This paper [pdf] explains why policy-makers choose regulation over taxes as a means of reducing negative externalities. (Read my post on Pigouvian taxes.)
Their main points are that taxes make it easy to shrink output (or the number of firms) to the correct level. Regulations do not and lead to too many firms competing inefficiently for a share of the regulated pie (e.g., for a share of permits in a cap and trade market for carbon).
Firms that prefer lobbying for a share of quotas to price competition, politicians who like to be lobbied and regulators who enjoy job security all prefer regulations. That "iron triangle" of interests explains why it's difficult to replace a complex system with a simpler one, and why the public gets screwed by regulatory failures.
But it gets worse: taxes are sometimes added to regulations, to either raise revenues or change relative prices. Those taxes may be a move in the right direction, but they usually just confuse things (e.g., taxing gasoline while regulating fuel efficiency). It would be better to have only taxes, but industry prefers regulations because -- as Buchanan and Tullock often discuss -- it can influence them.*
How could we overcome the iron triangle? We'd have to have politicians who cared more about the silent majority than special interests. That may happen in Northern Europe, but it's hard to see it happening in the US.
Bottom Line: Regulations may be the second- or third-best policy for society; we get them because they are the first best policy for special interests.
* This may explain why insurance companies lobbied so hard for Obamacare. Now we see that more people will be forced to pay more for coverage. FAIL.
Addendum: Tyler discusses the price of Obamacare in California.