SAMUEL CAMERON (*)
ABSTRACT. During the period when capital punishment was regularly used in England and Wales, the risk of self-execution from suicide, when suspected of murder, greatly dominated the risk of death at the hands of the state. Over the period 1900-1949, even with four years' data missing, there were 1,540 suicides by those suspected of murder. Using econometric analysis it is found that there is no significant relationship between self-execution and state execution.
THIS PAPER IS CONCERNED WITH a staggering oversight in the literature on capital punishment, found in England and Wales, during the period when it was regularly used: viz. that the risk of self-execution from suicide, when suspected of murder, greatly dominates the risk of death at the hands of the state. Over the period 1900-1949, even with four years' data missing, there were 1,540 suicides by those suspected of murder. Using a simple econometric model, I fail to find any relationship between self-execution and state execution. Self-execution follows suicides in general in showing a positive response to unemployment.
Murder and Suicide as Rational Choices
ECONOMISTS HAVE A TRADITION OF modelling both murder and suicide as rational choices. Becker (1968), in his classic article, was the first since Bentham to open up the field of crime to a rational choice approach. He did this with entirely conventional micro-economics as is common in the Chicagoan invasion of hitherto unexplored areas. He derived a supply of crime function from a conventional subjective expected utility approach to decision making under uncertainty. Crime is the outcome of rational utility maximizing behavior. Hence there are no "criminals" as such. A criminal is simply someone whose portfolio choice contains some activities that society has chosen to designate as illegal. As Chicagoan economists assume similarity in tastes, everyone is equally likely to commit crime if faced with the same constraints. The volume of crime therefore responds to the movements of underlying relative price variables. The relative price variables fall into two categories: those that are directly manipulated by governments in response to crime and those that are outside the control of the criminal justice agencies. The former are generally regarded as deterrents to crime and include the severity of punishment and the likelihood of being caught and punished, which is conditional on the amount of resources devoted to catching and trying criminals. The latter group of variables generally derives from the labor market. Labor market variables such as income and unemployment are taken to measure the opportunity cost of undertaking criminal activity. For example, if expected income in legitimate work is higher then we expect substitution away from risky illegitimate work to be reinforced by the greater loss of income experienced whilst in prison. The labor market will also tend to generate the expected pecuniary return from criminal acts. The more buoyant is the labor market, the greater will be the value of objects in the possession of potential victims of crime.
It is a straightforward matter to extend this approach to capital punishment as first proposed by Ehrlich (1975) in the following equation:
EU = (1-PCON) U(Co) + PCON(1-PE) U(C) + PCON.PE U(C) (1)
where EU = expected utility, PCON = probability of a murder conviction, PE = conditional probability of execution given a murder conviction, U(Co) = utility if not convicted of murder, U(C) = utility if convicted and not executed, U(C) = utility if executed.
There is an extensive literature on the impact of state execution on individual murder supply (see Cameron 1994), little of it using U.K. data with the exception of Wolpin (1977, 1978) and Deadman and Pyle (1993). Ehrlich and subsequent researchers of a Chicagoan …