Doing and refraining, and proportionality

In my previous post, I suggested that proportionality considerations in Double Effect work differently for positive actions (doings) than for negative ones (refrainings). One thing that is now striking me is that there is an interesting asymmetry with respect to relational features that is brought out by thinking about pairs of trolley cases with different groups of people on the two tracks, but where we vary which track the trolley is initially heading for.

For an initial pair of cases, suppose on one track is someone one has a close relationship with (one’s child, spouse, parent, sibling, close friend, etc.)—“friend” is the term I will use for convenience—and on the other track a stranger. Then:

• It’s completely clear that if the trolley is heading for the stranger, it is permissible not to redirect the trolley

• It’s significantly less clear but plausible that if the trolley is heading for the friend, it is permissible to redirect the trolley.

In this case, I already feel a moral difference between doing, i.e., redirecting the trolley, and refraining, i.e., leaving the trolley be, even though my permissibility judgment is the same in the two cases: redirecting the trolley towards the stranger and allowing the trolley to hit the stranger are both permissible. And yet regardless of where the trolley is initially heading, there are the same two outcomes: either a stranger dies or a friend dies. The difference between the cases seems to be solely grounded in which outcome is produced by doing (redirecting) and which by refraining (not redirecting).

Suppose we vary the ratio of strangers to friends in this case. At a 2:1 ratio of strangers to friends, my intuitions say:

• It’s very plausible that if the trolley is heading for the strangers, it is permissible not to redirect the trolley

• I can’t tell whether if the trolley is heading for the friends, it is permissible to redirect the trolley.

As the ratio of strangers to friends increases, my intuition shifts in favor of saving the greater number of strangers. But, nonetheless, my intuition consistently favors saving the strangers more strongly when this is done by refraining-from-redirecting than when this is done by redirecting. Thus, even at a 10:1 ratio of strangers to friends:

• It’s almost completely clear that if the trolley is heading for the strangers, it is morally required to redirect

• It’s completely clear that if the trolley is heading for the friends, it is morally forbidden to redirect.

In fact, I think there are points where the ratio of strangers to friends is both sufficiently high that:

• If the trolley is heading for the friends, it is forbidden to redirect

and yet still sufficiently low that:

• If the trolley is heading for the strangers, it is not required to redirect.

I feel that 3:2 may be such a ratio, though the details will depend on the exact nature of one’s relationship with the friends.

These cases suggest to me that the proportionality requirements governing refrainings and doings are different. It is consistently easier to justify refraining from redirect than to justify redirecting even when the consequences are the same. Nonetheless, even though the proportionality requirements are different, in the cases above they do not look qualitatively different, but only quantitatively so.