Do the elderly reduce housing equity? An international comparison

1. The Luxembourg Wealth Study : enhancing comparative research on household financeRome, 6th July 2007 Do the elderly reduce housing equity? An international comparison Maria Concetta Chiuri* and Tullio Jappelli** * Università di Bari, CSEF and CHILD ** Università di Napoli Federico II, CSEF and CEPR

2. Outline • Motivations • Main contribution of our paper • The evidence to date • The international dataset • Estimating ownership trajectories • Explaining international differences in ownership trajectories

3. Motivation 1 - Policy • Population aging makes of clear policy interest understanding the determinants of decisions over: • how much to save • how to save • should wealth be annuitized, … etc. as people get older. • Real estate is the largest component of total wealth (more than 70% of tot. wealth)

4. Motivation 2 - Theory • Life Cycle Hypothesis (LCH) predicts that with perfect markets selfish individuals run down their wealth in order to smooth consumption over their life-cycle from owning to renting or downsizing. • Bequest motives • Housing as a source of consumption itself

5. Main contribution of our paper • A systematic international comparison of age-trajectories of home-ownership 17 OECD countries, 59 national surveys, years 1974-2000, 300.000 obs. • Empirical test on whether they are explained by differences in financial markets, institutions and public policy.

6. The evidence to date -1 • US • Feinstein and McFadden (1989) use PSID and find transition from owning to renting of <0.3% per year. • Venti and Wise (2002; 2004) HRRS, SIPP, AHDAOO find 1.76%; but about 8% for those with precipitating shocks; they do not depend on h/h composition. Cohort effects are relevant. • Fisher, Johnson, Marchand, Smeeding and Torrey (2007) evidence from CEX that the elderly prefer to stay in their home. • Canada • Crossley and Ostrovsky (2003) use three Canadian surveys and find an annual decline of 0.6% from age 55 to 80.

7. The evidence to date - 2 • UK • Ermisch and Jenkins(1999) use five waves of BHPS and find only rare residential mobility. • Germany • Börsch-Supan (1994) compare data from Germany and the US and find that the decline is similar in the two countries.

8. Summing up – previous evidence • Decumulation, but slow • Importance of cohort effects

9. Table 1- The international dataset: the LIS project

10. -continued

11. Sample selection • Definition of h/h heads is biased: Individuals rather than h/hs e.g. if they move in their children’s place, treated as renters. • Mortality rate and potential entrance in a nursing home: women aged 50-80.

12. Table 2- Sample composition by age-groups

13. Further data treatment • Comparability in educational attainment: ISCED classification • Survey design can vary over time

14. Table3- Ownership by age-group (individuals)

15. Cross-sectional vs. cohort adj. profiles • In a cross-section individuals belong to different generations. • Repetead cross-sections allow to track cohorts over time. Avg. home-ownership rates for 30 age groups (from age 51-80).

16. Estimating ownership trajectories -1 • The restricted model: [1] where: f(a) is a common third order polynomial of age X= educ, marital and work status b=common cohort effect γ=country fixed eff. • Model [1] is estimated with WLS using a robust Var matrix to control for neighborhood effects.

17. Table 4- Regressions for homeownership (N=1595)

18. Figure 1. The cross-sectional and cohort-adjusted profile of homeownership (all countries)

19. Estimating ownership trajectories -2 • The assumption that age and cohort effects are the same in all countries is rather restrictive (an F-test rejected the null at 1 percent level). • A more generalmodel for each single country: [2] where: f(age)is a third-order polynomial in age. b=cohort effect • We plot the difference in cohort-adjusted ownership trajectories between 4 age groups (age 61-65, age 66-70, age 71-75 and age 76-80).

20. Figure 2. The cross-sectional and cohort-adjusted profiles of homeownership

21. Figure 3. Change in ownership: from age-group 61-65 to 66-70

22. Figure 4. Change in ownership: from age-group 66-70 to 71-75

23. Figure 5. Change in ownership: from age-group 71-75 to 76-80

24. Compare with previous findings • Negative values in all contries (after age 70) • Large differences across countries • Can we trace such differences to country characteristics?

25. Explaining international differences in ownership trajectories • Characteristics of rental market, moving costs. • Wealth taxes, property taxes and transaction costs • The generosity of Social Security systems • The local availability of long term care services • Financial markets development

26. Explaining international differences in ownership trajectories • The availability of financial instruments which help house-rich but cash poor old people to release housing equity: • Reverse mortgages • Mortgage equity withdrawal(trading-down, over-mortgaging, re-mortgaging or second-mortgage) • Regulation in financial mkt difficult to distinguish from other economy-wide regulation

27. Table5 - Index of mortgage market and economy-wide regulation, property taxes and no.of beds in nursing homes: international comparisons

28. Theindex of mortgage market regulation is taken from Tsatsaronis and Zhu (2004). The score adds one point for fulfilling each of the following five criteria: (i) Mortgage rate arrangement are primarily extended on the basis of fixed rate contracts; (ii) Mortgage equity withdrawals is absent or limited; (iii) The loan-to-value ratio does not exceed 75 percent, (iv) Valuation methods of property is based on historical values, rather than based on market values (v) Mortgage backed securitization is absent or limited. The index is then normalized to one. • Theindex of economy wide regulation is taken from Kaufman, Kraay and Zoldo Lobaton (1999). The index is a very wide indicator of the degree of economic regulation covering many different regulatory areas (state control, barriers to entrepreneurship, administrative regulations, tariff and non-tariff barriers, etc.) aggregated through factor analysis.

29. Figure 6. Change in ownership and mortgage market regulation Note. Cohort-adjusted change in ownership between age 71-75 and age 76-80.

30. Figure 7. Change in ownership and economy-wide regulation Note. Cohort-adjusted change in ownership between age 71-75 and age 76-80.

31. Table 6 Regressions for change in ownership

32. Sensitivity analysis • Lagged ownership • Overall ownership rate (proxy for thin rental mkt) • Property vs. transaction taxes • Social security income replacement rate • Price to income ratio

33. Conclusion • We estimated the home-ownership rate for the elderly using data from 17 OECD countries. • The analysis at the individual level • Controlling for cohort effects, the ownership rate falls after age 70; after age 75 falls at 1% per year. • Differences across countries are highly explained by the degree of morgage mkt regulation and by the economy-wide regulation. • Credit market imperfections are an explanatory factor for international differences in the aggregate saving rate.