Suicide rates reflect the proportion of a given population that commits or committed
suicide. The rates are generally so low that they are standardized by relating them
to population groups of size 100,000 each to make them comprehensible and to permit
meaningful and realistic comparisons between populations of equal size.
If the calculation of a suicide rate is based on the standardized population
of a given peer group, meaningful comparisons can be made between suicide rates of
sub-groups within an age group to permit comparisons of the likelihood of each sub-group
to commit suicide. These comparisons will permit to establish the determination of
accurate relative risks of each sub-group, such as how likely an average man is to commit
suicide as to the likelihood of an average woman in that group to kill herself, or, how
many times more likely is a man within a population to commit suicide than is a woman.
On the other hand, if the calculation of a suicide rate is based on the
overall population containing all sub-groups within the age group, such as members of both
sexes within the age group 85 and over, then all we will learn from the suicide rates
calculated for each sex is the likelihood that a given suicide victim within that age
group is a man or a woman.
The suicide data posted at the NCIPC web site contains
information on suicide rates that relates to the sexes in each age group.
Unfortunately, those rates are calculated on the basis of the total population segment
comprising both sexes in each age group. Therefore the suicide rates for men as well
as for women are not quite accurate in any of the age groups, but on account of the small
segment of men in the older age groups the rates are exceptionally misleading.
They'll show, with respect to the overall population in each group, what proportion
committed suicide but not how likely it is that a given man or woman commits suicide.
That may be acceptable for some demographers, but it isn't satisfactory
for anyone who wants to know what proportion of men within the population of men in that
age group committed suicide. Anyone who is truly concerned about the welfare of
older men would care enough to want to know accurate suicide rates and not rates that
contain a very serious bias.
To illustrate the fallacy in the "sex-specific" suicide rates at the NCIPC
web site, consider an age group that consists of 99 women and one man.
According to the rules that were used to calculate the suicide rates
shown at the NCIPC web site, if that single man were to kill himself, the corresponding
suicide rate for men would be 1,000 per a population of 100,000 men and women -- a reason
for concern. (As an aside, contrary to popular perception, the man would be very
likely to kill himself. Think about it.)
However, the true suicide rate for all men in that sector, extrapolated
for men only and not for the general population, should be 100,000.
If three of the 99 women were to kill themselves, that would
translate to a suicide rate of 3,000 when extrapolated to a population of 100,000 men and
women, no doubt reason for even greater concern for the fate of women. If the
suicide rate for the women were to be extrapolated to a population of 100,000 women only,
their standardized suicide rate is not much higher. It would then be 3,030.3 for a
population of 100,000 women.
The relative risk for suicide that the population of men poses in
relation to that of the population of women in the example case is correctly 100,000 :
3,030.3 or 33 : 1. Men in that population would be 33 times more likely than women
to commit suicide.
However, if the calculation of the respective suicide rates for the
sexes were based on the routine used to calculate the suicide rates shown at the NCIPC web
site, the relative risk for men in the population in the example would not be that but
rather 1,000 to 3,000 or 0.33 to one. That would underestimate the correct relative
risk for men by a factor of 99. That should be of great concern to anyone and be a
serious incentive to perform the calculations correctly instead.
The sex-specific suicide rates from the NCIPC web site were for that reason not
reflected in any graphs here. The graphs shown on this page do reflect correct and
true suicide rates that are truly specific for the population of each sex, calculated by
using census figures for each age group and for the population of each sex within that,
for each of the years of the period.