The productivity effects of regional anchors on local firms in Swedish regions between 2007 and 2019—evidence from an expert-informed machine learning approach

Abstract The concept of regional anchor firms remains under-investigated. We analyse the mechanisms by which anchors affect other regional firms, disentangling effects resulting from scale/size vis-à-vis knowledge spillovers. Departing from previous idiographic research, we adopt a nomothetic research design and develop a stepwise expert-informed supervised machine learning approach to identify all anchor firms in Sweden between 2007 and 19. We find support for positive anchor effects on the productivity of dependent regional firms. This effect is driven by factors reflecting scale/size, while anchors’ R&D intensity as a measure for knowledge spillovers does not drive productivity gains.