This chapter explores the essential characteristics of micro-level data that enable a robust analysis of productivity and support effective evidence-based policy making. Drawing on international best practices, it assesses the strengths and limitations of the microdata currently available in Egypt for this purpose and offers recommendations to enhance their quality. The chapter also highlights the critical importance of ensuring secure access to microdata for research and identifies key areas for improvement within the Egyptian context.
Productivity Review of Egypt
Focusing on the Manufacturing Sector
Abstract
Evidence-based policy making is the process of using high-quality information to inform the design and evaluation of government policies. It often starts with the systematic collection of representative and high‑quality data that are then used for policy analysis relying on rigorous research methods. The evidence generated from this policy analysis can inform how policies and programmes operate, when and where they work effectively, or trends in performance over time. Such evidence should be used in the decision‑making process, from the design and targeting of policies to their monitoring and evaluation.
For evidence-based pro-productivity policy making, the collection, maintenance and linking of firm-level data from different sources is key. These data may include administrative information about firm registrations, company tax records and export/import activities. Data can also be collected through surveys, e.g. on innovation activities (related to product or production process) or the digital activities of firms. As explained in Chapter 5, productivity drivers do not only relate to within-firm inputs such as investment, technology and skills, but also to competition and market conditions more broadly, i.e. the business environment. For this reason, evidence-based productivity policy making often starts with securely connecting and combining representative data from different sources on firms’ inputs and outputs, as well as the business environment, to produce information that informs complex questions of policy.
It is therefore important to have a rigorous evidence-based approach to public policy in Egypt to enhance the productivity of the Egyptian business sector and, in particular, its manufacturing sector.
In Egypt, some data commonly available in OECD countries for building a productivity evidence base – such as tax records or longitudinal production surveys (discussed in the next section) – are not yet accessible to analysts. In Egypt, other important data sources (summarised in Table 6.1), including the Economic Census, are available but cannot be linked over time or across different government providers. This limits the production of productivity statistics that require longitudinal data and restricts analysis of many drivers of productivity growth. These gaps highlight the need to strengthen data capacity to better monitor productivity growth in Egypt.
This chapter first discusses the data requirements to study productivity and the indicators widely used internationally. It then presents the Egyptian data suitable for these purposes, as well as their characteristics and limitations for productivity analysis. The chapter ends by proposing future data enhancements to better monitor productivity trends in Egypt.
This section describes the data and indicators widely used internationally to analyse productivity over time. It then summarises the currently available data in Egypt, discussing its strengths and limitations for productivity analysis, comparing Egypt against these international standards.
Necessary data and main indicators used internationally
Productivity measures the efficiency with which production inputs are used in the production process. It can be measured as single factor (i.e. labour or capital) or multifactor productivity.
Labour productivity is the simplest and most widely used measure of productivity. It captures how efficiently the labour input is used to produce output. It can be calculated using firms’ revenues as output. Yet, this measure does not control for the role of intermediates. As a result, a company with high reselling activity will probably rank very high in this measure. A measure of productivity that considers the role of intermediates is thus labour productivity calculated using value added as output, as value added is the difference between output and intermediates. The best measure of labour input is the number of hours worked, which accounts for differences in hours worked across countries and thus better estimates productivity. However, hours worked are often not available in microdata and can be replaced by employment (i.e. number of persons engaged or employees).
Multifactor productivity (MFP) reflects the efficiency at which production inputs (labour and capital) are used in the production process. It is a refined measure of productivity, as labour productivity does not account for capital input and thus does not measure differences in capital intensities across firms. MFP is measured as a residual, meaning that it represents the part of the growth that cannot be accounted for by labour and capital (OECD, 2023[1]). It thus captures disembodied technical change, meaning that changes in MFP reflect changes in management practices, organisational change, general knowledge, network effects, spillovers but also adjustment costs, economies of scale and measurement errors.
The choice of which indicator to utilise depends on the policy question and the availability of data (OECD, 2023[1]).
Productivity analysis requires detailed micro-level data, which should be representative of the targeted business population and should contain key variables for calculating productivity. Such data should contain, as a minimum, value added and employment at the unit level (either establishment of firm) to compute labour productivity, but also investment and/or capital to measure MFP. Gross output and intermediate inputs (e.g, raw material, energy) are needed in case of gross output measurements. Ideally, data need to be regularly collected with high frequency: this would ensure timely evidence that can support decision making.
Moreover, production units need to be followed over time (i.e. longitudinal data). This longitudinal component of data enables the implementation of a state-of-art methodology to estimate MFP and investigate within-business productivity growth (see Annex B). The main data used internationally for such purposes are either tax records or production survey data. Productivity analysis also requires detailed industry information, to examine productivity by detailed industries or control for unobserved heterogeneity and time trends across industries in econometric analyses. Moreover, the estimate of multifactor productivity at the micro level assumes that all units share the same production technology. This assumption is more likely to hold within industries.
As discussed, productivity is linked to factors internal and external to the firms which influence firm productivity. The analysis of such factors is key to examine micro-determinants of productivity growth. Such information could be exploited in the “primary source” of data, also linking it with other data sources that contain information on digital adoptions, research and development (R&D) expenditure and engagement in trade, for example.
Egyptian data: Strengths and limitations compared to international standards
This sub-section provides an overview on the main data available to study productivity in Egypt, discussing their strengths and limitations for productivity analysis based on the discussion of the previous sub-section.
Table 6.1 summarises the main takeaways for each type of Egyptian data suitable for productivity analysis, which are then discussed below in greater detail. The main data source is the Egyptian Economic Census 2022/2023, used to include Egypt in the OECD Multifactor Productivity (MultiProd) project and study productivity across countries in a comparable way (see Chapter 2 for the main findings using this methodology).
Egypt has rich data sources to study productivity. Building on this strong foundation, Egypt can further enhance its analytical potential by improving the time dimension of these datasets (i.e. providing a panel components) and introducing a unique business identifier. Strengthening these two aspects would significantly expand the ability to analyse firm growth, particularly within-firm productivity dynamics, and support continued improvements in data capacity.
Additionally, the development of a statistical business register (SBR) is needed for enhancing business data collection in Egypt. Such a register could be used to derive a sampling frame for business surveys (both by CAPMAS and other agencies), for improving data linkages across sources and to monitor regular business statistics. The need for an SBR is further confirmed by the fact that some surveys conducted by the Economic Research Forum (ERF) have used the Yellow Pages telephone directory of businesses to identify the business frame for Egypt.1 More details on how to initiate such processes are provided in the following section and in the OECD Business Dynamics Review of Egypt (2026[2]).
Finally, administrative data (e.g. business and tax records, social security data), which in other countries are used to conduct productivity analysis, remain largely unexploited in Egypt.
Note: * GOIEC time series refer to publicly available data accessed by the OECD. Longer time series are available from the GOEIC.
The CAPMAS Egyptian Economic Census
The Egyptian Economic Census is a large representative survey of Egyptian establishments measuring their economic performance and characteristics. It has been conducted by CAPMAS every five years since 1991/92. The CAPMAS website contains information and documentation on the census design and its characteristics (CAPMAS, 2020[3]), which are summarised below.
The census is designed to provide a comprehensive picture of establishments operating in Egypt thanks to detailed micro-level information on business outputs and inputs for establishments operating in different economic activities and operating in the country’s different geographic areas. The census represents an important source for building and adjusting national accounts and for studying the informal sector in Egypt. Regarding the latter, the 2022/23 wave gives information on the administrative registration of establishments, or whether they carry out their activities with permission/licence from the official authorities.2
The sample included in the census covers active establishments in all Egyptian governorates and economic activities (classified according to the International Standard Industrial Classification of All Economic Activities [ISIC] Rev.4).3 However, the data for 2022/23 contain only fixed establishments, defined as operating in a “fixed location where an economic activity is being carried out and is held by a natural or legal person”. The latter means that the census excludes establishments that operate in movable locations, such as most of the activity engaged in Agriculture (ISIC Rev.4, Section A), Construction (F) and Transportation (H) (World Bank, 2020[4]; Assaad et al., 2019[5]). Indeed, those sectors appear to suffer from a lack of representation in the census compared to national accounts in terms of value added and employment (Annex Figure E.1). Employment outside fixed establishments represents a large share of total employment in Egypt, accounting for 46% (Assaad et al., 2019[5]).4
The census covers both private and public establishments. For private establishments, the year of reference is the calendar year (e.g. Jan-Dec 2022), while for public establishments, the reference period is the fiscal year (e.g. July 2022 to June 2023). However, while waves contain information on whether the establishment is part of a business group or is a single branch, there is no detail to link establishments that are part of the same enterprise.
The Egyptian Economic Census 2022/2023 has been developed using the CAPMAS General Census for Population, Housing and Establishments 2017 as a framework.5 The census contains a wide set of variables measuring establishments’ activity and their characteristics. Core variables include: industry at the two-digit ISIC Rev.4. level; governorates; activity starting date; public/private ownership; and legal form. Labour and human capital variables include the number of workers, number of employees and further survey the gender of workers, their occupation and skills as well as detailed financial and performance variables (value added, revenue, production, intermediates, wages and fixed assets). The 2022/23 and 2017/18 wave contain additional key information that is not present in the 2012/13 census, including establishment registration status, the number of employees by education and future needs in terms of employee qualifications.6
The main limitation of the census is the lack of a full common firm identifier between waves, which prevents following establishments over time.7 The 2022/23 wave tracks some establishments also monitored in 2017/18, but they represent only a small share of the full sample.8 This means that the data can only be used in their cross-sectional nature without being able to exploit the potential panel structure of the census, with clear constraints for the depth and breadth of the analysis.9 Finally, the census can be further enriched by combining it with additional information, linked to digitalisation for example (e.g. CAPMAS ICT survey or ERF data) and trade activity (from the GOEIC export/import registers). However, the absence of a common establishment identifier across different data sources and agencies prevents the integration of the census (with census-specific identifiers) with other CAPMAS surveys or other administrative data.10
A small number of existing studies have exploited the census to examine productivity in the Egyptian economy. Most of these papers have only used the 2012/13 wave. Elshennawy and Bouaddi (2021[6]) study the determinants of labour productivity in the manufacturing private sector, distinguishing between low- and high-productive establishments. They provide evidence that capital intensity, larger size, high-quality management, formal status and higher wages explain high labour productivity across establishments. Zaki (2022[7]), instead, exploits the two most recent waves of the census to estimate total factor productivity (TFP) by economic activity, firm size, governorate, gender, formality status and sector. The paper shows that TFP declines between 2013 and 2018 – with the exception of crude petroleum, motor vehicles and some services – in all governorates and, in particular, in Alexandria and Cairo, and that larger establishments experienced a more severe drop in productivity than smaller ones. Formal establishments are found to be generally more productive than informal ones. Badr, Rizk and Zaki (2019[8]) examine the link between establishments’ TFP and agglomeration economies in Egypt. Using the 2012/13 census, they find that small and medium-sized enterprises (SMEs) benefit more from localisation and diversification compared to larger establishments and that activities located outside Cairo are more likely to gain from knowledge spillovers. Martinez-Zarzoso, Said and Zaki (2018[9]) study the link between trade liberalisation and establishment performance in Egypt, combining the 2012/13 census with trade macrodata. They show a positive link between imported inputs and value added, and a negative relationship between tariffs and TFP. Finally, using the 2012/13 survey, Krafft and Assaad (2018[10]) show a positive correlation between wages and productivity in the Egyptian private sector. They find a strong wage-productivity nexus for younger establishments and a weaker one for capital intensive establishments.
GOEIC import and export registers
The export and import registers provide customs data on Egyptian firms’ exports and imports for Harmonized System (HS) six-digit products. The data are collected by GOEIC and are available for dissemination through the ERF data portal.11 The publicly available data on the ERF portal covers only the 2005-16 period.12
The aim of the registries is to provide records of export/import value and volume, traded products and trading partners at the firm level. The firm is identified by its tax registration number in both export and import registers, covering all Egyptian trading firms listed in customs records.
The export (import) register includes information on export (import) value and volume, exported (imported) products and destination (origin) country. Yet, these data lack information on firm characteristics, such as name, location, age, number of workers, ownership structure, etc.: this means that the characteristics of Egyptian exporters/importers cannot be examined without linking these data with additional data sources. At present, this is not possible since the Egyptian Economic Census, for example, has its own specific identifier and the survey is collected at the establishment level (while GOEIC data are at the firm level), limiting the potential of these data sources to support evidence-based policy making. Moreover, the public data only cover the period up to 2016, which coincides with the reform on the liberalisation of the exchange rate in the country, resulting in a significant increase in the EGP-USD exchange rate in 2017 (when its annual average rose 7.69 in 2015 to 17.78 in 2017).13 More recent export/import data may enable the assessment of the impact of recent external and internal developments.
While existing literature utilising GOEIC data provides valuable insights into the effects of non-tariff measures, trade liberalisation and exchange rate fluctuations on Egyptian trading firms, it also highlights certain limitations in the scope and depth of the analysis. This dataset allows for the examination of both the intensive margin (export value) and the extensive margin (number of products, destinations and entry/exit probabilities). However, existing studies often focus on specific barriers without fully addressing the complexity of firms’ decisions across different markets or the dynamic nature of trade relationships. For example, El-Enbaby, Hendy and Zaki (2016[11]) examined the impact of sanitary and phytosanitary measures on Egyptian exporters, finding that they negatively affect the probability of exporting a new product to a new destination but do not impact export value. Kamal and Zaki (2018[12]) investigated the impact of technical barriers to trade (TBTs), showing that they reduce a firm’s probability of exporting to TBT-imposing markets and increase exit probabilities, with a more significant impact on smaller firms. However, their study did not capture how firms might adapt over time to mitigate these barriers. Hendy and Zaki (2021[13]) found that administrative barriers (proxied by the time to trade) have a negative effect on the value of exports, the number of exported products and the number of destination countries, with more pronounced effects for larger firms. Yet, their analysis overlooked potential changes in firm adaptation strategies or their long-term responses to such barriers. Finally, Zaki, Abdallah and Sami (2019[14]) used monthly data to examine the effect of the exchange rate fluctuations on Egyptian exports, showing that a real exchange rate depreciation increases the value of exports without affecting quantity, with the price effect being more significant than the quantity effect. While informative, this study did not consider potential variations in firm-level responses or the broader structural factors influencing trade. These studies collectively demonstrate the use of GOEIC data but also underscore the need for more analysis that incorporate dynamic, firm-level behaviours and account for long-term adjustments to trade barriers.
The IDOS-ERF 2020/21 Egyptian Industrial Firm Behavior Survey
The Egyptian Industrial Firm Behaviour Survey (EIFBS) is a recently conducted survey (between November 2020 and February 2021) that contains detailed data on Egyptian manufacturing establishments. It collects information pre- and post-COVID-19 pandemic, with the aim of understanding the characteristics of manufacturing establishments and examining their resilience to the crisis.14
The survey represents a crucial complement to the CAPMAS IPS to further deepen the understanding of the manufacturing sector in Egypt. Its publicly available feature (accessible on the ERF website upon request) makes the survey an important source of information for research and for supporting evidence‑based policy.15
The survey covers a sample of around 2 300 establishments in manufacturing industries, with five or more employees. Importantly, the sample was drawn from the Egyptian Economic Census 2017/2018, from a sample of 33 331 establishments that represent the more than 110 000 establishments in Egypt.16
The data have been collected using face-to-face surveys, with two questionnaires: one for active establishments and one for establishments exiting the market or temporarily inactive at the time of the interview.17 The survey contains standardised weights to report to the population of establishments.
The EIFBS collects detailed information on establishments’ activity and structure, including more than 800 questions. Among them, the survey collects data on the impact of COVID-19 on the establishment’s operation (such as closure, reduction in working hours, etc.), establishment characteristics (activity start year, single unit/branches, four-digit industry code, governorates), financial variables (revenue before and after the COVID-19 shock), production costs (prior the COVID-19, including wages, intermediates and land), employment (total employment in first year of operation, the year before the pandemic and the month of the interview, in 2020 or 2021, for formal and informal, part-time/full-time, male/female employees). It also contains other information such as ownership (foreign, governmental, gender, education level, including higher education) and management practices, as well as variables on innovation activity (including R&D activity and support) and technologies used (such as computer hardware, websites, accounting software and more advanced technologies as three-dimensional printing and industrial robots). Finally, the survey collects data on the main obstacles to operation and export activities (including non‑tariff measures), global value chain (GVC) participation and training provided to employees.
Even though the sample has been drawn from the Egyptian Economic Census 2017/2018, establishment identifiers are anonymised and cannot be matched with the census. This prevents integrating into the EIFBS detailed information measuring establishments’ performance before the shock, such as value added and fixed assets (that would allow to calculate MFP), but also worker education level (in addition to the manager’s available in the EIFBS). Moreover, the survey only covers industrial SMEs with five or more employees, meaning that informal establishments, generally smaller, are under-represented in the sample. Finally, the surveys only track manufacturing establishments, meaning that it would not be possible to benchmark manufacturing establishments with services, for example, and get a better picture of the resilience of industrial sectors to economic shocks.
Existing analyses have exploited the EIFBS to examine the determinants of manufacturing establishments’ resilience to the COVID-19 shock. For example, El-Haddad and Zaki (2022[15]) show that establishments which were private, smaller, informal and not located in industrial zones were less resilient to the COVID‑19 shock. Moreover, they show that pre-COVID‑19 characteristics such as technology adopted, R&D activity and managerial practices were associated with less likelihood of suffering from the crisis. Additionally, El‑Haddad and Zaki (2023[16]) show that firms engaged in GVCs had greater resilience and better performance during the COVID-19 pandemic. Finally, El-Haddad and Zaki (2023[17]) exploit the EIFBS to explore the allocation and effectiveness of government support to firms in Egypt during the pandemic crisis.
The ERF Digitalisation in the MENA Region Survey
In addition to the county-specific data presented above, the ERF has undertaken two surveys at the firm level in MENA economies (including Egypt). The surveys respectively measure firm digitalisation and resilience to the COVID-19 shock, and both are accessible through the ERF open data portal. While these surveys offer valuable insights into productivity analysis and resilience strategies adopted by Egyptian firms during the pandemic, particularly in comparison with peers across other MENA economies, they are less suited for analysing long-term structural patterns.
The Digitalisation in the MENA Region Survey is a harmonised survey that tracks firm digitalisation in Egypt, Jordan and Morocco. 18 It is a unique source to compare Egypt with peer countries in terms of technology adoption and use, and main bottlenecks to firm digitalisation. The survey has been conducted over two waves of around 1 000 firms per country (for Egypt, a total of 806 firms were successfully interviewed).
The data include general information about firms, such as their economic activity, location, size and ownership structure. It also captures details about the basic technology adopted by firms, for example whether they use the Internet, have their own website or engage in e-selling. In addition, it records the main obstacles to technology adoption, such as power outages or lack of available online services. The dataset also covers economic variables, including the number of workers (disaggregated by gender and education), as well as data on sales, export and imports, fixed assets and labour costs.
The data have been collected between April and August 2022 using computer-assisted telephone interviewing. In the absence of a business register, the Yellow Pages have been used as a frame for Egypt, including data for 300 000 enterprises.19 Target firms were all enterprises that began operating before 2022.20
The main limit of the data is that the survey is cross-sectional and does not contain a panel dimension. Moreover, conclusions drawn from data analysis might not be representative of all Egyptian firms. Additionally, the firm identifier is survey-specific, and data cannot be matched at the firm level with other sources. Finally, since the survey was conducted in 2022, some of the findings on digitalisation might be the results of technologies adopted post-COVID‑19 and so might reflect more the acceleration in adoption during the pandemic rather than more structural trends.
The survey has been used to examine the determinants of digital adoption in Egypt and Jordan. Zaki (2023[18]) shows that larger firms, those with highly-educated workers and women owners, those in service sectors and spending on R&D are more likely to be digitalised. Jordan seems to perform better than Egypt. Several institutional elements may affect this difference, e.g. lower digital infrastructure, higher power outages and higher time to get an electricity connection in Egypt. Zaki (2023[19]) instead provides evidence that, in Egypt, digitalisation is linked to firms with more women, less unpaid workers and more workers with a permanent contract.
The ERF COVID-19 MENA Surveys
The COVID-19 MENA Survey is a harmonised cross-country survey conducted by the ERF in 2021 measuring the impact of COVID-19 on SME activity in Egypt, Jordan, Morocco and Tunisia. The main objective of the survey is to understand firms’ strategies in coping with the crisis, the bottlenecks they were facing and their future expectations, and to compare firms’ responses to the crisis across MENA economies. Data collection began in February 2021 and was carried out in three waves, conducted roughly every two months. Similar to the Digitalisation in MENA Region Survey, the COVID-19 MENA Survey is valuable for analysing firm-level productivity patterns and resilience strategies during the pandemic, as well as benchmarking Egypt’s experience with other MENA economies. However, it is less suited for investigating long-term structural trends.
Egypt is only present in two out of the three waves: the first conducted in February-March 2021 and the second in June-July 2021. Some of the firms have been interviewed in subsequent waves, presenting a panel structure. For Egypt, the sample is composed of 852 enterprises interviewed over 2 waves, 149 of which were interviewed in both waves. The survey was run using computer-assisted telephone interviewing. The sample universe is composed of firms with 6 to 199 workers pre-COVID-19. Also, in this case for Egypt, the sample frame is the Yellow Pages. Four strata were considered: services; food and accommodation; trade, manufacturing and agriculture; and construction.
The survey contains information on firm performance before the COVID-19 pandemic, including firm characteristics (such as industry A10 – with details on manufacturing including textile, food, industry of mechanics or electronics, other – and foreign ownership) and performance (number of workers before the COVID-19 pandemic, revenue/sales in 2019, export/import in 2019). Moreover, the survey also asks questions related to the impact of the crisis on firm activity, firm status (temporary or permanently closed, decrease in salary, layoff of workers, hiring, changes in business model due to the pandemic), teleworking (before and during the pandemic), investment (in equipment, software or digital tools for teleworking) and the use of government support (by type of support, e.g. loan payment deferrals, cash transfer for unemployment; most needed policies to support during the pandemic).
The survey shares the same limitations as the Digitalisation in the MENA Region Survey above, regarding national representativeness and firm identifiers that are survey-specific. Among other papers that exploit the survey, Ebaidalla and Mahjoub (2022[20]) use these data from the three waves in Egypt, Jordan, Morocco and Tunisia to show that firms that were larger in size, with foreign ownership and receiving governmental support were more likely to digitalise as a consequence of the COVID-19 shock. Moreover, manufacturing firms were less likely to be digitalised compared to those in the service industry. Zhu and Luo (2023[21]) also examine the link between e-commerce and firm performance in the MENA region (Egypt, Jordan and Morocco). They show that e-commerce participation is relatively low in the three case study countries compared with regional peers or countries with similar levels of development. Large firms, young firms, firms in the information and communication sector and those with more educated workers are more likely to participate in e-commerce. They show that participation in e-commerce is also positively associated with firm export/import and innovation activities and that those firms performed better than others during the COVID-19 pandemic.
The World Bank Enterprise Surveys
The World Bank Enterprise Surveys (WBES) are firm-level surveys conducted by the World Bank Group that provide insights into the performance and characteristics of the private sector in several economies. These data are particularly useful to conduct cross-country comparisons in business performance and obstacles to business operation.
For Egypt, survey data were collected for 2013, 2016 and 2020, adhering to the global methodology implemented by the World Bank Enterprise Analysis Unit. Two other waves, namely 2007 and 2008, were conducted for Egypt but do not follow the global methodology.21 The questionnaires collected according to the global methodology rely on a standardised format that allows cross-country comparisons.
The WBES are carried out with formally registered private establishments, covering those with five or more employees. Thus, publicly owned, informal and micro firms are not included. The survey is based on a representative sample of the non-agricultural private economy, covering all manufacturing and most services sectors.22 The unit of analysis is the establishment, which is defined as “a business entity associated with a physical location with its own set of financial statements, including a balance sheet and income statement”. However, due to the lack of firm identifiers, establishments cannot be connected to their firms. Moreover, each firm receives a new identifier in every new wave, thus preventing longitudinal analysis.
The WBES collects data on the characteristics and performance of establishments. This includes variables such as age, ownership, sales, number of workers, innovation and management practices. The surveys also gather information on the main obstacles to operations, including issues related to taxes and regulations, corruption, crime, informality, access to finance and lack of adequate infrastructure. The WBES use a stratified random sampling methodology to ensure representative results.23
Seleem and Zaki (2021[22]) exploit the WBES to examine TFP micro-determinants and macro-determinants in MENA economies (including Egypt). They show that government ownership, foreign capital, female managers, foreign certification and the formal registration of establishments are all positively associated with TFP. Longer time contract enforcement times, high tax burdens and high lending rates all tend to have a significantly negative impact on TFP. Other studies use the survey to examine firms’ GVC participation in MENA economies and its association with a productivity premium (Ayadi et al., 2024[23]; Dovis and Zaki, 2020[24]) or to investigate the contribution of female labour participation as well female ownership and management to trade margins in MENA economies (Karam and Zaki, 2020[25]). One study also uses the WBES to explore the nexus between manufacturing firms’ export performance and components of the investment climate in Egypt (Aboushady and Zaki, 2019[26]).
Other data sources
Two other data sources relevant for examining productivity and its drivers in Egypt are included here. However, these data have either not been shared with the OECD or have only been provided for older years, and they are not publicly available. For this reason, they are not described in detail in this section.
First, the ICT survey conducted yearly by CAPMAS up to 2019. The sample is based on the Egyptian Economic Census 2017/2018 and covers around 9 000 enterprises (of which 2 000 are micro). To the OECD’s knowledge, the survey collects information on adoption of basic digital technologies, such as computer use and access to the internet.
Second, the Industrial Production Survey (IPS), which is a representative annual survey of Egyptian industrial establishments. The survey provides industrial statistics that are published yearly in two separate bulletins respectively for the public and private sectors, providing a detailed description of the industrial activity in Egypt, reflecting its structure and changes over time.24 The survey is conducted at the establishment level, covering businesses across all Egyptian governorates. It focuses on establishments engaged in specific industrial economic activities, classified according to ISIC Rev.4. The OECD accessed data only for 2013 and 2014, for both public and private establishments.25 The industrial survey is a valuable source for studying productivity in Egypt, but its current format – as accessed by the OECD – limits proper analysis. It lacks a unique establishment identifier: the form number included in the dataset is not unique and can be repeated, making it impossible to link establishments to firms or to use the survey reliably in cross‑sectional or panel form. This issue could be resolved if CAPMAS provided access to the raw data where correct identifiers should exist. A second major limitation is the absence of information on the economic activity of each establishment, even though this variable is collected in the original survey. CAPMAS should make it available to enable meaningful productivity analysis.26 While several studies have used the IPS and related data sources for productivity analysis in Egypt, these efforts also highlight key limitations in existing data.27
Enhancing Egyptian data capacity to monitor productivity trends is crucial to improve evidence-based policy making. To better monitor productivity growth, several data efforts are needed. These efforts need to be concentrated around three areas:
Creating of a comprehensive data infrastructure for productivity analysis. To achieve this, Egypt should focus on enhancing the quality and use of administrative data, creating an SBR, harmonising the data collection of business data around the SBR and integrating several data sources. This requires major commitment from all actors involved in the data collection and merging of business information in Egypt.
Improving the data quality of the Egyptian Economic Census. This should include unique establishment identifiers across census waves to ensure panel analysis for productivity estimates.
Enhancing microdata sharing for research. This can help improve the supply of quality evidence exploiting technical skills from outside.
These recommendations are related and complementary to recommendations proposed in the OECD Business Dynamics Review of Egypt (2026[2]) to enhance firm-level data in view of monitoring business dynamics and business demography indicators.
Alongside microdata improvements, Egypt could pursue improvements in aggregate data (mainly national accounts) to measure output, labour and capital and calculate productivity at the total economy and sectoral levels. For example, capital stock data, which are key to calculate TFP, are not publicly available, limiting productivity analysis. A discussion on these improvements is outside the scope of this report. Nonetheless, aggregate data enhancements are crucial to better monitor productivity at the aggregate level and ensure Egypt’s alignment with OECD productivity compendium indicators (OECD, 2024[27]).28
Creating a comprehensive data infrastructure can enhance productivity analysis
This section describes the main elements needed to build a comprehensive data infrastructure that merges different sources of data, which can be used to study the Egyptian business sector’s productivity and economic performance.
Improving the quality of administrative data by providing a unique business identifier and consistent classifications can enhance productivity analysis
Administrative data (e.g. business registries, tax records, social security data) are currently not fully exploited for statistical purposes and analysis in Egypt compared to international best practices (OECD, 2026[2]). Indeed, administrative data, such as tax records, are widely used in productivity analysis as they contain financial and other information on businesses, and they usually cover the whole business population of reference. Administrative data are also the primary sources of statistical business registers (SBRs), which are comprehensive lists of all (registered) businesses in a country, used to link data and conduct business surveys. Moreover, several countries are using value added tax and income tax records to improve the timeliness, coverage and quality of national account data (IMF, 2018[28]). Finally, some countries are switching to a register-based approach for the compilation of their censuses (UN, 2024[29]).29 This method consists of using administrative sources to provide a list of registered establishments for the census frame, or even a pre-filled questionnaire to businesses that appear in administrative sources (with information that businesses need to validate and update) and a blanked questionnaire to non-registered units. In the latter case, the register-assisted census reduces the form-filling information burden for registered firms and potentially provides a basis for improving its frequency.
Enhancing the collection and use of administrative sources in Egypt is thus not only key for producing an SBR (which is currently missing in Egypt, see (OECD, 2026[2])), but can also enhance the compilation of national accounts and the collection of the economic censuses.
To ensure a broader use of administrative data, their quality needs to be improved along two dimensions:
First, Egypt should develop and implement a consistent business identification system to enhance data linkages across administrative data sources. While a unified national number at the company level already exists in Egypt (Tax Identification Number, TIN), it is not systematically used by all public institutions. Moreover, a unique establishment number is currently missing. The Commercial Registry has recently finished creating a register of unified establishment identification numbers, but it is only used within the organisation. Such an environment limits data integration and usage. To address this challenge, the Egyptian government has recently decided to define a unique business identifier (UBI), to be used across all administrations. Because the tax number does not cover the universe of businesses, the UBI will be created ex-novo to ensure full applicability among all businesses (as in the case of Morocco presented in Box 6.1). Egypt should develop this national identification system for both companies (legal units) and their branches (local units) and ensure that these numbers are used by all public Egyptian entities. See also OECD Business Dynamics Review of Egypt (2026[2]) for further recommendations on this.
Second, Egypt should ensure that all administrative sources adhere to a consistent classification for industries and governorates. The industry classification, ideally, should align to the most recent international classification (currently ISIC Rev.4) to facilitate cross-country comparisons. For governorates, Egypt could define a unified governate classification to be followed by all public administrations. This harmonisation concerns all public entities collecting data on businesses, including administrative and statistical data. See OECD Business Dynamics Review of Egypt (2026[2]) for further development on this aspect.
Finally, Egypt should facilitate data sharing of administrative sources with CAPMAS for statistical purposes (see next sub-section).
In recent years, Morocco has completed a project setting up a national business identification system. In 2011, a Common Company Identifier (ICE) was established. The decree instituting this identifier was adopted and applied to all enterprises from July 2016 (Feddouli, n.d.[30]). This identification system, which includes a number identifying companies and their branches, is used by all administrations and businesses to facilitate the inter-administrative exchange of microdata on companies, simplifying the administrative procedures applied to enterprises. The identification number is generated in the early stages of a company’s creation. The ICE is composed of 15 digits: 9 for the company, 4 for its establishments and 2 as control digits. The resulting file, which contains companies’ key information, is called the ICE Central Database and is hosted by the tax administration. The constitution of such a system has also facilitated the creation of a single database on businesses, which is used by the Moroccan MSME Observatory to publish annual reports on the number and characteristics of MSMEs in Morocco.
Sources: OMTPME (n.d.[31]), L'Observatoire Marocain de la TPME, https://omtpme.ma/en/; Feddouli, M. (n.d.[30]), “Enhancing the cooperation with administrative data holders”, 30 September-2 October 2019, OECD Meeting of the Group of Experts on Business Registers, UNECE, Geneva, Switzerland.
Egypt can foster administrative data sharing by leveraging the Government-to-Government initiative and by means of law
Several public entities collect information on business operations in Egypt. However, data is not always easily shared across entities, which has relevant implications for policy analysis and evidence-based policy making. To help eliminate such data silos, procedures and regulations should be established and agreed across institutions and departments to share data and information while preserving the confidentiality of individuals and firms.
To improve data sharing between public agencies, the Egyptian state has established what is called the Government-to-Government (G2G) initiative. This initiative has recently been set in Egypt to facilitate inter‑administrative exchange of information on businesses for the purpose of business data collection and public work. The exchange of data between the various parties is organised through the signing of a joint co‑operation protocol that specifies the data that will be exchanged and the rules for data confidentiality and circulation: accordingly, data are exchanged between the two parties (see OECD Business Dynamics Review of Egypt (2026[2])).30
This initiative can be the building block to further develop data sharing among public institutions. The G2G search functionality could be improved gradually. As a first step, G2G can allow users to search firms by governorate or industry rather than by their identifier. As a second step, the G2G can be an avenue to share entire datasets across institutions. See the OECD Business Dynamics Review of Egypt (2026[2]) for further recommendations on how to enhance data sharing across institutions.
Importantly, administrative data (collected by the Egyptian Tax Authority, the Commercial Registry, the National Authority for Social Insurance, the General Authority for Investment and Free Zones (GAFI) and other entities) also need to be shared with CAPMAS for statistical purposes, i.e. to construct an SBR (see below). Several countries are facilitating access to administrative data for statistical purposes by means of law (UN, 2024[29]). Legislation could legitimise CAPMAS to access administrative data for statistical purposes while respecting data confidentiality rules. Developing a clear legal framework and good confidentiality rules for data sharing are key elements in achieving a smooth process and building trust with data providers.
Creating an SBR at CAPMAS could foster data linkages, business surveys and, ultimately, productivity analysis
Creating an SBR is a necessary step to enhance business data collection and more generally improve economic and policy analysis. Additionally, using data from SBRs is quicker and cheaper than conducting surveys, thus minimising the burden on businesses for data collection (OECD/Eurostat, 2008[32]).
A centralised and comprehensive statistical business register is currently unavailable in Egypt. Several entities are responsible for business registration and each has its own database and procedures. Nevertheless, Egypt has a strong foundation of business registration systems across multiple entities. This creates an opportunity to consolidate them into a centralised, comprehensive statistical business register that streamlines procedures and enables more connected insights across involved institutions.
There is no quick fix for establishing an SBR: rather, it is an ongoing process that demands significant efforts from all stakeholders involved in collecting business data in Egypt. It requires substantial investment in ICT infrastructure to ensure the successful implementation and maintenance of the SBR, as well as investments in data privacy and security. Egypt could initiate such a process, with the goal of obtaining full results in the medium term. Further guidelines and detailed recommendations on the development of these registers are elaborated in the OECD Business Dynamics Review of Egypt (2026[2]). The main takeaways are detailed below.
First, Egypt could establish a dedicated law, legitimising CAPMAS to create an SBR and regularly access administrative data from several stakeholders (see above). This has been done in several developed and peer countries. In Tunisia, for example, Decree No. 94-780 of 4 April 1994 entrusted the National Institute of Statistics with the creation of the SBR. The decree also stipulates the role of institute in the update, management, utilisation and dissemination of statistics (UN, 2024[29]) (see also Box 6.2).
Once CAPMAS has accessed several administrative sources, it should create an SBR by merging several data, storing detailed information about every registered unit in the country. A recent reform, backed by Presidential directives, empowers the Ministry of Investment and Foreign Trade and GAFI to lead the construction of an Economic Entities Platform to collects information on business establishment, licensing, operation, expansion and exit procedures under a single integrated system (following the “only-once” principle for data collection).31 Once created, this platform can be the backbone of the SBR.
The SBR should assign each unit a unique identifier (such as a registration number or unique alphanumeric code). The unit identification numbering system from an administrative source does not often correspond to or align with that used in the SBR: a link must therefore be established between the two (UN, 2024[29]).
A specific identifier should be assigned to each level of aggregation of statistical business unit: local unit (establishment identifier), enterprise level (firm identifier), business group (group identifier). As shown in Figure 6.1:
Each establishment should have a unique identifier.
Establishments belonging to the same enterprise should have the same firm identifier.
Firms belonging to the same business group should have the same group identifier.
Having different levels of unit-aggregation will help the design of pro-productivity policies and better support evidence-based policy making. As discussed in Chapter 5, the literature has identified several factors affecting productivity growth: factors within firms (such as human capital, management, intangibles, etc.) and factor outside the firm (including market context, infrastructure, etc.). Data at the establishment level may enable, for example, to observe workers skills and management structure of the units while studying local and regional differences (including the role of education and infrastructure); data at the enterprise level may help explain economies of scale and may allow focusing further on trade and the role of export; finally, business group information allows to further analyse the role of ownership, firms’ position in GVCs and network integration. Without placing units within these three layers of aggregation, productivity analysis may be limited and can miss several key insights into explaining productivity differences across firms.
One possible international reference for the identification of registered business units is the French business register. The register records both enterprises and their establishments. The SIREN number is a nine‑digit code used to identify enterprises. The SIRET code, instead, allows to identify the geographic location of any French establishment and is a 14-digit number composed of the SIREN number plus additional codes (SIREN code plus 5 digits specific to each establishment [NIC] and an internal National Institute of Statistics and Economic Studies classification number).32
The business register should be able to distinguish between true entry and entry associated with mergers and acquisitions, list the reason for a firm’s exit (bankruptcy, insolvency, etc.) and monitor changes in the ownership structure.
Once an SBR is created, data collection should be designed around it. Standardised data collection methods should be developed across all entities involved in the collection of business data. Such an objective requires commitment from all Egyptian entities involved in business data collection (including but not limited to CAPMAS) to ensure that data are harmonised and integrable around the same register and identification units.
Standardisation and harmonisation of data collection ensures the consistency and accuracy of the data collection process. It facilitates the integration of data from various sources, enhancing the overall quality and reliability of the information gathered.
The SBR (Répertoire national des entreprises, RNE) in Tunisia was created in accordance with Decree No. 94-780 of 4 April 1994, which entrusted the National Institute of Statistics with the register’s creation as well as its updating, management, use and dissemination of statistics (UN, 2024[29]). The decree also introduces the obligation to mention the company’s national identifier in all correspondence between public authorities and bodies, and the company.
The institute’s main partners in the management of the RNE are the Ministry of Finance (DGI) and the National Social Security Fund (CNSS). The RNE is indeed constructed from linking tax authority and social security files.
In Tunisia, there is an annual SBR update and administrative data must therefore be obtained in a regular and continuous manner (UN, 2024[29]). Agreements with partner administrations have been drawn up based on the statistical law and the decree establishing the SBR, which grant the National Institute of Statistics access to the various administrative sources containing useful information. Indeed, Article 6 of the decree also obliges public administration (notably the DGI, the CNSS and others) to regularly transfer to the their firm-related data related to the national statistical office. Moreover, the law relating to the National Statistical System (1999) establishes that personal data cannot be used for purposes related to fiscal, economic or social control (Article 5) and that public administrations must submit data to the national statistical office for statistical purposes only (Article 7).
Sources: UN (2024[29]), Guidelines on Statistical Business Registers, Statistics Division, United Nations Department of Economic and Social Affairs; Statistics Tunisia (n.d.[33]), Statistical Business Register, https://www.ins.tn/en/methode/statistical-business-register; Divay, J. et al. (2014[34]), “Evaluation globale adaptée du Système Statistique National de la Tunisie”, https://www.efta.int/sites/default/files/publications/statistics-eso/reports/2014-05-tunisia.pdf; Statistics Tunisia (n.d.[35]), Using Administrative Data for the Maintenance of SBR in Tunisia, https://www.unescwa.org/sites/default/files/event/materials/Session_3_Pres_F_Tunisia_Using_Administrative_Data.pdf; OECD (2026[2]), Business Dynamics Review of Egypt, OECD Publishing, Paris.
Integrating different data sources contributes to enhanced productivity analysis
The SBR should be the backbone for integrating various administrative and survey data to build a comprehensive data infrastructure that can be used to study business dynamics and, more generally, business performance over time. When data collection is harmonised around a common business register it becomes significantly easier to integrate different data sources, thus enriching any analysis. Unique identifiers for establishments, enterprises and business groups help ensure that units can be easily matched across datasets, allowing for a smoother data integration process.
Figure 6.2 provides an example of how such a data integration system could work in Egypt. Such an integrated data system can be built around an SBR merging several Egyptian data sources, including CAPMAS surveys, GOEIC trade register data, as well as social security and Ministry of Manpower data on workers. This infrastructure would merge data at the firm level (e.g. census, ICT survey), with data at the individual firm level (e.g. ownership structure or linked employer-employee data, which would also allow merging with individual data from the labour force survey in a second step) and product firm level (e.g. GOEIC) to obtain a complete picture of the characteristics and activity of firms. In the shorter term, Egypt could exploit the use of techniques like string matching to integrate data even in the absence of a business register.
Box 6.3 highlights the example of the Entrepreneur Information System in Türkiye, an integrated data system merging information from surveys and administrative data developed by the Ministry of Industry and Technology.
Note: The diagram should not be considered as exhaustive of all Egyptian data, but rather an example of data linkages with a business register.
The Ministry of Industry and Technology is responsible for the Entrepreneur Information System (EIS). The EIS is a company-based information system that integrates data from the administrative records of different institutions in Türkiye. The ministry is thus accessing data from several administrative sources including the Ministries of Trade, of Education and of Environment, as well as the Council of Higher Education, National Statistical Institute, Patent and Trademark Office, Public Procurement Authority, Scientific and Technological Research Council, SME Development Organization and Social Security Institution.
The EIS’ primary objective is to aggregate in a central database economic activity data linked to enterprises generating commercial income in the economy. The EIS has a dynamic and flexible structure that allows new datasets from different institutions to be integrated into the system. Currently, it has data on more than 3 million enterprises, mainly from the manufacturing sector from 2006 to 2022. From this integrated system, it is possible to examine employment, exports, industry support, intellectual property, inter-sector and inter-provincial trade, among other things. All data are classified according to sectoral, regional, firm-size, technology levels and spatial classifications, following both international and national classification standards.
Source: OECD (n.d.[36]), “1st OECD/EGY workshop on “Improving Data Capabilities for Effective Economic Analysis and Policymaking”, 8 March 2023.
Improving the data quality of the Egyptian Economic Census can enhance productivity analysis
The Egyptian Economic Census is a precious source to study productivity in the Egyptian economy, as it is representative of the business population (excluding outside fixed-location establishments). However, the data present some limitations for productivity analysis, notably the lack of a panel dimension and unique longitudinal establishment identifier. The paragraphs below detail the main recommendations to improve the quality of the census.
Providing a unique longitudinal establishment identifier across census waves can allow for panel analysis and within-firm productivity growth
Production surveys for productivity analysis must be longitudinal and must contain a panel dimension. This means that units need to be identified by a unique longitudinal identifier that must be constant over time. Although the Egyptian Economic Census contains an establishment identifier, such an identifier is wave-specific and units cannot be matched across the different waves of the survey, preventing analysts from studying productivity growth within establishments.
If a unique establishment identifier is already available in the census, this should be made accessible for research and policy analysis. If, instead, each wave presents a different identifier that cannot be linked over time, string-matching techniques can be exploited to match establishments at different points in time. These techniques use information on firms, such as their name, location, address, etc., to match the same units across different databases. This methodology would allow to assign ex post a unique longitudinal identifier to each unit of the census. Given confidentiality concerns, this matching would have to be conducted by CAPMAS and used to link other databases from administrative sources and surveys to the census and the SBR.
For the collection of future waves of the census, a unique identifier should be provided to each establishment at the beginning of the data collection process. Ideally, this identifier could be matched to the establishment identifier in the business register (see previous section). Also, each establishment should provide information on the enterprise to which they belong (firm identifier) and the business group (group identifier).
Providing a consistent set of core variables can improve cross-country comparison
The economic census contains a rich number of variables on establishment performance. For productivity analysis and comparison across countries, Egypt should be sure to collect core information in each wave. This information should include:
the establishment identifier
the three- or four-digit industry (following the most recent international classification)
the sector (public/private)
the governorate
basic inputs and outputs: value added, revenues, fixed assets, intermediate inputs, wages, employment; labour might be collected as number of workers in headcounts and full-time equivalent
the starting date of operation
foreign ownership (e.g. dummy variable)
an indicator of whether the unit is a single branch, or part of a multi-branch firm or business group.
In addition to this set of core variables, specific additional information might be collected to improve micro‑level analyses, particularly when related to:
export/import indicators
education/skills of workers
investment in digital (including more advanced) technologies
intangible assets (software, database, etc.)
informality (registration at the Commercial Registry, the Egyptian Tax Authority, holding necessary licences)
innovation activity (including R&D expenditure).
Note that some of the information above does not need to be collected within the census if the latter contains an identifier or if string matching is available to link information from the census with administrative sources (e.g. export/import or business registration). This would be important to increase efficiency in data collection and updating, and reduce the burden on firms having to reply to the census survey. Finally, to improve cross-country comparability with OECD countries, data should present a firm-level component. Thus, each establishment should be assigned to the firm they belong to in case of multi-establishment firms to allow aggregate information at the firm level.
Importantly, to examine the evolution of monetary variables over time, national account deflators should be available at the industry level (preferably ISIC two-digit) and should exist for value added, gross output, capital and intermediates. As of now, only value added deflators at the one-digit industry level are available from the website of Ministry of Planning and Economic Development (MPED).
Enhancing microdata sharing with external researchers can improve the supply of evidence to support policy making
Expanding access to various databases allows researchers to conduct more comprehensive and meaningful analyses. This can allow the government to gain insights from external experts who have the technical skills to produce useful evidence. Broader use of data by researchers would also enhance data credibility and quality. Consequently, this can contribute towards the development of better policies within the country.
Even though several major data sources are available upon request to researchers via the ERF Open Access Micro Data Initiative (OAMDI), some data sources still remain inaccessible.33 These include the ICT survey and more recent time series for GOEIC export/import data. Egyptian institutions and CAPMAS should ensure that the data provided through the OAMDI are available up to the most recent time series. Moreover, Egypt can consider allowing access to more data through this ERF portal and to other datasets that are currently unusable for research (e.g. ICT survey, GAFI data), maintaining the confidentiality of the information.
Several OECD countries are facilitating microdata access with standardised procedures that preserve confidentiality of information while easing the data access process (see the OECD Business Dynamics Review of Egypt (2026[2]) for country examples). Several developed and developing countries have indeed established physical rooms within their national statistical offices with dedicated computers where researchers can access microdata in a secure space. To enhance security, these rooms are under video surveillance and computers do not have access to Internet. More recently, countries are also setting up remote access solutions to allow researchers to access the data without the need to travel. Remote access solutions entail, for example, developing virtual private network (VPN) connections – or other end-to-end access mechanisms – wherein researchers can safely access the data remotely on their own computer. Alternatively, countries are also allowing remote execution solutions. In this method, researchers do not directly see the data: they just write the codes, which are then executed by statistical office officials, and researchers only receive the aggregate results. Box 6.4 presents the case of Italy, which is currently offering all three options via the National Institute of Statistics and the Bank of Italy (and their collaboration).
Following international practices, CAPMAS could establish physical locations at its offices and/or remote access solutions via a VPN connection, for example, to allow researchers to access its microdata, adequately anonymised when necessary.
The National Institute of Statistics ADELE Laboratory: A physical location and remote access at the Bank of Italy
The Laboratory for Elementary Data Analysis (ADELE Laboratory) is the data research centre where the National Institute of Statistics (Istat) provides access for scientific purposes to secure use files, microdata files from surveys without direct identifiers or special categories of personal data and/or data related to criminal convictions and offenses. Access to the ADELE Laboratory is granted at Istat headquarters and at all territorial Istat offices. Elementary data files are provided free of charge. Access to the ADELE Laboratory is also free of charge.34 In addition, microdata can be accessed under special research agreements within research projects conducted jointly by Istat and recognised research organisations. Istat completed an experimental project for remote access, leading to the opening of a remote elementary data access laboratory at the Bank of Italy headquarters, via VPN connection.
The Bank of Italy Research Data Centre (RDC): A remote execution system
The Bank of Italy RDC was established in 2019 to provide external researchers access to various granular data, including household data and firm-level data collected by the Bank of Italy. Depending on the characteristics of each database, the data is accessible either as public use files or through the Bank of Italy’s REX remote execution system. The access to firm-level data is provided through a remote execution solution, where researchers do not directly access the data but provide the code to the Bank of Italy and received the subsequent aggregate outputs. The RDC performs formal checks and grants data access. All outputs are consistently checked by RDC staff to avoid statistical disclosure and sending any data error messages to researchers. At the end of the process, researchers finally receive the final output. Data acquired through external institutions cannot be shared, only data directly collected by the Bank of Italy. To help users test their code, a fake dataset replicates the internal structure of the original data.
Sources: Istat (n.d.[37]), The Laboratory for Elementary Data Analysis (ADELE), https://www.istat.it/en/information-and-services-for-users/researchers/adele-laboratory/; Bank of Italy (n.d.[38]), Granular Data – Research Data Centre, https://www.bancaditalia.it/statistiche/basi-dati/rdc/index.html?com.dotmarketing.htmlpage.language=1.
References
[26] Aboushady, N. and C. Zaki (2019), “Investment climate and trade margins in Egypt: Which factors do matter?”, Economic Bulletin, Vol. 39/4, pp. 384-406, http://www.accessecon.com/Pubs/EB/2019/Volume39/EB-19-V39-I4-P213.pdf.
[40] Al-Ayouty, I. (2022), “Identifying activities for greater employment generation in Egypt: An input-output analysis”, African Economic Research Consortium.
[41] Al-Ayouty, I. and H. Hassaballa (2020), “Towards sustainable development: Measuring environmental total factor productivity in Egypt”, European Journal of Sustainable development, Vol. 9/2.
[5] Assaad, R. et al. (2019), “Job creation in Egypt: A sectoral and geographical analysis focusing on private establishments 1996-2017”, Economic Research Forum.
[23] Ayadi, R. et al. (2024), “Trade networks and the productivity of MENA firms in global value chains”, Structural Change and Economic Dynamics, Vol. 69, pp. 10-23.
[8] Badr, K., R. Rizk and C. Zaki (2019), “Firm productivity and agglomeration economies: Evidence from Egyptian data”, Applied Economics, Vol. 51/51, pp. 5528-5544, https://doi.org/10.1080/00036846.2019.1613506.
[38] Bank of Italy (n.d.), Granular Data – Research Data Centre, https://www.bancaditalia.it/statistiche/basi-dati/rdc/index.html?com.dotmarketing.htmlpage.language=1.
[3] CAPMAS (2020), A.R.E - Egyptian Economic Census 2017/2018, Egyptian Economic Census 2017/2018, https://censusinfo.capmas.gov.eg/Metadata-en-v4.2/index.php/catalog/405.
[34] Divay, J. et al. (2014), “Evaluation globale adaptée du Système Statistique National de la Tunisie”, https://www.efta.int/sites/default/files/publications/statistics-eso/reports/2014-05-tunisia.pdf.
[24] Dovis, M. and C. Zaki (2020), “Global value chains and local business environments: Which factors really matter in developing countries?”, Review of Industrial Organization, Vol. 57/2, pp. 481-513, https://doi.org/10.1007/s11151-020-09768-w.
[20] Ebaidalla, M. and E. Mahjoub (2022), “Does COVID-19 pandemic spur digital business transformation in the MENA region?”, International Journal of Social Economics, Vol. 50/3, https://doi.org/10.1108/IJSE-01-2023-0029.
[11] El-Enbaby, H., R. Hendy and C. Zaki (2016), “Do SPS measures matter for margins of trade? Evidence from firm-level data”, Applied Economics, Vol. 48/2, pp. 1949-1964, https://doi.org/10.1080/00036846.2015.1111987.
[17] El-Haddad, A. and C. Zaki (2023), “The role of political connections in COVID policy response: Effectiveness of firm-level government support in Egypt”, Journal of Development Studies, Vol. 59/8, pp. 1213-1235, https://doi.org/10.1080/00220388.2023.2204179.
[16] El-Haddad, A. and C. Zaki (2023), “Vulnerability and resilience: Crisis transmission channels for GVCs in COVID times”, Economics Bulletin, Vol. 43/3, pp. 1581-1592.
[15] El-Haddad, A. and C. Zaki (2022), “Firm dynamics in times of COVID: Evidence from Egyptian firms”, ERF Working Paper, No. 1586, Economic Research Forum, https://erf.org.eg/app/uploads/2022/09/1664179013_730_1270740_1586.pdf.
[6] Elshennawy, A. and M. Bouaddi (2021), “Sources of firm-level heterogeneity in labour productivity in Egypt’s manufacturing sector”, Empirical Economics, Vol. 60, pp. 2589-2612, https://doi.org/10.1007/s00181-020-01847-x.
[30] Feddouli, M. (n.d.), “Enhancing the cooperation with administrative data holders”, 30 September-2 October 2019, OECD Meeting of the Group of Experts on Business Registers, UNECE, Geneva, Switzerland.
[13] Hendy, R. and C. Zaki (2021), “Trade facilitation and firms exports: Evidence from customs data”, International Review of Economics and Finance, Vol. 75, pp. 197-209, https://doi.org/10.1016/j.iref.2021.03.023.
[28] IMF (2018), “Using administrative data to enhance policymaking in developing countries: Tax data and the national accounts”, IMF Working Paper, No. 2018/175, https://www.imf.org/en/Publications/WP/Issues/2018/08/02/Using-Administrative-Data-to-Enhance-Policymaking-in-Developing-Countries-Tax-Data-and-the-46054.
[37] Istat (n.d.), The Laboratory for Elementary Data Analysis (ADELE), Italian National Institute of Statistics, https://www.istat.it/en/information-and-services-for-users/researchers/adele-laboratory/.
[12] Kamal, Y. and C. Zaki (2018), “How do technical barriers to trade affect exports? Evidence from Egyptian firm-level data”, Journal of Economic Integration, Vol. 33/4, pp. 659-721, https://doi.org/10.11130/jei.2018.33.4.659.
[25] Karam, F. and C. Zaki (2020), “On women participation and empowerment in international trade: Impact on trade margins in the MENA region”, Journal of International Trade and Economic Development, Vol. 30/3, pp. 384-406, https://doi.org/10.1080/09638199.2020.1861067.
[10] Krafft, C. and R. Assaad (2018), “Do more productive firms pay workers more? Evidence from Egypt”, ESF Working Paper, No. 1222, Economic Research Forum, https://erf.org.eg/publications/do-more-productive-firms-pay-workers-more-evidence-from-egypt/.
[9] Martinez-Zarzoso, I., M. Said and C. Zaki (2018), “Trade policy and input liberalization: The effect on Egyptian firms’ productivity”, ERF Working Paper, Economic Research Forum, https://erf.org.eg/app/uploads/2018/10/1238.pdf.
[2] OECD (2026), Business Dynamics Review of Egypt, OECD Publishing, Paris, https://doi.org/10.1787/adf24f43-en.
[27] OECD (2024), OECD Compendium of Productivity Indicators 2024, OECD Publishing, Paris, https://doi.org/10.1787/b96cd88a-en.
[1] OECD (2023), OECD Compendium of Productivity Indicators 2023, OECD Publishing, Paris, https://doi.org/10.1787/74623e5b.
[36] OECD (n.d.), “1st OECD/EGY workshop on “Improving Data Capabilities for Effective Economic Analysis and Policymaking””, 8 March 2023, OECD, Paris.
[32] OECD/Eurostat (2008), Eurostat-OECD Manual on Business Demography Statistics, OECD Publishing, Paris, https://doi.org/10.1787/9789264041882-en.
[31] OMTPME (n.d.), Moroccan SME’s Observatory, OMTPME, Casablanca, https://omtpme.ma/en/.
[22] Seleem, N. and C. Zaki (2021), “On modelling the determinants of total factor productivity in the MENA region: A macro-micro firm-level evidence”, International Journal of Productivity and Quality Management, Vol. 32/2, pp. 147-178.
[33] Statistics Tunisia (n.d.), Statistical Business Register, https://www.ins.tn/en/methode/statistical-business-register.
[35] Statistics Tunisia (n.d.), Using Administrative Data for the Maintenance of SBR in Tunisia, https://www.unescwa.org/sites/default/files/event/materials/Session_3_Pres_F_Tunisia_Using_Administrative_Data.pdf.
[29] UN (2024), Guidelines on Statistical Business Registers, Statistics Division, United Nations Department of Economic and Social Affairs.
[4] World Bank (2020), “How do small formal and informal firms in the Arab Republic of Egypt compare?”, Policy Research Working Paper, No. 9423, World Bank, Washington, DC, https://documents1.worldbank.org/curated/en/324411601923707684/pdf/How-Do-Small-Formal-and-Informal-Firms-in-the-Arab-Republic-of-Egypt-Compare.pdf.
[39] World Bank (2014), More Jobs, Better Jobs: A Priority for Egypt, World Bank, Washington, DC, https://documents.worldbank.org/en/publication/documents-reports/documentdetail/926831468247461895.
[19] Zaki, C. (2023), “Does digitalization matter? Evidence from Egyptian and Jordanian firms”, ERF Working Paper, No. 1636, Economic Research Forum, https://erf.org.eg/publications/does-digitalization-matter-evidence-from-egyptian-and-jordanian-firms/.
[18] Zaki, C. (2023), “Which firms are more digitized? A comparative study between Egypt and Jordan”, ERF Working Paper, No. 1635, Economic Research Forum, https://erf.org.eg/app/uploads/2023/04/1682432769_852_760052_1635.pdf.
[7] Zaki, C. (2022), “Estimating total factor productivity in Egypt: Evidence from firm-level data”, ILO Working Paper, International Labour Organization, https://www.ilo.org/publications/estimating-total-factor-productivity-egypt-evidence-firm-level-data.
[14] Zaki, C., A. Abdallah and M. Sami (2019), “How do trade margins respond to exchange rate? The case of Egypt”, Journal of African Trade, Vol. 6/1, https://doi.org/10.2991/jat.k.190528.001.
[21] Zhu, N. and X. Luo (2023), “Digitalization and firm performance in the middle east and north africa: Case studies of Jordan, Morocco, and Egypt”, ERF Working Paper, No. 1637, Economic Research Forum, https://erf.org.eg/app/uploads/2023/05/1683004062_946_2882460_1637.pdf.
← 1. The Economic Research Forum (ERF) is a regional network dedicated to promoting high-quality economic research and contribute to sustainable development in the Arab countries, Iran and Türkiye.
← 2. Recently, the output estimated in Ministry of Planning and Economic Development (MPED) national accounts have been re-estimated based on the Egyptian Economic Census 2017/2018 to better reflect the size of the Egyptian economy and its different sectors.
← 3. The census excludes the public sector, the activity of households as employers and extraterritorial organisation, respectively letters O. Public administration and defence, compulsory social security, T. Activities of households as employers; undifferentiated goods- and services-producing activities of households for own use and U. Activities of extraterritorial organisation of the ISIC Rev.4 classification.
← 4. Additionally, informal establishments might be under-represented in the economic census, as informality is usually more present among workers who work outside establishments.
← 5. The Economic Census 2022/2023 sample design draws on five sample frames from the Population, Housing and Establishments’ Census 2017: i) the enumeration area sample (over 245 000 establishments); ii) the ten or more outside enumeration area sample (full coverage, over 54 000 establishments); iii) industrial zones (full coverage, over 36 000 establishments); iv) new urban areas (full coverage, 111 establishments); and v) other frames (over 58 000 establishments).
← 6. The education status of workers has been added based on the needs of data users.
← 7. For the 2023/24 wave, there are approximately 39 000 establishments included in the economic census whose data have already been completed in the previous census and can be tracked with tax and commercial identifiers. Those firms are mostly larger establishments.
← 8. The 2022/23 wave included about 39 000 establishments included in the Economic Census that were already covered in the previous census and can be tracked using the same identifier.
← 9. Moreover, the 2012/13 version accessed by the OECD has sampling weights that report only to half of the total population of firms in the same year, limiting the possibility of fully comparing the two waves, relying on different methodologies, based for example, on exploiting pseudo-panel analysis.
← 10. Other matching solutions, such as string-matching techniques, cannot be performed with the version of the census data made available to the OECD, as confidential information (establishment names, addresses, etc.) that could allow such matching was not shared.
← 12. This is the timeframe currently available on the ERF website. However, the GOEIC collects timely information and registers should be kept up to date.
← 13. See the International Monetary Fund international financial statistics database.
← 14. The German Institute of Development and Sustainability (IDOS) financed the survey.
← 16. Much like the census, it is not possible to know which establishments belong to the same business group.
← 17. The survey’s response rate was 75%. Stratified multistage sampling was used, considering the number of employees (5-9; 20-100; 101-400; 401 or more), region (urban governorates, lower and upper Egypt) and economic activity (at the two-digit level). First, the sample was allocated proportionally across the three regions. A systematic random sample was drawn to select three governorates from each region using probability proportional to size. The industrial establishments in each region were allocated among governorates proportionally to their size. Next, a systematic random sample was used to select the establishments in each governorate after sorting the establishments according to the number of employees and economic activity at the four-digit level. See https://www.erfdataportal.com/index.php/catalog/252 for the survey’s detailed metadata.
← 18. The survey has been harmonised by the ERF.
← 20. The sample includes both e-firms and non-e-firms, where e-firms are defined as enterprises that used Internet during the month before the interview to buy and/or sell goods or services. The survey includes standard weights.
← 21. Other surveys than the World Bank Enterprise Surveys tackled Egypt, including the Informality Survey in 2008 and the Innovation Survey in 2013.
← 22. For example, the 2016 data include: all manufacturing sectors according to the group classification of ISIC Rev.3.1: Manufacturing (D), Construction (F), Services (G and H), and Transport, storage, and communications (I). Note that this definition excludes the following sectors: financial intermediation (group J), real estate and renting activities (group K, except subsector 72, IT, which was added to the population under study) and all public or utilities-sectors.
← 23. There are three main strata, which are size, business sector and geographic region within a country. In terms of size, establishments are classified into small (5-19 employees), medium (20-99), and large (100 or more). Large establishments are oversampled in the survey since they usually are lower in number compared to SMEs but, at the same time, tend to be the engines of job creation. For Egypt, seven regions were selected: Frontier, Greater Cairo, Middle and East Delta, Northern Upper Egypt, Southern Upper Egypt, the Suez region and West Delta.
← 24. CAPMAS has conducted the IPS on an annual basis since 1957. Since 1989/90, the private sector has been separated from the public sector and each sector’s indicators published in a separate bulletin. Private sector data are based on the calendar year and public sector data on the fiscal year.
← 25. The OECD received less than half of the initial data collected by CAPMAS in both waves. This estimate is based on a comparison with the numbers published in CAPMAS Annual Bulletins on its website. Additionally, no information on sampling weights is available in the data made available to the OECD.
← 26. Other variables such as financial assets, production capacity and environment-related variables were also not found in the data accessed by the OECD. Some of the variables provided are entirely missing (such as fuel and electricity variables).
← 27. The World Bank (2014[39]) used IPS data from 2007 to 2011 to show that larger industrial establishments have higher labour productivity but are more capital intensive and less productive in terms of TFP, pointing to a misallocation of capital towards few larger and older firms. They also show that 1% higher initial TFP level in 2007 led to a 7.4% higher employment growth rate over the next 4-year period. More recently, Al-Ayouty (2020[41]) used annual industrial statistics from CAPMAS to estimate the environmental total factor productivity (ETFP) for Egypt’s energy intensive industries between 2002 and 2014. The study revealed that the ETFP remained largely unchanged. Gains from technical progress were offset by declines in efficiency. It also showed that excluding the environmental components leads to overestimated TFP, underlining the importance of incorporating environmental considerations. However, both studies were constrained by limitations in the available data, including the lack of dynamic firm-level information and limited coverage of informal sector activities. Lastly, Al-Ayouty (2022[40]) used aggregate output data by industry and governorate to identify industries with the potential to generate employment. Results indicated that industries such as food products, basic metals, motor vehicles, paper products, non-metallic mineral products, beverages, wearing apparel, and coke and refined petroleum products had the highest employment multipliers in 2016-17.
← 28. MPED is centred on compiling and producing the relevant national accounts indicators by applying an internationally recognised methodology, approved by the International Monetary Fund (IMF). However, the responsibility of conducting or undertaking primary data collection lies within the responsibility of CAPMAS.
← 29. See Thailand’s experience: https://www.unsiap.or.jp/sites/default/files/pdf/e-learning_el_material_4_eco_4_2_business_1805_sbr_mys06_case_study_from_thailand.pdf.
← 30. The General Authority for Investment and Free Zones (GAFI) has already exchanged data with the Commercial Registry, the Egyptian Tax Authority, the National Authority for Social Insurance and the Real Estate Registry. An electronic mechanism has been created that allows the authority to inquire about commercial registry data, tax data or insurance data, as well as power of attorney data issued by the Real Estate Registry. At the level of GAFI’s company incorporation system, the electronic form via the government digital switch (G2G) is being completed with the Commercial Registry, the Real Estate Registry and the Egyptian Tax Authority, to ensure real-time transmission of the established companies’ data and the speedy completion of the establishment’s procedures.
← 31. The platform is expected to be completed by the end of 2026, with technical co-leadership from the MCIT.
← 32. See https://entreprendre.service-public.fr/vosdroits/F32135 for more details.
← 34. Elementary data files appropriately processed through methods that, by limiting the original information content, make it possible to restrain the risk of breaching the confidentiality of data subjects.