Price discrimination occurs when a firm charges different prices to different customers for the same product, based on their willingness to pay. First-degree (perfect) price discrimination extracts all consumer surplus, eliminates deadweight loss, but transfers all gains to the producer. Second-degree discrimination uses quantity discounts or versioning; third-degree separates markets by observable characteristics (e.g., student discounts). Price discrimination requires market power, the ability to prevent resale, and identifiable differences in willingness to pay.
Compare the profit and welfare outcomes of standard monopoly pricing vs. each type of price discrimination graphically. The key insight is that first-degree discrimination is efficient (no DWL) but maximally inequitable.
A standard monopolist faces an inherent dilemma: any single price leaves money on the table. Charge high and you lose price-sensitive buyers who would have paid something. Charge low and you give a discount to buyers who would have happily paid more. From your study of consumer surplus, you know this gap — the difference between willingness to pay and the actual price — is a transfer from the monopolist to buyers. Price discrimination is the strategy of recovering that lost revenue by charging different prices to different buyers based on their willingness to pay. It requires market power (otherwise competitors undercut the price differences), the ability to prevent resale between groups, and some way to identify or induce separation of buyer types.
First-degree (perfect) price discrimination is the theoretical extreme: the firm knows every buyer's exact reservation price and charges each one precisely that amount. The demand curve and the marginal revenue curve become identical, because each unit is sold at its own maximum price — there is no need to lower the price on prior units to sell more. The result is efficient: every unit where value exceeds cost is sold, so deadweight loss disappears entirely. But every dollar of consumer surplus is extracted. The social total surplus equals the competitive outcome; it just flows entirely to the producer. Perfect price discrimination is rarely achievable because individual willingness to pay is unobservable, though personalized algorithms and targeted pricing are increasingly close approximations.
Second-degree discrimination doesn't require knowing individual buyer types — instead it uses self-selection mechanisms that induce buyers to reveal their type through their choices. Quantity discounts (bulk pricing), versioning (economy vs. business class, software editions), and tiered subscription menus are all examples. The firm designs a menu of options where each consumer type prefers the option targeted at them, not the option meant for someone else. This is an information design problem: the firm must ensure the higher-valuation option isn't so attractive that low-valuation buyers choose it, while keeping the lower-valuation option genuinely appealing to low-valuation buyers. The firm does not need to identify who you are; your choices do the identification for it.
Third-degree discrimination uses observable group characteristics — student status, age, geography, time of purchase — to segment markets directly. The profit-maximizing rule in each segment is the same: set MR equal to MC within that segment. Since MR = P(1 − 1/|ε|), the segment with less elastic demand faces a higher price. Airlines charging business travelers more than leisure travelers, and software companies charging educational institutions less than corporations, both follow this logic: the price-sensitive segment gets a lower price, the inelastic segment pays more, and total profit exceeds what any single price could achieve. The efficiency consequences depend on whether the discrimination expands total output; sometimes it does (by serving price-sensitive buyers who would be priced out under uniform monopoly pricing), sometimes it does not.