Scientometric laws connecting publication counts to national research funding

Scientometric laws like those of Lotka, Bradford, and Zipf provide useful models for the behavior of indicators. Here additional laws are proposed that connect scientific publication counts to national funding of research and development (GERD). The laws are based on experimental evidence of what is clearly a causal relationship: funding is necessary, if not always sufficient, to conduct publishable research. The evidence comes from integer and fractional counts from Scopus and Web of Science. The explanatory variables come from a UNESCO data set that provided data from 93 countries; a subset of 43 industrialized nations from OECD, and another of 50 less industrialized nations. Models were built from cross sectional data plus panel data models combining longitudinal and cross-sectional data. GERD was shown to be an excellent explanatory variable. If a second explanatory variable is added to the model, the number of researchers adds some precision. Applications include forecasting publication counts from published funding plans, estimating the funding required for a nation to improve its publication performance, and using models for “what-if” experimentation.