Retirement, Pension Reform, and Pension Transfer Wealth: An International Comparison Sang-Hyop Lee University of Hawaii at Manoa June 11, 21 Global NTA Conference, Honolulu, HI, USA 1
Research Questions What would be the effect of delaying retirement on economy (saving)? What if the delayed retirement is caused by a reduction in public transfers (e.g. an increase in normal retirement age)? 2
Motivation The age at retirement is usually fixed in saving literature (which focuses on demography). The effect of change in social security benefits on retirement has not been incorporated into a model. Bloom and Canning (27) find that response to a longer life span can take the form of a longer working life or increased savings, but depends on social security arrangements of a country. 3
An Economic Lifecycle Per Capita Consumption and Labor Income 6 5 4 3 2 1 Deficit Labor Income Surplus Large deficits at young and old ages. 2 4 6 8 Age Reallocations from surplus to deficit ages required. Consumption Deficit 4
Reallocations Transfers Public Transfers (Social Security System) Familial Transfers Asset-based Reallocations Interest, dividends, rent from personal assets Home Dis-saving 5
Background (Mason and Lee 27) Population aging can lead to an accumulation of wealth to meet pension needs for retirement (pension wealth) Pension wealth (Wp) is either Asset (A) or Transfer Wealth (Tp). If workers save more (A) in anticipation of aging, higher income is possible even after the first demographic dividend period has come to an end. Alternatively, workers can rely on transfer wealth (Tp), which has little implication on growth. τ = Tp/(Tp+A) plays an important role; countries with low τ leads to high aggregate savings. 6
Innovation What would be the effect of delaying retirement on old-age support? Reduce lifecycle deficit and pension wealth, unless retirees change the level of consumption. If transfer wealth is unchanged Should decrease savings and increases τ (new parameter) τ = Tp/(Tp+A) What if the delayed retirement is caused by a reduction in public transfers? It decreases τ. Delayed retirement increases τ. Depends on the degree of delayed retirement in response to the change in public transfers (σ ). 7
Normal Retirement Age (NRA) For example in the US, beginning with people born in 1938 or after, NRA gradually increases until it reaches 67 for people born after 1959. Can be an attractive option people live longer and healthier people retire early fiscal burden An alternative tool is reducing benefit. 8
Formulization: Basic Setup (Mason and Lee, 27) W( a, t) = PV[ C( a, t)] PV[ Y( a, t)] W() t = W( a,) t a= a W() t = A() t + T () t + T () t W () t = W() t T () t = A() t + T () t τ () t = T ()/ t W () t k At ( ) = (1 τ( t))(1 τ ( t)) W( t) k p k p p ω τ () t = T ()/ t W() t k p k p 9
Lifecycle Wealth Y( a, t) = y( a, t) L( a, t) yat (, ) = yatlat (, ) (, ) y( a, t + x) = y( a, t) G ( t, x) ω a PV[ Y(t) ] = y ( atd, ) ( xg ) ( xlat ) (, + x) x= a= a + x ω a PV[ C(t) ] = catd (, ) ( xg ) ( xnat ) (, + x) x= a= a + x x where D( x) = (1 + r) and G ( x) = (1 + g ) G y ( t, x) = (1 + gy ( t + z)) ω x 1 z = ω y y c y y c y x 1
Lifecycle Wealth (cont d) NTOT (, t x) = N( a, t + x) a= a + x LTOT (, t x) = L( a, t + x) ω ω a= a + x KNTOT(t,x) for children. KLTOT(t,x) ct () yt () = ω a ( )( τ ) ( ) Lt () r g 1 D( x) G (, t x) LTOT (, t x) + KLTOT(t,x) Y x= ω a ( )( τ ) ( ) Nt () r g 1 D( x) G (, t x) NTOT (, t x) + KNTOT(t,x) Y x= y c. 11
Steady State & Backward Recursion (1 + rat ) ( ) + (1 + r)[ Yt ( ) Ct ( )] = At ( + 1) = (1 + g) At ( ) 1+ r At (*) = [ c(*) t Nt (*) y(*) t Lt (*)] r g y ct (*) Lt (*) = 1( + r gy )(1 τ (*)) t wp ( t*) yt (*) Nt (*) ω a At () (1 + r)(1 τ) Dx ( ) c( t 1 + x) ( NTOT ( t 1, x) + KNTOT(t 1,x) ) x= 1 ω a + yt ( 1)(1 + r)(1 τ) Dx () G y() x LTOT ( t 1,) x + KLTOT(t 1,x) Lt ( 1) x= ct ( 1) = Nt ( 1)(1 τ)(1 + rd ) () 1 y ( ) ( ). 12
Data for simulation (195-23) and Assumptions Baseline assumptions Small open economy. Interest: 6% until 2 and decrease linearly to 4.75% until 23 No bequest, no crowing out Productivity growth: 1.5% Familial share to kids:.67 Population UN World Population Prospects 28 for most countries. Medium scenario (instead of high or low) Age profiles Activity rates: various sources National Transfer Accounts database (www.ntaccounts.org) Labor income Consumption Public transfers Public pension benefit, contribution Share of transfer wealth (tau) 13
1.2 1.8.6.4.2 -.2 -.4 Costa Rica 14 75 885 9 6 657 3 354 45 555 Consumption Labor Income Net public pension Age Net public transfers 51 15 225 Norm alized to YL 3-49
9 1.2 1.8.6.4.2 -.2 Indonesia 15 75 885 6 657 Net Public Pension 3 354 45 555 Age 15 225 51 Norm alized to YL 3-49
85 9 1.6 1.2.8.4 -.4 U.S. 16 75 8 7 65 6 55 45 5 Age 4 35 3 25 2 1 15 5 Normalized to YL 3-49
Delayed Retirement by 2 Years Germany (23) Indonesia (25) 35, 18,, 3, 25, 2, 15, 1, 5, P e r c a p it a la b o r in c o m e 16,, 14,, 12,, 1,, 8,, 6,, 4,, 2,, 5 1 1 5 2 2 5 3 3 5 4 4 5 5 5 5 6 6 5 7 7 5 8 8 5 9 1 2 3 4 5 6 7 8 9 17
Labor Income to Consumption after Delaying Retirement by 2 Years (for 65-74) Austria (2) Germany (23) Slovenia (24) Finland (24) Hungary (25) Taiwan (1998) Before After Spain (2) Sweden (23) Uruguay (1994) Japan (24) S.Korea (2) US (23) Chile (1997) Costa Rica (24) Mexico (24) India (24) 5 1 15 2 25 3 35 4 45 18
An Increase in NRA By 2 Years US 12, Per capita pension benefits 1, 8, 6, 4, 2, 1 2 3 4 5 6 7 8 9 Age 19
Countries without Full Pension Benefit S. Korea (2) 1,8 1,6 1,4 1,2 1, 8 6 4 2 1 4 7 1 13 16 19 22 25 28 31 34 37 4 43 46 49 52 55 58 61 64 67 7 73 76 79 82 85 88 91 P e r c a p it a p e n s io n b e n e f it s 18 16 14 12 1 8 6 4 2 S. Korea 1 2 3 4 5 6 7 8 9 2
% Change in Net Public Transfers after an Increase in NRA by 2 Years (for 6+) Indonesia (25) S.Korea (2) Mexico (24) Costa Rica (24) Hungary (25) Finland (24) US (23) Sweden (23) Spain (2) Japan (24). 2. 4. 6. 8. 1. 12. 21
Steady-State Results (Asset to Labor Income Ratio) Baseline Delayed Retirement Increased NRA (σ =) Increased NRA (σ =1) Increased NRA (σ =.5) Costa Rica. -16.9 43.1-7.4 6.1 (τ ) (.625) (.695) (.61) (.669) (.633) Finland. -44.5 32.2-18.9 1.8 (.74) (.814) (.684) (.752) (.717) Japan. -23.2 28.2 1.6 9.1 (.66) (.699) (.597) (.632) (.614) S. Korea. -14.8 11.6-4..1 (.67) (.692) (.642) (.663) (.652) Spain. -23.4 61.4-3.4 1.5 (.56) (.641) (.55) (.578) (.539) U.S.. -8.1 26.3 4.9 1.6 (.35) (.382) (.315) (.344) (.329) 22
Simulation Results (Backward Recursion) 23 South Korea 12 1 A/Yl Wp/Yl W/Yl Tk/Yl( ) 8 6 4 2 2 4 6 21 29 28 27 26 25 24 23 22 21 2 199 198 197 196 195
A/Yl Wp/Yl W/Yl Tk/Yl( ) 24 21 29 Simulation Results (cont d) US 25 2 15 1 5 5 28 27 26 25 24 23 22 21 2 199 198 197 196 195
Summary An increase in NRA raises the asset to labor income ratio, but delaying retirement lowers it. A lot of variation across countries. Age structure of population Labor productivity of older people Public transfers, public pension (Bloom and Canning 27) Under realistic assumptions, the combined effect will raise it. Value of σ: varies but usually range from.1-.2 (e.g. Burtless and Moffitt 1985; Krueger and Pischke 1992) Qualification: need more country data, relax assumptions on crowding-out, bequest, etc. 25
Thank you. 26