Patterns of diffusion of innovation

Grubler, Arnulf. 1996. “Time for a Change: on the Patterns of Diffusion of Innovation” Daedalus 125 (3): 19–42. doi:10.2307/20027369.

The spread of innovations in societies ranging from catalytic converters in automobiles to obligatory primary education follows similar rules. This is because random, complex and largely unpredictable micro-factors (fashion, cost, regulations, performance, habits etc.) aggregate to shape statistically recognisable macro-patterns.

After an invention becomes an innovation, it goes through a process of diffusion from the core to the periphery. Innovations achieve higher density in the core (e.g. cars in the US or Cisterian monasteries in France) but they are adopted faster on the periphery. Diffusion of innovation always follows an S-curve: slow ‘take-off’ in the beginning, followed by a rapid acceleration and then a saturation. However, the speed of adoption depends on the scope and scale of the innovation, for example:

  1. incremental improvements;
  2. radical change in technologies and artefacts;
  3. changes in technological systems (technologies + managerial and social systems);
  4. changes in clusters and families of technologies and associated institutional and social settings.

Whereas diffusion for (1) can take a decade or even a few years, (4) can take many decades or even centuries.

Comment: ‘regime changes’ addressed by the multi-level perspective fit (3) whereas energy transitions described by Smil fit (4) in the list above.

Towards the end, the article analyses the combined effect of many innovations and connects them to macro-economic and business cycles, particularly to Schumpeter’s theories.

POLET Observations

  • An individual country is the main unit for the analysis of the diffusion of innovations (the US is used most frequently). This is never explicitly stated but can probably be explained by the fact that most statistics by which such phenomena can be studied is assembled at the national level. Another explanation may be that factors influencing diffusion within a single country are more homogenous than internationally;
  • Political factors are among many other 'random' influences (ranging from fashions to technical breakthrough) that add up to the S-curves. The perspective does not attempt to single out or explain these.