Whenever I watch the US presidential election debates it always reminds me of the 1936 presidential election that saw Roosevelt going against Alfred Landon. This is a well-known election amongst statisticians. The interesting thing about this election is that a certain magazine called the Literary Digest undertook an ambitious pre-election poll in a bid to predict the result of the election. Indeed the Literary Digest had successfully predicted the winners of US presidential elections since the year 1906. That particular year however they decided to send out mock ballot papers to a whopping 10 million potential voters who they picked from the telephone directory. This was an expensive process which sadly yielded biased results, only 2.4 million ballots were returned and the Literary Digest predicted a landslide victory for Landon. If you know your history very well you will know than Roosevelt is the one who actually won that election – taking 57% of the votes.
This story bears resemblance to a looming disaster that is brewing in our life/funeral assurance and pension industries. These institutions are not in the business of predicting elections, but that of predicting how long (on average) their policyholders or pensioners will live. A statistical tool called a mortality table is used to tabulate the likelihood of a person dying based on their age and other factors such as whether they smoke and what their occupation is. This is the tool that actuaries use to calculate how much premium one should pay in order to secure a certain funeral death benefit or how much pension one should receive for the rest of their life. The majority of life assurance and pension funds in Zimbabwe have for a long time now used mortality tables from South Africa. This in itself is dangerous because the life expectancy and standards of living of Zimbabweans and South Africans are extremely different.
For a life assurance company, the consequences of not knowing the underlying mortality profile of its policyholders are far reaching. For starters, it may end up charging the wrong premiums to policyholders. Already we are seeing that it is possible for a 20 year old and a 55 year old to take out the same funeral assurance policy and pay the same premium. This is a direct result of the lack of investment in the creation of mortality tables that allow the life insurance company to quantify the life expectancies of its potential policyholders.Charging a flat premium means that life assurance becomes comparatively cheaper the older one gets. Such a situation is as absurd as seeing water flowing up a mountain. If we step into the world of economics and assume that policyholders are rational, such a situation means that the 55 year old is more likely to take out life assurance than the 20 year old. This could spell disaster down the road for the life assurance company since it would be accepting business from policyholders nearing the end of their lives but not charging appropriate premiums.
Have you ever wondered how life assurance providers calculate their annual profits? They take monthly premiums but may not need to pay out a benefit on those premiums for 20 to 30 years in the future. Every year, they need to declare profits (or losses) and pay taxes on those profits plus declare a dividend to their shareholders. Use of wrong mortality tables can lead to the company declaring too much (or too little)profit in the early years of its operation. This may lead to it run out of money a few years down the line or disadvantage the shareholders by retaining excess funds within the company.
While national statistics from organisations such as ZimStat or Aid agencies can provide useful insights into the mortality profile of the country as a whole, they are practically useless for a life insurance/pension fund. This is because less than 3% of the population has a life assurance policy or access to a pension fund. The national life expectancy of Zimbabwe is somewhere around 59 years while all pensioners are aged 60 and above. Simply put, the types of people who take out life assurance are not representative ofthe entire Zimbabwean population. Using national statistics to calculate mortality trends for pensioners would be like making the same mistake that was made by the Literary Digest in 1936 – using a telephone directory to sample a country at a time where the majority of the US population did not have a telephone.
The developed countries are investing a lot in studying how mortality is changing. What is evident is that mortality is improving – people are living longer. But the question is how much longer. Pension funds and life assurance companies in Zimbabwe will need to begin investing in understanding how the mortality of their beneficiaries is changing. The task at hand is formidable even for a single company/pension fund to attempt on its own. It will be more feasible to pool data from many providers and develop industry specific mortality tables. Failure to act now can spell disaster in the future.
Thomas Sithole is an Actuarial Analyst (Enterprise Risk Management) at Bluecroft Actuarial Solutions. Please refer to his corporate profile on this web address to contact him: thomas.bluecroftsolutions.com
Follow us on Twitter on @FingazLive and on Facebook – The Financial Gazette