Riester Pensions: the Matching Defined Contribution Savings Scheme in Germany

Transcription

1 Riester Pensions: the Matching Defined Contribution Savings Scheme in Germany Axel Börsch-Supan, Michela Coppola,Anette Reil-Held,Daniel Schunk Munich Center for the Economics of Aging Max Planck Institute for Social Law and Social Policy 1

2 Outline 1. How are Riester pensions designed? 2. How have Riester pensions developed since 2001? 3. Does the targeting to families with children and low-income individuals work? 4. Crowding in/out w.r.t. other savings (a) other private pension schemes (b) financial wealth, housing, bequests 2

3 1. Design of Riester pensions: Tax credits and deductions Table 1: State incentives for supplementary pension provision. Maximum contribution [percentage of gross earnings] Basic benefit [ p. a.] Child benefit [ p. a.] Maximum tax deduction [ p. a.] % /05 2% % %

4 1. Design of Riester pensions: Extent of matching Figure 5: Depth of subsidies Subsidy as to Riester percent pensions of total (!) contribution % 100 Mean= Euro Single, no children Single, one child Married, one income, two children Annual earnings (Euro) Note: Direct subsidy/the tax advantage as a percentage of savings in form of the new supplementary pensions. Source: Deutsche Bundesbank (2002). 4

5 2. Riester pensions: Uptake 16 mio contracts Home-annuity contracts (Wohn-Riester) Investment funds contracts Bank savings contracts Insurance contracts

6 2. Uptake relative to other private pension instruments 45% 40% 35% 30% 25% 20% 15% 10% 5% 0% Spread of private pension provision Source: SAVE % 31% 33% 26% 27% 23% 25% 23% 21% 19% 19% 23% 13% 13% 16% 15% 15% 15% 15% 7% 11% Riester-pension Occupational pensions Private-pensions 6

7 2. Multiple private instruments 80% 70% 60% 50% 40% 30% 20% 10% 0% 73% Households with multiple pension schemes 63% 53% 54% Source: SAVE % 49% 45% 28% 30% 27% 19% 23% 20% 19% 19% 18% 18% 20% 14% 13% 10% 9% 2% 4% 4% 5% 5% 5% No old-age provision One Instrument Two Instruments Three Instruments 7

8 2. Uptake by age Anteil der Haushalte % 20% 30% 40% Alter Age Age Alter Alter Age SAVE 2003 SAVE 2005 SAVE 2006 SAVE 2007 SAVE 2008 SAVE

9 16 mio contracts What happened in 2005?... and what in 2008? 2005: Design simplification 2008: Financial crisis, 4% limit reached Home-annuity contracts (Wohn-Riester) Investment funds contracts Bank savings contracts Insurance contracts

10 3. Targeting: Families with children 80% 70% 60% Spread of Riester-pensions by number of children Source: SAVE % 40% 30% 20% 10% 0% No children 1 Child 2 Children 3+ Children SAVE 2003 SAVE 2005 SAVE 2006 SAVE 2007 SAVE 2008 SAVE 2009 SAVE

11 3. Targeting: Families with children 80% 70% 60% 50% 40% 30% 20% 10% 0% Pension provision by number of children Proportion of households with private pension schemes Source: SAVE % 68% 38% 27% 25% 29% 30% 25% 14% 14% 18% 17% No children 1 Child 2 Children 3+ Children Riester-pension Occupational pension Other private pensions all instruments higher, but especially Riester 11

12 3. Targeting: Household income quintiles 60% 50% 5.3 times 5.6 times 4.6 times 40% 30% 20% 10% 0% Q1 Q2 Q3 Q4 Q5 SAVE 2003 SAVE 2005 SAVE 2006 SAVE 2007 SAVE 2008 SAVE 2009 SAVE

13 70% 60% 50% 40% 30% 20% 10% 0% 19% 3. Targeting: Household disposable income Private pension provision by disposable income Proportion of households with private pension schemes 53% 47% 48% Soource: SAVE % 4% 30% 15% 13% 35% 18% 23% > 4000 Riester-pension Occupational pension Other pension schemes -all instruments higher for the wealthier -Riester especially high among low income 13

14 4. Crowding in/out/what? Pay-as-you-go pensions (1) Total saving Retirement saving Other saving (2) (3) Figure 2: Substitution among savings types ( Crowding out )

15 4. Crowding in/out/what w.r.t. to other pension instruments Bivariate probit specification: Specification A: income in quintile dummies Specification B: income in quadratic Riester Other private pensions Riester Other private pensions (1) (2) (3) (4) Mc-Fadden R² Rho [Chi²(1)] [1.32] [1.54] Number of observations Absolute value of z statistic in parentheses, * significant at 10%, ** at 5%, *** at 1% confidence. Covariates include HH demographics, income, education, occupation, financial literacy et al. -positive correlation means crowding in! 15

16 4. Crowding in/out/what w.r.t. to non-pension saving Riester Specification A Other private pensions Riester Specification B Other private pensions Occupation: Self-employed (dummy) (0.88) (4.12)*** (0.81) (4.16)*** Unemployed (dummy) (0.67) (1.24) (0.35) (1.58)* Income: Disposable income: 1. Quintile (1.45) (2.69)*** Disposable income: 2. Quintile (0.25) (-1.91)* Disposable income: 3. Quintile Reference category Disposable income: 4. Quintile (1.41) (0.68) Disposable income: 5. Quintile (1.55) (0.04) Disposable income: (0.02) (1.81)* (Disposable income)^ (0.74) (0.99) Wealth: Net financial assets (0.76) (3.31)*** (0.85) (3.02)*** (Net financial assets) (0.46) (-2.52)** (0.46) (2.39)** Property owner (dummy) (0.93) (1.06) (0.77) (0.97) -crowding out for saving towards inheritance & housing -crowding in (if at all) for general saving 16

17 4. Crowding in/out/what w.r.t. to non-pension saving Aggregate private saving rate: ,4 9,9 10,3 10,4 10,5 10,6 10,8 11,7 11,1 11,4 Thus: evidence tends to support crowding in also for general saving

18 1. New instruments need time: Dynamism only after slow start and substantial simplification of the subsidy design 2. Depth of subsidies could not compensate for design flaws 3. Uptake only partially follows subsidy depth (families with children vs. low-income individuals) 4. Nevertheless. Uptake also increasing in lowest quintile 5. Crowding out: housing, bequests Policy conclusions 6. Crowding in: occupational pensions and other private pensions, and (tendency!) also general and thus total saving 18