Archived Content
Information identified as archived is provided for reference, research or recordkeeping purposes. It is not subject to the Government of Canada Web Standards and has not been altered or updated since it was archived. Please "contact us" to request a format other than those available.
| 39 |
Software has been essentially identified in merchandise
import statistics since 1996. Prior to then it is listed under
several HS codes going back to 1988, but can’t be separately
identified from recorded tapes/discs in general. Software wasn’t
even mentioned in earlier Canadian import classifications. Examination
of the import data 1996-1999 on software and those commodities
with which software had formerly been grouped (i.e., other recorded
tapes/discs) showed that software imports, on average, were growing
about 3.3 percentage points faster. This is assumed to have been
the case over the period 1988 through 1995, and a corresponding
adjustment is made in carrying back the software import series.
|
| 40 |
Two special steps should be noted. First, the 1981 estimate
is based on 1980 earnings data from the 1981 Census (the Census
asks for earnings in the previous year) indexed one year ahead
in order to start the investment series in 1981. Second, a switch
from the 1980 Standard Occupation Classification (SOC) to the
1991 version during the 1991 Census (which is double-coded to
both versions) shows lower estimates (10% lower) for 'computer
programmers' and 'systems analysts' as defined in the 1991 SOC.
Using the double-coded results, the 1981 and 1985 estimates are
ratio-adjusted (i.e., reduced) to be comparable with the subsequent
estimates on a 1991 SOC basis. |
| 41 |
The implicit assumption that this ‘ratio’ doesn’t
change significantly over time is particularly suspect in and
around the year 2000. Leading up to Y2K many own-account software
development projects were likely put on hold as programmers and
systems analysts became increasingly focussed on Y2K priorities.
Since then, there has been a return to business as usual, with
programmers/analysts devoting more of their time to development
projects as opposed to Y2K. An adjustment is however made to remove
a significant jump in employment of programmers and systems analysts
in the public sector in 1999, which was thought to be clearly
Y2K induced. This is tantamount to adjusting the ratio to remove
Y2K effects. |
| 42 |
The earnings of programmers and systems analysts in the computer
services industries is not included because, as mentioned earlier,
most of these earnings are kept out of the own-account software
estimates to avoid double-counting with purchased software. Insofar
as the preparation (assessment, testing, patching, etc.) for Y2K
was contracted out to computer services firms, this exclusion
serves to temper Y2K effects on the own-account estimates for
the business sector. The growth of private sector employment of
programmers and systems analysts more than tripled between the
first and second half of the nineties, from an average 5% per
year over 1990-1994 to 16% over 1995-1999. All of this increase
came in the computer services industries, where the growth of
employment of programmers and systems analysts more than doubled,
from 13% per year in the first half of the nineties to 27% in
the second. In contrast, the employment of programmers and systems
analysts in the rest of the private sector was relatively flat,
growing an average 1% per year in both the early and late nineties.
It is this latter, modest employment growth that drives the increase
in the business sector own-account software in the late nineties
(along with increases in the hourly earnings and the hours of
work of programmers and systems analysts outside of the computer
services industries). |
| 43 |
It was not possible to re-classify all software under the ‘software
products development’ commodity and at the same time maintain
the historical series on Canada’s trade in goods and in
services. To maintain these totals, some software is retained
under the good ‘recorded media’. |
| 44 |
All of the former intermediate use of pre-packaged software,
by industry, is reclassified as investment. However, because custom-design
software was formerly part of the broader commodity ‘professional
and processing computer services’, its use as an intermediate
input was not explicitly identified. The amounts removed from
intermediate use of this broader commodity are such that the national
totals for software investment are satisfied. The distribution
of this ‘custom-design’ component across industries
is proportional to hardware investment by industry. |
| 45 |
The software retained under these two commodities is identifiable
from other recorded media and other royalties and license fees
in that only the software is classified as investment. |
| 46 |
In order to determine software capital consumption, by province/territory,
for business sector industries (as well as for government sub-sectors),
the investment series (net of software already capitalized) are
carried back even further to 1976 (own-account and custom-design)
and 1978 (pre-packaged). This is done on the basis of labour income
by industry, province and territory. The results are constrained
to meet the national estimates for investment in each type of
software, by sector, established to calculate software capital
consumption at the national level. |
| 47 |
Yukon and the Northwest Territories are assumed to maintain
their 1998 shares (which are based on the census benchmark for
1995) through 1999-2000, and LFS shares for the provinces are
adjusted downwards accordingly. It might be noted that from 1999
on, estimates are made for the Northwest Territories (incl. Nunavut),
and then split into Northwest Territories (excl. Nunavut) and
Nunavut on the basis of each territories’ share of GDP,
by industry. |
