|
|
Universitą degli Studi di Salerno |
|
Economics
of Ageing in Europe (AGE)
RTN
European
Program HPRN-CT-2002-00235
The Sixth Workshop of the RTN on the Economics of Aging in Europe was held in Frankfurt on May 12-14 2005. It hosted by the Goethe University Frankfurt and sponsored by the European Union Marie Curie Research Training Networks and the Centre for Financial Studies (CFS), under its new program on Household Wealth Management. The conference organiser was Michael Haliassos of Goethe University Frankfurt and Director of the CFS program.
The first day of the workshop, Thursday May 12th, was devoted to training lectures for the young researchers of the AGE Network. The first lecture was given by Orazio Attanasio on “Global Demographic Trends and Social Security Reform” (co-authored paper with S. Kitao and G. Violante). Their work focused on the issue of the sustainability of PAYG systems and the transition to a privatized pension system. They explored the implications of different ways to pay for pensions and/or the transition for the welfare of different cohorts. Then, they presented how their results were affected by considering an open versus a closed economy. They showed that demographic trends make the current social security system unsustainable, but different policy reforms have different intragenerational distribution consequences. Their main conclusion was that privatizing the social security system has large implications and might be very costly for some generations.
The second lecture was given by Deborah Lucas on “Investing Public Pensions in the Stock Market: Implications for Risk Sharing and Asset Prices” (joint work with J. Heaton). They used a computational general equilibrium OLG model with many heterogeneous agents, calibrated to examine the various implications of pension investment policy changes similar to those currently under consideration in the U.S. They showed that predicted assets returns are fairly insensitive to whether stocks are held in the social security trust fund, independent of the details of policy implementation. However, the implications for risk sharing and welfare are quite sensitive to the details, and sometimes counter-intuitive. According to their analysis the risk redistribution goals can be accomplished by program rules that mimic financial derivatives (as an alternative to costly private accounts).
“How Do Household Portfolio Shares Vary with Age” was the third invited lecture, given by Stephen Zeldes (joint work with J. Ameriks). They examined the questions of how should portfolio allocations change with age and how do portfolio allocations change with age. In addressing the second issue, the authors explored the relative importance of age, time and cohort effects using US household level data. Their main finding was that there is no evidence supporting gradual decline in equity shares with age. There was only limited indication of people shifting completely out of equity around retirement as they begin to withdraw or annuitize their accumulation.
The second and third days of the workshop were devoted to contributed papers. The first paper was on “Understanding Saving and Portfolio Choices with Predictable Changes in Asset Returns”, and it was presented by Christian Gollier. The author described how existing results on household portfolio literature can be generalized to non-CRRA preferences. He also explored why unintuitive results hold when RRA is less than unity. He showed that both savings and portfolio choices are affected by the expectation of changes in the future opportunity set. Andre Masson discussed the relevance and usefulness of results in the paper. The second presentation was on “Preferences and the Dynamic Representative Consumer” by Christos Koulovatianos. He presented a model that leads to a representative consumer with a time-separable utility function, in a single-commodity-type deterministic dynamic environment, in the presence of consumer wealth-, labor-productivity, and preference heterogeneity. He showed that when the rates of time preference are heterogeneous across consumers, a representative consumer exists if, and only if, the momentary utility functions of all consumers are exponential. The discussant, Dirk Krueger, highlighted the usefulness of the results, cautioning also that if some strong conditions of the theory do not apply the results may not help to understand quasi-aggregation.
The next paper was on “Risk-Return Preferences in the Pension Domain: Are People able to Choose?” and given by Maarten van Rooij (joint work with C. Kool and G. Prast). They investigated respondents’ attitudes toward DB and DC pension schemes using data from Dutch households. In addition they explored whether respondents have consistent preferences with respect to portfolio choices for retirement saving. They found that the vast majority of households is in favor of compulsory saving for retirement and opts for a DB pension system. They also showed that given investor autonomy, agents face problems in bringing choices in line with their preferences. The discussant, Dimitris Georgarakos, pointed out the possible role of wealth effects that were not accounted for in the regression analysis and suggested some further comparisons with stock ownership and trading decisions. “How Do Risk Attitudes Change with Wealth? Nonparametric Evidence from a Hypothetical Gamble” was the fourth paper, presented by Juergen Maurer. The main aim of his study was to illuminate the relationship between risk aversion and wealth, using households’ response to a hypothetical gamble and nonparametric estimation techniques. He found that absolute risk aversion declines slowly with wealth. The discussant Antoine Bommier referred to the limitations that existing household data seem to impose on such an analysis.
The next presentation was on “Using Stated Preferences Data to Analyze Preferences for Full and Partial Retirement” given by Arthur van Soest (co-authored with A. Kapteyn and J. Zissimopoulos). They described an experiment with measuring retirement opportunities as perceived by the respondents, as well as preferences for retirement. For the latter, respondents evaluated how attractive they found a number of virtual, simplified, retirement trajectories involving early retirement, late retirement, and gradual retirement with different corresponding income paths. The data on preferences were used to estimate a stylized structural model of retirement decisions. They found that people are reasonably satisfied with retiring at the benchmark age and that there is no strong preference for phased retirement. In addition if late retirement were to be made more attractive by providing financial incentives, these incentives should be quite strong and should exceed by far the actuarially fair adjustments. The discussant, Ralf Wilke, pointed out that the results are based on subjective valuations and that further robustness checks should be made. Rob Euwals presented the next paper on “Early Retirement Behavior in the Netherlands: Evidence from a Policy Reform” (co-authored with D. van Vuuren and R. Wolthoff). Using Dutch panel data, the authors estimated hazard rate models for early retirement. They found evidence that the policy reform induces workers to postpone early retirement. They showed that the transitional scheme has already led to average retirement postponement by 8 months, which will become almost a year once the transition is completed. The discussant Bernd Fitzenberger referred to the relevance of the empirical findings and to some useful extensions.
Carolina Fugazza presented a paper on “Investing for the Long-run in European Real Estate: Does Predictability matter?” (joint work with M. Guidolin and G. Nicodano). The authors derived optimal portfolio shares under excess return predictability and parameter uncertainty, including real estate in the menu of assets. They found that the weight for real estate is between 10% and 30% for intermediate values of RRA and that the welfare costs of ignoring real estate are increasing with RRA and planning horizon. The discussant, Mario Padula, referred to the possible importance of transaction costs and no short sale constraints. The last paper was on “Temperant Portfolio Choice with a Correlated Background Risk” given by Hector Calvo (joint work with L. Arrondel). They investigated the impact of income risk on the demand for risky assets using data on French households. They showed that the probability of stock market participation increases for those with negatively correlated incomes while they did not identify any effect with regard to those with positively correlated or uncorrelated incomes. The discussant Christian Gollier showed that when the background risk is perfectly correlated with the return on the risky asset, the risky asset can be used to perfectly hedge the background risk.
On the 3rd day of the workshop, May 14th, the first paper was on “The Forgone Gains of Incomplete Portfolios” presented by Monica Paiella. She estimated a lower bound to the forgone gains of incomplete portfolios, which are in turn a lower bound to the entry costs that could rationalize non-participation to financial markets. She showed that such estimated bound can provide a heuristic test for the cost-based explanation of limited financial market participation, since high estimates would imply unrealistically high participation costs. The main finding was that the participation cost explanation of limited stock market participation can not be rejected. The discussant Arthur van Soest made some suggestions on the empirical strategy and the power of the test. “The Impact of Interest Rate Subsidies on Long-Term Household Debt: Evidence from a Large Program” was the second paper, presented by Ernesto Villanueva (co-authored with N. Martins). They explored whether mortgage interest rate subsidies affect the demand of long-term household debt, utilizing Portuguese micro data. They concluded that borrowing among groups on the margin of home-ownership responds to interest rates incentives. The discussant Jan van Ours suggested that the analysis adopts a static approach and that it is worth exploring dynamics. The next paper was on “Varying Life Expectancy and Social Security” given by Antoine Bommier (joint work with M. Leroux and J. Lozachmeur). They showed that heterogeneity of life expectancy plays a key role in the design of social security. In their framework the social optimum is obtained when individuals living longer retire later and consume less than short lived individuals. The discussant DIMITRIS Tsomocos made suggestions for the extension of their research.
The fourth paper was on “Obesity, Health and Socio-economic Status: an International Comparison” presented by Pierre-Carl Michaud (joint work with A. van Soest). They studied the association between obesity, health, use of health care, and socio-economic status and explored some explanations for the cross country differences. Their empirical findings based on SHARE data from 10 European countries and HRS from the US. Their main conclusion was that only a small share of the difference in average body mass index between the U.S. and Europe can be explained by differences in health behaviors and food consumption. The discussant Tullio Jappelli suggested the closer examination of reasons behind the observed heterogeneity across European countries. The last presentation was on “Demand Patterns around Retirement: Evidence from Spanish Panel Data” by Mette Lunde Christensen. She examined demand patterns before and after retirement, using a Spanish panel data set on household expenditures, in which households are followed across the retirement threshold. She found no evidence for an income fall for the retiring households. She also examined the effect of retirement on budget shares. She found no significant effect on any commodity groups, with only exception being the fall in share of medicine expenditures. The discussant, Arie Kapteyn, pointed out that pensions of retirees seem to be even higher than earnings and that the incidence of high unretirement may be due to noise.