Bars Not Dead Yet
New York Times
While it is hardly a new idea that bars’ can have both information and direct benefit network effects, these effects have now been amplified by new technologies particularly at Cornell. In this article, the writer chronicles the nights of several students whose choices are dictated principally by network based influences. It can be assumed that they could go to any of the the main bars in college town however they have prior information that informs them they will have free drinks at the bar with the mixer. There is then a direct benefit incentive to go to a different bar when they find out that their friends are at a different bar. These form a rather simple dynamic in this case. The way to save bars in Collegetown is to rely on the same network effects that fraternities employ to attract people to parties and apply them to bars.
Here at Cornell with the current rules, freshmen have had very limited access to parties. As a result many fraternities have to be extremely careful about who to invite and how to control the party while still having a good time. Usually when a freshman hears about a party they first look at the information that may sway him or her toward a given party. A good party may have one of several factors: hard alcohol, dancing, or attractive members of the opposite sex. Instead of random events acting as signals, there is usually a series of text messages that goes out. The probability of a good factor given a high signal is often unpredictable because this assumes truthfulness of college students. A cascade can occur if two people receive what they perceive as good signals. These two people may then describe party to others as a potential “good” party. They will then pass along the information concerning the “good party” to all others who will then accept the information having only the knowledge of the decision of others and not the signals.
The direct benefit effect is also at play in two distinct ways. The first is the total number of people at a party. Generally a poorly attended party of a few people will not be “fun” so if people think that more people are going to any one party then that party will have an increased attraction; however, due to overcrowding every party reaches a point where the direct benefit effect is actually negative because the more people at the party decreases the attraction of the party. There is also a more acute direct benefit effect that one party may suddenly become much more attractive to any one person if another specific person is at the party. For example, if John is at a party on West Campus and Lisa is at a party on North, John may be willing to traverse campus to find Lisa regardless of the current attractiveness of his party. Lisa may also be a person of particular network influence and her presence at one part or another may have a greater effect on the total attractiveness of a party.
The bars of Collegetown can now use the information effects and direct benefits that parties use to increase profitability. Many of the bars in Collegetown have had similar specials for decades. These specials no longer have significant appeal because they seem routine. A new special could cause an information cascade if enough people were informed. The probability of a good factor given a high signal at a bar is very high given that businesses cannot lie in advertising which can help attract students away from parties and towards bars. Bars can take advantage of the direct benefit effect by trying to attract specific members of the community who have a high degree of influence by making some sort incentive for “preferred customers” to return increasing the attractiveness for several influencing individuals. The bars in Collegetown need not be endangered. Instead they must adapt and rely on network effects to attract patrons.
gdh56