http://query.nytimes.com/gst/abstract.html?res=9A03EFDA1F3BE...
January 4, 1914, Sunday [...]
Several of New York City's skyscrapers are away out of plumb and have a tendency to emulate the famous leaning tower of Pisa according to the engineering department of the Building Bureau.
(Looking at the downmod pattern, though, there's someone who isn't liking the idea that regulation may not be perfect, and expressing that with downmods. Interestingly, my real point is that it can't be perfect for mathematical reasons, which is actually a completely different argument from whether or not it can improve things. But you do need to set out from a policy perspective with the goal of improving, not perfecting, because plans based on perfecting things always cause more trouble when they go wrong. The plan has no slack left over.)
An economy has a large number of intrinsically unstable parts. As they displace from the "norm", they will displace even further. For example, the mere fact of a commodity's price going up can cause it to go up further as others react. This is intrinsically unstable.
An unstable system can not run like a stable system. Jerks are mathematically inevitable. The only way to prevent it is to do one of two things: Remove all such things from the system, or create an instantaneous counter force for each of these things. Neither is possible economically. The second sounds appealing, but the thing is that it has to be instantaneous to work, and completely (inhumanly!) accurate, or it becomes simply one more destabilizing force.
You can see this even in economies with no human interaction; this is the source of nature's constant ebb and flow.
The only thing you can do is make things worse by convincing people that you can stop the ebb and flow of the economy, so they don't prepare, then the storm comes anyhow and is all the worse for people not being prepared.