Production networks, firm productivity and geography

Fattorini, Loredana (2018) Production networks, firm productivity and geography. Advisor: Rungi, Dott. Armando. pp. 174. [IMT PhD Thesis]

Text Fattorini_phdthesis.pdf - Published Version Available under License Creative Commons Attribution No Derivatives. Download (6MB)

Abstract

Modern economies are organised as webs of interconnected agents. Among others, companies represent the principal actors. In the recent decades, the emergence of supply networks, that is the organisation of technical processes in production stages involving specialised suppliers located in different countries, brings about an increasing complexity in the worldwide economic system. Moreover, empirical studies from firm-level data provide evidence of heterogeneous distribution in companies performance also within industries, regions and countries. However, when discussing policy-making in Europe this aspect is still neglected by many, and the impact of regional or industrial policies is evaluated at the macro level. Recently, geo-coded information on where firms run their activities is becoming available to the researchers, offering a good opportunity to improve the empirical design of specific research questions. In this thesis, we investigate these aspects in detail, showing the utility of adapting and implementing analytical tools from Network Science and Machine Learning along with ad hoc econometric techniques. Moreover, we contribute to the literature on international economics, industrial organisation and economic geography. With the first work, we propose an empirical tool, the Input Rank, adapted from the notorious PageRank, which is originally designed for social networks and search engines, and we test its empirical validity for choices of vertical integration. Then, we ambitiously test the effect of the EU Cohesion Policy on an own-built dataset with firm-level total factor productivities of European manufacturing companies. Finally, thanks to the implementation of the DBSCAN, a challenging density-based spatial clustering algorithm, we exploit firms’ location information to identify industrial clusters and examine the likelihood of firms’ survival according to their location in industrial clusters or more isolated areas.

Item Type:	IMT PhD Thesis
Subjects:	H Social Sciences > HB Economic Theory
PhD Course:	Economics
Identification Number:	10.6092/imtlucca/e-theses/260
NBN Number:	urn:nbn:it:imtlucca-27286
Date Deposited:	25 Jul 2019 08:09
URI:	http://e-theses.imtlucca.it/id/eprint/260

Actions (login required, only for staff repository)

View Item