But you're missing the point when you talk about the aggregate. When only certain firms are succeeding and certain ones failing, you don't try to prop up the failing in order to raise aggregate demand.
Until spillover occurs. Not all banks were failing this time last year, however if we were to let more banks fail, that would have caused even healthy institutions to become insolvent. Ever heard of the term "animal spirits"? I am a little bit confused in your "prop up" statement as i am unsure if it pertains to bailouts or fiscal stimulus (no they are not the same).
What we do know is this: spending decreases, hiring decreases, and job losses increase. This fall in spending sends ripple effects throughout the economy leading to even greater job loss until economic sentiment improves. Economic sentiment is largely psychological and can be steered by improving demand.
You allow production and investment to shift toward those firms that are succeeding.
What? Production and investment does shift to the acute percentages of business (who are weathering the storm). Your error comes in the form of proportional expectations. You are assuming that the fall in "others" production is made up proportionally by the "succeeding". It does not work this way. The "aggregate fall" exponentially exceeds the productive capacity of those who succeed. You are using static analysis to explain dynamic economic cycles.
If it's failing then people don't want it as much as was previously thought.
But why? Even if income remains the same, there is a serious demand gap of which can be illustrated by increased saving. Therefore the "fail" is not due to consumer sentiment, but fear of future economic expectations.
Here is the issue Tony: You might be unaware, but your idea draws from the classical case, in which in the LONG RUN markets self correct. I agree, they do self correct in the long run.
However in the short run, they do not self correct. As we wait for the self correction, many people will suffer in the form of job losses, poverty, poor health (stress), failed marriages, declining birth rates, falling wages, diminished human capital (in the form of skills), etc....
The question then shifts to: how much suffering are you willing to allow for the economy to self correct? Remember, in the long run we are all dead.