## wpiea2019206-print-pdf

## Source details

**Canonical URL:** [wpiea2019206-print-pdf](https://www.imf.org/-/media/files/publications/wp/2019/wpiea2019206-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2019/wpiea2019206-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2019/wpiea2019206-print-pdf.pdf.json)

---

### I. INTRODUCTION — Background and context
- Arab Spring began December 2010 in Tunisia; movement largely sustained by youth excluded from productive integration.
- 2010 MENA youth unemployment one of the highest worldwide; more than a quarter unemployed versus 20.4 percent on average in other emerging market economies.
- MENA economic model described as “inward-looking and state-centric,” producing protracted high unemployment and protecting well-connected players.
- Education geared toward public sector jobs; private sector characterized by:
  - Informal sector representing about ¼ of economic activity,
  - Concentration of small firms with few scaling to larger, more productive firms.
- Labor force participation (region-wide averages):
  - 2010: men 76 percent, women 26 percent.
  - 2018: men 75 percent, women 28 percent.
- Gallup polls (2005–2009): share “thriving”:
  - MENA countries outside the GCC: less than 14 percent on average.
  - Emerging markets (average): 27 percent.

### Research questions and contribution
- Objectives:
  - Identify main constraints to MENA firms’ employment creation compared to firms in other regions.
  - Analyze degree to which business climate is a constraint across firm types.
  - Identify constraints and policies most likely to influence job creation.
- Four core questions:
  (i) Which types of firms most affected by business climate constraints?  
  (ii) Are business climate constraints systematically different in MENA versus other regions?  
  (iii) How do firm characteristics, macroeconomic and structural variables affect job creation?  
  (iv) What priorities should policymakers set to design structural reforms and macroeconomic policies to support MENA private sectors?

### Data and sample
- Data source: World Bank Enterprise Survey (WBES).
- Sample: surveys conducted in 2013 for five MENA countries (Egypt, Jordan, Lebanon, Morocco and Tunisia) and surveys in 2013 or 2014 for a comparator group of 48 emerging and developing countries.
- Caveat: WBES includes only registered firms with at least five employees; informal firms excluded.

### Characteristics of MENA formal private sectors — Selected MENA (percent unless noted)
- Firm size distribution (Selected MENA):
  - Small (5–19): 58.5 percent
  - Medium (20–99): 33.2 percent
  - Large (100+): 8.3 percent
  - Number of firms in sample (Selected MENA): 4,426
- Country-specific sample shares (percent small / medium / large — number of firms):
  - Egypt: 60.8 / 32.3 / 6.9 — 2,443
  - Jordan: 67.5 / 26.2 / 6.3 — 548
  - Lebanon: 63.0 / 30.5 / 6.5 — 483
  - Morocco: 49.5 / 38.4 / 12.1 — 371
  - Tunisia: 57.7 / 33.7 / 11.6 — 580
- Distribution of permanent workers by firm size (Selected MENA):
  - Small: 11.6
  - Medium: 26.9
  - Large: 61.6
- Distribution of sales by firm size (Selected MENA):
  - Small: 24.5
  - Medium: 32.5
  - Large: 43.1
- Notable contrast: Large firms represented around 62 percent of employment but only about 43 percent of sales in Selected MENA.

### Productivity, employment growth, and firm dynamics — annual changes 2009–2012 (labor productivity growth; employment growth)
- Labor productivity growth (annual change, 2009–2012) — Selected MENA by initial firm size (small / medium / large / aggregate):
  - Selected MENA overall: -4.5 / -5.1 / -3.8 / -3.4
  - Egypt: -5.1 / -5.4 / -4.8 / -4.0
  - Jordan: -4.1 / -6.3 / -2.1 / -1.4
  - Lebanon: -1.3 / -5.0 / 4.5 / 3.1
  - Morocco: -2.3 / -1.8 / -2.9 / -2.5
  - Tunisia: -6.6 / -8.7 / -4.2 / -3.6
  - Comparators:
    - ECA: 3.9 / 3.4 / 4.7 / 7.1
    - SAR: -8.0 / -8.3 / -7.6 / -8.0
    - SSA: -13.0 / -14.4 / -9.4 / -3.1
- Annual employment growth (2009–2012) — Selected MENA by initial firm size (small / medium / large / aggregate):
  - Selected MENA overall: 0.8 / 2.2 / -1.3 / -0.9
  - Egypt: -1.3 / -0.5 / -2.6 / -2.0
  - Jordan: 3.8 / 5.8 / -3.0 / 1.8
  - Lebanon: 1.3 / 5.3 / -5.9 / -0.9
  - Morocco: 6.0 / 7.9 / 4.2 / 1.3
  - Tunisia: 0.4 / 2.9 / -3.6 / -0.2
  - Comparators:
    - ECA: 4.0 / 6.1 / -0.5 / -0.8
    - SAR: 5.4 / 6.8 / 4.1 / 3.7
    - SSA: 6.4 / 7.8 / -1.3 / -5.6
- Key patterns:
  - MENA firms generated fewer jobs than peers after controlling for firm characteristics.
  - Small MENA firms posted positive employment growth (~2 percent) but well below peers (6–8 percent elsewhere).
  - Medium-sized and large MENA firms experienced employment declines; medium-sized firms exhibited a sharp contraction — the “missing middle.”

- Trade, innovation, and age effects:
  - Exporters: labor productivity contraction -1.5 percent vs. -4.4 percent region-wide for non-exporters.
  - Employment growth: exporters 1.7 percent; fully integrated (exporters and importers) 3.2 percent.
  - Firms that did not innovate saw labor productivity decline twice as much as product innovators.
  - Process or product innovators generated up to three times more jobs than other firms.
  - Firm age:
    - Young firms: labor productivity contraction 2.1 percent; employment growth 4.7 percent.
    - Mature firms: labor productivity contraction 5.9 percent; employment growth 1.6 percent.
    - Old firms: labor productivity contraction 3.8 percent; employment growth -1.1 percent.

### Business climate constraints — methodology
- Subjective constraints: firms rated 15 obstacles (0=no obstacle to 4=very severe obstacle); analyzed with probit models for probability a firm reports an obstacle as major/very severe. Controls include firm size, age, ownership, trade integration, location, recent employment changes, sector, technology level, technology absorption, product diversification, respondent’s propensity to complain, and country fixed effects.
- Objective constraints: WBES measures (number of electricity outages, share of sales on security or bribes, share of management time on regulation); analyzed with OLS regressions with similar controls excluding propensity to complain.

### Business climate constraints — empirical findings (MENA firms reporting obstacle as major/very severe)
- Top reported constraints (percent reporting major/very severe):
  - Political instability: 52 percent (MENA) vs. 24 percent (non-MENA).
  - Corruption.
  - Electricity.
  - Informal practices.
  - Access to finance.
- Political instability:
  - MENA firms more likely to report political instability than firms elsewhere; highest concern across firm sizes, especially micro and large firms.
- Corruption:
  - 41 percent of MENA firms reported corruption as a major/very severe obstacle vs. 29 percent of non-MENA firms.
  - Within MENA: micro firms 39 percent, large firms 48 percent reported corruption as a constraint.
  - Large firms experienced higher levels of bribes (as percent of sales) than other firms in MENA and outside MENA.
- Access to finance:
  - 21 percent of MENA firms reported access to finance as a major/very severe obstacle vs. 18 percent elsewhere.
  - Large firms: 20 percent (MENA) vs. 14 percent (comparators) reported access to finance as major/very severe.
  - Firms that expanded employment in previous three years: 28 percent (MENA) vs. 17 percent (non-MENA) complained about access to finance.
  - Collateral: 83 percent of loans on firms’ books required collateral, with average collateral value just over twice the loan amount.
  - MENA firms (except large ones) financed a smaller share of working capital through financial institutions or externally than non-MENA firms; large MENA firms more likely to access external financing than other MENA firms.
- Infrastructure — electricity:
  - Subjective reports: 22 percent of MENA firms vs. 25 percent of non-MENA firms considered electricity a major/very severe obstacle.
  - Objective experience: MENA firms faced a greater number of outages than firms in other regions; micro and small firms faced about seven to eight times more outages on average.
  - Sales losses due to outages: MENA micro and small firms reported about 2 percent more of sales lost due to outages compared to similar non-MENA firms.

### Determinants of MENA firms’ weaker employment performance — key structural and macro factors
- Main conclusion: Worse employment performance of MENA firms relative to peers largely explained by macroeconomic and structural constraints rather than a unique “MENA specificity.”
- Most binding structural constraints identified:
  - Access to finance.
  - Access to electricity.
  - Corruption.
- Macroeconomic factors with largest effects:
  - Growth (GDP per capita growth positively associated with employment growth).
  - External competitiveness (REER appreciation has negative interactions with MENA firms in some specifications).

### Policy implications and priorities
- Address macroeconomic fundamentals and structural business environment weaknesses to improve job creation in MENA.
- Priority areas:
  - Improve access to finance (including credit information and collateral registries) to support firm expansion, especially for employment-growing firms.
  - Strengthen electricity infrastructure to reduce outages and sales losses for micro and small firms.
  - Tackle corruption and governance weaknesses that disproportionately affect firms, particularly large and expanding firms.
  - Promote trade integration and firm innovation to enhance productivity and employment generation; support exporters and firms integrated into global value chains.
  - Target policies to help medium-sized firms overcome constraints (the “missing middle”) to enable scaling and job creation.

### Box 1: Evaluating access to finance — measurement and approach
- WBES lacks data on original financing need, preventing direct assessment of unmet financing demand.
- Literature approaches summarized:
  - Fazzari, Hubbard, Petersen (1988): external financing need proxied by correlation between long-term investment and internal financing.
  - Kaplan and Zingales (1996): questioned investment-cash flow sensitivity as reliable.
  - Demirguc-Kunt and Maksimovic (1998): financial planning model to estimate excess growth if externally funded.
  - Rajan and Zingales (1998): use US industries' external financing access as benchmark.
- In WBES-based studies authors typically:
  - Use subjective constraint data only, or
  - Use share of firm financing from financial institutions as a proxy for “reliance” on financing type rather than direct “access to finance.”
- For objective constraint regression in this study:
  - Use share of investment and working capital financed through external financial institutions as proxies.

### Annex A–D — sample, variables, summary statistics, and regression highlights (selected exact figures)
- Annex A: Data sources
  - WBES used: 2013-14 World Bank Enterprise Surveys reflecting firm realities in years 2009 to 2012.
  - Total countries in sample: 53
    - SSA: 15 countries
    - ECA: 28 countries
    - SAR: 5 countries
    - MENA in sample: 5 countries (Egypt, Jordan, Lebanon, Morocco and Tunisia)
  - Income classification: 14 upper middle income, 19 lower middle income, 9 low income countries.
  - Total firms surveyed: 32,845 firms.
- Annex B: Summary statistics — objective constraint indicators (selected means)
  - Electricity: Obs 32,845; Mean 0.317339; Std. Dev. 0.465448; Min 0; Max 1
  - Telecom: Obs 32,455; Mean 0.087044; Std. Dev. 0.281903; Min 0; Max 1
  - Political instability: Obs 32,844; Mean 0.327244; Std. Dev. 0.469214; Min 0; Max 1
  - Access to finance: Obs 32,769; Mean 0.229424; Std. Dev. 0.420469; Min 0; Max 1
  - Corruption: Obs 32,844; Mean 0.357874; Std. Dev. 0.479382; Min 0; Max 1
- Annex B: Summary statistics — firms’ performance regressions (selected)
  - Annual employment growth: Obs 27,312; Mean 3.4; Std. Dev. 12.5; Min -42.9; Max 63.6
  - Fiscal balance: Obs 26,710; Mean -6.1; Std. Dev. 3.2; Min -9.5; Max 5.9
  - GDP per capita growth: Obs 26,710; Mean 2.0; Std. Dev. 5.8; Min -14.8; Max 17.5
  - Collateral need (WBES): Obs 26,309; Mean 214.9; Std. Dev. 71.0; Min 0.0; Max 600.0
  - Bribery depth (WBES): Obs 26,309; Mean 18.7; Std. Dev. 15.0; Min 0.0; Max 74.4
- Annex C: Objective constraint regressions — selected notable coefficients (coefficient with significance)
  - Fully integrated firms:
    - Number of power outages: -7.415*** 
    - Losses from outages as % of sales: -1.361*** 
    - % of management time spent on dealing with regulation: 11.88***
    - % of working capital financed by financial institutions: -4.173***
    - % of working capital financed externally: -4.357***
  - Large firms (100+ permanent employees):
    - Number of power outages: 4.728*** 
    - % of management time spent on dealing with regulation: -2.481** 
    - % of working capital financed by financial institutions: 6.053*** 
    - % of working capital financed externally: 3.681**
  - Located in MENA region:
    - Number of power outages: 6.623* 
    - Losses from outages as % of sales: 1.688* 
    - Losses due to theft, robbery, vandalism or arson as % of sales: 0.361** 
    - % of total sales spent on gifts: -0.828**
  - Interaction — Large firm in MENA:
    - Number of power outages: -7.545*** 
    - % of total sales spent on gifts: 1.121*** 
    - % of management time spent on dealing with regulation: 7.132*** 
    - % of working capital financed externally: 7.156***
  - Employment change interactions:
    - MENA firm that contracted employment — Losses from outages as % of sales: 2.837***
    - MENA firm that expanded employment — Losses from outages as % of sales: -1.610***; Payments due to security as % of sales: -0.348*; Losses due to theft, robbery, vandalism or arson as % of sales: -0.345**; % of total sales spent on gifts: 0.494*
  - Regression diagnostics (observations and R-squared by dependent variable column):
    - Column (1): Observations 22,073; R-squared 0.097; Adj. R2 0.0955
    - Column (2): Observations 21,132; R-squared 0.077; Adj. R2 0.0758
    - Column (7): Observations 20,909; R-squared 0.105; Adj. R2 0.104
- Annex D: Employment growth regressions — baseline and structural specifications (selected coefficients and statistics)
  - Composition of sample (percent):
    - Age: Young 9% (MENA), 9% (Others); Mature 34% (MENA), 43% (Others); Older 57% (MENA), 48% (Others).
    - Initial size: Small 40% (MENA), 50% (Others); Medium 38% (MENA), 35% (Others); Large 22% (MENA), 15% (Others).
    - Innovation: None 73% (MENA), 50% (Others); Moderate 8% (MENA), 28% (Others).
    - Trade Integration: Exporting and Importing 15% (MENA), 9% (Others).
  - Baseline regression sample size and fit:
    - Observations: 27,312 for columns (1)-(4); 26,710 for columns (5)-(6).
    - R-squared: 0.036 (1), 0.072 (2), 0.043 (3), 0.044 (4), 0.046 (5), 0.053 (6).
  - Selected firm-characteristics (coefficients with significance):
    - Age: Mature = -2.691*** (1); Age: Older = -4.773*** (1).
    - Size: Medium = -2.947*** (1); Size: Large = -3.143*** (1).
    - Innovation: Moderate to strong = 2.423*** (1).
    - Trade integration = 2.225*** (1).
    - Share of temporary workers = -0.021** (1).
  - Macro variables (selected):
    - GDP per capita growth = 0.153*** (column where shown).
    - REER appreciation interaction with MENA = -0.440*** [0.068] (specification noted).
  - Regressions with structural variables (Table D3) — selected structural interactions:
    - Interaction of Access to electricity obstacle with MENA = -10.120** [4.315] (where reported).
    - Collateral need interaction with MENA = -0.024** [0.010] (where included).
    - Access to finance obstacle = -3.172* [1.746] (where shown).

*Source — wpiea2019206-print-pdf.*

### REFERENCES _____________________________________________________________________________________ 24

### REFERENCES _____________________________________________________________________________________ 24

### I.   INTRODUCTION
- In December 2010, the Arab Spring started in Tunisia, when a small business owner set himself on fire to protest the harassment inflicted on him by municipal officials.
- What seemed at first like a single act of desperation spilled over into a wave of protests and resulted in a political movement that spread across the region; the movement was largely sustained by the youth that the region had failed to integrate productively into the economy.
- In 2010, the Middle East and North Africa region (MENA) had one of the highest unemployment rates worldwide, with more than a quarter of the youth unemployed versus

*Source: wpiea2019206-print-pdf - REFERENCES _____________________________________________________________________________________ 24*

### 20.4 percent on average in other emerging market economies.

### wpiea2019206-print-pdf - 20.4 percent on average in other emerging market economies.

### Background and context
- MENA’s economic model was “inward-looking and state-centric,” generating protracted high unemployment and protecting a few well-connected economic players rather than encouraging dynamic small and medium-size firms.  
- Education systems were geared toward public sector jobs rather than private sector employment.  
- The private sector was characterized by:
  - a large informal sector (representing about ¼ of economic activity),
  - concentration of small firms with few growing into larger, more productive firms.  
- Labor force participation rates (region-wide averages):
  - 2010: men 76 percent, women 26 percent.
  - 2018: men 75 percent, women 28 percent (trend has not improved since 2010).  
- Gallup polls (2005–2009): share of people who said they were “thriving”:
  - MENA countries outside the GCC: less than 14 percent on average.
  - Emerging markets (average): 27 percent.

### Research questions and contribution
- The paper seeks to identify main constraints to MENA firms’ employment creation compared to firms in other regions by:
  - analyzing the degree to which business climate is a constraint across firm types,
  - identifying constraints and policies most likely to influence job creation.  
- Four core questions addressed:
  (i) Which types of firms were most affected by business climate constraints?  
  (ii) Were business climate constraints systematically different in MENA compared to other regions?  
  (iii) How do firm characteristics, macroeconomic and structural variables affect job creation?  
  (iv) What priorities should policymakers set to design structural reforms and macroeconomic policies to support MENA private sectors?

### Data and sample
- Data source: World Bank Enterprise Survey (WBES).  
- Sample: surveys conducted in 2013 for five MENA countries (Egypt, Jordan, Lebanon, Morocco and Tunisia) and surveys in 2013 or 2014 for a comparator group of 48 emerging and developing countries.  
- Caveat: WBES includes only registered firms with at least five employees and therefore excludes informal firms.

### Characteristics of MENA formal private sectors
- Firm size distribution (Selected MENA, percent):
  - Small (5–19): 58.5 percent
  - Medium (20–99): 33.2 percent
  - Large (100+): 8.3 percent
  - Number of firms in sample (Selected MENA): 4,426
- Country-specific sample shares (number of firms in sample):
  - Egypt: 60.8 / 32.3 / 6.9 — 2,443
  - Jordan: 67.5 / 26.2 / 6.3 — 548
  - Lebanon: 63.0 / 30.5 / 6.5 — 483
  - Morocco: 49.5 / 38.4 / 12.1 — 371
  - Tunisia: 57.7 / 33.7 / 11.6 — 580
- Distribution of permanent workers by firm size (Selected MENA, percent):
  - Small: 11.6
  - Medium: 26.9
  - Large: 61.6
- Distribution of sales by firm size (Selected MENA, percent):
  - Small: 24.5
  - Medium: 32.5
  - Large: 43.1
- Notable contrasts:
  - Large firms represented around 62 percent of employment but only about 43 percent of sales in Selected MENA.
  - In other regions large firms often account for the majority of sales (ECA, SAR) or less than 20 percent (SSA).

### Productivity, employment growth, and firm dynamics
- Labor productivity growth (annual change, 2009–2012) — Selected MENA and by initial firm size:
  - Selected MENA overall: -4.5 (small), -5.1 (medium), -3.8 (large), -3.4 (aggregate)
  - Egypt: -5.1 / -5.4 / -4.8 / -4.0
  - Jordan: -4.1 / -6.3 / -2.1 / -1.4
  - Lebanon: -1.3 / -5.0 / 4.5 / 3.1
  - Morocco: -2.3 / -1.8 / -2.9 / -2.5
  - Tunisia: -6.6 / -8.7 / -4.2 / -3.6
  - Comparators: ECA 3.9 / 3.4 / 4.7 / 7.1; SAR -8.0 / -8.3 / -7.6 / -8.0; SSA -13.0 / -14.4 / -9.4 / -3.1
- Annual employment growth (2009–2012) — Selected MENA and by initial firm size:
  - Selected MENA overall: 0.8 (small), 2.2 (medium), -1.3 (large), -0.9 (aggregate)
  - Egypt: -1.3 / -0.5 / -2.6 / -2.0
  - Jordan: 3.8 / 5.8 / -3.0 / 1.8
  - Lebanon: 1.3 / 5.3 / -5.9 / -0.9
  - Morocco: 6.0 / 7.9 / 4.2 / 1.3
  - Tunisia: 0.4 / 2.9 / -3.6 / -0.2
  - Comparators: ECA 4.0 / 6.1 / -0.5 / -0.8; SAR 5.4 / 6.8 / 4.1 / 3.7; SSA 6.4 / 7.8 / -1.3 / -5.6
- Key patterns:
  - MENA firms generated fewer jobs than firms in other regions even after controlling for firm characteristics.
  - Small MENA firms posted positive employment growth (~2 percent) but well below peers (6–8 percent elsewhere).
  - Medium-sized and large MENA firms experienced employment declines; medium-sized firms exhibited a particularly sharp contraction, the “missing middle.”
- Trade, innovation, and age effects:
  - Exporters: labor productivity contraction of -1.5 percent vs. -4.4 percent region-wide for non-exporters.
  - Employment growth: exporters 1.7 percent; fully integrated (exporters and importers) 3.2 percent.
  - Firms that did not innovate saw labor productivity decline twice as much as those launching product innovations.
  - Process or product innovators generated up to three times more jobs than other firms.
  - Firm age:
    - Young firms: labor productivity contraction 2.1 percent; employment growth 4.7 percent.
    - Mature firms: labor productivity contraction 5.9 percent; employment growth 1.6 percent.
    - Old firms: labor productivity contraction 3.8 percent; employment growth -1.1 percent.

### Business climate constraints — methodology
- Subjective constraints: firms rated 15 common business environment obstacles (0=no obstacle to 4=very severe obstacle); referred to as “subjective” measures.  
- Objective constraints: WBES measures such as number of electricity outages, share of sales spent on security or bribes, share of management time spent on regulation.  
- Estimation approach:
  - Subjective constraints: probit models for probability a firm reports a constraint as major/very severe; controls include firm size, age, ownership, trade integration, location, recent employment changes, sector, technology level, technology absorption, product diversification, respondent’s average propensity to complain, and country fixed effects.
  - Objective constraints: OLS regressions with similar specifications excluding propensity to complain.

### Business climate constraints — empirical findings
- Overall ranking (MENA firms reporting constraint as major or very severe, top constraints):
  - Political instability: 52 percent (MENA) vs. 24 percent (non-MENA).
  - Corruption.
  - Electricity.
  - Informal practices.
  - Access to finance.
- Political instability:
  - MENA firms were more likely to complain about political instability than firms in other regions; highest concern across firm sizes, especially micro and large firms.
- Corruption:
  - 41 percent of MENA firms reported corruption as a major/very severe obstacle vs. 29 percent of non-MENA firms.
  - Within MENA: micro firms 39 percent, large firms 48 percent reporting corruption as a constraint.
  - Large firms experienced higher levels of bribes (as percent of sales) than other firms in MENA and outside MENA.
- Access to finance:
  - 21 percent of MENA firms reported access to finance as a major/very severe obstacle vs. 18 percent elsewhere.
  - Large firms: 20 percent (MENA) vs. 14 percent (comparator countries) reported access to finance as major/very severe.
  - Firms that expanded employment in previous three years: 28 percent (MENA) vs. 17 percent (non-MENA) complained about access to finance.
  - Collateral: 83 percent of loans on firms’ books required collateral, with average collateral value just over twice the loan amount.
  - MENA firms (except large ones) financed a smaller share of working capital through financial institutions or externally than non-MENA firms; large MENA firms were more likely to access external financing than other MENA firms.
- Infrastructure — electricity:
  - Subjective reports: 22 percent of MENA firms vs. 25 percent of non-MENA firms considered electricity a major/very severe obstacle.
    - The lower subjective reporting comes from large and micro MENA firms complaining less than peers; small and medium MENA firms complained at similar rates to peers.
  - Objective experience: MENA firms objectively faced a greater number of outages than firms in other regions, with micro and small firms facing about seven to eight times more outages on average.
  - Sales losses due to outages: MENA micro and small firms reported about 2 percent more of sales lost due to outages compared to similar non-MENA firms.

### Determinants of MENA firms’ weaker employment performance
- The paper finds that the worse employment performance of MENA firms relative to peers is largely explained by macroeconomic and structural constraints rather than a unique “MENA specificity.”  
- Most binding structural constraints identified:
  - Access to finance.
  - Access to electricity.
  - Corruption.
- Macroeconomic factors with largest effects:
  - Growth.
  - External competitiveness.

### Policy implications and priorities (as signaled in paper)
- Addressing macroeconomic fundamentals and structural business environment weaknesses is critical to improving job creation in MENA.  
- Priority areas for policymakers include:
  - Improving access to finance (including credit information and collateral registries) to support firm expansion, especially for firms that seek to grow employment.  
  - Strengthening electricity infrastructure to reduce outages and sales losses for micro and small firms.  
  - Tackling corruption and governance weaknesses that disproportionately affect firms, particularly large and expanding firms.  
  - Promoting trade integration and firm innovation to enhance productivity and employment generation, including supporting exporters and firms integrated into global value chains.  
  - Targeting policies to help medium-sized firms overcome constraints (the “missing middle”) to enable firm scaling and job creation.

*Italic: Source — wpiea2019206-print-pdf.*

### Box 1: Evaluating access to finance.

### Box 1: Evaluating access to finance.

### Measuring firms' financing constraints
- Reliable identification of whether firms are constrained in the quantities of financing they would like to access requires constructing a proxy of a firm’s financing needs and comparing it with the actual financing obtained.
- WBES surveys lack data on the original financing need, preventing direct assessment of unmet financing demand.
- While WBES provide information on sources of financing used and financing terms, they do not provide data on the original financing need.

### Approaches in the literature to overcome data gaps
- Fazzari, Hubbard, and Petersen (1988): external financing need depends on the magnitude of internal cash flows relative to investment opportunities; firms are considered credit constrained when a high correlation between long-term investment and internal financing is observed.
- Kaplan and Zingales (1996): questioned the above assumption, finding that in 85 percent of all cases firms could have increased their investment regardless of their internal financing position.
- Demirguc-Kunt and Maksimovic (1998): used a financial planning model to estimate firms’ credit constraints by estimating the excess growth rate firms could have achieved if funded beyond internal earnings.
- Rajan and Zingales (1998): used the ability of US industries to obtain external financing as a benchmark to assess how constrained comparable industries are in other countries.

### Use of WBES data and the objective constraint regression
- Authors using WBES data have acknowledged its limitations and typically:
  - Use subjective constraint data only, or
  - Use the share of firm financing received from financial institutions as a measure of “reliance on a particular type of financing” rather than as a direct measure of “access to finance” (Beck and others (2005), Love and Mylenko (2003)).
- For the objective constraint regression in this study, the approach followed is:
  - Use data on the share of investment and working capital financed through external financial institutions as proxies.  

*Source: wpiea2019206-print-pdf - Box 1: Evaluating access to finance.*

### REFERENCES

### wpiea2019206-print-pdf - REFERENCES

### References (bibliography)
- List of cited works covering firm growth, finance, governance, labor markets, and MENA-specific studies, including:
  - Acemoglu, D., and J. Simon, and J. Robinson, 2001, “The Colonial Origins of Comparative Development: An Empirical Investigation,” American Economic Review, Vol. 91, No. 5, pp. 1369-1401.
  - Aterido, R., M. Hallward-Driemeier and C. Pages, 2011. "Big Constraints to Small Firms' Growth? Business Environment and Employment Growth across Firms," Economic Development and Cultural Change, vol. 59(3), pages 609-647.
  - Alvarez de la Campa, Alejandro. 2011. “Increasing Access to Credit through Reforming Secured Transactions in the MENA Region.” Policy Research Working Paper 5613. World Bank.
  - Beck, T., A. Demirguc-Kunt, and V. Maksimovic, 2005, “Financial and Legal Constraints to Growth: Does firm size matter?”, Journal of Finance, Vol. 9, No. 1, pp. 137-177.
  - Blancher, N., M. Appendino, A. Bibolov, A. Fouejieu, J. Li, A. Ndoye, A. Panagiotakopoulou, W. Shi, and T. Sydorenko. “Financial Inclusion of Small and Medium-Sized Enterprises in the Middle East and Central Asia”. IMF Departmental Paper no. 19/02.
  - Claessens, S. and K. Tzioumis, 2006, “Measuring firms’ access to finance”, World Bank.
  - Demirguc-Kunt, A. and V. Maksimovic, 1998, “Law, Finance and Growth”, Journal of Finance, Vol. LIII, No. 6.
  - Dhillon N., and T. Yousef, (eds.) 2009. Generation in Waiting. The unfulfilled Promise of Young People in the Middle East. (Washington: Brookings Institution Press).
  - Fazzari, Steven M., R. Glenn Hubbard, and Bruce C. Petersen. "Financing Constraints and Corporate Investment." Brookings Papers on Economic Activity, Vol. 1, pp. 141-195, (1988).
  - Freund, C., B. Rijkers, 2012, “Employment Miracles.” Paper presented at the 13th Jacques Polak Annual Research Conference at the International Monetary Fund, November 8-9, 2012.
  - Freund, C., B. Rijkers, and A. Nucifora, 2014, “All in the Family: State Capture in Tunisia,” World Bank Policy Research Working Paper 6810.
  - Freund, C., B. Rijkers, H. Arouri, and A. Nucifora, 2013, “Structural Stagnation: Firm-level Evidence on Job Creation in Tunisia,” World Bank and Institut National de La Statistique Tunisie.
  - Gallup, 2010, “Gallup Global Wellbeing: The Behavior Economics of GDP Growth” (Washington, DC).
  - Hall, R., and C. Jones, 1999, “Why Do Some Countries Produce So Much More Output per Worker than Others?,” Quarterly Journal of Economics, Vol. 114, No.1, pp. 83–116.
  - Hallward-Driemeier, Mary and Reyes Aterido, 2009 “Comparing Apples with... Apples. How to Make (More) Sense of Subjective Rankings of Constraints to Business”, Policy Research Working Paper no. 5054, World Bank.
  - Harrison, A., J. Y. Lin and L. C. Xu, 2014, “Explaining Africa’s (Dis)Advantage” World Development, vol. 63(C), pages 59-77.
  - Helmy, O., 2009, “The Need for Insolvency Systems Reform in the MENA Region”, Survey on Insolvency Systems in the Middle East and North Africa.
  - Hoffman, M. and A. Jamal. “The Youth and the Arab Spring: Cohort Differences and Similarities.” Middle East Law and Governance, 2012.
  - Kaplan, S. and L. Zingales, 1997, “Do investment-cash flow sensitivities provide useful measures of financing constraints?”, Quarterly Journal of Economics, Vol. 112, Issue 1, pp. 169-215.
  - Kaufmann, D., and A. Kraay, and P. Zoido-Lobaton, 1999, “Governance matters,” World Bank Policy Research Working Paper 2196.
  - Love, Inessa and Nataliya Mylenko, 2003, “Credit reporting and financing constraints”, World Bank Policy Research Working Paper 3142.
  - Mauro, P., 1995, “Corruption and Growth,” The Quarterly Journal of Economics, Vol. 110, No. 3, pp. 681–712.
  - Pierre, Gaëlle and Stefano Scarpetta. 2004 “Employment Regulations through the eyes of employers. Do they matter and how do firms respond to them?” IZA Discussion Paper No. 1424.
  - Pierre, Gaëlle and Stefano Scarpetta. 2006 “Employment Protection: Do Firms’ Perceptions Match with Legislation?”, Economic Letters, 90, pp. 328-334.
  - Purfield Catriona et al., 2018, “Opportunity for All, Promoting Growth and Inclusiveness in the Middle East and North Africa”, IMF Departmental Paper no. 18/11.
  - Rajan, R. and L. Zingales, 1998, “Financial Dependence and Growth”, American Economic Review, Vol. 88, No. 3, pp. 559-586.
  - Rijkers, B., and H. Arouri, 2012, “Job Creation: A Big Role for Big Firms.” Blog, World Bank.
  - Rocha, R., F. Subika, R. Khouri, and D. Pearce, 2011, “The status of bank lending to SMEs in the Middle East and North Africa region: Results of a joint survey of the Union of Arab Bank and the World Bank,” World Bank Policy Research Working Paper 5607.
  - Rother, Bjorn et al., 2016, “The Economic Impact of Conflicts and the Refugee Crisis in the Middle East and North Africa”, IMF Staff Discussion Note no. 16/8.
  - Scarpetta, Stefano and Thierry Tressel, 2004, “Boosting Productivity via Innovation and Adoption of New Technologies: Any Role for Labor Market Institutions?”, Policy Research Working Paper 3273.
  - Schneider, Friedrich, 2012, “The Shadow Economy and Work in the Shadow: What Do We (Not) Know?”, IZA Discussion Paper No. 6423.
  - Svensson. J.,1998, “Investment, property rights, and political instability: theory and evidence,” European Economic Review, Vol.42, pp. 1317–1341.
  - World Bank, 2009, “From Privilege to Competition: Unlocking Private-Led Growth in the Middle East and North Africa”.
  - World Bank, 2013, “Jobs for Shared Prosperity: Time for Action in the Middle East and North Africa”.
  - World Bank, 2015, “Jobs or Privileges – Unleashing the Employment Potential of the Middle East and North Africa”.

### Annex A: Data sources (sample and datasets)
- WBES used: 2013-14 World Bank Enterprise Surveys reflecting firm realities in years 2009 to 2012.
- Sample composition:
  - Total countries in sample: 53
  - Sub-Saharan Africa (SSA): 15 countries
  - Eastern Europe and Central Asia (ECA): 28 countries
  - South Asia (SAR): 5 countries
  - Middle East and North Africa (MENA) in sample: 5 countries (Egypt, Jordan, Lebanon, Morocco and Tunisia)
  - Income classification: 14 upper middle income, 19 lower middle income, 9 low income countries.
- Total firms surveyed: 32,845 firms.
- Coverage note: Surveys include registered firms with at least five employees; informal firms excluded.
- Complementary datasets:
  - World Bank Doing Business (structural indicators)
  - IMF databases (macroeconomic data)

### Table A1: Variables (definitions and sources) — selected entries
- Annual employment growth — Source: WBES. Definition: Annualized growth of permanent full-time workers expressed as a percentage.
- Firm’s age group — Source: WBES. Definitions: “young” up to 5 years; “mature” 6–15 years; “older” 16 years.
- Initial firm’s size — Source: WBES. Definitions: “small” < 20 employees; “medium” 20–99 employees; “large” ≥ 100 employees. Size measured at beginning of period (2009).
- Firm’s sector — Source: WBES. Classified as Non-manufacturing, low-tech manufacturing, high-tech manufacturing (following Scarpetta and Tressel (2004)).
- Collateral need — Source: WBES. Value of collateral needed as a percentage of loan value; regional average used to avoid endogeneity.
- Credit line — Source: WBES. Takes value -1 if firm has a credit line; -2 otherwise; regional averages used to avoid endogeneity.
- Bribery depth — Source: WBES. The depth of Bribery (% of public transactions where a gift or informal payment was requested); regional averages used to avoid endogeneity.
- Employment regulation (EPL) — Authors’ calculations using World Bank data on labor market regulation for 2009 (index creation detailed in Pierre and Scarpetta (2004)).
- Fiscal balance — Source: WEO. Fiscal balance = revenue minus total expenditure, as a share of nominal GDP in 2009.
- GDP per capita growth — Source: WEO. Growth of GDP per person in constant national currency in 2009.
- REER appreciation — Source: INS. Percent change of the REER in 2009.

### Annex B: Summary statistics — objective constraint indicators (Table B1)
- Sample size for constraint indicators: 32,845 observations unless noted.
- Selected means (Obs, Mean, Std. Dev., Min, Max):
  - Electricity: Obs 32,845; Mean 0.317339; Std. Dev. 0.465448; Min 0; Max 1
  - Telecom: Obs 32,455; Mean 0.087044; Std. Dev. 0.281903; Min 0; Max 1
  - Customs regulation: Obs 32,842; Mean 0.121034; Std. Dev. 0.326172; Min 0; Max 1
  - Political instability: Obs 32,844; Mean 0.327244; Std. Dev. 0.469214; Min 0; Max 1
  - Access to finance: Obs 32,769; Mean 0.229424; Std. Dev. 0.420469; Min 0; Max 1
  - Corruption: Obs 32,844; Mean 0.357874; Std. Dev. 0.479382; Min 0; Max 1
  - Courts: Obs 32,528; Mean 0.084389; Std. Dev. 0.277974; Min 0; Max 1

### Annex B: Summary statistics — firms’ performance regressions (Table B2)
- Observations for performance regressions vary by variable; selected entries:
  - Annual employment growth: Obs 27,312; Mean 3.4; Std. Dev. 12.5; Min -42.9; Max 63.6
  - Share of temporary workers: Obs 27,312; Mean 4.6; Std. Dev. 12.0; Min 0.0; Max 97.9
  - Fiscal balance: Obs 26,710; Mean -6.1; Std. Dev. 3.2; Min -9.5; Max 5.9
  - GDP per capita growth: Obs 26,710; Mean 2.0; Std. Dev. 5.8; Min -14.8; Max 17.5
  - REER appreciation: Obs 26,710; Mean 1.4; Std. Dev. 6.9; Min -16.3; Max 14.9
  - Collateral need (WBES): Obs 26,309; Mean 214.9; Std. Dev. 71.0; Min 0.0; Max 600.0
  - Credit line (WBES): Obs 26,309; Mean -1.7; Std. Dev. 0.2; Min -2.0; Max -1.3
  - Bribery depth (WBES): Obs 26,309; Mean 18.7; Std. Dev. 15.0; Min 0.0; Max 74.4
  - Employment regulation (EPL): Obs 26,309; Mean 0.4; Std. Dev. 0.1; Min 0.1; Max 0.6
  - Ease of doing business (WB): Obs 26,309; Mean 51.2; Std. Dev. 8.4; Min 29.6; Max 75.2
  - Corruption obstacle (survey): Obs 26,710; Mean 0.4; Std. Dev. 0.2; Min 0; Max 0.9
  - Access to finance obstacle (survey): Obs 26,710; Mean 0.2; Std. Dev. 0.1; Min 0; Max 1
  - Access to electricity obstacle (survey): Obs 26,710; Mean 0.3; Std. Dev. 0.2; Min 0; Max 0.9

### Annex C: Factors associated with objective business climate constraints — key regression coefficient highlights (selected coefficients and significance)
- Regression setup: Columns correspond to dependent variables — (1) Number of power outages; (2) Losses from outages as % of sales; (3) Payments due to security as % of sales; (4) Losses due to theft, robbery, vandalism or arson as % of sales; (5) % of total sales spent on gifts; (6) % of management time spent on dealing with regulation; (7) % of working capital financed by financial institutions; (8) % of working capital financed externally.
- Notable coefficients (coefficient value with significance as reported):
  - Fully integrated firms:
    - Column (1) Number of power outages: -7.415*** 
    - Column (2) Losses from outages as % of sales: -1.361*** 
    - Column (6) % of management time spent on dealing with regulation: 11.88***
    - Column (7) % of working capital financed by financial institutions: -4.173***
    - Column (8) % of working capital financed externally: -4.357***
  - Importer only:
    - Column (2): 1.137** 
    - Column (3): 0.204* 
    - Column (4): 0.198*** 
    - Column (5): 0.360** 
    - Column (6): 5.550*** 
    - Column (7): -4.706*** 
    - Column (8): -5.105***
  - Private domestic ownership:
    - Column (1): -4.165*** 
    - Column (3): -0.353** 
    - Column (7): -7.365***
  - State-owned:
    - Column (1): -5.374*** 
    - Column (2): -1.376** 
    - Column (4): -0.263*** 
    - Column (6): 9.081** 
    - Column (7): -6.532*** 
    - Column (8): 6.147
  - Sectoral and firm-size effects (selected):
    - Low tech manufacturing:
      - Column (1): 4.873*** 
      - Column (2): 0.989*** 
      - Column (3): -0.292* 
      - Column (7): 4.260*** 
      - Column (8): 2.453*
    - Hotels and restaurants:
      - Column (1): -4.142*** 
      - Column (2): -1.375*** 
      - Column (4): 0.536** 
      - Column (6): 7.227*** 
      - Column (7): -10.62*** 
      - Column (8): -10.83**
    - Large firms (100+ permanent employees):
      - Column (1): 4.728*** 
      - Column (5): -0.245* 
      - Column (6): -2.481** 
      - Column (7): 6.053*** 
      - Column (8): 3.681**
  - Foreign ownership (partially or fully-owned by foreigners):
    - Column (1): -1.355*** 
    - Column (3): 0.220* 
    - Column (5): 0.628* 
    - Column (8): -3.902**
  - Innovation (process and product) — selected coefficients:
    - Degree of process innovation (Moderate):
      - Column (1): 4.932** 
      - Column (2): 0.824*** 
      - Column (7): 12.32*** 
      - Column (8): 12.77***
    - Degree of process innovation (Strong):
      - Column (1): -3.871*** 
      - Column (2): -0.645** 
      - Column (4): 0.490** 
      - Column (5): -0.455** 
      - Column (6): 5.384*** 
      - Column (8): 7.847***
  - MENA region effects:
    - Located in MENA region:
      - Column (1): 6.623* 
      - Column (2): 1.688* 
      - Column (4): 0.361** 
      - Column (5): -0.828**
    - Interaction — Large firm in MENA:
      - Column (1): -7.545*** 
      - Column (5): 1.121*** 
      - Column (6): 7.132*** 
      - Column (8): 7.156***
  - Employment change interactions:
    - MENA firm that contracted employment:
      - Column (2): 2.837*** 
    - MENA firm that expanded employment:
      - Column (2): -1.610*** 
      - Column (3): -0.348* 
      - Column (4): -0.345** 
      - Column (5): 0.494*
- Regression diagnostics (panel of columns):
  - Observations by column:
    - Column (1): Observations 22,073; R-squared 0.097; Adj. R2 0.0955
    - Column (2): Observations 21,132; R-squared 0.077; Adj. R2 0.0758
    - Column (3): Observations 21,472; R-squared 0.008; Adj. R2 0.00608
    - Column (4): Observations 21,908; R-squared 0.008; Adj. R2 0.00672
    - Column (5): Observations 18,658; R-squared 0.035; Adj. R2 0.0332
    - Column (6): Observations 20,639; R-squared 0.068; Adj. R2 0.0668
    - Column (7): Observations 20,909; R-squared 0.105; Adj. R2 0.104
    - Column (8): Observations 20,909; R-squared 0.056; Adj. R2 0.0547
  - Country fixed effects: YES for all regressions.
  - df reported: 31 or 32 depending on column.
  - Robust t-statistics in parentheses; significance notation: *** p<0.01, ** p<0.05, * p<0.1.
  - Omitted categories: small (10-19 permanent employees), not integrated at all, young less than 6 years, domestic firms, private domestic ownership, no change in permanent employment, high-tech manufacturing, no foreign technology, no innovation, single-product.

*Source: wpiea2019206-print-pdf - REFERENCES (IMF PDF).*

### Annex D: Factors Associated with Employment Growth

### Annex D: Factors Associated with Employment Growth

### Composition of Sample (Table D1)
- Age of Firms (Percent of firms in sample): Young 9% (MENA), 9% (Others); Mature 34% (MENA), 43% (Others); Older 57% (MENA), 48% (Others).
- Initial Size of Firms: Small 40% (MENA), 50% (Others); Medium 38% (MENA), 35% (Others); Large 22% (MENA), 15% (Others).
- State of Technological Advancement: Non-manufacturing 37% (MENA), 40% (Others); Low-tech manufacturing 50% (MENA), 42% (Others); High-tech manufacturing 13% (MENA), 18% (Others).
- Level of Firm Innovation: None 73% (MENA), 50% (Others); Weak 19% (MENA), 22% (Others); Moderate 8% (MENA), 28% (Others).
- Ownership of Firms: Private, domestic 93% (MENA), 93% (Others); Private, foreign 6% (MENA), 4% (Others); Public and mixed 1% (MENA), 3% (Others).
- Trade Integration: Exporting and Importing 15% (MENA), 9% (Others).
- Firms' Employees Status: Temporary workers 4% (MENA), 5% (Others).

### Baseline regressions and specifications with macroeconomic variables (Table D2)
- Sample size and fit:
  - Observations: 27,312 for columns (1)-(4); 26,710 for columns (5)-(6).
  - R-squared: 0.036 (1), 0.072 (2), 0.043 (3), 0.044 (4), 0.046 (5), 0.053 (6).
- Key firm-characteristics coefficient estimates (with robust standard errors in brackets; significance indicated):
  - Age: Mature = -2.691*** [0.879] (1); -2.272*** [0.795] (2); -2.749*** [0.884] (3); -2.765*** [0.885] (4); -2.542*** [0.905] (5); -2.591*** [0.928] (6).
  - Age: Older = -4.773*** [1.053] (1); -4.256*** [0.947] (2); -4.707*** [1.041] (3); -4.723*** [1.043] (4); -4.442*** [1.033] (5); -4.578*** [1.082] (6).
  - Size: Medium = -2.947*** [0.684] (1); -2.877*** [0.532] (2); -2.815*** [0.589] (3); -2.590*** [0.585] (4); -2.868*** [0.530] (5); -2.880*** [0.530] (6).
  - Size: Large = -3.143*** [0.536] (1); -3.076*** [0.418] (2); -2.906*** [0.445] (3); -2.978*** [0.518] (4); -3.029*** [0.413] (5); -3.042*** [0.412] (6).
  - Innovation: Weak = 1.193*** [0.346] (1); 0.744*** [0.229] (2); 0.960*** [0.264] (3); 0.966*** [0.263] (4); 0.784*** [0.243] (5); 0.719*** [0.243] (6).
  - Innovation: Moderate to strong = 2.423*** [0.814] (1); 1.733*** [0.529] (2); 1.945*** [0.561] (3); 1.938*** [0.563] (4); 1.690*** [0.538] (5); 1.574*** [0.496] (6).
  - Trade integration = 2.225*** [0.600] (1); 2.328*** [0.300] (2); 2.428*** [0.541] (3); 2.424*** [0.536] (4); 2.860*** [0.479] (5); 2.588*** [0.421] (6).
  - Share of temporary workers = -0.021** [0.009] (1); -0.032*** [0.006] (2); -0.022*** [0.008] (3); -0.022*** [0.008] (4); -0.030*** [0.007] (5); -0.032*** [0.007] (6).
- MENA region effects and interactions:
  - MENA coefficient entries: -3.048* [1.696] (column where shown); -2.489 [1.801]; -2.756** [1.324]; -0.000 [2.132] (as presented across specifications).
  - Interaction terms in specification (3) include: MENA * Medium firm = -1.580** [0.606]; MENA * Large firm = 0.179 [0.815].
- Macro variables (selected):
  - Fiscal balance coefficient = -0.048 [0.076] and -0.079 [0.071] in respective columns where included.
  - GDP per capita growth = 0.153*** [0.057] and 0.104** [0.046] in respective columns.
  - REER appreciation = -0.053 [0.061] and 0.038 [0.047] in respective columns.
  - Interaction of REER appreciation with MENA = -0.440*** [0.068] (in column where included).
- Constants across columns: 7.219*** [0.930] (1); 9.588*** [0.704] (2); 7.729*** [0.827] (3); 7.680*** [0.844] (4); 7.101*** [1.030] (5); 7.049*** [1.047] (6).
- Estimation notes:
  - Robust standard errors in brackets.
  - Significance notation: *** p<0.01, ** p<0.05, * p<0.1.
  - Column (2) estimated with country fixed effects.

### Regressions with structural variables (Table D3)
- Sample size and fit:
  - Observations: 26,309 for columns (1)-(3); 26,710 for columns (4)-(5).
  - R-squared: 0.048 (1), 0.047 (2), 0.055 (3), 0.056 (4), 0.057 (5).
- Core firm-characteristics (robust standard errors in brackets):
  - Age: Mature = -2.617** [0.976] (1); -2.606** [0.974] (2); -2.665*** [0.991] (3); -2.549*** [0.897] (4); -2.535*** [0.891] (5).
  - Age: Older = -4.586*** [1.127] (1); -4.535*** [1.120] (2); -4.544*** [1.118] (3); -4.506*** [1.097] (4); -4.448*** [1.084] (5).
  - Size: Medium = -2.802*** [0.568] (1); -2.717*** [0.544] (2); -2.736*** [0.525] (3); -2.759*** [0.512] (4); -2.738*** [0.503] (5).
  - Size: Large = -2.967*** [0.490] (1); -2.841*** [0.443] (2); -2.924*** [0.418] (3); -2.917*** [0.389] (4); -2.922*** [0.388] (5).
  - Innovation: Weak = 0.735*** [0.257] (1); 0.751*** [0.261] (2); 0.788*** [0.262] (3); 0.803*** [0.252] (4); 0.840*** [0.244] (5).
  - Innovation: Moderate to strong = 1.635*** [0.440] (1); 1.690*** [0.453] (2); 1.749*** [0.450] (3); 1.725*** [0.428] (4); 1.750*** [0.437] (5).
  - Trade integration = 2.639*** [0.402] (1); 2.738*** [0.481] (2); 2.729*** [0.507] (3); 2.621*** [0.439] (4); 2.591*** [0.424] (5).
  - Share of temporary workers = -0.030*** [0.007] (1); -0.030*** [0.007] (2); -0.030*** [0.007] (3); -0.036*** [0.007] (4); -0.038*** [0.007] (5).
- MENA region indicators:
  - MENA = -2.152 [1.291] (1); -2.364* [1.329] (2); 4.976 [26.401] (3); 0.735 [1.354] (4); 0.884 [2.351] (5).
- Macro variables (selected):
  - Fiscal balance = -0.107 [0.083] (column shown).
  - GDP per capita growth = 0.208*** [0.073] (1); 0.208*** [0.072] (2); 0.134** [0.065] (3); 0.129*** [0.043] (4); 0.130*** [0.044] (5).
  - REER appreciation = -0.052 [0.057] (1); -0.054 [0.058] (2); 0.035 [0.054] (3); 0.024 [0.046] (4); 0.023 [0.044] (5).
  - Interaction of REER appreciation with MENA: -0.412 [0.394] (3); -0.446*** [0.083] (4); -0.328** [0.122] (5).
- Structural variables (objective constraints) and interactions:
  - Collateral need (survey) = -0.004 [0.004] (1); -0.003 [0.005] (2); -0.001 [0.006] (3). Interaction with MENA = -0.024** [0.010] (reported in model where included).
  - Credit line (survey) = 2.005 [1.405] (1); 1.841 [1.352] (2); -0.636 [1.004] (3). Interaction with MENA = 1.334 [1.637].
  - Bribery depth (survey) = 0.010 [0.022] (1); 0.006 [0.021] (2); -0.011 [0.014] (3). Interaction with MENA = 0.114 [0.089].
  - Employment regulation (EPL) = -4.150 [2.914] (1); -3.074 [2.571] (2); -2.730 [2.655] (3). Interaction with MENA = 1.870 [29.005].
  - Ease of doing business (WB) = 0.017 [0.047] (1); 0.009 [0.046] (2); -0.020 [0.047] (3). Interaction with MENA = -0.007 [0.675].
- Structural variables (perception-based constraints) and interactions:
  - Corruption obstacle (survey) = -2.643 [1.749] (column where shown); -3.150* [1.581] (another column). Interaction with MENA = 3.227 [2.465].
  - Access to finance obstacle (survey) = -3.172* [1.746] and -3.172* [1.811] in columns shown. Interaction with MENA = -1.506 [5.103].
  - Access to electricity obstacle (survey) = 3.697** [1.552] and 4.500*** [1.483] in columns shown. Interaction with MENA = -10.120** [4.315].
- Constants in specifications: 11.471*** [3.508] (1); 11.819*** [3.469] (2); 9.027** [3.705] (3); 8.044*** [0.835] (4); 7.931*** [0.802] (5).
- Estimation notes:
  - Robust standard errors in brackets.
  - Significance notation: *** p<0.01, ** p<0.05, * p<0.1.
  - To gain degrees of liberty in regressions including structural variables, especially when interacted with MENA, the authors drop firms' characteristics and macroeconomic variables that were consistently non-significant in previous estimations. The baseline structural regression was estimated with and without the non-significant variables; results are shown in (1) and (2) and, as expected, the results are not affected when dropping the variables.

### Key empirical patterns and takeaways (as reported)
- Older and mature firms have consistently negative and statistically significant coefficients relative to the omitted age category, indicating lower employment growth associated with greater firm age across specifications.
- Medium and large firms have consistently negative and statistically significant coefficients relative to small firms, indicating lower employment growth for larger initial firm size across specifications.
- Innovation (both weak and moderate to strong) and trade integration are positively and significantly associated with employment growth across specifications.
- A higher share of temporary workers is consistently negatively and significantly associated with employment growth.
- GDP per capita growth is generally positive and significant in explaining firm-level employment growth.
- REER appreciation shows a negative and significant interaction with MENA in some specifications (e.g., -0.446*** [0.083] in Table D3 column where reported), indicating region-specific sensitivity to real exchange rate movements.
- Certain structural constraints show large differential interactions with MENA: e.g., Interaction of Access to electricity obstacle with MENA = -10.120** [4.315] where reported.

*Annex D: Factors Associated with Employment Growth — wpiea2019206-print-pdf*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2019/wpiea2019206-print-pdf.pdf_