Big Data for better innovation policy
What do governments need to know to get innovation policy right? And what can Big Data do to help? We're kicking off a set of interesting new projects to answer exactly this question.
Governments do many things to try and make innovation flourish: they fund research, they back entrepreneurs with grants and tax breaks, they try (not always successfully) to support clusters. They also gather data about what they need to do and whether it's working: R&D figures, business surveys, resources like the Gateway to Research.
Having the right information is really important if you're trying to help innovative businesses. Is the Midsomer widgets cluster really a cluster, or just a PR exercise? Is te Flugelbinder sector really competitive, or a lame duck that government should get its hands off? Is research into Unobtainium bringing any commercial benefits to the UK? Indeed, an argument used by opponents of innovation policy is that government can never know enough about what's going on to do the right thing.
Nesta has done its bit in the past to improve the information governments use to make innovation policy, measuring things from hidden innovation to high-growth firms to the absorptive capacity of different regions. But existing data is pretty bad when it comes to new, fast-moving businesses. There's a SIC code for whale oil production, but good luck finding one for video games development or graphene. And most official data comes out late, and only once a year.
The point of this new project is to test how new possibilities for understanding innovation are opened up by Big Data. We use the term "Big Data" cautiously, because it's been the beneficiary of a lot of hype in the past two years. But we don't believe it's going away. We take Big Data to include the use of new sources of large volumes of data (such as social media and online collaboration sites), \"data exhaust\" (interesting information generated as a by-product of doing something else), and new means of analysis that can deal with them (from text mining to social network analysis to sentiment analysis), often in an automated way in close to real time.
In theory, Big Data could provide information into all sorts of things that policymakers would be interested in. What types of new jobs are people hiring for and where? To what extent are coders at start-ups connected to large UK corporates? How are various clusters growing? How many firms are working on the government's "eight great technologies"?
We're not the only ones to be thinking about this. San Francisco-based Quid has been mining patent and start-up data to develop insights about innovation. Both East London and Cambridge have launched online maps to show the evolution of their high-tech clusters. Firms like Growth Intelligence and Mastodon C are working with public bodies to help them understand the landscape for their policies better. We've funded research on different quantitative methods businesses and governments have been using to try and understand the future (here's a preview). There's even been talk of an iPad data dashboard for the biggest UK policymaker of them all.
We've made four research grants to test out interesting methodologies for developing these kinds of insights. Maybe they won't work, maybe they will, but we hope to learn something from all of them, and to apply the lessons to the bigger questions of innovation policy. When it comes to the impact of data science on policy, we suspect that this is just the beginning.
If you're a researcher or a business working on these kinds of techniques, we'd love to hear from you to discuss your and our work. If you're a policymaker concerned with innovation or business, we'd be interested to hear from you too: what information could help you do your job better?
Here are the details of new projects and the partners we're working with, all of which are taking place over the coming 12 months.
Project title: "Mapping innovation and business growth in a strategic emerging technology: New data sources and methods for real-time intelligence on graphene enterprise development and commercialisation"
Principal investigator: Philip Shapira, Manchester Institute of Innovation Research, Manchester Business School, University of Manchester
Co-Principal Investigator: Sophia Ananiadou, School of Computer Science, Manchester Interdiscipinary Biocentre, National Centre for Text Mining, University of Manchester
This project aims to study the commercialisation and enterprise development associated with graphene, especially the innovation activity of smaller firms that are poorly covered by other data sources. This will involve combining information from three sources: unstructured enterprise webpages, unstructured data from Twitter; and data from established structured databases, including patenting data. The aim is to develop methods and data sources for real-time intelligence to understand and map enterprise development in a rapidly emerging and growing new technology.
Project title: "Using big data to find innovators"
Investigators: Francine Bennett, Mastodon C;
Chris Taggart, Open Corporates.
This project is investigating innovation and growth in high-technology software firms. Working with several sets of public data which offer a window into the software development activities of UK high-tech firms and the individuals involved in them, the project will use unstructured big data to build a more accessible picture of what is happening in these fast-moving companies. Data sources are likely to include professional programming resources such as Github and StackOverflow and social networks including twitter.
Project title: "Exploring the drivers of the UK digital economy"
Investigators: Dr Max Nathan, National Institute of Economic and Social Research (NIESR) and London School of Economics (LSE);
Tom Gatten, Growth Intelligence
This project is about mapping the digital economy across the UK using Growth Intelligence's unique real-time performance analysis for millions of businesses, especially its high-growth and innovative firms, and identifying determinants of their performance. This work builds on analysis by NIESR and Growth Intel for Google, and previous work by NIESR for Nesta on high-growth firms and city-regional performance. This work aims to combine very large datasets and cutting-edge econometric techniques to explore drivers of firms' innovative activity and growth. This team is also being co-funded by the Centre for London to develop work focused on the digital economy in the South-East.
Project title: "Using Want-Ad Data to Map Jobs and Economic Activity related to Innovative Technologies"
Investigators: Dr Michael Mandel, Progressive Policy Institute, Mack Center for Technological Innovation (Wharton) and South Mountain Economics
Judith Scherer, South Mountain Economics
This project will develop a UK-specific methodology based on pioneering work done in the U.S. on 'App Economy' jobs. It will use publicly available online data on help-wanted advertisements to address the problem of mapping job growth and economic impact for key innovative technologies. Government labour market surveys are not detailed enough to identify jobs in specific technology areas. The work will identify and map UK job clusters for two or three innovative technologies. In addition, it will show how this approach can be used to compare the number of jobs related to an innovative technology in the US and the UK.
For queries relating to the above please contact [email protected]