On the other hand, when it comes to the longevity risk it is very easy to make mistakes: Let’s take, though only as an ‘indirect’ example, forecasts made by the Italian Social Security Agency (INPS) less than twenty years ago12. With reference to the number of pensions of private employees in Italy, the forecast model supplied the following figures:
The actual figures so far were the following:
The gap was already significant only two years after the forecasts were published and revealed its systematic nature (error through weakness), even though regulatory amendments have progressively limited the access to pensions for younger people.
We think that the key aspect of demographic risk and longevity forecasts derives from the characteristics of scientific knowledge: indeterminate nature of the reality, the conjectural nature of science and, finally, the inevitably ‘unforeseeable’ nature of scientific discoveries13.
We do not have a theory which can explain precisely the longevity of a given population, even if the initial conditions are perfectly known (indeterminism); even if such a theory were available, it would still be always open to falsification and surpassed at a later stage (conjecturalism). However, let us pretend that we have a well-formed theory allowing us to make exact forecasts in terms of longevity and that it is the final and true theory.
Since the length of human life undoubtedly depends — mostly but not exclusively — on scientific progress, i.e. new discoveries and their applications – not only in the field of medicine – its clear that fairly approximate forecasts would require the perfect knowledge of future discoveries (and their impact), which is logically impossible and contradictory, since if perfectly known, they would no longer be future discoveries.
The logical corollary of this syllogism is that, although a structured and constant collaboration between the worlds of science and insurance is certainly useful for allowing full exploitation of at least all the information available, clearly it cannot logically solve the problem of forecasting the length of human life; the impact of research on longevity is assessable at most after events have taken place: we must be aware of how little we know.
Personally, I believe that this is the ultimate reason why statistical-actuarial sources and not medical-biological sources are (and will remain) sounder methodologies and hence provide more reliable forecasts on longevity.
The synthesis of the path of logic taken so far may be formulated as follows:
• From the age of Bismarck onwards, the state has taken on and maintained a kind of monopoly of the market in the longevity risk cover.
• This was due to both strong structural reasons (ex post financial flexibility, public yield expected to be greated than private yield and limited accountability of the political decision-makers) and ideological reasons (progressive weakening of the role of families and individual responsibility, while the role of the state was strengthened).
• In today’s context, the state still remains the main natural ‘owner’ of risk management.
• The powerful development of longevity, together with the ‘financial crisis’ of the State and the intrinsic limits of modern families as ‘risk absorbers’ prospectively lead to developing a – complementary and not substitutive – role for a private market in demographic longevity risk and the tools for managibg that risk.
Let us prepare ourselves therefore to enter a third phase in which:
1. the state, in its traditional role as pension supplier;
2. individual/family resources accumulated or freely available (precautionary financial savings);
3. the insurance system through life pensions deriving from accumulated savings will all be necessary — but none of them will be sufficient on their own — either individually or collectively to face the phenomenon of human longevity and its related challenges.
Many years will be necessary to transform the current system into a balanced three-pillars system, but finding a socially viable alternative in the long run is difficult.
Admittedly, in the Anglo-Saxon world this converging stage has already begun and some of the problems that emerged may provide those who follow with useful lessons for dealing with the issue.
Moreover, if the problem is presented in this way, the public sector may be persuaded to agree, since, apart from a few ideological extremists, and the usual — and inevitable — fluctuations and inconsistencies of the politique politicienne everyone agrees that current trends are unsustainable.
For clarity’s sake, it is worth mentioning that the ‘three pillars’ briefly outlined here do not match those normally referred to when one speaks of new social security systems, which are formed bearing in mind the stage at which resources are accumulated .
As a matter of fact, the third pillar of this model includes all forms of private social security however accumulated, whereas the second accounts for the accumulated financial resources — albeit not allocated — which can be spent since they are not destined to one’s heirs.
As regards the fourth pillar, we shall see below that even the slightest similarity between the two models fades out completely: the same number of current and prospective pillars during the accumulation and payment stages occurs by chance.
12 Taken from Palladino, 2003, “Le pensioni domani: si salvi chi può!”
13 There is no need to mention that this reflection is fully dependent upon the epistemological lessons of K.R. Popper.
Tags: Bundling Policies, four pillars, Longevity Market, Longevity Risk Cover, Reinsurance, Technical Reserves