2: Fundamentals of Price Discrimination
He then provides a Marketing 101 refresher on discrimination. A company may charge the same price to all customers (no discrimination), charge different amounts to different groups (third degree), create versions of a product to appeal to different segments (second degree) and individualized pricing for each customer (first degree). Essentially this is an attempt to capture more of what the market will pay (as revenue is lost when a customer pays less than he might have been willing). However, the effort of discriminating may exceed the additional revenue, and customers are often upset when they learn someone else paid a lower price.
Auctioning is an example of first-degree discrimination that is relatively easy to conduct: particularly when there are multiple similar items up for auction, the customers who are willing to pay the highest prices outbid the rest. The person who bid the most is not allowed to renege on his offer when the second "winner" pays a lower price. Essentially, all price discrimination is based on auction - though for the other kinds, the seller imagines what people might bid rather than allowing them to actually bid in an event.
He returns to second-degree price discrimination, creating a deluxe (Midas) version, a standard (Atlas) version, and an economy (Hermes) version to capture as much revenue from the three basic market segments. He mentioned this in his preface and says he'll have more to say later.
Back to third-degree price discrimination again, any distinction can fall into this category. If a product is sold at different prices in different locations, it's third-degree discrimination. Likewise, if groups of people get special offers (contractors get cheaper building supplies, teachers are given a group discount on school supplies, members of a certain club get a discount, business travelers pay more than leisure, etc.) This often seems like the company is being generous in discounting to a specific group, but in truth the group would be less likely than most people to purchase at the standard price.
Legal Issues
In mixed economies, governments intervene on behalf of customers to prevent sellers from conspiring to dominate the market, playing upon fear of monopoly, cartel, and oligarchy. Such things are not unheard of, but are extremely rare in free markets because there is little incentive for competing firms to collude - but there is nonetheless fear of this occurring.
One result of this is government intervention to prevent price discrimination at all: to insist that it is only "fair" for all customers to pay the same price for a given price, regardless of the cost of serving them. The problem with legislation of this nature is that it makes products less available - if a firm is unable to profitably serve a market (because shipping costs are too high, because selling to individuals is more costly than selling in bulk, etc.) then it chooses not to do business in that market at all and customers receive no service. Likewise, when a firm is forbidden to give a discount to favored groups (teachers, the elderly, etc.) the public is then outraged that sellers are forbidden to discriminate.
While certain legislation (notably the Robinson-Patman Act) have attempted to prevent discrimination, it has invariably been repealed, amended, or simply not enforced by the Federal Trade Commission and the Supreme Court has been notably "skeptical" of such legislation. As a result, sellers are still free to discriminate - but must be prepared to defend themselves in case any of this legislation is ever enforced by having a plausible justification for differentiating price.
Loose Bits
The author mentions eleasticity. An elastic product will experience a significant change in quantity demanded when there is a small change in price, and an elastic one will experience no significant change in demand when price is altered. This is plotted on a demand curve that typically shows that the lower the price, the greater quantity will be demanded.
Another stray topic is collaborative filtering, which is based on the assumption that people of similar qualities like similar products, so the behavior of some can be used to identify selling opportunities that will be attractive to others. Amazon currently does this extensively by analyzing shopping carts and making recommendations based on buyer behavior (people who bought this item also bought certain other items).
Another stray topic is bundling: offering customers special incentives to purchase multiple products at once (EN: there is also the black-hat version in which customers simply are not given the option and must buy them in a package, with no option to buy one without buying the others). Generally, this uses the strength of one item in the bundle in order to get the customer to accept less desirable items in addition.