Executing Justice

by Theodore Dalrymple (February 2026)

On the matter of capital punishment, as on many other questions, I face in two directions at once.

On the one hand, I think that there are some crimes so heinous that no other punishment is appropriate; on the other, that it is so brutal and dehumanising that it should never be employed.

Usually, arguments about capital punishment boil down to its efficacy or lack of it. If it deters murder, it is justified; if it does not, its is unjustified.

But the question of efficacy is far from the end of the discussion. The efficacy of a punishment is not by itself sufficient to justify it, and a very simple thought experiment will demonstrate this. Suppose that the death penalty were imposed for speeding: it would not have to be in every case but for, say, ten per cent of cases, chosen at random—but the sentences carried out. Would you expect fewer or more speeding offences? But even the most fanatical proponents of road safety would hardly approve of such severity, even if it could be proved with reasonable certainty that by imposing it more lives would be saved than lost. This, perhaps, is a refutation of crude utilitarianism.

But what of the opposite case, namely that of the inefficacy of a punishment: would that be sufficient to prove that it was unjustified?

The measure of efficacy and inefficacy depends crucially on what it is that one is trying to achieve. Philosophers have attacked the very notion of punishment because it assumes a philosophical anthropology that is untrue, namely that people can or could do other than what they actually do. (People who believe this usually go on to claim that punishment never achieves its aims, but this empirical claim is only secondary to their argument.)

It is difficult for anyone to be a thoroughgoing determinist: for if it were true that criminals could not act other than as they do, it would also be true that judges, lawyers, jurors and hangmen (or should it be hangpersons?) could not act other than as they do: unless, that is, humanity is divided into two categories, the kind of people who cannot help what they do, and those who can perfectly well help what they do. No idea could be more elitist, in the worst possible sense, than this.

The aim of punishment, however, is complex and even shifting, as therefore must be the estimation of its efficacy. In one situation, it may be mainly deterrent, in another declarative, a signal of social disapproval or that justice has been done. Those who deny that there should be punishment deny that there should be justice, denying in effect that anyone deserves anything. But if there are no just deserts, it cannot be that it is unjust for someone to get their just deserts, for the notion of injustice is parasitic on that of justice. It is therefore paradoxical (as well as empirically absurd) to argue that punishment is inherently wrong or unjustified.

The replacement of punishment by treatment, as if wrongdoing were a kind of illness, or illness-like phenomenon, does not have the liberal or lenient consequences that many people might suppose that it has, for it is as compatible with the most revolting severity as with the softest of mansuetude.

Let us suppose that a murder has been committed and that the authorities a) cannot find the murderer and b) have reason to suppose that it is better for the solidarity of society that a perpetrator be found and punished, even if he be only a scapegoat. So long as he is a plausible scapegoat, and no one discovers that in fact he is innocent, it is beneficial for society to punish him as if he were guilty. We should surely be horrified by this ruthless utilitarian calculation, and our objection to it would be that it is obscenely unjust. Justice is certainly not all-important, but it is important nonetheless.

Let us now turn to the contentious question of whether the death penalty is effective in deterring murder. People’s views on this question tend not to be strictly proportional to the evidence in their favour: they tend to become extremely passionate almost immediately. And it ought always to be borne in mind that the correct answer, of there is one, is dependent on the historical time and circumstances of the evidence. One would not expect the same answer in Norway and Albania or Pakistan.

That said, I discovered evidence that for Britain in the twentieth century the death penalty (which of course was not carried out in every case in which it was passed) had a deterrent effect on potential murderers.

In the forty years before the abolition of the death penalty, about a third of homicides were committed by people who then committed suicide. Most of these people were disturbed psychologically or psychiatrically, and therefore were not the kind of people to take careful note of any deterrent. Therefore, I hypothesised that, if the death penalty exerted any deterrent effect, the total number of homicides would rise but the number of homicides followed by suicide would remain more or less the same—and therefore the proportion fall.

My friend, the criminologist Mr David Fraser, kindly looked at the statistics for me and found that the hypothesis was borne out—albeit with a slight delay. This delay was easily explicable, I think, by an admittedly ad hoc additional hypothesis, that the death penalty had so long been a part of the national consciousness that it took a little while for it to be realised, not only factually but emotionally, that the penalty for murder could no longer be death.

I should also point out that after the penalty was abolished, the methods of surgery and resuscitation improved greatly, so that those who would have died of a violent attack before the abolition no longer did so. Not surprisingly, then, the number of such attacks climbed faster than the homicide rate itself. People were less frightened that their extreme violence would end in the death of their victim; and they realised that, if it did, the penalty would be some years in prison (with which a good proportion were already familiar) rather than death.

What, then, is to be said against the death penalty? A very important argument is that in all jurisdictions, no matter how scrupulous, mistakes have been made and the innocent put to death. In the last fifty years there have been about two hundred exonerations of condemned prisoners on Death Row in the United States in the light of new evidence, in general after about ten years of such incarceration. This is more than ten per cent of all those so condemned. It would be a foolhardy person who was certain that all the executions that were carried out were of the guilty.

For a state to put to death the innocent is to undermine the legitimacy of its criminal justice system. Of course, no such system can never be mistaken, since the possibility of error is conterminous with any conceivable system; but I need hardly point out that there is something rather final, irreversible about death as a mistake.

I have heard the following thought experiment tried. Suppose that of a thousand murderers put to death, five of them, in the event, are wrongly executed. Suppose also that, of a thousand murderers not put to death, ten go on to murder a second time. If that thousand had been put to death, ten people would not have been murdered. Thus, there would have been a net gain of five innocent lives by executing a thousand (or nine hundred and ninety-five) guilty people.

Quite apart from the inherent uncertainty of such a utilitarian calculation, its disregard of justice would strike most of us as horrific.

Another argument in favour of the death penalty is economic. To keep prisoners in modern states is fantastically expensive, vastly more so than it was in years gone by, no doubt because prison conditions have been greatly improved, but also because administrative and legal costs have had an inexorable tendency to rise.

Theoretically, the death penalty would cut the expenses: a dead man does not cost much to keep. But the condemned man who is subsequently executed in America today spends on average 22 years on Death Row (in 1984 it was 6 years), more than most murderers spend in prison in other jurisdictions, with lower murder rates. Taking into account the number of legal procedures which usually have to be gone through before a man is executed—a sign, incidentally, of a lack of faith in the competence or probity of lower courts—the expense incurred by the penalty of death cannot be very much lower than the penalty of life in prison without parole. Proponents of the death penalty might reply that they want these legal obstacles, which benefit lawyers as much as they benefit perpetrators, to be simplified or swept away, but that is easier said than done. (In Victorian times in England, the time between condemnation to death for murder and execution was three weeks, and it is far from certain that there were more miscarriages of justice.)

At any rate, to keep a man on Death Row for 22 years seems barbaric, though things in Japan (another democracy with the death penalty) are even worse. There, apparently, the condemned man may spend many years on Death Row and is never informed of the date of his execution. One day, at eight o’clock in the morning—without warning to himself or to anyone else—he is taken and hanged. If he survives eight o’clock in the morning, he is safe for the day, but not for the following day: and this can last for years. But Japan carries out very few executions.

A novel recently published in France by Constance Debré, the granddaughter of a former French Prime Minister, titled Protocoles, is in part the result of her lengthy research into the various methods used in the United States to put people to death. The larger implication of the book is that we increasingly go through life according to protocols and not according to real thought, morality or feelings.

But her account of the incompetence and cruelty that she found—all in official reports and documents—is startling. The best and surest method was by firing squad, but for some reason this was the least used. Since doctors and nurses now refuse to have any part in executions (though doctors certify death), death by lethal injection is often long drawn out. Death by nitrogen mask, tried out on humans because it was illegal to try it out on humans (or so says the author) sounds particularly horrible, and death by Zyclon B (cyanide gas) not very much better. She describes cases in which the convicted have waited so long for their execution that they have to be wheeled, now suffering from paralysis or some such, into the execution chamber.

In the prison in England in which I used to work, in the days of capital punishment, the doctor had to certify the condemned man fit for execution—a rather curious notion, with its implication that it was the executed man, not society, that was supposed to learn something from his forced demise. I would not have wanted to participate in the ceremonies of execution.

And yet, and yet … If someone were to express outrage that Rudolf Höss, for example, the former commandant of Auschwitz who knowingly, and on his own account, oversaw the murder of 3.5 million human beings, had been put to death, I should think him of a very peculiar moral sensibility indeed. And this drives a coach and horses through the deontological objection to the death penalty.

Persons such as Höss are fortunately, for one reason or another, not very frequently encountered, but if it is right that he should have been executed, the question then becomes at what level of evil should someone be executed. In pursuit of another literary and philosophical project, I recently read widely in the story of Joseph Vacher, the Jack the Ripper of the south-east of France, who towards the end of the nineteenth century raped, murdered and eviscerated at least eleven people, mainly young shepherds and shepherdesses, and possibly up to thirty such. Was he sane? Was he mad? No one can say: but who would lament that he fell victim to the guillotine on the grounds that the death penalty was wrong, deontologically wrong?

Table of Contents

Theodore Dalrymple’s latest books are Neither Trumpets nor Violins (with Kenneth Francis and Samuel Hux) and Ramses: A Memoir from New English Review Press.

Follow NER on Twitter @NERIconoclast