 Hello everyone. Today I'm going to tell you about location patterns of manufacturing industries in Tunisia. The literature on agglomeration economies shows that firms enjoy positive externalities from concentration of economic activities in a particular region. In most countries, concentration is too great to be random. Concentration is indeed determined by the complex interaction of manufacturers. We broadly distinguish between three main categories of factors. The first category is what we call first nature causes. These concern concentration attributed to natural cost advantages, including the proximity to the sea, the characteristics linked to the physical landscape, the abundance of natural resources, favorable climate conditions, and so on. The second category is what we call second nature causes or the whole market effect. These concern concentration attributed to the interaction between economic agents. They are divided into agglomeration forces operating via factors of production, including physical and knowledge spillovers and labor market pooling and agglomeration forces operating via goods related to demand and cost linkages. The third category is the in-between category is what might be called economic infrastructure. They include local public goods, such as transport networks, and factor endowments, such as industry specific skills. Let's introduce briefly agglomeration economies. A common classification of agglomeration economies distinguishes between localization and urbanization effects. Localization effects cause a high concentration of firms of the same industry. This is assimilated to externalities of specialization. And urbanization effects lead to clustering of firms of many different industries. This is assimilated to externalities of diversity. Both localization and urbanization effects can be divided into static and dynamic externalities. Dynamic externalities arise from technological and knowledge spillovers, and static externalities include natural cost advantages, local public goods, factor endowments, and pecuniary externalities related to demand and cost linkages. Both static and dynamic externalities explain the existing location of firms, but only dynamic externalities account for economic growth. In this presentation, I will try to shed some lights on broad spatial location patterns of the main manufacturing industries in Tunisia. My investigation will be based on three empirical studies and on general information about the characteristics of the various cities and main manufacturing industries involved. I will tackle mainly three issues. I will first answer the most crucial question of whether manufacturing industries are indeed agglomerated in Tunisia. Second, I will try to identify the main factors accounting for specialization patterns. And finally, I will try to suggest few facts about urbanization patterns based mostly on stylized facts, given that most studies have been conducted on the developed world, mainly on U.S. cities, so we still know very little about the developing world. The three empirical studies have mainly focused on testing whether dynamic externalities account for economic growth. They use data on Tunisian coastal governories and delegations, and the main manufacturing industries involved are divided into mature industries, including textile industries, leather and shoes industries, and agro food industries. And knowledge-intensive industries including chemical industries, electric and electronic industries, and mechanical and metallurgical industries. Let's start with the first issue. Manufacturing industries are indeed agglomerated in Tunisia. Mature industries are agglomerated along the whole coastal line. Leather and shoes industries are concentrated in the northeast region, notably in the governorates of Bizerte and Nebel. Tick-style industries are concentrated in the center east in the governorate of Monastir, and agro food industries are concentrated both in the northeast in the governorate of Nebel, and the center east in the governorate of Svex, the second central business district in the country. While knowledge-intensive industries are mainly concentrated in the northeast region, precisely in the governorate of Ben-Aruz, one of the four governorates of the Great Tunis. Now let's move to the second issue where I investigate specialization patterns. I will try in particular to broadly assess the importance of knowledge spillovers for each sector whenever it's possible. I start with mature industries. First, textile industries. This sector is less dependent upon local resources. It is one of the most exporting sectors in the country, and moreover, specialization is reinforced by foreign direct investment. The governorate of Monastir is characterized by its skilled labor pools specialized in textile due to the local transfer of knowledge from one generation to the next. It is also characterized by its good transport networks, including an international airport, and by its research infrastructure, including a competitiveness center specialized in textile. For this sector, it may be possible to broadly assess the importance of knowledge spillovers, which prevail based on empirical studies, mainly those using labor productivity-based regressions. Moreover, based on IEDI and Matusi, who found that 11% of new firms are created in Monastir, like the third after the two central business districts, Tunis and Svex, and moreover, most of new firms created at the national level are in the textile sector. This may suggest the likely importance of knowledge spillovers for these industries. One may then conclude that concentration of textile industries in Monastir is mainly attributable to skilled labor pool, to strong forward linkages mainly related to large foreign demand, and to the likely important knowledge spillovers. Now, let's move to the second sector, leather and shoes industries. This sector is largely dependent on tenant leader, so tannery is a crucial segment in the industry. Specialization is largely impassed by foreign direct investment in the proportion of 70% according to Kara Andris, and moreover, the Tunisian production is largely linked to the international market. Indeed, more than 95% of firms export 100% of their output. As for the two governments, Nebel and Bezert, they are characterized by their proximity to tanneries at least two, and to the most important ports in the country, the port of Redes for Nebel and the port of Bezert for Bezert, reducing thereby shipping goods to foreign consumers. Given that specialization in this sector is largely impassed by foreign direct investment, the question which arises why foreign direct investment does not favor the location of the governorate of Sfax, though it has similar characteristics as the first two governorates with its two tanneries and its important port, the second in the country. The answer is likely due mainly to two reasons. First, the attractiveness of the first two governorates, the beauty of their landscape, their mind weather, also for their touristic infrastructure, especially for the case of Nebel. And the second reason is the weak attractiveness of Sfax to foreign direct investment for many reasons, including its poor international accessibility. Sfax also exhibits these economies quite common to large towns and to central business districts, and it is also a quite capitalistic governorate. One may then conclude that concentration of leather and shoes industries is mainly attributable to input sharing, to strong forward linkages mainly related to large foreign demand, to the physical geography, the attractiveness of the city, to environment quality of life, and also to the prevalence of knowledge spillovers based on empirical studies, though it's not clear whether they are important or not here. Now let's move to the last mature sector, agro-food industries. This is a resource intensive sector, so the question which arises, shall we expect agro-food firms to concentrate more in governorates having natural cost advantages or in those contributing the most to the agricultural production at the national level? This sector is known to serve mainly the local market. Now let's describe the two governorates starting with Nebel, which is characterized by its natural cost advantages including climate conditions and land favorable for growing high value farm goods including citrus fruits, grapes and strawberries. Nebel is also characterized by its factor endowments including its skilled labor pool due to the transfer of know-how from the Andalusians since the 16th century and also to the availability of water resources due to the transfer of water from the northwest region region well endowed with water resources through the Mid-Jerda Cabin Canal. Nebel is also characterized by its proximity to technology parks specialized in plant biotechnology, water and environment and by its large market size and large local demand including strong seasonal tourist demand. One may then conclude that concentration of agro food industries in the governorate of Nebel is mainly attributable to natural cost advantages, climate conditions plus land to factor endowments, skilled labor pool plus availability of water resources to important pecuniary externalities related to demand and cost linkages and also to the prevalence of knowledge spillovers though we cannot assess their importance here. As for the governorate of specs, concentration of agro food firms there is attributable more to factor endowments and to strong agglomeration forces operating via goods than to natural cost advantages. As for agglomeration forces operating via goods similarly to the case of Nebel specs is characterized by its large market size and large local demand indeed more than 80% of its agro food firms serve the local market. As for its factor endowments in addition to its skilled labor pool as in the case of Nebel specs is characterized by an important by its important financial capacities allowing its investors to exploit economic policies which provide subsidies to encourage the cattle farming sector despite its poor grazing land. Now let's move to knowledge intensive industries which are mainly located in the governorate of Ben-Aruz. All what we can say or the most important we can say about these sectors is that one of these sectors notably electric and electronic industries is the first exporting sector in the country. As for the governorate of Ben-Aruz it is characterized by its skilled labor pool with higher human capital partly generated by its quite important research infrastructure including two technology parks specialized in biotechnology and pharmaceutical industries among other things and also an aerospace center. Ben-Aruz is also characterized by its well-designed industrial zones where a non-negligible portion of knowledge intensive industries firms are located. Ben-Aruz is one of the four governorates of the Great Jonas so it is characterized by its proximity to the first central business district Jonas meaning quite large land and housing prices so the location of knowledge intensive industries there is probably for the important knowledge spillovers which are likely to be more than enough to make up the high costs. One may then conclude that concentration of knowledge intensive industries in Ben-Aruz is mainly attributable to industry specific spillovers mostly knowledge spillovers to factor endowments including skilled labor pool with high human capital and well-designed industrial zones and to a lesser extent to forward linkages related to foreign demand for the aforementioned sector. Now I will suggest few facts about urbanization patterns. First it is well known in the literature that new firms generally locate in diverse environments. In Tunisia the largest portions of new firms are located are created in the two central business districts Jonas and Spaks. One may then conclude that Jonas and Spaks are two diversified governorates. As for Jonas there is an empirical evidence about this fact as for Spaks so far this fact is only based on observed patterns. Second though larger cities tend to be more diversified there is however a weak correspondence between city size and diversity. Ben-Aruz and Jonas are the most diversified governorates in the country however Ben-Aruz is more diversified but smaller than Jonas. After this investigation one may come up with the following conclusions. First the main manufacturing industries are all concentrated in Tunisia. Second labor market pooling is the most important agglomeration mechanism at work for both mature and knowledge intensive industries suggesting that industries partly locate in specialized environments to share a common pool of specialized workers. Second according to several studies localization effects are low and urbanization effects are high in knowledge intensive industries. This is not the case of Tunisia where both effects are high in knowledge intensive industries. Indeed Ben-Aruz the most diversified governorate is also the most specialized in knowledge intensive industries suggesting that the most important agglomeration mechanisms at work are labor market pooling and knowledge spillovers. The fourth conclusion there are no clear patterns for the correspondence between city size and specialization in Tunisia. Indeed several patterns coexist. First medium-sized cities the case of Bizarre and large cities the case of Monastir are specialized more in mature sectors and less in knowledge intensive industries. The second pattern large cities the case of Ben-Aruz are specialized more in knowledge intensive industries and less in mature industries. Third medium-sized cities the case of Sphax and large cities the case of Nebel are specialized in both knowledge in both mature and knowledge intensive industries. And I finish with the last conclusion. It is well known in the literature that urbanization effects are high in central business districts. This is the case of Tunisia where the two central business districts Tunis and Sphax are diversified however they exhibit different patterns. While Sphax is both diversified and specialized Tunis is diversified but has no single manufacturing industry. We should be here very cautious in drawing conclusions in general because this might not be the case if we consider service industries. Thank you very much for your attention.