Valuing Unnecessary Causal Contributions

In 'Why Citizens Should Vote: A Causal Responsibility Approach', Alvin Goldman argues that (i) there's a sense in which each vote for the winning party causally contributes to their victory, even if they receive many more votes than are necessary for victory (and similarly each vote against the winning party serves to causally "counteract" them), and (ii) you are morally responsible for outcomes that you are, in this way, causally responsible for.  So you get moral credit for voting for good parties, and against bad ones, and on this basis have (non-trivial) reasons to act accordingly.

I'm happy to grant this talk of "causal contribution", but I wonder about its normative significance. I'm more inclined towards Donald Regan's account of the ethics of cooperation and coordination. Roughly: we should be disposed to coordinate with like-minded others to bring about the best collectively possible results. But if others are not disposed to coordinate with you, then there's no point in pretending otherwise, or in valuing things (such as unnecessary "causal contributions") other than good consequences.

So, while many people who are unnecessary contributors to good or bad effects might indeed be acting well or poorly (respectively), I do not think there is any essential link between causal contribution (in this weak sense) and the moral status of one's action.  The mere fact that in Φ-ing one would be a causal contributor (albeit an unnecessary one) to some good outcome is not, I think, much if any reason to do it.

To bring this out, consider the following evil demon case:

An evil but trustworthy demon holds a referendum on whether to blow up the world. He explains there are two consequences:

(1) He will destroy the world unless at least 50 people vote against it, and
(2) He will torture 1 puppy for each person who votes against destroying the world.

You have x-ray vision, and as you enter the voting booth, you can see that there have already been the needed 50 votes against destroying the world. What should you do?

Surely you should not vote against destroying the world, for such a vote would (i) do no good, given that sufficient such votes have already been submitted, and (ii) do some harm, leading to additional puppy torture.  Here it seems fine to vote for the destruction of the world (knowing that there's no risk of the vote bringing about that outcome), or at least to spoil one's ballot or refuse to vote at all.

Now, you might object that puppy torture is a fairly extreme harm, a really significant reason that can outweigh our reasons to causally contribute to preventing the destruction of the world. I think the case still works if you make the cost here much weaker (a foregone lollipop, say). In any case, preventing the destruction of the world is a much bigger deal regardless, so if the value of causally contributing to an outcome is at all proportional to the value of the outcome to which one is contributing (as it surely should be), then the only way this could possibly be outweighed so easily here is if the value of unnecessary causal contributions is completely negligible (a near-zero portion of the value of the outcome to which one contributes).  But it is outweighed here.  And so the value of unnecessary causal contributions is completely negligible.

Sound right?

[Update: See also my follow-up post, The Best Case for Voting.]