2.2 Disaggregating Higher Education, Public, and Business R&D Expenditure
To isolate the structural determinants of sub-national economic performance outside core metropolitan agglomerations, the econometric methodology employs a longitudinal panel specification that systematically decomposes research and development expenditure into higher education, business, and public funding categories. Disaggregating R&D inputs proves essential because different institutional sources of research funding exert asymmetric effects on regional innovation capacity and localized economic output. Empirical evidence demonstrates that higher education R&D expenditure operates as a primary driver of patent production and regional innovation capacity over extended time horizons, whereas public and business research expenditures can exhibit divergent or statistically insignificant direct returns depending on the spatial context (crossref-10-71014-sieds-v80i4-562). Furthermore, persistent regional productivity disparities across the United Kingdom reflect acute imbalances in public innovation policy, wherein public support for research and development remains heavily concentrated in the South East, leaving non-core conurbations constrained by structural bottlenecks and a specific relative shortage of specialized STEM skills (W4323267612). By incorporating sectoral R&D disaggregation into the econometric panel model, the proposed research design explicitly controls for unobserved regional fixed effects and tests whether university-led knowledge creation compensates for the historical scarcity of concentrated public research investments in peripheral territories. Consequently, this empirical methodology isolates the specific elasticity of regional patenting and productivity relative to higher education research expenditure, providing a robust analytical framework for evaluating how targeted spatial rebalancing of public research resources can mitigate persistent economic divergence between leading metropolitan innovation hubs and peripheral industrial regions.