In critique, Robert Solow (1956) argued that the real-world economy is not on a “knife-edge” (except for the Great Depression). Solow said that there must be some market mechanism that brings an economy back to equilibrium and the warranted growth rate when it deviates from them. Harrod’s model assumed that the labor–capital ratio is constant over time. But “if this assumption is abandoned, the knife-edge notion of unstable balance seems to go with it” (Solow 1956: 65). Solow found this assumption inconsistent with neoclassical economics. When firms have excess capacity (excess investment), they substitute labor for capital—Solow essentially modified the Harrod–Domar model to allow neoclassical factor substitution in production. According to Solow, there can be stable equilibrium growth if the growth model is set up with this correct neoclassical assumption. Solow tried three production functions and picked one (the Cobb–Douglas production function) because it theoretically generated stable equilibrium. Solow argued that there exists a rate of investment—balanced investment—that keeps the growth of the capital stock equal to the growth of the labor force. If actual investment exceeds balanced investment, the amount of capital per worker grows until it reaches a level consistent with full employment—what Solow called the steady-state point. Hence, Solow showed that the neoclassical growth model is stable. It has the self-adjusting mechanism that guarantees a return to equilibrium. Solow’s model was an attack on the Keynesian explanation of unstable economic growth.