You are here

Working Papers

2018Sep
Aged care support

Aged care residents, residential care developers and government policy-makers need accurate information on likelihood of main events in residential care (i.e. residents’ functional decline and death). Since 20 March 2008 Australian government subsidies for residential care have been based on detailed assessments of individual care needs, and this generated 1.5 million assessment records by 30 June 2015. Four levels are assessed for three types of need - aids to daily living, behavioural needs, and complex health care. Logistic regression models are used to derive mortality and transition probabilities from these data. Backwards derivation was used to estimate mean life expectancies from these models, and microsimulation used to model distributions around means. As there has been continuing drift in assessed care needs, the mortality and transition assumptions estimates are based on the most recent year of experience. A microsimulation model of aged care residents, with all residents at 30 June 2015 as the initial population, has been constructed.

Key Words: ADL, Assistance needs, Life expectancy, Residential care, Australia

2018Aug
Colleagues discussing ageing research

Zvi Eckstein, Michael Keane and Osnat Lifshitz

Comparing the 1935 and 1975 U.S. birth cohorts, wages of married women grew twice as fast as for married men, and the wage gap between married and single women turned from negative to positive. The employment rate of married women also increased sharply, while that of other groups remained quite stable. To better understand these diverse patterns we develop a lifecycle model incorporating individual and household decisions about education, employment, marriage/divorce and fertility. The model provides an excellent fit to wage and employment patterns, along with changes in education, marriage/divorce rates, and fertility. We assume fixed preferences, but allow for four exogenously changing factors: (i) mother’s education, health and taxes/transfers; (ii) marriage market opportunities and divorce costs; (iii) the wage structure and job offers; (iv) contraception technology. We quantify how each factor contributed to changes across cohorts. We find that factor (iii) was the most important force driving the increase in relative wages of married women, but that all four factors are important for explaining the many socio-economic changes that occurred in the past 50 years. Finally, we use the model to simulate a shift from joint to individual taxation. In a revenue neutral simulation, we predict this would increase employment of married women by 9% and the marriage rate by 8.1%.

2018Aug
Pensioners enjoying retirement

Ermanno Pitacco

Heterogeneity of a population in respect of mortality is due to differences among the individuals, which are caused by various risk factors. Some risk factors are observable while others are unobservable. The set of observable risk factors clearly depends on the type of population addressed. The impact of observable risk factors on individual mortality, in particular when they also constitute “rating factors” in the calculation of premiums and other actuarial values, is usually expressed approximately, according to some pragmatic approach. For example, additive or multiplicative adjustments to the average age-specific mortality are frequently adopted. Heterogeneity due to unobservable risk factors can conversely be quantified by adopting the concept of individual “frailty”. However, individual frailty can be interpreted and consequently modeled in several ways, according to the causes which are considered as originating the frailty itself: congenital characteristics, environmental features, lifestyle aspects, etc. It follows that the individual frailty can, in particular, be assumed either constant or variable throughout the lifetime.

Keywords: Heterogeneity, Frailty, Risk factors, Force of mortality, Mortality laws, Parametric models, Special-rate annuities.

2018Aug
Financial independence

Natasha Ginnivan, Rafal Chomik, John Piggott, Tony Butler and Adrienne Withall

The Australian prisoner population has experienced a dramatic increase in the number of older inmates over the past decade. Modelling presented in this paper shows that these trends are likely to continue over the next decade and that they will result in higher health costs of prisons under different imprisonment scenarios. Taking into consideration the continued rise of incarceration rates, our calculations show that health costs of prisoners could increase by anywhere between 17% to 90% depending on whether the increase of older prisoners continues as it has in the last decade. Policy responses have been slow so far. We suggest that in the absence of a coordinated policy response, covering a range of interventions, costs will continue to increase, as this population continues to age more rapidly than the general population. Well-conceived interventions would be a worthwhile investment from both financial and social perspectives.

2018May
Financial independence

Fedor Iskhakov and Michael Keane

In this paper we structurally estimate a life-cycle model of consumption/savings, labor supply and retirement, using data from the Australian HILDA panel. We use the model to evaluate effects of the Australian aged pension system and tax policy on labor supply, consumption and retirement decisions. Our model accounts for human capital accumulation via learning by doing, as well as wealth accumulation and decumulation over the life cycle, uninsurable wage risk, credit constraints, a non-absorbing retirement decision, and labor market frictions. We account for the ’bunching’ of hours by discretizing job offers into several hours levels, allowing us to investigate labor supply on both intensive and extensive margins. Our model allows us to quantify the effects of anticipated and unanticipated tax and pension policy changes at different points of the life cycle. Our results imply that the Australian Aged Pension system as currently designed is very poorly targeted, so that means testing and other program rules could be improved.

2018Feb
Young family at home

Sarah Kaakai, Héloïse Labit Hardy, Séverine Arnold (-Gaille), Nicole El Karoui

Numerous studies have demonstrated the pervasive effect of socioeconomic status on mortality and cause-specific mortality. More recently, a growing number of studies have indicated a widening of socioeconomic inequalities. It has therefore become crucially important to understand and model the mortality of a heterogeneous population. Recent developments in multi-population mortality have improved the modeling and forecast of subpopulations mortality inside a national population. But this new class of models has raised a number of questions, among which the issue of consistency between subnational and national forecasts. Hence, modeling population dynamics provide complementary insights on subpopulations evolution as well as on aggregated quantities.
 

2018Feb
Aged care analysis

Yajing Xu, Michael Sherris and Jonathan Ziveyi

The pricing of longevity-linked securities depends not only on the stochastic uncertainty of the underlying risk factors, but also the attitude of investors towards those factors. In this research, we investigate how to estimate the market risk premium of longevity risk using investable retirement indexes, incorporating uncertain real interest rates using an affine dynamic Nelson-Siegel model. 

2018Jan
Pensioners enjoying a stroll

Shang Wu, Hazel Bateman, Ralph Stevens and Susan Thorp

We investigate whether a life care annuity - the integration of a life annuity with long-term care insurance (LTCI) - can enhance insurance participation to mitigate the economic puzzle of under- insurance in the longevity insurance and LTCI markets. Using an online choice experiment, we elicit individuals' preferences for consumption in different health conditions and their demand for a life care annuity and its health-contingent income feature. 

2018Jan
Data graphs

Michael Sherris, Yajing Xu and Jonathan Ziveyi

Multi-country risk management of longevity risk provides new opportunities to hedge mortality and interest rate risks in guaranteed lifetime income streams. This requires consideration of both interest rate and mortality risks in multiple countries. For this purpose, we develop value-based longevity indexes for multiple cohorts in two different countries that take into account the major sources of risks impacting life insurance portfolios, mortality and interest rates.