My claim that people pose the problem of evil as a basically moral objection to classical theism ought to be uncontroversial. The most common way of pressing the objection is to argue that God's permitting some people to suffer horribly beyond their deserts, when he could prevent it, is immoral. That is taken to be incompatible with God's being perfectly good, a quality classical theists ascribe to God. Of course, another way to pose the problem of evil as an objection is to argue that, although a perfectly good God might well want to prevent such suffering, he is powerless to do so. That is taken to be incompatible with omnipotence, another attribute classical theists ascribe to God. But that way of pressing the objection is fairly easy to answer. Although an omnipotent God could well have created a world in which suffering does not far outstrip deserts, God has not done so; given the natural order God has willed, it is logically impossible to prevent such presumptively disproportionate suffering without divine intervention so regular as to destroy the natural order of things. And omnipotence neither need nor should be thought to include the ability to do the logically impossible. So much is, or ought to be, obvious. Of course, the standard reply to that defense is to argue that God is immoral for creating and sustaining such a natural order of things in the first place when, as granted, God could have done otherwise. But that's essentially the same as the first way of pressing the problem of evil as an objection to classical theism.
My claim that the superfluity objection is also, at bottom, a moral one is much more controversial. Most of what follows provides my argument for it.
Metaphysical (as distinct from methodological) naturalists typically hold that the sorts of explanation of the world's existence proffered by classical theists—chiefly, by means of a posteriori cosmological and teleological reasoning—cannot do the sort of work that explanations in general are supposed to do. If so, then citing God as creator and/or designer of the world fails to explain anything; therefore, there is no reason to hold that God as explicans exists. But what does it mean to say that theistic explanation of the world's existence doesn't do the sort of work that explanations are supposed to do?
Classical theists should not, and the most intelligent among them do not, argue that presenting God as creator and/or designer does better explanatory work than the natural sciences. Those sciences have their own explanatory aims and canons which, though not immune to revision, remain exactly as they are whether or not classical theism is true. The theistic argument is, rather, that citing God as explicans does a different sort of explanatory work than natural science. The naturalist reply is that no such alleged "explanation" should be counted as explanation. What is the argument for that reply?
To explain something is to account for why it thus and not otherwise. In order do that, one must show that the explicandum would have been different if the explicans had been different. But classical theism does not claim that the world would have been different if God did not exist; the claim is that the world would not exist at all if God did not exist. That requires holding, among other things, that the world can and should be conceived as a certain totality which counts as an explicandum, such that only the action of something not comprised by that totality could account, in some non-trivial way, for why just that totality exists. But it will not do to characterize said totality simply as the physical universe studied by natural science, even at some hypothetical state before the Big Bang. For all we know, the primordial universe might have been the product of something else which could not qualify as the God of classical theism, but which might turn out to be identifiable by means of natural science—if not our science, then somebody else's. No, the totality that divine activity supposedly explains must be the totality-of-things-that-happen. Call that 'T'. Granted we do not know its full extent, T certainly exists. But such a totality, the naturalist would say, cannot admit of non-trivial explanation. We can explain its existence simply by noting that each of its constituents exist; but that would be trivial, and certainly not what the classical theist is after. And the reason we cannot have what the theist is after is that the theist cannot say what would have been different about T if God did not exist. T remains just as it is, whatever it is, whether or not God exists. Hence, goes the argument, citing divine activity to explain T's existence does not and could not really explain anything. There is no non-trivial explanation of T's existence. As Laplace said, there is no need of the God-hypothesis. It is superfluous.
From this point of view, it will not do to cite some version of "the principle of sufficent reason" as a premise in an argument for the claim that something not comprised by T accounts for T's existence as a totality. There's already "reason enough" for T's existence as a totality: that of each of its constitutents. But that is hardly germane. What the theist must do instead is show that T is the sort of entity whose existence calls for another sort of explanation altogether. Yet how is the theist to do that? The only way he can do it is to show that, whatever the extent of T as a totality, its constituents cannot, either individually or collectively, account for the general causal regularities that must be cited in some explanation of how things happen as they do—i.e., the constituents of T cannot account for the "laws of nature." But that sort of explanation would have to show that such laws would have been different if God as creator and/or designer did not exist. And that in turn would have to cite some sort of causal regularity. But given that such regularities are supposed to be part of what's being explained, such an explanation cannot qualify as an explanation at all.
It might seem that the way for the theist to begin countering that line is to point out that it premises scientism: the thesis that only what can be known scientifically can be known at all. He can then go on to argue that there is no good reason to believe scientism. And he would be quite right. Humans have always known various things non-scientifically, and no scientific argument for scientism can be given. But that will not suffice by itself. For the naturalist can always argue that, even if scientism is false, his point about explanation remains untouched. Even if there are things natural science cannot explain, and thus cannot know, that's no reason to believe that T's existence can be explained in some other way. Unless and until the theist can show that his "explanation" of T's existence does what explanations do, he hasn't explained anything.
As I've suggested, the debate is really about the nature of explanation. It is evident that there are successful explanations in the contexts of ordinary life and natural science, but it is by no means evident that there can be a kind of explanation which doesn't tell us how things would have gone differently if the explicans did not exist. To be sure, the theist must say that, if God did not exist, then T would not either—a conditional statement which, if true, would be a non-trivial truth. But that doesn't tell us that things would have been different if God did not exist; it only tells us that there would have been no "things" to be either the same or different if God did not exist. Absent some account of explanation which shows that such a peculiar result can function as explanations do, the theist has not established that he's explained anything. Nor will it do for the theist to insist that T is the sort of thing whose existence is explicable; for the only "sorts" of things we are familiar with are the sorts of things already comprised by T.
The only honest way for the theist to proceed is to argue that the question "Why does T exist?" is meaningful in such a way that one could reasonably entertain a non-trivial answer to it. That would show that we cannot rule out T's existence being explicable in terms of something which T does not comprise. And the only way to develop such an argument is to show that (a) one cannot rule out that T's existence embodies an intention, because (b) intentional explanations need not be thought reducible to causal explanations, which perforce cite natural regularities. That kind of argument has been given from time to time. In my hoary PhD thesis, I developed along such lines a book-length argument that it's more reasonable to allow for a unitary explicans of T's existence than to rule out the possibility of such an explicans on epistemological grounds. I still would argue to that effect.
As I've discovered over the years, however, the naturalist objection to that move is an essentially moral one. In ordinary life, natural science, and especially in formal disciplines such as logic and mathematics, there are reliable, agreed-upon methods for evaluating explanations as successful or unsuccessful. Prima facie at least, there are no such methods in natural theology—a discipline that not even the majority of religious believers find helpful. Given as much, naturalists typically argue that one ought not to expect people to find any of the putative explanations of natural theology cogent as explanations. Expecting people to do so is, in fact, morally defective. For such "explanations" necessarily transcend the sorts of considerations that it's reasonable to count as evidence; expecting people to go beyond the evidence in forming their beliefs is expecting what's unreasonable; and expecting from people what's unreasonable is a sign of disreputable motives that are themselves all too evident in the history of religion.
To judge from the recent literature of the "new atheism," which is really the old atheism with shoddier arguments, that's the kind of objection, other than that from the problem of evil, which motivates people to be atheists. I have no doubt, of course, that some atheists are such because they very much don't want to consider the implications for their lives if Christianity or some other form of classical theism is true. But that only serves to supply theists with a moral argument against atheism that is too ad hominem to be worth pressing. The real interest of the moral arguments against theism is that they steer the debate into a channel where the theist is on firmer ground. What is that ground?
Regarding the problem of evil, the theist can and ought to argue that the atheist has no moral legs to stand on. If a given atheist is an emotivist or some other sort of non-cognitivist in moral philosophy, he has no reason to believe that there are objectively binding moral norms which God fails to satisfy. If a given atheist is a utilitarian or some other sort of consequentialist in moral philosophy, he has no reason to believe that God's utility calculations, if there is a God, are inferior to his own. If a given atheist is some sort of deontologist in moral philosophy, he must show several things: that the moral norms he believes bind humans absolutely do so even though there is no God; that even if there were a God, those norms would bind God in pretty much the same fashion as us; and that God, if there were a God, cannot be said to observe them. All that is, at the very least, a tall order.
Regarding explanation and evidence, the atheist needs to show more than merely that it's unreasonable to expect people in general to find classical-theist natural theology persuasive. That people in general do not find such theology persuasive is easily accounted for by factors other than the objective quality of its arguments. Most people lack the happy combination of time, talent, and education to study and evaluate such arguments, so that whatever the reasons this-or-that person might have for believing in God, they cannot be faulted for leaving natural theology alone. For that reason, classical-theist philosophers don't expect most people to follow and evaluate their arguments. So the debate is really among philosophers, and the question whether one ought to go beyond what's generally recognized as evidence is a debate in moral philosophy and psychology.
About that debate, I shall conclude by noting that the atheist has a lot more work to do than simply pointing out that something called "religion" violates his moral norms. There are many different forms of religion, and some are more capable of moral self-reformation than others. But what is such "moral self-reformation" supposed to amount to? Before a charge of immorality can be made to stick, there has to be antecedent and common agreement about what morality requires. A person who wants to press a moral argument against theism, but who believes that the universe is morally indifferent and that no transcendent lawgiver underwrites morality, is burdened with showing that the moral norms he upholds are objectively binding as such. For unless and until he can do that, his moral arguments against theism can do no more than beg the question.