Romer's Endogenous Technological Progress Model

Core Idea

Romer's model features R&D firms that create new products or improve existing ones, generating increasing returns to scale and sustaining long-run growth. Monopolistic competition in the intermediate goods sector provides profit incentives for innovation while creating efficiency losses from market power and duplication. The model shows how market size, R&D intensity, innovation rates, and human capital jointly determine long-run growth and explains cross-country income differences through R&D investments.

Explainer

From endogenous growth theory, you know that long-run growth can arise from decisions within the economy rather than from exogenous technological improvement falling from the sky. Romer's 1990 model makes this concrete by asking: why would anyone invest resources in creating new ideas? The answer hinges on a crucial property of ideas — they are nonrival. A blueprint for a better engine can be used by any number of factories simultaneously without being depleted. This nonrivalry means that ideas generate increasing returns to scale at the economy-wide level, even if individual firms face constant or diminishing returns. But nonrivalry alone is not enough to motivate private investment. Ideas must also be at least partially excludable — inventors need some ability to profit from their creations, or no one would bother doing R&D.

Romer resolves this tension by splitting the economy into three sectors. A final goods sector uses labor and a variety of intermediate inputs to produce output under perfect competition. An intermediate goods sector consists of firms that each hold a patent on a unique input variety and sell it at a markup — this is where monopolistic competition enters. Each intermediate firm faces a downward-sloping demand curve because its product is differentiated, giving it pricing power and positive profits. Those profits are what attract resources into the third sector: the R&D sector, which uses human capital to produce new blueprints. When a researcher invents a new variety of intermediate good, they receive a patent and become a monopolist for that variety. The expected present value of future monopoly profits is what compensates them for the cost of research.

The growth mechanism works as follows. More researchers produce more blueprints, which means a greater variety of intermediate inputs available to the final goods sector. Greater variety raises total factor productivity — the economy produces more output from the same labor and capital. This is the "expanding variety" interpretation of technological progress. The steady-state growth rate of the economy depends on how much human capital the economy allocates to R&D versus production, which in turn depends on the size of the market (larger markets mean more profits from each invention), the productivity of the research sector, and the interest rate (which determines how heavily future profits are discounted).

A striking implication is the scale effect: larger economies grow faster because a larger market increases the return to inventing. This prediction has generated significant empirical debate, since countries like Luxembourg do not obviously grow slower than the United States. Later models by Jones and others modified Romer's framework to eliminate strong scale effects while preserving endogenous innovation. Nevertheless, Romer's core insight endures — growth is not manna from heaven but the result of purposeful, profit-motivated investment in ideas. The model also reveals a fundamental market failure: because inventors cannot capture all the spillover benefits their ideas create (other researchers build on existing knowledge), the decentralized equilibrium underinvests in R&D relative to the social optimum, providing a rationale for public subsidies to research.