Joyal,
J:—This
is
an
appeal
by
the
plaintiff
from
a
notice
of
reassessment
dated
March
9,
1981,
confirmed
by
the
defendant
on
May
3,
1983,
with
respect
to
certain
taxes
imposed
under
Part
XII
of
the
Income
Tax
Act.
The
issue
between
the
parties
is
exclusively
one
of
statute
interpretation
and
the
trial
proceeded
on
the
basis
of
an
agreed
statement
of
facts.
Part
XII
of
the
Income
Tax
Act
imposes
a
special
tax
on
life
insurance
companies.
The
charging
provision
in
Part
XII
is
subsection
208(1)
which
states:
208.
(1)
Every
life
insurer
shall
pay
a
tax
for
a
taxation
year
equal
to
15
per
cent
of
its
taxable
Canadian
life
investment
income
for
the
year.
It
sounds
simple
enough.
A
life
insurer
need
only
calculate
its
taxable
Canadian
life
investment
income
and
pay
a
15
per
cent
tax
thereon.
The
difficulty,
however,
is
in
determining
what
is
a
life
insurer’s
taxable
Canadian
life
investment
income
for
the
year.
This
is
where
the
issue
lies
between
the
parties.
Before
pursuing
the
several
intricate
provisions
of
Part
XII
or
before
attempting
to
make
one’s
way
through
its
labryinth
of
verbalized
mathematical
formulae,
a
quick
look
at
policy
considerations
behind
the
enactment
of
Part
XII
might
be
useful.
The
traditional
form
of
a
life
insurance
contract
contains
two
elements,
one
being
the
actuarily
quantified
risk
factor,
the
other
being
the
savings
or
investment
part
of
the
policyholder’s
annual
premium.
For
any
given
premium,
part
of
it
is
cost
of
the
risk
and
the
balance
is
savings.
The
savings
portion
is
invested
by
the
insurer
and
the
income
accrues
to
the
benefit
of
the
policyholder
from
year
to
year.
Historically,
this
income
has
not
been
subject
to
any
investment
income
tax,
a
situation
somewhat
unfair
to
other
taxpayers
who
preferred
to
purchase
term
insurance
from
an
insurer
at
lower
cost
and
invest
the
surplus
funds
privately.
The
income
from
such
investment
was
naturally
taxed
in
the
usual
way.
Both
the
plaintiff
and
the
defendant
in
this
appeal
agree
that
Part
XII
is
essentially
a
tax
on
the
investment
income
of
the
risk
cum
investment
type
of
insurance
policy.
For
obvious
reasons,
however,
it
would
have
been
administrative
chaos
to
attempt
to
impose
such
a
tax
directly
on
each
policyholder.
As
a
result,
the
Part
XII
tax
attempts
to
balance
the
equities,
so
to
speak,
in
charging
life
insurers
with
a
15
per
cent
tax
on
the
pool
of
investment
income
earned
from
such
a
policyholder
and
accruing
to
his
benefit.*
One
must
remember,
however,
that
life
insurance
companies
otherwise
pay
tax
on
their
profits,
a
tax
which
is
indirectly
borne
by
their
shareholders.
It
was
required,
therefore,
that
the
special
15
per
cent
tax
be
imposed
in
such
a
fashion
as
would
charge
the
policyholder
with
it
and
not
result
in
an
additional
tax
on
the
insurer.
Further
background
comment
is
required
before
analyzing
the
provision
of
Part
XII
of
the
Income
Tax
Act.
It
relates
to
the
peculiarities
of
life
insurers,
the
manner
in
which
they
categorize
the
funds
they
receive,
either
by
way
of
premiums
or
by
way
of
investment
returns
and
the
obligation
of
life
insurers
to
comply
not
only
with
the
more
general
provisions
of
the
Income
Tax
Act
but
with
the
more
specific
provisions
of
the
Canadian
and
British
Insurance
Compan-
les
Act.*
It
is
under
this
latter
statute
that
a
life
insurer
must
include
in
its
liabilities
a
reserve
for
all
unmatured
obligations
guaranteed
under
the
terms
of
insurance
policies
issued
by
it.
This
requires
the
establishment
by
the
life
insurer
of
“segregated”
funds
and,
as
a
corollary,
“non-segregated
funds”.
More
on
that
later.
Finally,
because
of
the
multitude
of
cross-references
upon
cross-references
in
specific
provisions
of
the
Income
Tax
Act
relating
to
life
insurers,
it
might
be
wise
to
outline
a
series
of
definitions
for
particular
words
or
expressions
found
in
Part
XII
of
the
statute
and
relating
to
the
issue
before
me.
List
of
Definitions
1.
Subsection
209(1)
defines
“Gross
Canadian
Life
Investment
Income”
as
follows:
209.
(1)
Gross
Canadian
life
investment
income.—A
life
insurer’s
gross
Canadian
life
investment
income
for
a
taxation
year
is
the
amount,
if
any,
by
which
the
aggregate
of
(a)
an
amount
equal
to
(i)
where
subsection
138(9)
applies
but
the
insurer
has
not
elected
under
that
subsection
in
respect
of
the
year,
such
part
of
its
gross
investment
revenue
for
the
year
from
its
non-segregated
property
as
is
required
by
paragraph
(b)
of
that
subsection
to
be
included
in
computing
its
income
for
the
year
from
carrying
on
its
life
insurance
business
in
Canada,
and
(ii)
in
any
other
case,
its
gross
investment
revenue
for
the
year
from
such
of
its
non-segregated
property
as
was
property
used
in
the
year
in,
or
held
in
the
year
in
the
course
of,
carrying
on
its
life
insurance
business
in
Canada,
and
(b)
the
amounts
required
by
subsection
138(4)
to
be
included
in
computing
the
insurer’s
income
for
the
year,
other
than
amounts
deducted
under
paragraph
138(3)(a)
in
computing
its
income
for
the
immediately
preceding
taxation
year,
exceeds
the
aggregate
of
the
amounts
deductible
under
paragraphs
138(3)(b),
(c)
and
(d)
in
computing
the
insurer’s
income
for
the
year.
2.
Subsection
209(2)
defines
“Net
Canadian
Life
Investment
Income”
as
follows:
209.
(2)
Net
Canadian
life
investment
income.—A
life
insurer’s
net
Canadian
life
investment
income
for
a
taxation
year
is
its
gross
Canadian
life
investment
income
for
the
year
minus
the
aggregate
of
(a)
its
outlays
or
expenses
that
were
deductible
under
Part
I
in
computing
its
income
for
the
year
to
the
extent
that
such
outlays
or
expenses
were
laid
out
or
incurred
by
it
for
the
purposes
of
managing
its
non-segregated
property,
the
gross
investment
revenue
from
which
may
reasonably
be
regarded
as
having
been
included
in
its
gross
Canadian
life
investment
income
for
the
year;
(b)
interest
payable
by
the
insurer
in
respect
of
the
year
pursuant
to
a
legal
obligation
incurred
by
it
in
the
course
of
carrying
on
its
life
insurance
business
in
Canada;
(c)
amounts
deductible
under
paragraph
20(1)(a)
in
computing
the
insurer’s
income
for
the
year
in
respect
of
a
depreciable
property
at
least
80
per
cent
of
which
was
used
regularly
by
it
for
the
purpose
of
earning
its
gross
Canadian
life
investment
income
for
the
year;
and
(d)
50
per
cent
of
the
aggregate
of
each
amount
deductible
under
Part
I
in
computing
the
insurer’s
income
for
the
year
from
carrying
on
its
life
insurance
business
in
Canada,
except
to
the
extent
that
such
amount
(i)
is
included
in
any
of
the
amounts
determined
in
respect
of
the
insurer
for
the
year
under
paragraph
(a),
(b)
or
(c),
(ii)
is
deductible
under
subsection
138(3)
in
computing
its
income
for
the
year
from
carrying
on
its
life
insurance
business
in
Canada,
(iii)
was
paid
or
payable
by
the
insurer
under
a
life
insurance
policy
before
the
end
of
the
year,
(iv)
was
an
outlay
or
expense
laid
out
or
incurred
by
it
for
the
purpose
of
earning
income
from
its
group
life
insurance
business,
or
(v)
was
payable
by
the
insurer
to
a
province
as
a
tax
in
respect
of
premiums
collected
by
it
in
the
year
under
life
insurance
policies.
3.
Subsection
209(3)
defines
“Taxable
Canadian
Life
Investment
Income”
as
follows:
209.
(3)
Taxable
Canadian
life
investment
income.—A
life
insurer’s
taxable
Canadian
life
investment
income
for
a
taxation
year
is
the
amount,
if
any,
by
which
its
net
Canadian
life
investment
income
for
the
year
exceeds
the
aggregate
of
(a)
the
interest
element
for
the
year
for
(i)
each
class
of
the
insurer’s
existing
fixed-premium
life
insurance
policies
in
Canada,
and
(ii)
each
class
of
the
insurer’s
life
insurance
policies
in
Canada
that
were
issued
or
effected
as
registered
retirement
savings
plans
or
pursuant
to
such
plans
or
to
deferred
profit
sharing
plans
or
registered
pension
funds
or
plans;
(b)
the
amount,
if
any,
by
which
the
insurer’s
income
for
the
year
from
carrying
on
its
life
insurance
business
in
Canada,
computed
in
accordance
with
Part
I,
exceeds
any
amount
deducted
by
the
insurer
under
paragraph
11
l(l)(a)
in
respect
of
noncapital
losses
in
computing
its
taxable
income
for
the
year;
and
(c)
the
aggregate
of
each
amount
(other
than
any
portion
thereof
that
is
determined
in
prescribed
manner
to
have
been
a
return
of
capital)
that
a
policyholder
became
entitled
to
receive
in
the
year
from
the
insurer
under
a
life
insurance
policy
in
Canada
other
than
a
policy
described
in
paragraph
(a),
to
the
extent
that
such
amount
is
required
by
paragraph
56(1)(d)
or
paragraph
148(1)(a)
to
be
included
in
computing
the
policyholder’s
income
or
is
an
amount
referred
to
in
paragraph
212(l)(o).
The
result
of
all
the
foregoing
calculations,
starting
from
gross,
to
net,
to
taxable,
determines
the
amount
on
which
the
15
per
cent
tax
under
subsection
208(1)
is
imposed.
It
is
the
position
of
the
plaintiff
that
in
filing
its
tax
return
for
the
year
1976
pursuant
to
the
formula
set
out
in
Part
XII,
it
has
properly
complied
with
its
requirements.
Counsel
for
the
plaintiff
suggests
that
a
proper
reading
and
analysis
of
Part
XII
stands
by
itself
in
the
calculation
of
tax
imposed
thereby,
that
it
constitutes
a
particular
code
containing
within
its
parameters
a
special,
unique
and
pragmatic
formula
to
determine
the
taxpayer’s
“taxable
Canadian
life
investment
income”
on
which
bottom
line
the
15
per
cent
is
levied.
Specifically,
counsel
for
the
plaintiff
argues
that
the
starting
point
to
get
through
the
maze
of
Part
XII
is
subsection
209(1)
which
defines
“gross
Canadian
life
investment
income”.
In
simpler
terms,
but
not
necessarily
all-inclusive
ones,
it
includes
the
gross
investment
revenue
from
the
insurer’s
gross
investment
in
non-segregated
property
(subparagraph
209(l)(a)(ii)).
Such
gross
investment
revenue
of
a
life
insurer
is
defined
in
paragraph
138(12)(e)
of
the
Act
to
mean
“the
aggregate
of
all
taxable
dividends,
interest,
rentals
and
royalties
included
in
its
gross
revenue
for
the
year,
.
.
.”.
Once
“gross
Canadian
life
investment
income”
as
above
calculated
is
determined,
the
next
step
is
to
establish
“net
Canadian
life
investment
income”.
The
process
for
this
is
found
in
subsection
209(2)
which
lists
in
paragraphs
(a)
to
(d)
thereof
certain
amounts
or
expenses
or
outlays
which
are
deducted
from
the
“gross”
income.
These
deductions
may
be
summarized
as
follows:
1.
outlays
or
expenses
deductible
under
Part
I
of
the
Act
to
the
extent
that
these
are
related
to
managing
non-segregated
property
and
the
gross
revenue
of
which
may
reasonably
have
been
included
in
the
insurer’s
gross
Canadian
life
investment
income,
(209(2)(a)
);
2.
all
interest
payments
incurred
in
the
course
of
carrying
on
the
insurer’s
life
insurance
business
in
Canada;
ie
whether
such
interest
payments
are
payable
with
respect
to
segregated
or
non-segregated
property,
(209(2)(b)
):
3.
capital
cost
allowances
on
depreciable
property
so
long
as
80
per
cent
of
it
was
used
to
earn
“gross
Canadian
investment
income’’
(paragraph
209(2)(c)),
including
therefore
income
from
segregated
funds
if
not
exceeding
20
per
cent
of
total
depreciable
property
(209(2)(c));
4.
one
half
or
50
per
cent
of
total
deductions
allowed
under
Part
I
in
computing
the
insurer’s
income
for
the
year
from
carrying
on
its
“life
insurance
business
in
Canada”
(209(2)(d)).
Included
as
a
deduction
made
by
the
plaintiff
under
the
last-mentioned
paragraph,
ie
paragraph
209(2)(d),
are
moneys
listed
as
‘’management
fees”
in
the
amount
of
$634,675,
an
amount
identical
to
that
claimed
by
the
plaintiff
in
its
Part
I
tax
return
and
not
otherwise
disputed
by
the
defendant.
It
is
this
amount
which
the
defendant
challenges
as
includible
in
the
calculation
of
the
tax
imposed
under
Part
XII.
Counsel
for
the
defendant
urges
a
more
selective
and,
at
the
same
time,
a
more
consistent
approach
to
such
outlays.
Counsel
argues
that
expenses
deductible
under
that
paragraph
must
be
limited
to
the
expenses
relating
to
gross
investment
income
from
non-segregated
property
and
not
to
similar
expenses
relating
to
segregated
property.
This
interpretation,
counsel
suggests,
is
more
consonant
with
the
scheme
of
the
Act
taken
as
a
whole
and
specifically
with
paragraph
4(1
)(a)
of
the
Act.
One
must
read
in
the
provisions
of
Part
XII,
he
says,
the
application
of
the
“source”
rule.
In
effect,
therefore,
there
must
be
a
matching
of
outlays
and
expenses
to
each
individual
source
of
income,
ie
income
from
“segregated”
and
“non-segregated”
property.
The
effect
of
this
interpretation
is
to
limit
the
management
fee
claimed
to
50
per
cent
of
such
fee
as
applicable
to
non-segregated
property
only.
To
advance
his
cause,
counsel
for
the
defendant
suggests
that
an
overriding
principle
of
the
Income
Tax
Act
which
assists
in
the
interpretation
of
many
of
its
provisions
is
found
in
Part
I
of
the
statute
where
the
element
of
“source”
is
established.
Subsection
4(1)
in
Part
I
is
a
general
provision.
It
speaks
“For
the
purposes
of
this
Act”
and
not
for
the
purposes
of
Part
I.
It
contains
no
exception
clause,
ie
an
expression
akin
to
“unless
otherwise
provided”.
Its
“sourcing”
rule
aids
us
in
interpreting
some
other
more
obscure
provisions
of
the
statute.
Counsel
quotes
from
Driedger*
and
from
Mr
Justice
Estey’s
decision
in
the
Stubart
Investment
caset
which
establish
that
the
whole
scheme
of
a
statute
must
be
respected
when
interpreting
a
particular
provision
thereof
and
an
interpretation
more
in
conformity
with
such
scheme
must
be
favoured.
In
essence,
the
case
for
the
defendant
is
that
segregated
funds
and
nonsegregated
funds
as
these
apply
to
life
insurers
are
equated
with
“sources”
of
income
under
subsection
4(1)
of
the
statute.
This
I
view
to
be
the
basic
argument
behind
the
defendant’s
assessment
and
the
defendant’s
own
interpretation
of
paragraph
209(2)(d)
of
the
statute.
It
is
a
tempting
argument
and
I
should
give
full
credit
to
counsel
for
its
able
presentation.
With
respect,
however,
I
must
disagree
with
it.
I
find
myself
much
more
readily
adopting
the
position
of
the
plaintiffs
counsel.
My
reasons
may
be
briefly
summarized
as
follows:
1.
An
analysis
of
what
constitutes
segregated
and
non-segregated
funds
applicable
to
life
insurers
discloses
that
there
are
included
in
their
respective
genus
any
number
of
income
sources.
These
sources
are
not
individually
related
to
a
“source”
as
that
word
has
been
adopted
or
interpreted
in
subsection
4(1).
Indeed,
each
fund
has
common
sources
ie
interest
income,
dividend
income,
royalty
and
rental
income,
and
other
income,
which
are
not
there
by
reason
of
the
“source”
provision
in
subsection
4(2)
but
by
reason
of
section
81
of
the
Canadian
and
British
Insurance
Companies
Act
which
imposes
on
an
insurer
the
obligation
to
segregate
certain
funds
to
meet
its
reserve
commitments
to
its
policy-
holders.
2.
I
should
not
believe
that
the
application
of
“source”
in
subsection
4(1)
to
the
genus
of
a
segregated
fund
“source”
in
the
narrow
confines
of
Part
XII
is
warranted.
In
that
regard,
I
hold
to
the
view
that
Part
XII
is
substantially
a
specific
tax
treatment
code
with
ingenious
and
quite
unique
calculations
set
out
for
the
purpose
of
imposing
some
tax
on
policyholders
whose
insurance
income
would
otherwise
be
tax
free.
The
intricate
detailing
of
the
whole
of
Part
XII
will
be
observed.
It
will
be
noted
that
Part
XII
provides
for
the
determination
of
“gross
Canadian
life
investment
income”,
of
“net
Canadian
life
investment
income”
and
finally,
of
“taxable
life
investment
income”.
It
seems
clear
from
the
care
with
which
these
provisions
are
drafted
that
the
bottom
line
on
which
the
charging
provision
of
subsection
208(1)
applies
is
the
result
of
a
distinct
and
calculated
approach
to
a
special
taxation
scheme.
3.
In
this
light,
the
scheme
or
framework
interpretation
rule
advanced
by
the
defendant
might
be
applicable,
but
not
in
the
manner
suggested
by
counsel.
I
would
think
that
such
a
rule
must
first
be
applied
to
an
analysis
of
Part
XII
as
a
whole.
Difficulties
in
the
interpretation
of
its
various
provisions
must
be
resolved
in
favour
of
conformity
within
the
context
of
that
Part
before
leaping
back
to
section
4
in
Part
I.
Such
an
internal
analysis
of
Part
XII
leads
me
to
conclude
that
paragraph
209(2)(d)
speaks
for
itself
in
allowing,
for
the
purposes
of
Part
XII,
half
of
the
deductions
allowed
in
Part
I
on
the
income
earned
by
the
life
insurer
on
its
life
insurance
business
in
Canada.
Perhaps
this
is
not
the
disposition
which
Parliament
intended
but
I
must
nevertheless
be
faithful
to
the
principle
that
Parliament’s
intention
is
determined
by
what
its
statute
says
and
not
by
what
Parliament
intended
to
say
in
its
statute.
4.
The
foregoing
observation
should
not
be
construed
as
an
acknowledgment
that
Parliament’s
intention
has
in
fact
been
frustrated
by
the
language
used
in
paragraph
209(2)(d).
Part
XII
of
the
statute
contains
a
number
of
key
expressions,
each
of
them
well-defined.
As
noted
earlier,
“gross
Canadian
life
investment”
is
defined,
“net
Canadian
life
investment
income”
is
defined
and
“taxable
Canadian
life
investment
income”
is
defined.
The
expression
“‘life
insurance
business
in
Canada”
is
not
defined
but
it
can
only
mean
the
total
operation
of
a
life
insurer’s
insurance
business
in
Canada
no
matter
what
aspects
of
it
might,
in
due
course,
come
under
a
“segregated”
or
“non-segregated”
umbrella.
Indeed,
Part
XII
contains
a
number
of
formulae
where
these
concepts
are
found.
Subparagraphs
209(l)(a)(i)
and
(ii)
specify
gross
investment
revenue
from
non-segregated
property.
Paragraph
209(2)(a)
speaks
of
outlays
for
the
purposes
of
managing
“non-segregated
property”.
It
seems
clear
in
these
provisions
that
the
“nonsegregated”
property
is
one
of
the
key
factors
to
be
fitted
into
the
calculations.
When
it
comes
to
paragraphs
209(2)(b)
and
209(2)(d),
however,
the
key
factor
appears
to
be
clearly
defined
as
the
insurer’s
“life
insurance
business
in
Canada’’.
It
will
be
noted
in
this
respect
that
for
purposes
of
Part
XII,
paragraph
209(2)(b)
allows
an
insurer
all
its
interest
expense
incurred
in
carrying
on
its
business,
whether
or
not
such
expense
is
attributable
to
segregated
as
against
non-segregated
property.
Such
a
selective
approach
in
building
up
a
whole
formula
on
which
the
special
15
per
cent
tax
is
to
be
levied
is
further
indication
to
me
that
the
paragraph
in
dispute
should,
if
at
all
possible,
be
construed
within
the
didactic
confines
of
Part
XII.
5.
If
the
calculations
required
to
determine
“net
Canadian
investment
income’’
allow
Life
insurance
business
income
deduction
under
Part
I,
the
50
per
cent
limitation
thereon
gives
it
some
credence.
This
is
the
point
forcefully
advanced
by
the
plaintiffs
counsel
both
in
written
brief
and
in
verbal
argument
before
me.
Such
a
50
per
cent
limitation
is
indicative
of
both
the
‘’arbitrariness”
of
many
provisions
of
Part
XII
as
well
as
the
pragmatism
applied
in
resolving
in
a
simple
and
expedient
way
the
complex
calculations
which
might
otherwise
have
been
required
to
determine
what
expenses
are
attributable
to
the
investment
or
“savings”
portion
of
the
insurer’s
business
and
the
“risk”
portion
in
its
business.
Cutting
all
deductions
by
half
is
consonant
with
the
scheme
or
framework
of
Part
XII
and
not
inconsistent
with
Part
I
of
the
statute.
6.
Buttressing
this
pragmatic
or
ad
hoc
approach
to
paragraph
209(2)(d)
are
other
provisions
of
subsection
209(2)
to
which
I
have
referred.
Paragraph
209(2)(b)
has
an
arbitrary
rule
with
respect
to
interest
expense.
paragraph
209(2)(c)
imposes
another
arbitrary
rule
with
respect
to
capital
cost
allowances
otherwise
allowable
under
paragraph
20(1
)(a)
of
the
statute.
When
it
comes
to
paragraph
209(2)(d),
the
expression
“life
insurance
business
in
Canada”
should
not,
in
my
view,
carry
a
connotation
different
to
that
found
in
paragraph
209(2)(b)
where
the
same
concept
is
found.
To
do
so
would
bring
into
the
wording
of
the
statute
something
which
is
not
there.
7.
Counsel
for
the
defendant
warned
me
against
adopting
any
interpretation
of
the
paragraph
which
would
strike
a
disharmonie
chord
with
Part
I
of
the
statute.
An
interpretation
which
is
on
key
with
the
statute
as
a
whole
is
to
be
preferred.
I
agree
with
counsel
in
respect
of
this
principle
but
I
do
not
agree
that
my
finding
brings
disharmony
to
the
statute.
It
merely
recognizes
a
minor
counterpoint
to
the
flowing
structure
of
the
whole
statute.
The
succession
of
the
Interprovincial
Pipe
Line*
cases
cited
by
counsel
is
of
limited
assistance
to
the
case
at
bar.
These
cases
deal
with
the
classical
“source
rule”
under
subsection
4(1)
of
Part
I
of
the
Income
Tax
Act
or,
to
be
more
accurate,
the
predecessor
to
that
Part.
Having
found
that
Part
XII
contains
within
its
ambit
its
own
internal
“source
rule”
(and
I
use
this
expression
with
some
reluctance),
I
must
rely
exclusively
on
the
wording
of
the
statutory
provision.
The
appeal
is
accordingly
allowed
and
the
plaintiffs
assessment
under
Part
XII
for
the
1976
taxation
year
is
referred
back
to
the
defendant
for
the
necessary
reassessment,
the
whole
with
costs.