WPS8218 Policy Research Working Paper 8218 Product and Factor Market Distortions The Case of the Manufacturing Sector in Morocco Jean-Pierre Chauffour Jose L. Diaz-Sanchez Macroeconomics and Fiscal Management Global Practice Group October 2017 Policy Research Working Paper 8218 Abstract This paper studies the effect of market distortions in the man- increase in TFP would be of 56 percent. The paper also ufacturing sector in Morocco. Recent microdata are used to finds that industries that are more opened to competition calculate the extent of resource misallocation associated to (international and domestic) such as machinery and textiles these distortions and the potential total factor productivity industries present lower levels of market distortions com- (TFP) gain resulting from their removal. Market distortions pared with more protected industries with relatively little in the manufacturing sector in Morocco are higher com- competition, such as the food industry. Besides, a higher pared with developed countries and slightly more important level of TFP can be achieved if more resources are allocated compared with other developing countries, such as China to “young” and “small” firms. The main results of the paper and India. These distortions decreased between 2007 and are robust to an alternative estimation that uses a different 2013. Full liberalization would raise TFP by about 84 methodological framework with a less extensive theoretical percent. If distortions are removed to the level of selected framework. The paper discusses policies to further limit the developed countries with better resource allocation, the extent of product and factor market distortions in Morocco. This paper is a product of the Macroeconomics and Fiscal Management Global Practice Group. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://econ.worldbank.org. The authors may be contacted at jdiazsanchez@worldbank.org and jchauffour@worldbank.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team Product and Factor Market Distortions: The Case of the Manufacturing Sector in Morocco Jean-Pierre Chauffour and Jose L. Diaz-Sanchez∗ The World Bank JEL Code: O11, 047, 055, L60 Keywords: TFP, Market Distortions, Resource Misallocation, Productivity, Manufacturing. ∗ Respectively Lead Economist for the Maghreb region and Economist, World Bank. The views expressed in this article are those of the authors and do not necessarily represent those of the World Bank. We would like to thank Abdessamad Saidi, Anta Ndoye, Fernando Parro, Addisu Lashitew, Ragbi Aziz, and participants of a research seminar of the Bank Al-Maghrib for useful comments and discussions. 1 Introduction To accelerate its pace of economic convergence, Morocco needs to achieve and sustain faster economic growth. Dabla-Norris et al. (2013) shows that episodes of growth take-off in emerging markets have been mostly associated with a strong pick-up in total factor productivity (TFP) growth. Yet, in the last two decades, average TFP growth in Morocco has been much lower than in other emerging mar- ket economies, especially in East and South Asia. Policies to lift-up TFP would appear therefore to be critical to accelerate the Moroccan economic convergence process (World Bank, 2017a). Productivity increases with the adoption and development of new technologies (Howitt, 2000; Klenow and Rodriguez-Clare, 2005; and Coe and Helpman, 1995) and with better institutions and higher social and human capital (World Bank, 2017a and 2017b). In addition, the structural change process of allocating re- sources from less productive to more productive sectors has also been found to be one of the main sources of TFP growth (McMillan, Rodrick, and Verduzco-Gallo, 2014). In this paper we focus on another significant determinant of aggregate productivity explored in the literature: resource misallocation across firms. Seminal papers in this literature include Restuccia and Rogerson (2008), Hsieh and Klenow (2009), Bartelsman, Haltiwanger, and Scarpetta (2013), and Asker, Collards-Wexler, and De Loecker (2014) and suggest that large heterogeneity in firm-level productivity within a sector may indicate misallocation of resources across firms, with a consequent negative and sizable effect on the aggregate TFP (see Restuccia and Rogerson, 2017 for a recent survey on this topic). In this paper we apply the Hsieh and Klenow’s (2009) framework (henceforth HK) to Morocco, which allows to first quantify the level of distortions in the Moroccan manufactur- ing sector, and second to compute the productivity gains from eliminating these distortions. In the HK framework, distortions to firm size (output distortions) and capital/labor ratio (factor distortions) lead to different marginal products of capital and labor across firms within industries. Productivity gains then can be quantified by equalizing marginal products across firms.1 Examples of output distortions include regulations restricting market access and/or firm size, bribes that have to be paid by firms in order to operate, pub- lic output subsidies, trade policies in the form of import protection and export subsidies (taxes), and distortions generated from exchange rate policies and price controls. Factor distortions are credit constraints, financial policies, tax policies, and labor regulations that differ across firms (for instance because of political connections) and affect the relative prices of capital and labor. Note that output and factor distortions can simultaneously affect both the size of a firm and the relative prices of capital and labor. The present paper proposes to investigate the effects of market distortions in the manufacturing sector in Morocco. It calculates the level of resource misallo- cation and the potential TFP gains associated to liberalization. It uses microdata 1 Note that this is an indirect approach to analyzing misallocation. This approach contrast with direct approaches, which analyze the impact of specific frictions (for instance see Buera, Kaboski, and Shin, 2011; and Moll, 2014, for the case of financial frictions). 2 from two World Bank’s Enterprise Surveys (2007 and 2013) for Morocco. Re- cently, Ragbi and Nihou (2016) undertake a similar study. However, even if they also apply the HK framework, their results can not be compared directly with ours.2 The paper finds that market distortions in the manufacturing sector in Mo- rocco are higher compared to developed countries and somewhat higher compared to other developing countries such as China and India. Empirical results provide evidence in the reduction of market distortions between 2007 and 2013. In ad- dition, we estimate about 84 percent TFP gains from a full liberalization and 56 percent from removing distortions to the level of selected developed countries with better resource allocation. Thus, the removal of economic distorsions has the potential to considerably accelerate the convergence process. The paper also finds that industries more opened to international and domestic competition (e.g. ma- chinery and textiles) present lower levels of output and factor distortions compared to industries where competition is more limited (e.g. food industry). Finally, the results show that a higher level of TFP can be obtained if more resources are allocated to younger and smaller firms. The paper’s main results are robust to an alternative estimation that uses a different methodological framework with a less extensive theoretical framework (Bartelsman, Haltiwanger, and Scarpetta, 2013). The remainder of the paper is organized as follows. Section 2 presents an overview of market distortions in Morocco. Section 3 gives a brief presentation of the HK methodology. Section 4 presents the data and describes the construction of the required variables for the study. The empirical analysis is contained in Section 5, where we quantify the level of distortions and compute the TFP gains from resource reallocation. Section 6 discusses policies to further limit the extent of product and factor market distortions in Morocco. Section 7 provides the final remarks. 2 Market Distortions in Morocco The free movement of capital and labor to the most productive manufacturing firms is restrained by market distortions, including distortions in Morocco’s in- ternational exchange system. In this section we present an overview of the main market imperfections that currently affect the optimal allocation of resources in both the tradable and nontradable sectors of the Moroccan economy. To be consis- tent with the grouping of market distortions in the paper’s theoretical framework section (Section 3), we describe here separately output and factor distortions. Out- put distortions affect the quantity of production leaving the input mix unaffected. These distortions are more related to limited competition, deteriorated business environment, and reduced international trade. Factor distortions affect the use of capital relative to labor and are mostly associated to high credit constraints and 2 In contrast to Ragbi and Nihou (2016), our empirical exercise takes into account a recent revision of the HK framework (see Hsieh and Klenow, 2013). Although most of their analysis is not affected by this change, the level of distortions and the TFP gains associated to a full liberalization cannot be compared with ours or with recent papers in this literature. 3 labor rigidities. 2.1 Competition, Business Environment, and International Trade There are many factors limiting competition in the Moroccan economy. According to a study by the Royal Institute of Strategic Studies (IRES), “by reducing compe- tition, protection allows for the emergence of monopolistic rents. Businesses that benefit from this situation have used price increases to collect super profits rather than investing in competitiveness through quality and new techniques” (IRES, 2014). Like in most countries in the Middle East and North Africa (MENA) re- gion (Rijkers, Freund, and Nucifora 2014, Diwan, Keefer, and Schiffbauer, 2016, Diwan and Hadar, 2016), Morocco’s capitalism is dominated by personal connec- tions and privileges, which are an integral part of a complex network of economic interests that hinder the emergence of new players (World Bank, 2009, 2015b). The effect is to maintain monopolies, impede the performance of the economy and weaken the growth of productivity and job creation. A recent study estimates that almost 13 percent of a representative sample of Moroccan manufacturing compa- nies are held by individuals connected with circles of economic influence (Saadi 2016). These businesses operate in various economic sectors and are apparently more effective than their unconnected counterparts owing, in part, to the privi- leges that they enjoy in terms of commercial protection, application of the rules and regulations and access to financing. Beyond considerations of equity, the ex- istence of connected enterprises can partially explain the slow overall structural transformation of the economy to the extent that these firms operate in sectors with low new business entry and exit rates and greater concentration. While some sectors have liberalized (mobile telephony, for example), this progress cannot mask the lack of real competition between economic agents in many other sectors.3 The regulatory framework governing competition has been updated, but its implementation has been suspended since 2014 pending the ac- tual appointment of the members to the new Competition Council. Despite the many reforms aimed at refocusing the government on its general interest missions, public enterprises and establishments (EEPs) continue to play a major role in the national economy, creating de facto entry barriers to key sectors (agriculture and marine fisheries; energy, mines, water and the environment; infrastructure and transportation; housing, urban planning and local and regional development). In- deed, because of their status, modes of governance, public financial support and often monopolistic position they can constitute an obstacle to the opening up of many sectors.4 3 The density of new businesses (a proxy for competition strength) in comparison with the total population is higher than in other countries in the region but low in comparison with many emerging market economies (Klapper and Love, 2010). 4 At the end of 2015, the EEP sector represented about 8 percent of gross domestic prod- uct (GDP) and almost 25 percent of the country’s total investment through 212 public estab- lishments, 44 companies in which the government had direct equity holdings and some 442 subsidiaries and public shareholdings. 4 Over the past 15 years, Morocco has made significant changes to its business environment and its public policies to modernize the economy and encourage effi- ciency. The efforts to improve the business climate have been recognized in many international rankings, such as the World Bank Doing Business, and owing in par- ticular to business start-up reforms (World Bank, 2015a).5 However, the business climate is still seen by most players as too unpredictable and bureaucratic and does not inspire the confidence that economic agents need to invest in the medium or long term. The importance of constraints such as burdensome, slow, complex and opaque administrative formalities and procedures is regularly confirmed in surveys of businesses. The constraints most often mentioned by the formal sec- tor are corruption, competition from the informal sector, low workforce education levels and difficulties in accessing financing (World Bank, 2013a). For very small enterprises and small and medium enterprises (VSEs and SMEs), these barriers can be insurmountable (World Bank, 2014a). The incidence of entrenched corruption affecting businesses in Morocco is high (EBRD, EIB, and World Bank 2016). The frequency and incidence of requests for benefits or informal payments faced by Moroccan businesses in their relations with government departments and public services, in their applications for licenses and construction permits, or in import operations or tax payments appear to be significantly higher than the MENA region average, which itself compares unfavorably to other middle-income countries that are Morocco’s competitors. Furthermore, while corruption involves the visible costs related to illicit payments and a poor allocation of resources, its highest cost is invisible and involves missed opportunities: the businesses that are not created, the investments that are not made, the loans that are not granted, and, ultimately, the jobs that are not created. The Moroccan economy has gradually opened up to international trade. Tariff and nontariff barriers have been lowered and international trade procedures have been simplified. Modern trade infrastructure has been developed for all types of transport (such as the Tanger-Med port) and the country has been fully up- dating its customs regulations and its transportation and trade logistics. This has facilitated the emergence of “new” industries in sectors such as automobile, aeronautics, and electronics that occupy a more central position in the global pro- duction space and are thus a potential source of higher revenues than traditional exports. Despite these efforts, Morocco’s share in international trade has tended to decline since the early 1980s, while most of its competitors have seen their shares increase. Morocco trade performance is slowed by high production factor costs owing to protected service sectors and rigid domestic prices, by a fixed exchange rate regime and by a skills and qualifications mismatch. To date, Morocco has been able to penetrate only a small portion of potential markets. One of the explanations for Morocco’s difficulties in benefiting more from the development of international trade lies in the delays and costs related to mer- 5 Morocco’s Doing Business ranking improved from the 129th position to the 68th position during 2010-2016. 5 chandise export and import logistics. According to the Doing Business indicator that measures distance to the frontier in terms of the accumulation of delays and costs (excluding customs duties) associated with three categories of procedures (documentary compliance, border compliance and domestic transport), Morocco scores better than other countries in the MENA region, but not as well as its main competitors, particularly Central and Eastern Europe. Morocco was ranked 86th out of a total of 160 countries in 2016 on the Logistic Performance Index, falling 24 places from its average ranking during the period 2007-14 (World Bank 2016b). Morocco ranks well for the quality of its infrastructure but loses in this classification owing to weaknesses and dysfunctions in foreign trade services. In fact, Morocco is ranked 124th in the provision of customs services (it was ranked 73rd from 2007 to 2014) and is ranked 122nd in terms of the ability to track and trace consignments. Other sources of market distortions are tariff and nontariff trade barriers. At 25 percent, maximum most-favored nation (MFN) tariffs on goods are still relatively high, particularly in comparison with the zero duty applied to preferential trade partners under the free trade agreements signed with the EU, the US and other countries. High duties on intermediate products tend to be an obstacle to the productivity of firms and the growth of Moroccan exports. This results from the fact that generally speaking a tax on imports ends up becoming a tax on exports, and this effect is becoming more important since the difference in export unit values between firms that both import and export and firms that export only has been increasing over time (World Bank, 2017a). Nontariff measures has a similar effect on resource misallocation. 2.2 Credit Constraints and Labor Rigidities The Moroccan financial system operates relatively effectively in mobilizing na- tional savings but has more difficulty in allocating it to the most productive activities. There are many reasons for this, including objective prudential con- straints related to the risk of concentration and maturity transformation and more subjective behaviors such as the reluctance and averseness of commercial banks to risk-taking. Another reason could be the still relatively low level of competition prevailing in the Moroccan banking sector (see Benazzi and Rouiessi, 2017). Ac- cording to the Doing Business indicator on getting credit, Morocco is ranked 101st in the world and is relatively far from the frontier, particularly in comparison with other emerging market economies such as Romania, Cambodia or Kenya. Fur- thermore, even though the government recently undertook reforms to promote the diversification of financial products and mechanisms for the allocation of capital, these reforms have thus far had few concrete results and there is still a significant gap between the advanced development of the banking system and the develop- ment of the equity market. The Casablanca Stock Exchange (CSE) suffers from a lack of new issuers, particularly key large institutional issuers. Family conglom- erates and medium-sized enterprises, which could develop more rapidly, generally seem reluctant to share information concerning their capital and governance. The limited convertibility of the dirham represents another source of capital 6 misallocation. Full convertibility of the dirham is prevented by capital controls still in place for domestic agents for the purchase of foreign exchange, which im- pede the development of national firms. Furthermore, the fixed exchange rate regime encourages, in the case of an overvaluation, economic agents to invest in non-export oriented firms, generating negative market distortions on export- oriented firms (which are often more productive than the average), and contribut- ing to the “anti-export bias” (World Bank, 2006). On the labor front, more than half of the working age Moroccans (15-64 years of age) does not participate in the economic activity of the country, which makes it one of the countries with the lowest employment rates in the MENA region, and across the globe. As a result, in 2016, some 2.7 million youth aged 15-29 (or almost one in three young people) were neither in education, employment or training (NEET). Moreover, most (90 percent) of the private jobs are in the informal sector. Social mobility and the ability to move from a less productive job to a more productive job are limited and take time, particularly for young people and women. Morocco has adopted labor legislation that is based on the conventions and recommendations of the International Labor Organization. The negotiation, adop- tion and implementation of this legislation on the basis of international conven- tions have, however, resulted in cumbersome and restrictive labor market reg- ulations. According to the 2017 Doing Business indicator “Hiring and firing”, the Moroccan labor market is particularly heavily regulated. Moroccan law pro- hibits fixed-term contracts for permanent tasks and limits their duration to 12 months, after which they cannot be renewed. Although the Moroccan law allows a certain amount of flexibility in terms of the hours worked, overtime and premi- ums for night work are expensive compared to international practices. Moroccan legislation on leave (annual leave, statutory leave, maternity/paternity leave) is more generous than that in other emerging market countries that are Morocco’s competitors. The labor regulations governing private sector terminations are re- strictive as only businesses with more than 10 employees may terminate jobs for economic, technical or structural reasons. Reducing staff for economic reasons is subject to a prior agreement with the regional authorities. As a result, the num- ber of employees terminated by medium-sized businesses is very low and almost a quarter of the businesses contacted for the Investment Climate Assessment (ICA) survey in 2008 indicated that they would have reduced their staff if redundancy conditions were less restrictive. The minimum wage in Morocco is high compared with the average national income or the average productivity of workers. In 2015, the minimum wage in urban areas was equal to approximately 100 percent of the national per capita income or more than 50 percent of the average wage in the formal private sector, rates that were extremely high not just in the region but also in comparison with international standards, including the OECD countries. Such high legal minimum wages deter the creation of jobs in the formal sector, particularly for young job- seekers with limited qualifications. Overall, the high cost of labor in the formal sector contributes to a lower demand for employees on the part of employers and 7 explains the low labor market participation rate and structural underemployment. 3 Theoretical Framework In this section we provide a sketch of the Hsieh and Klenow (2009) model, incor- porating their correction appendix (Hsieh and Klenow, 2013).6 This is a standard model of monopolistic competition with heterogeneous firms, a Melitz-type model without international trade. The key difference is that instead of Melitz (2003) firms are not only heterogeneous with respect to their productivity but also face idiosyncratic distortions to their output and factor prices. The HK framework is composed of a closed economy with two inputs, labor and capital. There are S differentiated intermediate goods industries and one single homogeneous final goods industry. Inputs are only used in the intermediate goods industries. The final goods industry is merely an assembly sector that combines the S different sets of differentiated intermediate goods to produce a homogeneous final output. Intermediate firms compete on monopolistic competitive markets to sell their differentiated outputs to final producers. Final producers compete in a perfectly competitive market to sell their products to a representative final consumer. Final producers of the final good Y combines the output Ys of S manufacturing industries using a Cobb-Douglas production technology: S S Y = Y s θs , where θs = 1 (1) s=1 s=1 Industry output Ys is itself a CES aggregate of Ms differentiated products: σ Ms σ −1 σ −1 Ys = Y si σ (2) i=1 Each differentiated firm uses a Cobb-Douglas technology function: αs 1−αs Ysi = Asi Ksi Lsi (3) It is important to note that capital and labor shares are allowed to differ across industries (but not across firms within an industry). Two types of distortions are introduced: output distortions that affect the quantity of production while leaving the input mix unaffected (denoted by τY ), and factor distortions that affect the use of capital relative to labor (denoted by τK ).7 The profit equation below of firm i operating in sector s make explicit the market distortions: 6 See their work for a more detailed exposition. 7 HK shows that the overall market distortion in this framework is equivalent to a modified model where instead of output and factor distortions there are only capital and labor distortions. 8 πsi = (1 − τYsi )Psi Ysi − ωLsi − (1 + τKsi )RKsi (4) where Psi is the output price fixed by each firm, R is the rental price of capital and ω is the wage common to all firms in a given sector. Profit maximization in a monopolistic competition setting yields the standard condition that the firm’s output price is a fixed markup over its marginal cost: αs 1−αs σ R ω (1 + τKsi )αs Psi = (5) σ−1 αs 1 − αs Asi (1 − τYsi ) Equation (6) and (7) shows the expression of τYsi and τKsi from the first-order conditions: αs ωLsi 1 + τKsi = (6) 1 − αs RKsi σ ωLsi 1 + τYsi = (7) σ − 1 (1 − αs )Psi Ysi Factor distortions (τKsi ) appear when the ratio of labor compensation to the capital stock differs relative to what one would expect from the degree of output elasticity with respect to capital and labor. τKsi is positive in a firm paying higher capital costs, and it is negative if the labor cost is relatively high.8 Similarly, output distortion (τYsi ) is measured when labor’s share is different compared with what one would expect given the degree of industry elasticity of output with respect to labor. The output distortion gives then an indication of how suboptimal is the firm size (in terms of labor). Also from the first-order conditions, Ysi is given by: Aσ σ si (1 − τYsi ) Ysi = (8) (1 + τKsi )αsσ Thus the production and allocation of resources across firms depends not only on firm TFP levels, but also on the output and capital distortions they face. In addition, if resource allocation is driven by distortions rather than firm TFP, this will result in differences in the marginal revenue products of labor and capital across firms. The marginal revenue products of labor and capital are respectively: σ − 1 Psi Ysi 1 M P RLsi (1 − αs ) =ω (9) σ Lsi 1 − τYsi 8 HK shows that it is straightforward to write the model with the capital and labor market distortion separately and to obtain the same results. However the output distortion will no longer be identified. 9 σ − 1 Psi Ysi 1 + τKsi M P RKsi αs =R (10) σ Ksi 1 − τYsi Distortions in this framework can be viewed as taxes. Thus, in equations (9) and (10) the after-tax marginal revenue products of capital and labor are equalized across firms. The before-tax marginal revenue products must be higher in firms that face disincentives, and can be lower in firms that benefit from subsidies. Now, HK shows that industry productivity (T F Ps ) is a function of output and factor distortions. HK uses the distinction between revenue productivity and physical productivity initially developed in Foster et al. (2008) and denoted respectively by T F P R and T F P Q.9 The use of a firm-specific deflator yields TFPQ, whereas using an industry deflator gives TFPR: Ysi T F P Qsi Asi = αs 1−αs (11) Ksi Lsi Psi Ysi T F P Rsi Psi Asi = αs 1−αs (12) Ksi Lsi In this framework, TFPR does not vary across firms within an industry unless plants face output and/or factor distortions. In the absence of distortions, more factors of production should be allocated to firms with higher TFPQ to the point where all firms have the same TFPR (this comes from a property of this model that higher output results in a lower price). From the marginal revenue products of labor and capital we obtain: αs 1−αs (1 + τKsi )αs T F P Rsi ∝ (M RP Ksi ) (M RP Lsi ) ∝ (13) 1 − τYsi Thus, we show first that TFPR is a function of output and factor distortions. Second, high firm TFPR is a sign that the firm confronts barriers that raise the firm’s marginal products of capital and labor, rendering the firm smaller than optimal. Having defined TFPR and TFPQ ( A) we can obtain the expression of industry total productivity (T F Ps ):10 1 Ms σ −1 σ −1 T F P Rs T F Ps = Asi (14) i=1 T F P Rsi 9 TFPQ reflects the quality and variety of a plants products, not just its physical productivity (see HK for the demonstration). 10 HK shows that a Lucas span-of-control model (diminishing returns in production rather than utility) and monopolistic competition are isomorphic for aggregate TFP if the number of firms is proportional to labor input. 10 where T F P Rs is the average TFPR of the sector s. Thus, if T F P Rsi were equalized across firms, T F Ps would be distortion-free and equal to: 1 Ms σ −1 −1 As = Aσ si (15) i=1 4 Data Description and Measurement This paper uses firm-level financial and economic data taken from the World Bank’s Enterprise Survey 2007 and 2013 (henceforth WBES2007 and WBES2013) for Morocco. To be sure, a more extensive survey (optimally a census) would have been preferable, the WBES however appears as a second-best option given the restricted access to the public of a recent government census of registered firms. Nevertheless, other papers in the literature have embraced this second-best option and it has been shown that WBES could give very similar results compared to census data (Inklaar, Lashitew, and Timmer, 2015). Manufacturing data from the WBES consists of formal manufacturing firms with at least 5 employees. Since the smallest formal firms and all informal man- ufacturing firms are not part of the data, our measure of misallocation might underestimate overall resource misallocation in manufacturing (this key issue will be examined later in the paper). Sampling for the WBES is conducted using strat- ified sampling procedures to ensure representativeness. Only the most important industries are selected for sampling. Since the relative importance of manufactur- ing sectors changed between the two surveys, the selected industries are not the same in the two surveys as well. In the second stage, the sample size is chosen to ensure a representative sample for the proportion of firms and the average sales in the industry. Further stratification is made on firm size and geographical location to select the firms that are covered by the survey. We regroup the selected manufacturing industries into 7 larger industries for each survey (see Annex). This regrouping represents a compromise between the required detailed manufacturing in the HK framework and limits to data avail- ability (Inklaar, Lashitew, and Timmer, 2015, follows the same approach). Our empirical approach requires firm-level data for production factors and value-added. Then, we only keep in our database firms with complete data on total production, cost of intermediate inputs, capital stock, and labor inputs. Value added is measured as the difference between sales and the cost of intermediate inputs. Cost of intermediate inputs is calculated by adding up three major cost categories: energy consumption (fuel, electricity and other energy costs), cost of raw materials, and overhead and other expenses. The capital stock is measured from the book value of assets, summed across the two categories “machinery, vehicles and equipment” and “land and buildings”. Following HK, we use labor cost rather than employment as a measure of labor inputs. The objective is to account for differences in both hours worked and human capital. In addition, we 11 also remove loss-making firms with negative valued and finally exclude industries with less than 4 observations. The two databases were reduced after the “cleaning” process and this was mainly due to a lack of capital stock data for a large group of firms. For instance, “Machinery” was dropped from both surveys, although this industry only rep- resents about 5 percent of the manufacturing sector’s value added (UNIDO). A critical point is to assess whether the final sample is not too different from the original (representative) sample. Inklaar, Lashitew, and Timmer (2015) proposes a rule of thumb to address this issue: if the remaining database has less than 40 observations or if it represents less than 40 percent of the original sample it should preferably not be used for empirical analysis. WBES2007 respects both criteria. The final database has 312 observations which represents more than 50 percent of the initial sample. Concerning WBES2013, the first criterion is fulfilled (61 ob- servations remained), however the second one is not (the final sample represents around 30 percent of the initial sample). We aknowledge that representativeness could have been importantly affected for the WBES2013 and these results should be taken with caution. For this reason, as a robustness exercise, we apply an alternative methodology to compute resource misallocation (Bartelsman, Halti- wanger, and Scarpetta, 2013) that uses less theoretical restrictions and does not require a capital stock measure, allowing then to use a higher share of the initial sample for the empirical exercise. 5 Computing Distortions and Potential Gain from Reallocation To be able to compute the degree of resource misallocation and the potential gain from resource reallocation a calibration exercise is required. Following HK and as is common in this literature, the rental price of capital (R) is set to 0.10 (the idea is to account for a 5% real interest rate and for a 5% depreciation rate). HK demonstrates that the level of R affects only the level of factor distortion but not the TFP gains from resource reallocation. The elasticity of substitution between firm value-added (σ ) is set to 3. HK shows that the gains from liberalization are increasing in σ so the choice of 3 is a conservative one (in the literature the estimates for σ go from 3 to 10, see for instance Broda and Weinstein, 2006, and Hendel and Nevo, 2006). As HK, α is set to be 1 minus the labor share in the corresponding industry in the United States. Since we cannot separately identify the average capital distortion and the capital production elasticity in each industry we need to pick these shares from a relatively undistorted economy. One more variable is needed to be able to compute the TFP gains associated to liberalization: Asi , defined in equation (11). The problem here is that we do not observe real output for each firm (Ysi ) but rather its nominal output (Psi Ysi ). HK notes however that firms with high real output must have a lower price to explain why buyers would demand the higher output. Then, Ysi can be simply 12 inferred from the observed Psi Ysi by assuming an elasticity of demand of −1/σ . Thus, we raise Psi Ysi to the power σ/(σ − 1) to arrive at Ysi . This shortcut is one of the reasons that has made this approach so popular. Henceforth, even without data on firm output prices, it is possible to compute both revenue and physical productivity for each firm. The final step is to calculate “efficient” output using equation (15) so we can compare it with actual output levels. For each industry, we calculate the ratio of actual TFP (equation 14) to this efficient level of TFP. Finally we aggregate this ratio across sectors using the Cobb-Douglas aggregator from equation (1): θ S Ms σ −1 σ −1 Y Asi T F P Rs = (16) Yef f icient s=1 i=1 As T F P Rsi Before computing the degree of distortions and the TFP gain from resource reallocation we follow HK and remove the top and bottom percentiles of TFPR and TFPQ across industries. At this stage we calculate the industry shares (θs = Psi Ysi /Y ) that will be used for the aggregation. 5.1 Results for Resource Misallocation Levels Table 1 provides a measure of dispersion (standard deviation) for TFPR and compare it with other developed and developing countries. In the HK frame- work dispersion of TFPR represents the level of market distortions. First, results show that dispersion of TFPR in Morocco decreased between 2007 and 2013. Then, this decrease represents an improvement in resource allocation, although as we already noted, this result should be taken with caution since results us- ing WBES2013 could not be fully representative of the Moroccan manufacturing sector in that year. In a subsequent sub-section we will show however that when using an alternative methodology to compute distortions, this result hold. Second, Table 1 shows that dispersion of TFPR in Morocco (whether we use WBES2007 or WBES2007) is not only higher compared to developed countries (US, France, South Korea) but it is also more important than the two largest developing coun- tries (India and China). An additional set of country comparators could have been regional peers. Unfortunately no analogous studies exist in the literature for other countries in the region.11 Table 2 shows the dispersion of TFPR by industry. Results are obtained us- ing the WBES2007 since results from this survey are more robust compared to WBES2013. We observe a large heterogeneity across sectors ranging from 0.60 to 1.13. “Textiles” and “Machinery” are the industries with the lowest TFPR dispersion and “Food” and “Chemicals” are the ones with the highest dispersion levels.12 This difference across industries can be explained first by sector-specific market distortions. Different degree of competition and concentration as well as 11 In the robustness sub-section we show that market distortions in Morocco are also higher than the average of a large group of developing economies. 12 Remark that in WBES2007 for Morocco “Confectionery” is treated as a separate industry and not included as usual in the “Food” industry. 13 Table 1. Dispersion of TFPR Country Morocco US France South Korea China India Data Year 2013 2007 1997 2005 2012 2005 1994 S.D ln(TFPR) 0.77 1.01 0.49 0.48 0.55 0.63 0.67 N ote: S.D. = standard deviation. Industry groups are weighted by their value-added shares. The database year is indicated next to each country name. Values for the U.S., China, and India comes from Hsieh and Klenow (2009), values for France are from Bellone and Mallen-Pisano (2013), and numbers for South Korea comes from Oh (2016). protection vis-`a-vis international competitors could play a sizable role. Indeed, industries more opened to competition (international and domestic) as “Machin- ery” and ”Textiles“ present lower levels of market distortions compared to more protected industries with relatively little competition such as “Food”.13 Second, the composition of firms for each sector can also explain the cross-sector differ- ence in market distortions. Thus, the share of firms that are exported-oriented, small, politically connected, public, and old, could play a key role to explain these results. For instance, the high level of market distortions in “Food” could be explained by its relative high share of small firms (which have on average higher credit constraints) and/or by the industry high percentage of politically connected firms (Saadi, 2016), which could generate negative distortions to the remaining firms. In addition, the high level of resource misallocation in “Chemicals” could be explained by the strong state ownership in the phosphate industry. Table 2. Dispersion of TFPR by Industries Industry Food Confectionery Textiles Chemicals Machinery Other S.D ln(TFPR) 1.06 0.85 0.62 1.13 0.60 0.87 N ote: S.D. = standard deviation. Results obtained using the WBES2007 database. Results for “Machinery” obtained using the WBES2004 database. To further investigate the role of firm characteristics and resource misallocation we run weighted least squares (WLS) regressions of TFPR (relative to industry means) on a range of dummies accounting separately for different firm features. This simple methodology has been used in this context by Hsieh and Klenow (2009) and Busson et al. (2013) and results should not be interpreted in terms of causation. Table 3 below present the estimation results. Regression (1) shows that “old firms” present 21 percent lower TFPR. High entry barriers and low competition would allow these businesses to continue to operate despite low profitability. Thus, more resources should be allocated to “young firms” to increase global TFP. In regression (2) results indicate that small companies have 18.9 percent higher TFPR than medium size firms and TFPR for large firms is not significantly different than the one for medium size firms. 13 In 2013, for example, the Competition Council identified a tacit pricing agreement between market participants in the dairy products sector and reported a high level of concentration. 14 Table 3. WLS Regressions of ln(T F P R)/T F P R) on Different Dummies (1) (2) (3) Old -0.210* (created ≤1985) (0.115) Small 0.189* (5 to 19 workers) (0.105) Large -0.007 (≥100 workers) (0.130) Exported oriented 0.077 (≥ 60% exported) (0.140) N ote: The dependent variable is the deviation of log TFPR from the industry mean. In (1) the independent variable is a dummy for export-oriented firms. In (2) the independent variable is a dummy for old firms. In (3) dummies for small and large firms are the independent variables and the omitted group is medium size firms. Regressions are weighted least squares with industry value-added shares as weights. Entries are the dummy coefficients, with standard errors in parentheses. * indicates significance at the 10 percent confidence level. Consequently, providing extra resources to small enterprises would produce higher returns than providing resources to larger ones. Credit constraints, which typi- cally have more incidence on small firms, can partly explain why small companies cannot acquire more resources to grow. Export-oriented firms exhibit 7.7 percent higher TFPR (regression 3). Keeping in mind the large standard errors associated to the coefficient, this result suggests that more resources should be allowed to export-oriented business. At first sight this seems a counter-intuitive result since it is normally expected that these firms face large positive distortions resulting from better access to credit or preferential treatment in export processing zones. Yet, this “positive” distortions can be more than compensated by other “nega- tive” distortions linked to the overall “anti-export” biais of economic policy (see Section 2 and 6). 5.2 TFP gain from Reallocation Before analyzing results in this sub-section, it is important to highlight that this exercise makes no allowance for measurement error or model misspecification. Such errors could lead us to inflate efficiency gains from resource reallocation. Table 4 provides TFP gains (in %) for Morocco and other developed and de- veloping countries from equalizing TFPR across firms in each industry group. Results are obtained using WBES2013 instead of the richer data in WBES2007 because the computation of TFP gain needs the most recent measures of physical productivity. Full liberalization would boost aggregate manufacturing TFP in Morocco by 83.8 percent. Morocco’s TFP gains from liberalization is well above the one in the US. This is not surprising since, as shown in Table 1, the level of market distortions in Morocco are much higher than in the US. Morocco’s benefits from liberalization are also very much in line with those 15 Table 4. TFP Gains from Equalizing TFPR within Industries Country Morocco US China India Mean of 52 MIC and LIC Data Year 2013 1997 2005 1994 2002-11 TFP gain in % 83.8 (55.6*) 42.9 86.6 127.5 113.4 N ote: Entries are 100 ∗ (Yef f icient /Y − 1). Numbers for Morocco are authors calculations. Numbers for the mean of 52 MIC and LIC are also authors calculations using results from Inklaar, Lashitew, and Timmer (2015). Values for the U.S., China, and India come from Hsieh and Klenow (2009), values for France are from Bellone and Mallen-Pisano (2013), and numbers for South Korea come from Oh (2016). *TFP gain (in %) from removing market distortions to the level of the less three distorted economies studied in this paper (US, France, and South Korea). observed in China, despite differing considerably in economic size and structure. Finally, we observe that Morocco’s potential TFP gain are below the mean of 52 middle and low income countries (MIC and LIC). This does not mean necessar- ily that the level of distortions in Morocco is lower than the average of these 52 economies and than in India. Indeed, India for instance, has less market distor- tions than Morocco and still has higher potential TFP gain. This is due to the key role that the level of physical productivity plays in computing efficiency gains in this framework. A full-fledged liberalization would imply radical policy changes. However, it remains purely theoretical. From a practical point of view, it would be more suitable to compute the efficiency gains obtained if resource misallocation in Mo- roccan manufacturing reaches the level of less distorted economies. To perform this simple counter-factual exercise we choose the three countries in Table 4 with the lowest dispersion in TFPR (US, France, and South Korea). Subsequently we replace the distortion level of Morocco with the average of these three countries and we proceed to compute again the TFP gains associated to this new level of resource misallocation. We obtain a TFP gain in the Moroccan manufacturing sector of 55.6 percent from removing distortions to the level of countries with the best allocation of resources. An ambitious liberalization program has then the potential to have a substan- tial impact in aggregate productivity. A full liberalization would imply that the manufacturing sector would almost double in size after the reform process. Fur- thermore, the strong increase in efficiency in this sector would lead potentially to a transfer of resources from other sectors of the economy (agricultural and ser- vices) to manufacturing. However, this broader effect would be damped by the extent to which distortions in the other sectors would be reduced as well causing an additional resource reallocation across sectors. Finally, in this empirical exercise the effect of the informal sector on market distortions in manufacturing is captured only indirectly. Indeed, lower labor cost in the informal sector would lead to less than optimal labor demand from for- 16 mal firms.14 Firm data accounting for the informal sector would allow to assess resource misallocation in this sector, which represents around 6 percent of the manufacturing sector as a whole (Haut Commissariat au Plan ). Intuitively, in- formal firms face more distortions than in the formal sector. Notably, these firms have incentives to remain small in order to avoid detection. In addition, informal firms face higher credit constraints.15 Busso et al. (2012) modify the HK frame- work to include the informal sector and apply the model to Mexico. They find an astonishing 200 percent increase in TFP from a full liberalization compared to 138 percent found in Inklaar, Lashitew, and Timmer (2015) applying the HK procedure to Mexico without the informal sector.16 5.3 Robustness: Alternative Measure of Market Distor- tions and New TFP Gain Calculation As a robustness exercise we compute market distortions and associated produc- tivity gains using Bartelsman, Haltiwanger, and Scarpetta’s 2013 methodology, which as previously mentioned, relies on a much less extensive theoretical frame- work than HK. The idea here is that low levels of market distortions would better ensure that more resources will be allocated to more productive firms, allowing them to grow in size. The approach uses a decomposition of labor productivity for a given industry initially presented in Olley and Pakes (1996): Y ≡ω= θi ωi = ω + (θi − θ)(ωi − ω ) (17) L i i Where Y and L represent respectively real output and labor for a given sec- tor. The symbol (ω ) denotes the weighted average of firm labor productivity with employment shares used as weights (θi =Li / i Li ). This can be written as the unweighted average labor productivity (indicated by the upper bar) plus the co- variance between firm size (in terms of employment) and firm labor productivity. We observe first that the more this covariance is higher the more the level of pro- ductivity increases. Second, a positive covariance indicates that more productive firms are also larger. (Bartelsman, Haltiwanger, and Scarpetta, 2013) develop a model where greater distortions to resource allocation lead to lower covariances. In contrast to the previous empirical analysis, this methodology does not re- quire data on capital at the firm level. This allows us to be able to use a larger share of each survey (eliminating the previous issue of representativeness). Thus, WBES2007 and WBES2013 contain 325 (about 55% of the initial sample) and 149 (around 70% of the initial sample) observations respectively. 14 The WBES2013 finds that firms consider the informal sector as one of the top obstacles they face. 15 Although this does not enjoy a full consensus in the empirical literature, see for instance Kaplan et al. (2011). 16 More precisely, 138 is a number obtained by the authors using results from Inklaar, Lashitew, and Timmer (2015). 17 We compute the covariance of productivity (in logs) and firm size by industry and we aggregate these covariance terms using employment shares. This aggre- gate covariance value represents the level of resource misallocation. We obtain a covariance term of 0.06 and 0.01 using the WBES2013 and WBES2007 respec- tively. This result implies a decrease in market distortions. Then, we obtain the same result than the one found using the HK methodology. Even though the covariance term increased between the two surveys, the value (0.06) is much lower (implying higher market distortions) than the US, Germany, and the Netherlands (respectively 0.51, 0.28, and 0.30; Bartelsman, Haltiwanger, and Scarpetta, 2013) and slightly lower compared to the mean covariance of 52 MIC and LIC (0.12; Inklaar, Lashitew, and Timmer, 2015). These two findings are in line with the results obtained using the HK approach. With respect to the TFP gain from reallocation, only a back-of-the-envelope calculation is possible. Bartelsman, Haltiwanger, and Scarpetta (2013) finds the aggregated covariance term for the US to be 0.51. The authors show that in an accounting sense, this implies that productivity in the average US manufacturing industry is 51 percent higher than it would be if the size of firms were determined randomly within industries. Analogously, the most recent covariance term value of 0.06 implies that productivity in the average Moroccan manufacturing industry is only 6 percent higher than it would be if the size of firms were randomly determined within industries. Although this is only an approximation, we could argue that if in the Morocco’s manufacturing sector the level of market distortions were reduced to the level found in the US, this sector will increase productivity by 45 percent (51-6). This value (45 percent) is very close to the value found using HK from eliminating distortions up to the level of three relatively undistorted economies (55.6 percent). 6 Policy Discussion Effective market institutions that limit the extent of product and factor market distortions could be gauged on the basis of at least three objectives: improvement in competition and business climate, low structural unemployment and high eco- nomic participation, and strong international economic and financial integration. In all these areas, there is scope for removing market distortions and gradually improving Morocco’s manufacturing TFP (World Bank, 2017a), which will help to reach the goals of the Industrial Acceleration Plan (2014-2020). (i) Improved Competition and Business Climate Morocco could strengthen its market institutions to promote and guarantee market openness as well as free competition. Morocco could act in the following two strategic areas to improve resource allocation: strengthen competition and tackle rent seeking; and improve the business environment. First, steps need to be taken to increase the competition regulatory authorities’ autonomy and clarify their powers, and reduce visible rents (for example, land, 18 approvals, licenses and administrative authorizations) and invisible rents (regu- latory loopholes). Access to urban and industrial property should be improved by a transparent, facilitating regulation. The role of the different public players operating in the tradable and non-tradable sectors needs to be reviewed to make sure they behave in keeping with the principles of the new business framework. The authorities should put an end to discretionary practices and ensure that the laws and regulations apply equally to all businesses. Second, business climate reforms call for close, efficient coordination among the different public and private players concerned. The strategic management and coordination of the programs and activities of the line ministries also pose a complex challenge, in particular in view of the proliferation of sectoral programs and public and private stakeholders. The success of the public initiatives will in fact hinge in large measure on the government’s capacity to coordinate, monitor implementation, and assess public policies at various levels: the government, cen- tral administrations, local administrations, autonomous agencies, private sector operators, and active civil society representatives. (ii) A More Efficient and Inclusive Labor Market Morocco’s poor labor outcomes are partly due to malfunctioning labor market institutions. The Moroccan Labor Code lays down cumbersome, restrictive reg- ulations unsuited to the country’s needs for structural change –especially to the ongoing job reallocation process required for efficiency purposes. Compared with many competitor countries, labor market regulations are particularly restrictive in terms of the use of fixed-term contracts, firing and working hours flexibility. The minimum wage is very high compared with the national average and median income, and collective agreements and seniority pay could end up pushing wage levels higher than staff productivity. Lastly, high social security contributions also help drive up the cost of labor and discourage formal employment, particularly for young people. As a result, only a minority of workers are covered by a social security system. All in all, the labor market rules fuel the vicious cycle of infor- mality that accounts for 9 in 10 private jobs. The active labor market policies and employment agencies cannot provide the expected outcomes. Given the demographic and social challenges, it would be in Morocco’s best interest to expand the current set of programs designed to place employment at the center of public policy by reshaping its labor market institutions on more sound, inclusive foundations. Estimates suggest that overhauling the labor code would significantly raise economic participation and employment, especially for- mal employment for young people and women, and reduce unemployment without jeopardizing wages (Angel-Urdinola, Barry, and Guennouni 2016). Like other re- forms conducted in a number of countries, this reform should be guided by the principles of flexibility of employment, income security for workers and effective active labor market policies. (iii) Increased Integration into the International Economy 19 The poor penetration of Moroccan exports, both goods and services, high- lights the country’s sizeable competitiveness problems. These problems concern not only the high costs of inputs, due to protection of the service sectors and domestic price rigidity, but also product quality and the quality of trade-related infrastructure and logistics. Despite the remarkable achievement of the Tanger Med Port and real progress with customs management, Morocco is handicapped by longer lead times and higher costs for its export and import logistics than its main competitors. These problems are exacerbated by a fixed exchange rate system and capital controls that distort the economy in favor of the nontradable goods sector, but which form a substantial obstacle to Morocco’s price compet- itiveness, product diversification and regional and global integration. Given the current state of market incentives, the lack of diversification into more sophis- ticated export products is also partly due to exporter risk aversion. Moroccan export firms are often old, small and less capable of supplying the global value chains than local subsidiaries of multinationals. It has been clearly established theoretically and empirically that the institu- tions governing a country’s foreign trade serve to capitalize on the benefits of spe- cialization and the international division of labor for greater economic efficiency, faster structural change and higher incomes. Morocco will raise its productivity (among others, through better resource allocation) if its foreign trade sector is able to develop and contribute more to growth. Greater integration of Morocco into the global economy would entail an end to the “anti-export bias” that contin- ues to be endemic to the institutions and policies governing Morocco’s exchange system, including a more flexible exchange rate regime, the liberalization of capi- tal controls, lower tariff and non-tariff barriers to trade, better trade facilitation and an improved investment regime (World Bank, 2006 and 2017a). The prospect of an ambitious Deep and Comprehensive Free Trade Agreement (DCFTA) with the European Union and its embedded potential for modernizing Morocco’s rules and regulations constitutes a strategic objective with a strong transformational potential for the economy. 7 Concluding Remarks We studied in this paper the effect of output and factor distortions in the manufac- turing sector in Morocco. We found that market distortions are higher in Morocco compared to developed countries and slightly more important compared to other developing countries such as China and India. We also provided evidence on the reduction of market distortions between 2007 and 2013 (although these economic distortions remain relatively high). We estimated about 84 percent TFP gains from a full liberalization and 56 percent from removing distortions to the level of selected developed countries with better resource allocation. The paper found that industries more opened to competition present lower levels of market distor- tions compared to more protected industries with relatively little competition. In addition, we showed that a higher level of TFP in the manufacturing sector can be reached if more resources are allocated to “young” and “small” firms. Finally, we 20 found that the main results of the paper are robust to an alternative estimation that is also used in the literature. A natural extension of this work would be to study the effect of resource mis- allocation in the service sector. Ragbi and Nihou (2016) compute the level of distortions in the service sector using the HK methodology. They find lower dis- tortion levels in the service sector compared to the manufacturing sector. This is especially true for the construction sector (a result somehow surprising). A more in depth analysis of the drivers of market distortions in this sector (espe- cially compared to manufacturing) and a computation of TFP gain associated to liberalization would be interesting to investigate. 21 References [1] Angel-Urdinola, Diego; Barry, Abdoul Gadiry and Guennouni, Jamal. 2016. “Are Minimum Wages and Payroll Taxes a Constraint to the Creation of Formal Jobs in Morocco?” World Bank Economic Policy Research Paper No. 7808. Washington D.C.: World Bank Group. [2] Asker, J., Collard-Wexler, A., De Loecker, J. (2014), “Dynamic Inputs and Resource (Mis)Allocation”, Journal of Political Economy, 122(5): 10131063. [3] Balassa, Bela, (1998). “The interaction of factor and product market distor- tions in developing countries.” World Development 16.4 (1988): 449-463. [4] Bartelsman, E., Haltiwanger, J., Scarpetta, S. (2013), “Cross-Country Dif- ferences in Productivity: The Role of Allocation and Selection”, American Economic Review, 103(1): 305334. [5] Bellone, Flora, and Jrmy Mallen, (2013), “Is Misallocation Higher in France than in the United States?” No. 2013-38. Groupe de Recherche en Droit, Economie, Gestion (GREDEG CNRS), University of Nice Sophia Antipolis, 2013. [6] Benazzi, Sara & Imane, Rouiessi, (2017), “Analyse de la concurrence bancaire au Maroc : Approche de Panzar et Rosse,” Document de travail No 1, Bank Al-Maghrib, Research Department, September 2017. [7] Broda, Christian, and David E. Weinstein, (2006) “Globalization and the Gains from Variety,” Quarterly Journal of Economics, 121, 541585. [8] Buera, Francisco J., Joseph P. Kaboski and Yongseok Shin (2011), “Finance and Development: A Tale of Two Sectors”, American Economic Review 101(5): 1964-2002. [9] Busso, M., Fazio, M. V., Levy Algazi, (2012). “(In) Formal and (Un) Pro- ductive: The Productivity Costs of Excessive Informality in Mexico”. IDB Working Papers Series IDB-WP-341. [10] Busso, M., L. Madrigal, and C. Pages, (2013) “Productivity and Resource Misallocation in Latin America”. The BE Journal of Macroeconomics 13.1: 903-932. [11] Calligaris, Sara. (2015) “Misallocation and Total Factor Productivity in Italy: Evidence from FirmLevel Data.” Labour 29.4 (2015): 367-393. [12] CGEM. 2014. “Etude sur les leviers de la comptitivit des entreprises maro- caines ”. Royaume du Maroc. [13] Claessens S, Erik Feijen, Luc Laeven, (2008). “Political Connections and Pref- erential Access to Finance: The Role of Campaign Contributions”. Journal of Financial Economics 88. 22 [14] Coe, D. and E. Helpman, (1995), “International R&D Spillovers,” European Economic Review, 39 , pp. 859-887. [15] Dabla-Norris, Era, Giang Ho, Kalpana Kochhar, Annette Kyobe, and Robert Tchaidze, (2013). “Anchoring Growth: The Importance of Productivity- Enhancing Reforms in Emerging Market and Developing Economies.” IMF Staff Discussion Note, SDN13/08. [16] D’Erasmo, Pablo N., and Hernan J. Moscoso Boedo, (2012). “Financial struc- ture, informality and development”, Journal of Monetary Economics 59.3 (2012): 286-302. [17] Diwan, Ishac and Jamal Ibrahim Haidar. 2016. “Do Political Connections Reduce Job Creation? Evidence from Lebanon”. Economic Research Forum Working Paper No. 1054. [18] Diwan, Ishac, Keefer, P. and Marc Schiffbauer, 2016. “Pyramid Capitalism: Cronyism, Regulation, and Firm Productivity in Egypt,” IDB Publications (Working Papers) 96038, Inter-American Development Bank. [19] EBRD, EIB, and World Bank. 2016. “Whats Holding Back the Private Sector in MENA? Lessons from the Enterprise Survey”. [20] Elbadawi, Ibrahim and Zaki, Chahir. 2016. “Does Exchange Rate Underval- uation Matter for Exports and Trade Margins? Evidence from Firm-Level Data”. Economic Research Forum Working Paper No. 1004. [21] Faccio, Mara, (2006), “Politically connected firms”, American Economic Re- view 96, 369386. [22] Faccio, M., R.W. Masulis, and J.J. McConnell, (2006), “Political Connections and Corporate Bailouts”, Journal of Finance 61, 2597-2635. [23] Faccio, Mara, (2010), “Differences between politically connected and non- connected firms: A crosscountry analysis”, Financial management 39.3 (2010): 905-928. [24] Foster, Lucia, John Haltiwanger, and Chad Syverson, (2008), “Reallocation, Firm Turnover, and Efficiency: Selection on Productivity or Profitability?”, American Economic Review, 98 (2008), 394425. [25] Hendel, Igal, and Aviv Nevo, (2006), “Measuring the Implications of Con- sumer Inventory Behavior,” Econometrica, 74, 16371673. [26] Howitt, Peter, (2000), “Endogenous Growth and Cross-Country Income Dif- ferences,” American Economic Review, 90 (2000), 829846. [27] Hsieh, Chang-Tai and Peter J. Klenow (2009), “Misallocation and Manufac- turing TFP in China and India”, Quarterly Journal of Economics 124(4): 1403-1448. 23 [28] Hsieh, Chang-Tai and Peter J. Klenow (2013), “Misallocation and Manufac- turing TFP in China and India: Correction Appendix”, Stanford University, unpublished. [29] Inklaar, Robert, Addisu A. Lashitew, and Marcel P. Timmer. “The Role of Resource Misallocation in Cross-country Differences in Manufacturing Pro- ductivity.” Macroeconomic Dynamics (2015): 1-24. [30] IRES. (2014), “Industrialisation et comptitivit globale du Maroc”. (p. 15). Royaume du Maroc. [31] Klapper, Leora & Love, Inessa, 2010. “The impact of business environment reforms on new firm registration,” Policy Research Working Paper Series 5493, The World Bank. [32] Kaplan, David S., Eduardo Piedra, and Enrique Seira, (2011). “Entry reg- ulation and business start-ups: Evidence from Mexico.” Journal of Public Economics 95.11 (2011): 1501-1515. [33] Klenow, Peter J., and Andres Rodriguez-Clare, (1997), “The Neoclassical Revival in Growth Economics: Has It Gone Too Far?” in NBER Macroeco- nomics Annual, B. Bernanke and J. Rotemberg, eds. (Cambridge, MA: MIT Press, 1997). [34] Leal, Julio (2014), “The informal sector in contemporary models of the ag- gregate economy”. No. 2014-24. Working Papers, Banco de Mxico, 2014. [35] McMillan, Margaret, Dani Rodrik, and igo Verduzco-Gallo, (2014), “Glob- alization, structural change, and productivity growth, with an update on Africa.” World Development 63.1 (2014): 11-32. [36] Moll, Benjamin (2014), “Productivity Losses from Financial Frictions: Can Self-Financing Undo Capital Misallocation?”, American Economic Review 104(10): 3186-3221. [37] Oh, Jiyoon, (2016). “Misallocation and Manufacturing TFP in Korea.” KDI Journal of Economic Policy (2016): 37-52. [38] Oubenal, Mohammed. 2016. “Crony Interlockers and the Centrality of Banks: The Network of Moroccan Listed Companies”. Economic Research Forum Working Paper No. 1066. [39] Ragbi, Aziz, and Abdelaziz Nihou, (2016) “Allocation inefficiente et produc- tivit manufacturire au Maroc.” in Equilibres externes, comptitivit et pro- cessus de transformation structurelle de lconomie Marocaine, OCP Policy Center. [40] Restuccia, D., Rogerson, R. (2008), “Policy distortions and aggregate produc- tivity with heterogeneous establishments”, Review of Economic Dynamics, 11(4): 707720. 24 [41] Restuccia, D., Rogerson, R. (2017), “The Causes and Costs of Misallocation”, Journal of Economic Perspectives, 31(3): 151174. [42] Rijkers, Bob; Freund, Caroline and Nucifora, Antonio (2014). “All in the Family: State Capture in Tunisia”. Policy Research Working Paper No. 6810. Washington D.C.: World Bank Group. [43] Saadi, Mohammed Sad (2016). “Moroccan cronyism: Facts, Mechanisms and Impact”. Economic research Forum Working Paper No. 1063. [44] World Bank, (2006). “Country Economic Memorandum Morocco 2006”. World Bank Group. Washington, D.C. [45] —————-, (2009). “From Privilege to Competition: Unlocking the Private- led Growth in the Middle East and North Africa”. MENA Development Re- port. Washington D.C.: World Bank Group. [46] —————-, (2013a). “Enterprise Surveys, Morocco Country Profile 2013”. Washington D.C.: World Bank Group. [47] —————-, (2014a). “Doing Business 2015: Going Beyond Efficiency - Mo- rocco. Doing Business 2015”. Washington, D.C.: World Bank Group. [48] —————-, (2015a). “Ease of Doing Business 2016”. Washington D.C.: World Bank Group. [49] —————-, (2015b). “Jobs or privileges : unleashing the employment po- tential of the Middle East and North Africa”. MENA Development Report. Washington D.C.: World Bank Group. [50] —————-, (2016b). “Connecting to Compete 2016: Trade Logistics in the Global Economy. The Logistics Performance Index and Its Indicators”. Wash- ington D.C.: World Bank Group. [51] —————-, (2017a), “Investing in Intangible Capital To Accelerate Eco- nomic Emergence”, Country Economic Memorandum, Washington, D.C.: World Bank Group. [52] —————-, (2017b). “Governance and the Law”. World Development Re- port. Washington, D.C.: World Bank Group. 25 ANNEX WBES2013: Industry Groups and Capital and Labor Shares Industry Groups Capital Share Labor Share Food, beverages and tobacco 0.489 0.511 Textiles, textile products and wearing apparel 0.234 0.766 Wood, paper, printing and publishing – – Chemicals, rubber and plastics 0.381 0.619 Basic and fabricated metal products 0.290 0.710 Machinery – – Miscellaneous manufacturing 0.406 0.594 Source: Bureau of Labor Statistics (average 2002-2010). N ote: To compute the capital and labor shares for each of the industry groups we compute a weighted average of the shares from the 2-digit industries that make up each of these industry groups. “Wood, paper, printing and publishing” and “Machinery” are dropped because of data availability. WBES2007: Industry Groups and Capital and Labor Shares Industry Groups Capital Share Labor Share Food 0.489 0.511 Confectionery 0.489 0.511 Textiles 0.306 0.694 Chemicals 0.281 0.719 Machinery – – Electronics – – Miscellaneous manufacturing 0.406 0.594 Source: Bureau of Labor Statistics (average 2002-2010). N ote: WBES2007 gives directly these aggregated industries. To compute the capital and labor shares for each of the industry groups we compute a weighted average of the shares from the 2-digit industries that make up each of these industry groups. “Machinery” and “Electronics” are dropped because of data availability. 26