## _wp1335

## Source details

**Canonical URL:** [_wp1335](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp1335.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp1335.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp1335.pdf.json)

---

### I. Introduction: objectives and approach
- Objective: examine the role of competitiveness in promoting strong growth and reducing imbalances for Ireland.
- Empirical approach:
  - Impulse response analysis and variance decomposition from an unrestricted VAR model.
  - Expanded VAR includes real effective exchange rate (REER) measures, external demand, and foreign direct investment (FDI).
  - Data: quarterly data covering 1980q1 to 2009q4.
- Key emerging conclusions:
  - Trading partners’ demand is the single largest driver of economic growth and government revenue.
  - REER appreciations constrain total exports.
  - FDI flows have a positive short-lived impact on exports and growth.
  - Exports boost GDP growth, which boosts revenue and reduces the debt burden.

### II. Debt, fiscal pressures, and macro constraints
- Ireland exemplifies the nexus between high debt and low growth; public debt levels post-2008/09 near or above thresholds characterized as harmful to growth.
- Financing and interest burdens:
  - Spread of a 10-year Irish government bond yield over the corresponding bund: almost nil during 2003−07; 3 percentage points at end-2010; almost 8 percentage points by end-2011.
  - Interest payments on public debt: €6 billion and accounted for 11 percent of total government revenue in 2011; 2⅓ percent of revenue in 2007.
  - Projected interest payments: annual average of €9¾ billion or 15½ percent of revenue during 2013–16.
- Implication: interest payments together with scheduled amortization leave little or no room for growth-enhancing investment spending.

### III. Pre-crisis reforms, competitiveness, and outcomes (1981–2007)
- Preconditions and reform actions:
  - Turnaround began near 1987 after stagnation; structural reforms and EU structural grants financed infrastructure and improved the socioeconomic environment.
  - Key reforms: spending cuts, tax amnesty, deregulation (telecommunication, road infrastructure, air travel), lowering corporate profit tax rates, and reorientation of tertiary education toward engineering, science, and information technology.
- Outcomes for trade, FDI, and productivity:
  - Falling domestic costs and REER depreciation helped attract FDI and boost investment.
  - Growth of real exports of goods and services (G&S): averaged 15 percent per year during 1994–2002 (vs about half as much during 1981–86).
  - Real GDP annual average growth, 1988–2007: 6 percent.
  - Unemployment: 17 percent (mid-1980s) to 4½ percent (2007).
  - Public debt fell from almost 110 percent (1987) to 25 percent (2007).
  - NIIP deficit: just over 20 percent of GDP (2007).
- Change in export composition:
  - Share of low-tech products in total exports declined to 14 percent in 2000–01 from almost 50 percent during 1985–89.
  - Share of high-tech exports almost doubled to 46 percent in 2000–01 (comparative: Greece 12 percent; Euro area average 21 percent).

### IV. Pre-crisis vulnerabilities and the disruption (2003–11)
- Demand- and supply-side shifts:
  - Supply: during 2003–07 average productivity growth fell as activity shifted to labor-intensive construction and services.
  - Demand: investment in buildings outpaced equipment; domestic demand share of GDP increased from 84 percent in 2002 to 91 percent in 2007.
- Fiscal and external deterioration:
  - Cyclically-adjusted balance deteriorated from 2¾ percent of GDP deficit in 2002 to over 7 percent of GDP deficit in 2007.
  - Current account deteriorated from balance in 2003 to a deficit of 5½ percent of GDP by 2007.
- FDI and competitiveness trends:
  - Stock of net FDI: almost 100 percent of GDP in 2002 → less than 20 percent of GDP by end-2007.
  - Over 2003–07: ULC-based REER appreciated by 10 percent; CPI-based REER appreciated by 32½ percent.
  - REER appreciations weakened total exports in the run-up to the crisis.
- Crisis outcomes:
  - End of housing boom early 2007 → contraction in housing sector, rising unemployment, deterioration in bank loan portfolios.
  - Exports-to-GDP ratio: 80 percent (2007).
  - Cumulative GDP growth 2008–11: almost negative 7 percentage points.
  - Public debt-to-GDP and NIIP deficit-to-GDP by end-2011: hovered 100 percent (from 25 percent and 21 percent, respectively, at end-2007).

### V. Econometric design and data (VAR framework)
- Variables (sample 1980Q1–2009Q4): GDP, exports, government revenue (real terms), REER (CPI-based and ULC-based), inward FDI (in 2005 Euros), trading partners’ demand (partners accounting for 95 percent of exports).
- VAR specifications:
  - VAR1 (baseline): DLNEXPGS, GRW, DLNREV; constant and EMU dummy (EMU = 1 from 1991Q1 onward); lag selection: 2 (AIC = 13.73; SC = 14.31*; HQ = 13.96*).
  - VAR2 (expanded): adds DLNMGS_TP, DINFDIR, DLNREER (ULC or CPI); lag selection varies (ULC specification chosen at 4 lags; CPI at 2 lags).
- Correlation and Granger-causality highlights:
  - LNREER_ULC with LNEXPGS: -0.92*** (p = 0.000).
  - LNMGS_TP with LNEXPGS: 1.00 (p = 0.000).
  - Selected Granger-causality: DINFDIR does not Granger Cause GRW — F-statistic: 3.56; P-value: 0.01. GRW does not Granger Cause DLNREV — F-statistic: 3.26; P-value: 0.01.

### VI. VAR1 key empirical results (exports → GDP → revenue)
- Impulse responses (one standard deviation shocks; variables in first differences):
  - Exports shock (equivalent to 6½ percent) → GDP: 1 percentage point after one quarter; cumulative 7 percent by quarter 8; just over 8 percent by quarter 16.
  - GDP shock (corresponding to 4 percent) → revenue: cumulative 2 percent after four quarters; peaks at 5 percent after eight quarters; ~3½ percent by quarter 16.
- FEVD (selected):
  - A shock to exports explains 34½ percent of the variation in GDP after one quarter; peaks at 42¾ percent after three quarters; ~39 percent after eight and 16 quarters.
  - Revenue’s variance explained predominantly by its own innovations: 98¾ percent after one quarter; 82¾ percent after eight quarters; 80¼ percent after 16 quarters.
  - GDP explains an increasing share of revenue variance over time (rises to 14 percent after 16 quarters).

### VII. VAR2 expanded-model findings (external demand, REER, FDI)
- Trading partners’ demand:
  - Inclusion of trading partners’ demand reduces the explanatory shares of exports for GDP and of GDP for revenue substantially relative to VAR1.
  - Trading partners’ demand becomes the single most important determinant of GDP and government revenue in VAR2.
  - Trading partners’ demand shock → exports: ⅓ percent (Q1); 4½ percent (Q4); 7 percent (Q8).
  - Corresponding GDP impacts: ⅓ percent (Q1); 3 percent (Q4); 6⅓ percent (Q8); 7¾ percent (Q16).
  - Revenue responses: ~0 (Q1); 3 percent (Q4); 4½ percent (Q8); 5⅓ percent (Q16).
- REER (competitiveness) shocks and elasticities:
  - ULC-based REER one SD shock (equivalent to 6¼ percent appreciation) → reduces GDP by almost 2 percent after one year.
  - CPI-based REER one SD shock (equivalent to 5 percent appreciation) → reduces GDP by marginally not significant 1 percent after one year.
  - ULC-based REER one SD shock → export decline: 1½ percent after one quarter; over 6 percent after eight quarters.
  - CPI-based REER one SD shock → export decline: ¾ percent after one quarter; 5½ percent after eight quarters.
  - Elasticities: a one percentage point increase in the ULC-based REER reduces exports of G&S by 0.6 percentage points; a one percentage point increase in the CPI-based REER reduces exports of G&S by 1 percentage point.
- Inward FDI shocks:
  - Boost exports by ¾ percent after one quarter, ~1 percent by the second quarter; effects are short-lived and lose statistical significance thereafter.
  - Increase GDP by about ⅓ percent one quarter ahead; thereafter impacts are not statistically significant.

### VIII. Variance decomposition insights (VAR2)
- Exports variance:
  - Own innovations explain ~69¼ percent (Q1), 40 percent (Q4), and 32 percent (Q16).
  - REER explains up to 29¼ percent in the first four quarters; from five quarters out trading partners’ demand becomes the second most important contributor (27 percent around Q5; 22 percent by Q16).
- GDP variance:
  - Own innovations: 81 percent (Q1), 36¼ percent (Q4), 19 percent (Q16).
  - Trading partners’ demand: 6½ percent (Q1) rising to 29 percent (Q4) and peak 34¾ percent (Q6); from Q5 it often dominates GDP’s own innovations.
  - Revenue’s share peaks at 25½ percent after five quarters; 21½ percent at Q16.
- Revenue variance:
  - Own innovations: 97 percent (Q1) decaying to 67 percent (Q8) and 64¾ percent (Q16).
  - Trading partners’ demand becomes significant after three quarters (13½ percent), peaks at 19¼ percent (Q5), and ~17¾ percent (Q16).
- Conclusion: government revenue depends more on components exogenous to the VAR2 system than GDP does (own innovations account for ~68 percent of revenue variance vs ~16½ percent for GDP at Q16).

### IX. Prospects, sensitivity analysis, and policy implications
- External demand projections and scenario sensitivity (illustrative):
  - Trading partners’ import demand projected to average 4¼ percent during 2013–17 (October 2012 WEO baseline used illustratively).
  - If trading partners’ demand grows at 6 percent (average of the five years leading into the crisis), Ireland’s odds of reducing public debt improve through faster GDP and revenue growth.
  - If external demand falls by 5½ percent (average of 2008–09), Ireland’s economic performance would worsen significantly and the country would remain under water.
- Competitiveness post-crisis and implications:
  - Over 2008–11 CPI-based REER depreciated by 10 percent and ULC-based REER depreciated by 28⅓ percent.
  - A 5-percentage point depreciation of the ULC-based REER would boost Ireland’s exports by 2½ percentage points.
  - An appreciation of 7 percent (matching pre-crisis average) would worsen export growth prospects.
- Domestic financing and FDI:
  - Bank lending tight: only 50 percent of business loan applications were successful in 2010 (down from 90 percent in 2007).
  - Net FDI-to-GDP: almost 100 percent (2002) → less than 20 percent (end-2007); net FDI-to-GDP: 4 percent (2010) and 13¾ percent (2011).
  - Shocks to FDI have positive but short-lived impacts on exports and GDP; persistence of FDI requires continuous investor-friendly conditions.
- Policy recommendations and priorities:
  - Boost and maintain price and non-price competitiveness to support export-led recovery.
  - Contain domestic prices and labor costs given inability to use national monetary or nominal exchange rate policy within EMU.
  - Nurture the knowledge economy, innovation, and technology uptake to increase technology content of exports and boost market share.
  - Promote competition and open markets in sheltered sectors; upgrade infrastructure for telephone, broadband communication, and goods distribution.
  - Keep public spending in check so improvements in government revenue can strengthen fiscal balance and reduce public debt.
- Caveats:
  - Analysis includes financing from FDI only, not from the domestic financial sector.
  - Post-crisis factors not included: high household debt, high unemployment, and bank fragility.
  - Public expenditure developments are not included; conclusions depend on fiscal restraint.

### X. Key numeric highlights (preserved exactly)
- ETCR index values: 5.67 in 1985 to 4.95 in 1988 (scale 0 to 6).
- REER movements:
  - ULC-based REER appreciation over 2003–07: 10 percent.
  - CPI-based REER appreciation over 2003–07: 32½ percent.
  - Post-crisis (2008–11) depreciations: CPI-based REER 10 percent; ULC-based REER 28⅓ percent.
- Macro aggregates and rates:
  - Real GDP annual average growth, 1988–2007: 6 percent.
  - Unemployment: 17 percent (mid-1980s) to 4½ percent (2007).
  - Public debt: almost 110 percent (1987) to 25 percent (2007).
  - NIIP deficit: just over 20 percent of GDP (2007).
  - Exports-to-GDP ratio: 80 percent (2007).
  - Cyclically-adjusted deficit: over 7 percent of GDP in 2007 (from 2¾ percent of GDP in 2002).
  - Current account: deficit of 5½ percent of GDP by 2007 (from balance in 2003).
- VAR shocks and elasticities:
  - Exports shock in VAR1: equivalent to 6½ percent → GDP: 1 percentage point (Q1); cumulative 7 percent (Q8); just over 8 percent (Q16).
  - GDP shock in VAR1: corresponding to 4 percent → revenue: 2 percent (Q4); peaks at 5 percent (Q8); ~3½ percent (Q16).
  - ULC-based REER one SD shock: equivalent to 6¼ percent → reduces GDP by almost 2 percent after a year.
  - CPI-based REER one SD shock: equivalent to 5 percent → reduces GDP by ~1 percent after a year (marginally not significant).
  - ULC-based REER elasticity: a one percentage point increase → exports of G&S decline by 0.6 percentage points.
  - CPI-based REER elasticity: a one percentage point increase → exports of G&S decline by 1 percentage point.
  - Trading partners’ demand projected average, 2013–17: 4¼ percent.
  - Net FDI-to-GDP: almost 100 percent (2002) to less than 20 percent (end-2007); net FDI-to-GDP: 4 percent (2010) and 13¾ percent (2011).
  - Cumulative GDP growth 2008–11: almost negative 7 percentage points.
  - Public debt-to-GDP and NIIP deficit-to-GDP by end-2011: hovered 100 percent.

*Source: _wp1335 - References (PDF).*

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

### _wp1335 - References

### I. INTRODUCTION
- Context:
  - The Euro area debt crisis raised concerns about promoting sustainable economic growth amid high debt and financial fragility.
  - Large fiscal and external imbalances in peripheral EMU countries (GIPS: Greece, Ireland, Portugal, and Spain) pose risks to recovery and disturb international financial markets.
  - High gross financing requirements and adverse investor sentiment threaten access to capital needed for investment and growth.

- Paper objective and approach:
  - Examine the role of competitiveness in promoting strong growth and reducing imbalances for Ireland.
  - Investigate linkages among exports, GDP, and government revenue using impulse response analysis and variance decomposition from an unrestricted VAR model.
  - Expand the VAR to include real effective exchange rate (REER) measures, external demand, and foreign direct investment (FDI).
  - Data: quarterly data covering 1980q1 to 2009q4.

- Relationship to prior literature:
  - Similar to Federici and Marconi (2002) for Italy (external demand, REER, exports, GDP).
  - Similar to Feasel, Kim and Smith (2001) for Korea (per capita output, exports, investment).
  - Builds on country-specific studies: Kanda (2008) (competitiveness and external demand among trading partners, but excludes exports) and Bermingham and Conefrey (2011) (external demand broken down among trading partners and includes exports but omits REER or competitiveness indicators).

- Key emerging conclusions (from the paper's analysis):
  - Trading partners’ demand is the single largest driver of economic growth and government revenue.
  - REER appreciations constrain total exports.
  - FDI flows have a positive short-lived impact on exports and growth.
  - Exports boost GDP growth, which boosts revenue and reduces the debt burden.

### II. DEBT AND GROWTH: IRELAND’S EXPERIENCE AND CURRENT STRUGGLE
- Overview:
  - Ireland exemplifies the nexus between high debt and low growth.
  - In the 1980s and early 1990s Ireland experienced high debt and low growth and later grew out of debt.
  - Post-2008/09 public debt levels are near or above thresholds characterized as harmful to growth (see footnote references).

- Debt-related financial constraints:
  - Large debt service liabilities and difficulty obtaining finance at steady low cost due to high risk premiums.
  - Spread of a 10-year Irish government bond yield over the corresponding bund:
    - almost nil during 2003−07
    - 3 percentage points at end-2010
    - almost 8 percentage points by end-2011
  - Interest payments on public debt:
    - €6 billion and accounted for 11 percent of total government revenue in 2011
    - 2⅓ percent of revenue in 2007
    - Projected to rise to an annual average of €9¾ billion or 15½ percent of revenue during 2013–16
  - Implication: interest payments together with scheduled amortization leave little or no room for growth-enhancing investment spending.

- Historical parallels and differences:
  - Post-crisis situation resembles the 1970s and 1980s in public and external debt levels, unemployment, and emigration.
  - Key difference: major structural rigidities of the past no longer exist.

A. Ireland’s Pre-crisis Reforms, Growth, and Public Debt: A Virtuous Cycle
- Preconditions and turning point:
  - Ireland’s economic turnaround began near 1987 after a prolonged period of stagnation with high unemployment, large fiscal and external deficits, and crippling public debt.
  - Structural reforms and EU structural grants financed infrastructure and improved the socioeconomic environment.

- Key reforms and implications:
  - Spending cuts and a tax amnesty that generated windfall revenue, helping fiscal consolidation.
  - Deregulation of telecommunication, road infrastructure, and air travel services promoted competition and lowered costs notably.
  - Lowering of corporate profit tax rates encouraged foreign investment.
  - Expansion and reorientation of state-funded tertiary education towards engineering, science, and information technology helped attract investment in high-tech sectors.

- Outcomes for competitiveness and trade:
  - Public policies enhanced competitiveness, attracted FDI, and boosted external trade.
  - Falling domestic costs underpinned a depreciation of the REER, which, together with incentives for inward FDI, boosted investment.
  - Growth of real exports of goods and services (G&S):
    - averaged 15 percent per year during 1994–2002
    - compared with about half as much during 1981–86
  - Imports grew in tandem with exports, reflecting MNCs’ reliance on imported inputs; nevertheless exports growth outpaced imports growth except in the five-year leading to the 2007/08 crisis.

- FDI and technological competitiveness:
  - Attractiveness to FDI was both a sign of competitiveness and a contributor to technological competitiveness.
  - MNCs concentrated funds in high-tech export sectors.
  - Change in export composition:
    - share of low-tech products in total exports declined to 14 percent in 2000–01 from almost 50 percent during 1985–89
    - share of high-tech exports almost doubled to 46 percent in 2000–01
    - comparative figures: Greece 12 percent (high-tech share) and Euro area average 21 percent (high-tech share) as of the same reference.

- Note on methodology and structure:
  - The remainder of the paper (not detailed here) covers:
    - Section 2: stylized facts about Ireland’s pre-crisis performance and post-crisis imbalances.
    - Section 3: econometric analysis of linkages between exports, output, and government revenue.
    - Section 4: prospects and challenges.
    - Section 5: conclusions and policy implications.

*Source: _wp1335 - References (content excerpt).*

### 5.67 in 1985 to 4.95 in 1988. The scale is 0 to 6, where ―6‖ is for the most regulated environment.

### _wp1335 - 5.67 in 1985 to 4.95 in 1988. The scale is 0 to 6, where ―6‖ is for the most regulated environment.

### Reforms, Productivity, and Macroeconomic Performance (1981–2007)
- Deregulation, reflected in a declining ETCR index, was accompanied by REER depreciation and rising FDI, supporting strong investment and productivity growth.
- Productivity (change in multi-factor productivity, MFP) became the main driver of growth for much of 1987–2002 but subsided subsequently.
- Real GDP expanded by an annual average of 6 percent during 1988 to 2007, reaching double digits during 1995–2000.
- Unemployment declined from 17 percent in the mid-1980s to 4½ percent by 2007.
- Public debt fell from almost 110 percent in 1987 to 25 percent in 2007.
- The deficit of the net international investment position (NIIP) stood at just over 20 percent of GDP in 2007.

### Pre-crisis vulnerabilities and the disruption of Ireland’s economic fabric
- Euro-area policies that contributed to the boom: low interest rates, integrated Euro-based wholesale funding markets, and integrated retail markets that heightened competition in mortgage lending.
- Ireland’s domestic financial regulation and supervision remained relatively lax, allowing financial excesses to build up in household and financial-sector balance sheets.
- Supply-side shift: during 2003–07 average productivity growth fell, reflecting a shift of activity to more labor-intensive real estate-related sectors (construction and services).
- Demand-side shift: investment in buildings outpaced equipment investment. Domestic demand share of GDP increased from 84 percent in 2002 to 91 percent in 2007.
- Fiscal indicators (headline) improved but cyclically-adjusted balance deteriorated from 2¾ percent of GDP deficit in 2002 to over 7 percent of GDP deficit in 2007.
- Current account deteriorated from balance in 2003 to a deficit of 5½ percent of GDP by 2007.
- NIIP composition shifted toward larger banks’ debt liabilities and a much lower net inward FDI position relative to 2002.

### FDI dynamics and competitiveness
- Stock of net FDI represented almost 100 percent of GDP in 2002 and fell to less than 20 percent of GDP by end-2007.
- Decline in net inward FDI occurred across export-contributing sectors, notably computer hardware and software.
- Rising domestic costs reduced Ireland’s attractiveness and contributed to adverse net FDI dynamics.
- Competitiveness indicators:
  - Manufacturing PPI trended down since the early 2000s relative to main trading partners.
  - Over the almost 20-year period from 1987, CPI-based REER appreciated while ULC-based REER depreciated.
  - Over 2003–07 both REER measures appreciated: ULC-based REER appreciated by 10 percent and CPI-based REER appreciated by 32½ percent.
- REER appreciations weakened total exports in the run-up to the crisis; response to REER changes is heterogeneous across goods and services exports.

### Crisis, post-crisis imbalances, and macro outcomes
- End of housing boom in early 2007 led to contraction of housing sector, rising unemployment, and deterioration in bank loan portfolios.
- Ireland’s exports-to-GDP ratio was 80 percent in 2007, contributing to strong exposure to external shocks.
- During 2008–11 Ireland registered a cumulative negative GDP growth rate of almost 7 percentage points.
- Public debt-to-GDP and NIIP deficit-to-GDP hovered around 100 percent by end-2011, from 25 percent and 21 percent, respectively, at end-2007.

### Econometric approach (VAR) and stylized facts (1980Q1–2009Q4)
- Data: quarterly GDP, exports, government revenue (real terms), REER (CPI-based and ULC-based), inward FDI (in 2005 Euros), and trading partners’ demand (partners accounting for 95 percent of Ireland’s exports).
- Bivariate findings:
  - Appreciation of the ULC-based REER is harmful to both FDI and exports.
  - The correlation between the CPI-based REER and exports is statistically significant and positive (counterintuitive).
- Two VAR models:
  - VAR1 (baseline): exports growth, GDP growth, revenue growth; includes constant and EMU dummy (EMU = 1 from 1991Q1 onward); four lags.
  - VAR2 (expanded): VAR1 plus DLNMGS_TP (trading partners’ demand), DINFDIR (change in inward FDI), and DLNREER (either ULC or CPI); four lags.

### Key VAR1 empirical results (exports → GDP → revenue)
- Impulse responses (one standard deviation shocks; variables in first differences):
  - A shock to exports of G&S (equivalent to 6½ percent) → 1 percentage point increase in GDP after one quarter; cumulative 7 percent by quarter 8; just over 8 percent by quarter 16.
  - A shock to GDP (corresponding to 4 percent) → cumulative impact on revenue of 2 percent after four quarters; peaks at 5 percent after eight quarters; falls to almost 3½ percent by quarter 16.
- Forecast error variance decomposition (FEVD):
  - A shock to exports explains 34½ percent of the variation in GDP after one quarter; peaks at 42¾ percent after three quarters; ~39 percent after eight and 16 quarters.
  - Revenue’s variance explained predominantly by its own innovations: 98¾ percent after one quarter; 82¾ percent after eight quarters; 80¼ percent after 16 quarters.
  - GDP explains an increasing share of revenue variance over time (rises to 14 percent after 16 quarters).

### VAR2 expanded-model findings (role of external demand, REER, FDI)
- Trading partners’ demand is a "game changer":
  - Inclusion of trading partners’ demand reduces the explanatory shares of exports for GDP and of GDP for revenue substantially relative to VAR1.
  - Trading partners’ demand becomes the single most important determinant of GDP and government revenue in VAR2.
- REER shocks (competitiveness effects):
  - One SD shock to ULC-based REER (equivalent to 6¼ percent appreciation) → reduces GDP by almost 2 percent after one year.
  - One SD shock to CPI-based REER (equivalent to 5 percent appreciation) → reduces GDP by marginally not significant 1 percent after one year.
  - One SD shock to ULC-based REER → export decline: 1½ percent after one quarter; over 6 percent after eight quarters.
  - One SD shock to CPI-based REER → export decline: ¾ percent after one quarter; 5½ percent after eight quarters.
  - Elasticities: a one percentage point increase in the ULC-based REER reduces exports of G&S by 0.6 percentage points; a one percentage point increase in the CPI-based REER reduces exports of G&S by 1 percentage point.
- Trading partners’ demand effects:
  - Shock to trading partners’ demand increases exports by ⅓ percent after one quarter; 4½ percent after four quarters; 7 percent after eight quarters.
  - Corresponding GDP impacts: ⅓ percent (Q1), 3 percent (Q4), 6⅓ percent (Q8), 7¾ percent (Q16).
  - Revenue responses: ~0 (Q1), 3 percent (Q4), 4½ percent (Q8), 5⅓ percent (Q16).
- Inward FDI shocks:
  - Boost exports by ¾ percent after one quarter, ~1 percent by the second quarter; effects are short-lived and lose statistical significance thereafter.
  - Increase GDP by about ⅓ percent one quarter ahead; thereafter impacts are not statistically significant.

### Variance decomposition insights in VAR2
- Exports variance:
  - Own innovations explain ~69¼ percent (Q1), 40 percent (Q4), and 32 percent (Q16).
  - REER explains up to 29¼ percent in the first four quarters; from five quarters out trading partners’ demand becomes the second most important contributor (27 percent around Q5; 22 percent by Q16).
- GDP variance:
  - Own innovations: 81 percent (Q1), 36¼ percent (Q4), 19 percent (Q16).
  - Trading partners’ demand: 6½ percent (Q1) rising to 29 percent (Q4) and peak 34¾ percent (Q6); from Q5 it often dominates GDP’s own innovations.
  - Revenue’s share peaks at 25½ percent after five quarters; 21½ percent at Q16.
- Revenue variance:
  - Own innovations: 97 percent (Q1) decaying to 67 percent (Q8) and 64¾ percent (Q16).
  - Trading partners’ demand becomes significant after three quarters (13½ percent), peaks at 19¼ percent (Q5), and ~17¾ percent (Q16).
- Conclusion: government revenue depends more on components exogenous to the VAR2 system than GDP does (own innovations account for ~68 percent of revenue variance vs ~16½ percent for GDP at Q16).

### Prospects, sensitivity analysis, and policy implications
- External demand projections and sensitivity:
  - Trading partners’ import demand projected to average 4¼ percent during 2013–17 (October 2012 WEO baseline used illustratively).
  - Sensitivity scenarios:
    - If trading partners’ demand grows at 6 percent (the average of the five years leading into the crisis), Ireland’s odds of reducing public debt improve through faster GDP and revenue growth.
    - If external demand falls by 5½ percent (average of 2008–09), Ireland’s economic performance would worsen significantly and the country would remain under water.
- Competitiveness post-crisis:
  - Over 2008–11 CPI-based REER depreciated by 10 percent and ULC-based REER depreciated by 28⅓ percent.
  - A 5-percentage point depreciation of the ULC-based REER would boost Ireland’s exports by 2½ percentage points.
  - An appreciation of 7 percent (matching pre-crisis average) would worsen export growth prospects.
- Domestic financing conditions:
  - Bank lending remains tight; only 50 percent of business loan applications were successful in 2010 (down from 90 percent in 2007).
  - Net FDI-to-GDP rose to 4 percent in 2010 and 13¾ percent in 2011, providing notable financing for firms and supporting exports growth.
  - Shocks to FDI have positive but short-lived impacts on exports and GDP; persistence of FDI requires continuous investor-friendly conditions.
- Policy recommendations and priorities:
  - Boost and maintain price and non-price competitiveness to support export-led recovery.
  - Contain domestic prices and labor costs given inability to use national monetary or nominal exchange rate policy within EMU.
  - Nurture the knowledge economy, innovation, and technology uptake to increase technology content of exports and boost market share.
  - Promote competition and open markets in sheltered sectors; upgrade infrastructure for telephone, broadband communication, and goods distribution.
  - Keep public spending in check so improvements in government revenue can strengthen fiscal balance and reduce public debt.
- Caveats:
  - Analysis includes financing from FDI only, not from the domestic financial sector (whose collapse was central to the crisis).
  - Post-crisis factors not included in the analysis: high household debt, high unemployment, and bank fragility.
  - Public expenditure developments are not included; conclusions depend on fiscal restraint.

### Key numeric highlights (preserved exactly as in the source)
- ETCR index values referenced: 5.67 in 1985 to 4.95 in 1988.
- REER ULC and CPI movements:
  - ULC-based REER appreciation over 2003–07: 10 percent.
  - CPI-based REER appreciation over 2003–07: 32½ percent.
  - Post-crisis (2008–11) depreciations: CPI-based REER 10 percent; ULC-based REER 28⅓ percent.
- Macro aggregates and rates:
  - Real GDP annual average growth, 1988–2007: 6 percent.
  - Unemployment: 17 percent (mid-1980s) to 4½ percent (2007).
  - Public debt: almost 110 percent (1987) to 25 percent (2007).
  - NIIP deficit: just over 20 percent of GDP (2007).
  - Exports-to-GDP ratio: 80 percent (2007).
  - Cyclically-adjusted deficit: over 7 percent of GDP in 2007 (from 2¾ percent of GDP in 2002).
  - Current account: deficit of 5½ percent of GDP by 2007 (from balance in 2003).
- VAR shock magnitudes and elasticities:
  - Exports shock in VAR1: equivalent to 6½ percent → GDP: 1 percentage point (Q1); cumulative 7 percent (Q8); just over 8 percent (Q16).
  - GDP shock in VAR1: corresponding to 4 percent → revenue: 2 percent (Q4); peaks at 5 percent (Q8); ~3½ percent (Q16).
  - ULC-based REER one SD shock: equivalent to 6¼ percent → reduces GDP by almost 2 percent after a year.
  - CPI-based REER one SD shock: equivalent to 5 percent → reduces GDP by ~1 percent after a year (marginally not significant).
  - ULC-based REER elasticity: a one percentage point increase → exports of G&S decline by 0.6 percentage points.
  - CPI-based REER elasticity: a one percentage point increase → exports of G&S decline by 1 percentage point.
  - Trading partners’ demand projected average, 2013–17 (used in sensitivity): 4¼ percent.
  - Net FDI-to-GDP: almost 100 percent (2002) to less than 20 percent (end-2007); net FDI-to-GDP: 4 percent (2010) and 13¾ percent (2011).
  - Cumulative GDP growth 2008–11: almost negative 7 percentage points.
  - Public debt-to-GDP and NIIP deficit-to-GDP by end-2011: hovered 100 percent.

*Source: IMF working paper content provided in the supplied PDF excerpt.*

### References

### _wp1335 - References

### Key references cited
- Barro, R., and X. Sala-I-Martin, Economic Growth, New York: McGraw-Hill, 1995.
- Bennett, H., Escolano and others, 2008, ―Competitiveness in the Southern Euro Area: France, Greece, Italy, Portugal and Spain,‖ IMF Working Paper 08/112 (Washington: International Monetary Fund).
- Bermingham, Colin and Thomas Conefrey, 2011, ―The Irish Macroeconomic Response to an External Shock with an Application to Stress Testing,‖ Research Technical Paper 10/RT/11 (Dublin: Central Bank of Ireland).
- Capell, Kerry, 2010, ―The New Generation Leaving Ireland,‖ Bloomberg Business Week Magazine, February 11, 2010.
- Cecchetti, S., M. S. Mohanty, and F. Zampolli, 2011, ―The Real Effects of Debt,‖ BIS Working Paper 352 (Basle: Bank for International Settlements).
- Central Bank of Ireland, 2011, Quarterly Bulletin No. 1 (January), (Dublin: Central Bank of Ireland).
- Central Bank of Ireland, 2012, Macro-financial Review 2012:II, (Dublin: Central Bank of Ireland).
- European Central Bank, 2005, ―Competitiveness and the Export Performance of the Euro Area,‖ ECB Occasional Paper 30.
- Everett, Mary, 2006, ―Foreign Direct Investment: An Analysis of its Significance,‖ Central Bank of Ireland, Quarterly Bulletin No 85(4).
- Forfás, 2008, Reconceptualising Ireland’s Competitiveness: A Summary of the 2008 National Competitiveness Council (Dublin: National Competitiveness Council).
- Forfás, 2011, ―Statement on Competitiveness Priorities,‖ (Dublin: National Competitiveness Council).
- Forfás, 2011, ―Ireland’s Competitiveness Scorecard,‖ September, (Dublin: National Competitiveness Council).
- Federici, Daniela and Daniela Marconi, 2002, ―On Exports and Economic Growth: The case of Italy,‖ Journal of International trade and Economic Development, Vol. 11(3), pp. 323–40.
- Holton, Sarah and M. O’Brien, 2011, ―Firms’ Financing during the Crisis: A Regional Analysis,‖ Central Bank of Ireland, Quarterly Bulletin No 89(1).
- Honohan, Patrick, 2009, ―What Went Wrong in Ireland,‖ Mimeo.
- Honohan, Patrick, and Brendan Walsh, 2002, ―Catching-up with the Leaders: The Irish Hare.‖ Brookings Papers on Economic Activity 2002(1): 1-79.
- Iacovone, Leonardo, and Veronika Zavacka, 2009, "Banking Crises and Exports: Lessons from the Past," Policy Research Working Paper Series, no. 5016 (Washington, D.C., World Bank, August).
- Kanda, Daniel, 2008, ―Spillovers to Ireland,‖ IMF Working Paper No. 08/2 (Washington: International Monetary Fund).
- Kumar, Manmohan, and Jaejoon Woo, 2010, ―Public Debt and Growth,‖ IMF Working Paper No. 10/174 (Washington: International Monetary Fund).
- Meese, Richard, and Kenneth Rogoff, 1983, ―The Out-of-Sample Failure of Empirical Exchange Rate Models: Sampling Error or Misspecification?‖ in Jacob A. Frenkel ed. Exchange Rates and International Macroeconomics, (Chicago: University of Chicago Press).
- O’Leary, D. and J. Tennent, 2011, ―Irish Competitiveness: Devaluation the Hard Way,‖ (Dublin: Goodbody Stockbrokers).
- Reinhart, Carmen, and K. Rogoff, 2010, ―Growth in a Time of Debt,‖ NBER WP/15639 (Cambridge: National Bureau of Economic Research).
- White, Mark C., 2003, ―Inward Investment, Firm Embeddedness and Place: An Assessment of Ireland’s Multinational Software Sector,‖ CISC Working Paper No 1 (Galway, Ireland: Centre for Innovation and Structural Change).

### Appendix — Variables used in econometric analysis: descriptions and data sources
- CPI-based real effective exchange rate (REER_CPI) — OECD and IMF, Information Notice System (INS).
- Unit labor cost-based real effective exchange rate (REER_ULC) — OECD and IMF, INS.
- Log of REER_CPI (LNREER_CPI) or REER_ULC (LNREER_ULC) — Calculated as noted in the description.
- First difference of LNREER_CPI (DLNREER_CPI) or LNREER_ULC (DLNREER_ULC) in percent — Calculated as noted in the description.
- Real exports of goods and services (EXPGS) — OECD.
- Log of exports of goods and services (LNEXPGS) — Calculated as noted in the description.
- First difference of LNEXPGS (DLNEXPGS) — Calculated as noted in the description.
- Real net exports of goods and services (NX) — OECD.
- Year-on-year change in real net exports of goods and services in percent (DLNNX) — Calculated as noted in the description.
- Real GDP (RGDP) — OECD.
- Real GDP growth (GRW) is year-on-year change in RGDP — Calculated as noted in the description.
- Real government revenue (REV) is government revenue deflated by the GDP deflator — Derived from OECD data as noted in the description.
- Log of real revenue (LNREV) — Calculated as noted in the description.
- First-difference of LNREV in percent (DLNREV) — Calculated as noted in the description.
- General government debt to GDP (DEBT_GDP) — OECD.
- Trading partners’ demand for goods and services (MGS_TP), 2005=100 — IMF, World Economic Outlook (WEO).
- Log of trading partners’ demand (LNMGS_TP) — Calculated as noted in the description.
- First-difference of LNMGS_TP in percent (DLNMGS_TP) — Calculated as noted in the description.
- Real foreign direct investment (INFDIR) is inward FDI Euros deflated by the GDP deflator with base 2005=100 — OECD and Ireland Central Statistics Office (CSO).
- Annual percentage change in INFDIR (DINFDIR) — Calculated as noted in the description.
- Note: LINFDIR is the logarithm of (100+INFDIR), a transformation aimed at avoiding taking logs of non-positive numbers.

### Correlation analysis (selected exact coefficients and p-values)
- Correlation matrix entries (correlation coefficient; p-value shown below each coefficient):
  - LNREER_ULC with LNREER_ULC: 1.00
  - LNREER_ULC with LNREER_CPI: -0.23*** (p = 0.01)
  - LNREER_ULC with LNMGS_TP: -0.89*** (p = 0.000)
  - LNREER_ULC with LNINFDIR: -0.27*** (p = 0.000)
  - LNREER_ULC with LNEXPGS: -0.92*** (p = 0.000)
  - LNREER_ULC with LNGDP_R: -0.92*** (p = 0.000)
  - LNREER_ULC with LNREV_R: -0.89*** (p = 0.000)
  - LNREER_CPI with LNMGS_TP: 0.56*** (p = 0.00)
  - LNMGS_TP with LNEXPGS: 1.00 (p = 0.000)
  - LNINFDIR with LNEXPGS: 0.12 (p = 0.21)
  - LNGDP_R with LNREV_R: 0.99*** (p = 0.000)
- *, **, and *** denote statistical significance at the 10 percent, 5 percent, and 1 percent significance level, respectively.

### VAR lag selection (sample 1980Q1–2009Q4)
- VAR1 (Endogenous: DLNEXPGS GRW DLNREV; Exogenous: C EMU): lag chosen by criteria: 2 (AIC = 13.73; SC = 14.31*; HQ = 13.96*).
- VAR2 with DLNREER_ULC (Endogenous: DLNMGS_TP DINFDIR DLNREER_ULC DLNEXPGS GRW DLNREV; Exogenous: C EMU): lag chosen by criteria: 4 (LR, FPE, AIC, SC, HQ indicate 4 with entries including highly large FPE values).
- VAR2 with DLNREER_CPI (same endogenous set but DLNREER_CPI): lag chosen by criteria: 2 (AIC = 28.56; SC = 30.60*; HQ = 29.39).

Notes for lag tests: LR denotes sequential modified LR test (each test at 5% level); FPE is Final prediction error; AIC is Akaike information criterion; SC is Schwarz information criterion; HQ is Hannan-Quinn information criterion.

### Bivariate Granger-causality tests (selected null hypotheses with F-statistic and P-value)
- DLNREER_ULC does not Granger Cause DINFDIR — F-statistic: 0.08; P-value: 0.99.
- DINFDIR does not Granger Cause DLNREER_ULC — F-statistic: 0.79; P-value: 0.53.
- DLNREER_ULC does not Granger Cause DLNEXPGS — F-statistic: 1.15; P-value: 0.34.
- DLNREER_ULC does not Granger Cause GRW — F-statistic: 2.83; P-value: 0.03.
- DINFDIR does not Granger Cause GRW — F-statistic: 3.56; P-value: 0.01.
- DINFDIR does not Granger Cause DLNREV — F-statistic: 5.60; P-value: 0.00.
- DLNREV does not Granger Cause DINFDIR — F-statistic: 3.83; P-value: 0.01.
- GRW does not Granger Cause DLNREV — F-statistic: 3.26; P-value: 0.01.
- DLNREV does not Granger Cause GRW — F-statistic: 10.80; P-value: 0.00.

Note: Tests based on 112 observations and four lags. Variables are defined in Table A1. Bold fonts denote statistical significance at the standard significance levels.

### Summary statistics for main VAR variables (all variables in percent; sample 1980Q1–2009Q4; Observations = 116 for listed series)
- DLNMGS_TP: Mean 4.96; Median 5.96; Std. Dev. 4.71; Observations 116.
- DINFDIR: Mean 1006.99; Median -5.98; Std. Dev. 1118.05; Observations 116.
- DLNREER_ULC: Mean -1.40; Median -1.35; Std. Dev. 1.05; Observations 116.
- DLNREER_CPI: Mean 1.04; Median 1.24; Std. Dev. 6.25; Observations 116.
- DLNEXPGS: Mean 8.85; Median 8.92; Std. Dev. 5.02; Observations 116.
- DLNNX: Mean 26.63; Median -9.68; Std. Dev. 64.84; Observations 116.
- GRW: Mean 4.34; Median 4.52; Std. Dev. 3.96; Observations 116.
- DLNREV: Mean 3.63; Median 4.35; Std. Dev. 4.77; Observations 116.
- Note: “The growth rate of net exports (DLNNX) is too volatile compared with other exports measures. A VAR framework including it in place of other export growth measures did not produce meaningful results.”

### Impulse response function highlights (generalized impulses with Monte Carlo standard errors, 1000 repetitions)
- Accumulated responses presented for VAR1 and two VAR2 specifications (including DLNREER_CPI or DLNREER_ULC). Selected accumulated responses (periods shown in tables):
  - Accumulated response of DLNMGS_TP (Period 16): VAR2 with DLNREER_CPI — 60.5; VAR2 with DLNREER_ULC — 69.0.
  - Accumulated response of DLNREER_CPI (Period 16): 0.0 2.6 6.2 -4.0 -1.5 -2.0 (table entries across variables).
  - Accumulated response of DLNREER_ULC (Period 16): -2.8 3.0 11.3 -8.8 -3.4 -4.0 (table entries across variables).
  - Accumulated response of DLNEXPGS (Period 16): VAR1 — 14.7 10.2 11.8 168.0 -3.1 -6.4; VAR2 (DLNREER_CPI) — 14.2 10.7 11.0 168.0 -3.1 -6.4; VAR2 (DLNREER_ULC) — 14.3 10.3 10.3 168.1 -3.7 -7.2.
  - Accumulated response of GRW (Period 16): VAR1 — 8.1 2.7 3.3 167.9 -3.1 -3.2; VAR2 results provided with analogous multi-variable entries.
  - Accumulated response of DLNREV (Period 16): VAR1 — 3.6 3.4 8.3 165.7 -3.3 -1.8; VAR2 entries analogous.
- Note: Bold fonts in tables denote statistical significance at the standard levels of 1 percent, 5 percent, and 10 percent.

### Forecast error variance decomposition (selected entries — percent contributions)
- Variance decomposition of DLNMGS_TP (Period 16): contributions from own shocks and others — for VAR2 (DLNREER_CPI) own-share 60.5; DINFDIR 5.1; DLNREER_CPI 3.3; DLNEXPGS 3.4; GRW 2.4; DLNREV 26.1.
- Variance decomposition of DINFDIR (Period 16): own-share 63.4; DLNMGS_TP 3.4; DLNREER_CPI 1.9; DLNEXPGS 2.3; GRW 1.2; DLNREV 13.3.
- Variance decomposition of DLNREER_CPI (Period 16): own-share 22.1; DLNMGS_TP 2.1; DINFDIR 3.5; DLNEXPGS 5.8; GRW 3.0; DLNREV 63.5.
- Variance decomposition of DLNEXPGS (Period 16): VAR1 (Period 16) — DLNEXPGS own 14.7; GRW 10.2; DLNREV 11.8; VAR2 entries show shifts with contributions from DLNMGS_TP, DINFDIR, DLNREER measures.
- Variance decomposition of GRW (Period 16): own-share ~38.9; DLNEXPGS ~28.0; DLNREV ~16.3 (approximate table entries preserved verbatim in full tables).
- Variance decomposition of DLNREV (Period 16): own-share ~63.6; DLNEXPGS ~3.3; GRW ~4.0; remaining shares across external variables.
- Note: Cholesky ordering reported: DLNEXPGS GRW DLNREV in VAR1; DLNMGS_TP DLNREER_CPI or DLNREER_ULC DINFDIR DLNEXPGS GRW DLNREV in VAR2. Bold fonts denote statistical significance at standard levels.

### Figures and additional empirical observations
- Figure A1 and Figure A2: Generalized impulse responses plotted for VAR1 and VAR2 respectively; impulses are generalized with Monte Carlo standard errors (1000 repetitions).
- Figure A3: Variance decomposition of exports, GDP, and revenue from VAR2 models. Cholesky ordering: DLNMGS_TP DINFDIR DLNREER_ULC (DLNREER_CPI) DLNEXPGS GRW DLNREV.
- Figure A4: Ireland foreign direct investment and credit market developments from Bank Lending Surveys:
  - Inward FDI picked up in 2009 and 2010 after registering a desinvestment in 2008.
  - Repatriation of outward FDI helped bring net FDI in percent of GDP to a 9-year high.
  - Lending standards eased more in Ireland compared with the Euro Area (EA) since mid-2010, except for a setback in January 2012.
  - For loan demand, the EA outperformed Ireland since mid-2007, though Ireland fared better on home loans since mid-2011.
  - Survey scoring: responses rated on a scale of 1 to 5, where “3” denotes unchanged conditions; figures below (above) 3 denote deterioration (improvement).
  - Sources cited for Figure A4: Ireland CSO, ECB, WEO, and author’s calculations.

*Italic: Source — _wp1335 - References (PDF)._

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp1335.pdf_
