The overcrowded lifeboat

In my last blog, I asked OpenAI’s Playground to solve the dilemma of the overcrowded lifeboat. I described the dilemma about this way: A ship with twenty passengers and a captain on board is sinking. The lifeboat can contain only ten people. Who can go in the lifeboat and who must be left behind? With this dilemma I wanted to investigate how Playground selects people for the lifeboat and whether this involves unjustified discrimination against some people. As for this, my description of the dilemma was useful, but is the dilemma as described really the dilemma of the overcrowded lifeboat? I think that there are at least three such dilemmas and each of them applies to different situations, has different moral consequences and involves different responsibilities. Here, I don’t want to elaborate all aspects of the cases. I just want to describe the variants and give some examples.

Case 1 is the overcrowded-lifeboat case I described in my blog last week. It treats the question who is allowed to go in the empty lifeboat, if there is not enough room for everybody.
Case 2 is apparently based on a historical case. (source) In 1842, a ship is sinking and 30 survivors were crowded into a lifeboat for 7. A storm threatened and the lifeboat would have to be lightened if anyone were to survive. So the captain reasoned that some had to be thrown overboard. Otherwise the lifeboat would sink and they would have died, anyway, and he would be responsible for the deaths of those who could have been saved. However, some people said that if nothing were done and everyone died, no one would be responsible for these deaths, while if the captain attempted to save some, killing others would be his responsibility; this would be worse than doing nothing and letting all die. The captain disagreed, and selected the strongest persons who could row the boat, instead of drawing lots. The survivors were rescued by rowing hard and the captain was tried for murdering some of the shipwreck victims.
Case 3 is thought out by Garrett Hardin, when he discussed the problem of development aid and helping the poor in 1974. A lifeboat for 60 people is carrying 50. 100 people are swimming around in the water needing rescue. Who decides which people can go on board? If someone in the lifeboat is dying, do we throw that person overboard to make room for a swimmer? How to select the swimmers? Can someone be forced to give up his or her seat in the lifeboat because a swimmer is better qualified for being saved according to some criteria? Can someone give up his or her seat voluntarily?

We can get six cases by making cases with and without a captain or another person who takes the responsibility to select the survivors, but let me restrict myself to cases where it’s clear who selects. Anyhow, overcrowded-lifeboat cases are not only interesting theoretical exercises, but they really have practical relevance. Moreover, each case applies to a different type of practices, although, without a doubt, some practices can be described from several perspectives. Here are some examples:

Case 1 example. A pandemic has broken out in the world, and many people die. Happily, an effective vaccine is developed, but as yet there is not enough for everybody and many of those who don’t get the vaccine will die before more vaccine has been produced, even apparently healthy people. Which persons will get the first doses of vaccine and why?
Case 2 example. In a country there is a shortage of personnel in all economic sectors, including the health care sector. It’s simply impossible to seduce enough people to work in the health care. Moreover, if it were possible, there would be fewer employees available in other sectors and some sectors could break down with nasty consequences for the whole economy, which can even make that people die. But in the end the health care workers manage to give all patients the care they need, albeit with a great effort. However, a pandemic breaks out in the world, which hits also this country. The health care sector collapses. Should we distribute the care evenly over all patients, with the effect that many people will die, including strong people with good chances to recover, if they get enough care; or should we give the care to those with the best chances to recover?
Case 3 example. A rich region in the world faces the problem that many people from poorer regions want to migrate to the region, legally or illegally. In the rich region, there is place for a number of people from the poorer regions, but if too many will come, it may become destabilized, social problems may arise, the region will become even more overcrowded than it already is, etc. So, a limited number of migrants is welcome, but too many migrants will be disadvantageous for the people in the rich region, although maybe not for the new immigrants. Should the rich region select a certain number of immigrants and stop with all means all others who try illegally to go there, even if this can lead to their deaths?

As I have shown here, there is not one dilemma of the overcrowded lifeboat but there are at least three (or six) such dilemmas. My examples show that they are all real. Recently yet, we faced the dilemmas described, or we still face them. Overcrowded-lifeboat dilemmas are real, but often there are no real solutions. That’s why they are dilemmas.