One of the real enduring puzzles of American public policy is why our infrastructure costs are so high by international standards. A very useful Stephen Smith column on Bloomberg.com details some of the reasons, which mostly seem to come down to very poor project management. But read him for the details.
My hazy generalization is that a lot of local government in the United States seems to me to reflect a misguided overemphasis on procedural barriers to corruption to the exclusion of relying on democratic accountability. You could imagine a policy framework in which mayors were supposed to finance infrastructure projects entirely out of slush funds. If you wanted to take your entire slush fund and give it to a company owned by your wife's brother in one giant contract, then you'd be free to do so. The problem is that your opponent at the next campaign might say, "Hey, let's elect a mayor who doesn't blow the whole infrastructure fund on a sweetheart contract for his brother-in-law." The same kind of system that would give the mayor flexibility to be corrupt would also give him the flexibility to try to manage projects more intelligently.
Obviously the slush-fund model is going to have some problems, and we probably shouldn't leap all the way in that direction. But as Smith argues, to a surprising extent, lowest-price bidding rules fail to achieve the goal of low prices. They leave you wide open to getting screwed over, which means your proposals need to be extremely detailed to avoid getting screwed over. Extremely detailed proposals make it difficult for new entrants to bid for contracts and also mean that you need to hire a lot of consultants to write the proposal in the first place. Next thing you know, the whole thing becomes an incestuous circle of "consultants who consultant with consultants and advisers who advise advisers" to put the thing together.