In his valuable exposition of neo-classical aggregative growth theory,' Professor Harry Johnson shows how his ingenious graphical and analytical apparatus can be extended to accommodate outside money, that is, monetary debt of the government to its citizens. He begins with the Keynesian assumption that saving is a constant fraction of disposable income. Saving means accumulation of money or of real capital; whatever the relative amounts of the two forms of saving their total must satisfy the fixed propensity to save. However, the shares of money and capital stocks in total private wealth must satisfy asset preferences which depend inter alia on the real rates of return on the two assets. The money price of goods is assumed flexible, and its correctly anticipated rate of change is the negative of the real rate of return on money. Johnson concludes that money is neutral in this model, in the following senses. (1) Unanticipated one-shot injections or withdrawals of nominal money do not alter the real development or equilibrium of the economy; they merely change the price level instantaneously in the same proportion. Neutrality in this sense is a trivial consequence of price flexibility and characterizes all models of this genre. (2) Equilibrium capital intensity, rate of return on capital, and per capita income are the same as in a non-monetary model with the same technology, population growth, and propensity to save. The presence of money alters only the approach to this equilibrium. (3) Likewise, the real equilibrium is independent of asset preferences and of the real rate of return on money. Thus it is not affected by the manner in which the growth in the real value of money balances is split between increasing nominal money and declining price level. The rate at which government creates new money affects only the path to the equilibrium, not the equilibrium itself. Conclusions (2) and (3) contradict the conclusions I reached for essentially the same model.2 Johnson notes this contradiction (footnote 1, p. 279) and suiggests the reason for my erroneous conclusion.