Artificial intelligence and regional economic development

Background and Objective

AI is considered a general-purpose technology and has been on everyone's lips at least since the market launch of the chatbot "ChatGPT". general-purpose technologies are technologies that are used in many different sectors of the economy. The development of general-purpose technologies can have a positive impact on all sectors that work with the new technology. The link with the application sectors in turn also enables improvements in the general-purpose technology itself. Consequently, there is a feedback loop through which all sectors in which the general-purpose technology is used can benefit economically and technologically.

The effects of AI technologies on the productivity and the technological development are not unambiguous and also the so far measured impact stays below general expectations. This contradiction is called “productivity-paradox“. Recent studies suggest that the weakness of the measured effects can be partially explained by the missing differentiation of AI. There are many technologies that are called AI, but these often function differently. Thus, the weakness of the effect could be attributed to difficulties of establishing the above-mentioned feedback loops, as AI as a technology is too divers. Another explanatory approach of the “productivity paradox” is the so far insufficient differentiation of AI users. Research indicates that AI innovation does not impact all firms and regions equally. The effects could be rather different depending of the size, the financial capabilities or the available knowledge base leading to even a negative impact of AI.

The project therefore pursues two aims. On the one hand, we want to systematically differentiate AI technologies depending on their impact on the technological and economic regional development. On the other hand, the regional prerequisites for a positive impact of AI should be identified and explained. The results in combination can help to improve the understanding of the importance of AI in regional development processes and to improve the efficiency and effectiveness of policy support for the usage of AI.

Approach

The project is divided into three major consecutive analysis steps. In the first step, we pursue the goal of a better differentiation and characterization of different AI technologies. We analyze the interaction of different AI technologies during the development process from academic knowledge to the entrepreneurial use of AI. In the second step, we investigate which region-specific characteristics promote or hinder the emergence of AI knowledge and the use of AI in an entrepreneurial context. Research into the effects of various AI technologies on regional economic development forms the third step of the analysis. Here we also look at whether the influence of AI varies between different types of regions.

Data and Methods

We collect various secondary data for the project. This includes publication data from scientific journals and patent data to quantify research on AI. Furthermore, secondary data from the Federal Statistical Office and the Statistical Office of the European Union are used to characterize regions based on various indicators, such as population density and regional GDP. Finally, data from the statistical offices of the federal states is used at company level to map the use of AI technologies by companies. If necessary, additional company data from the Amadeus and ORBIS databases will be used. The data is then analyzed using quantitative econometric methods (e.g., panel or negative binomial regressions) in order to identify the determinants of AI technologies and estimate their impact on regional development.

Our Research Questions

• Which framework conditions are particularly conducive or obstructive to the entrepreneurial use of AI?
• How does the transition from AI research to AI knowledge and AI use functions?
• How do AI technologies change the economic and technological structure of regions in which they are applied in or researched?
• Which regions benefit most or least from the use of or research into AI?