Kremer and Levy (2008) analyze the extent to which college students who consume alcohol influence their peers. College students could influence the alcohol consumption of their peers, which in turn could lead to harmful effects. For example, peers could potentially disrupt classes, be exposed to diseases and lower grade point averages, start binge drinking, abuse illegal drugs, and even die. These effects could generate multiplier effects in the future. Previous studies have examined problem drinking and negative peer effects among college students. Sacerdote (2001) examines peer effects in universities. He finds evidence to support this theory that students whose roommates reported high beer consumption were more likely to replicate the same behavior. Kremer and Levy (2003) noted that peers who had a roommate who drank could potentially increase the peer's preference for drinking alcohol. The Cooperative Institutional Research Program's (CIRP) Entering Student Survey demonstrated declining averages among peers whose roommates drank alcohol frequently, especially those who had attended high school. The survey classified males as particularly sensitive to peer influences. Empirical data presented by Wechsler, Lee, Kuo & Lee (2000) shows that alcohol consumption and abuse reported by 40% of college students had engaged in binge drinking at least once in the past two weeks and it was concluded that drinking of alcohol, and more specifically binge drinking is a social influence from peers. There are various economic theories presented in this article. For example, in the context of academic performance, hypotheses made are that peers might influence the preferences of others, since seeing their friends consume alcohol also stimulates the desire in the other student to consume alcohol. Many theoretical models assume that students' academic achievement is a linear function of their peers' ability to also influence their behavior. The theories presented consist of peer effects on influencing preferences, which leads to lower academic performance because the student's study is interrupted. The hypothesis that students who are frequent drinkers housed with another frequent drinker would increase their alcohol consumption and may lead to binge drinking based on variables, model preferences, and theoretical models that assume the student's academic achievement is a function linear influence of his peers, and multiplier effects. Kremer and Levy find from empirical work that some individuals are more vulnerable than others to the influence of increasing alcohol consumption. Increased alcohol consumption is most commonly linked to the desire for social acceptability.