Bequest Estimate and Wealth Impact in Japan: Based on a CGE model - - PowerPoint PPT Presentation

▶
bequest estimate and wealth impact in japan based on a
SMART_READER_LITE
LIVE PREVIEW

Bequest Estimate and Wealth Impact in Japan: Based on a CGE model - - PowerPoint PPT Presentation

Oct 27, 2022 368 likes •778 views

Max Planck Institute for Demographic Research Bequest Estimate and Wealth Impact in Japan: Based on a CGE model with realistic demography (Work-in-progress) Miguel Snchez-Romero Naohiro Ogawa Rikiya Matsukura Max Planck

slide-1
SLIDE 1

Max Planck Institute for Demographic Research

Bequest Estimate and Wealth Impact in Japan: Based

  • n a CGE model with realistic demography

(Work-in-progress)

Miguel Sánchez-Romero† Naohiro Ogawa‡ Rikiya Matsukura‡

† Max Planck Institute for Demographic Research (MPIDR) ‡ NUPRI, Nihon University

slide-2
SLIDE 2

Motivation

Japan is at the forefront of population aging ⇒ ↓ labor and production

1950 1975 2000 2025 2050 2075 2100 0.2 0.4 0.6 0.8 1 1.2

Year Aged dependency ratio

Austria Brasil Germany Japan Mexico Spain Sweden Taiwan USA

Source: Authors’ estimations based on local statistics, HFD, HMD, and UN Population Division. Notes: aged group (ages 62+), working group (ages 18-61).

() June 2-8th 2013 Global NTA Conference, Barcelona 2 / 31

slide-3
SLIDE 3

Motivation

Necessity of using additional resources for generating economic growth (mainly through physical capital and human capital) However, there aren’t estimations of bequest in Japan (micro-macro level)

() June 2-8th 2013 Global NTA Conference, Barcelona 3 / 31

slide-4
SLIDE 4

Motivation

Necessity of using additional resources for generating economic growth (mainly through physical capital and human capital) However, there aren’t estimations of bequest in Japan (micro-macro level)

() June 2-8th 2013 Global NTA Conference, Barcelona 3 / 31

slide-5
SLIDE 5

Motivation

Two main questions:

Can we estimate bequest?

◮ Macro and historical: Piketty (2011) for France 1820-2050 ◮ Lifecycle models: Kotlikoff and Summers (1981), Kotlikoff (1988), and Modigliani

(1986, 1988) applied to US

◮ Wealth inequality: general equilibrium models (see literature review by Cagetti and

Nardi (2008)) Can we use bequest to improve economic growth?

◮ Shall savings be annuitized? ◮ Who should receive bequest? ◮ “The tragedy of annuitization” by Heijdra et al. (2010) ⇒ wealth should not be

annuitized and it should be transferred to children

() June 2-8th 2013 Global NTA Conference, Barcelona 4 / 31

slide-6
SLIDE 6

Motivation

Two main questions:

Can we estimate bequest?

◮ Macro and historical: Piketty (2011) for France 1820-2050 ◮ Lifecycle models: Kotlikoff and Summers (1981), Kotlikoff (1988), and Modigliani

(1986, 1988) applied to US

◮ Wealth inequality: general equilibrium models (see literature review by Cagetti and

Nardi (2008)) Can we use bequest to improve economic growth?

◮ Shall savings be annuitized? ◮ Who should receive bequest? ◮ “The tragedy of annuitization” by Heijdra et al. (2010) ⇒ wealth should not be

annuitized and it should be transferred to children

() June 2-8th 2013 Global NTA Conference, Barcelona 4 / 31

slide-7
SLIDE 7

Research goals

Research goals

  • 1. Provide reliable estimates of bequest flows in Japan (using a CGE model with

realistic demography)

  • 2. Give insight on the observed inheritance profiles
  • 3. Give policy recommendations

() June 2-8th 2013 Global NTA Conference, Barcelona 5 / 31

slide-8
SLIDE 8

Research goals

Research goals

  • 1. Provide reliable estimates of bequest flows in Japan (using a CGE model with

realistic demography)

  • 2. Give insight on the observed inheritance profiles
  • 3. Give policy recommendations

() June 2-8th 2013 Global NTA Conference, Barcelona 5 / 31

slide-9
SLIDE 9

Research goals

Research goals

  • 1. Provide reliable estimates of bequest flows in Japan (using a CGE model with

realistic demography)

  • 2. Give insight on the observed inheritance profiles
  • 3. Give policy recommendations

() June 2-8th 2013 Global NTA Conference, Barcelona 5 / 31

slide-10
SLIDE 10

Model set-up

The model set-up

◮ Population ◮ Economic model () June 2-8th 2013 Global NTA Conference, Barcelona 6 / 31

slide-11
SLIDE 11

Model set-up Population reconstruction

  • Population

◮ Single sex model (“population reconstruction”)

  • Inverse projection, (Lee, 1985)
  • Generalized inverse population projection (Oeppen, 1993)

◮ Realistic fertility and mortality (exogenous) ◮ No migration ◮ Information derived from the population reconstruction:

⋆ Adults, children, expected parents, expected number of sibling, expected number of offspring

() June 2-8th 2013 Global NTA Conference, Barcelona 7 / 31

slide-12
SLIDE 12

Model set-up Population reconstruction

1700 1800 1900 2000 2100 2200 30 40 50 60 70 80 90 100

Year Life expectancy at birth

Simulation Data

Life expectancy

1700 1800 1900 2000 2100 2200 1 2 3 4 5 6

Year Total fertility rate

Simulation Data

Total fertility rate

1,880 1,900 1,920 1,940 1,960 1,980 2,000 2,020 2,040 −6 −4 −2 2 4 6

Year Rate (in %)

Net migration rate

20 40 60 80 100 0.5 1.0 1.5 2.0 2.5

Age Population (in millions)

Simulated pop. 1903 Simulated pop. 2000 Simulated pop. 2100 Census 1903 Census 2000

Population distribution Source: Authors’ calculations. UN Population Division, Ministry of Health and Labor of Japan, and Statistics Bureau

  • f Japan.

() June 2-8th 2013 Global NTA Conference, Barcelona 8 / 31

slide-13
SLIDE 13

Model set-up Economic model

Model: CGE OLG model with realistic demography Assumptions: Closed economy, perfect annuity market, no borrowing constraints, and competitive markets

  • Firm: Demands labor (H) and capital (K)
  • Government: Provides goods and services (G) and levies taxes on

{τct,τl,τk,τp,τb}. Our government runs an unbalanced social security pension system

  • Individuals: Maximum life span 120 years, (endog.) work effort, retirement,

saving/consumption (child-rearing cost), and bequest. Preferences similar to Braun et al. (2009) and ˙ Imrohoroˇ glu and Kitao (2012)

() June 2-8th 2013 Global NTA Conference, Barcelona 9 / 31

slide-14
SLIDE 14

Model set-up Economic model

Model: CGE OLG model with realistic demography Assumptions: Closed economy, perfect annuity market, no borrowing constraints, and competitive markets

  • Firm: Demands labor (H) and capital (K)
  • Government: Provides goods and services (G) and levies taxes on

{τct,τl,τk,τp,τb}. Our government runs an unbalanced social security pension system

  • Individuals: Maximum life span 120 years, (endog.) work effort, retirement,

saving/consumption (child-rearing cost), and bequest. Preferences similar to Braun et al. (2009) and ˙ Imrohoroˇ glu and Kitao (2012)

() June 2-8th 2013 Global NTA Conference, Barcelona 9 / 31

slide-15
SLIDE 15

Model set-up Economic model

Model: CGE OLG model with realistic demography Assumptions: Closed economy, perfect annuity market, no borrowing constraints, and competitive markets

  • Firm: Demands labor (H) and capital (K)
  • Government: Provides goods and services (G) and levies taxes on

{τct,τl,τk,τp,τb}. Our government runs an unbalanced social security pension system

  • Individuals: Maximum life span 120 years, (endog.) work effort, retirement,

saving/consumption (child-rearing cost), and bequest. Preferences similar to Braun et al. (2009) and ˙ Imrohoroˇ glu and Kitao (2012)

() June 2-8th 2013 Global NTA Conference, Barcelona 9 / 31

slide-16
SLIDE 16

Model set-up Economic model

Model: CGE OLG model with realistic demography Assumptions: Closed economy, perfect annuity market, no borrowing constraints, and competitive markets

  • Firm: Demands labor (H) and capital (K)
  • Government: Provides goods and services (G) and levies taxes on

{τct,τl,τk,τp,τb}. Our government runs an unbalanced social security pension system

  • Individuals: Maximum life span 120 years, (endog.) work effort, retirement,

saving/consumption (child-rearing cost), and bequest. Preferences similar to Braun et al. (2009) and ˙ Imrohoroˇ glu and Kitao (2012)

() June 2-8th 2013 Global NTA Conference, Barcelona 9 / 31

slide-17
SLIDE 17

Model set-up Economic model

⋆ Economic unit (double-head “pseudo-household”)

◮ Two adults (2 heads) ◮ Dependent children ◮ Economic decisions:

  • 1. Consumption/saving
  • 2. Intensive and extensive labor supply (work effort, retirement age)
  • 3. Bequest

◮ Assumptions:

  • 1. No economies of scale
  • 2. All resources are equally distributed within the heads
  • 3. All individuals are paired with an individual of the same age when they become adults
  • 4. Exit from marriage can only occur because of death

() June 2-8th 2013 Global NTA Conference, Barcelona 10 / 31

slide-18
SLIDE 18

Model set-up Economic model

Calibration

1900 1950 2000 2050 2100 1 2 3 4 5

Year Ratio

No bequest motive Bequest motive I Bequest motive II Bequest motive III Data Figure: Capital-output ratio, period 1885-2100, Japan

() June 2-8th 2013 Global NTA Conference, Barcelona 11 / 31

slide-19
SLIDE 19

Model set-up Economic model

20 40 60 80 100 0.5 1.0 1.5

Age Ratio to labor income ages 30−50

No bequest motive Bequest motive I Bequest motive II Bequest motive III NTA data

Consumption and labor income, 1984

20 40 60 80 100 0.5 1.0

Age Ratio to labor income ages 30−50

No bequest motive Bequest motive I Bequest motive II Bequest motive III NTA data

Consumption and labor income, 1994

20 40 60 80 100 0.5 1.0 1.5

Age Ratio to labor income ages 30−50

No bequest motive Bequest motive I Bequest motive II Bequest motive III NTA data

Consumption and labor income, 2004

() June 2-8th 2013 Global NTA Conference, Barcelona 12 / 31

slide-20
SLIDE 20

Bequest estimation

Comparison of our model to JSTAR data

() June 2-8th 2013 Global NTA Conference, Barcelona 13 / 31

slide-21
SLIDE 21

Bequest estimation

20 40 60 80 100 120 0.02 0.04 0.06 0.08 0.10

Age Hazard rate

First parent Both parents Second parent Spouse Total (sum) JSTAR raw data JSTAR smoothed Cases = 216 Exposures = 3946 Figure: Inheritance hazard rate, year 2009

() June 2-8th 2013 Global NTA Conference, Barcelona 14 / 31

slide-22
SLIDE 22

Bequest estimation

20 40 60 80 100 120 0.1 0.2 0.3

Age Relative to labor income ages 30−50

Bequest motive I Bequest motive II Bequest motive III JSTAR raw data JSTAR smoothed Cases = 216 Exposures = 3946 Figure: Average bequest received, year 2009

() June 2-8th 2013 Global NTA Conference, Barcelona 15 / 31

slide-23
SLIDE 23

Bequest estimation

20 30 40 50 60 70 80 90 100 110 −2 2 4 6 8 10

Age Ratio to labor income ages 30−50

Bequest motive I Bequest motive II Bequest motive III Data Figure: Assets profile, year 2009

() June 2-8th 2013 Global NTA Conference, Barcelona 16 / 31

slide-24
SLIDE 24

Results

The estimation of bequest in Japan from year 1885 to 2100

() June 2-8th 2013 Global NTA Conference, Barcelona 17 / 31

slide-25
SLIDE 25

Results 1900 1950 2000 2050 2100 5 10 15

Year Rate (in %)

Bequest motive I Bequest motive II Bequest motive III

Figure: Bequest to output ratio (period 1885-2100), Japan

U-shaped pattern

◮ Piketty (2011, QJE): r > n +ρ logic ◮ Alternative and complementary

reasons from demography:

  • Decline
  • Rapid population growth ↓ K/N
  • “Tempo effect” postponement of

inheritance

  • ↓ precautionary saving (↓ variability of

the age at death)

  • Increase
  • Declining population ↑ K/N
  • ↑ saving for retirement motive (↑ eR)

() June 2-8th 2013 Global NTA Conference, Barcelona 18 / 31

slide-26
SLIDE 26

Results

20 40 60 80 100 120 0.05 0.10 0.15 0.20 0.25

Age Ratio to labor income ages 30−50

Bequest motive I Bequest motive II Bequest motive III

Year 1900

20 40 60 80 100 120 0.05 0.10 0.15 0.20 0.25

Age Ratio to labor income ages 30−50

Bequest motive I Bequest motive II Bequest motive III

Year 2000

20 40 60 80 100 120 0.05 0.10 0.15 0.20 0.25

Age Ratio to labor income ages 30−50

Bequest motive I Bequest motive II Bequest motive III

Year 2100

Figure: Simulated evolution of the bequest profile by bequest motive (selected years), Japan

() June 2-8th 2013 Global NTA Conference, Barcelona 19 / 31

slide-27
SLIDE 27

Results

Counterfactual experiment I/II

() June 2-8th 2013 Global NTA Conference, Barcelona 20 / 31

slide-28
SLIDE 28

Results

Inheritance law change in year 2015

  • Three alternatives
  • 1. Offspring-Spouse (O-S) ⇒ 100% - 0%
  • 2. Offspring-Spouse (O-S) ⇒ 50% - 50%
  • 3. Offspring-Spouse (O-S) ⇒ 0% - 100%

() June 2-8th 2013 Global NTA Conference, Barcelona 21 / 31

slide-29
SLIDE 29

Results

20 40 60 80 100 120 0.05 0.10 0.15 0.20 0.25

Age Ratio to labor income ages 30−50

O−S 100%−0% O−S 50%−50% O−S 0%−100%

Bequest profile, year 2015

1900 1950 2000 2050 2100 2 3 4 5

Year Ratio

O−S 100%−0% O−S 50%−50% O−S 0%−100%

Capital-output ratio, period 1885-2100

() June 2-8th 2013 Global NTA Conference, Barcelona 22 / 31

slide-30
SLIDE 30

Results

Counterfactual experiment II/II

“tragedy of annuitization: although full annuitization of assets is privately

  • ptimal it may not be socially beneficial due to adverse general equilibrium

repercussions” [Heijdra et al. (2010), p. 3] Thought experiment: mandatory annuitization of 50% of private assets from year 2015 onwards

() June 2-8th 2013 Global NTA Conference, Barcelona 23 / 31

slide-31
SLIDE 31

Results

1900 1950 2000 2050 2100 2150 2200 2 3 4 5

Age Ratio

Benchmark 50% of private annuitization

Bequest motive I

1900 1950 2000 2050 2100 2150 2200 2 3 4 5

Year Ratio

Benchmark 50% of private annuitization

Bequest motive II

1900 1950 2000 2050 2100 2150 2200 2 3 4 5

Year Ratio

Benchmark 50% of private annuitization

Bequest motive III

() June 2-8th 2013 Global NTA Conference, Barcelona 24 / 31

slide-32
SLIDE 32

Conclusion

Conclusions

◮ Bequest profiles can be estimated using CGE models with realistic demography ◮ Inheritance in Japan also presents a U-shaped pattern similar to that in France (≈

10% before 1950, 5% 1970-2000, 7%-12% from 2050-)

◮ We provide an alternative and complementary explanation based on demography

for the U-shaped pattern given by Piketty (2011)

◮ We find similar results shown by Heijdra et al. (2010), known as “The tragedy of

annuitization” → no annuitization and ↑ share of transfers to children

() June 2-8th 2013 Global NTA Conference, Barcelona 25 / 31

slide-33
SLIDE 33

Conclusion

Thank you

The authors would like to thank Ronald Lee, Andrew Mason, and Hidehiko Ichimura for valuable comments.

() June 2-8th 2013 Global NTA Conference, Barcelona 26 / 31

slide-34
SLIDE 34

Bequest estimation

Estimation of bequest

() June 2-8th 2013 Global NTA Conference, Barcelona 27 / 31

slide-35
SLIDE 35

Bequest estimation

Bequest: Part I/II

...

pt(x)pt(x) 2pt(x)qt(x) pt(x) qt(x)qt(x) qt(x)

Figure: Expected bequest given, by partnership status and age

Bequest given at age x depends on

◮ Age ◮ Partnership status {married,

widow/er}

◮ Number of eligible offspring ◮ Assets holding ◮ Inheritance law () June 2-8th 2013 Global NTA Conference, Barcelona 28 / 31

slide-36
SLIDE 36

Bequest estimation

Bequest: Part II/II

...

  • ✁
✂ ✄

ft(x + α) ft(x + α + 1) ft(x + α + 2) ft(x + β)

Figure: Expected bequest received from parent(s), by age

Bequest received at age x depends on

◮ Age of the expected parent ◮ Status of the parent {married,

widow/er}

◮ Assets held by parent(s) ◮ Own marriage status ◮ Assets held by spouse ◮ Inheritance law () June 2-8th 2013 Global NTA Conference, Barcelona 29 / 31

slide-37
SLIDE 37

Bequest estimation

“Head’s” problem

V(ax;z) = max

cx ,ℓx

  • u(cx,1−ℓx;ηc

x ,ηℓ x)+β

  • px+1V(ax+1;z)+(1−px+1)UB (˜

ax+1)

  • (1)

s.t. ax+1 =       

  • Rx
  • 1+γ qx

px

  • −τp
  • ax +(Rx −τb)Bx +(1−τl)(1−ςτs,x)ωεxℓx −(1+τc,x)cx

if working,

  • Rx
  • 1+γ qx

px

  • −τp
  • ax +(Rx −τb)Bx +(1−τl)bx(z)−(1+τc,x)cx

if retired, where ˜ a is the effective bequest left (or (1−γ)(1−τb)a), R is the compound (real) interest rate net

  • f capital income tax, or 1+r(1−τk), and γ ∈ [0,1] is the percentage of private savings that are

annuitized.

First-order conditions

  • Optimal consumption (Euler equation)

uc(x) uc(x +1) = βpx+1

  • Rx+1
  • 1+γ qx+1

px+1

  • −τp

1+τc,x 1+τc,x+1 +β(1+τc,x) ˜ ax+1 ax+1 UB

a (x +1)

uc(x +1)

  • Optimal work effort

u1−ℓ(x)/uc(x) = ωεx(1−tx), where tx = (1−τl)(1−ςτs,x)/(1+τc,x)

  • Optimal retirement age

z∗ = arg max

z∈Z V(ax0;z) () June 2-8th 2013 Global NTA Conference, Barcelona 30 / 31

slide-38
SLIDE 38

Bequest estimation

“Head’s” problem

V(ax;z) = max

cx ,ℓx

  • u(cx,1−ℓx;ηc

x ,ηℓ x)+β

  • px+1V(ax+1;z)+(1−px+1)UB (˜

ax+1)

  • (1)

s.t. ax+1 =       

  • Rx
  • 1+γ qx

px

  • −τp
  • ax +(Rx −τb)Bx +(1−τl)(1−ςτs,x)ωεxℓx −(1+τc,x)cx

if working,

  • Rx
  • 1+γ qx

px

  • −τp
  • ax +(Rx −τb)Bx +(1−τl)bx(z)−(1+τc,x)cx

if retired, where ˜ a is the effective bequest left (or (1−γ)(1−τb)a), R is the compound (real) interest rate net

  • f capital income tax, or 1+r(1−τk), and γ ∈ [0,1] is the percentage of private savings that are

annuitized.

First-order conditions

  • Optimal consumption (Euler equation)

uc(x) uc(x +1) = βpx+1

  • Rx+1
  • 1+γ qx+1

px+1

  • −τp

1+τc,x 1+τc,x+1 +β(1+τc,x) ˜ ax+1 ax+1 UB

a (x +1)

uc(x +1)

  • Optimal work effort

u1−ℓ(x)/uc(x) = ωεx(1−tx), where tx = (1−τl)(1−ςτs,x)/(1+τc,x)

  • Optimal retirement age

z∗ = arg max

z∈Z V(ax0;z) () June 2-8th 2013 Global NTA Conference, Barcelona 30 / 31

slide-39
SLIDE 39

Calibration

Table: Model economy parameters

Symbol Value Source Household heads Risk aversion parameter σ {2.5;3.0;3.50} Weight on consumption φ 0.35 Weight on bequest utility ψ1 {0;20;40;60} Curvature of bequest utility ψ2 0.40AΩ Subjective discount factor β 1.00 Age at leaving parent’s home x0 20 Employee social contribution share ς 0.50 Technology Capital share α 0.363 Hayashi and Prescott (2002), Chen et al. (2007), Braun et al. (2009) Depreciation rate δ 5.00% National accounts Future labor-aug. techn. progress dAt/At 1.00% Labor efficiency profile εx Braun et al. (2009) Government Public consumption to output G/Y 0.12 National accounts Capital income tax rate τk 0.150 OECD Labor income tax rate τl 0.075 OECD Property tax rate τp 0.005 OECD Bequest tax rate τb 0.100 OECD

() June 2-8th 2013 Global NTA Conference, Barcelona 31 / 31