I've used variations on the same idea idea in two different contexts (abortion, theodicy). Now I'd like to combine them in reference to theodicy. When atheists raise the problem of evil, the unspoken assumption is that a better world is possible. They can imagine various ways of improving the world we inhabit. If, therefore, an omniscient, omnipotent, and benevolent God really exists, the actual world would correspond to the better world an atheist imagines. Or so goes the argument.
But as I've often pointed out, that's shortsighted. There is no best possible world. For alternate timelines have unique goods. Some goods are contingent on prior evils. Preventing the evil prevents the resultant good.
Just about every human life produces a chain reaction. Whether or not a particular individual exists will affect the course of history in complex ways. If he exists, history will go one way. If he doesn't exist, history will go another way. Time-travel stories illustrate the principle of tradeoffs in that regard. Our individual lives may seem insignificant, but lives have long-range consequences–as does their absence, counterfactually speaking. The upshot is that when an atheist imagines a better possible world, there are losers in that scenario. Indeed, people who are winners in one possible world may well miss out in the "superior" alternative. It's no improvement for them.
One objection I encounter to this observation is that people who never exist in the first place have nothing to lose. That may be true given the status quo, but the objection is superficial and misses the point. Nonexistence is the greatest conceivable deprivation. Every lesser deprivation is a matter of degree. But this is a lost opportunity in the most absolute sense.
If, for some odd reason, it just isn't possible for someone to exist, then there's nothing to lose. There was no alternative. If, however, the alternatives are existence and nonexistence, and those are both live possibilities, then to be denied the opportunity to exist when that was feasible is a genuine loss.
To take a comparison, when a teenager dies, we consider that an "untimely" death. He "died before his time". We lament the death of the young because they had their "whole life ahead of them".
The sense of loss is based on wasted potential. Lost opportunities. The future he never had. He missed out on so much.
But if that's valid for a teenager, we can extend the same principle back in time. If a 16-year-old has so much to lose, doesn't a 6-year-old have at least as much to lose, if not more? What about a 2-year-old? Or a 6th-month old baby? Or a 3-month-old baby in utero? At each stage of premature death, there's lost potential. And the further back you go, the greater the deprivation. The greater the unrealized potential.
What about a minute before conception compared to a minute after conception? If that really different in kind? You may say that prior to conception, he doesn't exist, but isn't one minute's difference either way rather arbitrary? Since the principle concerns potential futures, it ranges along a continuum. There's no intrinsic cutoff at any point along the continuum. Suppose you make the cut at 20. But you could just as easily make cut at 19. You could make the cut a moment earlier, or an hour earlier, or a day earlier, or a week earlier, or a month earlier, or a year earlier. The sooner the cut-off, the more there is to lose. The lost opportunity is that much more extensive.
Notice, I'm not saying that possible people who never exist were wronged or harmed by never existing. But there's a weighty sense in which some people are better off existing than not existing. Given the opportunity, they'd enjoy that.
Before an atheist complains about how God could make a better world, the atheist needs to think several moves deep. Like a chess game, changing one move changes subsequent moves.