Notice, first, that with a two-part tariff, it didn't pay for the firm to set its per-unit price above cost. Why not? The reason is that if it sets the per-unit price greater than cost, it immediately creates some dead-weight loss. Some units for which WTP is greater than cost will not be purchased. In other words, some value that could have been captured by the firm is now entirely lost. A two-part tariff allows the firm to reduce the dead-weight loss (and maximize the total value transacted on), but also to capture more value! How does it do this? By using the lump-sum fixed fee to capture the consumer surplus. The result is that, just as with perfect price discrimination, the customer is left with zero surplus, and dead-weight loss is also eliminated. In effect, two-part tariffs highlight a key point about pricing: in setting prices, a firm should try to separate "cost recovery" from value capture. A low per-unit price allows the store to recover its costs, while a high fixed fee allows it to capture surplus. This suggests an interesting possibility: what if the firm's per-unit cost were zero and it only incurred fixed costs? In that case, the firm should set a per-unit price of zero as well and just charge a fixed fee! By the way, that's the reason we often see theme parks (such as Disney World) charge a high entry fee but no charge for each additional ride: the marginal cost of serving an additional ride to a customer, once they are in the park, is close to zero. Subscription pricing also takes this form: pay a subscription and then "eat all you can." (The original paper exploring two-part tariffs was, not surprisingly, about Disney theme parks: it's referred to as the "Disneyland Dilemma" and it's available for reference here. Please note that it is not necessary to purchase this resource in order to proceed in the course.) But How Much Information Do Firms Require? To choose the right two-part tariff in our example, the store needed to know, as before, the customer's WTP for every unit. Of course, this is hard to know in practice. Let's think about the "informational burden" on firms with two-part tariffs more generally. Do firms require more information about customers, or less, to effectively implement two-part tariffs (compared to the case of perfect price discrimination)? ( more or lless )