumers’ budgets, and many being

staple products that are frequently purchased.

In the next chapter, we have a similar discussion of supply concepts and theory of the

firm.

34
References

Alston, J. M., and J. A. Chalfant. 1991. “Can We Take the Con Out of Meat Demand Studies?”
West. J. Ag. Econ. 16:36-45.

Buse, R. C. 1958. “Total Elasticities–A Predictive Device,” J. Farm Econ. 40:881-891.

Capps, O., Jr., and J. Havlicek, Jr. 1987. “Concepts of Consumer Demand Theory,” in Food
Demand Analysis, R. Raunikar and C-L. Huang, eds. Iowa State Univ. Press. Chapter 1.

Chung, C. and H. M. Kaiser. “Determinants of Temporal Variations in Generic Advertising

Effectiveness.” Agribusiness: An International Journal. 16(2000):197-214.

Davis, L. G. 1997. “The Logic of Testing Structural Change in Meat Demand: A Methodological
Analysis and Appraisal,” Am. J. Ag. Econ. 79:1186-1192.

Kaiser, H. M. “2016. An Economic Analysis of the Cattlemen’s Beef Promotion and Research
Board Demand-Enhancing Programs.” Applied Economics. 48:312-320.

Kaiser, H. M. 2022. An Economic Analysis of the National Pork Board Checkoff Program.
Unpublished Study Conducted for the National Pork Board.

Okrent, A. M., and J. M. Alston. 2012. The Demand for Disaggregated Food- Away-From-Home
and Food-at-Home Products in the United States. Washington, D.C.: U.S. Department of
Agriculture, Economic Research Service, Economic Research Report Number 139

Schmit, T. and H. M. Kaiser. “Egg Advertising, Dietary Cholesterol Concerns, and U.S.
Consumer Demand.” Agricultural and Resource Economics Review. 27(1998):43-52.

Schmit, Todd M. and Harry M. Kaiser. “Forecasting Fluid Milk and Cheese Demands for the
Next Decade.” Journal of Dairy Science. 89(2006):4924-4936.

Thaler, R. 2019. Misbehaving: The Making of Behavioral Economics. New York: W.W.
Norton and Company.

Tomek, W. G., and H. M. Kaiser. 2014. Agricultural Product Prices, 5th Edition. Ithaca: Cornell
University Press.

Tomek, W. G. 2000. “Commodity Prices Revisited,” Ag. & Res. Econ. Rev. 29:125-137.

United Nations. 2019. World Population Prospects. New York: https://www.un.org/en/global-

issues/population

Zheng, Y. and H. M. Kaiser. 2008. “Advertising and U.S. Nonalcoholic Beverage Demand.”
Agricultural and Resource Economics Review. 37:147-159.

35
Restaurant Food

10 U3

U2
L
U1

8 At-Home Food

Figure 2-1. Hypothetical indifference curves for at-home and

restaurant foods.

36
Restaurant Food

50 Z

26 K
25 B

U2
U1

A C
50 60 100 At-Home Food 125
Price At-Home
Food

4
Dh

50 60 At-Home Food

Figure 2-2. Derivation of demand curve for at-home food.

37
Restaurant Food

50 Z

K
26
25 Q
B
U2
U1
M
A C
50 54 60 100 125
At-Home Food

Figure 2-3. Substitution and income effects due to a price decrease

for at-home food.

38
 Figure 2.4. Per capita beverage milk consumption in the
United States, 1975-2020.
300

250

200
Pounds per person

150

100

50

0
1975 1980 1985 1990 1995 2000 2005 2010 2015 2020
Year

39
 Price

Elastic

Inelastic

Quantity
Total
Revenue

Quantity

Demand
Elasticity

0
Quantity
-1

Figure 2.5. Relationship between demand, total revenue,

and price elasticity of demand.

40
Price

Total Response

D1
D2
D3
Quantity

Figure 2.6. Total Response Curve.

41
 Table 2-1. Estimated elasticities of demand, selected foods and beverages, United
States

Product Own-price elasticity Income (expenditure)

elasticity
Beefa −0.43 0.69
Porkb −0.34 0.51
Poultryc −0.81 0.03
Fishc -0.84 0.03
Eggsd −0.08 0.18
Applesc −0.58 0.03
Bananasc −1.01 0.05
Citrusc -1.10 0.06
Other fresh fruitc −0.90 0.04
Potatoesc −0.42 0.03
Lettucec −0.84 0.04
Tomatoesc -0.58 0.06
Other fresh vegetablesc -0.94 0.04
Cheesef -0.35 0.09
Milke −0.15 0.61
Juicee −0.27 0.66
Coffee/Teae −0.46 3.14
a
Kaiser (2016).
b
Kaiser (2022).
c
Orkent and Alston (2012).
d
Schmit and Kaiser (1998).
e
Zheng and Kaiser (2008).
f
Schmit and Kaiser (2006).

42