## 3. Effects of Pension Reforms on Household’s Labor Effort Profile by Age

## Source details

**Canonical URL:** [3. Effects of Pension Reforms on Household’s Labor Effort Profile by Age](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12283.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12283.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12283.pdf.json)

---

### I. Introduction
- Major policy challenges for economies in the periphery of the euro area:
  - preserve fiscal sustainability, improve the external balance, and enhance future growth prospects.
- Pension reform summary (conditional on reforms achieving similar fiscal targets):
  - Increasing the retirement age: large positive effect on output, small (and often negative) effect on the current account.
  - Cutting pension benefits: large positive effect on the current account, reduces output.
  - Mixed reforms (extend working life and cut benefits) can in some cases increase output and improve the current account.
- Principal conclusion:
  - Increasing the size of the current account effect can only be achieved through a reduction in the size of the output effect; this entails larger pension benefit cuts and shorter extensions of the working life.

### II. The Model
- Structure:
  - Overlapping generations of finitely-lived households, atomistic firms, infinitely-lived government.
  - Households: work during youth (T years), retire for R years; utility depends on consumption (c) and leisure (l) with preference parameters (β, γ).
  - Firms: constant-returns-to-scale production with labor-augmenting technological progress Z, capital share α, depreciation δ.
  - Government: collects payroll (τ), income (τI), and consumption (τc) taxes to finance G, pension benefits, and redeem debt D; households and government trade in international capital markets at interest rate r.
- Key mechanisms:
  - Pension benefit at retirement based on average past wage earnings indexed by wage growth; replacement parameter ψ determines replacement behavior.
  - Households internalize that higher wage earnings increase pension benefits, affecting labor incentives.
  - Stationary-transformed model accounts for labor-augmenting technological progress (ξ) and population size (P). Interest rate (r) and labor effort (n) are not detrended.
- Balanced Growth Equilibrium maintained with constant p, ξ, T, R, and fiscal policy.
- Parameterization (Table 1 values):
  - α = 0.300
  - γ = 3.050
  - β = 0.985
  - δ = 0.060
  - ξ = 0.015
  - p = 0.008
  - Work life T = 40.000 (entry to labor force at natural age 23)
  - Retirement life R = 18.000 (retirement at natural age 63, death at natural age 81)
  - r = 0.050
  - ψ = 0.330
  - Social security payroll tax rate = 0.142
  - Capital-income tax rate = 0.130
  - G/Y = 0.230
  - D/Y = 0.750
- Calibration notes:
  - γ set so household fraction of time worked = 0.293.
  - ξ and p match historical Spain.
  - r and β chosen to obtain baseline current account deficit of 3 percent of GDP.
  - Fiscal/pension targets guided by Spanish historical observations: social security revenue = 9.5 percent of GDP, pension spending = 8 percent of GDP.

### III. Results
- Simulation design:
  - Baseline and pension reform scenarios simulated.
  - Income and payroll tax rates unchanged; consumption tax rate adjusts endogenously to satisfy government budget constraint.

- A. Baseline
  - Household wage income profile by age: hump-shaped; income falls at retirement because pension benefit partially replaces wage income and pension benefits are tax-exempt while wages are taxed.
  - Consumption profile: rising lifetime consumption; households incur debt early in working life and accumulate assets mid-life to supplement pension income.

- B. Reform I: Increasing the Retirement Age (from 63 to 66 years)
  - Four forces affecting household disposable income (ordered by quantitative importance):
    1. Higher wage income at ages 63-65 because wage > pension benefit (boosts disposable income).
    2. Lower consumption taxes resulting from the reform (boosts disposable income).
    3. Inter-temporal substitution of leisure: forced delay of retirement leads households to substitute leisure inter-temporally and work fewer hours at ages 23-62 (reduces disposable income; small relative to other effects).
    4. Lower average productivity at retirement because extending working life shifts retirement into lower-skilled ages given hump-shaped skill profile (small effect).
  - Net household effect: increase in lifetime disposable income and higher consumption.
  - Saving behavior:
    - Households reduce saving during most working life (ages 40-62); lower pension benefits partially offset this by encouraging saving.
    - Household saving also falls during retirement due to higher consumption and lower pension benefit.
  - Macroeconomic effects:
    - Effective labor increases (old-age dependency declines).
    - Capital increases.
    - Output (per capita) increases.
    - Consumption tax rate falls (fiscal space from higher payroll revenue and lower aggregate pension spending).
    - Current account balance declines relative to baseline due to lower aggregate saving as fraction of output and higher domestic investment in line with output.
  - Tradeoff outcome: reform increases output but worsens current account (point A in Figure 4).

- C. Reform II: Cutting Pension Benefits (reducing ψ)
  - Reform calibrated to deliver same pension expenditure reduction (percent of GDP) as Reform I.
  - Household response:
    - Reduction in effective return on work effort → labor effort by households aged 40-62 declines.
    - Households increase saving during working lives to smooth lifetime consumption given lower retirement income.
    - Faster asset accumulation and lower consumption taxes raise disposable income and lifetime consumption.
  - Macroeconomic effects:
    - Aggregate labor, capital, and output decline in per capita terms.
    - Pension spending falls; consumption tax rate reduced.
    - Current account improves: large increase in saving as share of output and decline in domestic investment in line with output.
  - Tradeoff outcome: reform reduces output but improves current account (point D in Figure 4).

- D. The Long-Run Tradeoff between Output and the Current Account
  - Mixed reforms evaluated (increase retirement age and cut benefits while achieving same expenditure reductions):
    - Point B: increase retirement age by two years and cut benefits.
    - Point C: increase retirement age by one year and rely more on benefit cuts.
  - A-D curve: long-run tradeoff — for a given expenditure reduction target (percent of GDP), larger current account improvements require smaller output gains (left-upward movements along A-D curve), implying larger benefit cuts and shorter working-life extensions.
  - Additional observations:
    - With a larger initial pension system (example: ψ = 0.4 and pension expenditure = 9.7 percent of GDP), A’-D’ curve lies above A-D:
      - Tradeoff remains; larger initial pension system magnifies reform effects.
      - Example point A’ can yield simultaneous gains in output and the current account (a reform that only increases retirement age can boost both in that case).
    - Tradeoff robust when fiscal objective is a given consumption tax rate target (A-E curve) instead of pension spending-to-GDP target.
  - Welfare (normalized utility = 1 in baseline):
    - Reform A: 1.0035 (0.35 percent gain)
    - Reform B: 1.0046
    - Reform C: 1.0058
    - Reform D: 1.0069
  - Concluding result: robust long-term tradeoff between output and current account effects of pension reforms, conditional on similar fiscal targets.
  - Note: short- and medium-term presence of the tradeoff may depend on grandfathering and sequencing of fiscal targets (beyond scope).

### IV. Quantitative Results and Key Statistics (selected values from Table 2; variables expressed as fraction of de-trended output in the baseline unless indicated)
- Scenarios: Baseline | Increasing the Retirement Age | Reducing Pension Benefits (Exp. Target) | Reducing Pension Benefits (Tax Target)
- Key aggregate levels and indexes:
  - Capital:
    - Baseline: 3.000
    - Increasing the Retirement Age: 3.050
    - Reducing Pension Benefits (Exp. Target): 2.984
    - Reducing Pension Benefits (Tax Target): 2.988
  - Effective Labor:
    - Baseline: 1.217
    - Increasing the Retirement Age: 1.237
    - Reducing Pension Benefits (Exp. Target): 1.211
    - Reducing Pension Benefits (Tax Target): 1.212
  - Output (index, baseline = 1.000 / 100.0):
    - Baseline: 1.000 / 100.0
    - Increasing the Retirement Age: 1.017 / 100.0
    - Reducing Pension Benefits (Exp. Target): 0.995 / 100.0
    - Reducing Pension Benefits (Tax Target): 0.996 / 100.0
  - Net Factor Income:
    - Baseline: -0.064 / -6.4
    - Increasing the Retirement Age: -0.067 / -6.5
    - Reducing Pension Benefits (Exp. Target): -0.043 / -4.4
    - Reducing Pension Benefits (Tax Target): -0.048 / -4.8
  - Gross National Income:
    - Baseline: 0.936 / 93.6
    - Increasing the Retirement Age: 0.950 / 93.5
    - Reducing Pension Benefits (Exp. Target): 0.951 / 95.6
    - Reducing Pension Benefits (Tax Target): 0.948 / 95.2
  - Private Consumption:
    - Baseline: 0.486 / 48.6
    - Increasing the Retirement Age: 0.493 / 48.5
    - Reducing Pension Benefits (Exp. Target): 0.495 / 49.7
    - Reducing Pension Benefits (Tax Target): 0.493 / 49.5
  - Investment:
    - Baseline: 0.249 / 24.9
    - Increasing the Retirement Age: 0.253 / 24.9
    - Reducing Pension Benefits (Exp. Target): 0.248 / 24.9
    - Reducing Pension Benefits (Tax Target): 0.248 / 24.9
  - Current Account:
    - Baseline: -0.030 / -3.0
    - Increasing the Retirement Age: -0.031 / -3.0
    - Reducing Pension Benefits (Exp. Target): -0.020 / -2.0
    - Reducing Pension Benefits (Tax Target): -0.022 / -2.2
  - National Saving:
    - Baseline: 0.220 / 22.0
    - Increasing the Retirement Age: 0.223 / 21.9
    - Reducing Pension Benefits (Exp. Target): 0.228 / 22.9
    - Reducing Pension Benefits (Tax Target): 0.226 / 22.7
  - Net Exports:
    - Baseline: 0.021 / 2.1
    - Increasing the Retirement Age: 0.036 / 3.5
    - Reducing Pension Benefits (Exp. Target): 0.023 / 2.3
    - Reducing Pension Benefits (Tax Target): 0.026 / 2.6
  - Consumption Tax Rate (percent):
    - Baseline: 27.1
    - Increasing the Retirement Age: 23.4
    - Reducing Pension Benefits (Exp. Target): 22.3
    - Reducing Pension Benefits (Tax Target): 23.4
  - Payroll Tax Revenue:
    - All scenarios: 0.095 / 9.5
  - Pension Expenditure:
    - Baseline: 0.080 / 8.0
    - Increasing the Retirement Age: 0.062 / 6.1
    - Reducing Pension Benefits (Exp. Target): 0.061 / 6.1
    - Reducing Pension Benefits (Tax Target): 0.066 / 6.6
  - Balance (pension system):
    - Baseline: 0.015 / 1.5
    - Increasing the Retirement Age: 0.034 / 3.4
    - Reducing Pension Benefits (Exp. Target): 0.034 / 3.4
    - Reducing Pension Benefits (Tax Target): 0.029 / 2.9
  - Revenue (overall):
    - Baseline: 0.330 / 33.0
    - Increasing the Retirement Age: 0.317 / 31.2
    - Reducing Pension Benefits (Exp. Target): 0.310 / 31.2
    - Reducing Pension Benefits (Tax Target): 0.315 / 31.6
  - Expenditure (overall):
    - Baseline: 0.347 / 34.7
    - Increasing the Retirement Age: 0.334 / 32.9
    - Reducing Pension Benefits (Exp. Target): 0.327 / 32.9
    - Reducing Pension Benefits (Tax Target): 0.332 / 33.3
  - Overall Balance:
    - All columns: -0.017 / -1.7
- Asset decomposition (selected):
  - Assets:
    - Baseline: 2.471 / 247.1
    - Increasing the Retirement Age: 2.481 / 244.1
    - Reducing Pension Benefits (Exp. Target): 2.862 / 287.7
    - Reducing Pension Benefits (Tax Target): 2.769 / 278.1
  - Net Foreign Assets:
    - Baseline: -1.279 / -127.9
    - Increasing the Retirement Age: -1.331 / -130.9
    - Reducing Pension Benefits (Exp. Target): -0.868 / -87.3
    - Reducing Pension Benefits (Tax Target): -0.965 / -96.9
- Pension system metrics:
  - Pension benefit at retirement (percent of output per capita):
    - Baseline: 30.4
    - Increasing the Retirement Age: 28.9
    - Reducing Pension Benefits (Exp. Target): 23.3
    - Reducing Pension Benefits (Tax Target): 25.0
  - Replacement rate (percent of average wage income):
    - Baseline: 33.0
    - Increasing the Retirement Age: 33.0
    - Reducing Pension Benefits (Exp. Target): 25.4
    - Reducing Pension Benefits (Tax Target): 27.2
  - Replacement rate (percent of last wage income):
    - Baseline: 43.3
    - Increasing the Retirement Age: 56.9
    - Reducing Pension Benefits (Exp. Target): 35.1
    - Reducing Pension Benefits (Tax Target): 37.1

### Appendix: Household’s Optimization Problem (summary)
- Household chooses sequences of consumption, leisure, and asset holdings to maximize lifetime utility (log utility) subject to budget constraint, pension formula, and no intergenerational bequest condition (A_{t+T+1} = 0).
- Optimization expressed as backward-recursive dynamic problems (pre-retirement and post-retirement); value function at retirement V(A, b) has closed-form solution due to log utility.
- First-order conditions yield:
  - Intratemporal condition (consumption-leisure) for working age and retirement (see tabular summary of first-order conditions).
  - Intertemporal condition (consumption-saving) for working age and retirement with derivatives of V with respect to A and b featuring in Euler-like equations.
- Recursive derivations produce constants (e.g., Ω) aggregating log and parameter terms across future dates and closed-form expressions for derivatives A V and b V used in FOCs.

*Source: IMF Working Paper chapter "3. Effects of Pension Reforms on Household’s Labor Effort Profile by Age" (extracted from _wp12283).*

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

### References      .............................................................................................................................22

### Tables
- 1. Baseline Parameter Values ................................................................................................... 13
- 2. Results ................................................................................................................................... 16

### Figures
- 1. Baseline Household’s Wage, Pension, Disposable Income, Consumption, and Saving Profiles by Age ....................................................................................................13
- 2. Effects of Pension Reforms o Household’s Disposable Income and Consumption Profiles by Age ................................................................................................................ 14

*Source: _wp12283 - References (PDF); https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12283.pdf*

### 3. Effects of Pension Reforms on Household’s Labor Effort Profile by Age ........................... 14

### 3. Effects of Pension Reforms on Household’s Labor Effort Profile by Age

### I. Introduction
- Economies in the periphery of the euro area face three major policy challenges: preserve fiscal sustainability, improve the external balance, and enhance future growth prospects.
- Pension reforms that increase the retirement age or cut pension benefits clearly improve fiscal balance; their long-run effects on output and the current account are less evident.
- Using an overlapping-generations small open economy model with a pay-as-you-go pension system, the paper shows a long-run tradeoff between a domestic objective (output) and an external objective (current account balance), conditional on reforms achieving similar fiscal targets:
  - Reforms that increase the retirement age: large positive effect on output, small (and often negative) effect on the current account.
  - Reforms that cut pension benefits: large positive effect on the current account, reduce output.
  - Mixed reforms (extend working life and cut benefits) can in some cases increase output and improve the current account.
- Principal conclusion: Increasing the size of the current account effect can only be achieved through a reduction in the size of the output effect; this entails larger pension benefit cuts and shorter extensions of the working life.

### II. The Model
- Structure:
  - Overlapping generations of finitely-lived households, atomistic firms, infinitely-lived government.
  - Households: work during youth (T years), retire for R years; utility depends on consumption (c) and leisure (l):
    - Utility: sum over life of log(c) and log(l) terms with preference parameters (β, γ) as in equation (1).
  - Budget constraint and wage/pension specification given in equation (2) and pension formulae.
  - Firms: constant-returns-to-scale production with labor-augmenting technological progress Z, capital share α, depreciation δ.
  - Government: collects payroll (τ), income (τI), and consumption (τc) taxes to finance G, pension benefits, and redeem debt D (government budget constraint shown).
  - Households and government trade in international capital markets at prevailing interest rate r.
- Model features and key mechanisms:
  - Pension benefit at retirement based on average past wage earnings indexed by wage growth; replacement parameter ψ determines replacement behavior.
  - Households internalize that higher wage earnings increase pension benefits, affecting labor incentives.
  - Stationary-transformed model accounts for labor-augmenting technological progress (ξ) and population size (P). Interest rate (r) and labor effort (n) are not detrended.
- Balanced Growth Equilibrium:
  - Constant p (population growth), ξ, working life T, retirement R, and fiscal policy (constant tax rates, public expenditure, debt-to-output).
- Parameterization (Table 1 highlights):
  - α = 0.300
  - γ = 3.050
  - β = 0.985
  - δ = 0.060
  - ξ = 0.015
  - p = 0.008
  - Work life T = 40.000 (implying entry to labor force at natural age 23)
  - Retirement life R = 18.000 (implying retirement at natural age 63, death at natural age 81)
  - r = 0.050
  - ψ = 0.330
  - Social security payroll tax rate = 0.142
  - Capital-income tax rate = 0.130
  - G/Y = 0.230
  - D/Y = 0.750
- Calibration:
  - γ set so representative household fraction of time worked = 0.293.
  - ξ and p match historical Spain.
  - r and β chosen to obtain baseline current account deficit of 3 percent of GDP (Table 2).
  - Fiscal and pension parameters guided by Spanish historical observations: social security revenue = 9.5 percent of GDP, pension spending = 8 percent of GDP.

### III. Results
- Simulation design:
  - Baseline and pension reform scenarios simulated.
  - Income and payroll tax rates unchanged; consumption tax rate adjusts endogenously to satisfy government budget constraint.
- A. Baseline
  - Household wage income profile by age is hump-shaped; income falls at retirement because pension benefit partially replaces wage income and pension benefits are tax-exempt while wages are taxed (Figure 1).
  - Households choose a rising lifetime consumption profile; incur debt early in working life and accumulate assets mid-life to supplement pension income.
- B. Reform I: Increasing the Retirement Age (from 63 to 66 years)
  - Four forces affect household disposable income (ordered by quantitative importance):
    1. Higher wage income at ages 63-65 because wage > pension benefit (boosts disposable income).
    2. Lower consumption taxes resulting from the reform (boosts disposable income).
    3. Inter-temporal substitution of leisure: forced delay of retirement leads households to substitute leisure inter-temporally and work fewer hours at ages 23-62 (reduces disposable income; small relative to other effects).
    4. Lower average productivity at retirement because extending working life shifts retirement into lower-skilled ages given hump-shaped skill profile (small effect).
  - Net household effect: increase in lifetime disposable income and higher consumption.
  - Saving behavior:
    - Households reduce saving during most working life (ages 40-62); lower pension benefits partially offset this by encouraging saving.
    - Household saving also falls during retirement due to higher consumption and lower pension benefit.
  - Macroeconomic effects (Table 2 summary):
    - Effective labor increases (old-age dependency declines; more working cohorts dominate leisure substitution effect).
    - Capital increases (capital per effective worker necessary to complement higher labor).
    - Output (per capita) increases.
    - Consumption tax rate falls (fiscal space from higher payroll revenue and lower aggregate pension spending).
    - Current account balance declines relative to baseline due to:
      - Lower aggregate saving as fraction of output (saving decline of cohorts aged 40-62 and 66-80 dominates increases elsewhere).
      - Domestic investment rises in line with output; investment as fraction of output remains roughly unchanged.
  - Tradeoff outcome: reform increases output but worsens current account (point A in Figure 4).
- C. Reform II: Cutting Pension Benefits (reducing ψ)
  - Reform calibrated to deliver same pension expenditure reduction (percent of GDP) as Reform I.
  - Household response:
    - Reduction in effective return on work effort → labor effort by households aged 40-62 declines (Figure 3).
    - Households increase saving during working lives to smooth lifetime consumption given lower retirement income.
    - Faster asset accumulation and lower consumption taxes raise disposable income and lifetime consumption (Figure 2).
  - Macroeconomic effects (Table 2 summary):
    - Aggregate labor, capital, and output decline in per capita terms.
    - Pension spending falls; consumption tax rate reduced.
    - Current account improves: large increase in saving as share of output and decline in domestic investment in line with output.
  - Tradeoff outcome: reform reduces output but improves current account (point D in Figure 4).
- D. The Long-Run Tradeoff between Output and the Current Account
  - Evaluating mixed reforms that increase retirement age and cut pension benefits while achieving the same expenditure reductions:
    - Point B: increase retirement age by two years and cut benefits.
    - Point C: increase retirement age by one year and rely more on benefit cuts.
  - A-D curve demonstrates the long-run tradeoff: for a given expenditure reduction target (percent of GDP), larger current account improvements require smaller output gains (left-upward movements along A-D curve), implying larger benefit cuts and shorter working-life extensions.
  - Additional observations:
    - In economies with larger initial pension systems (example: ψ = 0.4 and pension expenditure = 9.7 percent of GDP), the A’-D’ curve lies above A-D:
      - Tradeoff remains.
      - Larger initial pension system magnifies reform effects.
      - Example point A’ can yield simultaneous gains in output and the current account (reform that only increases retirement age can boost both in that case).
    - The tradeoff is robust when the fiscal objective is a given consumption tax rate target (A-E curve) instead of a pension spending-to-GDP target.
  - Welfare implications (footnote values, normalized utility = 1 in baseline):
    - Reform A: 1.0035 (0.35 percent gain)
    - Reform B: 1.0046
    - Reform C: 1.0058
    - Reform D: 1.0069
  - Concluding result: robust long-term tradeoff between output and current account effects of pension reforms, conditional on similar fiscal targets.
  - Note: Short- and medium-term presence of the tradeoff may depend on grandfathering and sequencing of fiscal targets (beyond scope).

### IV. Quantitative Results and Key Statistics (from Table 2; all variables expressed as fraction of de-trended output in the baseline unless indicated)
- Baseline, Increasing the Retirement Age, Reducing Pension Benefits (Exp. Target), Reducing Pension Benefits (Tax Target) — selected rows and exact values:
  - Capital:
    - Baseline: 3.000
    - Increasing the Retirement Age: 3.050
    - Reducing Pension Benefits (Exp. Target): 2.984
    - Reducing Pension Benefits (Tax Target): 2.988
  - Effective Labor:
    - Baseline: 1.217
    - Increasing the Retirement Age: 1.237
    - Reducing Pension Benefits (Exp. Target): 1.211
    - Reducing Pension Benefits (Tax Target): 1.212
  - Output (index, baseline = 1.000 = 100.0):
    - Baseline: 1.000 / 100.0
    - Increasing the Retirement Age: 1.017 / 100.0
    - Reducing Pension Benefits (Exp. Target): 0.995 / 100.0
    - Reducing Pension Benefits (Tax Target): 0.996 / 100.0
  - Net Factor Income:
    - Baseline: -0.064 / -6.4
    - Increasing the Retirement Age: -0.067 / -6.5
    - Reducing Pension Benefits (Exp. Target): -0.043 / -4.4
    - Reducing Pension Benefits (Tax Target): -0.048 / -4.8
  - Gross National Income:
    - Baseline: 0.936 / 93.6
    - Increasing the Retirement Age: 0.950 / 93.5
    - Reducing Pension Benefits (Exp. Target): 0.951 / 95.6
    - Reducing Pension Benefits (Tax Target): 0.948 / 95.2
  - Private Consumption:
    - Baseline: 0.486 / 48.6
    - Increasing the Retirement Age: 0.493 / 48.5
    - Reducing Pension Benefits (Exp. Target): 0.495 / 49.7
    - Reducing Pension Benefits (Tax Target): 0.493 / 49.5
  - Investment:
    - Baseline: 0.249 / 24.9
    - Increasing the Retirement Age: 0.253 / 24.9
    - Reducing Pension Benefits (Exp. Target): 0.248 / 24.9
    - Reducing Pension Benefits (Tax Target): 0.248 / 24.9
  - Current Account:
    - Baseline: -0.030 / -3.0
    - Increasing the Retirement Age: -0.031 / -3.0
    - Reducing Pension Benefits (Exp. Target): -0.020 / -2.0
    - Reducing Pension Benefits (Tax Target): -0.022 / -2.2
  - National Saving:
    - Baseline: 0.220 / 22.0
    - Increasing the Retirement Age: 0.223 / 21.9
    - Reducing Pension Benefits (Exp. Target): 0.228 / 22.9
    - Reducing Pension Benefits (Tax Target): 0.226 / 22.7
  - Net Exports:
    - Baseline: 0.021 / 2.1
    - Increasing the Retirement Age: 0.036 / 3.5
    - Reducing Pension Benefits (Exp. Target): 0.023 / 2.3
    - Reducing Pension Benefits (Tax Target): 0.026 / 2.6
  - Consumption Tax Rate (percent):
    - Baseline: 27.1
    - Increasing the Retirement Age: 23.4
    - Reducing Pension Benefits (Exp. Target): 22.3
    - Reducing Pension Benefits (Tax Target): 23.4
  - Payroll Tax Revenue:
    - All scenarios: 0.095 / 9.5 (unchanged across columns shown)
  - Pension Expenditure:
    - Baseline: 0.080 / 8.0
    - Increasing the Retirement Age: 0.062 / 6.1
    - Reducing Pension Benefits (Exp. Target): 0.061 / 6.1
    - Reducing Pension Benefits (Tax Target): 0.066 / 6.6
  - Balance (pension system):
    - Baseline: 0.015 / 1.5
    - Increasing the Retirement Age: 0.034 / 3.4
    - Reducing Pension Benefits (Exp. Target): 0.034 / 3.4
    - Reducing Pension Benefits (Tax Target): 0.029 / 2.9
  - Revenue (overall):
    - Baseline: 0.330 / 33.0
    - Increasing the Retirement Age: 0.317 / 31.2
    - Reducing Pension Benefits (Exp. Target): 0.310 / 31.2
    - Reducing Pension Benefits (Tax Target): 0.315 / 31.6
  - Expenditure (overall):
    - Baseline: 0.347 / 34.7
    - Increasing the Retirement Age: 0.334 / 32.9
    - Reducing Pension Benefits (Exp. Target): 0.327 / 32.9
    - Reducing Pension Benefits (Tax Target): 0.332 / 33.3
  - Overall Balance:
    - All columns: -0.017 / -1.7
  - Asset decomposition (selected):
    - Assets:
      - Baseline: 2.471 / 247.1
      - Increasing the Retirement Age: 2.481 / 244.1
      - Reducing Pension Benefits (Exp. Target): 2.862 / 287.7
      - Reducing Pension Benefits (Tax Target): 2.769 / 278.1
    - Net Foreign Assets:
      - Baseline: -1.279 / -127.9
      - Increasing the Retirement Age: -1.331 / -130.9
      - Reducing Pension Benefits (Exp. Target): -0.868 / -87.3
      - Reducing Pension Benefits (Tax Target): -0.965 / -96.9
  - Pension system metrics:
    - Pension benefit at retirement (percent of output per capita):
      - Baseline: 30.4
      - Increasing the Retirement Age: 28.9
      - Reducing Pension Benefits (Exp. Target): 23.3
      - Reducing Pension Benefits (Tax Target): 25.0
    - Replacement rate (percent of average wage income):
      - Baseline: 33.0
      - Increasing the Retirement Age: 33.0
      - Reducing Pension Benefits (Exp. Target): 25.4
      - Reducing Pension Benefits (Tax Target): 27.2
    - Replacement rate (percent of last wage income):
      - Baseline: 43.3
      - Increasing the Retirement Age: 56.9
      - Reducing Pension Benefits (Exp. Target): 35.1
      - Reducing Pension Benefits (Tax Target): 37.1

### Appendix: Household’s Optimization Problem (summary)
- Household chooses sequences of consumption, leisure, and asset holdings to maximize lifetime utility (log utility) subject to budget constraint, pension formula, and no intergenerational bequest condition (A_{t+T+1} = 0).
- Optimization expressed as two dynamic problems: pre-retirement and post-retirement recursion.
- Value function at retirement V(A, b) has closed-form solution due to log utility; first-order conditions derive consumption-leisure condition and labor supply implications (see Table A1 and appendix derivation).

*IMF Working Paper chapter: "3. Effects of Pension Reforms on Household’s Labor Effort Profile by Age" (extracted content)._

### 1. The household’s problem at date 1

### 1. The household’s problem at date 1

### Problem formulation at date 1
- The household at date R_tt + T (household’s age is R_tt sT T = +) maximizes the value function:
  - 1 111 111 ,log11. RRR tttttttt RRRR R tt tttttt TT tt R tT T tt TTTTTTTT tTtT tT TtT TtT TtT T A VAbMax rAbA            
- Note: The value function (.)V depends also on future interest rates and income tax rates.

### Terminal condition and no-Ponzi constraint
- The household consumes all remaining assets in its last period of life, leaving no bequests. The no-Ponzi condition is:
  - 1 0 R tt R tt TT tT T A   
- Thus the solution in the terminal period is:
  - 11 111 ,log1. RR tttttt RRR tt tttttt TTTTTT tTtT tT TtT TtT T A V a b r           

### Recursive nature emphasized
- The formulation sets up a backward-recursive optimization where the terminal-period solution feeds into earlier-date problems.

---

### Date 2: optimization and first-order condition
- Household at date R_tt + T + 1 (age 1 R_tt TT + -) faces:
  - 1 111 1 2221 1 1 ,log11 ,. RRR tttttttt RRRR R tt tttttttt TT tt R tT T tt R ttt R t tt TTTTTTTT tTtT tT TtT TtT TtT T A TTT tT tT T VAbMax rAbA VAb                
- Plugging the terminal-period solution yields:
  - 1 111 1 2221 1 11 ,log11 log 1. RRR tttttttt RRRR R tt tttttttt TT tt R tT T tt R ttt RR t tttt TTTTTTTT tTtT tT TtT TtT T A TTT tT tT TtT T VAbMax rAbA rAb                      
- First-order condition solved for 1 R tt R tt TT tT T A    :
  - 2 11 21 1 1 1 111 ; 11 1 R ttt RRR t tttttt R tt R tt R tt TTT tT tT T itT TtT T TT i tT T tT T rAr b A r                         
- Plugging that into the date-2 value function gives:
  -   2 1111 1 221 1 ,1log12, RR tttttt RRRR tt ttttttt TTTTTT tTtT tTTtTT itTTtTT i VAbrArb                
- Constant 1 defined as:
  - 1 1 log 11log 1log 1log. R tt tT T r               

---

### Date 3: optimization and FOC
- Household at date R_tt + T + 2 (age 2 R_tt sT T = + -) faces:
  - 1 2 212 11 3332 11 2 ,log11 ,. RRR tttttttt RRRR R tt tttttttt TT tt R tT T tt R ttt R t tt TTTTTTTT tTtT tT TtT TtT T A TTT tT tT T VAbMax rAbA VAb                  
- Replacing the date-2 value function yields an expression with nested terms and rArb interactions:
  -  1 2 212 11 3332 2 11 21 1 ,log11 1log12 RR tttttttt RRRR R tt tttttttt TT tt R tT T tt R ttt RRR t tttttt TTTTTTTT tTtT tT TtT TtT T A TTT tT tT T itT TtT T i VAbMax rAbA rArb                           1 .    
- First-order condition solved for 1 2 R tt R tt TT tT T A   :
  - 32 21 31 111 2 2 2 1 1111(1)2 . 111 R ttt RRRR t ttttttt R tt R tt R tt TTT tT tT T itT TtT T itT T T T T ii tT T tT T i i rArrb A r                               
- Plugging that into the date-3 value function yields:
  - 32 2121 2 2 331 11 ,1 log112, RR ttttttt RRRRR tt tttttttttt TTTTTT tTtT tTTtTT itTTtTT itTT ii VAbrArrb                       
- Constant 2 is defined as:
  - 22 2 12 log 1log 11log 1 RR tttt tT TtT T rr            
  - Additional composition: 1 1log11log1           .

---

### General backward recursion and value function at generic date t
- Repeating the backward procedure, the value function at date t (household’s age is 1 t T +) is:
  -   1 2 1111 1 11 111 ,log11111, RR RR tt tt tttt RRR tttt tttttt TT TT j TT T j tTtTtTtT tT T itT T itT T i jj iii VA bArbrr                               
- Where  is a constant aggregating log and parameter terms across future dates.

### Derivatives of the value function
- Derivative with respect to asset holdings A (A V):
  - 1 1 1 1 2 11 1 111 1 (.), 11(1)11 R R t t R tt RR R tt t tt RRR tt tttttt T T j tT T i j i A TT T j TT tTtT tTTitTTitTTi j iii r V Ar b rr                                           
- Derivative with respect to pension benefits b (b V):
  -  1 2 1 11 11 1 11 111 1(1)1 1 (.) 11(1)11 R RR t tt RR tttt RR tt tt RRR tt ttttt T TT j j tT T itT T i jj ii b TT j TT tTtT tTTitTTitTTi j iii rr V Ar b rr                                             2 1 . R t T          

---

### First-order conditions — tabular summary (as presented)
- Table A1 provides first-order conditions separated into:
  - Consumption-Leisure Decision (Intratemporal condition)
  - Consumption-Saving Decision (Intertemporal condition)
- Conditions are provided for:
  - Working Age (t sT = 1,..., T-1)
  - Retirement (sT = 1,..., 1 R tt T T = + + -)
- Representative intratemporal condition (working age) includes terms:
  - 1 1 11 111 111 ) ( , ) (1 (1 ) T tT ttt tt ts sI Ts T s T s s btT s s c ts ts ts t V We Ab We lc T                         
- Representative intertemporal condition (working age) includes:
  - 1 11 [1 (1 )] (1 ) (1 ) (1 ) I ts ts sc sc ts ts ts ts r cc               
- Representative retirement intratemporal and intertemporal conditions similarly provided with  and  interactions and derivatives of V with respect to A and b.

*Source: _wp12283 - 1. The household’s problem at date 1*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12283.pdf_