As long as the smoker-to-non-smoker mortality ratio remains relatively stable across impairment severity levels, smoking and the impairment should be considered independent, and no additional interaction should be reflected in underwriting. Only when the smoking multiplier changes materially by impairment severity is there evidence of dependence that may justify different ratings for smokers and non-smokers.
A common challenge in life insurance underwriting occurs when an applicant has both a specific impairment, such as diabetes or heart disease, and a history of smoking. The underwriter must determine whether smoking adds risk beyond what is already reflected in smoker-distinct pricing.
This decision produces a fundamental question: Do smoking and the impairment affect mortality independently?
If they do, smoking should not change the impairment rating. If they do not, smoking may require additional consideration because it modifies the impairment’s effect on mortality.
Although this principle seems straightforward, independence is often misunderstood in practice. Much of the confusion arises from different interpretations of how smoking affects mortality and whether those effects should be assessed using an additive or multiplicative framework.
This article explains the distinction and shows why a multiplicative model is the appropriate framework for assessing smoking interactions in life insurance underwriting.
Consider a situation in which an applicant has an impairment and is a smoker.
Under this approach, smoking is assumed to affect mortality independently of the impairment. Because smoking-related mortality risk is already reflected in the insurer’s smoker pricing structure, no additional adjustment is made to the impairment rating.
As a result, the impairment receives the same rating regardless of smoking status.
Under this approach, smoking is assumed to modify the impact of the impairment on mortality.
As a result, smokers may receive a more severe underwriting classification for the impairment than non-smokers.
The key question, therefore, is how to determine whether smoking and the impairment are truly independent.
Suppose a hypothetical impairment has five severity levels, where:
The figure shows mortality rates across the five severity levels of the hypothetical impairment, stratified by smoking status.
The correct interpretation is that smoking and the impairment are independent.
This conclusion follows because:
In other words, the effect of smoking is constant across all severity levels, and the effect of the impairment is constant across smoking groups.
The relative risks remain unchanged regardless of smoking status. This is the defining feature of independence in a multiplicative mortality framework.
Many underwriters reviewing such a figure may conclude that smoking and impairment severity are not independent.
The reasoning typically goes as follows:
This interpretation, however, implicitly assumes an additive model of mortality.
In an additive framework, smoking is assumed to increase mortality by a fixed absolute amount.
For example:
Under this model, the key measure is the mortality difference between smokers and non-smokers.
If that difference increases as impairment severity rises, one might conclude that smoking and the impairment interact synergistically and are therefore dependent.
In a multiplicative framework, smoking is assumed to increase mortality by a fixed relative amount.
For example:
In our example:
The relevant measure is not the mortality difference but the mortality ratio.
Although the absolute mortality gap widens at higher impairment severity levels, the mortality ratio remains constant at 2.1 across all levels. Similarly, impairment severity has a consistent relative effect: Mortality increases by 1.5 times for each severity level, regardless of smoking status.
Therefore, smoking and the impairment remain independent in the multiplicative sense.
The distinction between additive and multiplicative effects is not merely academic. It directly affects underwriting decisions.
Life insurance pricing generally treats smoking as a multiplier of baseline mortality, not as a fixed additive increase. Smoker rates are built around the relative increase in mortality risk associated with smoking.
Consequently, when evaluating whether smoking interacts with another impairment, the assessment should be made using the same multiplicative framework that underlies pricing.
From this perspective:
For example, if smokers exhibit a mortality ratio of 2.1 at low severity levels of an impairment but a ratio of 3.5 at high severity levels, this suggests a meaningful interaction between smoking and the impairment severity. In such cases, additional underwriting consideration may be warranted.
Conversely, if the mortality ratio remains close to 2.1 across all severity levels, then smoking has already been adequately reflected in smoker pricing and no additional adjustment to the impairment rating is justified.
In practice, smoking can interact multiplicatively with certain impairments, often referred to as smoking-related conditions, including asthma, chronic obstructive pulmonary disease (COPD), severe coronary artery disease, and peripheral arterial disease. Evidence indicates that the mortality ratio between smokers and non-smokers is substantially higher among individuals with these conditions than among those without them. Accordingly, additional underwriting debits are warranted, and in some cases a decline may be appropriate when smoking coexists with these conditions.
The determination of whether smoking should modify the underwriting assessment of another impairment depends on the concept of independence. Independence is often misunderstood because observers may focus on absolute mortality differences rather than relative mortality relationships.
In life insurance underwriting, smoking is incorporated into pricing as a multiplicative adjustment to mortality. Therefore, the appropriate framework for evaluating interactions between smoking and other impairment is also multiplicative.
By adopting a multiplicative perspective, underwriters can make more consistent, transparent, and actuarially sound decisions when evaluating the combined effects of smoking and other mortality risks.
RGA can help you put these insights into action across your underwriting operation. Contact us today to learn more.
