The Royal Prerogative of Mercy is an ancient right traditionally bestowed upon sovereigns, and vested in the common law. In Canada, this right as applied to murderers was preserved in the criminal code. The code did not specify criteria under which clemency might be granted, nor were the authorities required to explain or defend its use. Prior to the last legal executions in Canada (1962), it appears that the Royal Prerogative was utilized on a case-by-case basis; between 1963 and the de jure end of capital punishment in Canada in 1976, it was used by the various governments to impose a de facto policy of abolition.’ The primary purpose of this study is to examine the historical use in Canada of the Royal Prerogative in the context of different theories of punishment. Two theories in particular invite interest: retributive theory, and the economic theory of punishment. According to John Rawls,’ retribution rests on the simple proposition that wrong-doing merits punishment. Two principal tenets are that the offender should suffer in proportion to his guilt, and that the severity of the punishment depends upon the depravity of the act. The costs imposed upon the convicted offender should approximate the direct costs of the crime. The intention of providing a deterrent is not a purpose of punishment in retributive theory per se. Thus a justification for capital punishment is provided for those who do not believe, or are not sure, that a deterrent effect exists.’ The economic theory of punishment is one part of the economic model of crime. Society is assumed to attempt to minimize the social cost of crime, which includes the direct harm of offenses, plus the indirect costs which include the expense of a protection system (police, prisons, etc.). Crime is viewed as an externality-generating activity, and, consistent with the economic theory of externalities, the state attempts to induce the offender to internalize the externality by threatening to impose a cost. Central to the model is an incentive response, i.e., the assumption that some prospective offenders might be deterred from their felonious enterprises by an appropriate increase in the perceived expected cost of punishment. The economic model, then, is essentially forward-looking, whereas the retributive model, which examines the particulars of individual crimes to determine moral culpability, is backward-looking. These two views of punishment suggest some differences in the organization of society’s legal system, as well as in its actual administration. With respect to capital punishment, an ungraded definition of murder with a centralized commutation body is particularly desirable from the economic standpoint. This gives the authorities maximum flexibility to respond to a change in some exogenous shock to the system. To be more concrete, suppose that the optimal response is perceived to be an increase in the proportion of murderers who are executed.4