William Rowe, like most philosophers of religion today, rejects the notion that "God exits" is logically incompatible with pain and suffering. I will delve into the logical problem of suffering at a later time, but for now I want to focus on the implications of Rowe's argument. "God" is to be understood not only as a Creator/Designer/Sustainer of the universe, but as a maximally great being (all-powerful, all-knowing, and morally perfect). Rowe himself argues like this:
1. Pain and suffering exist. (Premise)
2. If God exists, we would most likely know what the morally sufficient reasons God has for allowing pain and suffering are. (Premise)
3. We do not know what the morally sufficient reasons are. (Premise (3)
4. Therefore, God most likely does not exist. (From 1 - 3)
Again, Rowe isn't arguing in favor of atheism. Even if he is correct above, there could still exist a very powerful personal Creator, and so forth. Rowe is only arguing against God as a maximally great being.
First, the theist could challenge any of the premises. Premise (1) should be granted because of its perspicuity. (3) is debatable. The Free Will Defense serves as a refutation that we do not know why God would allow pain and suffering. I don't want to focus on this too much at this time, though. I'm much more interested in (2).
It has become almost a mantra among some of our secular friends that evidence is needed in order to conclude that something exists. As much as I disagree with this, it may be granted for the sake of argument. For instance, suppose that you are accused of a crime and all the available evidence suggests you are in fact guilty. Yet, you know you are innocent. Surely you are not compelled to believe that you committed the crime. This is because your experience of the incident provides you with warrant to maintain your innocence. This warrant is an overwhelming defeater of the evidence against you. Now, you may not be able to persuade law enforcement that you are innocent, but that point is moot, since we are focusing on your own epistemic justification.
Now, for the one who has heard the problem of suffering as defended by Rowe, but is also aware of the argument of natural theology for God's existence, could we not have a similar case? Take the ontological argument:
1*. It is possible that a maximally great being exists. (Premise)
2*. Necessarily, a maximally great being is maximally excellent in every possible world. (Premise)
3*. Necessarily, a maximally excellent being is omnipotent, omniscient, and morally perfect. (Premise)
4*. Hence, an omnipotent, omniscient, and morally perfect being exists in all possible worlds. (From 1 - 3, S5)
5*. Therefore, an omnipotent, omniscient, and morally perfect being exists. (From 4 - the actual world is a member of the set of all possible worlds)
In short, God's possible existence entails God's actual existence. This means that the possibility premise is the most controversial of the ontological argument. In fact, (2*) and (3*) are virtually uncontroversial, leaving (1*) as the only premise to be viably rejected.
Yet, we can actually reverse Rowe's argument against God and turn it into support for the ontological argument's premise (1*). Consider the following:
1'. To say that God is impossible is to say that God's existence is not possible. (Premise)
2'. If God's existence is impossible, we would most likely know what the reasons are for God to not possibly exist. (Premise)
3'. We do not know what the reasons are. (Premise)
4'. Therefore, it is most likely the case that God's existence is possible. (From 1' - 3')
If at this point the skeptic wishes to reject either (2') or (3'), then the theist is given at least as much justification for rejecting the original (2) and (3) of Rowe's argument. Do we not have enough information yet to conclude that God's existence is possible or impossible? Then why should we expect to have enough information to say that pain and suffering undermine God's existence? The skeptic should be careful to not commit a double standard.