To the title question, I would say that sometimes inaction is morally wrong, but not always. If, for example, we change the trolley problem such that no one is on the side track, and, assuming that one knows how to throw the switch and is capable of doing it (little details that make it in practice unrealistic, because most people don't know how to throw train switches), then I think it would be wrong to not throw the switch. And one should be blamed for this callousness in not saving 5 people's lives with so little cost to oneself. (As to whether there should be any legal responsibility, I have no wish at present to comment on that; I am confining my remarks to moral or ethical responsibility.)
But I do not accept the idea that it is morally okay to murder one person to save five people. Although most people are ready to say that they should do so in the original trolley problem, they often do not keep to that principle with the variations to the problem. Such as:
Marvin_Edwards wrote: ↑May 29th, 2020, 2:23 am.... The trolley example is a bit deceptive, because it suggests at the outset that we can save 5 people by killing 1 person. So, the immediate response is well, saving 5 must be better than saving 1.
In the case of the trolley we are tempted to do that. But if we switch to the doctor wanting to save 5 people by killing 1 person and harvesting their organs, the situation is slightly different.
....
I suspect that most people do not think it would be morally okay for a surgeon to go out and grab some person off the street to harvest their organs to save five people.
That is a more extreme case than the case of pushing a fat man off of a bridge, whose body causes the trolley to stop, to stop the trolley from running over the five people. More people are willing to go along with killing the one with the original problem than with the fat man version, and more people are willing to go along with killing the fat man than will go along with the surgeon harvesting the organs of one (thereby killing the one) in order to save five people.
In short, people who affirm that one should murder one person to save five often do not follow that principle in all of the scenarios in which that idea is applicable, where killing one person would save five people. Clearly, there is something else that is involved in their opinions of how to deal with the various scenarios, if they do not stick with the principle that it is better to kill one to save five.
Also, it is worth mentioning that such thought experiments as the trolley problem are not meant to be realistic. They are idealized scenarios intended to illustrate ethical principles that people follow (or say they follow). In realistic scenarios, there are typically too many uncertainties to make it clear what principles people are applying to the problem, and what principles they should apply. Like, for example, jumping in a river to save a drowning person. If one contemplates that, often, one is uncertain that one will actually succeed in saving the person, and one may also drown oneself. Some years ago, four people drowned in a fountain at the Fort Worth Water Gardens, the one that was used as a set on the film
Logan's Run. I read that one person fell into the fountain, and the other three entered to try to get the one out. (Not all at the same time; one entered, and then could not rescue the one who fell in and also could not get out, and then the others followed.) They all died, whereas only one would have died if they had not tried to save the one who fell in. In real life, the question of what to do to save someone is typically filled with uncertainty about the outcome. This is why thought experiments are used, so that one will not get sidetracked by the uncertainties and can focus on the relevant ethical issues.
In other words, the idealized situations are unrealistic on purpose. So complaining about it being unrealistic means that one is missing the point of the thought experiment.