Abstract

In the last two decades, technological advancements have led to a large availability of detailed data gathered on a small regional scale. In regional and urban economics, the use of social media data, mobile phone or satellite imagery gave the opportunity to break through regional aggregation of official statistics. These secondary data sources to use for economic research is peculiar for the era of Big Data. Compared with primary official statistics about regional economic development the creation is comparable cheap and with advancements in statistical learning the data sources help in predicting local economic outcomes. So far, research is aiming primarily on data gaps in developing countries. However, also in industrial countries with reliable statistical offices these new data sources give new research opportunities, e.g., in the observations of urban sprawl pattern from satellite imagery (Bruchfield et al., 2006). Evidence on predicting specific regional economic development is rather scare. In this project, we use satellite imagery of build-up intensity and employment data based on grid cells for Germany to predict regional employment pattern.