Effect of changing business R&D expenditure mix on productivity of the South African national system of innovation

The paper examines the effect of the sharp reallocation of business R&D expenditure from experimental research to applied research, since the global financial crisis, on the productivity of the South African national system of innovation (NSI). Business R&D expenditure declined sharply, without reversal, which suggests structural change. The paper applies linear regression models to data on trademarks and patents obtained from the World Intellectual Property Office and business R&D expenditure obtained from the Human Sciences Research Council, South Africa, for the period 2004–2018. The paper finds that reallocation of R&D expenditure from experimental to applied research had detrimental effects on the productivity of the NSI. Counterfactually, reallocation of R&D spending from experimental research to basic research had positive effects on innovation performance. The paper also establishes structural change in the innovation-business R&D spending regime. Policymakers, through tax incentives and nudging, should incentivise reallocation of business R&D spending towards basic research, while adjusting downwards the allocation to experimental and applied research until they reach their optimal levels. Discovery of structural change in the innovation-business R&D expenditure regime, the identification of sub-optimality in business R&D spending allocation, and identifying distorted incentive structures, all bring new knowledge to scholarship on the national system of innovation.