Where innovation develops

New knowledge, which is the basis for innovation, is created where regions create the prerequisites for them. This know-how is not easily transferable, as current research shows. The graph shows research networks in Europe. Source: AIT/Scherngell

What induces firms to settle in certain areas and what effects this has on the region is a central topic in the field of economic geography. The issues involved are of great political relevance in the context of long-term job security. According to the traditional view of economic geography, labour and capital are the two factors which determine the appeal of a business location, i.e. what costs companies must expect and whether they will be able to find manpower under favourable conditions. For some time now, another factor has become very relevant, namely the ability to produce new knowledge and translate it into innovation. Despite the fact that communication media are improving constantly, a great deal of valuable knowledge – particularly specialist know-how relating to certain branches of industry – is confined to specific regions, since this type of knowledge can only be transferred through personal exchanges.

Ready examples are Silicon Valley or, in Austria, the auto-cluster in Graz and the Vienna biotechnology cluster. In a project funded by the Austrian Science Fund FWF, a research group from the Austrian Institute of Technology (AIT) in Vienna has now developed a new model that simulates the factors relevant for regional knowledge creation and maps the processes involved in greater detail than ever before. The model is said to be able to better assess the effect of innovation-promoting measures, to give one example. Principal investigator Thomas Scherngell explains the approach.

Knowledge for innovation confined to specific regions

“Regions that are particularly innovative have a long-term competitive edge and exhibit a more positive socio-economic development than less innovative regions,” explains Scherngell. “There is strong empirical evidence in innovation research of the fact – and there is also general agreement on this – that new knowledge, which is the basis for innovation, is highly localized. Such an innovation-friendly milieu is very difficult to transfer to other geographical areas. “It is confined to the minds of specific researchers, if you like,” says Scherngell. “This leads to that pronounced localization effect, resulting in competitive advantages of regions in a particular economic sector.”

Patent applications as a measure of innovation

Now that the localized geographical scope of innovation and its importance for the sustainable development of regions is known, researchers have sought to understand why innovation is generated better and more sustainably in some regions than in others. The aim is to identify the most important factors influencing the innovation capacity of companies, universities and applied research institutions in a region. In this context, the number of patent applications serves as a measure for innovative output. “Although a patent does not per se constitute an innovation, patents are the best indicator of the emergence of new knowledge with the aim of commercial exploitation,” explains Scherngell.

Models that simulated these effects were already developed in the past. However, these models did not consider individual firms but, rather, a region as a whole. According to Scherngell, the areas studied were about the size of the Austrian provinces. The FWF project has now attempted to develop a model that drills down to a deeper level, as Scherngell’s colleague Manfred Paier explains.

Agent-based model

“The focus on the regional at an aggregate level is a weak point of the previous models,” says Paier. The resulting picture was not detailed enough. “We have tried to go one level deeper and look at the organisations in each region.” Paier speaks of an “agent-based” model. In this modelling context, an agent is a unit that can exchange information with other units, such as a firm that exchanges information with another firm. “Agent-based models normally deal with narrowly defined application areas such as traffic flows, material flows or the behaviour of pedestrians. In our case, it’s different. We model knowledge repositories and knowledge flows.” As a result, the model reveals the patent activity of a region – how many patents are submitted by the firms and what technological specifications are involved.

“In this way, we can both capture greater heterogeneity in a region and get a better representation of the dynamics of developments. We study individual processes at firm level and individual decisions by firms to cooperate with others,” notes Paier. He points out, however, that the results cannot be interpreted at the level of individual firms, but that conclusions are only valid at the regional level.

Paier emphasizes that the parameters of the model are based on empirical data. Martina Neuländtner, a member of the project team, investigated this aspect. She has intensively explored ways of achieving the empirical initialization, calibration and validation of the model – a practice hitherto uncommon in agent-based models. “We have developed a process for inserting the empirical data into the model and calibrating it,” explains Paier. “The latter in particular enables us to apply the model in a practice-oriented manner, for example with regard to capturing the effects of political interventions on the quantity and quality of regional knowledge creation.”

Follow-up project

The aim of the two-and-a-half-year project, which was completed at the end of 2018, was therefore to develop a new model that could serve as a decision-making aid for politicians. In a direct follow-up project, which is also funded by the FWF, the research group is now focusing more on application-oriented issues by comparing regional innovation in Europe and China. Here, the new model is to prove itself in a practical setting.

Personal details

Thomas Scherngell is an economic geographer at the Austrian Institute of Technology (AIT). His interests include the geography of innovation, in particular innovation networks, and the statistical analysis of the relationship between innovation and socio-economic development.

Manfred Paier is an innovation researcher at the Austrian Institute of Technology (AIT). A graduate in physics, he is interested in innovation, regional questions of research and development, as well as agent-based models for the simulation of economic systems.


Dünser, M., M. Paier, A. Unger, M. Barber and T. Scherngell: Regional Knowledge Creation and R&D Collaboration in Europe: Specification of an empirical agent-based model. SSRN Working Paper No. 3456085, 2018
Vermeulen, B. and M. Paier, (Eds.): Innovation Networks for Regional Development. Concepts, Case studies, and Agent-Based Models. Economic Complexity and Evolution, Springer International Publishing 2017
Dünser, M., M. Paier and A. Unger: Knowledge creation and diffusion in a multi-regional setting. Conceptual foundations for an agent-based model. SSRN Working Paper No. 3455965, 2017

Comments (0)

Currently no comments for this article.

Write a comment

Your email address will not be published. Required fields are marked *