Inside the black box: Paul Romer’s endogenous growth theory

Earlier this month, American economist Paul Romer jointly won the Nobel Prize for economics with fellow American William Nordhaus for his contribution to integrating technological innovation with economic growth in the form of endogenous growth theory.  As an earlier prize winner Robert Lucas famously observed “Once you start thinking about growth, it’s hard to think about anything else.”

In response to the failure of neoclassical growth theory such as Robert Solow (1957) to adequately account for the ‘black box’ of innovation within growth, a group of economists in the 1990s including Paul Romer (1990), Robert Lucas (1988), Philippe Aghion and Peter Howitt (1992) sought to extend neoclassical models to incorporate endogenous innovation, hence “endogenous” growth theory.

Romer’s analysis begins with the insight that innovations are based on ideas and an intrinsic characteristic of ideas is that they are non-rivalrous.  That is, once an idea has been produced, anyone with knowledge of the idea can utilise it.  Non-rivalrous goods need to be produced only once because, by definition, more than one consumer can utilise the goods, even simultaneously.  This means non-rivalrous goods involve a fixed cost of production and zero marginal cost.  For example, it can cost a great deal to produce the first unit of a software program but, equipped with that knowledge, subsequent units cost little to reproduce.

Romer’s analysis led to a simple but powerful insight: as innovations are based on ideas and ideas are non-rivalrous, their producers cannot capture the full benefits or value unless they can capture monopoly rents for their ideas through devices such as patents.  No firm would invest the often vast resources required to develop ideas in a perfectly competitive market because it would not be able to capture the returns required to justify the investment.  Therefore, the fact that producers of ideas can price them above the marginal cost of production means that they receive increasing (not constant or diminishing) returns to scale.  The presence of increasing returns to scale, according to economic theory, indicates the presence of imperfect competition.  To continue with the software example: whilst the very first unit of the program is expensive to produce, each additional copy costs very little but is sold at a high mark-up.

Furthermore, argues Romer, ideas can be a classic positive externality because, once released, their producers may not be able to exclude others from utilising them.  Consequently, the market will tend to supply a less than socially optimal level, as markets tend to do with positive externalities, resulting in market failure.  Therefore, as the long run rate of economic growth depends on the rate at which innovation and knowledge grow, the public sector has a vital role to play in supplementing the private provision of research and education to correct this market failure.

Endogenous growth theory can be understood as the addition of human capital as a separate form of capital to Solow’s (1957) growth model.  This appears as the addition of human capital (H) to the statement: Output (Y) is a function (f) of capital (K), labour (L), and innovation (A) so that Y = Af(K, H, L).

However, the decisive difference between neoclassical growth theory and endogenous growth theory is that the latter allows for the possibility of increasing returns, whereby if all inputs are doubled, output may more than double.  As Romer explains, this is because knowledge is not completely excludable, so that a doubling in the stock of knowledge can result in more than doubling of its productive utilisations by firms.

Despite its intuitive appeal, endogenous growth theory is not without its critics.  Perhaps its strongest is Setterfield (2003) who observes that since endogenous growth theorists can connect capital, labour, innovation and human capital to the rate of growth, there is little that cannot be connected to growth, making it difficult to establish precisely what the determinants of growth are.