Another mass shooting. A devastating flood. Famine. War. How could a good and all-powerful God let this happen? This, in essence, is the ‘Problem of Evil’. It is a challenge for any theology that posits a God who is omniscient, omnipotent, and perfectly benevolent. It begins by acknowledging the vast amount of suffering in the world. Some of it is due to natural causes: earthquakes, tsunamis, disease (‘natural evil’). Some of it is suffering we inflict on each other (‘moral evil’). The existence of evil – both natural and moral – is, on the face of it, reason to think that God does not exist, or at least that God must fall short in some way. Of all the worlds God had the power to create, why create this one, with its many flaws?

What would an adequate solution to the Problem of Evil look like? To begin with, we need to show that the existence of evil is at least logically compatible with the existence of an omniscient, omnipotent, and benevolent God. One way to do that is to describe a possible scenario in which God and evil coexist. Note that we do not need to establish that the scenario in fact obtains; we need only show that the scenario is possible. Consider an old riddle. Suppose we want to show that the existence of my four-sided house is compatible with the possibility that each side of my house faces south. To do that, we need only describe a scenario in which both are true: in this case, the scenario in which my house is centred on the North Pole. We do not need to establish that my house is in fact on the North Pole (it is not), only that it is possible. In the same way, a possible scenario in which God and evil coexist would show, at least, that God could possibly exist in a world such as ours. Note, though, that even if the existence of God is logically compatible with the existence of evil, we might still worry that the evils in this world render it less likely that God exists. (Compare: even if it is possible for all four sides of my house to face south, it is nevertheless improbable.) For this reason, philosophers distinguish the Logical Problem of Evil from the arguably more difficult Evidential Problem of Evil. We will see, however, that even the Logical Problem of Evil is not so easy to solve.

Let us look, then, at a possible scenario in which God and evil coexist. Notice, first, that while there is much evil in the world, there is also a lot of good: acts of kindness, as well as moments of bravery, compassion, and altruism. It may even be that, over the course of human history, there is more good than evil over all. Note, though, that for our actions to be morally significant – that is, to be the sorts of actions that could even count as good or evil – we must perform those actions freely. If someone makes you commit a crime – say, by hypnotizing you – we do not assign blame because you did not commit the crime freely. The same goes for good acts: you will not get any credit for good acts you were forced to perform. Doing good or evil things requires free will.

We are only capable of performing morally good acts if we have free will. But if we are genuinely free, then we are also capable of performing morally evil acts. God therefore cannot create a world with free creatures, and hence a world with moral good, without allowing for the possibility of moral evil. But a world containing free agents who do both good and evil things is better and more valuable than a world without freedom that doesn't allow for the possibility of any moral good whatsoever.

Philosophers and theologians call this proposed solution the ‘Free Will Defence’. It lets God off the hook by placing the blame for evil squarely on us. Note that the primary focus here is on moral evil. Proponents of the Free Will Defence sometimes try to explain away natural evil by blaming it on demons or fallen angels, so that natural evil is in fact a species of moral evil, performed by supernatural agents.

The most serious challenge for the Free Will Defence was raised by the twentieth-century philosopher J. L. Mackie. Mackie asks, why didn't God make us all such that we always freely choose the good? The Free Will Defence assumes that God must either create a world with no free people at all, or create a world with people who sometimes do evil things. But are these really the only two options? After all, it seems at least possible for there to be people with free will, but who happen never to choose to do anything wrong. And a truly omnipotent God can do anything that is possible: God could have created a world whose inhabitants have free will, but who always make good choices. So why didn't God create a world like that? This is sometimes called the Utopia Thesis.

An important response to the Utopia Thesis takes its inspiration from a picture of divine foreknowledge that dates to the medieval philosopher Luis de Molina. Imagine God at the moment of creation. Molina argued that at this juncture, God has three types of foreknowledge: natural knowledge, free knowledge, and middle knowledge. Natural knowledge is knowledge of truths that hold necessarily, independently of any decisions God makes about what to create. Take for example, 7 + 5 = 12. No matter what God decides, 7 + 5 will still equal 12. (Some philosophers hold that God has the power to decide even these sorts of truths; that God could have made it so that 7 + 5 = 13. But this is a minority view.) Free knowledge consists of truths God knows because God has decided to make them true. God knows that the Earth will have one moon, and not three, because God has decided to create the Earth with one moon. God has creative control over the domain of free knowledge because these truths depend on what God chooses to create.

Molina proposed that, in between natural knowledge and free knowledge, there is another type of knowledge: middle knowledge. This is knowledge about what any possible person would freely choose to do in any possible circumstance. We often talk about what would happen in various circumstances. Even if these circumstances never actually obtain, we think there is a fact of the matter about what would happen if they had. Take the jar of salt in my pantry. If I were to pour it into a glass of warm water, the salt would dissolve. I may never do this; the jar of salt might sit unopened in my pantry forever. But it is nevertheless true of the salt that if I were to combine it with water, it would dissolve. In the same way, it may be true that if you were to win the lottery, you would quit your job. It could be true of you even if you never end up winning the lottery. This is what Molinists call a counterfactual of freedom: it is a fact about what you would freely do if you were to win the lottery. Molina's point is that God knows, pre-creation, all the counterfactuals of freedom, not only about you or me, but also about any possible creature God could create.

Both natural knowledge and middle knowledge put constraints on God's choices about what to create. Consider natural knowledge. God could choose to create a world with three moons instead of one. But God cannot create things that are impossible. God cannot create a round square, or a world in which 7 + 5 = 13. Middle knowledge also constrains what kinds of worlds God could create. If it is true of you that you would quit your job upon winning the lottery, then God cannot create a world in which you win the lottery but decline to quit your job. (Assuming we keep everything else constant: it may be that if lottery winnings were taxed at 95 per cent, you would continue to work.) The point is that if you have free will, God cannot give you a choice between keeping or quitting your job, and then make you keep it. That said, God can use middle knowledge as a guide in decisions about what to create. Suppose God wants you to quit your job. Then God can fix things so that you win the lottery. In this way, God can indirectly cause you to quit your job by placing you in a situation in which you would freely choose to quit. And God can do this without undermining your freedom.

We use our knowledge of counterfactuals of freedom to get people to do things all the time. You know that it is true of Uber drivers that if you pay them the appropriate amount of money, they will drive you to the airport. Relying on that knowledge, you offer to pay them, and they drive you to the airport. In this way, you indirectly cause your Uber driver to drive you to the airport. Yet none of this prevents the driver from freely driving you to the airport. In the same way, God can use middle knowledge to maximize the good in the world by putting us in situations where God knows we will choose to do morally good things. Once God decides which world to create, God knows what we will do – what will happen – down to the smallest detail.

In his 1974 book, God, Freedom, and Evil, Alvin Plantinga applies this Molinist framework to Mackie's Utopia Thesis. Plantinga asks us to imagine two men, Curley and Smedes. Curley is the notoriously corrupt mayor of Boston from the first half of the twentieth century. Smedes is the director of highways. Smedes offers Curley a bribe of $35,000 to drop his opposition to a proposed freeway that would require the destruction of the Old North Church. Curley accepts. Later Smedes wonders, ‘Would Curley have accepted $20,000?’ Suppose the answer is yes: if Smedes had offered Curley a $20,000 bribe, he would have accepted it. This is a counterfactual of freedom true of Curley. And it means that God cannot create a world in which Curley is offered the $20,000 bribe but turns it down.

What if God doesn't want Curley to accept Smedes's bribe? Well, God, knowing full well that Curley would accept the bribe in those circumstances, might create the world so that those circumstances do not obtain. God could ensure that Smedes does not put Curley in that position. There are various ways this could be done. God could cause Smedes to slip and break his leg on the way to his meeting with Curley. Or God could make Curley so wealthy that he would not bother to accept a paltry $20,000 bribe from Smedes.

What if God doesn't want Curley to do anything wrong? Could God engineer a world in which Curley is never in a situation where he makes a bad choice? This will depend on what other counterfactuals of freedom are true of Curley. It could be that if Curley were wealthy, he would not accept Smedes's bribe. However, it might also be true that if Curley were wealthy, he would do something else wrong: perhaps Curley would use his wealth to make corrupt bribes of his own. That is, Curley might suffer from what Plantinga calls transworld depravity. Curley's counterfactuals of freedom could be so bad that any world with Curley in it is a world in which Curley does something wrong. If Curley has transworld depravity, then God cannot create a world in which Curley has free will, but never does anything wrong.

Armed with middle knowledge about Curley's unfortunate counterfactuals of freedom, God might throw in the towel and not create Curley at all. Better to have no Curley than a Curley who will inevitably do bad things. But what if it is not just Curley who has transworld depravity: what if we all suffer from it? What if every potential person God could possibly create suffers from transworld depravity? In that case, God cannot create a world with free creatures who never do anything wrong. No matter who God creates and no matter what situations God puts them in, they will do something wrong. If every possible person suffers from transworld depravity, then it could be that this world – with its mixture of good and evil – is the best option open to God.

We have seen how Plantinga uses the Molinist conception of middle knowledge to describe a scenario in which God and evil coexist. But many philosophers find the notion of middle knowledge suspect. First, remember that these counterfactuals of freedom are supposed to be true before God creates anything. So they are true of people who do not yet exist, or who may not ever exist at all. But how could there be truths about what non-existent people would do? It is one thing to talk about what you or I would do in different circumstances, but how can we talk about what someone would do if that person does not even exist? Second, if God is truly omnipotent, then how can middle knowledge be out of God's control? Why can't God choose which counterfactuals of freedom are true in the same way that God can choose how many moons the Earth will have? The idea that middle knowledge puts constraints on what God can create seems at odds with the idea that God is all-powerful. Finally, remember that middle knowledge is a kind of foreknowledge. God, having created us, knows what we will go on to do. But if God knows what we will do ahead of time, and God is omniscient, then how is it that we do anything freely? If God knows thousands of years before Curley was born that Curley would accept Smedes's bribe, and assuming God is infallible, then how could Curley do anything but accept Smedes's bribe? And if Curley cannot do anything else but accept the bribe, then how is it that he accepts the bribe freely?

These sorts of worries have led some philosophers of religion – most notably the philosopher William Hasker – to suggest that God does not have advance knowledge of what we will freely do. That is, if we are genuinely free, then it must be an open question what choices we will make right up to the moment that we make them. And if it is an open question what we will choose to do, then it is not something God can know ahead of time: there is nothing yet for God to know. These Open Theists claim it was not possible for God to foresee the evil that would result from creating free creatures. God may have hoped that we would make good choices but did not know for certain that we would. Creating free creatures therefore involves taking a risk. The late philosopher David Lewis describes the God of Open Theism as an unlucky gambler: God made a gamble on how we would use our freedom, and lost the bet. But the bet was nevertheless a wise one to take since a world with no free creatures, and thus no morally good actions, would be, as Lewis puts it, mediocre.

This is all well and good, but why, then, doesn't God intervene when we begin to go astray? Both Mackie and Lewis raise this question in somewhat different ways. Mackie asks, why doesn't God control evil wills, leaving us free to will rightly, and taking away our freedom when we are about to will wrongly? Lewis asks, why not make all evil victimless? God could have put us in a metaphorical playpen, engineering things so that whenever we make a bad choice, no one suffers as a result. In both cases, God would be a kind of helicopter parent: God would hover over us, micromanaging every aspect of our lives so that we are never able to make mistakes, or at least arranging it so that no bad consequences ever result from the mistakes we do happen to make.

Parents often agonize about how much to intervene in their children's lives. Should I wake up my sleepy teenager when he forgets to set his alarm, or do I let him be late for school? Should I e-mail my ten-year-old's teacher about a conflict with a classmate? Should I monitor my children's Instagram accounts? We typically hear that helicopter parenting will result in children who are less self-reliant, who are prone to anxiety, or who are unable to self-regulate. Though perhaps this is only a problem because our power over our children's lives is not absolute. Eventually, we must cut the cord and let our kids strike out on their own. And when that happens, they will probably face hardships we cannot protect them from. There is a reason they need to learn self-reliance and perseverance in the face of adversity. God, on the other hand, is supposed to be omnipotent. If God puts all of us in a permanent playpen, we would have no need for perseverance in the first place. There would be no adversity to persevere through. And in any case, if God can intervene when we will wrongly, God can presumably also intervene to forestall any bad effects that might result from all this intervention.

Nevertheless, there are some potential problems with the helicopter model of divine providence. First, as Lewis notes, if God puts us all in a playpen, if all evil is victimless, would our choices still have the same moral significance? Perhaps our moral decisions matter because they have consequences. If we are in playpens, there would be no real consequences for our decisions. Any ‘good’ decisions would be hollow, since nothing bad would have happened if we had decided differently. Second, some argue that the possibility of suffering and loss play an important role in a meaningful life. Perhaps the projects and the personal relationships that give shape and meaning to our lives would no longer grip us in the right way if nothing bad could ever happen. A life of meaning and value requires that there be something at stake, and if God is constantly shielding us from the possibility of anything going wrong, nothing would really matter.

Some have suggested that we should think of God as a free-range parent instead. Indeed, there are interesting parallels between the Open Theist conception of divine providence and the free-range model. The Open Theist presents God as more risk tolerant. The free-range parent makes a judgement that a child's freedom and autonomy is worth the risk that the child might be abducted while taking the subway home alone. God likewise reasons that our ability to make choices of moral significance is worth the risk that we might make the wrong choices. The free-range parenting literature also emphasizes the importance of failure. In Lenore Skenazy's influential book on free-range parenting, there is a chapter entitled: ‘Fail! It's the New Succeed’. It is in this context that we hear about successful people who experienced failure along the way. Michael Jordan was cut from his high school team. Abraham Lincoln failed in business and lost several elections before becoming president. When your children fail at something, they learn to pick themselves up and try again, or to try something different. Likewise, God may refrain from intervening in our lives because it is only by failing that we learn to pick ourselves up.

At the same time, we should be careful not to push the analogy too far. For one thing, free range parents stress the importance of being mindful about how much freedom their children are ready to take on. A preternaturally mature nine-year-old may be ready for a solo subway ride. A toddler – obviously not. But looking around at the horrors in the world, we might worry that some of us have been given more freedom than we can handle.

Another theme in the free-range parent literature is that many of the fears we have for our children's safety are overblown. The probability that your child will be abducted on her way home from school is astronomically low. But the way the media sensationalizes every tragedy makes them play an outsized role in our imagination. However, even the most resolute free-range parent will presumably intervene when the situation is truly dangerous. And while it is important to let a child fail, I would think that most free-range parents would intervene if their children were about to do something that will cause great harm. As we reflect again on the amount and degree of suffering in the world, we might wonder why we are still waiting for God to intervene.