Bjørn Bo Sørensen defends his PhD thesis at the Department of Economics


Bjørn Bo Sørensen, Department of Economics, University of Copenhagen


Essays on Industrial Development in Africa and Asia


  • John Rand, Professor, Department of Economics, University of Copenhagen
  • Henrik Hansen, Professor, Department of Economics, University of Copenhagen

Assessment Committee:

  • Finn Tarp, Professor, Department of Economics, University of Copenhagen
  • Carol Newman, Professor, Department of Economics, Trinity College Dublin
  • Måns Söderbom, Professor, Department of Economics, University of Gothenburg


This Ph.D. dissertation consists of three self-contained chapters analysing different aspects of industrial development and structural transformation in Africa and Asia.

In the first chapter (co-authored with Henrik Hansen, John Rand, and Helge Zille), we study the ownership of “pioneer firms” that establish industries in regions where they did not previously exist. Using the Vietnamese Enterprise Survey (2001-2017), which allows us to track close to one million formal firms over time, we explore whether Vietnam’s remarkable industrial diversification during the past two decades was driven by state-owned, multinational, or domestic and privately owned enterprises (SOEs, MNEs, or PDEs). We document significantly higher pioneering frequencies in SOEs and MNEs compared to PDEs, also after controlling for a broad spectrum of observable characteristics. Second, we show that MNEs tend to pioneer in complex industries in Vietnam’s economic centres, while SOEs pioneer in unsophisticated industries in regions with less economic activity. Finally, we investigate the dynamic employment effects in industries pioneered by different firms. We find that MNE-pioneered industries generate significantly more jobs, which is primarily driven by employment in pioneering MNEs and other MNEs following the pioneer in subsequent years. In contrast, SOEs create no additional employment in the industries they pioneer compared to industries pioneered by PDEs.

In the second chapter, I study how granular climate hazards affect local production, and what role spillovers and general equilibrium adjustments play in shaping local resilience. I combine high-resolution satellite imagery of floods with geo-coded panel data on Chinese manufacturing firms between 2000 and 2007 to investigate these questions in an event study analysis. I first show that 1-in-50-years floods cause large and persistent declines in aggregate manufacturing output in affected counties. I then analyse the role of direct and indirect flood exposure in shaping the aggregate outcome. Within flooded counties, firms that are actually flooded experience large declines in output compared to nearby firms. Yet, nearby firms in flooded counties simultaneously perform worse than similar firms in non-flooded counties. The results point to the existence of local spillovers and equilibrium adjustments. I show that floods lower the cost of labour and that firms in tradable and labour-intensive industries expand, whereas other firms contract. The results are consistent with a simple model of local spillovers ala Moretti (2010), and they suggest that a common factor affecting both flooded and nearby firms may be a fall in local demand.

The third and final chapter consists of two self-contained articles (co-authored with Christian Estmann, Benno Ndulu, Enilde Sarmento, and John Rand) introducing a methodological framework for identifying attractive product sectors that can be targeted by industrial policy aiming to (i) promote economic upgrading and diversification and (ii) increase export revenue. In a supply-side analysis, we first use the economic complexity methodology to identify product sectors that are important for structural transformation. In a demand-side analysis, we then use gravity models to rank the trade potential of these target products across different markets. The first article applies the framework to guide industrial policy in Mozambique—one of the world’s least complex economies. We identify unexploited opportunities in machinery, vehicles, and transport equipment. The second article applies the framework to the case of Tanzania, where we find a high potential in complex sectors such as machinery and chemicals. We further extend the framework to consider sectors’ capacity to absorb labour, and find that less complex sectors such as agro-processing, metal, and wood, have a higher potential to create employment.

An electronic copy of the thesis can be requested here: