This paper argues that if the main purpose of punitive damages in product liability cases is to deter conduct that results in unsafe consumer products, the threat of punitive damages ought to be better targeted at the conduct of senior management, rather than the misconduct of employees. In most jurisdictions, punitive damages may be imposed on a firm for misconduct by employees who are not part of senior management. In a sizable number of these cases, punitive damages are imposed even though the employees whose conduct led to litigation were acting contrary to firm policy. Although the imposition of compensatory liability is clearly appropriate when a firm's products injure consumers, allowing punitive damages for conduct contrary to management's direction dilutes management's incentive to exercise appropriate control over the behavior of employees. Management can't guarantee that every employee will make appropriate decisions regarding safety issues, but it can establish well-designed policies and procedures to guide these decisions. A standard designed to influence the behavior of management should provide an incentive for management to take actions within its power to assure that its firm's products are safe. Building on concepts enunciated in a recent Supreme Court case involving employment discrimination, this paper argues that the deterrent effect of punitive damages would be enhanced by a rule that provides an affirmative defense against punitive damage claims based on reckless conduct of employees, when that conduct is contrary to a firm's good-faith efforts to employ well-designed safety policies and procedures. This article further argues that the so-called "government standards" defense against punitive damages should be defined in a way that both promotes compliance with government standards and helps to assure that the standards provide adequate public safety. This goal could be accomplished by providing that compliance with an applicable federal safety standard constitutes prima facie evidence of good faith and shifts the burden of proof to the plaintiff to show, by clear and convincing evidence that either (1) the defendant was guilty of fraud on a regulatory agency; (2) clear advances in the state of the art, implemented after the standard had been adopted but before the defendant's product was designed, had rendered the product obsolete; or (3) the standard itself reflects a flagrant indifference to safety.