Bertrand Differentiated Products
Setup
- Many partial (but imperfect) substitute products with market power
- Akin to brand name drug (i.e., with patent protection) with some substitutes
- Simplify to 2 drugs from two companies:
- Denote demand for drug 1 by \(q_{1}(p_{1},p_{2})\)
- Denote demand for drug 2 by \(q_{2}(p_{1},p_{2})\)
- \(\frac{d q_{i}}{d p_{j}} > 0\), so that products are substitutes
- \(\frac{d q_{i}}{d p_{i}} < 0\), as usual (downward sloping demand)
Specific Bertrand Setup
- Players: Firm 1 and Firm 2
- Actions: Each firm chooses a price \(p_{i}\) for its product.
- Demand: The demand for firm \(i\)’s product: \[q_{i} = a - bp_{i} + cp_{j}\]
- Profits: Each firm’s profit: \[\pi_{i} = (p_{i} - mc_{i})q_{i}\]
Key Points:
- \(a\) is the maximum quantity when the price is zero
- \(b\) represents how quantity demanded changes with the product’s price
- \(c\) captures the effect of the rival firm’s price on demand
Best Response Functions
Firm 1: \[p_{1}^*(p_{2}) = \frac{a + cp_{2} + b\,mc_{1}}{2b}\] Firm 2: \[p_{2}^*(p_{1}) = \frac{a + cp_{1} + b\,mc_{2}}{2b}\]
Key Points:
- Each firm’s price is a function of the other firm’s price
- Best response functions represent the optimal price choice in response to the rival’s price
Nash Equilibrium
Equating the two best response functions to solve for equilibrium prices:
\[\begin{align*} p_{1}^* &= \frac{a(2b+c) + 2b^2\,mc_{1} + bc\,mc_{2}}{4b^2 - c^2} \\
p_{2}^* &= \frac{a(2b+c) + 2b^2\,mc_{2} + bc\,mc_{1}}{4b^2 - c^2} \end{align*}\]
Key Points:
- Nash equilibrium prices are where neither firm has an incentive to change its price unilaterally
- Both firms maximize their profit given the other firm’s price
Comparison to Monopoly Pricing
Monopoly Price (previously derived): \[p_{m} = \frac{a + b\,mc_{m}}{2b}\]
Comparison:
- Classical Monopoly vs. Competition: Monopolists set higher prices than competitive firms
- Differentiation: In Bertrand, differentiation can lead to prices closer to monopoly levels, depending on the substitution effect (parameter \(c\))
- Cost Structures: Bertrand prices depend on both firms’ costs; monopoly price only depends on its own cost
Bertrand Pricing with Insurance
- Stage 1: Insurer chooses one drug with a copayment to be on formulary. Other drug will not be covered at all
- Stage 2: Bertrand pricing with “bids” to insurer
- Stage 3: Insurer selects drug to be on formulary
Solution Idea
- Imagine firm 1 set price to original bertrand price, \(p_{1}^{b}\)
- Firm 2 can then set price to \(p_{2}^{b}=p_{1}^{b}-\epsilon\) and get most or all of the market
- So, \(p_{1}^{b}, p_{2}^{b}\) is not an equilibrium
- Firms lower price until extra sales on formulary are less than profit of drug off formulary
- Final prediction: Insurers used tiered formularies and copayments for differentiated (but substitutable) on-patent drugs
Summary from joint hearings in FTC and DOJ
Unique drugs with patent protection:
Under these circumstances, there is little opportunity for a purchaser to stimulate competition among manufacturers. Manufacturers are roughly free to set launch prices, they rarely discount those prices, and purchasers are price takers.
Differentiated drugs with imperfect substitutes:
This stage represents the lion’s share of the market at any given time…Depending on how similar the drugs are…organized purchasers have the ability to either switch patients in a medically appropriate way…or at least start new patients on a preferred drug…This is the area where formularies can be applied for the greatest effect on overall costs.