When talking to people in favour of the death penalty, a common argument one hears is that it shouldn’t be common, but a possibility for the “rarest of rare” scenarios. What is the rarest of rare? To be true to definition, it is impossible for there to ever be a “rarest of rare” scenario because it could ALWAYS have been rarer. It becomes apparent that this “rarest of rare” is awfully illusory.
Now, to those who accept that by “rarest of rare” they don’t literally mean the rarest case, but mean cases where humans have crossed the threshold and done something unimaginably inhumane (an interesting term in itself, but that’s a topic for another day), I would ask: what is that threshold? 1 premeditated murder? 5? 100? Another issue arises. Even if someone managed, in reptilian fashion, to provide a number, I would ask: why that number? The category does not have clearly defined boundaries. Human wrongdoing exists on a continuous and multidimensional spectrum. This makes a single universal threshold exceedingly difficult to define.
The problem is not merely that the line is extraordinarily hard to draw, but that there is no obvious reason that the line should exist where the state draws it.
Let us entertain the idea that this isn’t really a problem. One could point out that laws need not account for every conceivable form of human behaviour; they merely need to establish a reasonable threshold. We do already do this in countless areas of criminal law - a drinking age, a blood-alcohol limit, sentencing guidelines. Arbitrariness at the margin is something that the law tolerates everywhere. However, it tolerates that arbitrariness because the cost of landing on the wrong side of an ordinary line is bounded and correctable. The death penalty, on the other hand, is unique. It is irreversible.
A prison sentence can be corrected.
A death sentence cannot.
This is why the threshold problem, which might otherwise be dismissed as a feature of law in general, becomes disqualifying here: an arbitrary line is only tolerable when being wrong about it can be fixed. Once the consequence of the line is death, arbitrariness is no longer just a tolerable inconvenience, but the problem.
This logic applies even more sharply when we talk about certainty. Even if we were somehow able to establish a threshold, another problem remains.
How certain must the state be before it acquires the right to kill someone?
99%? 99.9%? 99.999%?
There is no percentage point at which human institutions become infallible.
If the state cannot be perfectly certain that it has identified the right person, and execution is irreversible, then the burden of proof becomes extraordinarily high. If life imprisonment can protect society without creating the possibility of an extraordinarily morally costly mistake, why would the state ever choose the irreversible option?