In Defence of Intuition

Should we execute writers of computer viruses? (link via Amit.) I believe that we shouldn’t, but let me dispose of the invalid reason first – the reason that goes “We cannot put a number on human lives!” We make choices about our own lives all the time. Every time I spend a rupee on pleasure, I take it away from my future medical fund, money that could save my life some day. “Putting numbers” on human lives just involves quantifying choices we make routinely, not descending to a new level of moral depravity.

On the other hand, I think there are some good reasons why our intuition in this case is right:

  • Our natural sense of justice demands proportionality. Even if that demand is not justified, it exists, and it will not be changed when faced with cost-benefit analysis. If we execute virus-writers,  most of us will be unhappy, and this unhappiness should in turn be factored into our cost-benefit analysis.
  • In a just society that executes only murders, executions will be rare. The kind of people we execute will tend to be the worst of the lot, people who will usually not have many friends or relatives who will cry over them. On the other hand, a writer of a virus will probably be a functioning member of society and will have people who will be desperately unhappy to see him dead. That should factor into our costs.
  • A society that executes “petty” criminals is a very different kind of society from one that kills only butchers. A society that metes out proportional justice will also be one where the law is generally respected. If you start killing 20-year old college students for petty crimes, people will start considering the law their enemy and disobeying the law to save him. A society where good and bad does not correspond to lawful and unlawful is not a good place to live.

There could be many more. While it is a good idea to subject our moral intuitions to these tests, it is not a good idea to be so eager to accept counterintuitive conclusions just because they are counterintuitive.

7 thoughts on “In Defence of Intuition”

  1. Each one of the reasons you mention is true. Apart from that, there are many why that particular analysis is terribly flawed.

    1) He assumes the utility of extra protection to be a linear function. It is surprising that a trained economist, who should be given to thinking in marginal terms, makes this misleading assumption.

    2) The cost of a computer virus attack is a directly measurable figure, while the cost of human life is taken as a pure behavioural function (what value do you place on the extra safety?). Present value of future production over the working life may have been a better attempt at an estimate that is consistent with the methodology.

    All this is, of course, over and above the central point which Amit already pointed out – deterrence is not the only justification of capital punishment.

  2. I am puzzled by your first point. Where has he made that assumption? The value of a human life has been deciphered from the choices that humans are making right now about their own lives. These choices being made, like all choices, at the margin.

    Your second point – yes, it is a behavioural function, but what’s wrong with that? This is a value that a human puts over his own life. That is the best measure of what we are measuring.

    Present value of future production assumes that the human is a production machine, and the only loss from his death is the loss of what he could have earned from his life. This calculation is indeed more measurable, but it will not measure what we want to measure.

  3. point 1 – value of human life is extrapolated from an extra 1 in 10 millionth chance of safety. from this, the figure of 33 cents is calculated. This is for one execution. It is then compared with 83 cents, which is the average savings provided by one execution of a ‘vermiscripter’. This comparison is true for one murderer’s execution vs one virus writer’s execution. It can be true for all executions in a given time period only if you assume the utility of safety fucntion to be linear, becuase the 33 cents figure is calculated using that. The vermiscripters exceution function will obviously be linear (because it is back calculated using total figures), but the same is not true for execution of murderers.

    point 2 – It is an incomplete behavioural function, for other human beings also place a value on a perosn’s life. (similar to point 2) I am the last person to suggest that a human being be considered a prduction machine – all I’m saying is that if you’re going to effectively calculate the cost of my life to compare it with the cost of computer virus attacks (which are primarily on organizations), it is only logical to first think about the direct, measurable, non-behavioural value that my organization derives from me. I have a feeling that his calculations will get hit with even this much of a change, without even taking into account any behavioural factors(which are definitely paramount).

  4. Ritwik – if not deterrence ( weighted by the intensity of the crime and/or cost on society ) what else is the justification for capital punishment?

  7. I can ( or so I think ) understand the feelings of those who want retributive justice. But the question is whether law should encourage that. Even in the article Ravi quoted, the disagreement with Amit’s views is in the interest of deterrence, right? Also, wouldn’t any punishment more than what is necessary for deterrence ( as per our vague guesstimates ) amount to consciously imposing morality based on the beliefs of part of the population without being of any particular benefit to the society?

Comments are closed.