Rob Bauer, Piet Eichholtz, Frank de Jong, Peter. Researchers from Maastricht and Tilburg. New hires: Paul Sengmüller (UvT), Schotman, Bas Werker

Transcription

1 ICPM / Netspar / UM forum Introducing ourselves Maastricht 31 October 2007 Netspar funded research theme Private Retirement Provision Three main research areas Design of pension plans and products Performance of providers The choices people make Private Retirement Provision Researchers from Maastricht and Tilburg Rob Bauer, Piet Eichholtz, Frank de Jong, Peter Schotman, Bas Werker New hires: Paul Sengmüller (UvT), Arvid Hoffmann (UM) More theme presentations later today Other recent topics Managing house price risk and the role of housing futures (De Jong, Driessen and Van Hemert) Long-term house prices and rents (Ambrose, Eichholtz and Lindenthal) Costs and performance of pension funds versus mutual funds (Bauer, Frehen, Lum and Otten) Valuation of pension claims in incomplete markets (De Jong) Demand for and design of annuities (Koijen, Nijman and Werker) Annuity guarantees (Frehen, Hoevenaars, Palm and Schotman) Alternative asset classes (Hoevenaars, Molenaar, Schotman and Steenkamp) Subjective financial well-being (Nies and Schotman)...

2 The choices people make: LevensLoop participation Joint work with Kathrin Nies Very preliminary: no paper yet Todays focus for discussion: Can individuals take care of themselves? Dutch Policy Intervention on Early Retirement Schemes In 2005 the Dutch government abolished the favorable tax treatment of early retirement schemes Discourage early retirement Low labor force partcipation rates of 55+ Can individuals take care of themselves? Behavioral biases in recent literature Massive inertia Tired of making choices Financial illiteracy Procrastination, hyberbolic discounting Mistakes by poor and low educated households LevensLoop arrangement All schemes were replaced by the Levensloop (Life course) arrangement Voluntary Max 12% saving out of gross income Build up wealth up to 210% of income Above 210% no further taxfree savings allowed, only return on wealth in Levensloop account Tax bonus for participating Can be used to finance unpaid leave

3 Why do people participate? NIBUD survey 2007: Early retirement: 67% 25%, predominantly women, for leave of absence Two years after introduction the LevensLoop arrangement is not a success Less than 10% participation Alternative tax-exempt savings arrangements Our focus is on first motive Life-cycle model with early retirement option Who participates and who should participate? If participate: how is money invested? Age and income effects Income effects (+) tax effects (different marginal tax rates) subsistence consumption For low income groups less restrictive tax-friendly savings schemes are more attractive (With standard CRRA utility, everyting is scalable) Age effects (+) Precautionary savings and liquidity more important at younger age Hurst (2003): Preference for leisure dependent on age, marital status and work satisfaction LevensLoop for early retirement The 210% savings allow early retirement by three years (age 62 instead of 65) Additional motive: No early retirement, but saving to fill a pension gap Favorable tax treatment versus reduced liquidity of wealth in LevensLoop account Attractive for people with strong preference for leisure Saving 12% of gross income reduces available income for consumption Early retirement: foregone earnings Data Participation decisions of civil servants in different sectors Background characteristics (age, gender, partner, wage) Zip code: related to wealth Investment decisions of participants Riskfree rate, equity, or mixed Excluded data Special sectors with special arrangements where participation is almost default Wage above 200,000

4 12% 10% Levensloop Participation by Age Methodology Reduced form Probit regressions for participation in LevensLoop 8% 6% LL = f ( age, wage )+ controls + e 4% 2% 0% %LL Controls include dummies on gender, partner, sector, type of offer, zipcode, part-time factor and interactions Age LL = ( α 0 +α 1 ln( wage )) ln( age ) +β ln( wage ) + controls + e Excluding the young Age Effect by Income all PROBIT parameter Age effect is negative for low incomes Income Age Effect by Income all > PROBIT parameter Income Remember the secondary motive!

5 Why are the young different? Remember the secondary motive Maternity leave by females At older age saving for maternity less likely Low income groups have large proportion of parttime workers, mostly females Investment decisions of partcipants Individuals have three basic options Savings account at adjustable rate Mixed guarantee fund Equity Conditional on participation, younger people will choose more equity Some other results Parttime workers Less than 50% less likely to participate Liquidity motive Parttime factor % more likely to participate Preference for leisure Couples less likely to participate Consistent with Dutch evidence that singles are more likely to retire early (Mastrogiacomo, Alessie and Lindeboom, 2003) Younger people hold more equity Equity Investment by Age % holding equity 0.0 <25 [25,30) [30,35) [35,40) [40,45) [45,50) [50,55) [55,60)

6 Who holds equity? Multivariate Probit regressions for holdings in each of three possible categories Results for equity Strongest effects for Income (+)! Age ( ) Parttime workers (+) Men (+) Military (+) Concluding remarks Preliminary first peek at the data Some life-cyle theory predictions seem to hold in broad terms Consistent with LevensLoop as an arrangement for early retirement Missing data Education Pension rights Household income and wealth