January 2013

Living to 100 Monograph – Mortality Articles

By Raymond D. Berry

The next Living to 100 Symposium is scheduled for January 8-10, 2014 in Orlando, Fla. This is the fifth symposium in this triennial event. The symposium is supported by over 50 organizations from around the world, including the SOA. Thought leaders provide insights into various issues related to the increased number of retirees such as mortality projections and the benefits and risks associated with the increases in retirees. Details concerning this upcoming symposium are here: Livingto100.soa.org/symposium.aspx.

One of the positive outcomes of these events is the accumulation of a wide array of papers on various topics related to longevity, mortality projection, patterns of aging and so on. Below are abstracts of four papers regarding mortality which may be of interest to pension actuaries. PDF’s of the complete papers are available at the following link: SOA.org/library/monographs/life/living-to-100/2011/2011-toc.aspx

How to Survive Living to 100: Ways to Improve the U.S. Retirement System
By Beverly Orth

Workers in the U.S., along with their counterparts around the world, face significantchallenges in saving enough funds to last a lifetime. Many who plan for increased longevity andpurchase insurance products to protect their assets may still have difficulties if they live to bevery old or require extended periods of long-term care.

The U.S. retirement system has many defects that affect individuals’ ability to surviveliving to 100. This paper explores some of the problems that individuals face and recommends changes that could make the U.S. system work better.

Pension Reform in Canada-An Actuarial Perspective
By Robert Brown

This paper is written in two parts. In the first section, we give background materialon the existing Canadian Income Security system, including both government‐sponsoredsystems and private sector supplements. This is to give the rest of the paper its propercontext.

Part two of the paper is the report from the Canadian Institute of Actuaries’ TaskForce on Government‐Facilitated Retirement Income Plans, published in March 2010 to provide a response from the Canadian actuarial profession on the continuing pension reform debate in Canada.

Mortality Compression and Longevity Risk
By Jack Yue

Mortality improvements, especially of the elderly, have been a common phenomenon in many countries since the end of World War II, and many believe life expectancy will continue to increase. As a result, longevity risk becomes an essential component in designing annuity products, as overestimating mortality rates may result in financial insolvency for a life insurance company. Stochastic mortality models have been a popular choice for addressing this risk. However, the longevity predictions of these models are based on historical data and the predictions behave somewhat like extrapolation. But there are no guarantees that future longevity will follow historical trends.

Instead of fitting stochastic models for mortality rates, this study explores increasing life expectancy by examining the basic properties of survival curves. Specifically, we shall check if there are signs of mortality compression (i.e., rectangularization of the survival curve) and evaluate what it means to designing annuity products. Note that the majority of past studies for mortality compression use graduated mortality rates and their results are likely influenced by the graduation methods used. Based on the raw mortality rates, we propose an alternative approach and also propose some measurements to verify if there is mortality compression. We then apply the proposed method to the mortality rates of Japan, Sweden and the United States (data source: Human Mortality Database). Unlike the previous results using the graduated mortality rates, we found there are no obvious signs that mortality improvements are slowing down. This indicates that human longevity is likely to increase, at least for a while, and longevity risk should be seriously considered in pricing annuity products. In addition to verifying the mortality compression, we also propose some suggestions for dealing with the longevity risk.

Mortality Rates at Oldest Ages
By Bob Howard

Because of a lack of data, the highest age mortality rates in most tables are conjectural. This paper presents a method for using death records to infer exposure on nonextinguished cohorts, thereby allowing the development of a credible table for high ages. The method uses Whittaker-Henderson graduation in a number of unusual ways. The paper also validates the method by applying it to stochastically generated sets of death records for which the underlying mortality and improvement tables are known. There are some surprising results.

Recent Adult Mortality Trends in Canada, the United States and Other Low Mortality Countries
By Nadine Ouellette, Robert Bourbeau

This paper examines recent changes in the age-at-death distribution at older ages in Canada, the U.S. and eight additional low-mortality countries. Data starting in 1950 are taken from the Human Mortality Database and a flexible two-dimensional smoothing approach based on P-splines is used to monitor these changes. The U.S. displayed the most worrying picture for the latest two decades. Indeed, for several consecutive years in that timeframe, US females and males have both recorded important declines in their modal age at death and their level of old-age mortality remains high compared to the other countries. Thus, although Canada and the US are neighboring countries, the findings for the former regarding recent old-age mortality trends rather resemble those obtained for the remaining eight low mortality countries studied. Further analysis of changes in the age-at-death distribution at older ages by socioeconomic group or by region could improve our current understanding of the latest mortality dynamics recorded among US adults.

Is Raising the Age of Eligibility Fair to All? An Investigation of Socioeconomic Differences in Mortality Using Population Data
By Geoff Rashbrooke

Constraining the cost of pay‐as‐you‐go financing of social security age pensions isbecoming an increasingly important issue as population’s age. One option, increasing the age ofeligibility, raises the issue of the impact of differential mortality by socio‐economic status onthe fairness of such a change. Analysis using panel data may not be fully reliable due todefinitional issues. Population mortality data by ethnicity is available but subdivision by socioeconomic status is more problematic.

This paper draws on New Zealand research that has matched individual death records tocensus records. Using this research, the paper derives mortality tables by adapting New ZealandMāori and non‐Māori population mortality data to reflect differences in socio‐economic status.This adapted data is used as a basis to explore the implications of differential mortality inassessing the equity of increase in the pension age of eligibility.As convergence of mortality rates over time would remove the impact of differential mortality,a brief discussion is included on the prospects for convergence, and some conditionsconsidered necessary for this to occur described. The paper concludes with suggestions as tohow the imperatives for fiscal sustainability might be tempered with actions designed tomitigate the equity shortcomings indicated by the paper’s analysis.

Raymond Berry, ASA, EA, MAAA, MSPA can be reached at raymonddberry@yahoo.com.