Diffusion of innovation and innovation systems is a family of concepts that explain how new ideas move from invention into everyday use and who gains along the way. An invention changes little until firms and households adopt it, and adoption is usually slow, uneven and shaped by knowledge that is hard to copy. The family covers the spread of new technology, the special case of general-purpose technologies, tacit knowledge, the temporary profits that innovators earn, and the networks of institutions that create and spread technology in a country.
Economists, historians and policy makers use these ideas to explain why productivity growth can lag behind visible technical change and why some countries catch up faster than others. They sit alongside the wiki's pages on innovation (economics), types of innovation and economic rent. In finance, the same patterns describe the spread of card payments, online banking and mobile money.
Diffusion of Innovation
Diffusion of innovation is the process by which a new product, process or practice spreads through an economy over time. Adoption typically follows an S-shaped curve: a few early users, then rapid take-up as costs fall and the technology improves, then a slowdown as the remaining non-users are the hardest to convert. Zvi Griliches's 1957 study of hybrid corn in the United States showed such curves differing by region according to expected profit. The sociologist Everett Rogers, in Diffusion of Innovations (1962), grouped adopters into innovators, early adopters, early majority, late majority and laggards.
The gap between invention and wide use, sometimes called the diffusion lag, matters because an economy's productivity depends on how widely the best methods are used, not on whether they exist somewhere. Diffusion speed depends on price, complementary skills and infrastructure, and on network effects. Kenya's M-Pesa, launched in 2007, spread quickly because each new user made the service more useful to others.
General-Purpose Technology
A general-purpose technology (GPT) is one that is used across many sectors, keeps improving and sparks further inventions in the industries that adopt it. The term was set out by Timothy Bresnahan and Manuel Trajtenberg in a 1995 paper. Standard examples are the steam engine, electricity and information and communication technology; many economists argue that artificial intelligence may become one.
GPTs tend to raise productivity slowly at first. Firms must reorganise work, retrain staff and invest in complementary equipment before the gains show. The economic historian Paul David's 1990 essay "The Dynamo and the Computer" compared the decades-long lag in factory electrification with the slow payoff from computers, which Robert Solow had summed up in 1987 by remarking that the computer age was visible everywhere except in the productivity statistics. The Industrial Revolution itself is often described as the first GPT-driven transformation.
Tacit Knowledge
Tacit knowledge is know-how that people possess but cannot easily write down or pass on through documents. The philosopher Michael Polanyi summed up the idea in the observation that people know more than they can tell, developed in The Tacit Dimension (1966). Riding a bicycle, judging a borrower's character or running a complex production line all rely on skills learned by practice and observation.
The contrast is codified knowledge, which can be recorded in manuals, blueprints, software or patents and transferred at low cost. Tacit knowledge explains why copying a successful firm is harder than reading its patents, why skilled staff are valuable and why industries cluster in places where workers can learn from each other. It also slows the diffusion of innovation across borders, because the know-how travels with people rather than with documents. Firms protect some of it as a trade secret, and much of it forms part of human capital.
Innovation Rent
An innovation rent is the temporary above-normal profit that a firm earns from a new product or process before rivals can imitate it. Because the idea is associated with Joseph Schumpeter, such profits are also known as Schumpeterian rents. In Schumpeter's account of creative destruction, the prospect of these rents is what drives firms to innovate, and their erosion is what spreads the benefits to consumers.
Rents last as long as the innovator can hold off competitors. A patent gives a legal monopoly, normally for 20 years from filing in countries that follow the WTO's TRIPS agreement. Other sources include first-mover advantage, brand loyalty, tacit know-how and network effects. Once imitators enter, prices fall and the rent is competed away, which is why innovation rents differ from a lasting monopoly rent. In payments, the early margins on a new service often shrink once rival schemes copy its features.
National Innovation System
A national innovation system is the network of firms, universities, public research laboratories, funders, standards bodies and policies that together create, adapt and spread new technology within a country. The concept was developed in the late 1980s and early 1990s by Christopher Freeman, Bengt-Åke Lundvall and Richard Nelson, who argued that innovation depends on links between institutions, not just on the effort of individual firms.
Studies of national systems compare how countries organise research and development, technical education, finance for new firms and the flow of people between universities and industry. Germany is often cited for its applied research institutes and vocational training, South Korea for close coordination between government and large industrial groups, and the United States for university research and deep venture capital markets. The idea links to endogenous growth theory and to the study of catch-up development.
How the members relate
Innovation rents give firms a reason to invent. General-purpose technologies create the largest opportunities for such rents across the economy. Tacit knowledge slows imitation, which protects rents but also slows diffusion. National innovation systems shape how fast new knowledge is produced and how widely it spreads, and diffusion determines when the gains finally appear in aggregate productivity and economic growth.
Ref 1100001663