David
Wingfield wrote:
I
think that my last email got lost in the shuffle (or perhaps I did
not send it correctly). But here it is again -- sort of.
The
coat versus child debate reveals that vicarious liability analysis
is really a very poor way of thinking about liability. That is why,
amongst other reasons, the SCC got it right in Bazley
v. Curry.
When
we are debating vicarious liability we are (implicitly or expressly)
debating the nature of the relationship between the person who suffered
a loss and the person who is responsible for making good that loss.
Confusion sets in when we focus on the person who caused the loss
instead of the person who is responsible for making the loss good.
We can illustrate this with the coat versus the child examples.
If
I am in the business of looking after your coat, I must make good
your loss if my employee steals the coat or damages it or what ever.
I am liable, not because my employee was engaged in an unauthorised
mode of carrying out an authorised act or any such thing, but for
the simple and straightforward reason that I am responsible for
your coat and I failed in my responsibility to care for it. Why
I failed (my employee's theft or negligence) is really quite irrelevant.
The same goes for a teacher (or gardener) abusing a child. The school
is responsible for caring for the child. The school failed to do
so whilst the child was in its care. It should be irrelevant that
the child was injured by an abusing teacher, or gardener or by falling
bricks.
The
real issue, as I see it, was whether the injury (loss or damage
to a coat, loss caused by abuse) was something the person being
sued was supposed to avoid causing, either because the person expressly
or impliedly promised not to cause the loss or damage or because
as a matter of social policy the person is required not to do so.
Conventional
vicarious liability analysis works best when the underlying relationship
between the injured person and the responsible person is contractual
or is based on property rights. In those cases we know or can easily
determine the limits of liability by looking to a contract (including
any limitation of liability clauses) or to property law (bailment,
occupier's liability). The ease and familiarity with which we analyse
problems of liability where the relationship is voluntary and rooted
in contractual or property dealings drives us to analogise other
relationships with those (eg the example of bailing a child).
The
real question, therefore, is whether the injured party and the responsible
party were in a voluntary relationship such that their rights and
immunities etc can be analysed in the conventional way of those
relationships or whether their relationship is involuntary. If it
is the latter, then social policy must, as it always does, determine
the limits of liability for the harm caused by the relationship.
Vicarious liability is a pointless way of analysing the first and
is crude and ineffective way of analysing the second.
Since
there is a difference between liability voluntarily assumed and
liability imposed by law, the reasons for imposing liability in
the latter case cannot be derived from analogies with the former.
This is why the SCC in Bazley
v. Curry said, correctly in my view, that where the relationship
of the injured and responsible parties is involuntary, policy determines
the limits of liability (and by the way, vicarious liability itself
was a creature of policy too, remember) based on whether the body
sought to be responsible created or enhanced the risk of the harm.
Though this is the right way to analyse the problem of liability
in an involuntary relationship (again in my view) it would have
been better had the SCC simply jettisoned vicarious liability language
entirely.
