New Data for Innovation Policy

We contend that existing datasets based on survey or administrative data queried via standard industrial and occupational standards are ill suited for the purposes of innovation policy, and therefore, of its ability to attain its goals to address market, system and emergence failures preventing new ideas from being applied. These datasets are constrained in their ability to identify novel sectors, map innovation networks, characterise complex ecosystems, and operationally target innovation interventions. New data – big data, web data and open data – data combinations and interactive mapping and reporting tools could help address some of the limitations. We review recent studies that have used these methods, and our own experience with them. We also discuss challenges – related to data quality and trust, relevance and the need to develop complementary capabilities and processes in government, which stand on the way of their more extensive application and impact. We hypothesise that although new data can greatly improve the effectiveness of innovation policy, some defining features of this policy domain – low volume of decisions, and high levels of uncertainty – may limit its transformational impact compared to other policy areas.