## _wp10238

## Source details

**Canonical URL:** [_wp10238](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp10238.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp10238.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp10238.pdf.json)

---

### Introduction and main findings
- Paper examines effect of growth shocks in European countries on economic activity in Morocco and Tunisia using structural vector autoregressions (SVARs) following Bayoumi and Swiston (2009).
- Key empirical findings:
  - Positive shocks to a growth composite of trade partners have significant expansionary hump-shaped effects on growth in Morocco and Tunisia.
  - Transmission channels differ: in Tunisia, exports (and to a much lesser extent tourism) are the major channels; in Morocco, exports, remittances, and tourism all play important roles.
  - Sectoral VARs: export-oriented sectors in Tunisia respond significantly to external growth shocks; service-related sectors in Morocco are most sensitive.
  - Results are consistent with the evolution of macroeconomic aggregates during the 2008 recession in Europe.

### Background: trade liberalization and association agreements
- Tunisia:
  - Signed an association agreement with the EU in July 1995 and implemented it in March 1998; amendment covering agricultural products implemented on January 1, 2001.
- Morocco:
  - Signed an association agreement in February 1996, which entered into force in March 2000.
- Agreements aimed at gradual creation of a free trade area for industrial products, liberalization of services and capital, and harmonization of regulatory frameworks.
- Trade flows increased exponentially since implementation of association agreements.
- Exports, tourism, and remittances are sizable in both countries but vary in composition.

### External receipts and partner dependence (exact figures preserved)
- Shares of Exports, Tourism, and Remittances (2008, in percent of GDP):
  - Morocco: Exports 22.4; Tourism 8.0; Remittances 8.7.
  - Tunisia: Exports 42.7; Tourism 6.1; Remittances 4.4.
- EU shares in Exports, Tourism, and Remittances (average 2005–08, in percent):
  - Morocco: Exports EU 70; Tourism EU 79; Remittances EU 87.
  - Tunisia: Exports EU 77; Tourism EU 85; Remittances EU 88.
- EU average share in Moroccan exports for 2005–08: 70 percent.
- EU average share in Tunisian exports for 2005–08: 77 percent.
- Tourism and remittances originating from the EU: average shares between 80 percent and 90 percent (2005–08), and relatively stable over time.

### FDI and other external channels (exact figures preserved)
- FDI in percent of GDP (average 2005–08):
  - Morocco 4.9; Tunisia 6.2.
- EU share in total FDI (panel values):
  - Morocco EU 80; Tunisia EU 58.
- Interpretation: FDI is a potentially important transmission channel but tends to operate mainly in the medium to long term rather than the business-cycle horizon.

### Financial linkages and banking sector exposure (exact figures preserved)
- Banking activities funded with domestic deposits:
  - Morocco 75 percent of banks’ resources; Tunisia 80 percent of banks’ resources.
- Banks’ external liabilities to BIS banks (selected 2009 values, in percent of GDP):
  - Morocco short-term liabilities to BIS banks (2009): 3.3.
  - Tunisia short-term liabilities to BIS banks (2009): 1.2.
  - Overall liabilities to BIS banks (2009): Morocco 9.8; Tunisia 4.8.
- Foreign bank presence (share of total banking assets):
  - Morocco 21.4 percent (2008); Tunisia 29.4 percent (2009).
- Capital account openness: only partially open; several restrictions on foreign currency transactions, foreign borrowing and foreign investment by domestic firms.

### Data, methodology, and structure
- Data: real nonagricultural year-on-year growth rates of GDP series (seasonally adjusted, quarterly) for Morocco (MORNA_t_g) and Tunisia (TUNNA_t_g).
- EU composite: country-specific composite GDP growth rate (EU_t_gγ) calculated as weighted average of growth rates of main EU partners based on export weights from IMF GEE.
- Empirical approach: SVARs with robustness checks using alternative identification schemes; decomposition of contributions of trade, tourism, and remittances following Bayoumi and Swiston (2009).
- Paper organization:
  - Section II: background and data.
  - Section III: SVAR estimation and robustness.
  - Section IV: transmission channels.
  - Section V: sectoral SVARs.
  - Section VI: policy issues given European growth forecasts.
  - Section VII: conclusions.

### Correlation statistics with European partners (exact figures preserved)
- Sample period for correlations: 2000Q1–2009Q4.
- Standard deviations (diagonal):
  - MORNA_t_g = 1.69; (EU_t_gMOR) = 2.08; TUNNA_t_g = 1.59; (EU_t_gTUN) = 2.05; (t_gOil) = 14.60.
- Correlations (off-diagonal, with T-stats in brackets):
  - Correlation between (EU_t_gMOR) and MORNA_t_g: 0.46 (3.21)
  - Correlation between (EU_t_gTUN) and TUNNA_t_g: 0.63 (5.06)
  - Correlation between (t_gOil) and MORNA_t_g: 0.16 (0.98)
  - Correlation between (t_gOil) and (EU_t_gMOR): -0.07 (-0.43)
- Table 2 conclusion: correlation is 46 percent for Morocco and 63 percent for Tunisia; statistical tests reject the hypothesis that these correlations are equal to zero.
- No strong correlation between growth in real petroleum spot prices (t_gOil) and Moroccan or Tunisian growth rates.

### Estimating the spillovers — Identification and impulse responses
- Method: VAR-based identification using reduced-form VAR and structural mapping with Choleski inverse of variance-covariance matrix; assumption that structural shocks in Tunisia have zero contemporaneous effect on EU shocks.
- Sample periods:
  - Morocco: 1990Q1 to 2009Q4.
  - Tunisia: 2000Q1 to 2009Q4.
- Two lags included in VAR (optimal by Schwarz information criteria).
- Impulse Response Function (IRF) findings:
  - The impulse response of both Morocco and Tunisia GDP growth to EU GDP growth is significantly different from zero in the first four quarters.
  - A variation of one percentage point in EU GDP growth produces peak responses:
    - Morocco: 0.59 in the second quarter.
    - Tunisia: 0.62 in the third quarter.
  - Morocco response: more abrupt and short lived.
  - Tunisia response: more persistent.
- Robustness:
  - Reversing order of growth rates produced similar results.
  - Growth shocks in Morocco and Tunisia generate nonsignificant responses on growth in the EU partner countries.

### Variance decomposition (share of country GDP growth forecast error variance explained by (EU_t_g·))
- Morocco:
  - Quarter 1: 0.49%
  - Quarter 2: 6.01%
  - Quarter 3: 8.12%
  - Quarter 4: 9.91%
  - Quarter 5: 10.51%
  - Quarter 6: 10.71%
  - Quarter 7: 10.74%
  - Quarter 8: 10.74%
  - Quarter 9: 10.74%
  - Quarter 10: 10.75%
- Tunisia:
  - Quarter 1: 3.78%
  - Quarter 2: 17.20%
  - Quarter 3: 28.94%
  - Quarter 4: 34.55%
  - Quarter 5: 36.71%
  - Quarter 6: 37.28%
  - Quarter 7: 37.31%
  - Quarter 8: 37.36%
  - Quarter 9: 37.58%
  - Quarter 10: 37.90%
- Interpretation: At a 10-quarter horizon, shocks to (EU_t_gMOR) explain 11 percent of Morocco’s GDP growth whereas shocks to (EU_t_gTUN) explain 38 percent of Tunisia’s GDP growth.

### Alternative identification (long-run restriction)
- Long-run identification assumes growth in Morocco or Tunisia has no long-run impact on GDP in partner countries (Blanchard and Quah, 1989).
- IRFs with long-run restriction resemble those from the short-run identification scheme.
- IRFs to one standard deviation shock of the reduced-form residual (no identifying restriction) suggest short-run and long-run restrictions do not impose big transformations.

### Decomposing the spillovers by transmission channels (Bayoumi and Swiston, 2009 approach)
- Method: SVARs augmented one at a time with trade, tourism and remittances as exogenous variables; contribution of channel k calculated as ck = r − rk (r = response of growth to partner growth; rk = response when channel k included as exogenous).
- Caveats:
  - Does not exclude other relevant channels.
  - Does not account for collinearity between channels.
  - Contributions do not necessarily sum to the impulse response r.
- Findings (cumulative contributions):
  - Tunisia:
    - Major transmission channels: exports, and to a much lesser extent tourism.
    - Remittances play a very limited role at these frequencies.
  - Morocco:
    - Exports play an important role.
    - Remittances and tourism are also major transmission channels.
    - Cross-correlations between the three flows are significant in Morocco.
- Context during the EU downturn:
  - In Morocco all three variables (exports, tourism, remittances) strongly responded to the sharp drop in EU GDP.
  - In Tunisia, exports were the most affected; tourism receipts were less affected; remittances showed strong resilience.
- Morocco phosphate specifics:
  - Phosphate share in total exports:
    - 2007: 8 percent
    - 2008: 14 percent
    - 2009: 7 percent
  - Phosphate quantities:
    - Decreased 16 percent in 2008 and 50 percent in 2009.
  - Phosphate prices:
    - More than tripled in 2008 and fell 41 percent in 2009.

### Sectoral analysis (SVARs with sectoral value-added growth)
- Method: replace nonagricultural GDP growth with sectoral GDP growth rates (jSECTOR_t_g) to measure sector reactions to country-specific EU GDP composite.
- Morocco results:
  - Service-related sectors are more responsive to the European cycle.
  - Responsive sectors: Commerce, Other Services, Transports & Telecom.
  - Non-responsive sectors: Manufacturing, Agriculture, Mining, Construction, Energy.
- Tunisia results:
  - Export-oriented sectors are most responsive.
  - Responsive sectors: Manufacturing, Energy, Transports & Telecom, Construction.
  - Non-responsive sectors: Agriculture, Mining, Other services.

### Policy implications and projected changes
- The EU plays a key role for both Tunisia and Morocco; analysis confirms robustness of this connection and relative importance of transmission channels.
- Euro Area context:
  - Real output dropped by more than 4 percent in the Euro Area in 2009.
  - Current forecasts (at time of analysis) indicate a modest and slow rebound in the EU area; growth in domestic demand and real GDP will stay below pre-crisis levels in the years to come.
  - Recent concerns about sovereign debt and sustainability of public finances in Southern European countries add downside risks.
- Projected changes:
  - Engines of growth in Europe expected to change, with less contribution from domestic demand.
  - Growth in European imports (main source of external demand for Morocco and Tunisia) projected to be less than half compared to the pre-crisis period in 2010 and 2011.
- Market share observations:
  - Morocco and Tunisia market shares in EU merchandise imports (EU (27 countries), percent) series 1999–2008:
    - Morocco: 0.8, 0.6, 0.7, 0.7, 0.7, 0.6, 0.8, 0.5, 0.6, 0.5
    - Tunisia: 0.7, 0.6, 0.6, 0.7, 0.7, 0.7, 0.6, 0.6, 0.6, 0.6
  - Market shares have not increased since 1999; remained fairly stable or slightly decreased.
  - Factors: increased liberalization under WTO, competition from fast-growing exporters such as China and Eastern Europe.
- Domestic resilience and challenge:
  - Morocco and Tunisia weathered recent slowdown due to relatively insulated financial systems, exceptional agricultural production and strong domestic demand in 2009.
  - Growth accounting suggests domestic demand became a more dynamic source of growth from 2008 onwards.
  - Challenge: enhancing competitiveness and further diversification of trade flows essential to sustain future growth.
  - Trade with other Maghreb countries is marginal and represents unexploited potential given subdued European growth projections.

### Conclusion
- Both Morocco and Tunisia nonagricultural GDP growth rates are significantly influenced by variations in GDP of their main European partners.
- Transmission channels differ:
  - Tunisia: growth shocks mainly transmitted through exports and, to a much lesser extent, tourism.
  - Morocco: exports, tourism receipts and remittances play equally important roles.
- Sectoral results:
  - Morocco: service-related sectors most responsive to EU GDP.
  - Tunisia: most responsive sectors are export-oriented.
- Limitations:
  - VAR analysis indicates sensitivity to external shocks, not a growth accounting decomposition.
  - Break-down by transmission channel is tentative; other channels (FDI, imports, financial sector linkages) warrant further study.
- Policy takeaway: Given the important role of the EU, enhancing competitiveness and diversifying trade flows is essential for sustaining future growth.

### Appendix — Trade weights from the Global Economic Environment (GEE)
- GEE and weight construction:
  - GEE data are group composites based on weighted averages of selected World Economic Outlook (WEO) variables.
  - Trade weights derived from IMF’s Direction of Trade (DOT) database.
  - Every country has four sets of trade weights: imports and exports weights for all partners, and imports and exports weights for only advanced economies.
  - Weights calculated from the reporter’s trade with partners comprising 95 percent of total exports or imports; the remaining 5 percent not included.
  - GEE constructs a reporting country's series from aggregated sum of weighted percentage changes of its partners' data for a given WEO series.
- GEE weights used to calculate country-specific composite EU growth rates (EU_t_gγ) (EU countries’ weights normalized to sum to 100):
  - Morocco weights:
    - France 36.36
    - Spain 33.42
    - United Kingdom 8.22
    - Italy 7.87
    - Belgium 5.33
    - Germany 5.20
    - Netherlands 3.60
    - Total 100
  - Tunisia weights:
    - France 42.2
    - Italy 29.1
    - Germany 10.9
    - Spain 7.6
    - Belgium 3.7
    - United Kingdom 3.5
    - Netherlands 2.9
    - Total 100
- Composite GDP growth rate formula (based on export weights αi and country growth rates gi):
  - (EU_t_gγ) = Σ i αi gi  (weights normalized so sum of αi = 1).

*Source: Country’s authorities; and IMF staff estimates.*

### References .............................................................................................................

### _wp10238 - References

### Introduction and main findings
- Paper examines effect of growth shocks in European countries on economic activity in Morocco and Tunisia.
- Methodology: structural vector autoregressions (SVARs) to measure response of Tunisian and Moroccan growth to an innovation in the growth rate of key partner countries, following Bayoumi and Swiston (2009).
- Key empirical findings:
  - Positive shocks to a growth composite of trade partners have significant expansionary hump-shaped effects on growth in Morocco and Tunisia.
  - Transmission channels differ: in Tunisia, exports (and to a much lesser extent tourism) are the major channels; in Morocco, exports, remittances, and tourism all play important roles.
  - Sectoral VARs: export-oriented sectors in Tunisia respond significantly to external growth shocks; service-related sectors in Morocco are most sensitive.
  - Results are consistent with the evolution of macroeconomic aggregates during the 2008 recession in Europe.

### Background: trade liberalization and association agreements
- Association Agreements:
  - Tunisia signed an association agreement with the EU in July 1995 and implemented it in March 1998; an amendment covering agricultural products was implemented on January 1, 2001.
  - Morocco signed an association agreement in February 1996, which entered into force in March 2000.
  - Agreements aimed at gradual creation of a free trade area for industrial products, liberalization of services and capital, and harmonization of regulatory frameworks.
- Trade impacts:
  - Since implementation of association agreements, trade flows increased exponentially (Figure 1).
  - By exports of goods definition: exports, tourism, and remittances are sizable in both countries but vary in composition.

### External receipts and partner dependence (exact figures preserved)
- Shares of Exports, Tourism, and Remittances (2008, in percent of GDP):
  - Morocco: Exports 22.4; Tourism 8.0; Remittances 8.7.
  - Tunisia: Exports 42.7; Tourism 6.1; Remittances 4.4.
- EU shares in Exports, Tourism, and Remittances (average 2005–08, in percent):
  - Morocco: Exports EU 70; Tourism EU 79; Remittances EU 87.
  - Tunisia: Exports EU 77; Tourism EU 85; Remittances EU 88.
- EU as main export destination (text summary):
  - EU average share in Moroccan exports for 2005–08: 70 percent.
  - EU average share in Tunisian exports for 2005–08: 77 percent.
  - Tourism and remittances originating from the EU: average shares between 80 percent and 90 percent (2005–08), and relatively stable over time.

### FDI and other external channels (exact figures preserved)
- FDI importance (average 2005–08):
  - FDI in percent of GDP: Morocco 4.9; Tunisia 6.2.
  - EU share in total FDI (panel values): Morocco EU 80; Tunisia EU 58.
- Interpretation: FDI represents a potentially important transmission channel but tends to operate mainly in the medium to long term rather than the business-cycle horizon.

### Financial linkages and banking sector exposure (exact figures preserved)
- Domestic funding and cross-border financing:
  - Banking activities funded largely with domestic deposits: Morocco 75 percent of banks’ resources; Tunisia 80 percent of banks’ resources.
  - Cross-border financing is small.
- Banks’ external liabilities to BIS banks (selected 2009 values, in percent of GDP):
  - Morocco short-term liabilities to BIS banks (2009): 3.3.
  - Tunisia short-term liabilities to BIS banks (2009): 1.2.
  - Overall liabilities to BIS banks (2009): Morocco 9.8; Tunisia 4.8.
- Foreign bank presence:
  - Majority-owned foreign banks account for 21.4 percent of total banking assets in Morocco (2008) and 29.4 percent in Tunisia (2009).
- Capital account openness: capital account is only partially open; several restrictions apply to transactions in foreign currency, foreign borrowing and foreign investment by domestic firms.

### Data, robustness, and structure of the paper
- Data and empirical approach:
  - Paper describes GDP data, contemporaneous correlations, and estimates spillovers using SVARs with robustness checks using alternative identification schemes.
  - Follows Bayoumi and Swiston approach to decompose contributions of trade, tourism revenues, and remittances to spillovers.
- Paper organization (section overview):
  - Section II: background and data description.
  - Section III: SVAR estimation of spillovers and robustness checks.
  - Section IV: role of transmission channels.
  - Section V: SVARs with sectoral data.
  - Section VI: policy issues given European growth forecasts.
  - Section VII: conclusions.

*Source: Country’s authorities; and IMF staff estimates.*

### Box 2. Some of the Remaining Restrictions on Foreign Currency

### Box 2. Some of the Remaining Restrictions on Foreign Currency Transactions and Capital flows

### (a) Morocco
- Moroccan exporters of goods and services may open foreign exchange accounts domestically and deposit up to 50 percent of the repatriated foreign exchange receipts from exports.
- Foreign residents and corporations may open foreign exchange accounts without limitations.
- Commercial banks may borrow abroad only to finance foreign trade or investment operations on behalf of customers or as part of the introduction of the forward cover for customers’ foreign exchange swaps, with maturities equal to the duration of the hedging operations offered.
- There are no controls on inward direct investment.
- Outward direct investments are subject to approval, but resident corporations in operation for more than three years benefit from certain facilities.
- Resident foreign nationals are free to invest abroad, provided the operations are financed from their own funds abroad or from their holdings denominated in convertible dirham or foreign exchange.
- A number of measures were taken during 2009 to further liberalize the capital account.

### (b) Tunisia
- Professional accounts in foreign currency may be opened by resident individuals, Tunisian legal entity and foreign legal entity in Tunisia with foreign currency assets in connection with its activities.
- Authorization by the Central Bank of Tunisia (BCT) is required for residents to open accounts abroad.
- Export proceeds must be repatriated within a time limit from the payment due date. However, resident exporters may retain up to 100 percent of their foreign exchange export proceeds.
- The acquisition of assets (capital and money market instruments) abroad by residents is subject to authorization.
- Residents are required to repatriate all income from investment abroad, as well as all proceeds from the divestiture or liquidation of such investments.
- Stocks in existing companies in Tunisia may be acquired freely with foreign exchange transferred from abroad by foreign nonresidents.
- The acquisition by foreigners of shares with voting rights is subject to the approval of the High Investment Commission.
- Nonresidents may sell freely shares of companies established in Tunisia and transfer freely net real proceeds from the sale of shares that were purchased with foreign exchange transferred from abroad.

*Source: International Monetary Fund, 2009, Annual Report on Exchange Arrangements and Exchange Restrictions (AREAER).*

### B. Correlation of Growth Rates with European Partners
- Growth rates of Morocco and Tunisia and their key European partners are strongly positively correlated in the last decade.
- The econometric analysis uses real nonagricultural year-on-year growth rates of GDP series for Morocco ,MORNA t g and Tunisia ,TUNNA t g. These series are seasonally adjusted and available on a quarterly basis.
- For the EU, a country-specific composite GDP growth rate ()EU t gγ is calculated as a weighted average of growth rates of the main EU partners based on the weights these countries take in Morocco’s and Tunisia’s exports of goods respectively.
- The country weights are constructed by the IMF for the Global Economic Environment (GEE).

Key statistics (2000Q1–2009Q4):
- Standard deviations (diagonal, in table): MORNA t g = 1.69; ()EU t gMOR = 2.08; TUNNA t g = 1.59; ()EU t gTUN = 2.05; ()t gOil = 14.60.
- Correlations (off-diagonal, with T-stats in brackets):
  - Correlation between ()EU t gMOR and MORNA t g: 0.46 (3.21)
  - Correlation between ()EU t gTUN and TUNNA t g: 0.63 (5.06)
  - Correlation between ()t gOil and MORNA t g: 0.16 (0.98)
  - Correlation between ()t gOil and ()EU t gMOR: -0.07 (-0.43)
- Table 2 conclusion: correlation is 46 percent for Morocco and 63 percent for Tunisia; statistical tests reject the hypothesis that these correlations are equal to zero.
- No strong correlation between growth in real petroleum spot prices, ()t gOil, and Moroccan or Tunisian growth rates, suggesting correlations with European partners are not hiding a global cycle.

Notes:
- Agricultural output excluded because it remains largely dependent on climatic factors.
- Similar numbers found when calculating correlation with eurozone GDP reported by Eurostat.

### III. Estimating the Spillovers — Identification Scheme
- Method: VAR-based identification scheme to estimate impact of an innovation in EU growth on Morocco and Tunisia nonagricultural GDP growth.
- Baseline VAR variables for country j: ,(),'. EUjNA ttt Zgjg = [ ].
- Reduced-form VAR: ,)'(, 1 VuuEuZBcZ p j tttjtjt     where c is a constant and t Z is the matrix with variables.
- Residuals mapping: t u are mapped into structural shocks t ε by structural matrix A0: 0. tt Auε
- Identification: inverse of the Choleski factor of the variance-covariance matrix V to identify orthogonal structural shocks [ ]IE tt 'εε.
- Assumption: on impact structural shocks in Tunisia have a zero effect on shocks in European partner countries.
- Ordering of variables in Z did not change results.

### III.B Responses to External Growth Shock
- Sample periods:
  - Morocco: 1990Q1 to 2009Q4.
  - Tunisia: 2000Q1 to 2009Q4.
- Two lags included in VAR (optimal by Schwarz information criteria).
- Impulse Response Function (IRF) findings:
  - The impulse response of both Morocco and Tunisia GDP growth to EU GDP growth is significantly different from zero in the first four quarters.
  - A variation of one percentage point in EU GDP growth produces peak responses:
    - Morocco: 0.59 in the second quarter.
    - Tunisia: 0.62 in the third quarter.
  - Morocco response: more abrupt and short lived.
  - Tunisia response: more persistent.
- Robustness checks:
  - Reversing order of growth rates in VAR produced similar results.
  - Growth shocks in Morocco and Tunisia generate nonsignificant responses on growth in the EU partner countries.

Variance decomposition (share of country GDP growth forecast error variance explained by ()EU t g·):
- Morocco:
  - Quarter 1: 0.49%
  - Quarter 2: 6.01%
  - Quarter 3: 8.12%
  - Quarter 4: 9.91%
  - Quarter 5: 10.51%
  - Quarter 6: 10.71%
  - Quarter 7: 10.74%
  - Quarter 8: 10.74%
  - Quarter 9: 10.74%
  - Quarter 10: 10.75%
- Tunisia:
  - Quarter 1: 3.78%
  - Quarter 2: 17.20%
  - Quarter 3: 28.94%
  - Quarter 4: 34.55%
  - Quarter 5: 36.71%
  - Quarter 6: 37.28%
  - Quarter 7: 37.31%
  - Quarter 8: 37.36%
  - Quarter 9: 37.58%
  - Quarter 10: 37.90%

- At a 10-quarter horizon, shocks to ()EU t gMOR explain 11 percent of Morocco’s GDP growth whereas shocks to ()EU t gTUN explain 38 percent of Tunisia’s GDP growth.

### III.C Alternative Identification of Shocks (Long-run restriction)
- Long-run identification assumes growth in Morocco or Tunisia has no long-run impact on GDP in partner countries (Blanchard and Quah, 1989).
- Variables ordered as before.
- Long-run effects: t Zε = Θ.
- Identification restriction: (1, :)[ (1, 1)0 ]Θ = Θ.
- IRFs with long-run restriction resemble those from the short-run identification scheme.
- IRFs to one standard deviation shock of the reduced-form residual (no identifying restriction) suggest short-run and long-run restrictions do not impose big transformations.

### IV. Decomposing the Spillovers by Transmission Channels
- Method: Bayoumi and Swiston (2009) approach.
- SVARs augmented one at a time with data on trade, tourism and remittances as exogenous variables.
- Contribution of channel k (ck) calculated as ck = r − rk, where r is response of growth to partner growth and rk is response when channel k included as exogenous.
- Caveats:
  - Does not exclude other relevant channels.
  - Does not account for collinearity between channels.
  - Contributions do not necessarily sum to the impulse response r.

Findings (cumulative contributions):
- Tunisia:
  - Major transmission channels: exports, and to a much lesser extent tourism.
  - Remittances play a very limited role at these frequencies.
- Morocco:
  - Exports play an important role.
  - Remittances and tourism are also major transmission channels.
  - Cross-correlations between the three flows are significant in Morocco.

Contextual observations:
- In Morocco all three variables (exports, tourism, remittances) strongly responded to the sharp drop in EU GDP.
- In Tunisia, exports were the most affected; tourism receipts were less affected; remittances showed strong resilience.

Additional note on Morocco exports:
- Phosphate share in total exports:
  - 2007: 8 percent
  - 2008: 14 percent
  - 2009: 7 percent
- Phosphate quantities:
  - Decreased 16 percent in 2008 and 50 percent in 2009.
- Phosphate prices:
  - More than tripled in 2008 and fell 41 percent in 2009.

### V. Sectoral Analysis
- Approach: re-run SVAR replacing nonagricultural GDP growth rates with sectoral GDP growth rates ,jSECTOR t g.
- Impulse response measures reaction of sector value added growth to country-specific EU GDP composite.

Results:
- Morocco:
  - Service-related sectors (sensitive to tourism and transfers) are more responsive to the European cycle.
  - Non-responsive sectors: Manufacturing, Agriculture, Mining, Construction, Energy.
  - Responsive sectors highlighted: Commerce, Other Services, Transports & Telecom.
- Tunisia:
  - Export-oriented sectors (Manufacturing and Energy) are most responsive.
  - Non-responsive sectors: Agriculture, Mining, Other services.
  - Other responsive sectors: Transports & Telecom, Construction.

### VI. Policy Implications
- The EU plays a key role for both Tunisia and Morocco; the analysis confirms robustness of this connection and relative importance of transmission channels.
- Close ties with the EU have contributed to opening up the economies and improving competitiveness but may pose future challenges.
- Euro Area context:
  - Real output dropped by more than 4 percent in the Euro Area in 2009.
  - Current forecasts (at time of analysis) indicate a modest and slow rebound in the EU area; growth in domestic demand and real GDP will stay below pre-crisis levels in the years to come.
  - Recent concerns about sovereign debt and sustainability of public finances in Southern European countries add downside risks.
- Projected changes:
  - Engines of growth in Europe expected to change, with less contribution from domestic demand.
  - Growth in European imports (main source of external demand for Morocco and Tunisia) projected to be less than half compared to the pre-crisis period in 2010 and 2011.
- Market share observations:
  - Following association agreements, market shares of Morocco and Tunisia in EU merchandise imports have not increased since 1999; market shares remained fairly stable or slightly decreased.
  - Factors: increased liberalization under WTO, competition from fast-growing exporters such as China and Eastern Europe.
- Table 4: Morocco and Tunisia Shares in EU Imports (EU (27 countries), percent)
  - Morocco series (1999–2008): 0.8, 0.6, 0.7, 0.7, 0.7, 0.6, 0.8, 0.5, 0.6, 0.5
  - Tunisia series (1999–2008): 0.7, 0.6, 0.6, 0.7, 0.7, 0.7, 0.6, 0.6, 0.6, 0.6
- Morocco and Tunisia weathered recent slowdown due to relatively insulated financial systems, exceptional agricultural production and strong domestic demand in 2009.
- Growth accounting suggests domestic demand became a more dynamic source of growth from 2008 onwards.
- Challenge: enhancing competitiveness and further diversification of trade flows essential to sustain future growth.
- Trade with other Maghreb countries is marginal for both Morocco and Tunisia and represents unexploited potential in the context of subdued European growth projections.

### VII. Conclusion
- Both Morocco and Tunisia nonagricultural GDP growth rates are significantly influenced by variations in GDP of their main European partners.
- Transmission channels differ:
  - Tunisia: growth shocks mainly transmitted through exports and, to a much lesser extent, tourism.
  - Morocco: exports, tourism receipts and remittances play equally important roles.
- Sectoral results:
  - Morocco: service-related sectors most responsive to EU GDP.
  - Tunisia: most responsive sectors are export-oriented.
- Limitations:
  - VAR analysis shows sensitivity to external shocks, not a growth accounting decomposition.
  - Break-down by transmission channel is tentative; other channels (FDI, imports, financial sector linkages) warrant further study.
- Policy takeaway: Given the important role of the EU, enhancing competitiveness and diversifying trade flows is essential for sustaining future growth.

### Appendix. Trade Weights from the Global Economic Environment (GEE)
- GEE data are group composites based on weighted averages of selected World Economic Outlook (WEO) variables.
- Trade weights derived from IMF’s Direction of Trade (DOT) database.
- Every country has four sets of trade weights: imports and exports weights for all partners, and imports and exports weights for only advanced economies.
- Weights calculated from the reporter’s trade with partners comprising 95 percent of total exports or imports; the remaining 5 percent not included.
- GEE constructs a reporting country's series from aggregated sum of weighted percentage changes of its partners' data for a given WEO series.

GEE weights used to calculate country-specific composite EU growth rates ()EU t gγ (EU countries’ weights normalized to sum to 100):

- Morocco weights:
  - France 36.36
  - Spain 33.42
  - United Kingdom 8.22
  - Italy 7.87
  - Belgium 5.33
  - Germany 5.20
  - Netherlands 3.60
  - Total 100

- Tunisia weights:
  - France 42.2
  - Italy 29.1
  - Germany 10.9
  - Spain 7.6
  - Belgium 3.7
  - United Kingdom 3.5
  - Netherlands 2.9
  - Total 100

- Composite GDP growth rate formula (based on export weights αi and country growth rates gi):
  - ( ) EU t gγ = Σ i αi gi  (weights normalized so sum of αi = 1). 

*Source: IMF staff estimates and GEE trade-weight methodology (Direction of Trade database).*

### References

### _wp10238 - References

### References listed
- Arezki, R., R. Cherif, and J. Piotrowski, 2009, “Tourism Specialization and Economic 
Development: Evidence from the UNESCO World Heritage List,” IMF Working Paper 
09/176 (Washington: International Monetary Fund).

- Baxter, M., and M. Kouparitsas, 2005, “Determinants of Business Cycle Comovement: a Robust 
Analysis,” Journal of Monetary Economics, Vol. 52(1), January, pp.113–57.

- Bayoumi, T. and A. Swiston, 2009, “Foreign Entanglements; Estimating the Source and Size of 
Spillovers Across Industrial Countries,” IMF Staff Papers, Vol. 56(2), pp. 353–83.

- Blanchard, O. and D. Quah (1989), “The Dynamic Effects of Aggregate Demand and Supply 
Disturbances,” American Economic Review, Vol. 79(4), pp. 655–73.

- Canova, F., and H. Dellas, 1993, “Trade Interdependence and the International Business Cycle,” 
Journal of International Economics, Vol. 34 (February), pp. 23–47.

- Gapen M. T., A. and others, 2009, “Do Workers’ Remittances Promote Economic Growth?”
 IMF 
Working Paper 09/153 (Washington: International Monetary Fund).

- Imbs, Jean (2004), “Trade, Finance, Specialization, and Synchronization,” Review of Economics 
and Statistics, Vol. 863, pp.723–34.

- Kose, M.A., C. Otrok, and C.H. Whiteman, 2003, “International Business Cycles: World, 
Region, and Country-specific Factors,” American Economic Review, Vol.93 (4), pp. 1216–
39.

- Toujas-Bernaté, J., B. Loko, and D. Simard, 2009, “Investigating Growth Spillovers from 
Europe,” Tunisia Selected Issues Paper, IMF Country Report  
No. 10/109.

*Source: _wp10238 - References*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp10238.pdf_