The "money" in the economy is actually mostly credit,
    and during a recession it contracts

What arguments are there to support the theory that a recession (decline in overall productivity) causes the amount of credit in the system to shrink?

From a look at the M2 and M3 money supply, there never seems to have been a contraction:

https://fred.stlouisfed.org/series/M2

https://fred.stlouisfed.org/series/MABMM301USM189S

That seems to be a pretty strong argument against the money contraction theory.