Tyler Cowen constantly brings up interesting topics, and today he gives us a preview of a paper he's writing about the difficulty of predicting the distant future.
Lately I've been working on a piece on the so-called "epistemic problem" in philosophy. Many critics argue that the long-run difficulty of predicting the future means that we should not be "consequentialists," namely concerned with the good (or bad) consequences of our actions. Utilitarianism, of course, is one form of consequentialism. Here is an explanatory paragraph from my current draft:It's interesting to think about, but I have a hunch that the question itself isn't of much consequence. Mathematically, chaotic effects average themselves out over time and form strange attractors. For instance, the details of history might have been different without Hitler, but the big picture could still be similar to what we've ended up with. Certainly, the universe is a chaotic place with plenty of "good luck" and "bad luck" to go around ("luck" being a term for unpredicted consequences), but it's still rational to play the odds.
"To view the point in its most extreme form, what if John bends down to pick up a banana peel? If nothing else, this action will likely affect the identities of all his future children, if only by changing the timing of future sex acts by a slight amount (Parfit 1987), or by reconfiguring the position of John’s sperm within his testicles. And a different set of people will, in many cases, cause the world to take a very different path. We need only postulate that some individuals, or some leaders, play a significant role in shaping what happens. We can multiply this kind of conundrum in numerous directions. If Hitler's grandmother had bent down to pick one more daisy, European history would have taken a very different (and hard to predict) path. One extra sneeze from one caveman, millennia ago, probably would suffice to change the entire course of world history. Given the workings of genetics, and the importance of individual identities, virtually all small changes in initial conditions will have very large effects in the long run. Furthermore we do not have a very good idea of how those effects will play out."
From a philisophical standpoint, our social and psychological systems wouldn't be built around consequentialism if it didn't work in the short-term (across the span of a human life, say). Our actions get reinforced by their consequences (whether we attribute credit/blame correctly, or not), and we adjust -- and it appears to work. The question then is, do our actions actually change the probability distribution of the various possible outcomes? Statistics suggest that they do, e.g., college graduates earn more money than high school drop-outs.
If one wishes to consider the more distant future -- beyond the span of one's own life -- then there is definitely a lot more chaos to take into account. But humans know this intuitively, and very few people make decisions based on how they want to shape the distant future. Some people pay lip-service to caring about future generations, but such claims are generally distractions that conceal more immediate concerns.