After I-35 bridge collapse in Minneapolis on August 1, 2007 Geoff Ficke wrote an article accusing the government in inability to properly build and maintain infrastructure on the whole and bridges in particular. He contrasted it with the Great Northern railroad built after the Civil War by a tycoon James J. Hill which bed, bridges and tunnels are still in use and none collapsed so far.
I do not think that the government is entirely to blame here especially because the government itself did not construct the bridge on I-35. It was done by a private company, which won a government contract. The root cause of the catastrophe is deeper than the banal government inaptitude.
There were many accidents and catastrophes (including collapse of bridges) on railroads in the first half of XIX century. Every catastrophe caused engineers to increase the safety factor until it eventually became so big that railroads became pretty reliable. In contrast, there were not many collapses of automobile bridges in the beginning of XX century when their construction commenced. This led engineers to believe that the safety factors they used in automobile bridges design are pretty high and need not be further increased. With safety thereby secured priority was given to cost-efficiency.
This is a typical example of changing design priorities in the course of a system's evolution. At the early stages of any system's evolution the priority is given to its reliability and performance characteristics which distinguish the new system from the previous ones. Just recall the motto of the early days of aviation: "go farther, higher, faster".
Then the quest for maximizing absolute performance characteristics (such as speed, power, etc.) gives way to optimizing specific performance characteristics (such as specific power, etc.) and cost efficiency. Since the latter and the former are contradictory parameters, change in priorities result in lowering reliability. This is the real cause of I-35 bridge collapse and not inability of the government to do anything well.
Evolution of design priorities in the course of a system's evolution has never been the topic of a research in TRIZ. Moreover, this type of evolution was overlooked by TRIZ. The time has come to start researching it.