Induced Innovations and Directed Technical Change

The question of how a change of the economic environment acts upon the process of technical change revolved around the idea that relative factor prices, conceived of as scarcity indexes, do have a significant impact on the rate and bias of innovations and technical change.

The real reason for the predominance of labour saving inventions is surely that... [a] change in the relative prices of factors of production is itself a spur to invention, and to invention of a particular kind - directed to economizing the use of a factor which has become relatively expensive. (Hicks 1932: 124)

Salter (1960) criticized the hypothesis and argued that firms implement innovations that reduce total costs and are not concerned with a particular innovation bias. In a similar vein Samuelson (1965: 354-5) stressed that “since at minimum-cost equilibrium all inputs are equally marginally dear or productive, the simplest theory of induced innova­tion is ill-founded.”

Despite these objections the induced innovation hypothesis found its way into growth theory in the 1960s (Kennedy 1964; von Weizsacker 1965; Samuelson 1965). Central to these models is the innovation possibility frontier (IPF), introduced by Kennedy (1964). The IPF describes technological restrictions on innovation possibilities. It is assumed to have the properties of a neoclassical transformation function and thus poses a funda­mental trade-off between the rate of reducing labour and capital requirements for unit production. Based on this concept, Kennedy showed that firms, seeking to maximize overall unit cost reduction, choose a combination of capital and labour saving innova­tions that depends on relative factor shares.

Nordhaus (1973) questioned the usefulness of Kennedy’s model, because it assumes (1) that innovating is a costless activity and (2) that the IPF is exogenously imposed and static such that the rate of reducing capital requirements does not depend on the level of reduction of labour requirements.

About 30 years later, Daron Acemoglu (2002) revived the idea of induced innovations within an endogenous growth model. He investigates the determinants and consequences of directed technical change: innovative activities, responding to profit incentives, are biased towards different factors of production. This is so because firms are likely to invest in the development of the most promising type of innovation (embodied in an intermediate good, for example, a machine) in terms of profitability. The latter is determined by two forces working in opposite directions: (1) the price effect, that is, the incentives to invest in a specific technology increase with the price of the good in whose production the technology is deployed; thus, the price effect implies technical change to be biased towards scarce factors of production; (2) the market size effect, that is, the larger the supply of a specific production factor, the more likely the development of a new technology that embodies a relatively large amount of this factor. This effect is similar to the market forces at work in the demand-pull model by Schmookler (1966), except that innovative activities are driven by factor markets, since the market size effect promotes technical change in favour of the (relatively) abundant factors.

Which of the two effects ultimately prevails depends on the substitutability between the different technologies (respective factors). For a sufficiently large elasticity of sub­stitution, the demand curve for an abundant factor might even be upward sloping. This was the case with regard to skilled labour in the US in the second half of the twentieth century, Acemoglu contended: the increase in qualified workers was a strong incen­tive to develop a technology that embodies a lot of human capital (and not the other way round). The resulting rate of skill-biased technical change induced a rise in the skill premium, outweighing the usual substitution effect between skilled and unskilled workers owing to this change in the relative price of the two kinds of labour. The model, also entailing the innovation possibility frontier, is said to also contribute to an explana­tion of income disparities between rich and poor countries and the potential innovation bias triggered by international trade.