 Good evening everybody. Thank you. I'm really very proud to be here and present the paper, the disengling of patterns of geographic concentration in Tunisian manufacturers that I written with my colleague with Ed Matusi and I think I want to to have a very great to current human who helped us to to have a good presentation. Okay, our motivation are from written papers and we started with Paul Krugman who asked that Grugmanation may be more the rule than an exception and these results have been proved by a study of World Bank reshaping economic geography when you demonstrate that market favor some places about our others and places cities coastal areas and connected countries are favored by producers then the choice of the places it isn't it isn't done exigenously but it is not an undeniable decision and then analyzing the agglomeration and the situation and the effect of the geographic localization is so important and theory you demonstrate that productivity spolovers is so important why because an increase in a firm productivity can have a positive effect a significant fact on the neighborhood of first productivity as have demonstrated mrs. Howard and and also there is in the all the territory you you are speaking about Mar models Jacobs models and porters and in this model who have a great aggregation have an effect and this effect help to reduce cost and this a cost reduction has a leading are really to efficiency gain of the first this is this is why studying agglomeration it will be very interesting and the question is how can the Tunisian industry concentration be measured how to measure this aggregation have to measure this concentration and to respond to this main question you have three parts the first one is whether firms cluster in Tunisia yes or no and to analyze this you use some summary statistics sample some graphs also some aggregation indices who help us to test and to confirm this analysis and then the second part will be why firms clusters and to respond to the second question we we will use two econometrics models in the first one we analyze factors driving forms location choice and the second one you analyze factors driving employment increases employment increases it is used in the literature as the measure of the the firms magnitude the firms growth and the third question is what are the benefits of clustering why you cluster and in order to analyze its effect you we will analyze also an econometric model measuring the effects of location on productivity growth then the the outline of paper will be responding the three questions but at the end you have two parts the first one is economic externalities and when you return to the first paper in this you have Anderson and you have crew man Anderson analyze the different phenomena what is urbanization and this is why you you try to to give an explanation is it localization or an urbanization and you demonstrate that in less developed the less developed countries essentially it is a phenomena of urbanization not a localization at you find in developed and developed countries and then you analyze to to continue have conclusion and policy decision you you make a correlation with previous decision made in Tunisia and the results that first question where from cluster we started with just a map graphs and today they will have some duality between Eastern which is the literal zones and Western who it is the anterior zones at what we remark to remark that there is a gap in this gap is going huge a user and high and high and other things we you have six region in Tunisia and what we precise okay the east regions but between the east regions there is two principle region which is the northeast and the center is what are the more important than the others now you we consider the two more important regions where the northeast and what we remark remark that we have a huge gap between Tunis capital and the other governorates there is a very huge part the other things the trends for the central east also you have sparks we about is different source we are coming and the other governorate then what so what we say are there grass drink yes there is clustering effect and in Tunisia as you look you have two things 83% of firms are concentrated in the east regions and the other things there is another phenomena that what is the dark and hard here it is Tunis and governor at us facts and there were these 40% of firms and Tunis as facts are the two huge CBDs in Tunisia this is why you are asking about urbanization which is a phenomenon as important in less developed countries there is another kind of heterogeneity it is a sector of disparities in Tunisia you have the more essential firms are in textile and now there is an electric and electronic who the number is all is going and he increases more and more from year to year and then what to have to analyze don't you consider three sectors the textile what we consider that textile it is essentially specifies here what is the monastere there is one the third of firms are here electric electronics are concentrated in Tunis and in governorate of facts however when you consider agro food there is more more diversification and the gap it is lesser the chemicals also you have this and what we do you say that where firms are cluster and what our first conclusion is that exporting sex per essentially electronics textile and chemicals are concentrated in lateral region that we give them up and only products associated with local demand okay essentially agro food are more diversified and the anterior governorate are limited in order to analyze don't we use specialization decks that are the metrics when we use this this one you have a phenomenon is very specific that the what what we found we found that some interior zones have a greater specialization index why because simply this kind of index specialization is not appropriate there is the problem of of monopoly and when we have a limited number of our firms when you calculate this index you have a wrong signal then we use alisonic leisure who essentially correct for the fact that firms consisting of your relatively countries and when we use the alisonic leisure index you find result that are coherent with our summary results and you say that located industries are textile and leather and less locate our agro food products why from cluster here we use a firm location model where we consider unknown the business variables is the the variation of the rate of variation of the number of firms and you have two kinds of exploratory variables you have economic variables and you have regional characteristic variables with the X and W that you have capital revenue exporting share something like this I'm sorry what we find in our estimation that specialization index is not significant however competition is positive and significant and you say that the number of firms tend to increase in more competitive areas rather than especially as one another results what is also important associated with aggregation phenomena that is dummies associated with the lateral or interior okay with that as fax dummy and you say that small size firms are mainly concentrated around literal and literal zones when we found CVDs and this is why I ask the question location choice may rather be considered as urbanization externalities choice however what we find that growth on firm create creation decrease if initial number of firms is important and governance industries with initial high level of employment will have lower with this contradictory with the other firms and you say that France capital income and employment are not significant what it is surprising this is why because in the first model we don't integrate the economic effects this is this is why you use the second model where the variables is the employment growth of the firm and what you find to find something that initial high level of employment lead to a lower to lower slow growth in the industry and employment growth decrease as productivity and skill workers improved surprise and employment growth increase also for countries who are near to us and what we find with that that specialization have negative effect a significant one so just and what is the result different from the model competition however has a positive effect and competition also leave leads to higher government employment growth what are the benefits of clustering here we use a model of productivity growth that car where we consider the rate of productivity increase and what we found to find that high ratio productivity in government rate reduced productivity growth productivity decrease if government rate are exporters and literal dummies have all-time positive effects specialization has a positive effective on productivity which is agree with more or where we can we differentiate the on the genius variables if you use the growth of employment this you have the opposite and governor rate our competition reduce competitiveness however when we consider the two indices competition and specialization effect the competition effect become non-significant and then we consider that there is dynamic of personalities may not be appropriate if we restrict to classical mar and porter models and this is why we try to analyze the feminine of urbanization just two minutes why analyze differentiation for the argumentation of localization you have localized because we have resources or transport advantages you localized also because it helped us to reduce the cost of getting supply and the flow of formation and localization also can help us to to have more stable demands this is what the theory says if we consider urbanization you you can say that firms localizing government because of high demand local high local demand and because they can reduce their their cost and then in our model we found that location greater toness or location in literal help to location greater toness have to affect on growth of unemployment growth and location in literal improve productivity also then you can modify the model a consider the what Henderson consider as urbanization conclusion and policy decision not not policy recommendation just you say that Tunisia's structural adjustment program is done in 1984 has increased firm performance but he has an effect that growing inequality between coastal and region that it is as you demonstrate and the elitonic laser index depict that textile electronics and chemical are more limited agglomerated sectors that various agglomeration specification however and competition location have different effects if we modify the the model if you modify the specification because specification has no significant effect on the number of firms have or a wrong sign or a wrong effect on employment growth but he's increased productivity that it is a coherent with mar model essentially when we use the right model competition also has a positive effect on the number of firms it includes employment but he has a wrong sign for productivity however locating in greater toness results in firm growth and proof and locating in literal grenade enhanced productivity also why just the last slide yeah historically CBD is offered better economic incentives that are for small firms but there is lack of courage from the Tunisian government you have no strong political action that have taken to develop new cities never and this is why you have a problem of urbanization because there isn't courage policy and this is why exporting industries are located new older cities and no export industries are less located by pre for eastern regions thank you very much