The post office (PO) responded by analyzing the methods of attack, concluding that existing anti-bandit counter screens were inadequate and security procedures were deficient. The PO embarked on a prevention initiative that included upgrading of screens and staff training. In 1982, the robbery rate at sub-post offices peaked at 266, and then steadily declined to 121 in 1985. An analysis of the effectiveness of the initiative indicates that, with a wide margin of error, it cut the rate of successful robberies by about 40 percent of the level expected. However, when failed robbery attempts are included, the preventive effect was smaller in proportional terms, but larger in absolute terms (incidents prevented per year). This pattern of effects can be explained in terms of displacement effects. The method of robbery attacks changed from one involving removal of screens to increased use of firearm threats at the screens, but this latter method was more likely to result in a failed robbery attempt than the previous method. There also may have been some displacement from robberies inside the facilities to attacks on deliveries. However, these displacement effects were not sufficient to nullify the preventive gains of the initiative. 1 figure and 8 references.