I mostly agree, though I don't have a strong position on the correct answer. But as someone who does study complex systems sometimes in my day job, I find
that argument for libertarianism extremely unconvincing. Most complex systems you find do
not, left alone, function in any sort of reasonable way. They are full of crazy oscillating attractors, pathological feedback loops, more-stable and less-stable states that you (or an actor in the system) can deliberately manipulate it into, etc., etc. "Well-behaved" dynamical systems are comparatively hard to design even when you're deliberately designing, and ones you just find somewhere are usually not well behaved.
Which makes me particularly skeptical of "the market just works" type analysis, which claims that this particular complex dynamical system does not have all those pathologies, and in fact can be analyzed with fairly high-level, aggregate laws like the classical economics analysis of supply/demand adjustment. Heck, even fluid dynamics (another complex system) doesn't work that way--- all our general laws about how pressure/flow/velocity/etc. relate only apply to specific flow regimes, and can have different or even opposite-of-normal relationships in different flow regimes (especially anything turbulent).
One argument for why there should be some intervention would be: the classical-economics analysis of how economies should work is an appealing one, but is not in fact how they always work, so the government's job should be to dampen out the pathologies and turbulence that would take us out of the "normal" regime where the classical laws do apply. That's sort of the line taken by ordoliberalism. The question would be whether it can, on the whole, do more good than harm. But I wouldn't start from the assumption that what you get without government intervention operates "well" or "ideally".
Which makes me particularly skeptical of "the market just works" type analysis, which claims that this particular complex dynamical system does not have all those pathologies, and in fact can be analyzed with fairly high-level, aggregate laws like the classical economics analysis of supply/demand adjustment. Heck, even fluid dynamics (another complex system) doesn't work that way--- all our general laws about how pressure/flow/velocity/etc. relate only apply to specific flow regimes, and can have different or even opposite-of-normal relationships in different flow regimes (especially anything turbulent).
One argument for why there should be some intervention would be: the classical-economics analysis of how economies should work is an appealing one, but is not in fact how they always work, so the government's job should be to dampen out the pathologies and turbulence that would take us out of the "normal" regime where the classical laws do apply. That's sort of the line taken by ordoliberalism. The question would be whether it can, on the whole, do more good than harm. But I wouldn't start from the assumption that what you get without government intervention operates "well" or "ideally".