1. Alternative approaches to Long Term Care financing. Distributive implications and sustainability for Italy Massimo Baldini and Luca Beltrametti European Conference on Long-Term Care Mannheim, October 21-22, 2005

2. TOPICS What are the distributive implications of alternative options for LTC financing? –Among income groups –Among age groups What about intergenerational distribution? –Financial sustainability and intergenerational fairness

3. Six (stylised) policy options 1) Germany 1: 1,7% tax on income, with a ceiling 2) Germany 2: 1,7% with children, 2% without 3) Germany 3: flat tax (a fixed amount not dependent on income) 4) Luxembourg: a tax proportional to income, without ceilings 5) Japan: only taxpayers aged 40+ pay 6) Ise: a tax proportional to the “Indicatore della situazione economica” (income and wealth)

4. What is ISE? Ise is an indicator of economic situation introduced to means test the access to welfare services Why a tax on Ise? –Wealth increases with age –The older generations pay more not because of their higher level of risk, but in relation to their economic strenght

5. The distributional analysis A sample of microdata representative of the Italian population taken from the Bank of Italy survey, 2002 Microsimulation of disposable incomes The distributive implications are evaluated on disposable equivalent income: YEQ = Y / (family members) 0,65 The tax unit is always the individual All simulations are carried out imposing the same total revenue of the first simulation (9.2 billion euro). All the other stylized differences across the six cases are preserved

6. RateCeiling % exempt taxpayers Average amount paid for taxpayers with positive tax (€) % exempt households Average amount paid for households with positive tax (€) 1) Germany 11.70%€711/year32%34519%517 2) Germany 2 1.54% with children 1.81% no children €711/year32%34519%517 3) Germany 3345 euro/year/32%34519%517 4) Luxembourg1.49%No32%34519%517 5) Japan2.18%No54%51236%660 6) Ise2.31%€711/year32%34518%511 Table 1: Some basic characteristics of the simulated contributions (p.10)

7. Chart 1: Share of households which do not pay the tax, by deciles of equivalent disposable income (p.10)

8. Chart 2.1: Average incidence by deciles of equivalent disposable income all households

9. Chart 2.2: Average incidence by deciles of equivalent disposable income all households

10. Age class% taxpayersGermany 1Germany 2Germany 3LuxembourgJapanIse <201.30.50.60.80.50.00.2 21-254.23.13.34.12.90.02.3 26-306.6 6.87.26.20.05.4 31-359.010.310.410.010.40.09.5 36-4010.411.911.510.611.60.010.4 41-4510.613.112.611.513.519.811.3 46-509.411.811.49.912.217.810.2 51-559.011.010.79.211.717.010.4 56-607.58.18.27.88.812.98.4 61-657.87.17.37.76.910.18.9 66-707.85.75.97.15.68.27.3 71-756.64.64.85.94.26.26.4 76-805.13.23.44.22.94.24.7 80+4.72.93.14.02.63.84.6 Total100 <4031.532.432.632.731.60.027.9 41-6544.351.150.246.153.177.749.2 65+24.216.517.321.315.322.323.0 Table 2: Percentage composition of taxpayers by age classes and contribution of each class to total revenue

11. Main results  The poorest decile woud be exempt under all alternatives  A significant share of the richest 50% of the popolutaion exempt only in the Japanese case (Chart 1).  In the Japanese case, more than 1/3 of all households would be exempt (Chart 1).  On average, the incidence would be around 1-1,5% of disposable income  The Japanese tax and the tax on Ise shift the burden on the elderly (see table 2)

12. Why LTC expenditure grows The growth of LTC expenditure is driven by: 1) demographic change; 2) the evolution of disability incidence for each age group and gender; 3) the real increase in per unit cost of care (increase in expenditure for any given number and distribution of beneficiaries); 4) changes in the distribution of beneficiaries across different disability levels.

13. A numerical exercise Common assumptions: –German incidence of disability per age group and gender applied to Italian population –Initial number of beneficiaries =100 –Initial benefit (the same for all beneficiaries) =100 –Initial expenditure is therefore 10,000 –Initial tax revenue is 10,000 and is set to grow at the alternative projected growth rates of GNP –When a decrease in the disability incidence is considered, n° of beneficiaries decreased by 0.08% per year and 0.2% per year, respectively, for beneficiaries in the 65-74 and 75+ age group. –Inflation rate is set at 2% per year.

14. Main results 1 With no indexation and 1% GDP growth: – financial sustainability is guaranteed (Chart 3.1) even in a scenario in which GDP growth is modest (1%) – benefit reduced in real terms by almost 10% in 5 years, by almost 20% in 11 years, halved in 35 years With indexation (Chart 3.2), –With 1% growth and declining disability: deficit in percentage of revenues of about 10% by 2013 and more than 20% by 2022; that is 0.08% or 0.16% of GDP. –We need a 2% GDP growth to achieve sustainability (Chart 4).

15. Main results 2 With indexation to both inflation and 1.5% unit cost : –even a robust 3% average growth of GDP cannot prevent a prolonged and substantial deficit in the LTC program (Chart 5).

16. Chart 3.1: No indexation of benefits, 1% GDP growth (year 1=2005, year 47=2051).

17. Chart 3.2: Indexation of benefits to inflation, 1% GDP growth (year 1=2005, year 47=2051)

18. Chart 4: Indexation of benefits to inflation, 2% GDP growth (year 1=2005, year 47=2051)

19. Chart 5: Full indexation of benefits (inflation +1.5% growth of unit costs of care); 3% GDP growth (year 1=2005, year 47=2051).

20. What about a “LTC trust fund”? Difficult to achieve jointly sustainability and fairness with pure payg Is it fair that those that are rich and old in first generation of beneficiaries do not co-pay? What about using proceeds of temporary co- payment in order to build up a trust fund? An LTC trust fund in order to smooth the effects of demography?

21. LTC trust fund: a numerical exercise Beneficiaries in the first x years of working of the scheme are subject to a means-tested co- payment that covers y% of total expenditure. In case total contributions exceed current expenditures (thanks to the first years co- payment) the surpluses go to a trust fund that yields a 1.5% real interest rate. In case expenditure exceeds current contributions, the deficit is financed by taking from the trust fund (as long it is sufficient to) or by issuing debt at a 1.5% real interest rate.

22. LTC trust fund: main results 1 With 3% GDP growth and a 1.5% unit cost dynamics and 20% co-payment for 10 years (see Chart 6.1) the scheme runs a surplus up to 2015 (1.3% of GDP) and is financially sustainable up to 2028 or 2032 in case of constant or decreasing incidence of disability, respectively. With 1.5% GDP growth (1.5%), with 30% co- payment for 20 years, financial sustainability could be guaranteed only up to 2029-2030 (see Chart 7.1 and 7.2).

23. LTC trust fund: main results 2 With no increase in per unit costs (or indexation only with respect to the general price index), even the modest growth scenario of Chart 7 appears more positive (see Chart 8.1 and 8.2). In such a scenario we would actually need – in the constant incidence of disability case - a lower co- payment (25%) for a shorter period (15 years) in order to have long term sustainability (see Chart 8.3). If the incidence of disability decreases we need a 20% co-payment for “only” 12 years (see Chart 8.4); trust fund 1.7% of GDP in the period 2016-2018.

24. Chart 6.1: Net balance of LTC scheme with: full indexation of benefits (inflation + 1.5% growth of unit costs of care), 3% GDP growth, 20% average co-payment for the first 10 years (year 1=2005, year 47=2051).

25. Chart 6.2: Trust fund: full indexation of benefits (inflation + 1.5% growth of unit costs of care), 3% GDP growth, 20% average co- payment for the first 10 years (same as in chart 6.1) (year 1=2005, year 47=2051).

26. Chart 7.1: Net balance of LTC scheme with: full indexation of benefits (inflation + 1.5% growth of unit costs of care), 1.5% GDP growth, 30% average co-payment for the first 20 years (year 1=2005, year 47=2051).

27. Chart 7.2: LTC trust fund with: full indexation of benefits (inflation + 1.5% growth of unit costs of care), 1.5% GDP growth, 30% average co-payment for the first 20 years. (same as in Chart 7.1) (year 1=2005, year 47=2051).

28. Chart 8.1: Net balance of LTC scheme with: indexation of benefits to inflation, 1.5% GDP growth, 30% average co-payment for the first 20 years (year 1=2005, year 47=2051).

29. Chart 8.2: Trust fund with: indexation of benefits to inflation, 1.5% GDP growth, 30% average co-payment for the first 20 years (same as in chart 8.1) (year 1=2005, year 47=2051).

30. Chart 8.3: Trust fund with: indexation of benefits to inflation, 1.5% GDP growth, 25% average co-payment for the first 15 years (year 1=2005, year 47=2051).

31. Chart 8.4: Trust fund with: indexation of benefits to inflation, 1.5% GDP growth, 10% average co-payment for the first 12 years (year 1=2005, year 47=2051).