Geography of innovation is another interesting topic. Here we show this might be studied, using the SMS platform. The example we chose is a core element of current science and innovation policy in the Netherlands, where a very large part of public research money is distributed to the so-called ‘top sectors’, that is the economic sectors of which Dutch government expects that they will be the core of future economic growth. Money is competitively distributed among consortia that focus on one of the top sectors (such as energy, water, chemistry, life sciences, logistics, etc.), and within the consortia companies have a leading role. Some information about the grated research consortia is available in the RVO project database.
Fig 36. Geography of innovation: data processing
As the ‘top sector policy’ is considered core in Dutch STI-policy, we are interested in how these research and innovation activities are distributed over the country. How could this be done using the SMS platform? Figure 36 represents the steps, and we will discuss them in some detail.
Figure 37 (left) shows the FUAs in the Netherlands. The map is produced using open data in the SMS platform. Figure 37 (right) shows how the projects are distributed over the FUAs, and the darker the the color the higher the number of projects.
Fig 37. FUAs in the Netherlands (left), and the distribution of projects over the FUAs (right)
Underlying the Functional Urban Areas is an idea of what are meaningful definitions of geographical boundaries. The FUA idea is based on the assumed importance of the distribution of people - population density and commuting patterns. However, a researcher may have good (theoretical) arguments for other geographical delineation. So why wouldn’t other characteristics not be more important? the SMS platform has several services to support researchers to use their own definition or geographical areas. These deploy to types of open data. Firstly, we use the availability of statistical data at different levels of aggregation. Second we use the availability of so-called ‘shape files’ - both available as open data.
Fig 38. Producing dedicated geo-boundaries
The idea boils down that the researcher defines the property or properties of regions he/she is interested in. An example could be ‘population density’. As statistical data are available, the combination of those data with the available geo-boundary data results in a ‘population density’ geography. Figure 39 (right) shows again the OECD-FUA geography of the Netherlands, whereas figure 39 (left) shows the population density (above a threshold). The two geographies are similar (as population plays an important role in the FUA), but not identical. One may, however, also use a different property such as ‘density of companies’, or a hybrid definition using as well population density and company density. These definitions result in different geographies, as figure 39 shows.
Fig 39. Alternative geo-boundaries
Fig 40: Geography of innovation depends on the selected geo-boundaries
The selection of the definition of geographical boundaries is not neutral when studying the geography of innovation. In figure 40, we show the density of innovation projects (from the ‘topsectoren’ database) for regions. And obviously, there are regions that do play a role in the innovation projects that are found when using our hybrid definition, but that would have remained invisible using the population density or the FUAs.
The total process for this analysis (see figure 41 for an overview) is rather complex, and requires several computational steps. So doing this may require collaboration between the social scientist who wants to investigate the geography of innovation, and a data scientists to support with the production and retrieval of the required data. But it gives also a flavor of the new possibilities that the SMS platform hopes to provide.
Fig 41. Overview of the data processing