Epistemic institutional sclerosis and the middle-income trap: evidence from China’s intellectual capital imbalance

The dataset is a cross-country panel-style comparative dataset covering 62 countries and 58 academic disciplines, aggregated into three epistemic domains—sciences, social sciences, and creative arts—and then further collapsed into two broad categories, sciences versus arts, to examine how national intellectual capital composition relates to development outcomes. The main independent variables are per-capita measures of discipline-based publication output, alongside country dummy variables for China and the USA relative to the rest of the world, while the dependent variables are GDP per capita as a measure of economic quantity and anxiety prevalence (reversed in some models as a proxy for mental well-being) as a measure of economic quality. The study also reports strong internal consistency for these knowledge domains, with Cronbach’s alpha of 0.93 for sciences, 0.97 for social sciences, 0.91 for creative arts, and 0.99 for the combined arts category, supporting the aggregation strategy. The data are analysed using GLM with robust standard errors and multivariate regression to compare whether China and the USA differ from the world average in the extent to which their growth and mental-health outcomes are associated with science-heavy or arts-inclusive intellectual capital structures.