1) Not subsidizing anyone, ever. Actually not subsidizing can be non-optimal in a wide range of situations, although in general intervention is questionable.

1) Not subsidizing anyone, ever. Actually not subsidizing can be non-optimal in a wide range of situations, although in general intervention is questionable. We'd have no tech industry at all in Asia if it weren't for HEAVY industry subsidization in early stages in Japan, Korea, and Taiwan, followed by the wildly successful Chinese special export zones (which are, to some extent, just a loosening of laws, but also included economic stimulus). This same pattern has emerged worldwide in a number of other industries (Taiwanese Orchids and African textiles off the top of my head) that otherwise wouldn't have appeared at all. If you'd like some better cites on this you could do worse than to start with the work of Dani Rodrik: http://ksghome.harvard.edu/~drodrik/papers.html http://ksghome.harvard.edu/~drodrik/papers.html The government doesn't have to pick winners in terms of single companies or technologies, but subsidizing whole sectors is often quite useful. 2) Increasing transparency of its laws Totally agree here 3) Taxing negative externalities rather than positive ones (so tax carbon instead of capital gains) Negative Externalities are often extremely difficult to value in the free market (the original Cap&Trade systems had wild swings in price until the market matured via gov't intervention), so that can be a sticking point. Not everything can be cleanly auctioned for value, whereas positive externalities in the form of profit are usually clearly valued. 4) Removing market distortions like the immigration issue discussed in the article. Agreed 1) Subsidizing some part of the industry which therefore damages the innovation that would out-compete it. Which is why an approach is needed that doesn't pick winners, but is equally helpful to a wide range of (generally small) companies. This is particularly true if you can't get competing countries to stop subsidizing. The 'second kind' of policies arise not necessarily from a desire to be 'world leading', but because lawmakers have a different criteria of what is positive than you do. Many things that are strictly profit increasing are not welfare increasing, whether for the better (Save AIG because they reinsure half of the damn country and insurance rates would soar), or for the worse (Save GM because unions are politically powerful and raise my chances of re-election). I feel like many HN'ers (although not necessarily you) went out and took econ 101 and 102, decided it was mathematical proof for strict Randian libertarianism, and now deride any government. This simplified view isn't well supported even in economics, particularly since so many problems in real life fall into either collective action or non-rational behavior categories. </rant>