COMMERCE EXTÉRIEUR ET TRANSFORMATION STRUCTURELLE DE L’ÉCONOMIE MAROCAINE 2. Imports contents, value added generation and structural change in Morocco: an input-output analysis 43 Elhadj EZZAHID, Enseignant chercheur en économie, FSJES/LEA - Mohammed V University-Rabat Abdellatif CHATRI, Enseignant chercheur en économie, FSJES/LEA - Mohammed V University-Rabat Abstract Our main goal in this paper is to classify productive sectors according to the combination of two effects of a unitary final demand shock. On the one hand, the impact of that shock on the external dependency of the economy (more imported inputs by unit of final demand) and, on the other hand, its impact on the efficiency of the economy (more VA generated by an extra unit of final demand). To perform this ordering, we domesticate a national I-O tables for the period 1999-2009 and we develop a convenient arrangement of the standard input-output model. Two of our results worth to be highlighted. First, the imports elasticity with respect to growth is superior to unity. Second, there are no productive sectors that belong to the most virtuous classes of sectors able to accelerate structural change, ie those sectors that are likely to increase the capacity of the economy to generate more value added and reduce its external dependency. Key words: Input-output analysis, Backward linkages, Leakages, Structural change, Value added, Imports, Morocco. JEL classification : C67, D57 43  This paper is part of a research project that received financial support from MESRC and CNRS. We thank the OCP Policy Center and the participants in the workshop organized May 24th, 2016 for their comments, remarks, and suggestions. EQUILIBRES EXTERNES, COMPÉTITIVITÉ ET PROCESSUS DE TRANSFORMATION STRUCTURELLE DE L’ÉCONOMIE MAROCAINE 91 PARTIE I Contenu en import, génération de valeur ajoutée et transformation structurelle au Maroc : une analyse input-output Résumé Ce papier vise à hiérarchiser les secteurs productifs en combinant les deux effets d’un choc unitaire de demande finale. D’une part, l’impact de ce choc sur le contenu en importation de la production domestique (effet de dépendance externe), et d’autre part, son impact sur la capacité de l’économie à générer davantage de valeur ajoutée (effet d’efficience). A cet effet, nous avons réarrangé le modèle input output classique et l’avons appliqué aux tableaux I-O domestiqués sur la période 1999-2009. Deux de nos résultats méritent d’être soulignés. Tout d’abord, l’élasticité des importations par rapport à la croissance est supérieure à l’unité. Le deuxième résultat est qu’aucun secteur n’est localisé dans les classes/zones les plus avantageuses et les mieux à même d’accélérer la transformation structurelle, ie celles abritant les secteurs qui sont susceptibles d’accroître la capacité de l’économie à générer de la valeur ajoutée et de réduire sa dépendance externe. Mots-clefs : Modèle Input-output, liens en amant, fuites, transformation structurelle, importations, valeur ajoutée, Maroc. 92 EQUILIBRES EXTERNES, COMPÉTITIVITÉ ET PROCESSUS DE TRANSFORMATION STRUCTURELLE DE L’ÉCONOMIE MAROCAINE COMMERCE EXTÉRIEUR ET TRANSFORMATION STRUCTURELLE DE L’ÉCONOMIE MAROCAINE Introduction Most previous studies dealing with economic structural change using input-output analysis have paid little attention to the distinction between imported and domestically produced inputs. They have simply compared backward and forward linkages over time without domesticating the technical coefficients matrix (Guo and Planting, 2000). Indeed, in an increasingly integrated World economy, where inter-countries ties have become more and more dense, measuring the leakages for individual sectors and for the economy, as a whole, is obviously a relevant question. This is especially true for small open economies such as the Moroccan one. The analysis of the sources of economic growth in Morocco in recent years brings out two important facts. The first is the strong reliance of growth on domestic demand and the negative contribution of its external component to growth (among others BAD and al., 2015; Agénor and El Aynaoui, 2015). This is due to both the weight of incompressible imports and the low competitiveness of exportable supply. The second fact shows that the Moroccan process of economy’s openness and its greater integration into its regional and international environment benefits mostly to partner countries and it has not yet allowed a rebalancing of growth toward external demand (IRES, 2013). The result is, in particular, a high domestic production’s content in imports and limited links between domestic productive sectors. This may affect the process of structural change of the Moroccan economy. In this paper, we develop a convenient arrangement of the basic input-output model to appreciate the extent of structural change in the Moroccan economy. Indeed, we show that the increase of final demand addressed to sector j can influence both the ability of the economy as whole to generate more value added as well as its external dependency. Therefore, the dominance of the one or other effect permits to classify productive sectors and thereby to propose a new kind of “key sectors” according to their respective contributions to accelerate the process of structural transformation of the Moroccan economy. The remainder of the paper is as follows. The next section discusses some stylized facts about the Moroccan economy especially related to the importance of imports. The second section presents the methodology. We showed that the overall effect of a unit change of final demand derived from the Leontief inverse matrix must be the sum of three terms: interindustry flows, distributed value added, and imported inputs. The third section presents data and the methodology of domestication of the Moroccan Input-output tables. This domestication is necessary in order to separate imported and domestically produced inputs. The empirical results are exposed and discussed in the fourth section. The last section concludes. EQUILIBRES EXTERNES, COMPÉTITIVITÉ ET PROCESSUS DE TRANSFORMATION STRUCTURELLE DE L’ÉCONOMIE MAROCAINE 93 PARTIE I I. Stylized facts on the Moroccan economy: increasing dependency on imports The Moroccan economy is a small and an open one. Imports include strategic items such as equipment, machines, and energy products. Thus, Moroccan productive sectors use intensively imported inputs. Recently, apprehensions were raised due to the growing dependency of the Moroccan economy on imports. The rate of penetration, calculated as the part of domestic demand satisfied by imported goods, have depicted a growing trend since mid-1990s (figure 1). Figure 1. Evolution of the imports penetration rate Source: Author’s elaboration from HCP data The increase of imports’ penetration rate implies that the representative consumed basket of Moroccan consumers (households, government …) tends to include a growing share of imported goods. Indeed, the average share of imported goods in the representative basket of households and government increased from 11.4% over the period 1998-2006 to 15.8% over the period 2007-2012. Two points worth highlighting in this regard. First, the structural reliance of the Moroccan economy on imported inputs increased due to the expansionist fiscal policy conducted especially during the first decade of the 2000s. This policy was designed to boost the economy and to satisfy social demand. It manifested by public projects aiming to increase the stock of public infrastructure and to fill in social gaps. Unfortunately, this policy generated more imports. This raises serious doubts concerning the sustainability of current account deficits. Second, the growing content of Moroccan exports in imported inputs could be explained by the implementation of new sectoral strategies that promote less integrated productive sectors to the economy, such as cars making, electric and metallurgical industries, whose production depends heavily on imported inputs. 94 EQUILIBRES EXTERNES, COMPÉTITIVITÉ ET PROCESSUS DE TRANSFORMATION STRUCTURELLE DE L’ÉCONOMIE MAROCAINE COMMERCE EXTÉRIEUR ET TRANSFORMATION STRUCTURELLE DE L’ÉCONOMIE MAROCAINE In all over the world, GDP dynamics are highly instrumental in the evolution of imports. In empirical literature, imports (M) are frequently, and mainly, explained by GDP and their price compared to domestically produced goods price (price of imports/price of GDP- denoted PM/PY). The first variable captures the income effect and the second variable captures the terms of trade effect. The responsiveness of imports to GDP is measured by the arc elasticity44 or by running equation (1). lnMt = α1+ α2lnGDPt+ α2log(PM/PY)+ ut (1) The elasticity of imports is ∂lnMt / ∂lnGDPt. Remark that we have not used a full specification of the demand equation of M. The objective is to gauge the magnitude of the responsiveness of Moroccan imports to GDP. It is worth to signal the scarcity of research done on the elasticity of imports to GDP in the case of Morocco even if this issue is at the heart of the current account sustainability. The national board of foreign trade (Conseil National du Commerce Extérieur –CNCE- (2013) estimated econometrically this elasticity to be about 1.5. That means that a 1% growth rate of GDP induces a 1.5% increase of imports. For the CNCE, this is due to the incompressibility of Moroccan imports and to the increasing rate of penetration (imports divided by domestic demand). Using data from national accounts, we find that imports elasticity to growth are much higher than CNCE estimates. For the period 1990-2013, 1% increase of GDP induced a 2.19% increase of imports, taking only the period 1998-2013, the nominal elasticity of imports to GDP jumped to 2.57%. It is necessary to signal that Moroccan imports recorded important increases in the period 2002-2008 mainly because of the unusual increases of the price of commodities (energy, metals, and foods). Therefore, it appears that an important component of Moroccan imports’ increase is attributed to the price effect and not to the volume effect45. Running equation (1) provides evidence that real GDP (RGDP) is instrumental in the evolution of Moroccan imports. It appears that 1% growth of GDP produced 1.32% growth of imports. Remark that in the equation we introduced only the two major variables commonly used as drivers of imports of a country. This is the full estimated equation : lnMt = -5.37 + 1.32 lnRGDPt + 0.054 log(PM/PY)+ ut (2) (-9.77) (31.03) (0.29) R2=0.96 (between parentheses are t-statistics) 44  The arc definition of elasticity of imports with respect to GDP (εM/GDP) is defined as follows: εM/GDP= gM/ gGDP ; gM and gGDP are, respectivelly, proportional growth rates of real M and of real GDP. 45  For more details about the evolution of Moroccan imports during the period 2002-2012, see the Annual Report of Bank Al-Maghrib (2013) EQUILIBRES EXTERNES, COMPÉTITIVITÉ ET PROCESSUS DE TRANSFORMATION STRUCTURELLE DE L’ÉCONOMIE MAROCAINE 95 PARTIE I Figure 2. Evolution of GDP elasticity of imports Source: Author’s elaboration (HCP data) II. Methodology 1. Input-Output model and production’s multipliers Input-output methodology provides a useful tool to quantify intersectors’ relationships in the economy (Miller and Blair, 2009). It shows how the ith sector output xi is used by other productive sectors as intermediate input and by final users as final demand. For clarity, consider an economy consisting of n productive sectors. Let xi be the total output of sector i, and zij be the ith sector production used as input by the jth sector and yi be the total final demand of goods produced by this sector46. We can represent the way in which sector i distributes its production through sales to other productive sectors and to final users as follows: (3) It should be noted that the I-O model relies on some strong assumptions (Bess and Ambargis, 2011). One of these fundamental assumptions is that the production process operates under constant returns to scale. In addition, in a given period, the interindustry flow from sector i to sector j depends entirely on the total output of the jth sector. It assumes also that inputs are used in fixed proportion, without any possible substitution. The fixed relationship between a sector’s 46  Final consumption comprises goods that would not reappear in the economy in another form. They either are used by domestic final users of goods who are households (C), private investors (I), and governments (G) or exported abroad (E). Aggregately, final demand decomposes into domestic final demand (C+I+G) and foreign final demand (E). 96 EQUILIBRES EXTERNES, COMPÉTITIVITÉ ET PROCESSUS DE TRANSFORMATION STRUCTURELLE DE L’ÉCONOMIE MAROCAINE COMMERCE EXTÉRIEUR ET TRANSFORMATION STRUCTURELLE DE L’ÉCONOMIE MAROCAINE output and its intermediate consumption are measured by technical coefficients47 aij. They are obtained by dividing the inputs zij by the production xj of a given sector/industry j (aij=zij/xj and consequently zij=aijxj). Thus, the distribution of a sector i production, as intermediate inputs final demand, may be represented by the following equation : (4) This equation depicts the dependency of interindustry flows on total output of each sector. If we denote the matrix of technical coefficients by A, the vector of total output by X, and the vector of final demand by Y, the matrix form of that system is X=AX+Y. Traditionally, within the I-O framework, final demand is considered as exogenous to the production process. Then, the model is used to determine the vector of production that is necessary to satisfy a given final demand vector. Indeed, equation X=AX+Y can be re-written as : Y = (I-A)X (5) If the matrix I-A is invertible then this is a linear system of n equations with a unique solution. In this case, we can find the output necessary from each sector and the economy as whole to supply a given increase of final demand of the exogenous sectors (consumers). Indeed, the solution of the system is X=(I-A)-1Y. The matrix (I-A)-1 is the Leontief inverse matrix denoted by L=(I-A)-1. The matrix L is also known as the total requirements matrix. It gives valuable information to assess the effect on the economy of changes in elements that are exogenous to the model. Each element of it (lij) presents the total direct and indirect effect on the production of sector i of a unit increase in final demand of sector j. The sum of terms of the jth column of the total requirements matrix gives the total effect (direct and indirect) on overall domestic production of a unitary change in the final demand addressed to the jth sector. This is the so called (simple) output multiplier of sector j. It is measured as follows: (6) 2. Measuring leakages and structural change The simple output multiplier (showed by equation 6) for a given sector gives the well known backward linkages between this sector and others productive sectors. In fact, within input-output framework, there are two kinds of economic linkages between productive sectors: backward- 47  Called also input–output coefficients, direct input coefficients or direct requirements. EQUILIBRES EXTERNES, COMPÉTITIVITÉ ET PROCESSUS DE TRANSFORMATION STRUCTURELLE DE L’ÉCONOMIE MAROCAINE 97 PARTIE I linkages48 and forward-linkages49. Backward and forward linkages indicators quantify economic sectors’ “connectedness”50. Comparisons of the strengths of sectors’ backward and forward linkages, in absolute or normalized terms, provide an interesting tool to identify “key” or “leading” sectors in a given economy51. In addition, the track of the change of these indicators, in a dynamic or comparative static, is valuable to detect possible structural change in that economy (Guo and Planting, 2000). In this paper, we propose a particular approach to order productive sectors in Morocco. Indeed, starting from the evolution of the backward linkage of each sector, we show that the increase of final demand addressed to the concerned sector exercises two main effects on the economy as whole. The first is the change on the ability of that economy to generate value- added (by unit of final demand increase). The second is the change on imported inputs content of domestic production (in response to the same shock). The combination of these effects (the dominance of the one or other effect) permits to distinguish between sectors involving a positive “efficiency” change, ie permitting to the economy to increase the generated value added and those involving a negative “dependency” change, ie exacerbating the external dependence of the economy. This allows, consequently, appreciating the structural change of the economy as a whole. Indeed, if more and more sectors record a positive efficiency change, one can deduce that resources are used where they are most productive (McMillan and Rodrik, 2011). For this purpose, remark first that the jth sector’s output requires interindustry inputs z1j, z2j, … and znj from other productive sectors, inputs from payment sectors, represented by distributed value added (vj)52, and imported inputs (mj)53. Thus, total jth sector’s output can be written as follows: (7) 48  If an industry i increases its production, there will be increased demand for the other industries whose products are used as inputs by industry i. Models that measure impacts based on this type of relationship are called backward models. An industry with higher backward linkages than other industries means that its expansion is more beneficial to the economy compared to other productive activities. 49  Increased output in sector i also means that additional amounts of product i are available to be used as inputs by other sectors for their own production. That is, there will be an increased supply from sector i (as a seller) to the sectors that use good i in their production. Models that mesure impacts based on this type of relationship are called forward-linkage models. 50  Since the seminal and pioneering works of Rasmussen (1956) and Hirschman (1958), several more sophisticated measures have been proposed in the literature, especially for advanced economies, where the objective is not to reinforce the linkages with existent activities, but to explore new emerging activities. For more details on this question, see for example (Drejer, 2002) 51  Those sectors that are most connected and therefore, in a sense, are the most ‘important. 52  The value added is the sum of incomes disturbed as compensation of services offered by primary sectors (labor and capital, notably) and used by productive sectors as inputs. 53  The second payment sector concerns imported goods used as inputs by productive sectors. This is what we call foreign intermediate consumption. 98 EQUILIBRES EXTERNES, COMPÉTITIVITÉ ET PROCESSUS DE TRANSFORMATION STRUCTURELLE DE L’ÉCONOMIE MAROCAINE COMMERCE EXTÉRIEUR ET TRANSFORMATION STRUCTURELLE DE L’ÉCONOMIE MAROCAINE Consequently, one can deduce, as observed by Amaral and al. (2011), that the overall effect of a unit change of final demand is equal to the sum of three terms: interindustry flows, value added, and imported inputs. On the other hand, given equations (3) and (7), one can deduce that, from a macro- economic point of view, total final demand is equal to the sum of total value added and total imports. (8) If we assume that value added coefficients vci=vi/xi and imports coefficients mci=mi/xi are constant and knowing that consequently to a unit increase in the final demand addressed to the jth sector the output of each sector increased by ∆xi=lij, we can write: (9) And if normalized by l.j: (10) This equation, representing the weighted average of value-added and imported inputs coefficients, can also be expressed as follows: (11) Vi and Mi represent the normalized values lij*vci/l.j and lij*mci/l.j. In comparative static, equation (11) permits to detect the change in the productive structure of an economy. Precisely, it permits to evaluate the gain in the capacity of an economy to generate value added (growth effect) and the associated propensity to increase or to decrease external dependency (external dependency effect) from one point of time to another. For each productive sector, two situations are possible. The first is the case where the total effect of a unitary change of sector j final demand decreases between two given years (Δl.j ˂ 0). In this case, as the increase in the global production of the economy in order to satisfy a unitary increase in final demand addressed to sector j must be smaller, the sum ΔVi+ΔMi must be positive. The sector’s performance can thereby be located in one of the four following areas, EQUILIBRES EXTERNES, COMPÉTITIVITÉ ET PROCESSUS DE TRANSFORMATION STRUCTURELLE DE L’ÉCONOMIE MAROCAINE 99 PARTIE I which are ordered from the most advantageous to the worst one. • Area A: the most virtuous area, where are located sectors that are likely to increase the capacity of the economy to generate more value added (ΔVi >0 ) and to reduce at the same time its need for imported inputs (ΔMi< 0). Of course, in this situation, we have ΔVi > ǀ ΔMiǀ. • Area B: where are located sectors whose efficiency effect, i.e. increase of distributed value added (ΔVi >0 ), dominates the increase of imported inputs (ΔMi >0). In this situation, we have ΔVi > ΔMi . • Area C: where are located sectors whose value added and imports increase (ΔVi>0, ΔMi >0), but external dependency effect dominates net efficiency effect (ΔMi > ΔVi). • Area D: the disadvantageous area, where are located sectors that decrease the capacity of the economy to generate value added (ΔVi <0) and increase its external dependency (ΔMi >0). Of course, in this situation we have ǀΔViǀ<ΔMi. The second situation occurs when the total effect on the output of a final demand increases between two given years. In this case, as the increase in the total output of the economy in order to satisfy a unitary increase in final demand must be higher, the sum ΔVi +ΔMi must be negative. Here also, productive sectors can be located in one of the four following areas, ordered bellow in a descending order. • Area A*: grouping sectors, with ΔVi > 0 and ΔMi< 0 and ΔVi < ǀΔMiǀ; • Area B*: grouping sectors, with ΔVi < 0 and ΔMi< 0 and ǀΔVi ǀ < ǀΔMiǀ; • Area C*: grouping sectors, with ΔVi <0 and ΔMi< 0 and ǀΔViǀ > ǀΔMiǀ; • Area D*: grouping sectors, with ΔVi <0 and ΔMi> 0 and ǀΔViǀ > ΔMi. The table below summarizes the characteristics of each area. Table 1. Sectors’ classification according to efficiency and external dependence effect SITUANTION 1 SITUATION 2 Area A: ΔVi > 0 and ΔMi< 0 and ΔVi > ǀΔMiǀ Area A*: ΔVi > 0 and ΔMi< 0 and ΔVi < ǀΔMiǀ Area B: ΔVi > 0 and ΔMi> 0 and ΔVi > ΔMi Area B*: ΔVi < 0 and ΔMi< 0 and ǀΔVi ǀ < ǀΔMiǀ Area C: ΔVi > 0 and ΔMi> 0 and ΔVi < ΔMi Area C*:ΔVi <0 and ΔMi< 0 and ǀΔViǀ > ǀΔMiǀ Area D: ΔVi <0 and ΔMi> 0 and ǀΔVi ǀ < ΔMi Area D*: ΔVi <0 and ΔMi> 0 and ǀΔViǀ > ΔMi The changing distribution of productive sectors of an economy among the different areas permits to appreciate the extent of its structural change. Precisely, if there is a tendency of the net gain on efficiency effect to dominate the external dependency effect (more sectors in area A for instance), we can conjecture that there is an ongoing structural change and that the concerned economy develops over time. 100 EQUILIBRES EXTERNES, COMPÉTITIVITÉ ET PROCESSUS DE TRANSFORMATION STRUCTURELLE DE L’ÉCONOMIE MAROCAINE COMMERCE EXTÉRIEUR ET TRANSFORMATION STRUCTURELLE DE L’ÉCONOMIE MAROCAINE III. Data and domestication of Moroccan input output tables The application of the method described above for the Moroccan case is hampered by two major problems related to the construction of the Moroccan input-output tables. The first problem lies in the fact that Moroccan I-O tables provide data on production and imports on the basis of the base price, while their distribution to intermediate and final uses is given on the basis of the purchase price. Given our objective in this paper, it is necessary to express the vector of imports on the basis of the purchase price. This requires allocating different costs between internally generated resources and those of external origin. Some costs are easily and entirely allocated to one or the other origin. This is the case of the VAT where we can distinguish VAT on imports and domestic VAT. For the allocation of other costs and margins, we use the import coefficient, i.e. the share of imports in total resources, as a distribution key. The second and main problem lies in the treatment of imported inputs. Generally, there are two configurations in this respect. The first case is when a separate import matrix is available, in addition to the domestic one. In this case, the application of the traditional input-output analysis is straightforward and poses no particular problems. The second case, which is the most common one, both imports and domestically produced inputs are included in the intermediate transactions matrix. This is the case of Moroccan input-output tables, which do not give separate matrices for inputs produced locally and those imported. Only their sum is known54. In this situation, the multipliers derived from Leontief inverse matrix tend to overestimate the impact of a change in final demand on domestic industry outputs (Dietzenbacher et al. 2005; Reis and Rua, 2006). In reality, imported inputs represent leakages that reduce industries’ linkages and thereby the magnitude of multipliers. One way to escape these problems is to separate domestic from imported inputs. We apply in this paper a domestication technique based on an assumption commonly referred to as import similarity where for each product the mix of imports and domestically produced goods is the same across all productive sectors, but may be different for each product (see Miller and Blair, 2009). Let us suppose that the fraction of a given input supplied by imports is the same for each industry and that fraction also applies to final users, then that same proportionality holds 54  In this model there is the interesting implicit assumption of the substitutability of imported and domestically produced intermediate inputs. Indeed, instead of assuming that both Ad (domestic inputs matrix) and Am (imported inputs matrix) are constant, only A is required to be constant, which is a reasonable approximation for the short- run. It is straightforward that: A=Ad+Am. EQUILIBRES EXTERNES, COMPÉTITIVITÉ ET PROCESSUS DE TRANSFORMATION STRUCTURELLE DE L’ÉCONOMIE MAROCAINE 101 PARTIE I between total output and total imports55. If we denoted that fraction by θi we can write mi=θi * xi and given equation (3), we can deduce: (12) On the other hand, we know that imported goods are used by productive or by final demand sectors. So, we can write . Consequently, the estimated matrix of imported intermediate consumption and the estimated vector of imports used directly by final sectors is obtained from the following equations and . Finally, we can domesticate the Moroccan I-O tables by subtracting the imported inputs and imports used by final users from the initial table, as follows: (14) (13) Where zijd (respectively ydi) are goods produced locally by sector i and used as inputs by productive sectors (respectively used by final consumers for final consumption). Note that there are other alternatives for the methodology presented here to domesticate the I-O tables. One of them is to assume implicitly that there are no imports consumed directly by final demand, which is probably seldom the case (Dietzenbacher and al. 2005; Reis and Rua, 2006; and Miller and Blair, 2009). IV. Results and discussion We have applied the method described above to the Moroccan economy in the period 1999- 2009, after domestication of the national Input-Output Tables with 20 sectors. The data are from national statistics authority (Haut commissariat au plan). The table 2 below shows the evolution of backward linkages of each industry over the period 1999-2009. It ranks sectors according to the intensity of the backward linkages. 55  This assumption may not be very realistic in many developed economies as the Moroccan one but is often necessary due to the limits on available data. 102 EQUILIBRES EXTERNES, COMPÉTITIVITÉ ET PROCESSUS DE TRANSFORMATION STRUCTURELLE DE L’ÉCONOMIE MAROCAINE COMMERCE EXTÉRIEUR ET TRANSFORMATION STRUCTURELLE DE L’ÉCONOMIE MAROCAINE Table 2. Evolution of backward linkages and ranking of productive sectors (1999-2009) Two main conclusions are to draw from table 2. First, it appears that the hierarchy of sectors has not changed significantly over the studied period. Few key sectors have emerged from this point of view56. The second conclusion is the weaknesses of backward linkages and the apparent general deterioration of integration of the Moroccan productive system between 1999 and 2009. Indeed, the intensity of backward linkages decreases for almost all sectors. This means that, in order to satisfy a unitary increase of final demand addressed to those sectors, a smaller increase in the global domestic production of the economy is necessary57. This is not automatically a bad news provided that this decrease in production has not been substituted by imports. Table 3 shows, for each sector over the period 1999-2009, the changes of backward linkages, external dependency and value added generation. It gives also the area of localization of each 56  This includes, in particular, the sector “Refined petroleum and other energy products”. The activity of this sector is related, at least in part, to the increase of international prices and and some services as transport and the posts and telecommunications sector. In 2009, the sector of transport emerged as the leading sector, while it was ranked third in 1999, behind “Food and tobacco” and “textile and leather industries”. “Posts and telecommunications” sector gained five places, to the 13th place in 2009. “Financial and insurance activities” showed the same performance, to 11th place instead of 16th in 1999. 57  The average backward linkage dropped from 1.38 in 1999 to 1.31 in 2009. EQUILIBRES EXTERNES, COMPÉTITIVITÉ ET PROCESSUS DE TRANSFORMATION STRUCTURELLE DE L’ÉCONOMIE MAROCAINE 103 PARTIE I sector (Table 1). Table 3. Changes of sector’s backward linkages, imports and generated value added generation between 1999 and 2009 Code Sector ΔBL ΔM ΔV Area A00 Agriculture, forestry, hunting and exploitation -0,12 0,00 0,07 Area B B05 Fishing and aquaculture -0,12 0,03 0,06 Area B C00 Extractive industry (Mining) -0,14 0,26 -0,16 Area D D01 Food and tobacco industry -0,16 0,06 0,01 Area C D02 Textile and leather industries -0,13 0,02 0,04 Area B D03 Chemical and Para-chemical industries -0,18 0,23 -0,13 Area D D04 Mechanical, metallurgical and electrical industries -0,08 0,37 -0,32 Area D D05 Other manufacturing industries (outside petroleum refining) -0,17 0,17 -0,08 Area D D06 Refined petroleum and other energy products 0,12 1,10 -1,23 Area D* E00 Electricity and Water -0,09 0,10 -0,04 Area D F45 Building and public works -0,20 0,00 0,12 Area B G00 Commerce and repair -0,02 0,00 0,01 Area B H55 Hotels and restaurants -0,16 0,03 0,04 Area B I01 Transport 0,02 0,11 -0,12 Area D* I02 Posts and Telecommunications 0,11 -0,02 -0,07 Area C* J00 Financial activities and insurance 0,00 0,01 -0,02 Area D* K00 Real estate, rental and services to companies -0,03 0,03 -0,01 Area D L75 General Public Administration and Social Security -0,02 0,00 0,01 Area B MNO Education, health and social action -0,01 0,00 0,00 Area B OP0 Other non-financial services -0,08 0,00 0,06 Area B Our results show, that over the period studied, an increase of final demand addressed to domestic production generates leaks on imports and benefits to foreign agents. The striking result that emerges from this table is that no productive sector is located in the two most virtuous areas (areas A and A*) including, as described in the methodology before, sectors allowing an increase of the value added generated by the economy and a decrease of its propensity to import. This confirms the weaknesses of the Moroccan economy and its increasing reliance on imported products and consequently the seriousness of threats on its current account. In addition, there are more sectors with decreasing backward linkages (16 sectors) than with increasing backward linkages (4 sectors). Does this fact reflects improvement of economic efficiency or, conversely, increased leakage and an aggravation of external dependency? Unfortunately, the available evidence suggests that the external dependency increase is most likely occurring. We showed that when backward linkages decrease, the sum of the change in value added 104 EQUILIBRES EXTERNES, COMPÉTITIVITÉ ET PROCESSUS DE TRANSFORMATION STRUCTURELLE DE L’ÉCONOMIE MAROCAINE COMMERCE EXTÉRIEUR ET TRANSFORMATION STRUCTURELLE DE L’ÉCONOMIE MAROCAINE and the imports change is positive. In this situation, our results show that nine (9) sectors are located in area B (value added higher than imports leakage). The nature of sectors included in area gives an idea about the “leading” sectors in Morocco. These sectors includes: 1. Utilities and low technological content sectors (Agriculture, Forestry, Hunting and Exploitation, Fishing and Aquaculture, Commerce and Repair,...), 2. Protected and domestically oriented sectors (Building and public works, General Public Administration and Social Security, Education, Health and Social Action) and traditional productive sectors (Hotels and restaurants, Textile and leather industries). However, the majority of sectors with some technological content, as Mechanical, metallurgical and electrical industries, Chemical and Para-chemical industries… are located in the most disadvantageous area (Area D), involving a greater external dependency and lower value added generation. Another negative tendency in the Moroccan economy appeared when we analyze the performance of sectors that their backward linkage indicators increase between 1999 and 2009. This positive variation means, to reiterate, that in order to satisfy a unitary increase in final demand of each sector for them, it is necessary a higher increase in the global production of the economy. This is not necessarily good fact if this increase in production reflects a decline of its ability to generate value added or has also induced a rise of imports. We showed before that in this situation, the sum of generated value added and the induced imports leakage is negative. The results shows, nevertheless, that no sector is located in the area A* (more value added and less imports) nor the area B* (lower decrease in value added than imports). This is an important concern to the extent that the sectors whose backward linkages increased over the period (Transport, Financial activities, Posts and Telecommunications and Refined Petroleum…) do not necessarily promote structural transformations in Morocco. Their apparent dynamism is more due to leaks they are causing to the economy without being able to generate more value added. The table below summarizes the results and gives the shares of each area in total output, total value added, and total employment for the year 2009. Table 4. Share of each area in total output, total value added and total employment-2009 Areas A A* B C B* C* D* D Nb of sectors 0 0 9 1 0 1 3 6 % to total output 0 0 38,8 10,7 0 2,6 10,2 37,3 % to total outputValue added 0 0 56,9 4,4 0 3,2 9,2 25,8 % to total employment 0 0 86,4 1,5 0 0,6 1,2 7,3 Table 4 shows in particular that almost the half of Moroccan productive sectors (9 of 20) are located in the two most disadvantageous areas (D and D*) including sectors that influence EQUILIBRES EXTERNES, COMPÉTITIVITÉ ET PROCESSUS DE TRANSFORMATION STRUCTURELLE DE L’ÉCONOMIE MAROCAINE 105 PARTIE I negatively the ability of the national economy to generate more value added and increase, conversely, its external dependency. These 9 sectors represent, at the end of 2009, 47,5% of total output, 35% of GDP and 8,5% of total employment. It is worth to note that the concerned sectors are those with significant capital intensity. To refine our analysis, we have distinguished between two subperiods: 1999-2004 and 2004- 2009, where the annual average growth rate of Moroccan real GDP was similar (respectively 4% and 4,5%). The results are given in the appendix. Our previous conclusions remain broadly the same and confirm the negative impact of international trade on the integration of Moroccan productive sectors. The backward linkages declined for most of productive sectors. This reinforces their external dependency. During the second subperiod (2004-2009), backward linkages increased only for three sectors (General public administration and social Security, Education, health and social action, Refined petroleum and other energy products) against six sectors in the first subperiod (Extractive industry-Mining, Commerce and repair, Transport, Posts and telecommunications, Financial activities and insurance, Refined petroleum and other energy products). More importantly, over the two sub-periods, these sectors belong to the disadvantageous area (Area D*), where external dependency increases and ability to generate value added declines. The increase of external dependency of the Moroccan economy and the increase of the number of sectors generating less value added is also noticeable when we analyze the performance of productive sectors with negative variation of backward linkages. In this case, the number of sectors included in the most disadvantageous area (area D) extended from three sectors during the period 1999-2004 (Chemical and Para-chemical industries, Mechanical, metallurgical and electrical industries, Other manufacturing industries) to eight sectors during the period 2004-2009. More importantly, the latter concern all branches of industry (except Textile and leather industries) and some services with reasonable technological content and high positive externalities (Telecommunications, financial activities, Real estate, rental and services to companies). It is true that the sectors belonging to the most virtuous area (Area A) have also increased from two sectors (Electricity and Water, Hotels and restaurants) to five (Extractive industry- Mining, Agriculture, Transport, Textile and leather industries, Other Non-Financial Services). In fact, only the Extractive industry that really emerged, while others sectors value added increase only slightly. The dark side of this situation is that these sectors remained traditional with low technological content; the modern and higher productivity sectors are localized elsewhere. The comparison of the shares of each area in total output, total value added, and total employment gives more evidence on the shortcomings recorded by the Moroccan economy in reducing its external dependency and of using its resources where they are most productive. For example, during the period 2004-2009, the productive sectors that induce more imports leakage and generate less value added represent on average of 57% of total output against 30% for the 106 EQUILIBRES EXTERNES, COMPÉTITIVITÉ ET PROCESSUS DE TRANSFORMATION STRUCTURELLE DE L’ÉCONOMIE MAROCAINE COMMERCE EXTÉRIEUR ET TRANSFORMATION STRUCTURELLE DE L’ÉCONOMIE MAROCAINE ones that generates more value and lower imports (36% against 4% during the period 1999- 2004). The two tables below give more details in this respect. Table 5. Average share of each area in total output, total value added and total employment during the period 1999-2004  Areas A A* B C B* C* D* D Nb of sectors 2 0 6 3 0 1 5 3 % to total output 4,33 0 30,9 26,6 0 1,9 13,9 22,4 % to total value added 5,24 0 30,6 28,8 0 2,8 24,4 8,2 % to total employment 2,05 0 22,2 53,2 0 0,3 17,8 4,3 Table 6. Average share of each area in total output, total value added, and total em ployment (2004-2009)  Areas A A* B C B* C* D* D Nb of sectors 5 0 3 1 0 0 3 8 % of total output 29,9 0 11,0 2,2 0 0,0 13,3 43,5 % of total value added 24,6 0 18,5 2,5 0 0,0 18,5 35,9 % of total employment 57,8 0 22,9 2,0 0 0,0 8,8 8,6 Finally, it should be noted that in Morocco there is no domestic input-output flows at constant prices. Input-output tables at current prices do not allow to separate prices effects from quantity/ real effects. Dietzenbacher and Temurshoev (2012) showed that the impact’s analysis differs when constant prices are used instead of current prices. EQUILIBRES EXTERNES, COMPÉTITIVITÉ ET PROCESSUS DE TRANSFORMATION STRUCTURELLE DE L’ÉCONOMIE MAROCAINE 107 PARTIE I Concluding remarks The Input-Output Model (IO) provides a powerful tool to analyze interindustry linkages and to identify key sectors in a given economy. It is also increasingly used in the analysis of the structural change of economies (Sonis and al. (1995), Bureau of Economic Analysis (2000)) Dietzenbacher and van der Linden (1997)), Dridi and Hewings (2002). We have attempted to assess the external dependency of Moroccan productive sectors and the associated low value added generated in domestic production by using the Leontief inverse matrix. Our strategy here is simple and draws heavily on the work of Amaral, Lopez and Dias (2011). The results reveal the reduced capacity of the majority of Moroccan productive sectors to generate value added and to reduce external dependency, particularly in recent years. During the period 1999-2009, backward linkages decreased for 16 of 20 Moroccan productive sectors. This decrease in production needed to satisfy the increase in domestic demand (reduction of the output multiplier) does not reflect gains in efficiency but an increase of external dependency in that imported inputs have supplanted domestic ones. As a result, no Moroccan productive sector is located in the most virtuous areas that includes sectors recording an increase in their ability to generate more value added and to rely less on imports. This confirms the failure of the Moroccan economy to catch up emerging countries. Naturally, we expect that as an economy develops most sectors should be located in the virtuous areas. In addition, our results show that sectors with some content of technology, especially industrial activities and some services, belongs to the disadvantageous areas of less value added generation and increasing external dependency. The importance of these results paves the way for other additional analyzes to better assess the process of structural transformation of the Moroccan economy. Two perspectives are, from our point of view, crucial. First, it is important to use disaggregated data over a long enough period, as that process is inherently complex and secular. Second, the comparative static analysis developed here should be completed by a dynamic approach, which is more able to provide relevant insights into the structural changes in the national economy. 108 EQUILIBRES EXTERNES, COMPÉTITIVITÉ ET PROCESSUS DE TRANSFORMATION STRUCTURELLE DE L’ÉCONOMIE MAROCAINE COMMERCE EXTÉRIEUR ET TRANSFORMATION STRUCTURELLE DE L’ÉCONOMIE MAROCAINE References • Agenor P.R, & El Aynaoui, K. (2015). Maroc : Stratégie de croissance à l’horizon 2025 dans un environnement international en mutation, OCP Policy Center • Amaral, J. F, lopes, J.C. & Dias, J. (2011). External dependency, value added generation and structural change: an inter-industry approach, Notas economicas, 11, pp: 6-19 • Avonds, L, Seguel, V et Gilot, A (2003)  . Quelques applications à l’aide du tableau entrées-sorties 1995, Working paper, n°18-03,: pp 1-50 • Banque Africaine de Developpement, Gouvernement du Maroc, MCC (2015). Diagnostic de croissance du Maroc : Analyse des contraintes à une croissance large et inclusive. http://www.afdb.org/fileadmin/uploads/afdb/Documents/Generic- Documents/Diagnostic_de_croissance_du_Maroc_%E2%80%93_Analyse_des_ contraintes_%C3%A0_une_croissance_large_et_inclusive.pdf • Bess, R. & Ambargis, Z. O. (2011). Input-output models for impact analysis: Suggestions for practitioners using rims ii multipliers, Paper presented at the 50th Southern Regional Science Association Conference, New Orleans, Louisiana: pp: 1-28 • Bureau Federal du Plan (2012).  Analyse entrées-sorties : Modèles, multiplicateurs, linkages. Working paper n°12-12, pp : 1-57 • Chafiki, M, (2014). L’économie marocaine à l’épreuve de l’ouverture et de la transformation structurelle : Défis et enjeux. Ministère de l’économie et des finances du Maroc • Dietzenbacher, E. & Van Der Linden, J., (1997). Sectoral and spatial linkages in the EC production structure, Journal of Regional Science, 37, pp: 235-257. • Dietzenbacher, E, Albino, V & Kuhtz, S. (2005). The fallacy of using US-type input-output tables, Mimeo (Paper presented at the 15th International Conference on Input-Output Techniques) pp: 1-15 • Dietzenbacher, E. & Temurshoev, U, (2012). Input-output impact analysis in current or constant prices: Does it matter?, Journal of Economic Structures, Vol1:4, pp:1-18 • Drejer, I. (2002). Input-output based measures of interindustry linkages revisited: A survey and discussion, Paper presented at the 14th International Conference on Input- Output Techniques, Montréal, Canada. pp: 1-35 • Dridi, C. & Hewings, G. (2002). An investigation of industry associations, association loops and economic complexity: Application to Canada and the United-States, Economic Systems Research, 14, pp: 275-296. • Guo, J & Planting, M.A. (2000). Using input-output analysis to measure U.S. economic structural change over a 24-year period. WP2000-01 pp: 1-23 • Hampÿe, C, (2013). Les multiplicateurs de production, de revenu et d’emploi 1995- 2005 : Une analyse entrées-sorties à prix constants, Bureau fédéral du Plan Analyses et Prévisions économiques, Working Paper 8-13, pp :1-62 EQUILIBRES EXTERNES, COMPÉTITIVITÉ ET PROCESSUS DE TRANSFORMATION STRUCTURELLE DE L’ÉCONOMIE MAROCAINE 109 PARTIE I • Haut-Commissariat au Plan–Royaume du Maroc (HCP) (2009). Les comptes nationaux 1999–2009 (base 1998). • Institut Royal Des Etudes Strategiques (2013). Les accords de libre échange conclus par le Maroc: Quelles incidences sur la compétitivité globale du Maroc?, Rapport entrant dans le cadre du programme d’études «Compétitivité globale et positionnement du Maroc dans le système mondialisé» • Mcmillan, M. & Rodrik.,D (2011). Globalization, structural change, and productivity growth, NBER Working Paper no. 17143, pp: 1-54 • Miller, R, E. and Blair,P. D. (2009). Input-output analysis: Foundations and extensions, Cambridge, Cambridge University Press • Ministere de L’industrie, de Commerce et des Nouvelles Technologies, (2012). Estimation des élasticités du commerce extérieur du Maroc. Etude réalisée dans le cadre de la coopération technique avec la Banque Mondiale • Muniz, G. & Salome, (2013). Modelling linkages versus leakages networks : The case of Spain, Regional and Sectoral Economic Studies, Vol. 13-1 pp: 1-13 • Reis, H. & Rua, A. (2006). An input-output analysis: Linkages vs leakages. Banco de Portugal, Working Papers,n° 17, pp: 1-31 • Sonis, M, Hewings, G. J. D. & Guo, J. (1996). Sources of structural change in input- output systems: a field of influence approach, Economic Systems Research, Vol. 8, n°1, pp: 15-32 • Voyvoda, E. (2009). Sources of structural change and its impact on interdependence: An input-output perspective for the post-1980 Turkish economy, Economic Research Forum, Working Paper n°507, pp :1-15 110 EQUILIBRES EXTERNES, COMPÉTITIVITÉ ET PROCESSUS DE TRANSFORMATION STRUCTURELLE DE L’ÉCONOMIE MAROCAINE COMMERCE EXTÉRIEUR ET TRANSFORMATION STRUCTURELLE DE L’ÉCONOMIE MAROCAINE Appendix Table 7. Changes of imports and value added for sectors with negative variation of backward linkages (1999 – 2004) ΔBL ΔM ΔV AREA A00 -0,04 0,02 0,00 AREA C B05 -0,09 0,03 0,04 AREA B D01 -0,06 0,01 0,02 AREA B D02 -0,10 0,02 0,02 AREA C D03 -0,11 0,20 -0,13 AREA D D04 -0,03 0,02 -0,01 AREA D D05 -0,08 0,05 -0,01 AREA D E00 -0,03 -0,04 0,06 AREA A F45 -0,13 0,00 0,08 AREA B H55 -0,07 -0,01 0,04 AREA A K00 -0,01 0,00 0,00 AREA C L75 -0,02 0,00 0,02 AREA B MNO -0,03 0,00 0,02 AREA B OP0 -0,05 0,01 0,03 AREA B Table 8. Changes of imports and value added for sectors with negative variation of backward linkages (2004– 2009) ΔBL ΔM ΔV AREA C00 0,01 0,70 -0,71 Area D* D06 0,05 0,31 -0,34 Area D* G00 0,01 0,00 -0,01 Area D* I01 0,08 0,12 -0,17 Area D* I02 0,12 -0,03 -0,07 Area C* J00 0,00 0,00 -0,02 Area D* Table 9. Changes of imports and value added for sectors with positive variation of backward linkages (2004– 2009) ΔBL ΔM ΔV AREA D06 0,0760 0,7963 -0,8892 AREA D* L75 0,0047 0,0000 -0,0102 AREA D* MNO 0,0158 0,0000 -0,0174 AREA D* EQUILIBRES EXTERNES, COMPÉTITIVITÉ ET PROCESSUS DE TRANSFORMATION STRUCTURELLE DE L’ÉCONOMIE MAROCAINE 111 PARTIE I Table 10. Changes of imports and value added for sectors with positive variation of backward linkages (2004– 2009) ΔBL ΔM ΔV AREA A00 -0,080 -0,021 0,071 AREA A B05 -0,029 0,004 0,018 AREA B C00 -0,147 -0,446 0,549 AREA A D01 -0,095 0,048 -0,011 AREA D D02 -0,030 -0,008 0,017 AREA A D03 -0,062 0,032 -0,004 AREA D D04 -0,047 0,343 -0,317 AREA D D05 -0,084 0,120 -0,070 AREA D E00 -0,057 0,145 -0,106 AREA D F45 -0,072 0,000 0,038 AREA B G00 -0,033 0,000 0,020 AREA B H55 -0,083 0,038 0,001 AREA C I01 -0,061 -0,004 0,052 AREA A I02 -0,017 0,016 -0,007 AREA D J00 -0,003 0,003 -0,002 AREA D K00 -0,016 0,025 -0,013 AREA D OP0 -0,029 -0,002 0,025 AREA A 112 EQUILIBRES EXTERNES, COMPÉTITIVITÉ ET PROCESSUS DE TRANSFORMATION STRUCTURELLE DE L’ÉCONOMIE MAROCAINE ISBN : 978-9954-9636-2-3 396 EQUILIBRES EXTERNES, COMPÉTITIVITÉ ET PROCESSUS DE TRANSFORMATION STRUCTURELLE DE L’ÉCONOMIE MAROCAINE