15-001-XIE
Gross
Domestic Product by Industry
November
2001 |
Appendices
Appendix IIa - GDP by Industry: Overview
1. Gross Domestic Product by Industry
at Basic Prices, in Constant Dollars
Definition
Gross Domestic Product (GDP) by industry at basic prices is a measure
of the economic production which takes place within the geographical
boundaries of Canada. As the name suggests, it is designed to show the
industrial distribution of total output. GDP for a single industry is
also referred to as net output, or value added. It is equal to the total
value of production by the industry (gross output) less the value of
inputs purchased from other industries or imported (intermediate
inputs). Primary factors of production, such as labour and capital,
produce gross output by adding value to the intermediate inputs. Value
added can therefore also be derived as the remuneration to the primary
factors of production, that is as the sum of labour income,
depreciation, and profits. GDP at basic prices for the economy as a
whole, is the sum of net output, or value added, for all industries.
Summing net rather than gross outputs avoids double counting the output
of one industry which is an intermediate input to another.
The term "gross" in Gross Domestic Product means that capital
consumption costs, that is the costs associated with the depreciation of
capital assets (buildings, machinery and equipment), are included.
"Constant dollars" means that the estimates are valued at the prices
that prevailed in some base year rather than those of the current year.
As a result, changes in the estimates of value added by industry reflect
changes in the physical volume of production, and do not reflect changes
that are simply due to price fluctuations. With this publication, the
base year for constant price estimates is 1997 (see section 2.2).
"By industry" implies a classification system. The North American
Industrial Classification System (NAICS) replaced the Standard
Industrial Classification (SIC) with the release of November 2001 data (see
Section 3).
2. Major Changes and Differences
Between Industry-Based and Expenditure-Based GDP
2.1 Differences in Objectives
The Canadian System of National Accounts (CSNA) publishes two
constant dollar subannual measures of Gross Domestic Product: (1) the
monthly GDP by industry and (2) the quarterly expenditure-based GDP.
Both measures are designed to measure total production in Canada, but
serve different analytical purposes. GDP by industry provides
information about the sources of output, while the expenditure-based GDP
shows the disposition of output among the various categories of final
demand. The two series yield similar results, but because of conceptual,
statistical and methodological differences, they are not identical.
2.2 Rebasing and Deflation Methods: Fixed-Weighted Laspeyres
versus Chain-Weighted Fisher
In order to analyse the behaviour of GDP independently of the
influence of price changes, GDP is calculated for each industry in
constant prices. Constant price measures inform about the volume of
goods and services produced, independently of changes in prices. Such
series are generally calculated by choosing a reference period in the
past, called the base year, and valuing current production of goods and
services in the prices of that year.
With the release of the November 2001 data, a new base year was adopted.
The year 1997 (previously 1992) became the new base year for the
constant dollar estimates.
Currently, real GDP by industry estimates are calculated using a
Laspeyres formula chained periodically. The Laspeyres formula is
fixed-weighted using weights based on some fixed point in the past. This
formula basically adds up the volume changes in GDP by using the price
levels of the base year as the weights. It is chained by changing the
base year periodically. The base years are 1961, 1971, 1981, 1986, 1992
and 1997. Each time span is added up in its own base year prices and
then the spans are linked together by measuring each base year on two
price levels (old base and new base) and using the ratio of the two to
link them together. One of the effects of linking together series in
this manner is the loss of additivity in the period prior to the new
base year (i.e., prior to 1997, aggregates will not generally be equal
to the sum of their components).
The selection of the base year can have significant consequences, as
different base years may yield different growth rates in total GDP and
other aggregates. Consider an industry whose output price has declined
relative to that of other industries, between two years. The
contribution of this industry to total output will be larger when valued
in prices of the earlier period since the relative price was larger in
that period, and movements in this industry will have more impact on
movements in total output. If the industry is growing faster than
average, valuing output in prices of the earlier period will result in a
total GDP measure that grows faster than it would were output valued in
prices of the later period.
Ideally the base year is a typical year, followed by a number of
years in which the relative prices of commodities remain stable. In a
dynamic economy, however, relative prices constantly shift due to such
factors as uneven technological developments in different industries,
variations in productivity, shifts in consumer demand, cycles in
economic growth and so on. The more remote a base year becomes in time,
the more today's relative prices will have changed compared to those of
the base year, and the less prices of the base year will be relevant for
the current period. The usefulness of constant price estimates therefore
diminishes as we move away from the base year. The rate of obsolescence
depends on the degree of relative price change.
Recently, with the rapid expansion of the Information and
Communication Technology (ICT) industries in Canada, the current
Laspeyres volume measure has produced significantly biased results.
Growth has been overestimated because prices of equipment and services
related to this fast-growing sector of the economy have declined
dramatically since the base year 1992 because of rapid technological
change. The Laspeyres index measures changes in GDP by adding up the
quantities produced of these commodities using 1992 price levels as
weights. This is the equivalent of giving these commodities roughly four
times the weight they would have at price levels of 2000. This
over-weighting of the Laspeyres index, called "substitution bias", is
why the Laspeyres formula produces a growth rate of GDP that is at the
upper limit of possible measures. It does not compensate for
substitution to lower-priced commodities.
Changing the index formula to a Paasche index, where current prices
would always be used as a weight base, is not a solution. This results
in a bias opposite to that of the Laspeyres, with a tendency to
understate growth in GDP as ICT prices fall monotonically. Growth in the
base year would be measured using current prices, which is as faulty as
the reverse. So with a Laspeyres index producing an upper bound to
measurement of economic growth and a Paasche index producing the lower
bound, the Fisher index, which is the geometric average of the two,
follows a more stable middle path.
The expenditure-based quarterly GDP adopted a quarterly chained
Fisher index as the official measure of economic growth at the time of
their annual revision in May 2001. The new Fisher formula, which is the
middle ground between the Laspeyres and the Paasche, is rebased each
quarter in order to minimize the bias introduced by dispersion.
In order to produce chained Fisher indices for the industry-based GDP
measures, extensive development work is required that will take place
between now and the Fall of 2003. In the interim, steps will be taken to
improve the timeliness of rebasing in an effort to retain reasonable
comparability between the expenditure- and industry-based measures.
Provincial GDP by industry measures (annual) will also be rebased to
1997 for the 2001 Fall release, and will be converted to an annual
chained Fisher in two years, in tandem with the expenditure-based
Provincial GDP estimates.
2.3 Market Price and New Basic Prices Valuation
Because expenditure-based GDP is viewed from the perspective of
purchasers, the analysis is concerned with final demand categories such
as personal expenditure, government expenditure, capital formation, and
exports and imports. Output is valued at market prices, which reflect
what purchasers actually pay for the goods and services. Market price
valuation includes indirect taxes such as sales and excise taxes,
licences, property taxes, etc., and excludes subsidies.
Since its inception, GDP by industry has been measured at factor
cost. This measure differed from the market price measure by its
exclusion of taxes on production (formerly called indirect taxes) and
the inclusion of subsidies. While the market price measure represents
the value of GDP as paid for by final consumers, the factor cost measure
took the point of view of producers.
With this revision, value added is no longer measured at factor cost,
but instead at basic prices. This new measure adds to the factor cost
measure some taxes on production (such as property and payroll taxes,
but not federal or provincial sales taxes), and subtracts some subsidies
(such as labour-related subsidies, but not product-related subsidies).
The end result is that the new basic prices measure of GDP stands
somewhere in between the lower and upper bounds defined by the factor
cost and market price measures, respectively.
The difference between the three measures can be
illustrated using data from the 1997 Input-Output tables in constant
prices for total GDP.
| 1. Value of output at modified basic prices for
total economy (in billions): |
1664 |
| 2. Plus value of subsidies on products |
8 |
| 3. Less value of intermediate goods at purchasers' prices
(including taxes) |
855 |
| equals |
| 4. Gross Domestic Product at basic prices |
817 |
| 5. Less other taxes on production (i.e., excluding taxes
on products) |
49 |
| 6. Plus other subsidies on production |
1 |
| equals |
| 7. Gross Domestic Product at factor cost |
769 |
| 8. Plus net taxes on production (taxes less subsidies) |
116 |
| equals |
| 9. Gross Domestic Product at market prices
|
885 |
Indirect taxes fall more heavily on some products (for example liquor
or tobacco) than on others, and consequently the contribution of
specific industries to total GDP is more accurate if such taxes are not
taken into consideration. GDP by industry is therefore estimated at
basic prices.
2.4 Statistical Differences
The estimates of monthly GDP by industry and the quarterly
expenditure-based GDP are built up somewhat independently. The estimates
are prepared using various data sources which, in addition to being
weighted differently, also result from different estimation methods. For
example, the industry-based estimates do not rely on data for exports
but are based on manufacturers' shipments data. The expenditure-based
estimates of GDP use exports, but not shipments. Since exports by
manufacturers must have been shipped, both systems will reflect the same
underlying activity. However, since they do so via different statistical
vehicles, there will be statistical differences between the two
systems.
3. Other Major Changes Introduced with
this Revision
3.1 Classification: Adoption of the North American Industrial
Classification System (NAICS)
With the November 2001 release, the 1997 North American Industry
Classification System (NAICS-1997) replaced the 1980 Standard Industrial
Classification (SIC-1980). NAICS is an industry classification system
developed by the statistical agencies of Canada, Mexico and the United
States. Created against the background of the North American Free Trade
Agreement, it was designed to provide common definitions of the
industrial structure and a common statistical framework to facilitate
the analysis of the three economies. The adoption of a common system
ensures that statistical agencies in the three countries can produce
information on inputs and outputs, industrial performance, productivity,
unit labour costs, employment, and other statistics that reflect
structural changes occurring in the three economies.
NAICS is based on a production-oriented or supply-based conceptual
framework. This means that producing establishments are grouped into
industries according to similarity in the production processes (i.e.,
the similarity of input structures, labour skills, etc.) used to produce
goods and services. This is a marked departure from previous
classifications systems, where the guiding principle in delineating
industry boundaries was the commodity itself (i.e., establishments were
grouped together based on which commodities they produced or,
alternately, the end use to which they were put). For example, under the
SIC-1980, manufacturing of motor vehicle plastic parts (SIC 3256) was a
subdivision of motor vehicle parts production (SIC 325), because of
similarity of product/end use. However, under NAICS-1997, Motor Vehicle
Plastic Parts Manufacturing (NAICS 326193) has been classified under
Plastic Product Manufacturing (NAICS 3261), reflecting the shared
similarity of production process, which involves the extensive use of
compression, extrusion and injection moulding techniques.
The numbering system that has been adopted is a six-digit code, of
which the first five digits are used to describe the NAICS levels that
will be used by the three countries to produce comparable data. The
first two digits designate the sector, the third digit designates the
subsector, the fourth digit designates the industry group and the fifth
digit designates industries. The sixth digit is used to designate
national industries. The Canadian version of NAICS consists of 20
sectors, 99 subsectors, 321 industry groups, 734 industries and 921
national industries. The 20 sectors are listed below.
Sectors
| 11 |
Agriculture, forestry, fishing and hunting |
| 21 |
Mining and oil and gas extraction |
| 22 |
Utilities |
| 23 |
Construction |
| 31-33 |
Manufacturing |
| 41 |
Wholesale trade |
| 44-45 |
Retail trade |
| 48-49 |
Transportation and warehousing |
| 51 |
Information and cultural industries |
| 52 |
Finance and insurance |
| 53 |
Real estate and rental and leasing |
| 54 |
Professional, scientific and technical
services |
| 55 |
Management of companies and enterprises |
| 56 |
Administrative and support, waste management and
remediation services |
| 61 |
Educational services |
| 62 |
Health care and social assistance |
| 71 |
Arts, entertainment and recreation |
| 72 |
Accommodation and food services |
| 81 |
Other services (except public
administration) |
| 91 |
Public administration |
The Canadian System of National Accounts only publishes 19 of these
sectors. The twentieth, Sector 55 (Management of Companies and
Enterprises) has, for reasons of historical continuity and data quality,
been grouped together with Sectors 52 and 53 into an aggregate called
Finance, Insurance and Real Estate.
Monthly GDP estimates are compiled back to 1981. At the lowest level
of aggregation of the current industrial classification framework, it is
not always possible to have an homogeneous series from 1981 to the
present. Data gaps as well as changes in the industry classifications of
1980 and 1997 are responsible for this.
With the introduction of the new 1997 NAICS, establishments are
sometimes reclassified from one industry to another. Thus a NAICS
industry with a name similar to that of the 1980 SIC may have a somewhat
different composition in terms of establishments.
In some cases the differences will be small, and in others large. In
general there is greater such definitional discontinuity at lower levels
of industrial dis-aggregation than at higher levels. For some purposes
it is useful to have definitional continuity at as low a level of
dis-aggregation as possible. For this reason, special industry groupings
that provide relatively good continuity were created back to 1981. These
groupings are identified with an asterisk (*) in Appendix I.
Special attention was given in NAICS to developing classifications
for new and emerging industries, services industries in general, and
industries engaged in the production of advanced technologies. As much
as possible, time series continuity was maintained. However, owing to
fundamental changes in the economy, to the classification's strong
production orientation and to proposals from data users, this has not
always been possible. There are such aggregates including goods
producing industries, services producing industries, industrial
production and others. They are identified in Appendix I with an
industry code prefixed by a ''T''.
3.2 Capitalization of Software
All spending by businesses and governments on software is now treated
as a capital expenditure. Previously, only a small portion, less than
20% of these expenditures, was treated as such. This will be a net
addition to GDP, both in level and in growth, in the years in which
software expenditures grow faster than other components of final
expenditure. Treating software as a capital expenditure has been
introduced by most countries, including the United States, within the
last couple of years.
3.3 Treatment of Rent Paid by Farmers
Another conceptual change concerns the treatment of rent in the Crop
Production Industry. Previously, the rent paid by farmers was considered
as an intermediate input; now it is treated as mixed income, which is a
primary input (i.e., part of GDP). This treatment conforms more closely
to the international standards as described in System of National
Accounts 1993. The end result of this change is to increase GDP in the
Crop Production Industry. However, this is offset by a decrease in the
GDP of the Lessors of Real Estate Industry, where the aforementioned
output of rent was formerly classified. This redistribution of output,
which has no net effect at the level of the total economy, brings land
use for agriculture in line with the treatment of royalties on natural
resources.
4. Methodological Overview: GDP by
Industry
4.1 Annual Benchmarks
For all but the most recent two years, the annual estimates of GDP by
industry are derived within the framework of the Input-Output accounts
by subtracting the intermediate inputs from the gross output of
industries. The data sources are typically annual surveys or censuses.
For the most recent two full years and part of the current year, GDP is
estimated using the monthly methodology described below.
4.2 Monthly Estimates
On a monthly basis information on industry activity, especially the
consumption of inputs, is not as complete as it is annually. As a
result, the monthly values of GDP by industry are projections, estimated
using proxy indicators such as gross output or employment, usually
obtained from monthly surveys see (Appendix
IIb). The main assumption inherent in the monthly projector system
is that the relationship between inputs and outputs changes slowly
enough that it can be regarded as constant over short spans of time.
This assumption will be true if both technology and the product mix
change slowly. It is therefore more useful for projecting GDP over short
time spans than over long ones, and at lower levels of industrial
disaggregation, where product mix is relatively homogenous and stable.
Prices of inputs and outputs do not necessarily evolve at the same pace
during a given period. As a result, the relationship between inputs and
gross output can be distorted at current prices. For that reason, the
assumption above only applies to the calculation of GDP at constant
prices.
4.3 Benchmarking
This is the process by which monthly estimates are adjusted to the
new benchmark levels. Since the monthly estimates are projections, they
are adjusted to sum to the annual benchmark levels once they become
available. This adjustment is based on a quadratic minimization
technique, which preserves the month-to-month movements in the original
series as much as possible, subject to the constraint that the monthly
estimates sum to the annual benchmarks.
4.4 Seasonal Adjustment
To aid interpretation of economic developments the estimates of GDP
by industry are seasonally adjusted and are published in both seasonally
adjusted and unadjusted forms. Seasonal adjustment removes movements
caused by repetitive climatic or institutional events, thereby making
the underlying activity more apparent.
The seasonal adjustment technique relies heavily on the use of moving
averages, that become less reliable towards the end of the series. To
minimize this problem, the series are projected one more year so that
the last real data point is 12 months from the end point of the extended
series. An ARIMA (autoregressive integrated moving average) model is
used for this purpose.
4.5 Trading-Day Adjustment
In many monthly economic time series a significant portion of the
monthly growth rate is associated with differences in the composition of
calendar months. For example, the volume of production usually varies
with the different days of the week, and since consecutive months
contain different combinations of days, a significant portion of the
month-to-month changes may be caused strictly by the calendar. To give a
more precise idea of economic developments, the calendar months are put
on an equalized basis. This is done by applying a smoothing procedure,
known as trading-day adjustment. Trading-day adjustment alters the
distribution of monthly output to reflect changes in production that
would occur if all months contained the same number and type of
days.
5. Dissemination
Vehicles
5.1 The Daily/CANSIM
As with many Statistics Canada series, the monthly estimates of GDP
are officially released to the public simultaneously through the
Statistics Canada Daily Bulletin and CANSIM II (Statistics Canada's
machine-readable database). The CANSIM II database is available on
Internet at the following address: www.statcan.ca. The estimates are
published approximately 60 days after the end of the reference month.
Data can also be obtained directly from the division on diskette,
print-out or other.
| Reference |
month |
Release dates |
| August |
2001 |
October 31, 2001 |
| September |
2001 |
November 30, 2001* |
|
|
| October |
2001 |
December 24, 2001 |
| November |
2001 |
January 31, 2002 |
| December |
2001 |
February 28, 2002* |
| January |
2002 |
March 28, 2002 |
| February |
2002 |
April 30, 2002 |
| March |
2002 |
May 31, 2002* |
| April |
2002 |
June 28, 2002 |
| May |
2002 |
July 31, 2002 |
| June |
2002 |
August 30, 2002* |
| July |
2002 |
September 30, 2002 |
| August |
2002 |
October 31, 2002 |
| September |
2002 |
November 29, 2002* |
| October |
2002 |
December 24, 2002 |
| November |
2002 |
January 31, 2003 |
| December |
2002 |
February 28, 2003* |
* The quarterly income and expenditure-based GDP at market price will
also be released on these dates. Release dates for the upcoming year are
published, along with those for other major economic indicators, in
December.
The data on CANSIM II appear in greater industrial detail than in the
Daily and are also in both seasonally adjusted and in seasonally
unadjusted form. They are available historically from January 1981, and
may be retrieved from the following tables (see Appendix
III):
| Frequency and adjustment of data |
Main tables |
Special aggregations |
| Annual |
3790017 |
3790020 |
| Quarterly (seasonally adjusted and unadjusted) |
3790018 |
3790021 |
| Monthly (seasonally adjusted and unadjusted) |
3790019 |
3790022 |
Information can also be obtained by calling any Statistics Canada
regional office or Industry Measures and Analysis Division
(1-800-877-IMAD).
5.2 Publication
The monthly, quarterly and annual constant price estimates of Gross
Domestic Product by industry are published in Gross Domestic Product by
Industry, Statistics Canada, Catalogue no. 15-001-XIE, available
approximately five working days after the data have been released. This
publication contains seasonally adjusted data extending back to 1997,
the current base year.
[Main menu |
Highlights |
Tables and charts |
Appendices |
User information |
Products and services |
Contact us |
Français]
All rights reserved Statistics Canada
