Keynes turned the theory of interest inside out by denying its premise. Interest, he argued, is not the reward for saving, and not the reward for waiting — not a real phenomenon arising from impatience or the productivity of capital — but a monetary one: the reward for parting with liquidity, the price exacted for giving up ready money.
The argument begins from a fact the real theories had set aside: that money is the most liquid of assets, holdable against an uncertain future, and that people, facing that uncertainty, prefer to keep some of their wealth in liquid form. They hold money for three reasons — to make ordinary transactions, as a precaution against the unexpected, and to speculate on the future course of asset prices. To induce someone to part with money — to lend it, to lock it into a bond — he must be paid, and the interest rate is the payment: the price that brings the public's demand to hold money, its liquidity preference, into balance with the quantity of money in existence. Interest, on this account, is determined in the market for money, not in the market for goods; it is the price of liquidity, not the reward for thrift.1
The break with the classical theory is sharp and consequential. The classical and Fisherian view made interest the price that equilibrates saving and investment — the reward that calls forth abstinence and rations it among investment projects. Keynes denied this whole function. Saving and investment, he argued, are brought into equality not by the interest rate but by changes in income: when investment rises, income rises through the multiplier until the extra saving it generates matches the investment, with no necessary work left for the rate to do. The rate is thereby freed from the task of balancing thrift against productive opportunity, and assigned instead to the money market. And because it is set there, by liquidity preference and the quantity of money, it can settle at a level the real economy cannot abide — too high to call forth the investment that full employment would require — with no automatic force to bring it down. The rate, for Keynes, could simply be wrong.2
Three implications follow, each carrying the rate further from a fact of nature. First, interest is monetary: it can be altered by changing the quantity of money, with no change whatever in thrift or in the productivity of capital — a possibility unintelligible on the real theories. Second, the rate is in part conventional, sustained by belief. Liquidity preference depends on expectations of the future rate: if holders expect the rate to rise, and bond prices to fall, they hold money rather than bonds, and their expectation, by raising the demand for money, itself helps hold the rate up — so that the rate rests partly on the mere expectation that it will stay where it is, a highly conventional magnitude, in Keynes's phrase, held aloft by opinion about itself. Third, in the extreme, this yields the liquidity trap: when the rate is already very low, almost everyone expects it to rise, the demand to hold money grows nearly insatiable, and the rate cannot be pushed lower however much money is created — the point at which the monetary lever loses its grip.3
Keynes's account did not go unchallenged. The loanable-funds theory, defended by Robertson and the classics, held that interest remains a real price after all — the price that clears a market for loanable funds, supplied by saving and demanded for investment. A long technical debate followed over which market truly sets the rate, the market for money or the market for loanable funds; the eventual textbook settlement, Hicks's IS–LM apparatus, showed that the rate and the level of income are jointly determined by the real side, saving and investment, and the monetary side, liquidity preference and the money supply, together — so that neither theory is complete alone. But the distinctively Keynesian point survived the synthesis intact: the rate has a monetary, conventional, expectational dimension that the real theories had altogether missed, and it is not, at bottom, simply the price of waiting.4
And this is the hinge on which the whole volume now turns. Once interest is monetary — once it is the price of parting with liquidity, settled where the demand to hold money meets the supply of it — a conclusion follows that the real theories could never have reached: that whoever controls the supply of money controls the rate. If the rate is set in the money market, then an institution that can create and withdraw money at will can move the rate wherever it chooses, by making liquidity abundant or scarce. Keynes made interest a monetary phenomenon; the central bank, master of the money supply, makes it an administered one. The price of liquidity becomes a lever, and the hand on the lever is a public authority's.5
Keynes severed interest from thrift and from the productivity of capital and tied it instead to money and to expectation. The price of time became the price of liquidity — a price that could be moved by changing the quantity of money, and held up by nothing more solid than the belief that it would hold. With that severing, the rate ceased to be a fact the market quietly discovered about the deep forces of the real economy, and became something far closer to a policy lever. The next plate follows the lever to the hand that works it, and finds that the price of time, in the modern economy, is not discovered at all. It is set.