Clustering, international networks and performance of firms: some complement approaches for MENA’s convergence

FEM33-17 | March 2011

Title

« Clustering, international networks and performance of firms: some complement approaches for MENA's convergence »

By

Juliette Milgram, University of Granada, Spain

Contributeurs

ACHY Lahcen INSEA, Rabat Morocco; BASRI Amine Ministère de l’Industrie et du Commerce, Rabat, Morocco. ; GEORGANTZIS Nikolaos, University of Granada, Spain; LACOMBA Juan Antonio University of Granada, Spain; LAGOS Francisco Miguel University of Granada, Spain; PAMUKCU Teoman Middle East Technical University, Ankara, Turkey; TAYMAZ Erol Middle East Technical University, Ankara, Turkey; YILMAZ Kamil Koç University, Istambul, Turkey

Note :

This document has been produced with the financial assistance of the European Union within the context of the FEMISE program. The contents of this document are the sole responsibility of the authors and can under no circumstances be regarded as reflecting the position of the European Union.

Summary :

Our first objective is to study if networks effectively ameliorate the TFP of firms and to compare international networks (imports of intermediate and capital goods, vertical or horizontal cooperation) and local networks (experience at exporting and producing of other firms at the regional level). Since these phenomena are essentially microeconomic in essence, firm-level datasets offer a very good opportunity to deep in their analysis. Our analysis is based on a sample of Spanish and Turkish firms. We modify the Olley and Pakes (1996) approach to control for endogeneity bias and consider different agglomeration indicators as an endogenous input of the production function.An important contribution of this study is to show that Spanish and Turkish firms share important characteristics. Namely, traders are bigger and more productive. Small plants that trade are more productive than large plants that do not trade. Foreign firms are larger and trade more than domestic firms in both countries. Regional location doesn’t make a so clear distinction among firms than their international activities, in terms of productivity.In the Turkish case, we obtain more evidence of congestion problems than evidence in favour of positive spillovers. Concentration of exporters, regardless to the sectors of activity, appears as the sole source of positive externalities for the productivity of firms located nearby. In Spain, we obtain more evidence about the positive effect that firms can obtained from their localisation in terms of TFP but some congestion’s problems are also evidenced. There is a risk of congestion costs if the concentration of production sharing horizontal linkages is too dense. Unlike Turkey, Spanish manufacturing firms benefit from positive spillovers from concentration of workers and importers conducing similar activities in their vicinity. Small plants are the firms that learn more from other firms in Spain or are less affected by congestion costs in Turkey. Small plants in both countries productivity are positively impacted by the amount of import of the region.The second objective of this report is to verify if the proximity of other exporting firms increases the probability of a firm to become an exporter. To this aim, we use a huge datasets for Moroccan manufacturing firms and estimate a model for the decision o export that takes into account agglomeration variables. The paper reveals that regional concentration of exports either in the same industry or in all industries exerts a positive and statistically significant effect on the firm’s decision to export. This finding corroborates the hypothesis that agglomeration offers opportunities for firms to interact and exchange information and knowledge on foreign markets, and that such exchange is much more rewarding ?in the sense of leading to export? when interacting firms belong to the same industry.In the third part of this study, we use experimental economic technique to study how trust among business partners is affected by the information about the residence’s country of the partner. The experiment examines a modified version of the Trust Game where players come from four countries: Turkey, Morocco, France and Spain. Our results show that Moroccan are significantly more trusting than players from other countries, except toward Spain. They obtain in turn less reciprocity. Overall, participants exhibit low level of reciprocity but this behavior does not discriminate according to the countries’ receptor, except Morocco that is more selfish with Spanish people. Spanish and Moroccan seem to reflect a mutual distrust that could be dangerous for their economic and political relations.