CEPAR

You are here

Working Papers

2020Dec

Dandan Yu, Bei Lu, and John Piggott

Abstract: Background: Results of research into the effects of alcohol drinking on population health are inconsistent. Some studies suggest that light to moderate alcohol consump- tion can have a protective effect on morbidity and mortality. But others challenge this view and claim that alcohol use could lead to health loss regardless of the amount. We contribute evidence to this debate by investigating the association between alcohol drinking and all-cause mortality among older adults in China.

Methods: We use nationally representative samples from the Chinese Longitudinal Healthy Longevity Survey and the China Health and Retirement Longitudinal Study. Cox regressions compare ever drinkers who had been exposed to alcohol at some time and lifelong abstainers. We then subdivide ever drinkers into former drinkers who had stopped drinking and current drinkers who were still consuming alcohol. Cox results are supplemented with the interpolated Markov chain approach to calculate total and disability-free life expectancy.

Results: Among older males, ever drinkers seemed to have similar mortality risks to lifelong abstainers. Compared to abstinence, mortality was elevated for former male drinkers, although the effects were generally insignificant. Current male drinkers aged between 65 and 85 had a significantly lower risk of death. With the adjustment for so- cioeconomic status, an average current male drinker at age 65 could expect to live 1.65 years longer in total and 1.84 years longer without disability than lifelong abstainers. However, we can’t reliably estimate alcohol effects on older Chinese females.

Conclusions: Since there is little doubt that heavy drinking is detrimental to health, our results provide evidence supporting an association between light to moderate alcohol consumption and reduced mortality. The recommendation of zero alcohol consump- tion might not be well-justified in the contemporary Chinese context. The importance of alcohol intake in evaluating population health should be taken into account when predicting future health care burdens.

Keywords: Alcohol drinking; All-cause mortality; IMaCh; Life expectancy

 

2020Dec

Phitawat Poonpolkul

Abstract: Background: Results of research into the effects of alcohol drinking on population health are inconsistent. Some studies suggest that light to moderate alcohol consump- tion can have a protective effect on morbidity and mortality. But others challenge this view and claim that alcohol use could lead to health loss regardless of the amount. We contribute evidence to this debate by investigating the association between alcohol drinking and all-cause mortality among older adults in China.

Methods: We use nationally representative samples from the Chinese Longitudinal Healthy Longevity Survey and the China Health and Retirement Longitudinal Study. Cox regressions compare ever drinkers who had been exposed to alcohol at some time and lifelong abstainers. We then subdivide ever drinkers into former drinkers who had stopped drinking and current drinkers who were still consuming alcohol. Cox results are supplemented with the interpolated Markov chain approach to calculate total and disability-free life expectancy.

Results: Among older males, ever drinkers seemed to have similar mortality risks to lifelong abstainers. Compared to abstinence, mortality was elevated for former male drinkers, although the effects were generally insignificant. Current male drinkers aged between 65 and 85 had a significantly lower risk of death. With the adjustment for so- cioeconomic status, an average current male drinker at age 65 could expect to live 1.65 years longer in total and 1.84 years longer without disability than lifelong abstainers. However, we can’t reliably estimate alcohol effects on older Chinese females.

Conclusions: Since there is little doubt that heavy drinking is detrimental to health, our results provide evidence supporting an association between light to moderate alcohol consumption and reduced mortality. The recommendation of zero alcohol consump- tion might not be well-justified in the contemporary Chinese context. The importance of alcohol intake in evaluating population health should be taken into account when predicting future health care burdens.

Keywords: Alcohol drinking; All-cause mortality; IMaCh; Life expectancy

 

2020Dec

Chung Tran and Sebastian Wende

Abstract: We study the incidence of capital income taxation in a dynamic general equilibrium model with heterogeneous firms and lifecycle households. In this incomplete market setting, marginal excess burdens of three capital taxes, namely corporate income, dividend and capital gains taxes, are vastly different due to heterogeneous responses of firms and households, and heterogeneous effects of general equilibrium adjustments. It is indeed important to account for firm heterogeneity in productivity and investment financing as well as household heterogeneity in age and skill. Overall, taxing capital with a corporate income tax at the firm level results in higher excess burden than taxing capital with dividend and capital gains taxes at the household level. Given the existing U.S. tax treatment for capital income, reforms that shift tax burden from the firm to household side potentially result in efficiency gains and overall welfare improving. However, the welfare benefits of the tax reforms are quite different across households and generations over transition time, depending on skill, age-cohort and budget balancing tax instruments. In particular, majority of currently alive households, especially retirees, experience welfare gains under moderate corporate income tax cuts, but suffer from welfare losses under more radical tax cuts.

Keywords: Excess burden; Tax incidence; Distributional effects; Overlapping generations; Dynamic general equilibrium

 

2020Dec
Dr Miguel Olivo-Villabrille

Miguel Olivo-Villabrille 

Abstract:  Numerous studies find that married men earn more than single men. However, identifying whether and why marriage a↵ects earnings is complicated by the fact that marriage market outcomes are jointly determined with potential earnings. As such, I exploit exogenous variation in marriage induced by the introduction of no-fault divorce laws in the US. I find a 38% causal increase of marriage on earnings of husbands. This increase in earnings is explained by a large increase in labor market work after marriage. My findings are robust to the possibility of unobserved heterogeneity in the effect of marriage on earnings across individuals.

Keywords: Marital earnings premium, marriage, divorce laws, local average treatment effects.

2020Nov

Andrew Hunt and Andrés M. Villegas

Abstract: Rather than looking at the mortality rates directly, a number of recent academic studies have looked at modeling rates of improvement in mortality when making mortality projections. Although relatively new in the academic literature, the use of mortality improvement rates has a long-standing tradition in actuarial practice when allowing for improvements in mortality from standard motality tables. However, mortality improvement rates are dificult to estimate robustly and models of them are subject to high levels of parameter uncertainty, since they are derived by dividing one uncertain quantity by another. Despite this, the studies of mortality improvement rates to date have not investigated parameter uncertainty due to the ad hoc methods used to fit the models to historical data. In this study, we adapt the Poisson model for the numbers of deaths at each age and year, proposed in Brouhns et al. [Insurance: Mathematics and Economics 3 (2002) 31] to model mortality improvement rates. This enables models of improvement rates to be fitted using standard maximum likelihood techniques and allows parameter uncetainty to be investigated using a standard bootstrapping approach. We illustrate the proposed modeling approach using data for the England and Wales population.

Keywords: Mortality Improvements; Mortality forecasting; Parameter uncertainty; Robustness

2020Oct
George Kudrna

George Kudrna, Trang Le and John Piggott

Abstract: This report documents and studies demographic and household survey data in Indonesia. The two key objectives are to provide (i) data for the calibration of the economic model that will be developed in this project; and (ii) stylized facts for the Indonesian household sector and economic behaviour of Indonesian households over their life cycle that will be closely captured by the economic model.

The focus of this report is on the demographic change, labour force and older people (and their resources) in Indonesia, using the United Nations demographic data (UN 2019) and the Indonesian Family Life Survey (IFLS) (documented by Strauss et al., 2016). In this report, “older people” can be thought of as those 50 and above.

We show that:

·      Indonesia will undergo pronounced population ageing driven by a reduction in total fertility rate. For example, the aged dependency ratio (65+/15-64) is projected to increase from less than 10% (in 2020) to over 46% in 2100. This is also attributed to an increasing life expectancy, particularly at older ages. For those at age 65, life expectancy is projected to increase by almost 20 years in 2100 (which is almost double the expected lifespan in the middle of the 20th century). Indonesia’s total population has also quadrupled to 273 million since 1950 and is projected to increase to 320 million in 2100. However, the annual population growth rate will become negative, reaching
-0.3% in 2100 due to population ageing.

·      Importantly, drawing on IFLS household survey data, this demographic transition is occurring in an economy where the majority of the labour force operates in informal employment not covered by a formal retirement income policy.

·      At older ages, people continue to derive their income mainly from employment, along with private transfers from their adult children and these two income sources will be impacted by fewer adult children (to provide private transfers) and longer lifespans (affecting the labour supply of older people).

These findings indicate a pressing need for major social policy development over the next two decades to mitigate negative social and economic implications of this demographic shift and to avert large-scale poverty among older cohorts.

2020Oct
Bei Lu

Bei Lu, Mi Hong, Guanggang Feng, John Piggott and Guy Mayraz

Abstract: This paper uses a unique dataset of seriously ill patients in China across the retirement window to analyse the impact of a change in the co-pay ratio at retirement on inpatient expenditures. We find that a decrease in the co-pay ratio (that is, a lower proportion of cost borne by the user) leads to an increase in medical insurance spending. Surprisingly, out-of-pocket spending also increases.  Individuals' Medical Saving Account (MSA) balances are associated with higher inpatient expenditures. Results indicate that cost-sharing arrangements in China are very sensitive to changes in the co-pay ratio, an effect which appears to be magnified by significant MSA balances. The reduction in the co-pay at retirement leads to substantial increases in medical expenditures at that time. If policy reform is aimed at containing aggregate health expenditures, the retirement age change in the co-pay rate should be re-visited.

JEL Classification Numbers: I13, I14, I15

Keywords: health policy, co-pay; cost-sharing, medical savings accounts, medical expenditure

2020Sep
cepar award

Qiqi Wang, Katja Hanewald and Xiaojun Wang

Abstract: This article proposes a new model that combines a neural network with a generalized linear model (GLM) to estimate and predict health transition intensities. The model allows for socioeconomic and lifestyle factors to impact the health transition processes and captures linear and nonlinear relationships. A key innovation is that the model features transfer learning between different transition rates. It autonomously finds the relationships between factors and the links between the transition processes. We apply the model to individual-level data from the Chinese Longitudinal Healthy Longevity Survey from 1998–2018. The results show that our model performs better in estimation and prediction than standalone GLM and neural network models. We thus provide new estimates of the life expectancies for a range of population subgroups. The model can be easily applied to other datasets, and our results confirm that machine learning techniques are promising tools to model insurance risks.

Keywords: Neural networks, Transfer learning, Multi-state health transitions

2020Sep

Hazel Bateman and Inka Eberhardt

Abstract: Voluntary annuitization from defined contribution pension plans is uncommon, and in many countries, retirees self-insure against retirement risks by holding on to and even building up assets. Lack of awareness of retirement income products and their design and financial impact is a key reason for low take-up of annuity products. Using an online discrete choice experiment we test how a Fact Sheet presenting standardised information on key product features - income, risk, access to capital and death benefits - affects stated choices from a menu of annuity, phased withdrawal and bundled retirement income products. Our setting is Australia where retirees can choose how to decumulate their retirement savings. When using the Fact Sheet, participants chose the lifetime annuity and bundled annuity products most often, which is contrary to the actual behaviour of Australian retirees who predominantly take phased withdrawal products. Of five Fact Sheet information items, choices were mostly driven by the Product Rating (a 1-7 rating of protection against a fall in income due to inflation, market and longevity risk) and Average Annual Income. The lifetime annuity and the bundled lifetime annuity/phased withdrawal products were more likely to be chosen where Fact Sheets used graphs and tables to present information, and where the Product Rating is more salient. However perceptions of risk and control were more important to product choices than actual product knowledge or understanding of the Fact Sheet information. Our findings suggest that Fact Sheet information items, especially the prescribed Product Rating and the associated information on inflation, longevity and market risk decisions drive both perceptions and choice of retirement income product and must be carefully designed.

Keywords: information disclosure, retirement income products, discrete choice experiment, product perceptions