I floated a few Trolley-problem-type hypotheticals yesterday that highlighted some difficult choices that sometimes have to be made when it comes to punishing wrongdoers. Specifically, what to do in cases when rewarding (say, via gentle punishment that is actually desirable, or via outright rewards) a wrongdoer turns out to be better for society at large than punishing that wrongdoer. The example I gave was the bailout of banks in the recent financial crisis, which could have limited damage to the economy but would have rewarded those who helped cause the crisis.
Expanding on this, there are related questions worth pondering (if you have nothing better to do):
1. In the bailout scenario, what if helping the banks only improves the odds of helping the economy by 90%, and the bailout money is not recoverable? 75%? 50%? 20%?
2. In the bailout scenario, what if helping the banks can help the economy, but also increases the risk of future crises by 10%? 100%? 1000%?
3. In the context of a financial crisis caused by criminal wrongdoing, what if maintaining confidence in the banking system requires a bailout and non-prosecution of criminal wrongdoing?
4. In the context of a financial crisis caused by criminal wrongdoing, what if maintaining confidence in the banking system requires a bailout actively enriching those who caused the crisis – basically, paying them to undo what they did?
Let’s make it more difficult and go from income statements to more graphic situations, and go in different directions. Imagine for now that the analysis below can only be done after the fact, not before.
5. A genetic analysis determines that criminals with a certain gene variant who commit sexual assault do not commit any other crimes for the rest of their lives if they aren’t punished. If they are punished, their recidivism rates are the same as for all other rapists. Do you punish them for their first rape, if you catch them, or do you let them go?
6. A genetic analysis determines that criminals with a certain gene variant who commit sexual assault commit crimes at a half the rate than average citizens if they get away with their first crime, but if punished, their recidivism rates are the same as for all other rapists. Do you punish them? What if the rate is the same as for average citizens but lower than other criminals?
7. A genetic analysis determines that criminals with a certain gene variant who commit sexual assault are completely rehabilitated and commit no crimes again if they spend 17 years in prison, which is much longer than such criminals usually spend in prison. Do you keep them in prison for the longer term?
8. Imagine a scenario combining 1 and 6, where two 18-year-olds commit the same crime, perhaps even together. Can you justify letting one go and imprisoning one for 17 years, if their gene variants differ?
These are hard questions, I think, in part because they run counter to some very strong innate instincts of fairness and justice. However, we’re bound to encounter issues of this kind, and it’s worth thinking about them. Obviously there are serious real-world complications to consider: for example, knowing your own gene variant in the latter scenarios can affect your behavior. If you know you have the gene in #5, you basically get a freebie if you want to commit sexual assault. This is clearly a terrible incentive, but look at the alternative, in which a punish assailant is punished, goes free, and then likely offends again (recidivism rates are high). These scenarios don’t present good options, and I hope we don’t have to confront their analogues in the real world. Chances are, however, that we’ll have to.
I don’t have any answers here, but feel free to share your instincts.