On Luigi Mangione and morality

Talk of the town recently has all been about the killing of the UnitedHealthcare CEO. Far more discussed than the killing itself is its justification, whether it was good or bad. The other day my sister asked me my opinion on it, which I was unsure of at the time, but have since solidified.

My immediate reaction to the event was positive. The issue, however, is that the act of murder goes fairly clearly against really the only value I hold dear, life being better than death. Though if I were to disavow all murder, many other issues arise. Primarily the one confronted here, the issue of revolt: when confronted with a murderous system (this is a wide category of organisms: governments, corporations, people, etc.) in which the only immediate remedy is its death, is its death preferable to its life? In other words, is murder justifiable when it leads to less net death?

Before tackling that question we first must bring into question the context of life > death. On an individual level this statement applies only to, well, the individual, since the individual only knows for certain their own liveliness, and not if their perception sees verily other lives like theirs or just figments of the mind or simulations or etc. The life of the individual correlates however with the life of the system it lives in, a healthy system ("healthy" meaning a high life expectancy, since a high life expectancy means there is more life and less death, thus good under the value of life being better than death) containing more healthy individuals and thus greater good (life) for each individual. So as long as one lives in a certain system, the "health" of this system is interlinked with what is good and bad.

So back to the question of if murder is justifiable when it leads to less net death, the answer is yes (assuming the murderer and murdered and observer are all part of the same system) as it would lead to a healthier system. And so to return to that initial question of if Mangione was justified, we can now answer that question by determining if his actions have/will result in a healthier system.

Note that under this system of morality there is no direct consideration of past actions. While the murdered may have committed bad actions (actions resulting in positive net death) in the past, the death of the murdered would only increase net death, not bring any justice to those already affected. The death is only justified insofar as it will result in less future death, so in the future there will be less net death as result of the murder.

The past must still be recounted for the purposes of predicting the future, however. Let's take the classic example of someone stealing bread to feed their family: this has the immediate effects of feeding the hungry, lengthening the time they can live, thus increasing the health of the system. On the other hand, the person who had the bread now does not, so if they are not able to acquire more this may result in them becoming hungry themselves. If they were to starve, this would not be good for the health of the system. We are unable to determine precisely the future effects though, so we estimate by looking at his past (he does not need too much bread) and his present (he has plenty of bread) the chance of the shopkeeper starving. Let's estimate 10%. In that case, it seems that we can calculate the morality of an action by comparing the immediate effects with the future ones, in a fashion similar to M = ∑i + ∑(c∑f), where M is the morality of an action (positive thus being good and negative bad), ∑i the immediate net life from the action, and∑c∑f the sum of the predicted net lifes multiplied by the chance of those predictions. So for this example, we can say that M = 3 + -1 * 10% for a result of 2.9, indicating that the action was good (this is arbitrarily assuming the family was of three people).

There are a few issues with that calculation, however. First of all, we treat the one loaf of bread as being able to fully save the lives of three people, by putting in 3 as the net life of the family. This greatly overestimates the true value of the bread, because it does not forever save them, it merely postpones their starvation. In order to equate for this, we have to add in the future effects of them starving, resulting in M = 3 + (-1 * 10%) + (-3 * 90%) = 0.2 (once again arbitrarily assuming that the chance of them starving later is 90%.) Another issue with this calculation is that it fails to account for the effects each death may have. For instance, let's say that the shopkeeper had a wife and son of their own, and that the death of the shopkeeper could result in their starvation. Let's say their chance of dying in event of the shopkeeper's death was 50%. We need to equate for the chance of their death, so M = 3 + ((-1 + -2 * 50%) * 10%) + (-3 * 90%) = 0.1. But we can't forget the effects that their deaths cause, and the deaths caused by deaths caused by their deaths, and the deaths caused by deaths caused by deaths caused by their deaths, and…

Clearly our original equation is not going to cut it. Since we are examining morality through a collection of individuals, the result of the action has to be examined by the effect it has on each individual, and each effect on each individual examined by the effect that effect has, resulting in a circular calculation. We can represent this with a small series of recursive functions, still keeping M equal to immediate plus future, M = I(i) + F(f, c), where I(i) = ∑[i + F(i₂, c)], i is the immediate consequences, and i₂ is the potential consequences of i. F(f, c) is similarly defined as F(f, c) = ∑[c∑f + F(f₂, c₂)], where f and c are the same as before, f₂ is the consequences of f, and c₂ is the chances of f₂.

Despite the precision of this equation, we are still just guessing all its data, so its accuracy is quite dubious, but regardless let us try and now apply it to the question we first set out to answer. Immediate results is easy enough, -1, the death of Thompson, and his death will most likely not directly lead to any other results, but we can add in a small chance of suicide in his immediate family, so I(i) = -1.002. Future consequences is where it gets a bit more complicated. The main justification for the murder is the high rate of claim denials UnitedHealthcare makes, which is claimed to be 32%, but the question is whether or not the murder will lead to lower denial rates. I'll be quite optimistic and say they will be terrorized enough and say they will in the future only deny 30% of claims. If they have around 50 million users, each approved claim gives around a year more of life, and life expectancy is around 77.5 years, that means that terrorizing UnitedHealthcare will lead to 2% * 50000000 * (1 / 77.5) ≈ 12903 more life. That's pretty good. There are some more upsides, such as other insurance coverers also being terrorized, and also more downsides, such as rises in extremism, and then there are also a million other things from the individuals who would then get to live. But I think just going surface-deep can answer if M > 0, which I say it is, and so the murder is just!

Actually doing calculations like I did above is thoroughly pointless, since most of the numbers we come up with are arbitrary, but having a rough model is nice. The calculations done would also have to vary depending on the person doing them: for I, someone not directly affected by the event, it makes sense to calculate it as the overall effect on my society, but for someone more directly affected, for instance the murderer, murdered, or someone who uses UnitedHealthcare, it would have to be done differently, prioritizing their life in the processing.