Definition
Learning effects refer to efficiency or performance gains that arise from cumulative experiences and a prolonged learning process1. They explain the reduction in friction in a task that happens over time as a result of the technological improvements and economies of scale that are achieved over time. In digital markets, they operate through self-reinforcing feedback loops wherein increased data generation leads to improvement in service, which attracts more users, which in turn leads to further improvement as more user interaction generates more data.
Commentary
Origin of the term
The concept of learning effects has its roots in the learning curve literature, primarily in German psychologist Hermann Ebbinghaus’s work, who described what we call the ‘learning curve’ and highlighted that the time taken to relearn something decreases each time the relearning occurs, which happens to be a foundational principle of the learning curve theory.2 This was further developed by a prominent U.S. aeronautical engineer, Theodore Paul Wright, in his 1936 paper, ‘Factors Affecting the Cost of Airplanes’, which explained an inverse relation between aeroplane labour hours and production costs and production quantities because of factors like increased worker proficiency, reduced waste generation and material costs.3
The concept has been further developed by an American economist and mathematician Kenneth J. Arrow in his 1962 paper, ‘Economic Implications of Learning by Doing’, in which he argued that economic growth is driven by learning generated by production and investment, and that it generates both private and social benefits.4
Learning effects may arise across a wide range of sectors, including traditional industries, digital markets, and AI downstream markets.
Relevance vis-a-vis Competition Law
Learning effects hold significant relevance in the competition law realm as they lead to market inefficiencies, tilting the market in favour of an entity not because of better products or innovation compared to its competitors, but because of the disproportionate access to resources that it has in terms of the database, which helps it gain more value and deny other competitors a level-playing field. This also has repercussions for the consumers who, while interacting with digital platforms, are consciously or unconsciously forced into adverse selection. It also reinforces consumer lock-in by making switching less desirable and disincentivising for veteran clients and long-standing users. All this cumulatively leads to creating a hindrance in undistorted market competition and has the potential to significantly increase static entry barriers for the prospective and desirous entrants, while strengthening the position of the existing player who is improving every day because of data aggregation.
Successfully combating the ill effects of learning effects requires that the regulators understand how the traditional digital platform markets and the new AI downstream markets work (which unlike the former might have diminishing returns on more data), and craft regulations that do not merely impose blanket prohibitions on data combinations and access, but integrate computational techniques to adequately balance both the improvement of the model and healthy market competition.
The concept of Learning Effects is closely related and interwoven with other concepts such as Learning Curve Effect5, Learning by Doing Effects6, Experience Curve Effect.7
Tejvan Pettinger, ‘Learning effect- definition, explanation’ (Economics Help, 1 April 2002) https://www.economicshelp.org/blog/168243/economics/learning-effect-definition-explanation/↩︎
Ungvarsky, Janine, ‘Learning curve theory’ (EBSCO Research Staters,2020) https://www.ebsco.com/research-starters/education/learning-curve-theory↩︎
T.P. Wright, ‘Factors Affecting the Cost of Airplanes’ (1936) 3 Journal of the Aeronautical Sciences https://pdodds.w3.uvm.edu/research/papers/others/1936/wright1936a.pdf↩︎
Kenneth J Arrow, ‘The Economic Implications of Learning by Doing’ (1962) 29(3) Review of Economic Studies 155 https://www.jstor.org/stable/2295952↩︎
ACCA, ‘The learning rate and learning effect’ (ACCA Global) https://www.accaglobal.com/gb/en/student/exam-support-resources/fundamentals-exams-study-resources/f5/technical-articles/the-learning-rate-and-learning-effect.html↩︎
Zimmer, Scott, JD, ‘Learning-by-doing (economics)’ (EBSCO Research Starters, 2025) https://www.ebsco.com/research-starters/economics/learning-doing-economics↩︎
‘Experience Curve Effect’ (Universal Marketing Dictionary) https://marketing-dictionary.org/e/experience-curve-effect/↩︎
