Transcript Empirical Regularities in Pricing

Strategic Pricing:
Theory, Practice and Policy
Professor John W. Mayo
[email protected]
Raise your prices!!!!
• ….Ugh?
• Identify where you can do a better job than
competitors of meeting customer needs
• Shape products to dominate these segments
• Manage cost and prices to maximize profit
2
• Identify value opportunities
• How can we modify our products to create value to
customers
• Set priorities
• To be a candidate for “performance pricing” product
had to have strong competitive position
• Align Price and Value
• Document benefits from your product that accrue to
customer
• Get Cooperation
• Communicate value to customer
3
Empirical Regularities in Pricing
• Hosken and Reiffen (RAND, 2004) - Most prices stay at their modal
price most of the time. Deviations are asymmetrical, with more down
than upward pricing periods.
• Kashyap (QJE, 1995) – nominal prices of catalog items often remain
fixed for more than one year. Inflation is associated with more
frequent price changes. Changes in price, when they do occur, are
disparate.
• Blinder (AER, 1991) Survey. If we had to pick a single abstraction to
represent the whole economy, annual price changes (the median
response) would appear to be the “right” model
• Chevalier, et al (AER, 2003) – Behavior of prices in peak demand
periods
Source: Blinder (1991)
Source: Blinder (2001)
Hosken and Reiffen
• Uses data from CPI surveys
• 8400 items, 5 years, 20 product categories
• Most products have a “regular” (high modal) price
• When not at regular price p*, p<p*
• But most price reductions are temporary
• What is the strategic implication of this?
Prices and Demand
Chevalier, et al (AER 2003)
• How do prices change with demand
changes?
• Competitive market theory suggests that price
changes are non-negative to demand changes
Competition, Prices and Demand Changes
What happens to firm
price when market demand
Increases?
$/q
D0
D1
S
P1
P0
Q0
Q1
q
• But Imperfectly Competitive Models of Price
responsiveness to demand suggest the
possibility of counter-cyclical pricing
Alternative models of counter-cyclical pricing
• Because of fixed costs of search, consumers will
be more responsive to prices during peak
demand periods [ε increases]
• Multi-round pricing models of tacit collusion
suggest the probability of defection is highest in
peak demand periods
• “Loss-leader” models in which retailers compete
for consumers via advertised prices
Chevalier, Kashyap and Rossi (2003)
• Examine retail prices of grocery store chain
in Chicago, 1989-1997
• Data on retail transaction price, wholesale
price, quantities and advertising, available
at website
• Seek to differentiate among theories of
counter-cyclical pricing
Methodology and Findings
• Dominick’s Finer Foods - Items chosen include
those with seasonal demands and nonseasonal product categories
• Examine prices and their responsiveness to
specific peaks demands [e.g., beer on July 4th,
Tuna in Lent]
• Find that prices for season items
idiosyncratically fall with peak demands, NOT
general demand elevations. HANDOUT
• Conclusion: Loss-leader models best explain
pricing patterns
Pricing and Costs
• How, in practice, do cost changes affect prices?
• Consider the impact of minimum wage changes on
prices in retail food outlets.
• BLS field personnel collected prices during 19951997 for nearly 7,500 food items at over 1,000
different outlets.
• Both “Limited Service” (LS) and “Full Service” (FS)
outlets were surveyed
14
Retail Price changes at Retail Restaurants
Percentage of items with unchanged price, bimonthly
1995–97
15
(Source: MacDonald and Aaronsen, Am. J. Ag. Econ. 2006)
Minimum Wages and Retail prices
• Aug 1996, President Clinton signed bill enacting
minimum wage increases
• Oct, 1, 1996 ($4.25 to $4. 75)
• September 1, 1997 ($4.75 to $5.15)
• Both geographic and industry differences in
impact
• Limited Service (LS) v. Full Service (FS)
• Some states have own supra-federal minimum wage
• Hawaii, Massachusetts, New Jersey, Oregon, Vermont, and
Washington. Alaska, Connecticut, and the District of Columbia are
constant % above federal.
16
Empirical Findings
• Price Clusters and the Cluster of Price Changes
• 1/8 of LS item prices ends in 99, >30% end in 9
• Twelve and three times those that would appear by
random chance if all endings were equally likely.
• 25% of all LS price changes are 10 cents
• 50% of LS price change are 5, 10, 20, and 30 cents.
•
• Price Lags and Leads
• Both leads and lags occur
17
Magnitude of Price Changes
• Regression model
• ΔP = f (PPI, MEALTYPE ,Pup, Pdown, MW)
• Estimates: 1.56 % increase for 10% increase in
minimum wage
• Payroll = 25 % of sales in LS restaurants
• Expected pass-through?
• The Distribution of price changes
• Not across the board changes
• Outlets most often change prices of a few items
• Which items?
18
Changing Market Structure and Prices:
Northwest/Republic Merger
Markup Relative to industry average prices (%)
40
NW or Rep
NW + REP
30
20
NW or Republic + Other
10
NW +Rep + Other
1985
1986
1987
Source: S. Borenstein, AER, May 1990
Pricing in online markets
• Suppose the price you pay to acquire a good for
sales is $50.
• A comparison site charges $.50 per click
• You have other incremental costs of $3 per order
shipped
• The comparison site has a conversion rate of 5%
• Prior experimentation indicates that your firm’s
product has a price elasticity of -2.
• What price do you charge?
Pricing in online markets
• Good news & bad news
• Bad news: consumer search costs fall
dramatically (optimal consumer search
increases;, markets, product lives, prices move
very rapidly
• Good news:
• “menu costs” of changing prices are reduced
• changing prices/tactics can occur more quickly
• Analysis: must be quicker
• (Baye recommends Cournot (P – MC)/P = 1/(nε) rule of
thumb