population aging, government policy and the postwar ... · the postwar japanese economy aging and...

Population Aging, Government Policy and the PostwarJapanese Economy

Keisuke Otsu and Katsuyuki Shibayama

University of Kent

27 December 2016

Otsu & Shibayama (Kent) Aging 12/27 1 / 33

IntroductionMotivation

Japan has gone through rapid population aging over the past fewdecades

decline in the share of working age population (15-64) among adults(15+)decline in adult population growth rate

How important is population aging and related government policies inaccounting for postwar Japanese growth?


IntroductionPostwar Japanese Growth

1960 1970 1980 1990 2000 2010

Out

put p

er A

dult

500,000

1,000,000

5,000,000

Figure: Real GDP per adult


IntroductionPopulation Aging

1960 1970 1980 1990 2000 2010

Popu

latio

n (1

0 Th

ousa

nds)

0

2000

4000

6000

8000

10000

12000

1564 Years Old65+ Years Old

Figure: Population Share 1955-2014


IntroductionRelated Studies

Neoclassical model for postwar Japanese growth

Christiano (1989), King and Rebelo (1993): capital destruction andsubsistence consumptionHayashi and Prescott (2002), Chen, Imrohoruglu and Imrohoroglu(2006): productivity growth

Demographic effects on Japanese output

Chen, Imrohoruglu and Imrohoroglu (2007), Braun, Ikeda and Joines(2009), Yamada (2012): productivity dominates demographic effects inOLG

Labor decline

Hayashi and Prescott (2002), Yamada (2012): cut in the workweekduring the lost decadeBraun, Ikeda and Joines (2009): reduction in family size


IntroductionContribution

This paper

constructs a parsimoneous neoclassical growth model with young andold adultsquantitatively decomposes Japanese growth 1975-2014 into the effectsof

population agingproductivitygovernment distortions


The Postwar Japanese EconomyDemographic, Productiviy and Government Variables

196 0 197 0 198 0 199 0 200 0 201 0 202 00 .55

0 .6

0 .65Demographics

0 .0 2

0

0 .02

a g i n gp o p u l a t i o n g ro wth

197 5 198 0 198 5 199 0 199 5 200 0 200 5 201 0 201 5750

800

850

900

950

100 0

105 0

110 0Productivity

197 5 198 0 198 5 199 0 199 5 200 0 200 5 201 0 201 50 .12

0 .14

0 .16

0 .18

0 .2

0 .22

0 .24

0 .26

0 .28

0 .3Fiscal Policy

g o v e rn m e n t c o n s u m p ti o nl a b o r ta xc a p i ta l ta x

197 5 198 0 198 5 199 0 199 5 200 0 200 5 201 0 201 570

71

72

73

74

75

76

77Workweek Policy

Figure: Exogenous VariablesOtsu & Shibayama (Kent) Aging 12/27 7 / 33

The Postwar Japanese EconomyAging and Decline in Employment Rate

1960 1970 1980 1990 2000 20100.55

0.6

0.65

0.7

DataProjected

Figure: Aging Effect on Employment Share


ModelOverview

Representative household consists of young and old adults (no kids)

a fraction ηt are young and have high employment rate πya fraction 1− ηt are old and have low employment rate πohead of household allocates resources among the familyThe number of households Nt increases over time at the rate nt

Firm hires labor and capital to produce output

Government taxes the household by labor and capital income tax andlumpsum tax


ModelHousehold Problem

Preferences

U = max∑t

βt[Ψ ln ct + et (1−Ψ) ln

(ht − ht

)], (1)

whereet = ηtπy + (1− ηt )πo

Budget constraint

ct + it = (1− τl ,t )wthtet + (1− τk ,t ) rtkt + ζt , (2)

Capital law of motion

(1+ nt )kt+1 = it + (1− δ)kt , (3)


ModelHousehold Problem

Weekly leisure of the workers (et )

leisuret = ψ ln(restt ) + (1− ψ) ln(weekendt )

where

restt = (ω−ωt )× workweektweekendt = ω× (7− workweekt )

therefore

leisuret = ψ ln(ht − ht

)+ (1− ψ) ln (ω× (7− workweekt ))

whereht = ω× workweekt , ht = ωt × workweekt

weekend is exogenous and does not affect choices (due to seperability)


ModelFirm Problem

ProductionYt = AtK θ

t (htetNt )1−θ ,

soπtNt = Yt − wthtetNt − rtKt ,

or in per family terms

πt = yt − wthtet − rtkt .


ModelGovernment

Government budget constraint

Gt = τl ,twthtetNt + τk ,t rtKt − ζtNt . (4)

where assumeGt = gtYt .

so that(1− gt )yt = ct + it (5)


ModelEquilibirum

Equilibirum conditions

Ψct

= µt

1−Ψht − ht

= µt (1− τl ,t )wt

(1+ nt )µt = βµt+1 {(1− τk ,t+1)rt+1 + 1− δ}

rt = θytkt

wt = (1− θ)ythtet

(1+ nt )kt+1 = it + (1− δ)kt ,

yt = Atkθt (htet )

1−θ

(1− gt )yt = ct + it


Quantitative AnalysisAlgorithm

Shooting algorithm

8 variables {kt+1, µt , ht , yt , ct , it , rt ,wt} ,8 equilibirum conditions for1975-2014specify initial and terminal conditions

initial capital = data in 1975terminal capital = steady state capital given constant productivitygrowth, taxes etc. after terminal period

search for the tragectory of capital that satisfies all equilibirumconditions and the terminal condition


Quantitative AnalysisCalibration

Table 7. Parameter Values

θ Capital Income Share 0.404β Subjective Discount Factor 0.956δ Capital Depreciation Rate 0.088Ψ Preference Weight 0.494


Quantitative AnalysisBenchmark Simulation

197 5 198 0 198 5 199 0 199 5 200 0 200 5 201 0 201 5

106

2 .8

3

3 .2

3 .4

3 .6

3 .8

4

4 .2

4 .4

4 .6

4 .8Output

d a tab e n c h m a rk m o d e l

197 5 198 0 198 5 199 0 199 5 200 0 200 5 201 0 201 538

39

40

41

42

43

44

45

46

47Hours Worked

197 5 198 0 198 5 199 0 199 5 200 0 200 5 201 0 201 5

106

1 .5

2

2 .5

3Consumption

197 5 198 0 198 5 199 0 199 5 200 0 200 5 201 0 201 5

106

0 .5

0 .6

0 .7

0 .8

0 .9

1

1 .1

1 .2

1 .3

1 .4

1 .5Investment

Figure: Simulated Variables: BenchmarkOtsu & Shibayama (Kent) Aging 12/27 17 / 33

Quantitative AnalysisCounterfactual Simulation: Constant Demographics

197 5 198 0 198 5 199 0 199 5 200 0 200 5 201 0 201 5

106

2 .5

3

3 .5

4

4 .5

5

5 .5Output

b e n c h m a rk m o d e lc o n s ta n t a g i n gc o n s ta n t p o p u l a ti o n g ro wth

197 5 198 0 198 5 199 0 199 5 200 0 200 5 201 0 201 536

37

38

39

40

41

42

43

44

45Hours Worked

197 5 198 0 198 5 199 0 199 5 200 0 200 5 201 0 201 5

106

1 .6

1 .8

2

2 .2

2 .4

2 .6

2 .8

3Consumption

197 5 198 0 198 5 199 0 199 5 200 0 200 5 201 0 201 5

105

5

6

7

8

9

10

11

12

13Investment

Figure: Simulated Variables: No Demographic TransitionOtsu & Shibayama (Kent) Aging 12/27 18 / 33

Quantitative AnalysisCounterfactual Simulation: Constant Productivity

197 5 198 0 198 5 199 0 199 5 200 0 200 5 201 0 201 5

106

2 .8

3

3 .2

3 .4

3 .6

3 .8

4

4 .2

4 .4

4 .6

4 .8Output

b e n c h m a rk m o d e lc o n s ta n t p ro d u c t i v i ty

197 5 198 0 198 5 199 0 199 5 200 0 200 5 201 0 201 538

39

40

41

42

43

44

45

46Hours Worked

197 5 198 0 198 5 199 0 199 5 200 0 200 5 201 0 201 5

106

1 .6

1 .8

2

2 .2

2 .4

2 .6

2 .8Consumption

197 5 198 0 198 5 199 0 199 5 200 0 200 5 201 0 201 5

105

5

6

7

8

9

10

11

12Investment

Figure: Simulated Variables: Constant ProductivityOtsu & Shibayama (Kent) Aging 12/27 19 / 33

Quantitative AnalysisCounterfactual Simulation: Constant Fiscal Policy

197 5 198 0 198 5 199 0 199 5 200 0 200 5 201 0 201 5

106

2 .5

3

3 .5

4

4 .5

5

5 .5Output

b e n c h m a rk m o d e lc o n s ta n t g o v e rn m e n t c o n s u m p ti o nc o n s ta n t l a b o r ta xc o n s ta n t c a p i ta l ta x

197 5 198 0 198 5 199 0 199 5 200 0 200 5 201 0 201 536

37

38

39

40

41

42

43

44

45

46Hours Worked

197 5 198 0 198 5 199 0 199 5 200 0 200 5 201 0 201 5

106

1 .5

2

2 .5

3Consumption

197 5 198 0 198 5 199 0 199 5 200 0 200 5 201 0 201 5

105

5

6

7

8

9

10

11

12

13

14Investment

Figure: Simulated Variables: Constant Fiscal PolicyOtsu & Shibayama (Kent) Aging 12/27 20 / 33

Quantitative AnalysisCounterfactual Simulation: Constant Workweek

197 5 198 0 198 5 199 0 199 5 200 0 200 5 201 0 201 5

106

2 .5

3

3 .5

4

4 .5

5

5 .5Output

b e n c h m a rk m o d e lc o n s ta n t wo rk we e k

197 5 198 0 198 5 199 0 199 5 200 0 200 5 201 0 201 538

39

40

41

42

43

44

45Hours Worked

197 5 198 0 198 5 199 0 199 5 200 0 200 5 201 0 201 5

106

1 .6

1 .8

2

2 .2

2 .4

2 .6

2 .8

3Consumption

197 5 198 0 198 5 199 0 199 5 200 0 200 5 201 0 201 5

105

5

6

7

8

9

10

11

12

13

14Investment

Figure: Simulated Variables: Constant WorkweekOtsu & Shibayama (Kent) Aging 12/27 21 / 33

Quantitative AnalysisSummary

Productivity growth is by far the most important driver of growth

Population aging increases hours worked but reduces total labor andhence output by 8.4%

Population shrinking reduced capital dilution and increased output by5.9%

Government consumption increased otuput by 3.9%

Labor income tax reduced output by 8.1%

Workweek shortening reduced output by 9.6%


Population Aging and Structural TransformationData

1975 1980 1985 1990 1995 2000 2005 2010 20150.7

0.8

0.9

1

1.1

1.2

1.3

1.4

1.5

1.6

1.7

Nominal Expenditure RatioRelative Price

Figure: Structural Change Data


Population Aging and Structural TransformationOverview

Representative household consumes goods and services

old relatively prefers services more than younggovernment subsidizes service consumption

Firm produces goods and services

Government taxes the household by labor and capital income tax andlumpsum tax and subsidezes service consumption


Population Aging and Structural TransformationHousehold

Consumption

cy ,t =(

ω1εy c

ε−1ε

yg ,t + (1−ωy )1ε c

ε−1ε

ys ,t

) εε−1,

co ,t =(

ω1εo c

ε−1ε

og ,t + (1−ωo )1ε c

ε−1ε

os ,t

) εε−1,

Budget constraint

ηt (cyg ,t + (1− sy )ptcys ,t ) + (1− ηt )(cog ,t + (1− so )ptcos ,t ) + it= (1− τl ,t )wthtet + (1− τk ,t ) rtkt + ζt ,


Population Aging and Structural TransformationFirms

Production

yg ,t = Ag ,tkθg ,t (hg ,teg ,t )

1−θ ,

ys ,t = As ,tkθs ,t (hs ,tes ,t )

1−θ .

Relative price of services

pt =Ag ,tAs ,t

.


Population Aging and Structural TransformationGovernment

Government budget constraint

Gt = St + G̃t= τl ,twthtetNt + τk ,t rtKt − ζtNt .

whereSt = ηtsyptcys ,t + (1− ηt )soptcos ,t .


Population Aging and Structural TransformationQuantitative Exercise

Table 8. Parameter Values II

ε Consumption Elasticity 0.3ωy Preference Weight Young 0.55ωo Preference Weight Old 0.2sy Subsidy Rate Young 0.1so Subsidy Rate Old 0.25



Nominal expenditure share

ptcs ,tcg ,t

=

(ηt1−so1−sy

ωo1−ωo

((1−so )pt )ε−1+1ωy1−ωy

((1−sy )pt )ε−1+1+ 1− ηt

)(

ηt1+ 1−ωo

ωo((1−so )pt )1−ε

1+ 1−ωyωy

((1−sy )pt )1−ε + 1− ηt

) pt(ωo1−ωo

((1− so )pt )ε) .

where pt and ηt from data



Subsidy share of government consumption

φt =StCt× CtGt

where

StCt

= ηtsy

1− sy1

1+ ωy1−ωy

((1− sy )pt )ε−1

+(1− ηt )so

1− so1

1+ ωo1−ωo

((1− so )pt )ε−1 .

and CtGtfrom data



197 5 198 0 198 5 199 0 199 5 200 0 200 5 201 0 201 50 .7

0 .8

0 .9

1

1 .1

1 .2

1 .3

1 .4

1 .5Nominal Consumption Ratio

Da taBe n c h m a rkNo Ag i n g

197 5 198 0 198 5 199 0 199 5 200 0 200 5 201 0 201 50 .2

0 .25

0 .3

0 .35Government Subsidy Share

Da taBe n c h m a rkNo Ag i n g

Figure: Structural Change SimulationOtsu & Shibayama (Kent) Aging 12/27 31 / 33

Conclusion

A parsimoneous model can capture the effects of demographics,government policy and productivity

Population aging harms growth through

decline in labor participationincrease in social security tax burden

Population aging contributes to structural transformation by

increasing the share of services relative to goodsincrease in government expenditure


Extensions

OLG?: intertemporal inequality

Non-separable utility?: intratemporal inequality

Variable employment rate?: should amplify the result

Population aging and productivity?: endogenous growth?


population aging, government policy and the postwar ... · the postwar japanese economy aging and...

Documents