The paper studies the relative effectiveness of extrinsic monetary disincentives and intrinsic non-monetary disincentives to corruption. In doing so, we also test the Beckarian prediction that at the same level of expected payoff, a low probability of detection with high penalty is a stronger deterrent to corruption than a high probability of detection with low penalty. In Experiment 1, two treatments are designed to study the effect of a low probability of detection with high penalty, and a high probability of detection with low penalty, on bribe taking behavior in a harassment bribery game. In Experiment 2, subjects participate in the same baseline harassment bribery game either without or after having gone through a four-week ethics education program. Results show that a) a low probability of detection with high penalty reduces both the amount and the likelihood of bribe demand, b) a high probability of detection with low penalty has no effect on bribe demand behavior, c) normative appeals of ethics education has a small effect on the likelihood but not on the amount of bribe demand, when measured immediately after the intervention, d) the effect of ethics education vanishes when measured four weeks after the intervention, e) extrinsic monetary intervention, particularly low probability of detection with high penalty, is more effective than normative appeal driven non-monetary intervention that aim to increase intrinsic moral cost, f) analysis of belief about acceptability of bribe demand indicates that the underlying channels through which monetary and non-monetary interventions work are very different.