## _wp10297

## Source details

**Canonical URL:** [_wp10297](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp10297.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp10297.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp10297.pdf.json)

---

### Introduction — Context and Motivation
- The fiscal impact of the global crisis has reinforced the urgency of pension and health entitlement reform.
- Staff projections suggest that age-related outlays (pensions and health spending) will rise by 4 to 5 percent of GDP in the advanced economies over the next 20 years.
- Pension and health entitlements already represent over a third of total spending in G-7 countries.
- The net present value of pension spending increases alone is estimated at about 8 percent of GDP in advanced and emerging countries.
- The paper emphasizes the short-run macro impact of pension reforms given that the economic recovery is not yet fully established.

### Aim and Approach
- Uses the IMF's Global Integrated Monetary and Fiscal (GIMF) model parameterized on data for five regions: United States (US), the euro area (EU), Japan (JA), emerging Asia (AS), and remaining countries (RC).
- Considers three pay-as-you-go public pension policy reform options:
  - Raising the retirement age: modeled via χ and N; reduces lifetime benefits; encourages longer working lives; may reduce saving and increase consumption during working years; increases fiscal saving with long-run positive effects on output through lower cost of capital and crowding in investment.
  - Reducing pension benefits: modeled as non-distortionary lump-sum transfers to all households; increases incentives to raise savings; reduces consumption in the short to medium run; increases investment over the long run.
  - Increasing contribution rates: modeled as an increase in the labor income tax rate; leads to distortionary supply-side effects on labor and negative aggregate demand effects on real disposable income; depresses real activity in both the short and long run.
- Benchmark starting point: 2014 for reform scenarios; scenarios include monetary policy sensitivity where policy remains accommodative for one or two years.

### Model Overview (Appendix 1 and model features)
- GIMF: dynamic stochastic general equilibrium model with overlapping generations (OLG) and finite horizons.
- Key structural features:
  - Overlapping generations and finite horizons produce medium- and long-run crowding-out effects of government debt and capture life-cycle income patterns, including age-dependent labor productivity.
  - Endogenous labor and capital markets; labor income taxes are distortionary; government debt crowds out economic activity via capital markets.
  - Multi-country trade in intermediate and final goods with bilateral flows; international asset trade limited to nominally non-contingent bonds denominated in U.S. dollars.
  - Financial spillovers across regions are a key channel for aging-related spending effects; spillovers likely dominate trade channel because of compounding effects on real interest rates.
  - Monetary policy: inflation-forecast-based interest rate reaction function targeting announced inflation.
  - Fiscal policy: fiscal rule stabilizes the government deficit-to-GDP ratio at a long-run target and lets the deficit fall with the output gap; rules out default and fiscal dominance.
- Agent groups and household parameters:
  - Three groups: households (OLG and LIQ), firms, government.
  - ψ: share of liquidity-constrained households (LIQ). LIQ consume after-tax income every period.
  - θ: survival parameter; average planning horizon = 1/(1 − θ).
  - χ: constant rate of decline in labor productivity over lifetimes (life-cycle income profile).
  - Working-age population index: N corresponds to workforce ages 15 to 64.
- Calibration globals:
  - Real global growth rate: 2.5 percent
  - Global population growth: 0.5 percent
  - Long-run global real interest rate: 4.0 percent

### Quantifying Public Pension Reforms (Appendix 1.B)
- Reform implementation specifics:
  - Targeted design: retirement age increases sufficient to stabilize pension spending as a share of GDP at its 2014 level over the next three to four decades; differences across regions reflect baseline pension gaps.
  - Two-year global-average retirement-age extension modeled by lengthening income profile (χ) and assuming increase in working-age population (N); a two-year extension ≈ 4 percent increase in working-age population (ages 15–64).
  - Income profile increase occurs immediately at reform start; labor force increases gradually over next 15 years.
- Table 1 indicative required pension age extensions across regions (preserved numeric entries as presented):
  - 2015–2030: US 2.5; JA 1.0; AS 1.5; EU 3.0
  - 2030–2050: US 0.0; JA 0.5; AS 0.5; EU 1.5
- Caveat on contribution-rate hike scenario: modeling via higher labor income tax captures labor supply effects but not direct labor demand effects from higher employer contributions (seen as a lower bound).

### Results: Region-by-Region Benchmark Public Pension Reform Scenarios (Section III)
- General result ranking:
  - Increasing the retirement age has the largest positive impact on real GDP.
  - Increasing contribution rates has the least favorable effect on output.
  - Reducing benefits can increase private savings but may weaken fragile domestic demand in the short run.
- Quantified effects of increasing retirement age (average two-year increase across regions):
  - Raises GDP by almost 1 percent in the short to medium run on average across regions.
  - Raises GDP by 4¼ percent by 2050 above the baseline.
  - Reduces the debt-to-GDP ratio by 30 percentage points over the same period.
- United States example (retirement age increase alone, keeping public pension spending constant):
  - Private saving rate as a ratio of GDP declines immediately by 0.2 percentage points.
  - Consumption rises above baseline by close to 2 percentage points in the short run.
  - Short-run increase in real GDP is 0.75 percent above baseline in period 2.
  - Public finances improve slightly: debt-to-GDP ratio is 4 percentage points below baseline after 20 periods (when transfers constant).
- United States full reform including concurrent cuts in public pension spending (complete fiscal consolidation):
  - Budget deficit improves by close to 3 percentage points of GDP after 30 years and settles around 2.2 percentage points in the long run.
  - Debt-to-GDP ratio declines by roughly 43 percentage points in the long run.
  - Short-run tightening of interest rates by 80 basis points in year 4 in response to inflationary pressure from short-run demand increase.
  - Medium- to long-run: investment is boosted as real interest rates fall in response to fiscal consolidation; output rises further with increased labor supply and falling marginal cost (real wage).

### Effects by Reform Type — Summarized Outcomes
- Increasing retirement age (United States focus and region specifics):
  - Without fiscal adjustment: real exchange rate appreciates; trade balance and current account deteriorate; real wages decline as more labor is supplied; higher return on capital attracts capital inflows.
  - With fiscal consolidation: lower real interest rates from increased world saving crowd in investment in external assets, leading to accumulation of net foreign assets; in very long run, current balances above baseline and REER begins appreciating again.
  - Japan: no pension age extension was needed.
  - Euro area: consumption weaker in the short run relative to the United States due to more rigid prices and more aggressive monetary rule; long-run consumption improves more; real GDP rises 5.8 percent above baseline in euro area under the retirement-age package.
  - Emerging Asia: fixed nominal exchange rate peg vis-à-vis the United States constrains short-run interest rate profile; real exchange rate depreciates initially; peg requires disinflation to obtain long-run real depreciation.
- Reducing pension benefits (United States case):
  - Short-run: consumption drops by about 1 percent below baseline; short-lived increase in GDP due to rapid fall in imports.
  - Long-run: real GDP rises and settles at almost 0.5 percent above baseline; world real interest rates decline moderately beginning in period 10, hitting a trough close to -0.4 percentage points below the baseline after 40 years.
  - Investment booms across countries; U.S. fiscal saving increases U.S. and world savings; real interest rates decline by close to 40 basis points leading to crowding in of investment.
- Raising contribution rates (proxied by increase in labor income tax; United States case):
  - Mechanism: higher labor income tax → decline in labor supply → negative effects on actual and potential output; declines in real disposable income reduce demand.
  - Quantitative outcomes (United States): real GDP declines by about 0.7 percent below baseline by period 10; long-run real GDP remains close to 0.4 percent below baseline.
  - Average required increase in contribution rates: 2¾ percentage points or over 10 percent (text preserved).

### Cooperative Global Reform Versus Regional Reform (Benchmark global scenario)
- Cooperative global reform yields larger impact on output and fiscal sustainability than isolated regional reform; cooperative benefit is greater than sum of individual benefits.
- Mechanism: larger decline in the world real interest rate under cooperation → larger compounding effect on world savings → stronger crowd-in of investment and higher potential and actual output in the long run.
- Representative cooperative outcomes reported (exact table extracts preserved):
  - Table 2 — Increase in Retirement Age (Global Scenario):
    - Real GDP (Global Scenario): US 5.4; EU 7.9; JA 2.1; AS 7.2; RC 13.5
    - Consumption (Global Scenario): US 6.4; EU 10.5; JA 2.1; AS 8.2; RC 14.3
    - Real Interest Rate (Global Scenario): -0.72 for US, EU, JA, AS, RC
    - Government Debt to GDP (Global Scenario): US -53.6; EU -64.8; JA -0.0; AS -21.2; RC -77.1
  - Table 3 — Reducing Pension Benefits (Global Scenario):
    - Real GDP (Global Scenario): US 1.6; EU 1.7; JA 1.7; AS 2.0; RC 1.9
    - Consumption (Global Scenario): US 1.3; EU 1.7; JA -0.3; AS 0.8; RC 2.4
    - Real Interest Rate (Global Scenario): -0.82 for US, EU, JA, AS, RC
    - Government Debt to GDP (Global Scenario): US -53.2; EU -58.1; JA -0.0; AS -21.3; RC -68.8
  - Table 4 — Raising Contribution Rates (Global Scenario):
    - Real GDP (Global Scenario): US 0.5; EU 0.0; JA 1.2; AS 1.0; RC 0.0
    - Consumption (Global Scenario): US 0.1; EU -0.5; JA -0.6; AS -0.3; RC -0.1
    - Real Interest Rate (Global Scenario): -0.62 for US, EU, JA, AS, RC
    - Government Debt to GDP (Global Scenario): US -45.9; EU -50.2; JA -0.0; AS -19.2; RC -62.3
- Cooperative quantitative claims (text preserved exactly):
  - Under cooperative case, real GDP is 50 percent higher in the long run (text statement preserved).
  - Permanent expansion in world real GDP (average over five regions) in the order of 7.2 percent above baseline (text preserved).
  - Cooperative improvement is about 40 percent larger than the sum of benefits from individual country reforms (text preserved).
- External balances under cooperation:
  - Cooperative scenario yields a weaker external balance in each country and less accumulation of net foreign assets versus unilateral action.
  - If only one country does not reform, the non-reformer’s balance would deteriorate.

### Sensitivity Analysis (Appendix 2 summary)
- Tests vary agents’ impatience/myopia (θ), share of LIQ households (ψ), timing of fiscal consolidation, labor supply response to retirement-age increases, and monetary policy accommodation.
- Key sensitivity findings:
  - Smaller increase in labor supplied in response to retirement-age increase reduces real activity; benchmark assumes no change in hours worked and effort.
  - More myopic agents (shorter planning horizon) → larger initial consumption drop but higher medium-run real GDP via investment from increased labor supply.
  - Much longer planning horizon (more Ricardian) → demand effect weaker; real interest rate does not respond in long run to global reform that lowers debt permanently.
  - Increasing LIQ share to 50 percent in all regions → consumption declines in US and EU relative to benchmark; after 10 periods interest rates fall more than benchmark, boosting investment and long-run real GDP.
  - Timing/profile of fiscal consolidation:
    - Immediate fiscal adjustment depresses short-run consumption for most regions (larger in RC due to larger LIQ share) but speeds investment pickup and earlier GDP recovery.
  - Monetary policy accommodation for one or two years:
    - Boosts consumption and real GDP in short run; improves public finances; when accommodation ends, larger policy tightening needed; long-run path unaffected.
  - Inflation persistence and nominal rigidities:
    - Lower inflation persistence or altered inflation formulation can make short-run consumption stronger, but affects realism of other dynamics (e.g., REER movements) and weakens monetary accommodation effectiveness.

### Main Conclusions and Policy Implications
- Reforms to pension systems can help ensure long-run viability of public finances while considering short-run effects on economic activity.
- Reforms that produce short-run adverse effects on real GDP (e.g., benefit reductions) are largely outweighed by benefits from declining real interest rates and positive effects on future potential productive capacity.
- Lengthening working lives (raising retirement age) yields the most positive long-run effects:
  - Boosts domestic demand in the short run.
  - Eases pressure on governments to cut pension benefits or to raise payroll and labor income taxes.
- Coordinated age-related fiscal reforms across regions produce greater positive impact on real GDP than unilateral reforms.
- Stabilizing the GDP share of age-related expenditures generally leads to sizable declines in debt-to-GDP ratios (declines by 40 to 50 percentage points depending on reform; improvement of approximately 30 percent relative to a non-cooperative strategy).
- Policy design considerations:
  - Timing/profile of consolidation affects short-run demand and recovery speed; gradual, long-run-tilted cuts reduce short-run negative consumption effects for non-Ricardian agents.
  - Temporary monetary accommodation can extend short-run boosts to activity and support confidence without altering long-run reform outcomes, but requires careful exit to re-anchor inflation.
  - Heterogeneity in liquidity constraints and planning horizons matters for short- and medium-run adjustment dynamics.

### Selected Calibration and Steady-State Values (preserved exactly)
- Intertemporal elasticity of substitution = 0.5 (γ = 2).
- Wage elasticity of labor supply = 0.5.
- Shares of LIQ consumers: 25 percent in US, EU, JA; 50 percent in AS and RC.
- Planning horizon: 20 years for 0.95θ = 0.95.
- Production elasticities: capital-labor substitution = 1; domestic-foreign goods = 0.75; tradables-nontradables = 0.5.
- Selected steady-state expenditure shares (In percent of GDP):
  - US: Consumption 65.3; OLG Consumption 51.8; LIQ Consumption 13.5; Private Investment 17.2; Government Expenditures 17.5; Exports 11.5; Imports 11.5.
  - AS: Consumption 59.0; OLG Consumption 33.7; LIQ Consumption 25.3; Private Investment 25.0; Government Expenditures 16.0; Exports 27.0; Imports 27.0.
  - EU: Consumption 58.2; OLG Consumption 46.5; LIQ Consumption 11.7; Private Investment 18.3; Government Expenditures 23.5; Exports 17.4; Imports 17.4.
  - JA: Consumption 59.5; OLG Consumption 46.9; LIQ Consumption 12.6; Private Investment 21.0; Government Expenditures 19.5; Exports 11.0; Imports 11.0.
  - RC: Consumption 59.0; OLG Consumption 34.0; LIQ Consumption 25.0; Private Investment 19.0; Government Expenditures 22.0; Exports 21.9; Imports 21.9.
- Selected steady-state deficit, debt, taxes, transfers (exact figures preserved):
  - US: Net Debt / GDP 128.7; Deficit / GDP 6.8; Labor Income Tax Rate 29.3; Capital Income Tax Rate 15.8; Consumption Tax Rate 7.0; Lump-Sum Taxes / Total Tax Revenue 10.2; Government General Transfers / GDP 9.5.
  - AS: Net Debt / GDP 48.5; Deficit / GDP 2.6; Labor Income Tax Rate 8.3; Capital Income Tax Rate 12.2; Consumption Tax Rate 17.0; Lump-Sum Taxes / Total Tax Revenue 18.9; Government General Transfers / GDP 5.1.
  - EU: Net Debt / GDP 136.2; Deficit / GDP 6.6; Labor Income Tax Rate 40.8; Capital Income Tax Rate 13.4; Consumption Tax Rate 18.4; Lump-Sum Taxes / Total Tax Revenue 6.1; Government General Transfers / GDP 15.6.
  - JA: Net Debt / GDP 156.0; Deficit / GDP 6.1; Labor Income Tax Rate 26.3; Capital Income Tax Rate 19.3; Consumption Tax Rate 8.7; Lump-Sum Taxes / Total Tax Revenue 9.1; Government General Transfers / GDP 7.5.
  - RC: Net Debt / GDP 94.4; Deficit / GDP 4.6; Labor Income Tax Rate 22.3; Capital Income Tax Rate 13.4; Consumption Tax Rate 17.0; Lump-Sum Taxes / Total Tax Revenue 33.8; Government General Transfers / GDP 16.7.

*Source: _wp10297 (IMF working paper content provided).*

### Introduction ...........................................................................................................

### Introduction

### Context and motivation
- The fiscal impact of the global crisis has reinforced the urgency of pension and health entitlement reform.
- Staff projections suggest that age-related outlays (pensions and health spending) will rise by 4 to 5 percent of GDP in the advanced economies over the next 20 years, underscoring the need to stabilize these outlays in relation to GDP.
- Pension and health entitlements already represent over a third of total spending in G-7 countries.
- The net present value of pension spending increases alone is estimated at about 8 percent of GDP in advanced and emerging countries.
- The paper emphasizes the short-run macro impact of pension reforms given that the economic recovery is not yet fully established.

### Aim and approach
- The paper examines public pension reforms using the IMF's Global Integrated Monetary and Fiscal (GIMF) model parameterized on data for five regions representing the world.
- It considers three policy reform options relating to pay-as-you-go public pension systems:
  - Raising the retirement age: reduces lifetime benefits paid to pensioners; encourages longer working lives; may reduce saving and increase consumption during working years; increases fiscal saving with long-run positive effects on output through lower cost of capital and crowding in investment.
  - Reducing pension benefits: increases incentives to raise savings to avoid sharper income declines in retirement; reduces consumption in the short to medium run; increases investment over the long run.
  - Increasing contribution rates: leads to distortionary supply-side effects on labor and negative aggregate demand effects on real disposable income; depresses real activity in both the short and long run.

### Summary of key results
- Increasing the retirement age has the largest positive impact on growth compared to reducing benefits.
- Increasing contribution rates (approximated by higher taxes on labor income) has the least favorable effect on output.
- Lengthening working lives:
  - Boosts domestic demand in the short run.
  - Reduces pressure on governments to cut pension benefits significantly or to raise payroll and labor income taxes.
- Reducing benefits can increase private savings but may weaken fragile domestic demand in the short run.
- Raising taxes can distort labor supply incentives.
- Coordinated reforms across regions produce larger impacts than unilateral regional reforms.
- Early and resolute action to reduce future age-related spending or finance the spending can improve fiscal sustainability over the medium run, with greater gains if reforms are enacted cooperatively.

*Introduction section of _wp10297 - Introduction*

### Appendix 1 provides further explanation of the sectors and the optimization involved. For a fuller exposition

### _wp10297 - Appendix 1 provides further explanation of the sectors and the optimization involved. For a fuller exposition

### A. Overview of the Model’s Key Features
- Model: GIMF (global integrated monetary–fiscal model), a dynamic stochastic general equilibrium model with overlapping generations (OLG) and finite horizons.
- Key structural features:
  - Overlapping generations and finite horizons produce medium- and long-run crowding-out effects of government debt and capture life-cycle income patterns, including age-dependent labor productivity.
  - Endogenous labor and capital markets: labor income taxes are distortionary; government debt crowds out economic activity via capital markets.
  - Multi-country structure: five regions — United States (US), the euro area (EU), Japan (JA), emerging Asia (AS), and remaining countries (RC).
  - Trade: intermediate and final goods trade with bilateral trade flows based on recent historical averages.
  - International asset trade limited to nominally non-contingent bonds denominated in U.S. dollars.
  - Financial spillovers across regions provide a key channel for aging-related spending effects; financial spillovers likely dominate the trade channel because of compounding effects on real interest rates (see Section IV.C.).
  - Monetary policy: inflation-forecast-based interest rate reaction function; central bank sets interest rates to stabilize inflation at an announced target.
  - Fiscal policy: fiscal rule stabilizes the government deficit-to-GDP ratio at a long-run target (structural) level and lets the deficit fall with the output gap to stabilize the business cycle; rules out default and fiscal dominance.

- Agent groups and important household parameters:
  - Three groups of agents and sectors: households, firms, government.
  - Household types and non-Ricardian parameters: ψ, θ, and χ.
    - ψ: share of liquidity-constrained households (LIQ) with no access to financial markets; consume after-tax income every period.
    - Remainder: overlapping generations (OLG) households, fully optimizing but face constant probability of death (1 − θ) each period, implying an average planning horizon of 1/(1 − θ).
    - χ: constant rate of decline in labor productivity over lifetimes (life-cycle income profile).
  - Working-age population index: N corresponds to population of workforce ages 15 to 64.
  - Firms are managed according to preferences of myopic OLG households and thus have finite planning horizons.

### B. Quantifying Public Pension Reforms
- Benchmark starting point: 2014 as the starting point for reform scenarios.
- Monetary policy sensitivity: scenarios include benchmark where monetary policy responds normally and sensitivity cases where monetary policy remains accommodative for a year or two.
- Reform options considered (pay-as-you-go public pension systems):
  - Increase in retirement age (modeled via χ and N).
  - Reduction in pension benefit payments (modeled as non-distortionary lump-sum transfers to all households).
  - Increase in contribution rates (modeled as an increase in the labor income tax rate).
- Targeted design: differentiated retirement age increases sufficient to stabilize pension spending as a share of GDP at its 2014 level over the next three to four decades. Differences across regions reflect different baseline pension gaps.
- Table 1 (required pension age extensions across regions):
  - Row labels and entries preserved as in source:
    - 2015–2030: US 2.5 ... JA 1.0 AS 1.5 EU 3.0
    - 2030–2050: US 0.0 ... JA 0.5 AS 0.5 EU 1.5
  - (RC entries are included in the table layout in the source; the above preserves the numeric entries as presented.)
- Implementation specifics for a global average two-year extension:
  - Modeled by lengthening agents‘ income profile (χ) and assuming an increase in working-age population (N).
  - Income profile increase occurs immediately at reform start; labor force increases gradually over next 15 years.
  - In GIMF, a two-year extension is roughly equivalent to a 4 percent increase in the working-age population (ages 15–64).
  - The income profile increase is consistent with a decrease of private saving as a percent of GDP (as found in related European studies referenced).
- Contribution-rate hike scenario caveat: modeling via higher labor income tax captures labor supply effects but not direct labor demand effects from higher employer contributions (seen as a lower bound).

### C. Caveats and Qualifications
- GIMF fiscal block limitations:
  - Does not allow for explicit breakdown of working-age and retired population nor an elaborate pay-as-you-go pension regime.
  - Rising dependency ratio cannot be explicitly decomposed into fertility versus longevity changes in this framework.
  - Labor force perception of pensions (tax-and-transfer versus insurance) and shifts to actuarially-based programs are not modeled; these could affect saving and labor supply behavior.
  - Extension of working lives is represented by changes to income profile (χ) and working-age population index (N) rather than explicit retirement cohort mechanics.
- Contribution-rate scenario limitation: increase in labor income tax is a lower-bound representation; employer contribution effects on labor demand are not captured.

### D. Calibration
- Calibration uses regional steady-state ratios and behavioral parameters consistent across regions unless stated otherwise; detailed tables in KLMM (2010).
- Annual version of the model used.
- Calibrated government debt-to-GDP ratios based on 2014 net debt projections from IMF's World Economic Outlook Update, February 2010, adjusted for pension gaps.
- Global parameter values:
  - Real global growth rate: 2.5 percent
  - Global population growth: 0.5 percent
  - Long-run global real interest rate: 4.0 percent

### III. Results: Public Pension Reforms
#### A. Baseline
- Baseline based on IMF‘s February 2010 World Economic Outlook for public debt in G-20 countries up to 2014–15 and Fiscal Affairs Department staff projections of public pension spending and primary fiscal balances over the next four decades.
- Baseline implies higher steady-state debt-to-GDP ratios and a higher world real interest rate than usually used in GIMF.

#### B. Region-by-Region Benchmark Public Pension Reform Scenarios
- Three reform options are broadly equivalent in fiscal impact (sufficient to offset projected long-run pension spending increases) but differ in growth and short-to-medium-run output effects.
- Type of reform matters:
  - Increasing retirement age has the largest positive impact on real GDP.
  - Increasing contribution rates has the least favorable effect on output.

- Quantified effects of increasing retirement age (average two-year increase across regions):
  - On average across regions:
    - Raises GDP by almost 1 percent in the short to medium run.
    - Raises GDP by 4¼ percent by 2050 above the baseline scenario.
    - Reduces the debt-to-GDP ratio by 30 percentage points over the same period.
  - United States example (increase in retirement age in US alone, first keeping public pension spending constant):
    - Private saving rate as a ratio of GDP declines immediately by 0.2 percentage points.
    - Consumption rises above baseline by close to 2 percentage points in the short run.
    - Short-run increase in real GDP is 0.75 percent above baseline in period 2.
    - Public finances improve slightly: debt-to-GDP ratio is 4 percentage points below baseline after 20 periods (when transfers constant).
  - Full reform including concurrent cuts in public pension spending (complete fiscal consolidation):
    - Budget deficit improves by close to 3 percentage points of GDP after 30 years and settles around 2.2 percentage points in the long run (given the long-run target).
    - Debt-to-GDP ratio declines by roughly 43 percentage points in the long run.
    - Consumption is depressed relative to the partial-analysis case (Figure 2) because lower transfers and perceived decline in OLG agents‘ net wealth affect LIQ agents more.
    - Short-run tightening of interest rates by 80 basis points in year 4 in response to inflationary pressure from short-run demand increase; monetary authority maintains a tight stance for a long period.
    - Medium- to long-run: investment is boosted as real interest rates fall in response to fiscal consolidation; output rises further with increased labor supply and falling marginal cost (real wage).

- Short-run versus long-run dynamics:
  - Short-run: reform can be stimulative (higher consumption, output) prompting monetary tightening; dynamics influenced by monetary–fiscal interaction and non-Ricardian features.
  - Medium/long-run: fiscal consolidation effects dominate the trajectory of debt, real interest rates, investment, and output.

*Source: Excerpts from the IMF working paper content provided in the source PDF.*

### 3.5 percent above baseline in the long run.

### _wp10297 - 3.5 percent above baseline in the long run.

### Effects of increasing the retirement age (United States focus)
- Increase in retirement age without fiscal adjustment:
  - Real exchange rate appreciates.
  - Trade balance and current account deteriorate.
  - Real wages decline as more labor is supplied; higher return on capital attracts capital inflows.
- When retirement-age increase is combined with fiscal consolidation:
  - Fiscal consolidation dominates external-variable effects.
  - Lower real interest rates from increased world saving crowd in investment in external assets, leading to an accumulation of net foreign assets.
  - In the very long run, declining interest payments to foreigners lead to current balances above baseline and the real effective exchange rate begins appreciating again.
  - Relative to baseline, the real effective exchange rate has depreciated, but the depreciation is smaller at 50 years than at 30 years.
- Short-run dynamics:
  - Private capital rises less than employment due to an expected marginal upward pressure on interest rates from increased demand which may lead to a partial crowding-out. (Footnote text preserved exactly.)

### Region-specific results for the retirement-age package (Figures 4–6)
- Euro area:
  - Qualitatively similar to the United States.
  - Smaller required pension-age increase to attain given budgetary saving (pensioners receive on average larger benefits).
  - More rigid prices and a more aggressive monetary rule → weaker short-run consumption profile relative to the United States.
  - Over the long run: consumption improves more as pension transfers are cut more aggressively in later periods, bringing a larger drop in interest rates and lower debt level (close to 47 percentage points below baseline).
  - Driven by higher domestic demand, real GDP rises 5.8 percent above baseline.
- Emerging Asia:
  - Fixed nominal exchange rate peg vis-à-vis the United States (instead of inflation targeting).
  - Real exchange rate depreciates in initial cases similar to United States and euro area.
  - Peg constrains emerging Asia to importing the short-run interest rate profile from the United States (as uncovered interest rate parity holds in GIMF).
  - To obtain real effective exchange rate depreciation in the long run, downward pressure on domestic prices yields sustained disinflation relative to baseline.
  - Model caveat: GIMF excludes capital or credit controls which may affect actual exchange rate policy conduct.
- Remaining countries block and Japan:
  - Japan: no pension age extension was needed (explicitly stated).
  - Quantitative results differ across regions but the intrinsic mechanism is similar to the United States case, with qualifications noted above.

### Reducing pension benefits (United States case; Figure 7)
- Short-run:
  - Consumption drops by about 1 percent below baseline.
  - Short-lived increase in GDP occurs due to rapid fall in imports relative to domestic demand (imports fall more quickly than consumption and investment).
- Long-run:
  - Real GDP rises and settles at almost 0.5 percent above the baseline scenario.
  - World real interest rates decline moderately beginning in period 10, hitting a trough close to -0.4 percentage points below the baseline after 40 years.
  - All countries experience a boom in investment; investment varies between 0.5 and 2 percent in the euro area and emerging Asia, respectively.
  - Permanent expansion in real GDP in trading partners varies between 0.3 to 0.5 percent.
  - U.S. fiscal saving increases U.S. and world savings; real interest rates decline by close to 40 basis points leading to crowding in of investment.
  - Government debt crowding-in mechanism frees resources from government debt instruments to other investments including foreign assets → current balances improve and gradual real appreciation follows.
- Cross-region spillovers:
  - Spillover on other regions’ real GDP from a large region (United States or euro area) is about four times that from a smaller region (emerging Asia calibrated at 13 percent of world nominal GDP versus 27 percent for the United States).

- Additional model notes preserved:
  - The average reduction in benefit payments is high exceeding 20 percent in the remaining countries block (footnote).
  - Potentially reduced sovereign risk premia associated with favorable market responses to improved public finances are not taken into account.

### Raising contribution rates (proxied by increase in labor income tax; United States case; Figure 11)
- Mechanisms and effects:
  - Increase in labor income tax rate → decline in supply of labor → negative effects on actual and potential output.
  - Decline in households’ real disposable income reduces demand; important given myopic LIQ and OLG agents (finite remaining working life of 20 years on average).
  - LIQ households consume at most their after-tax current income; share of LIQ households matters regionally.
  - Decline in potential output likely exerts upward pressure on inflation.
- Quantitative outcomes (United States):
  - Short-run losses larger than benefit-reduction scenario: real GDP declines by about 0.7 percent below baseline by period 10.
  - Long-run: real GDP remains close to 0.4 percent below baseline versus an increase of 0.4 percent when cuts in pension benefits are used instead.
- Average required increase in contribution rates:
  - On average, the necessary increase in contribution rates is 2¾ percentage points or over 10 percent (footnote).

### Benchmark global scenario — simultaneous reforms in all regions (cooperative strategy)
- Cooperative global reform yields larger impact on output and fiscal sustainability than isolated regional reform; cooperative benefit is greater than sum of individual benefits.
- Mechanism:
  - Larger decline in the world real interest rate under cooperation → larger compounding effect on world savings → stronger crowd-in of investment and higher potential and actual output in the long run.
- United States comparison (Figure 15; starting in 2014):
  - Under cooperative case, real GDP is 50 percent higher in the long run (text statement preserved exactly).
  - Interest rate decline about five times that under individual action.
  - Permanent expansion in world real GDP (average over five regions) in the order of 7.2 percent above baseline (text preserved).
  - Cooperative improvement is about 40 percent larger than the sum of benefits from individual country reforms (text preserved).
- External balances under cooperation:
  - Cooperative scenario yields a weaker external balance in each country and less accumulation of net foreign assets.
  - Current account improves by less under cooperative action than when policy is taken by each country alone.
  - If only one country does not reform, the non-reformer’s balance would deteriorate (private saving declines or consumption rises due to appreciating real exchange rate; investment rises due to lower interest rates).
- Fiscal sustainability:
  - Stabilizing GDP share of age-related (pension) expenditures leads to sizable declines in debt-to-GDP ratios.
  - Debt-to-GDP ratios decline by 40 to 50 percentage points (depending on reform) below baseline on average across regions, an improvement of approximately 30 percent relative to a non-cooperative strategy (text preserved).

### Quantitative impacts from Tables (selected exact figures)
- Table 2. Cooperative Versus Regional Public Pension Reform — Increase in Retirement Age (Global Scenario rows shown):
  - Real GDP (Global Scenario): US 5.4; EU 7.9; JA 2.1; AS 7.2; RC 13.5
  - Consumption (Global Scenario): US 6.4; EU 10.5; JA 2.1; AS 8.2; RC 14.3
  - Real Interest Rate (Global Scenario): -0.72 for US, EU, JA, AS, RC
  - Government Debt to GDP (Global Scenario): US -53.6; EU -64.8; JA -0.0; AS -21.2; RC -77.1
- Table 3. Cooperative Versus Regional Public Pension Reform — Reducing Pension Benefits (Global Scenario rows shown):
  - Real GDP (Global Scenario): US 1.6; EU 1.7; JA 1.7; AS 2.0; RC 1.9
  - Consumption (Global Scenario): US 1.3; EU 1.7; JA -0.3; AS 0.8; RC 2.4
  - Real Interest Rate (Global Scenario): -0.82 for US, EU, JA, AS, RC
  - Government Debt to GDP (Global Scenario): US -53.2; EU -58.1; JA -0.0; AS -21.3; RC -68.8
- Table 4. Cooperative Versus Regional Public Pension Reform — Raising Contribution Rates (Global Scenario rows shown):
  - Real GDP (Global Scenario): US 0.5; EU 0.0; JA 1.2; AS 1.0; RC 0.0
  - Consumption (Global Scenario): US 0.1; EU -0.5; JA -0.6; AS -0.3; RC -0.1
  - Real Interest Rate (Global Scenario): -0.62 for US, EU, JA, AS, RC
  - Government Debt to GDP (Global Scenario): US -45.9; EU -50.2; JA -0.0; AS -19.2; RC -62.3
- Selected regional scenario rows (United States Scenario rows preserved as in tables):
  - Increase in retirement age — United States Scenario:
    - Real GDP: US 3.6; EU 0.3; JA 0.4; AS 0.5; RC 0.5
    - Consumption: US 4.0; EU 0.2; JA 0.2; AS 0.4; RC 0.6
    - Real Interest Rate: -0.13 for US, EU, JA, AS, RC
    - Government Debt to GDP: US -43.0; EU 0.0; JA -0.0; AS 0.0; RC -0.0

### Sensitivity analysis (summary of approach and focus)
- Benchmark cooperative global reform scenarios are tested for sensitivity to key parameters, including agents’ impatience and myopia, timing of fiscal consolidation, and selected other values.
- Tests apply to the benefit-cut scenario except for two tests: sensitivity of hours worked to a raise in retirement age, and role of monetary policy accommodation in the short run.
- Paper reports qualitative results of main tests (text ends with referral to further sections; content preserved exactly as provided).

*Italic attribution: Source: _wp10297 - 3.5 percent above baseline in the long run.*

### Appendix 2 for a quantitative assessment accompanied by Figures 18 to 24. The sensitivity

### Appendix 2 for a quantitative assessment accompanied by Figures 18 to 24. The sensitivity

### Sensitivity tests and scenario findings
- A smaller increase in labor supplied in response to an increase in retirement age reduces real activity. The benchmark case assumes no change in hours worked and effort.
- A shorter planning horizon with more myopic agents:
  - leads to a larger drop in consumption initially;
  - leads to higher real GDP in the medium run driven by a boost in investment resulting from an increased labor supply.
- Under a much longer planning horizon, agents become far more Ricardian in their behavior:
  - demand effect is far weaker with consumption behavior barely changing in response to reform in both the short and long run;
  - supply and the capital intensity of production respond a lot less which reduces the effect of fiscal reform on real interest rates—the latter does not respond in the long run to the global reform which lowers debt permanently.
- Increasing the share of liquidity-constrained agents in the population to 50 percent in all regions:
  - leads to a reduction in consumption in the United States and the euro area (regions where this share has doubled), contrary to the benchmark results;
  - over the medium run, interest rates fall relative to benchmark, boosting investment and the long run level of real GDP;
  - Emerging Asia and the block of remaining countries (regions for which this share does not change) benefit from lower world interest rates, which leads to a small improvement in consumption starting in the first year.
- Timing of fiscal consolidation:
  - having public pension spending adjust immediately to its lower long-run level depresses consumption in the short run for most regions;
  - the difference is larger in the remaining countries block due to a larger share of LIQ households;
  - the increase in government savings crowds out the need for private savings much quicker, but investment picks up faster, meaning the recovery in real GDP happens much earlier than in the benchmark reform scenario.
- Monetary policy accommodation for one or two years:
  - boosts consumption and real GDP at the expense of added inflation volatility;
  - short-run demand pressure moves inflation away from target and further lowers real interest rates, prolonging the initial positive impact of public pension reforms;
  - public finances improve short-run as government deficits decline faster due to lower debt service payments;
  - when normal conduct of monetary policy returns, there is the usual dampening effect in the medium run; in the long run, the economy follows the same path as in the benchmark scenario.
- Inflation formulation effects:
  - if the United States alone carries out public pension reform, there is a short-run decline in consumption because inflationary pressures lead monetary authority to increase interest rates;
  - by cutting the degree of inflation persistence, or altering how inflation persistence is determined, consumption will be positive;
  - such a change to calibration would be at the expense of other properties such as exchange rate volatility and the possibility for effective monetary accommodation.
- Note references in sensitivity discussion: footnote markers 26 and 27 appear in the text.

### Main conclusions drawn in the paper
- Reforms to pension systems can help ensure long-run viability of public finances while considering short-run effects on economic activity during a global financial crisis; analysis uses a dynamic general equilibrium model (GIMF).
- Reforms that produce short-run adverse effects on real GDP (i.e., benefit reductions) are largely outweighed by benefits from declining real interest rates and positive effects on future potential productive capacity.
- The reform with the most positive long-run effects is lengthening working lives (raising the size of the active labor force relative to retirees):
  - helps boost domestic demand in the short run;
  - eases pressure on governments to cut pension benefits alone (which can otherwise raise private savings and depress fragile domestic demand in the short run) or to raise payroll and labor income taxes (which can distort labor supply incentives).
- Cooperative age-related fiscal reforms across regions produce a greater positive impact on real GDP relative to unilateral reform by one region.
- Stabilizing the GDP share of age-related expenditures generally leads to a sizable decline in the debt-to-GDP ratio.
- Early and resolute action to reduce future age-related spending or finance spending through measures (preferably increasing retirement age) can significantly improve fiscal sustainability over the medium run, more so if enacted cooperatively.

### GIMF: main model features (households, firms, government, calibration)
- Sectors and externalities:
  - model economies feature households, firms (and entrepreneurs), government, and a fully developed external sector;
  - real aggregate variables rescaled by labor-augmenting world technology tT and by population growth, notation ( ) t t t x x T n = ;
  - world technology grows at constant rate 1tt g T T   and world population grows at constant rate n.
- Households:
  - share of liquidity-constrained (LIQ) households in population is ()z ψ where z refers to a country;
  - in each period ( ) ( ) 1 t N z nz n θ ψ     of such individuals are born, where ()Nz indexes absolute population sizes in period 0;
  - OLG households have finite planning horizons with share 1 ( )z ψ ; in each period ( ) (1 ( )) 1 t N z nz n θ ψ     of such individuals are born;
  - each agent faces constant probability of death ( ) 1 θ  in each period, implying average planning horizon ( ) 1/ 1 θ  ;
  - households experience labor productivity that declines at constant rate χ over their lifetimes, reaching near zero in 20 years.
- OLG households — optimality and consumption:
  - key relation: consumption equals marginal propensity to consume out of wealth times real wealth;
  - current aggregate consumption expressed as function of real aggregate financial wealth and human wealth (human wealth composed of expected present discounted value of time endowments at after-tax real wage and expected present discounted value of capital or dividend income net of lump-sum transfers);
  - government debt adds to agents’ net worth under myopia; with finite planning horizon agents treat government debt as net wealth because they underweight future tax liabilities needed to service debt (myopia 1 θ  and labor income decline 1 χ );
  - fiscal consolidation through higher taxes tilts tax profile toward near future reducing debt stock but decreases human wealth and consumption for given marginal propensity to consume.
  - intertemporal substitution in consumption ( ) 1 γ affects sensitivity to real interest rate; for conventional 1 γ  income effect of higher real interest rate is stronger than substitution effect and increases marginal propensity to consume and tends to produce larger interest rate changes to fiscal shocks.
- Key expressions and equations cited in text:
  - aggregate consumption relation: OLGLK ttttt cfwhwhw Θ = + + (equation (6))
  - expressions for fw, hw components and marginal propensity Θ given by formulae labeled (7), (8), (9), (10), (11) in the source.
- LIQ households:
  - can consume at most current income (after-tax wage income plus net government transfers);
  - marginal propensity to consume out of present income is unity, making revenue-based fiscal stimulus measures have particularly high fiscal multipliers when LIQ share is large;
  - objective function assumed identical to OLG households; LIQ households accept wages as given and supply labor proportionate to OLG response.
  - aggregate consumption and labor supply: OLGLIQ t t t C c c = + and OLGLIQ t t t L l l = + .
- Firms:
  - managed according to preferences of myopic OLG household owners (finite planning horizons);
  - productive sector with tradables, nontradables, intermediate goods, final consumption and investment goods; financial accelerator amplifies propagation of shocks;
  - monopolistic competition and nominal rigidities in price setting except capital goods producers, entrepreneurs and retailers;
  - firms maximize present discounted value of net cash flow or dividends with survival probability θ in discount factors.
- Government and fiscal policy:
  - government consumption and investment: consinv ttt G G G = + ; lump-sum transfers to both LIQ and OLG denoted OLGLIQ ttt  =  +  ;
  - taxes: labor income τ ,Lt consumption τ ,ct and corporate income τ ,kt ; social security contribution rates embedded in taxes on labor income; pensions treated as untargeted lump-sum transfers to households;
  - real government budget constraint given by equation (1) in source:
    11 11 tt tttttt tt ii bbGbs gngn          
    where government issues nominal one-period debt t b at gross nominal rate t i and t s is primary surplus.
  - transfers partly compensate for lack of asset ownership by LIQ agents by redistributing fraction of OLG dividend income to LIQ agents; lump-sum transfers and taxes received/paid by LIQ in proportion to their share in aggregate consumption (but rule can be changed).
- Fiscal policy rule:
  - stabilizes interest-inclusive government deficit to GDP ratio rat t gd at long-run target rat t gdss and stabilizes business cycle by letting deficit fall with output gap:
    ln rat rat gdp t tt pot gdp gdgdssd gdp        (equation (2) in source)
  - rat t gd defined via expression (3) in source involving debt, interest and inflation terms;
  - relationship between debt target rat t bss and deficit target rat t gdss: 1 rat rat tt gn bss gdss gn      (equation (4) in source);
  - parameter gdp d quantifies automatic stabilizers: for a given one percent increase in the output gap the fiscal authority increases its target surplus by gdp d percent;
  - default setting in this paper is that lump-sum transfers are the instrument used to satisfy the fiscal targeting rule.
- Monetary policy:
  - stylized Taylor-type interest rate reaction function where central bank sets rates on basis of deviation of current and one-year ahead inflation from inflation target to stabilize inflation.
- Calibration:
  - steady state ratios of particular importance for exercise are presented in subsequent calibration materials (not repeated here).

*Source: Appendix 2 and accompanying text from the GIMF-based quantitative assessment (document content provided).*

### Appendix Table 1. Steady state GDP decompositions and trade flows are based on recent

### _wp10297 - Appendix Table 1. Steady state GDP decompositions and trade flows are based on recent

### Calibration and key parameter choices
- Household utility functions are equal across countries.
- Intertemporal elasticity of substitution is 0.5 (describing the sensitivity of consumption to changes in the real interest rate), or  = 2.
- Wage elasticity of labor supply is 0.5.
- Parameters ,  and  govern non-Ricardian behavior in the model.
- Shares of LIQ consumers:
  - 25 percent in the United States, the euro area, and Japan.
  - 50 percent in emerging Asia and other countries.
- Planning horizon and retention parameters:
  - Planning horizon is equal to 20 years for 0.95.
  - Average remaining time at work is 20 years for 0.95.
- Criterion for choosing  and : empirical evidence for the effect of government debt on real interest rates (model calibrated so that a 1 percentage point increase in the government debt to GDP ratio in the United States leads to a long-run increase of approximately 1 to 2 basis points in the U.S. (and world) real interest rate).
- Production technology elasticities:
  - Elasticity of substitution between capital and labor = 1.
  - Elasticity between domestic and foreign goods = 0.75.
  - Elasticity between tradables and nontradables = 0.5.
- Steady-state markups:
  - Manufacturing and wage setting = 1.1.
  - Retailing, investment and consumption goods production = 1.05.
  - Import agents = 1.025.
- Fiscal rule calibration:
  - Use OECD estimates for the output gap coefficients, gdp_d, as proxies for automatic stabilizers.
  - Overall target deficit-to-GDP ratios equal projected 2014 government debt to GDP ratios (IMF, 2010c).
- Monetary rule parameters: conventional for an annual model.
- For emerging Asia, assume a fixed nominal exchange rate vis-à-vis the U.S. dollar.

### Selected steady-state values (preserved exactly)
- A. Expenditure Shares (In percent of GDP)
  - US: Consumption 65.3; OLG Consumption 51.8; LIQ Consumption 13.5; Private Investment 17.2; Government Expenditures 17.5; Government Consumption 15.0; Government Investment 2.5; Exports 11.5; Imports 11.5.
  - AS: Consumption 59.0; OLG Consumption 33.7; LIQ Consumption 25.3; Private Investment 25.0; Government Expenditures 16.0; Government Consumption 12.0; Government Investment 4.0; Exports 27.0; Imports 27.0.
  - EU: Consumption 58.2; OLG Consumption 46.5; LIQ Consumption 11.7; Private Investment 18.3; Government Expenditures 23.5; Government Consumption 20.5; Government Investment 3.0; Exports 17.4; Imports 17.4.
  - JA: Consumption 59.5; OLG Consumption 46.9; LIQ Consumption 12.6; Private Investment 21.0; Government Expenditures 19.5; Government Consumption 17.0; Government Investment 2.5; Exports 11.0; Imports 11.0.
  - RC: Consumption 59.0; OLG Consumption 34.0; LIQ Consumption 25.0; Private Investment 19.0; Government Expenditures 22.0; Government Consumption 20.0; Government Investment 2.0; Exports 21.9; Imports 21.9.

- B. Deficit, Debt, Taxes, and Transfers
  - US: Net Debt / GDP 128.7; Deficit / GDP 6.8; Labor Income Tax Rate 29.3; share in tax revenues 61.9; Capital Income Tax Rate 15.8; share in tax revenues 11.3; Consumption Tax Rate 7.0; share in tax revenues 16.3; Lump-Sum Taxes / Total Tax Revenue 10.2; Government General Transfers / GDP 9.5.
  - AS: Net Debt / GDP 48.5; Deficit / GDP 2.6; Labor Income Tax Rate 8.3; share in tax revenues 20.1; Capital Income Tax Rate 12.2; share in tax revenues 13.9; Consumption Tax Rate 17.0; share in tax revenues 46.3; Lump-Sum Taxes / Total Tax Revenue 18.9; Government General Transfers / GDP 5.1.
  - EU: Net Debt / GDP 136.2; Deficit / GDP 6.6; Labor Income Tax Rate 40.8; share in tax revenues 60.5; Capital Income Tax Rate 13.4; share in tax revenues 6.9; Consumption Tax Rate 18.4; share in tax revenues 26.4; Lump-Sum Taxes / Total Tax Revenue 6.1; Government General Transfers / GDP 15.6.
  - JA: Net Debt / GDP 156.0; Deficit / GDP 6.1; Labor Income Tax Rate 26.3; share in tax revenues 55.2; Capital Income Tax Rate 19.3; share in tax revenues 17.5; Consumption Tax Rate 8.7; share in tax revenues 18.2; Lump-Sum Taxes / Total Tax Revenue 9.1; Government General Transfers / GDP 7.5.
  - RC: Net Debt / GDP 94.4; Deficit / GDP 4.6; Labor Income Tax Rate 22.3; share in tax revenues 33.8; Capital Income Tax Rate 13.4; share in tax revenues 7.3; Consumption Tax Rate 17.0; share in tax revenues 25.2; Lump-Sum Taxes / Total Tax Revenue 33.8; Government General Transfers / GDP 16.7.

- C. Factor Shares (In percent of GDP)
  - Labor Income: US 60.0; AS 54.0; EU 60.0; JA 60.0; RC 60.0.

- Source line included in table: KLMM (2010); IMF (2010c); and staff estimates.

### Sensitivity analysis — main themes and findings
- Labor supply response to an increase in retirement age:
  - If increases in hours worked by older workers are smaller (cases labeled 23 and 34), real GDP outcomes decline relative to benchmark:
    - In the 34 case, real GDP reduces by roughly 20 to 25 percent worldwide.
    - In the 23 case, real GDP reduces by 27 to 33 percent worldwide.
  - Mechanisms: direct reduction in production; indirect reduction via lower marginal product of capital leading to negative effects on investment growth; cumulative decrease in human wealth and consumption given unchanged marginal propensity to consume.

- Role of the degree of myopia / planning horizon ():
  - Benchmark: 20-year planning horizon (0.95).
  - More myopic (0.90, 10-year horizon):
    - Consumption initially drops notably for the United States and the euro area (lower share of LIQ households).
    - OLG agents do not internalize crowding-in benefits and higher future income, so they consume less and react stronger to pension benefit cuts.
    - Labor supply rises more (agents consume less leisure), investment increases substantially, and in the long run real GDP boost lowers debt to GDP ratios and real interest rates.
  - Much longer horizon (0.98, 50-year horizon):
    - Agents behave far more Ricardian; debt holdings are not treated as wealth.
    - Demand effects are far weaker; saving and consumption behavior barely change.
    - Supply and capital intensity respond less; fiscal reform has smaller effect on real interest rates.
    - Real interest rate does not respond in the long run to global reform that lowers debt permanently.

- Share of liquidity-constrained (LIQ) households:
  - Experiment: set LIQ share = 50 percent for all countries.
  - Results:
    - Consumption declines for the United States and the euro area, contrary to benchmark results.
    - Supply side largely unaffected in short and medium run.
    - After 10 periods, interest rates fall more than in benchmark, propelling investment and long-run growth.
    - Emerging Asia and remaining countries benefit from lower world interest rates, with small improvements in consumption in early periods.

- Timing/profile of fiscal consolidation:
  - Benchmark: pension spending cuts are more tilted toward the longer run (less medium-run adjustment).
    - Non-Ricardian agents see attenuated short-run wealth reduction and smaller negative impact on consumption.
    - Less crowding-out on private saving vs. front-loaded consolidation.
  - Rapid, front-loaded consolidation (immediate and permanent decline in public pension spending preserving long-run magnitude):
    - Consumption drops in short run for United States, euro area, and remaining countries (larger drop for remaining countries due to higher LIQ share).
    - OLG agents' consumption less affected.
    - Government savings crowd out private savings need faster; investment picks up faster; real GDP recovery happens earlier than benchmark.
  - If all countries mimic U.S. consolidation profile:
    - Magnitude of responses in real GDP, consumption, and private saving become much closer to U.S. case in long run.
    - External adjustment patterns (imports, exports, REER) remain unchanged because they are dictated by trade linkages calibrated in GIMF.

- Monetary policy accommodation:
  - Pre-announced temporary monetary accommodation (nominal rates fixed at initial levels for one or two years) reduces real interest rates (nominal lower than otherwise plus higher inflation).
  - Effects:
    - Crowding in of consumption and investment; higher real GDP.
    - Improved public finances: public debt declines further and deficits improve due to lower debt service and larger GDP tax base.
    - When accommodation ends, policy rates must rise by larger magnitude than under benchmark to restore inflation stabilization.
  - Conclusion: monetary accommodation can extend short-run boost in activity to sustain confidence in entitlement reforms without long deviations from price stability; long-run impact of reforms unchanged by accommodation.

- Behavior of inflation and nominal rigidities:
  - Experiments:
    - Cut level of nominal rigidities by half.
    - Use hybrid nominal rigidity formulation: half Ireland (year-to-year inflation inertia) and half Rotemberg (adjustment costs depending on deviation from inflation target).
  - Findings:
    - Under benchmark, consumption weakens during years 2 to 7.
    - With lower inflation persistence (lower rigidities or hybrid formulation), consumption can strengthen in those years.
    - Mechanisms: faster inflation response shortens horizon for monetary offset; inflation more responsive to interest rates so interest rates need to increase less to achieve same inflation movement; lower interest rate increases exert less downward pressure on real economy in short run.
    - Caveats: changing inflation process calibration affects overall model properties; monetary accommodation would be less effective in these experiments; some model variable responses become less sensible (e.g., REER moves much less), at odds with observed data.

### Policy implications and recommendations (as reflected in the analysis)
- Design fiscal consolidations with attention to timing/profile:
  - More gradual, long-run tilted pension spending cuts reduce short-run negative consumption effects for non-Ricardian agents and limit crowding out of private saving.
  - Front-loaded consolidations speed up investment-led recoveries but cause larger short-run consumption drops, particularly in high-LIQ-share regions.
- Consider coordinated monetary policy when implementing structural entitlement reforms:
  - Temporary monetary accommodation can sustain confidence effects and extend short-run boosts to real activity without changing long-run reform outcomes, but requires careful exit to re-anchor inflation.
- Account for heterogeneity in household liquidity constraints and planning horizons:
  - Higher shares of LIQ households amplify short-run negative consumption effects in some regions; more myopic agents change short- and long-run adjustment dynamics.
- Be cautious when altering nominal rigidity calibrations:
  - While lower inflation persistence can improve short-run domestic activity responses, it may reduce realism of other model dynamics and weaken monetary accommodation effectiveness.

*Source: KLMM (2010); IMF (2010c); and staff estimates.*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp10297.pdf_