I just finished reading The Rational Optimist, by Matt Ridley, and I’m finding it both interesting and interestingly unsettling. Its subtitle is “How Prosperity Evolves”, and it’s a look back at various aspects of human development from an optimistic libertarian point of view. Basically, his thesis is that new ideas lead to new niches for specialization, which generates time savings that give more room for new ideas to grow, creating a virtuous cycle; to make things even better, ideas turn out to be surprisingly good at mixing and recombining to give new areas for improvement. So, if you put all this together (and don’t get in its way; governments are Ridley’s favorite target), you have an incredible flourishing of prosperity; and there’s nothing magic here, so no reason to believe that this flourishing won’t continue over the coming centuries.

Which I found it really refreshing to read. It is, to me, undeniable that the world has gotten hugely better over the last few centuries; and I like seeing somebody say “why assume that that progress will stop now?”, backing it up with plausible reasons why that might be the case. As he says, there’s a lot of pessimism out there; if there’s reason for that, fine, but is there?

His focus on specializing was a bit of a wakeup. Over the last several years, I’ve spent time in intellectual climates that warn of the dangers of specializing; certainly I prefer the role of generalizing specialist myself. Having said that, there’s no particular reason why what I prefer should be good for the progress of society as a whole! He points out Ricardo’s law of comparative advantage: even if person A is better at person B at producing both X and Y, if it happens to be the case that A can produce X faster than Y while B can produce Y faster than X, it’s still best for everybody if A specializes in producing X and B specializes in producing Y, because that leads to the most production for the least effort, despite Y being produced less efficiently than it could be. (That sentence is pretty abstract, so follow the above link for some numeric examples.)

Given this, why not specialize wherever possible? Lean and ToC give one answer, namely the dangers of suboptimizing. And Ridley gives another answer: new ideas appear through the mixture of other ideas. (Among other means.) So I don’t have to feel bad about my penchant for sticking my nose into random areas of thought; but the flip side is that what I do professionally (and, to a large extent, in my free time) is solidly grounded in the details of programming, and that’s healthy. (And, looking back over the course of human history, my profession certainly qualifies me as highly specialized.)

He also had some interesting negative things to say about movements that I’m ideologically sympathetic to, namely the organic food movement and the movement against global warming. I don’t entirely agree with his criticisms of both of those movements—in particular, I think he both overestimates how seriously people took some previous disaster scenarios (e.g. acid rain) and underestimates how serious some other previous disaster scenarios actually were (nuclear war; actually strike the word “were” there and replace it by “are”), fitting them to a bit of a narrative Procrustean bed. And, to some extent, the reason why some of these didn’t turn out badly seems to me to be because of the sort of government intervention that Ridley dislikes; in general, he doesn’t seem to me to spend enough time on Tragedy of the Commons problems.

But there’s a lot in his criticisms of those movements to think about, and I agree whole-heartedly with one part of his criticism and approach to criticism, namely that you should get your metrics right. Metrics aren’t everything, but they’re important, and you shouldn’t cherry pick them to lead in the direction where you’d like the answer to be. (And his other point that there’s a lot of unwarranted pessimism going around is probably true, too.) I hope that the environmental movement isn’t as bad in this regard as Ridley claims (I certainly hear a lot more bad said than good about corn-based ethanol, for example), and the Azimuth Project (including blog posts like this one) makes me optimistic that good use of metrics will be (hopefully, already has been) a strong part of the discussion. But it’s an important enough point to deserve to be hammered home.

Certainly a book I’m glad to have read. (And it’s a fast read, too.) A hat tip to Jesse Schell for the recommendation.

Post Revisions:

This post has not been revised since publication.