[284]In order to qualify as a permissible limitation in terms of section 33 the State must therefore establish that the invasions on the right to life, the right to be protected from unfair discrimination, the right to dignity and the right to be protected from cruel, inhuman or degrading punishment, which the application of the death penalty causes, satisfy at least the three separate elements specified in sections 33(1)(a)(i), (ii) and 33(1)(b). In the case of a limitation on the right to dignity and the right to be protected from cruel, inhuman or degrading punishment, the fourth element of "necessity" contained in section 33(1)(aa) must further be satisfied.
[285]The most plausible argument in support of the submission that the death penalty does satisfy these onerous conditions prescribed by section 33 is the submission that it acts as a deterrent. That argument has dominated perceptions in support of the death penalty, both in South Africa and abroad.
[286]It must readily be conceded that if it could be established that the death sentence does indeed deter the commission of serious offences in respect of which the death penalty is a competent sentence, it would indeed be a very relevant and at least a potentially persuasive consideration in support of its justification in terms of section 33. There are, however, some serious difficulties involved in the acceptance of the proposition that the death penalty is, or ever has been, a demonstrable deterrence.
[287]The legitimacy of the argument must to a substantial degree be premised on an assumption which appears to me to be fallacious and at the least, highly speculative and rationally unconvincing. That assumption is that a criminal, contemplating the commission of a serious offence, weighs the risk that he might be sentenced to death against the risk that he might not be sentenced to death but only to a long term of imprisonment of twenty years or more. The assumption is that he would decide to commit the offence even at the risk of receiving a long term of imprisonment but that if the death sentence was the risk, he would refrain from committing the offence at all. I have serious difficulties with these assumptions. In the first place they are not supported by any empirical evidence or research in this country or abroad. Secondly, this argument attributes to the offender a capacity for reflection and contemplation and a maturity of analysis which appears to me to be unrealistic. Thirdly, and more fundamentally, it ignores what is possibly the real factor in any risk assessment which might activate a potentially serious offender: the risk which he considers is that he will not be caught. If he believed that there was a real risk of being apprehended, charged and convicted he would not willingly assume the prospect of many years of quite punishing imprisonment.
[288]If, as I believe, such offenders commit the crimes contemplated because of a belief that they will probably not be apprehended at all, it is a belief which is regrettably justified. On the information that was common cause in argument before us, sixty or seventy percent of offenders who commit serious crimes are not apprehended at all and a substantial proportion of those who are, are never convicted. The risk is therefore worth taking, not because the death penalty would, in the perception of the offender, not be imposed but because no punishment is likely to result at all. The levels of serious crimes committed in South Africa are indeed disturbing. For many in the community, life has become dangerous and intolerable. Criminals do need vigorously to be deterred from conduct which endangers the security and freedom of citizens to a very distressing degree but, on the available evidence, it is facile to assume that the retention of the death penalty will
