Measuring the Private and Social Returns to R&D: Unintended Spillovers versus Technology Markets

Innovation and Economic Growth, Intellectual Property, Patents, Interoperability and Standards

Article Snapshot


Pere Arqué-Castells and Daniel Spulber


Journal of Political Economy, Vol. 130, No. 7, pp. 1860–1918, 2022


Traditional methods of estimating returns to research and development (R&D) ignore gains from selling or licensing intellectual property. Counting these gains shows they are an important source of returns to R&D.

Policy Relevance

Policymakers could consider revenues from technology markets in setting R&D subsidies.

Main Points

  • Firms share technology through voluntary transfers in technology markets (such as IP licensing) and involuntary spillovers (such as observations about other firms' technologies gleaned from conference presentations or patent applications).
  • The dominant approach for estimating returns to R&D assumes that a technology provider’s R&D and downstream product production are integrated, and ignores technology markets; however, increasingly, firms’ R&D and production are not integrated, and firms monetize R&D by selling or licensing intellectual property rights to other firms.
  • If spillovers are the main method of generating returns to R&D, policymakers could support R&D subsidies and tax credits to increase incentives to innovate; however, if tech transfers are the main method, policymakers might reduce R&D subsidies to avoid reinforcing a few “superstar” firms.
  • Returns to R&D include private returns (the revenues accruing to a technology provider because of its R&D) and social returns (the total increase in output generated by the R&D, including the technology provider’s revenue gains and technology adopters’ revenue gains after spillovers).
  • Updating the traditional model to include technology markets means that both private and social returns are larger.
    • Private returns to the technology provider also include revenues from technology markets.
    • The social returns to adopters include benefits of the technology transfer that the provider fails to capture.
    • The traditional model underestimates both private and social returns.
  • Data from technology providers and technology adopters collected from 1990 to 2014 shows that only a few firms adopt from or provide technologies to more than 50 different firms; a small number of these firms are responsible for most R&D investment.
  • According to traditional measures, the private returns and social returns to R&D differ by 39 percent; however, once technology markets are counted, both private and social returns are higher, and the wedge between social and private returns is only 30 percent.
  • More than 10 percent of total revenues gained through R&D are generated by exchange in technology markets; technology markets serve an important function in developing knowledge.

Get The Article

Find the full article online

Search for Full Article