Poolability and Aggregation Problems of Regional Innovation Data: An Application to Nanomaterial Patenting

Research and development (R&D) in the field of nanomaterials is expected to be a major driver of innovation and economic growth. In this respect, many countries, as national systems of innovation, have established support programs offering subsidies for industry- and government-funded R&D. Consequently, it is of great interest to understand which factors facilitate the creation of new technological knowledge. The existing literature has typically addressed this question by employing a knowledge production function based on firm-, regional- or even country-level data. Estimating the effects for the entire national system of innovation, however, implicitly assumes poolability of regional data. We apply our reasoning to Germany, which has well-known – and wide – regional disparities, for example between the former East and West. Based on analyses at the level of NUTS-3 regions, we find different knowledge production functions for the East and the West. Moreover, we investigate how our results are affected by the adoption of alternative aggregation levels. Our findings have implications for further research in the field, that is, a careful evaluation of poolability and aggregation is required before estimating knowledge production functions at the regional level. Policy considerations are offered as well.nanotechnology, patents, poolability, aggregation, Germany, spatial autocorrelation, spatial filtering