It's really hard; that's basically my point. But I think a good starting place is to model the expected effects on various populations of all of these various

It's really hard; that's basically my point. But I think a good starting place is to model the expected effects on various populations of all of these various factors (increased floods, increased crop yields, etc). If the vast majority of these models are tilted overwhelmingly negative or positive, we can somewhat adjust our confidence levels. Of course, we'll never be able to be that confident when the time horizon stretches so long (100+ years) and the effect sizes are so small (e.g., sea levels are expected to rise by a lower amount on average than the existing low-tide to high-tide variance). But..that's also my point: it's not merited to be so extremely confident that the effects of global warming are going to be so negative (or positive) over 100+ years. People focus extremely hard on the negatives of global warming, don't focus much at all on the positives (my guess is most people living in urban areas aren't even aware there even exist any positives), and focus their debating efforts on the (as far as I can tell) long ago settled question of whether or not the climate warming is happening, or is caused by humans. My guess as to why this happens is that considering whether global warming could be a positive thing for the world conflicts with an anti-market bias that people in urban areas frequently share.