This paper deals with the integration of Walras' law into Keynesian macroeconomics and the attempts at a consistent specification of period models (beginning- vs. end-of-period-equilibrium). Three examples are examined where neglect of a consistent specification led to erroneous results: (1) the identification of the IS-condition with equilibrium of the "flow market" for bonds, (2) superficial treatments of the liquidity trap, and (3) the assumptions on the stochastic structure of monetary and real shocks in determining the optimal monetary instrument. (author's abstract) / Series: Department of Economics Working Paper Series