## Leveraging Trade to Boost Growth in the MENAP and CCA Regions

## Source details

**Canonical URL:** [Leveraging Trade to Boost Growth in the MENAP and CCA Regions](https://www.imf.org/-/media/files/publications/reo/mcd-cca/2017/october/menap/10-chapter-4.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/reo/mcd-cca/2017/october/menap/10-chapter-4.pdf.md)
- [Structured JSON version](/-/media/files/publications/reo/mcd-cca/2017/october/menap/10-chapter-4.pdf.json)

---

### Trade Helps Boost Growth
- Structural reforms to foster investment and job creation, and targeted fiscal policies to mitigate adjustment costs, may be needed to relieve negative consequences of increased openness and ensure inclusive growth.
- Empirical evidence:
  - Trade has a substantial and robust positive effect on growth and income.
  - Faster growth when countries:
    - have a more diversified export structure (Lederman and Maloney 2003);
    - upgrade the quality of their exports (Henn, Papageorgiou, and Spatafora 2015);
    - are well integrated into GVCs (Didier and Pinat 2017).
- IMF staff analysis (131 countries, 20 from MENAP or CCA) shows:
  - investment in infrastructure, foreign direct investment (FDI), and overall openness to trade (sum of exports and imports relative to GDP) all help increase growth in real per capita terms;
  - export diversification, the quality of exports, and participation in GVCs—particularly the share of domestic value added in exports—appear important for growth;
  - a negative relationship between growth and the initial level of per capita GDP, suggesting convergence over time;
  - the positive impact of labor force education on growth found in advanced and emerging market economies becomes ambiguous once low-income countries with very low levels of completed postsecondary education are included.
- On inclusiveness:
  - Impact of trade openness on inequality is less clear; some analysis suggests no significant impact on inequality (Box 4.1).
  - Trade-driven higher incomes help reduce poverty (IMF, World Bank, and WTO 2017) and narrow wage gaps (Council of Economic Advisers 2015).
  - Trade can reduce inequality by lowering prices for food and beverages consumed mainly by the poor (Faijgelbaum and Khandelwal 2016).
  - Increased openness may impose adjustment costs that hurt some communities or groups of workers; Helpman (2016) finds trade has had a modest impact on wage inequality, highlighting the role for domestic policies to mitigate adjustment costs and ensure equitable sharing of benefits.

### Trade Openness Has Declined in Recent Years
- Drivers of decline (consistent with international trends):
  - overall weakness in international economic activity, particularly investment;
  - waning pace of trade liberalization;
  - decline in commodity prices including for oil;
  - slower growth of GVCs.
- Regional patterns:
  - Decline in oil importers has been relatively faster in MENAP than in CCA, reflecting regional conflicts and geopolitical tensions.
  - For MENAP and CCA oil exporters, weakened trade openness largely reflects stagnant or declining oil and gas exports and lower prices.
  - Excluding oil, trade openness has increased slightly in MENAP oil exporters, while it has been broadly stable in CCA oil exporters.
  - For MENAP oil importers, slower export growth than the average emerging market and developing economy explains the trend.
  - In CCA oil importers, relatively fast export growth but import compression driven by a decline in remittances has driven the overall decline in openness.
- Services exports:
  - Services represented, on average, 44 percent of total exports in MENAP and CCA oil importers in 2015 but less than 15 percent in oil exporters.
  - Services exports are dominated by tourism, especially in MENAP oil importers where it represents 51 percent of total services exported.
  - Many countries in the MENAP and CCA regions have restrictive policies in services; professional and transportation services are among the most protected.
  - Restrictions often apply to entry, ownership, and operations; market access can be unpredictable due to opaque and discretionary allocation of licenses.
  - Technological innovations such as e-commerce could help businesses reach international markets at minimum cost and facilitate insertion into GVCs.

### Integration with Global Trade and Partners
- Integration assessment:
  - MENAP and CCA are relatively weakly integrated into the global trade network in both export flows and number of trading partners.
  - MENAP oil exporters are substantially more important in global trade due to a significant share of the global oil market; exports dominated by oil mainly to Asian countries and the United States.
  - CCA oil exporters export mainly to major European countries.
  - MENAP oil importers direct exports primarily to large European markets and the United States.
  - CCA oil importers export mainly to Russia and nearby European countries, such as Bulgaria, as entry points to EU markets.
  - China is a major trading partner; virtually every MENAP and CCA country exports to China.
- Integration metrics:
  - On average, MENAP countries export to about 70 percent of potential trading partners; CCA countries export to about 50 percent of potential markets.
  - Trade penetration (export value per trading partner) is relatively weak in both MENAP and CCA, signaling scope to enhance export quality and domestic value added.

### Export Diversification and Product Quality
- Diversification:
  - Export diversification in MENAP and CCA underperforms relative to EMDEs.
  - Oil exporters are the least diversified and likely to face higher output volatility.
  - Levels of diversification among oil importers approach the EMDE average; MENAP countries fare better due to geographic access to European markets and FDI inflows from the euro area and Gulf Cooperation Council countries.
  - In CCA oil importers, diversification progress stalled recently partly because FDI slowed (Tajikistan).
- Export quality:
  - Most MENAP and CCA countries produce lower-quality exports than other EMDEs; only Jordan and Tunisia have export quality above the EMDE average.
  - Some oil importers improved export quality recently, mainly in apparel (Egypt, Jordan, Pakistan) and manufacturing (Armenia, Georgia, Jordan, Morocco, Tunisia).
  - Quality of oil exports remains relatively low, with deterioration in Algeria and Azerbaijan.
- Economic complexity:
  - Export complexity among oil exporters is low compared with oil importers.
  - Complexity among oil importers is higher for MENAP than for CCA oil importers, reflecting supply chain links with manufacturing firms in the euro area.

### Global Value Chains (GVCs) Potential
- Current integration:
  - Current levels of GVC integration do not allow full exploitation of potential.
  - Oil importers are generally better integrated into GVCs than oil exporters.
  - Backward integration (share of foreign value added used in production of exports) is relatively high in Jordan, Lebanon, and Tunisia (MENAP) and the Kyrgyz Republic (CCA) compared with EMDEs.
  - Forward integration (share of value added used in destination country’s production) is above EMDE average in Egypt, Mauritania, and Morocco (MENAP) and in Armenia (CCA).
  - In oil exporters, backward integration is particularly low, indicating imports are mainly finished products for consumption and investment; forward integration is relatively high primarily due to crude oil exports processed abroad.
- Trends and potential:
  - Most MENAP oil importers improved both backward and forward GVC participation, aided by diversification (light manufacturing in Jordan, Morocco, Tunisia).
  - CCA oil importers improved mainly in backward integration, partly by becoming a hub for transport of Chinese products to Russia and the rest of CCA; forward integration declined, consistent with slowing export diversification.
  - Additional progress in GVC integration, particularly for oil importers in both regions, may be possible in the context of the Belt and Road Initiative.

### The Trade Environment Needs Upgrading
- Geopolitical tensions and conflicts (October 2016 World Economic Outlook) have negatively affected the trade environment through:
  - Disruption of economic activity and infrastructure and death or displacement of people active in the labor force in Afghanistan, Iraq, Libya, Somalia, Syria, and Yemen.
  - Conflict spillovers to neighboring countries, reducing cross-border trade, tourism, and generating refugee inflows (Jordan, Lebanon, Pakistan; Rother and others 2016).
  - Specific disruptions: conflict in Syria disrupted Iraq’s main trading link to the Mediterranean; conflict in Iraq affected Jordan’s export route to Iran.
  - Elevated uncertainty and rising insecurity weakened FDI inflows.
  - The diplomatic rift between Qatar and other MENAP countries is affecting trade and financial flows (see Box 1.1 in Chapter 1).
- Business climate and logistics:
  - Procedures for trading across borders have been comparable to or better than the EMDE average, but the trading environment has deteriorated in MENAP countries while improving in CCA countries.
  - Need for structural reforms to improve efficiency and reduce costs associated with compliance procedures and domestic transportation.
- Foreign exchange restrictions:
  - Foreign exchange restrictions in some MENAP and CCA countries hinder trade expansion.
  - Empirical evidence (Wei and Zhang 2007): a one standard deviation increase in controls on trade payments or foreign exchange transactions reduces trade by the same amount as tariff increases of 11 to 14 percentage points.
- Exchange rate movements and GVCs:
  - Some studies find a 10 percent real effective depreciation is associated with a rise in real net exports of, on average, 1.5 percent of GDP, with substantial cross-country variation.
  - Increased participation in GVCs has reduced the relevance of exchange rate movements for trade flows.

### Participation in Trade Agreements and Trade Facilitation
- Use of trade agreements:
  - Only MENAP oil importers are notably active in bilateral and regional trade agreements; most other countries in MENAP and CCA have substantially fewer trade agreements than the EMDE average.
  - MENAP’s least-developed countries (Afghanistan, Djibouti, Mauritania, Yemen) enjoy duty-free and quota-free access to all or almost all export markets under WTO rules.
- Trade Facilitation Agreement:
  - The Trade Facilitation Agreement came into force in early 2017.
  - WTO estimates implementation would cut customs-related costs of merchandise trade by 14 percent, particularly for developing economies, and could lead to a $1 trillion annual increase in global trade.
- Accession and bilateral agreements:
  - Eight MENAP countries (Algeria, Iran, Iraq, Libya, Lebanon, Somalia, Sudan, Syria) and two CCA countries (Azerbaijan, Uzbekistan) have been negotiating WTO accession, with slow progress.
  - Several countries have concluded association agreements with the European Union (Algeria, Egypt, Jordan, Lebanon, Morocco, Syria, Tunisia) and Georgia signed a free trade agreement with China.
  - Some CCA countries (Armenia, Azerbaijan) and Iraq have signed cooperation and partnership agreements with the European Union.
  - Morocco and Tunisia joined the Compact with Africa.

### Participation in Global Value Chains (policy context)
- The Belt and Road Initiative is likely to increase backward integration further.
- Foreign investment triggered by the Compact with Africa may support forward integration for MENAP oil importers.
- Most oil exporters in MENAP and CCA reduced backward integration but improved forward integration, reflecting efforts to increase oil processing and refining and increase value added.
- Oil importers are generally better placed to benefit from improved global growth momentum, but scope remains to improve export quality and reverse CCA oil importers’ decline in forward integration.
- Oil exporters need to increase export diversification and improve export quality.

### Conclusion and Policy Options
- Simulated growth impacts (based on econometric analysis in Annex 4.1):
  - A sustained increase in trade openness equal to the best historical period-over-period improvement observed in the region could add 1 percentage point to the average growth rate over the following five-year period.
  - If greater trade openness is supported by enhanced diversification, improved export quality, or more active participation in GVCs, the impact would be even higher.
  - Implementing reforms that boost trade could raise the level of income by 5–10 percent within the following five to ten years.
- Specific numeric benchmarks from Figure 4.9 (increase equal to best historical period-over-period improvement in last 20 years):
  - For trade openness: 7.7 percentage points (pps).
  - For global value chain: 4 pps.
  - For diversification: 2.4 pps.
  - For quality: 1.5 pps.
- Policy recommendations:
  - Increase trade openness, participation in global value chains, export diversification, and product quality.
  - Diversify into sectors with substantial job-creating potential.
  - Upgrade export quality by improving access to finance, education, and technologies.
  - Use fiscal policies to mitigate transitional costs of greater openness and to create fiscal space for lower trade tax revenue and increased public investment in infrastructure.
  - Pursue broader structural reforms to improve the business and investment environment.

### The Trade and Inclusiveness Nexus (Box 4.1)
- Empirical findings (106 countries, 1980–2013, including 11 from MENAP and CCA):
  - Inclusiveness, as measured by the Gini inequality index, does not seem to be directly affected by trade openness.
  - Other variables appear more important: financial deepening, education, and employment shares.
- Regression summary (panel fixed effects with time and country fixed effects; robust standard errors clustered at the country level):
  - Observations: 435.
  - R-squared: 0.237 for Market Gini; 0.161 for Net Gini.
  - Number of countries: 106.
  - Coefficients of note (with significance):
    - Financial deepening (t 21): 0.0276*** in Market Gini (0.0102).
    - Education (t 21): −0.736* in Market Gini (0.398); −0.725** in Net Gini (0.349).
    - Agriculture Employment share (t 21): −0.124*** in Market Gini (0.0388); −0.0830** in Net Gini (0.0353).
    - Industry Employment share (t 21): −0.199*** in Market Gini (0.0712); −0.0841 in Net Gini (0.0602).
    - Trade Openness (t21): −0.00140 (0.00869) in Market Gini; 0.00605 (0.00884) in Net Gini (not statistically significant).
- Policy implications:
  - Public policies are important to manage adverse side effects for specific workers and communities.
  - Fiscal redistribution policies can be effective in offsetting transitional costs, especially for industrial sector job losses.
  - Domestic policies to mitigate trade-related adjustments may include:
    1. Active labor market policies: job search assistance, training programs, carefully designed wage insurance.
    2. Unemployment insurance, employment protection, and other passive labor policies.
    3. Complementary policies in education, housing, credit, infrastructure, and place-based measures.
  - Reforms to the business environment that support private sector development are also important.

### Leveraging Trade Agreements for Growth (Box 4.2)
- Multilateral level:
  - Many MENAP countries (Algeria, Iran, Iraq, Libya, Lebanon, Somalia, Sudan, Syria) and some CCA countries (Azerbaijan, Turkmenistan, Uzbekistan) are not WTO members; several others have joined recently (Afghanistan, Kazakhstan, Tajikistan, Yemen).
  - Recent WTO accession evidence (Kireyev 2016): among recently acceded members (1996–2015), impact was neutral in 63 percent of cases, positive in 24 percent, and negative in 13 percent.
  - New WTO members achieved substantial positive results in openness, diversification, growth, inflation control, fiscal deficits, and attracting FDI in many cases.
- Regional level:
  - Number of PTAs notified to the WTO rose from about 50 in 1990 to about 300 in 2017, many involving MENAP and CCA countries.
  - PTAs have expanded scope to include customs regulations, export taxes, countervailing measures, and technical barriers to trade.
  - Most PTAs involving MENAP and CCA remain relatively shallow; few are “deep” agreements effective for GVC integration and attracting FDI.
  - The Eurasian Economic Union is an exception, including three CCA countries (Armenia, Kazakhstan, the Kyrgyz Republic).
  - Example: a Deep and Comprehensive Free Trade Area under negotiation between Morocco, Tunisia, and the EU could yield potential long-term GDP gains of up to 1.6 percent for Morocco and 7.4 percent for Tunisia (EC 2013), with associated export expansion and trade balance improvements for MENAP oil-importing countries.

### Annex 4.1 — Openness and Growth: empirical specification and data
- Baseline regression:
  - ∆ y_{i,t} = α1 y_{i,t−1} + α2 Controls_{i,t} + α3 TC_{i,t} + δ_{t} + γ_{i} + ε_{i,t}
  - ∆ y_{i,t} is growth of real GDP per capita at time t for country i; y_{i,t−1} is the log of real GDP per capita.
  - Controls_{i,t} includes logs of terms of trade, level of education, a proxy for public infrastructure development, and the ratio of foreign direct investment (FDI) to GDP.
  - TC_{i,t} is a set of trade characteristics included sequentially.
  - Period dummies indicate a declining trend in global growth since 1960.
- Sample and aggregation:
  - Nonoverlapping five-year period averages for 131 countries covering 1960 to 2013.
- Estimation:
  - System GMM combining levels and differences specifications.
  - Validation approaches: Hansen test of over-identifying restrictions, incremental Hansen test, and test for serial correlation of the error terms.
- Caveats:
  - For specifications with many instruments, a restricted sample of control variables is used to reduce instruments.
  - Limitations of using lags as instruments in a trade-growth context are acknowledged; interpret results with caution.

### Estimated coefficients and key statistics (selected from Table 4.1)
- Dependent variable: growth in GDP per capita. Sample: 131 countries (varies by specification).
- Selected estimated coefficients (coefficient (standard error)) across columns (1)–(6):
  - Initial GDP per Capita:
    - Column (1): 25.014*** (0.892)
    - Column (2): 24.618*** (0.827)
    - Column (3): 25.061*** (0.697)
    - Column (4): 25.664*** (0.909)
    - Column (5): 25.741*** (1.344)
    - Column (6): 25.730*** (1.516)
  - Labor Force Education:
    - Column (1): 0.357 (0.468)
    - Column (2): 0.104 (0.594)
    - Column (3): 0.429 (0.474)
    - Column (4): −0.358 (0.486)
    - Column (5): −0.580 (0.498)
    - Column (6): −0.436 (0.346)
  - Infrastructure:
    - Column (1): 2.241*** (0.640)
    - Column (2): 2.087*** (0.540)
    - Column (3): 2.458*** (0.411)
    - Column (4): 2.574*** (0.697)
    - Column (5): 3.285*** (1.086)
    - Column (6): 2.908** (1.414)
  - Inflows of FDI/GDP:
    - Column (1): 0.667** (0.297)
    - Column (2): 0.761* (0.434)
    - Column (3): 1.313** (0.635)
    - Column (4): 1.102* (0.593)
  - Terms of Trade:
    - Column (2): −7.435 (6.520)
    - Column (3): −0.0830 (10.75)
    - Column (4): 3.335 (10.17)
  - Trade Openness:
    - Column (1): 2.446*** (0.781)
    - Column (2): 1.919*** (0.633)
    - Column (3): 1.996*** (0.620)
    - Column (4): 2.796*** (0.924)
    - Column (5): 2.240 (1.549)
    - Column (6): 2.122 (1.471)
  - Export diversification:
    - Column (4): 4.249** (1.839)
  - Export Quality:
    - Column (5): 5.034* (2.892)
  - Participation in GVC:
    - Column (4): 9.170** (4.075)
  - Backward Integration:
    - Column (5): 6.642 (7.586)
    - Column (6): 9.771 (6.924)
  - Forward Integration:
    - Column (6): 11.72** (5.356)
  - Domestic Value Added:
    - Column (5): 48.01* (24.90)
- Sample and instrumentation:
  - Number of Observations:
    - Column (1): 1,030
    - Column (2): 1,021
    - Column (3): 1,037
    - Column (4): 641
    - Column (5): 641
    - Column (6): 641
  - Number of Countries:
    - Column (1): 131
    - Column (2): 123
    - Column (3): 129
    - Column (4): 131
    - Column (5): 131
    - Column (6): 131
  - Number of Instruments:
    - Column (1): 118
    - Column (2): 123
    - Column (3): 123
    - Column (4): 83
    - Column (5): 52
    - Column (6): 52
  - Period dummies: yes (in all specifications)

### Validation tests and robustness
- Full Hansen Test (p-values):
  - Column (1): 0.293
  - Column (2): 0.240
  - Column (3): 0.246
  - Column (4): 0.197
  - Column (5): 0.112
  - Column (6): 0.153
- Incremental Hansen Test (p-values):
  - Column (1): 0.805
  - Column (2): 0.632
  - Column (3): 0.398
  - Column (4): 0.365
  - Column (5): 0.105
  - Column (6): 0.108
- p-value of AR(2) statistic:
  - Column (1): 0.236
  - Column (2): 0.470
  - Column (3): 0.392
  - Column (4): 0.511
  - Column (5): 0.495
  - Column (6): 0.557
- Coefficient interpretations:
  - Initial GDP per capita has a negative and statistically significant coefficient, suggesting conditional convergence across countries in GDP per capita.
  - Labor force education coefficient is not statistically significant in the full sample but is positive and significant in a smaller sample of 82 advanced and emerging market economies.
  - Infrastructure and inflows of FDI have positive and statistically significant coefficients.
  - Volatility of terms of trade is not statistically significant in five-year period averages.

*Prepared by Alexei Kireyev (lead author), Maxym Kryshko, Boaz Nandwa, and Magali Pinat, with research assistance by James Aylward and Samira Kalla.*

### Box 2.3 in Chapter 2). Structural reforms to foster

### Leveraging Trade to Boost Growth in the MENAP and CCA Regions

### Trade Helps Boost Growth
- Structural reforms to foster investment and job creation, as well as targeted fiscal policies to mitigate adjustment costs, may be needed to relieve any negative consequences of increased openness and to ensure the resulting boost to growth is as inclusive as possible.
- Empirical evidence confirms a substantial and robust positive effect of trade on growth and income.
- Findings indicate faster growth when countries:
  - have a more diversified export structure (Lederman and Maloney 2003);
  - upgrade the quality of their exports (Henn, Papageorgiou, and Spatafora 2015);
  - are well integrated into GVCs (Didier and Pinat 2017).
- IMF staff analysis covering 131 countries, 20 of which were from the MENAP or CCA regions, shows:
  - investment in infrastructure, foreign direct investment (FDI), and overall openness to trade (sum of exports and imports relative to GDP) all help increase growth in real per capita terms;
  - export diversification, the quality of exports, and participation in GVCs—particularly the share of domestic value added in exports—appear important for growth;
  - a negative relationship between growth and the initial level of per capita GDP, suggesting convergence over time;
  - the positive impact of labor force education on growth found in advanced and emerging market economies becomes ambiguous once low-income countries with very low levels of completed postsecondary education are included.
- On inclusiveness:
  - the impact of trade openness on inequality is less clear; some analysis suggests no significant impact on inequality (Box 4.1);
  - trade-driven higher incomes help reduce poverty (IMF, World Bank, and WTO 2017) and narrow wage gaps (Council of Economic Advisers 2015);
  - trade can reduce inequality by lowering prices for food and beverages consumed mainly by the poor (Faijgelbaum and Khandelwal 2016);
  - increased openness may impose adjustment costs that hurt some communities or groups of workers; Helpman (2016) finds trade has had a modest impact on wage inequality, highlighting the role for domestic policies to mitigate adjustment costs and ensure equitable sharing of benefits.

### Trade Openness Has Declined in Recent Years
- Trade openness has declined significantly across the MENAP and CCA regions, consistent with international trends driven by:
  - overall weakness in international economic activity, particularly investment;
  - waning pace of trade liberalization;
  - decline in commodity prices including for oil;
  - slower growth of GVCs.
- Regional patterns:
  - Decline in oil importers has been relatively faster in MENAP than in CCA, reflecting regional conflicts and geopolitical tensions.
  - For MENAP and CCA oil exporters, weakened trade openness largely reflects stagnant or declining oil and gas exports and lower prices.
  - Excluding oil, trade openness has increased slightly in MENAP oil exporters, while it has been broadly stable in CCA oil exporters.
  - For MENAP oil importers, slower export growth than the average emerging market and developing economy explains the trend.
  - In CCA oil importers, relatively fast export growth but import compression driven by a decline in remittances has driven the overall decline in openness.
- Services exports:
  - Services represented, on average, 44 percent of total exports in MENAP and CCA oil importers in 2015 but less than 15 percent in oil exporters.
  - In the region, services exports are dominated by tourism, especially in MENAP oil importers where it represents 51 percent of total services exported.
  - Many countries in the MENAP and CCA regions have restrictive policies in services; professional and transportation services are among the most protected.
  - Restrictions often apply to entry, ownership, and operations; market access can be unpredictable due to opaque and discretionary allocation of licenses.
  - Technological innovations such as e-commerce could help businesses reach international markets at minimum cost and facilitate insertion into GVCs for goods and services.

### Integration with Global Trade and Partners
- MENAP and CCA appear relatively weakly integrated into the global trade network in both export flows and number of trading partners.
- MENAP oil exporters are substantially more important in global trade by virtue of controlling a significant share of the global oil market; their exports are dominated by oil exports mainly to Asian countries and the United States.
- CCA oil exporters export mainly to major European countries.
- MENAP oil importers direct exports primarily to large European markets and the United States.
- CCA oil importers export mainly to Russia and nearby European countries, such as Bulgaria, as entry points to EU markets.
- China has emerged as a major trading partner; virtually every MENAP and CCA country exports to China.
- Integration metrics:
  - On average, MENAP countries export to about 70 percent of potential trading partners; CCA countries export to about 50 percent of potential markets—indicating room to expand export markets.
  - Trade penetration (export value per trading partner) is relatively weak in both MENAP and CCA, signaling scope to enhance export quality and domestic value added.

### Export Diversification and Product Quality
- Export diversification in MENAP and CCA underperforms relative to emerging market and developing economies.
  - Oil exporters are the least diversified and likely to face higher output volatility.
  - Levels of diversification among oil importers approach the EMDE average, with MENAP countries faring better due to geographic access to European markets and inflows of FDI from the euro area and Gulf Cooperation Council countries.
  - In CCA oil importers, diversification progress stalled in recent years in part because FDI slowed (Tajikistan).
- Export quality:
  - Most MENAP and CCA countries produce lower-quality exports than other EMDEs; only Jordan and Tunisia have export quality above the EMDE average.
  - Some oil importers improved export quality recently, mainly in apparel (Egypt, Jordan, Pakistan) and manufacturing (Armenia, Georgia, Jordan, Morocco, Tunisia).
  - Quality of oil exports remains relatively low, with deterioration in Algeria and Azerbaijan.
- Economic complexity:
  - Export complexity among oil exporters is low compared with oil importers.
  - Complexity among oil importers is higher for MENAP than for CCA oil importers, reflecting supply chain links with manufacturing firms in the euro area.

### Global Value Chains (GVCs) Potential
- Current levels of GVC integration do not allow full exploitation of potential.
  - Oil importers are generally better integrated into GVCs than oil exporters.
  - Backward integration (share of foreign value added used in production of exports) is relatively high in Jordan, Lebanon, and Tunisia (MENAP) and the Kyrgyz Republic (CCA) compared with EMDEs.
  - Forward integration (share of value added used in destination country’s production) is above EMDE average in Egypt, Mauritania, and Morocco (MENAP) and in Armenia (CCA).
  - In oil exporters, backward integration is particularly low, indicating imports are mainly finished products for consumption and investment; forward integration is relatively high primarily due to crude oil exports that are processed abroad.
- Trends:
  - Most MENAP oil importers improved both backward and forward GVC participation, aided by diversification (light manufacturing in Jordan, Morocco, Tunisia).
  - CCA oil importers improved mainly in backward integration, partly by becoming a hub for transport of Chinese products to Russia and the rest of CCA; forward integration declined, consistent with slowing export diversification.
  - Additional progress in GVC integration, particularly for oil importers in both regions, may be possible in the context of the Belt and Road Initiative.

*Prepared by Alexei Kireyev (lead author), Maxym Kryshko, Boaz Nandwa, and Magali Pinat, with research assistance by James Aylward and Samira Kalla.*

### Chapter 2). The Belt and Road Initiative is likely

### Chapter 2). The Belt and Road Initiative is likely

### The Trade Environment Needs Upgrading
- Geopolitical tensions and conflicts (October 2016 World Economic Outlook) have negatively affected the trade environment through:
  - Disruption of economic activity and infrastructure and death or displacement of people active in the labor force in Afghanistan, Iraq, Libya, Somalia, Syria, and Yemen.
  - Conflict spillovers to neighboring countries, reducing cross-border trade, tourism, and generating refugee inflows (Jordan, Lebanon, Pakistan; Rother and others 2016).
  - Specific disruptions: conflict in Syria disrupted Iraq’s main trading link to the Mediterranean; conflict in Iraq affected Jordan’s export route to Iran.
  - Elevated uncertainty and rising insecurity weakened FDI inflows.
  - The diplomatic rift between Qatar and other MENAP countries is affecting trade and financial flows (see Box 1.1 in Chapter 1).
- Business climate and logistical performance:
  - Procedures for trading across borders have been comparable to or better than the EMDE average, but the trading environment has deteriorated in MENAP countries while improving in CCA countries (Figure 4.7).
  - Need for structural reforms to improve efficiency and reduce costs associated with compliance procedures and domestic transportation.
- Foreign exchange restrictions:
  - Foreign exchange restrictions in some MENAP and CCA countries hinder trade expansion (Figure 4.7).
  - Empirical evidence (Wei and Zhang 2007): a one standard deviation increase in controls on trade payments or foreign exchange transactions reduces trade by the same amount as tariff increases of 11 to 14 percentage points.
- Exchange rate movements and GVCs:
  - Historical evidence: some studies find a 10 percent real effective depreciation is associated with a rise in real net exports of, on average, 1.5 percent of GDP, with substantial cross-country variation.
  - Increased participation in GVCs has reduced the relevance of exchange rate movements for trade flows, pointing to the need to improve the overall trade environment (October 2015 World Economic Outlook).

### Participation in Trade Agreements and Trade Facilitation
- Use of trade agreements:
  - Only MENAP oil importers are notably active in bilateral and regional trade agreements; most other countries in MENAP and CCA have substantially fewer trade agreements than the EMDE average (Figure 4.8).
  - MENAP’s least-developed countries (Afghanistan, Djibouti, Mauritania, Yemen) enjoy duty-free and quota-free access to all or almost all export markets under WTO rules.
- Trade Facilitation Agreement:
  - The Trade Facilitation Agreement came into force in early 2017.
  - WTO estimates implementation would cut customs-related costs of merchandise trade by 14 percent, particularly for developing economies, and could lead to a $1 trillion annual increase in global trade.
- Recent accession and bilateral agreements:
  - Eight MENAP countries (Algeria, Iran, Iraq, Libya, Lebanon, Somalia, Sudan, Syria) and two CCA countries (Azerbaijan, Uzbekistan) have been negotiating WTO accession, with slow progress.
  - Several countries have concluded association agreements with the European Union (Algeria, Egypt, Jordan, Lebanon, Morocco, Syria, Tunisia) and Georgia signed a free trade agreement with China.
  - Some CCA countries (Armenia, Azerbaijan) and Iraq have signed cooperation and partnership agreements with the European Union.
  - Morocco and Tunisia joined the Compact with Africa (see Box 2.3 in Chapter 2).

### Participation in Global Value Chains (GVCs)
- Trends:
  - The Belt and Road Initiative is likely to increase backward integration further.
  - Foreign investment triggered by the Compact with Africa may support forward integration for MENAP oil importers.
  - Most oil exporters in MENAP and CCA reduced backward integration but improved forward integration, reflecting efforts to increase oil processing and refining and increase value added.
- Regional strengths and weaknesses:
  - Oil importers are generally better placed to benefit from improved global growth momentum, but scope remains to improve export quality and reverse CCA oil importers’ decline in forward integration.
  - Oil exporters need to increase export diversification and improve export quality.

### Conclusion and Policy Options
- Simulated growth impacts (based on econometric analysis in Annex 4.1):
  - A sustained increase in trade openness equal to the best historical period-over-period improvement observed in the region could add 1 percentage point to the average growth rate over the following five-year period (Figure 4.9).
  - If greater trade openness is supported by enhanced diversification, improved export quality, or more active participation in GVCs, the impact would be even higher.
  - Implementing reforms that boost trade could raise the level of income by 5–10 percent within the following five to ten years.
- Specific numeric benchmarks from Figure 4.9 (the growth increase is conditional on an increase in the given trade measure equal to the best historical period-over-period improvement observed in the region in the last 20 years):
  - For trade openness: 7.7 percentage points (pps).
  - For global value chain: 4 pps.
  - For diversification: 2.4 pps.
  - For quality: 1.5 pps.
- Policy recommendations:
  - Increase trade openness, participation in global value chains, export diversification, and product quality.
  - Diversify into sectors with substantial job-creating potential.
  - Upgrade export quality by improving access to finance, education, and technologies.
  - Use fiscal policies to mitigate transitional costs of greater openness and to create fiscal space for lower trade tax revenue and increased public investment in infrastructure.
  - Pursue broader structural reforms to improve the business and investment environment.

### The Trade and Inclusiveness Nexus (Box 4.1)
- Empirical findings (data for 106 countries, 1980–2013, including 11 from MENAP and CCA):
  - Inclusiveness, as measured by the Gini inequality index, does not seem to be directly affected by trade openness.
  - Other variables appear more important, such as financial deepening, education, and employment shares.
- Regression results (Table 4.1.1 highlights; panel fixed effects with time and country fixed effects; robust standard errors clustered at the country level):
  - Observations: 435.
  - R-squared: 0.237 for Market Gini; 0.161 for Net Gini.
  - Number of countries: 106.
  - Coefficients of note (with significance):
    - Financial deepening (t 21): 0.0276*** in Market Gini (0.0102).
    - Education (t 21): −0.736* in Market Gini (0.398); −0.725** in Net Gini (0.349).
    - Agriculture Employment share (t 21): −0.124*** in Market Gini (0.0388); −0.0830** in Net Gini (0.0353).
    - Industry Employment share (t 21): −0.199*** in Market Gini (0.0712); −0.0841 in Net Gini (0.0602).
  - Trade Openness (t21): −0.00140 (0.00869) in Market Gini; 0.00605 (0.00884) in Net Gini (not statistically significant).
- Policy implications:
  - Public policies are important to manage adverse side effects for specific workers and communities.
  - Fiscal redistribution policies can be effective in offsetting transitional costs, especially for industrial sector job losses.
  - Domestic policies to mitigate trade-related adjustments may include:
    1. Active labor market policies: job search assistance, training programs, carefully designed wage insurance.
    2. Unemployment insurance, employment protection, and other passive labor policies.
    3. Complementary policies in education, housing, credit, infrastructure, and place-based measures.
  - Reforms to the business environment that support private sector development are also important.

### Leveraging Trade Agreements for Growth (Box 4.2)
- Multilateral level:
  - Many MENAP countries (Algeria, Iran, Iraq, Libya, Lebanon, Somalia, Sudan, Syria) and some CCA countries (Azerbaijan, Turkmenistan, Uzbekistan) are not WTO members; several others have joined recently (Afghanistan, Kazakhstan, Tajikistan, Yemen).
  - Recent WTO accession evidence (Kireyev 2016): among recently acceded members (1996–2015), impact was neutral in 63 percent of cases, positive in 24 percent, and negative in 13 percent.
  - New WTO members achieved substantial positive results in openness, diversification, growth, inflation control, fiscal deficits, and attracting FDI in many cases.
- Regional level:
  - Number of PTAs notified to the WTO rose from about 50 in 1990 to about 300 in 2017, many involving MENAP and CCA countries.
  - PTAs have expanded scope to include customs regulations, export taxes, countervailing measures, and technical barriers to trade.
  - Most PTAs involving MENAP and CCA remain relatively shallow; few are “deep” agreements effective for GVC integration and attracting FDI.
  - The Eurasian Economic Union is an exception, including three CCA countries (Armenia, Kazakhstan, the Kyrgyz Republic).
  - Example: a Deep and Comprehensive Free Trade Area under negotiation between Morocco, Tunisia, and the EU could yield potential long-term GDP gains of up to 1.6 percent for Morocco and 7.4 percent for Tunisia (EC 2013), with associated export expansion and trade balance improvements for MENAP oil-importing countries.

*International Monetary Fund | October 2017*

### Annex 4.1. Trade

### Annex 4.1. Trade

### Openness and Growth — empirical specification and data
- Baseline regression:
  - ∆ y_{i,t} = α1 y_{i,t−1} + α2 Controls_{i,t} + α3 TC_{i,t} + δ_{t} + γ_{i} + ε_{i,t}
  - ∆ y_{i,t} is growth of real GDP per capita at time t for country i; y_{i,t−1} is the log of real GDP per capita.
  - Controls_{i,t} includes logs of terms of trade, level of education, a proxy for public infrastructure development, and the ratio of foreign direct investment (FDI) to GDP.
  - TC_{i,t} is a set of trade characteristics included sequentially.
  - Period dummies indicate a declining trend in global growth since 1960.
- Sample and aggregation:
  - Nonoverlapping five-year period averages for 131 countries covering 1960 to 2013.
- Estimation method:
  - System generalized method of moments (system GMM) combining levels and differences specifications to address unobserved country effects and endogeneity.
  - Three validation approaches: the Hansen test of over-identifying restrictions, the incremental Hansen test, and the test for serial correlation of the error terms.
- Caveats:
  - For specifications where the number of instruments is close to or larger than the number of countries, a restricted sample of control variables is used to reduce instruments.
  - Limitations of using lags as instruments in a trade-growth context are acknowledged; results should be interpreted with caution.

### Estimated coefficients and key statistics (Table 4.1)
- Dependent variable: growth in GDP per capita.
- Sample: 131 countries (different estimations use subsets as noted).
- Robust standard errors are reported in parentheses. Significance codes: * p < 0.1; ** p < 0.05; *** p < 0.01.
- Estimated coefficients (selected variables shown with coefficients and standard errors as in columns (1)–(6)):
  - Initial GDP per Capita:
    - Column (1): 25.014*** (0.892)
    - Column (2): 24.618*** (0.827)
    - Column (3): 25.061*** (0.697)
    - Column (4): 25.664*** (0.909)
    - Column (5): 25.741*** (1.344)
    - Column (6): 25.730*** (1.516)
  - Labor Force Education:
    - Column (1): 0.357 (0.468)
    - Column (2): 0.104 (0.594)
    - Column (3): 0.429 (0.474)
    - Column (4): −0.358 (0.486)
    - Column (5): −0.580 (0.498)
    - Column (6): −0.436 (0.346)
  - Infrastructure:
    - Column (1): 2.241*** (0.640)
    - Column (2): 2.087*** (0.540)
    - Column (3): 2.458*** (0.411)
    - Column (4): 2.574*** (0.697)
    - Column (5): 3.285*** (1.086)
    - Column (6): 2.908** (1.414)
  - Inflows of FDI/GDP:
    - Column (1): 0.667** (0.297)
    - Column (2): 0.761* (0.434)
    - Column (3): 1.313** (0.635)
    - Column (4): 1.102* (0.593)
  - Terms of Trade:
    - Column (2): −7.435 (6.520)
    - Column (3): −0.0830 (10.75)
    - Column (4): 3.335 (10.17)
  - Trade Openness:
    - Column (1): 2.446*** (0.781)
    - Column (2): 1.919*** (0.633)
    - Column (3): 1.996*** (0.620)
    - Column (4): 2.796*** (0.924)
    - Column (5): 2.240 (1.549)
    - Column (6): 2.122 (1.471)
  - Export diversification:
    - Column (4): 4.249** (1.839)
  - Export Quality:
    - Column (5): 5.034* (2.892)
  - Participation in GVC:
    - Column (4): 9.170** (4.075)
  - Backward Integration:
    - Column (5): 6.642 (7.586)
    - Column (6): 9.771 (6.924)
  - Forward Integration:
    - Column (6): 11.72** (5.356)
  - Domestic Value Added:
    - Column (5): 48.01* (24.90)
  - Constant:
    - Column (1): 30.66*** (6.340)
    - Column (2): 20.39*** (7.321)
    - Column (3): 15.01 (10.13)
    - Column (4): 68.21** (30.83)
    - Column (5): 35.92 (53.80)
    - Column (6): 19.05 (49.75)
- Sample and instrumentation details:
  - Number of Observations:
    - Column (1): 1,030
    - Column (2): 1,021
    - Column (3): 1,037
    - Column (4): 641
    - Column (5): 641
    - Column (6): 641
  - Number of Countries:
    - Column (1): 131
    - Column (2): 123
    - Column (3): 129
    - Column (4): 131
    - Column (5): 131
    - Column (6): 131
  - Number of Instruments:
    - Column (1): 118
    - Column (2): 123
    - Column (3): 123
    - Column (4): 83
    - Column (5): 52
    - Column (6): 52
  - Period dummies: yes (in all specifications)

### Validation tests and robustness
- Full Hansen Test (p-values):
  - Column (1): 0.293
  - Column (2): 0.240
  - Column (3): 0.246
  - Column (4): 0.197
  - Column (5): 0.112
  - Column (6): 0.153
- Incremental Hansen Test (p-values):
  - Column (1): 0.805
  - Column (2): 0.632
  - Column (3): 0.398
  - Column (4): 0.365
  - Column (5): 0.105
  - Column (6): 0.108
- p-value of AR(2) statistic:
  - Column (1): 0.236
  - Column (2): 0.470
  - Column (3): 0.392
  - Column (4): 0.511
  - Column (5): 0.495
  - Column (6): 0.557
- Interpretation of coefficients:
  - Initial GDP per capita has a negative and statistically significant coefficient, suggesting conditional convergence across countries in GDP per capita.
  - Labor force education coefficient is not statistically significant in the full sample, consistent with findings when using a wide country sample (but noted to be positive and significant in a smaller sample of 82 advanced and emerging market economies).
  - Infrastructure and inflows of FDI have positive and statistically significant coefficients.
  - Volatility of terms of trade is not statistically significant in five-year period averages, potentially due to reduced volatility from averaging.

*Source: IMF staff calculations.*

---


_Source: https://www.imf.org/-/media/files/publications/reo/mcd-cca/2017/october/menap/10-chapter-4.pdf_