Economics at a turning point(3)

After the previous 2 posts on my view on the economics science, it is time to present my view on propelling it forward. I will present a far more humble and empiric science of economy rather than the abstract science that it has become. My ideas are still in a very early stage so they may lack a full and thourough basis and therefore i would like to invite the readers of this blog to comment on the concept.

My general idea is to approach the economy as a set of different sectors (production consumption government etc.) as is common within economics. Here is where the twist is: instead of theorizing how the interaction between these blocks is and fitting the data to such an interaction, i would like to compute response functions between them. The response function, as borrowed from electrical enginering, does not need to be a neat mathematical sound function as all to often happens in classical economics. It would be a function that has probably lags and maybe even cases in which a system responds before the signal actually was given. This could happen with consumption when people, worry that a crisis might occur, adjust their spending accordingly.

This does not mean that the theory of economy would go to the trash, rather it could explain and smooth en the function. This will be very necessary because it is impossible to observe an economic impulse in abstraction, rather we will have to get it from observing the entire system. This means that there will be a lot of noise, that economic theory of all branches, be it classical evolutionary or something else, could help over come.

When the response functions of all important components from the economic system are know, we could use simulation (which we probably need because the system will be far to complex to analyze only analytically) to answer questions on all sorts of economic questions. Interestingly we could use a system like this also to test the robustness of our predictions by varying the initial conditions slightly (an approach commonly used in simulation). This means that we can predict the likelihood of major events like another financial crisis much better.

As I've said before, this idea is still in the early stages so i would like everybody to comment and expand on it.