Take by analogy this account of my trip home from my parents' house this afternoon.
I was driving home from my parents' house this afternoon and I was pulled over by a police officer and given a ticket. (Thankfully this last bit did not actually happen. At least, it did not happen today.)
The ticket was for speeding. I will now have to appear in court or pay a fine.
All these things are observations that illustrate operational facts about the relationship between a certain member of the public (me) and certain institutions of government.
If a police officer sees me driving over a certain speed, she will stop me and give me a ticket that will result in my having certain liabilities toward the government.
I may try to come up with some ideas, based on my experience, how I can avoid getting a ticket in the future. Any such analysis should be based on the operational facts above. If I am unaware as to the operational reason I incurred these liabilities I will be in trouble.
What does this have to do with macroeconomics? Well, a commentator on Yglesias' blog called halfkidding said recently:
So-called monetary policy with all its tools and mechanisms is used to do one thing. That is to increase or decrease the amount of credit in the system.
It's fine to believe, as many fine economists do, the theoretical assertion that monetary policy can increase or decrease the amount of credit in the banking system. However, it's crucial to retain the knowledge that in operational fact what the central bank is doing is setting the interest rate. There are various theoretical mechanisms by which this can increase or decrease the amount of lending in the system, but the magic Fed lever controls the PRICE of funds, not the AMOUNT of funds.
As a matter of operational fact banks determine the amount of credit (thus money) in the system by their lending decisions, which are influenced by the prevailing interest rate but also by the availability of creditworthy borrowers.
*At least I think it's strange. Researchers of other disciplines could probably say better than I can.