4 & 5. Price discrimination Flashcards

1
Q

What is Price discrimination?

A

Price discrimination is the capacity of firms to price consumers differently considering what they want to purchase.
The increasing use of Internet as a market place has led to think that we would enter to an area of perfectly competitive market. The two main reasons lie in the demand side, where we can observe information enhancements, and in the supply side, with information improvements.
.
Why should firm price discriminate?
We are saying firms have market power, downward sloping demand curve like a monopoly.
.
There are five central assumptions in a perfectly competitive environment:
1. Atomicity;
2. Product homogeneity;
3. Perfect information: every agent (firms, consumers) knows the price charged by every firm
within the market;
4. Technological symmetry: every firm has access to the available production technologies;
5. No entry and exit barriers.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Perfect price discrimination – 1st grade

A

It consists in pricing each consumer with a different price for each unit sold, price that exactly matches his/her willingness to pay (under the hypothesis that each consumer buys only 1 unit).
This way, the firm is able to recover both losses: the profit surplus is maximized, the social welfare the same (so this is an efficient solution). The problem is that there is not the 􏰃o􏰂su􏰄e􏰖􏰗s surplus: there are no differences between price and willingness to pay. Since all the pie goes to the firm, this type of discrimination is considered a sort of utopia, a highly theoretical possibility but hard to be developed in reality. This is because a firm is unable to know in advance the willingness to pay of each consumer. Secondly, even if this becomes possible, it is impossible to prevent bad behaviors of consumers (for instance, they can buy at a low price and then resell the good at a higher one).
Internet has made this discrimination less utopist, because through the web it is easier to know the willingness to pay.
.
Look at graph.
.
Discussion:
Optimal pricing solution. Inefficient in terms of social welfare in the viewpoint of monopolies too. Because choosing uniform pricing is leaving money on the table. In particular it loose the consumer surplus A but also the deadweight loss B

Why is it leaving money in the table bcz there are ppl who will pay more than what they actually paid (A)

Will pay more than marginal cost for the good

A way to sell to those customers for a profit..

Price discrimination will try to recoup the quantities A and B. We will analyse 3 ways,
!st price discrimination: perfect from pop of social welfare and


Profit is not just the level before but this whole area. A consumer surplus, B wipe out deadweight loss, producer surplus is at the max level and equal to social welfare. Issue is ethics because Consumer surplus = 0.

This is a theoretical concept. (Like the pizza example, 1st slice, 2nd slice) Considered impossible, was difficult to avoid resale phenomenon. Ppl lower on the demand curve may sell for a higher price to the ppl higher in the demand curve.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Perfect price discrimination (I°)

A

(Pigou, 1920):
One consumer - One price for each unit sold
Each customer is charged a different price – exactly matching his/her willingness to pay for each unit
.
Max Producer Surplus.
Max Social Welfare.
Equity problems: no Consumer’s Surplus.
Problems:
Which is the wtp of each consumer is difficult to know
Difficult to avoid arbitrage (absence of resale).
.
Example: Consumer profiling (or price steering) via devices already exists(search made by myself through Booking.com via smartphone (left) and desktop (right))

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

Extra: From “shopping-then-shipping” to “shipping-then-shopping”?

A

Most shoppers have noticed Amazon’s recommendation engine while they shop. At present, Amazon’s AI does a reasonable job suggesting items for customers to buy, particularly considering the millions of items on offer. However, the company’s predictions are far from perfect. […] Now, imagine that Amazon’s AI collects more information about us: In addition to our searching and purchasing behavior, it also collects other data it finds online, including social media, as well as offline, such as our shopping behavior at Whole Foods. It knows not only what we buy, but also what time we shop, which location we shop at, how we pay and more. What if the AI uses that data to improve its predictions? What happens to Amazon’s strategy as its data scientists, engineers and machine learning experts work tirelessly to dial up the accuracy on the prediction machine?
At some point, the AI’s prediction accuracy crosses a threshold, such that it becomes in Amazon’s interest to change its entire business model. The predictions become sufficiently accurate that it becomes more profitable for Amazon to ship you the goods that it predicts you’ll want rather than wait for you to order them.[…..]
By Ajay Agrawal, Joshua Gans and Avi Goldfarb – Harvard Business Review, Oct 4, 2017.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

3° Price discrimination

A

“Group” Pricing: different prices for different groups of consumers, same price within the same group.
Selection by (exogeneous) indicators:
Age
Occupation
Geography
(Internet) Device
Examples: geographical market segmentation (books in India and UK); special discounts (senior, student, etc); mode of buying.
.
Extra: It consists in setting different prices for different groups of customers (group pricing). Within the same group, prices are homogeneous.
Sometimes it is convenient to subdivide the market considering some exogenous indicators:

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

3 degree price discrimination: The trick is to apply the elasticity rule to each market segment:

A

The trick is to apply the elasticity rule to each market segment:
.
- In sub-market i: max πi = TR(qi) – TC(qi) o pi(qi) · qi – TC(qi)
.
First order condition formula: pi+qidpi/dqi=MC
pi[1+(qi/pi)
(dpi/dqi)]=MC
pi[1-(1/|ei|]=MC
(pi-MC)/pi=1/|ei|
.
- In sub-market j repeat the same and obtain: (pj-MC)/pj=1/|ej|
.
Of course absence of resale possibilities is needed also in this case
.
Implications:
Rule: different elasticities = different prices.
Specifically, higher prices in less elastic markets

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

Exercise on 3° price discrimination

A

Exercise on 3° price discrimination
No fixed costs, marginal costs constant and equal to 2

Total demand curve represented by uptown market
.
LEt’s suppose firm has total cost function = 2q
Analyse 2 scenarios,
1st w price discrimination (can subsegment markets), 2nd no price discrimination

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

2° Price discrimination

A

Self-selection by consumers
seller cannot directly identify consumer type, but can still induce consumers to distinguish themselves. This selection may be based on the willingness of consumers to consume:
different quantities (so price paid by consumers depends on the quantity of the good consumed: non linear-pricing, bundling)
Different versions of the same product (Versioning): (often) case of VERTICAL DIFFERENTIATION (rather than HORIZONTAL)
.
Extra:
It is characterized by self-selection by consumers: seller cannot directly identify consumer type, but can still induce consumers to distinguish themselves. This selection may be based on the willingness of consumers to consume:
- Different quantities (so price paid by consumers depends on the quantity of the good consumed: non-linear pricing);
- Different versions of the same product (Versioning).

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

A typical non-linear pricing technique is the two-part tariff:

A
  • With identical consumers (same demand curve), the firm may obtain the maximum surplus possible (same as perfect discrimination)
  • With heterogeneous consumers, the firm will opt for multiple two-part tariffs. If there are 2 typologies of consumers with CS2(p) > CS1(p), we will have: A1 < A2 e p1 > p2
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Versioning

A

The aim is to sell more than one version of the same product at different prices targeting different segments of consumers.
.
This applies also when the low quality version has the same cost of production of the high quality one (or even higher):
o Software: basic version obtained by degrading the premium through the disablement (bearing some costs) of some functions;
.
o Information services about share prices: the delayed version is produced with some additional costs with respect to the immediate version.
Companies leave that consumers choose the version they prefer (self-selection).
.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

The art of Profitable versioning:

A

In order to make a profitable versioning, there is a key constraint (incentive): you cannot make the inexpensive version too attractive to those willing to pay more. Therefore, companies need to lower the price of the premium version and lower the quality of the basic one.
.
Key constraint (incentive): you can’t make the inexpensive version too attractive to those willing to pay more.
Need to: Lower price of the premium version
Lower quality of the basic version
.
.
One additional constraint (participation) refers to the fact that the cheap version must be sufficiently cheap that low types are willing to purchase.
.
Additional constraint (participation): basic version (of decent quality) must be relatively inexpensive that those with relatively low (but still sufficient) willingness to pay wish to purchase (to “be in the market”).
.
.
The second additional constraint (good design): it is impossible for consumers to transform the basic into the premium version. they need to be really different versions: don’t upset “high-wtp” customers.
.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

Versioning (and PD) and Social Welfare

A

What about consumer welfare? Is versioning good or bad for consumers?
It is difficult to say a priori. The answer crucially depends on whether versioning is able to enlarge the customer base or not:
- If the number of customers increase, the consumer’s surplus may not be inferior with versioning;
- If the number of consumers does not increase, then versioning enables the producer to gain more at the expense of consumers. In this case, versioning leads to a lower consumers’ surplus.
.
(Discussion in class)
sales are increasing with versioning) or if more products are sold. Social welfare is low.
.
.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Versioning on Quantity

A

Second degree price discrimination (“sort of” versioning on quantity)
.
TWO-PART TARIFF:
- Tariff entails a fixed entry fee (f) and a per-unit price (p) (Abbonamento)
- Total average unit price depends on quantity
- Customers self-select depending on their preferences
- Simplifying, it is like having 2 main dichotomous options: a) consume few units (and spend less in absolute terms) but paying a high average price; b) consume more units (and spend more in absolute terms) but paying a lower average price.
.
Beware of the participation constraint:
Excursus: To better meet the participation constraint the seller may also propose different combinations (f,p), i.e. a (relatively) low f and high p for “low” demanders (those who consume few units), and a (relatively) high f and low p for “high” demanders (see the graphical representation in the next slide)
.
.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

VERSIONING(functionalities/ product features) How many versions?
No general rule

A

There are not general rules to decide how many versions a company should implement.
More versions lead to:
- More possibilities to capture all the value from consumers;
- More personalization costs (which presumably are convex in the number of versions);
- Risk of cluttering effect for potential consumers.
.
To alleviate cluttering (especially when the number of versions increases) many firms think is important to describe functionalities in a very detailed way for each version.
.
Shapiro and varian said that “if you cannot decided how many versions to have, choose 3.”
3 can be better than 2 for the phenomenon of extremeness aversion: that is the risk that 2 versions (heavy and light) are felt by potential consumers as too big or too small, with the risk that a high percentage will opt for the light version, thus generating less revenues for the company.
.
Adding another category (super gold premium) and making the previous premium version as the medium one can produce some advantages.
.
If switching from 2 to 3 is not possible for some reasons, note that a firm might always artificially increase the number of versions and sell the preferred version exploiting the cognitive inclination of consumers.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

Simonson e Tversky experiment (1992, JMR): Extremeness Aversion (and cluttering); Decoy effect

A

Microwave oven. Two possible set of choices for two identical groups of people (n = 60):
- 1° group:
1) basic (Emerson): 110 $ 57% of consumers
2) premium (Panasonic): 180$ 43% of consumers
- 2° group:
basic (same): 110 $ 27%
premium (same): 180$ 60%
High-premium (Panasonic): 200$ 13%
.
Annotation1: note that with the “extremeness aversion” (and cluttering) argument we are now diverging from the rational “homo oeconomicus” implied by the neoclassical theory who always knows exactly and with no doubt what she/he wants
.
Intro to behavioural economics: Thinking fast and slow; Misbehaving
.
.
Annotation2: if switching from 2 to 3 is not possible for whatever reason, note that a firm might always artificially increase the number of versions and sell the preferred version exploiting the cognitive biases of consumers
Dan Ariely and the Decoy effect(in Predictably Irrational, 2010)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

We can introduce the Decoy effect:

A

Extra on pg 64 of studocu notes

17
Q

Prospect theory

A

Generally, individuals (and so consumers) are much more sensitive to losses than to gains
.
The decoy effect exploits this psychological trait: the average consumer does not know if s/he “wins” by choosing “3” but s/he is sure (or at least more confident) that s/he does not lose by choosing that option.
.
Thus, more generally, our choices may depend on how outcomes are presented (“framed”) to us. Invariance (and possibly other axioms) of “rational choice” may not hold
( Kahneman & Tversky (several articles, see in particular to “Choices, values, and frames”, 1984, American Psychologist, 39(4), pp. 341-350)
.
.
Slide 32, 5. Price discrimination part 2 lecture. 9th march.
(p. 343) “The formulation of Problem 1 implicitly adopts as a reference point a state of affairs in which the disease is allowed to take its toll of 600 lives. The outcomes of the programs include the reference point and two possible gains, measured by the number of lives saved. In this case preferences are risk averse. A clear majority of respondents prefer saving 200 lives for sure over a gamble of that offers a one third chance of saving 600 lives.”
.
“[…] Now consider another problem in which the same cover story is followed by a different description of the prospects associated with the two programs:
“It is easy to very that options C and D in Problem 2 are undistinguishable in real terms from options A and B in Problem 1, respectively. The second version, however, assumes a reference state in which no one dies for the disease. The best outcome is the maintenance of this state and the alternatives are losses measured by the number of people that will die of the disease.” People in this case are risk seekers.
.

18
Q

Discussion in class: participation constraint, 2 degree price discrimination

A

Profit won’t be 5000 but 3000 because everyone will buy the delayed version. All 100 consumer,
Why?
Because difference is high between the two;
There is a consumer surplus for the impatient. By buying the delayed version of 30, consumer surplus is 10 (because willingness to pay is 40) They have utility of 10 euro.
Consumer surplus will be 0 if they pay 100. They will self select.
Is there a way to make this version function well?

Sell at 89. Opportunity cost?
Price immediate version: 89.99




If price same for delayed version


Still not profit as high as before which is unrealistic. But still it’s positioned in between uniform pricing and

Incentive compatibility constraint when the version j is effectively both by typology of consumer j. j customers, They do not self select for i not thought for them.

19
Q

References for price discrimination lecture

A

Lecture 4: price discrimination part 1
- Cabral Industrial Organization I edition, 2000: chapter 10 [or Industrial Organization II edition, 2018: chapter 6.1-6.3].
- Varian, Intermediate Microeconomics, chap. 26.7.
.
Further reading:
- Cabral Industrial Organization I edition, 2000: chapter 6.3, 6.4 [or Industrial Organization II edition, 2018: chapter 4.2, 4.3].
- Varian, Intermediate Microeconomics, chap. “Monopoly Behavior”, 26.1, 26.2, 26.3, 26.4, 26.5, 26.6)
- Shapiro & Varian, Information Rules, chapter 3.
.
.
Lecture 2: PD part 2
Cabral Industrial Organization I edition, 2000: chapter 10 [or Industrial Organization II edition, 2018: chapter 6.1-6.3].
Varian, Intermediate Microeconomics, chap. 26.7.
.
Further reading:
Cabral Industrial Organization I edition, 2000: chapter 6.3, 6.4 [or Industrial Organization II edition, 2018: chapter 4.2, 4.3].
Varian, Intermediate Microeconomics, chap. “Monopoly Behavior”, 26.1, 26.2, 26.3, 26.4, 26.5, 26.6)
Shapiro & Varian, Information Rules, chapter 3.