Why do manufacturers issue
coupons? An empirical analysis of
breakfast cereals
Nevo and Wolfram
Presented by Huanren (Warren) Zhang
2/22/2012
Static Monopoly Price Discrimination
• Couponing is a tool for price discrimination
• Only more price-sensitive customers bother to
clip, save and use coupons
• Manufacturers can use coupons to sort
customers into groups with distinct price
elasticities.
However…
• Procter & Gamble’s senior vice president of
advertising once said
“I don’t like couponing. Period.”
Static Monopoly Price Discrimination
• 𝜋 𝑝1 , 𝑝2 : profit when the consumers are
faced with full price 𝑝1 and price with coupons
𝑝2
• 𝜋 𝑝1 , 𝑝2 is continuous and twice
differentiable with a unique optimum at
∗ ∗
(𝑝1 , 𝑝2 )
• 𝜋 𝑢 𝑝 = 𝜋 𝑝, 𝑝 : profit with uniform price,
peaked with optimum at 𝑝𝑢∗
• Proposition: 𝑝1∗ ≤ 𝑝𝑢∗ ≤ 𝑝2∗
Price Differentiation
A direct implication of
the proposition is that
coupons and shelf
prices are positively
correlated
Relevant Liturature
• Existing work uncovered patterns consistent with the
price-discrimination interpretation of coupons:
– Coupon users have more elastic demand than nonusers
(Narasimhan 1984)
– Low-priced generic products have lower market shares if
the brand-name manufacturers coupon heavily
(Sethuraman and Mittelstaedt, 1992)
– Larger percentage of consumers use coupons for brands
with higher shelf prices (Vilcassim and Wittink, 1987)
• Few work is done on the relationship between shelf
prices and coupons
Static Monopoly Price Discrimination
• Unrealistic for cereal markets
– Not monopoly
– Ignores the changes in demand over time
– Manufacturers do not sell the shelf price to retail
consumers
– Managers set coupon policies that may not fully
internalize profit-maximizing incentives
Other models
•
•
•
•
Oligopoly Price Discrimination
Dynamic Demand Effects
Retailers’ Objectives
Retailer or manufacturer costs
Oligopoly Price Discrimination
• Under symmetry assumptions, the conclusion
of monopoly price discrimination can be
carried onto oligopolistic industries (Holmes
1989)
• BUT, under certain conditions, price
discrimination may lead to lower prices and
profits (Corts 1998)
Oligopoly Price Discrimination
• Professors Prefer Raisin Bran
• Students prefer Cheerios
Oligopoly Price Discrimination
P↓
P↓
To
To compete
increase with
Cheerios,
Raisin
profit, Raisin
Bran may
offersalso
reduce
couponsitstoshelf
price
students
To keep the
market share
and profit,
Cheerios may
reduce the shelf
price
Oligopoly Price Discrimination
• Prices fall for all consumers if the coupon
users and nonusers have different brand
preferences (“best-response asymmetry”)
• Assess the influence of strategic interaction by
investigating the effect of the presence of
coupons for competing brands
Dynamic Demand Effects
• Low-valuation consumers are willing to
postpone purchases
• Sellers periodically lower prices to clear out
low-valuation consumers
• Coupons are issued in response to intertemporal patterns in demand (accumulation
of low-valuation consumers)
Dynamic Demand Effects
• Coupons tend to follow periods of low-volume
sales
• The quantity demanded would be lower
following a period when coupons were issues
Dynamic Demand Effects
• Coupons tend to follow periods of low-volume
sales
• The quantity demanded would be lower
following a period when coupons were issues
• Coupons can also be used to induce repeated
purchase
Retailers’ Objective Functions
• When they have market power, the retailers
(e.g. supermarkets) may not change the shelf
prices according to the whole sale prices set
by the manufacturers
• Can use wholesale prices to examine whether
the changes on shelf prices are driven by
retailer or manufacturer behavior
Retailer or Manufacturer Costs
• In periods when demand is expected to be
low, manufacturers may simultaneously issue
coupons and reduce prices to generate more
sales
• Managers may reduce price and issue coupon
to achieve market share goals by the end of
the company fiscal year
Data
• Cereal Price Data: IRI Infoscan Data Base
collected by a marketing firm in Chicago
• Coupon Data: research company, Promotion
Information Management (PIM)
The Model
• 𝑆𝐻𝐸𝐿𝐹 𝑃𝑅𝐼𝐶𝐸𝑏𝑐𝑡 = 𝛾𝑏(𝑐) + 𝜙𝑐(𝑡) + 𝛿𝑡(𝑏) + 𝜃 𝐷𝑂𝐿𝐿𝐴𝑅𝑆 𝑂𝐹𝐹𝑏𝑐𝑡 + 𝜖𝑏𝑐𝑡
• 𝑆𝐻𝐸𝐿𝐹 𝑃𝑅𝐼𝐶𝐸𝑏𝑐𝑡 = 𝛾𝑏(𝑐) + 𝜙𝑐(𝑡) + 𝛿𝑡(𝑏) + 𝜃 𝑃𝑅𝑂𝐵 𝑂𝐹 𝐶𝑂𝑈𝑃𝑂𝑁𝑏𝑐𝑡 + 𝜖𝑏𝑐𝑡
•
•
•
•
•
•
𝛾𝑏 : brand fixed effects
𝜙𝑐 : city fixed effects
𝛿𝑡 : time trend of in prices
𝛾𝑏𝑐 : brand-fixed effects to vary by city
𝜙𝑐𝑡 : city-fixed effects to vary across quarters
𝛿𝑡𝑏 : the quarter effects to vary by brand
We need to understand the negative
correlations between prices and coupons
Cross-Brand Effects
The negative coefficients are driven by the
interaction between manufacturers’ and their
competitors’ couponing
Dynamic Effects
• Reduced-form Vector Autoregressive (VAR)
model
Dynamic Effects
Dynamic Effects
Coupons and prices Granger-cause volume
but not vice versa
Dynamic Effects
• Coupons and prices Granger-cause volume but
no vice versa
• Manufacturers’ decisions to coupon are not a
function of previous quantities sold
• Coupons may induce consumers to try new
brands
Conclusion: Why Coupons?
• Strategic interactions between manufacturers
• Incentives given to the people within firms
who make decisions about coupons
• The effects of coupons on repeat purchases