9.3 Gross fixed capital formation in non-residential structures

General concepts

9.52 Gross fixed capital formation in non-residential structures represents about 32% of investment or 6% of GDP (in 2000). It is a key variable in the macroeconomic system. It tends to be cyclical, as business investment in non-residential structures tends to lag the business cycle, largely because of its long-term nature involving major contractual commitments. The time that elapses between a businesses' decision to invest and the deployment of new capital results in a cascading lag.

9.53 When the economy is expanding, businesses make profits and begin to consider increasing their capacity either by building new plants, purchasing equipment or by improving existing facilities to meet growing demand. Before new capital can go into operation, there is the planning and design phase, followed by the construction phase. Depending on the nature and size of the investment, this process can take anywhere from a few months to many years, even as much as a decade. This cascading lag is a major determinant of the cyclical nature of investment and many economists argue that it is a major determinant of the cyclical nature of the economy.

9.54 The component non-residential structures includes the gross value of:

  • all new non-residential construction put in place with the exception of defence installations (treated as government current expenditure on goods and services);
  • all additions and major renovations; and
  • all conversions and alterations that extend the life of an existing asset.

9.55 In the IEA, these estimates are grouped into two categories. The first category, building construction, refers to investment in industrial, commercial and institutional buildings. It includes any permanent structure with walls and a roof, any attachment that forms part of the structure, plumbing, electrical wiring, air conditioning or elevator installations. It covers hotels, office buildings, railway stations, schools and shopping centres.

9.56 The second category, engineering construction, includes all capitalized costs: legal fees, engineering and architectural fees, capitalized interest and work completed for the business by its own employees. It includes items such as bridges, roads, highways, waterworks, sewage systems, airports, seaways, power line construction, oil well drilling, mine development, dams, street lighting, railway tracks and pipelines.

9.57 The estimates include both contract work and work completed by the firm's own employees. The cost of site preparation and land improvement is included; however the purchase value of land is excluded. Transfer costs associated with the purchase of existing fixed assets are included. The estimates cover the value of work put in place (WPIP) during the reference period.

Table 9.5 Gross fixed capital formation in non-residential structures in the Income and Expenditure Accounts, 2000. Opens in a new browser window.

Table 9.5
Gross fixed capital formation in non-residential structures in the Income and Expenditure Accounts, 2000

Annual estimation methods and data sources

9.58 The annual benchmark estimates (for t-4 and t-3 years) for gross fixed capital formation in non-residential building and engineering construction are reconciled to the final demand matrix of the Input-Output Tables (IOT). The Industry Accounts Division (IAD) primarily uses information from the Investment and Capital Stock Division's Capital Repair Expenditures Survey1 (CES) by asset type. This information is based on the North American Industry Classification System (NAICS). The data are produced on a January-December basis and are consistent (in most cases) with national accounts' concepts.

9.59 The Capital Expenditure Survey totals are not identical to IEA's gross fixed capital formation measures because a number of adjustments are made to ensure consistency with the CSNA definitions. One difference between the survey and IEA data is related to sectoring. The survey data are reclassified because the business and government sectors are defined differently in the CSNA.

9.60 In addition, IEA totals are based on other sources of information, such as the Quarterly Survey of Financial Statistics for Enterprises.2 This survey focuses on revenues, profits and assets, but also includes a number of questions on capital spending. It differs from the Capital Expenditure Survey in coverage (based on enterprises rather than establishments) and content (e.g., includes the purchase price of land and used buildings).

9.61 Information is also taken from Investment and Capital Stock Division's Investment in Non-residential Building Construction data based on the Building Permits Survey for comparison with Capital Expenditure Survey data on the value of some investment projects.

9.62 The Input-Output supply-disposition framework, which is used to refine the estimates for the benchmark years, and the values for the various sub-items that are generated with the methodologies developed for the production of quarterly estimates, also contribute to the accuracy or completeness of the investment estimates.

9.63 As in the case of the annual benchmark estimates, the Capital Expenditure Survey is the main source of information. For the t-2 period, actual data3 from the survey are used whereas for the t-1 period, the preliminary actual estimates are used as the main indicator along with the quarterly estimates based on related indicators.

Quarterly estimation methods and data sources

9.64 In the absence of a quarterly capital expenditure survey, related indicators must be used. Non-residential building and engineering construction estimates are published separately on a quarterly basis. The methodologies used to produce these aggregates are closely related.

9.65 There are essentially two approaches: a detailed approach and a global approach. The results of both approaches are carefully examined and analysed in order to arrive at an acceptable estimate. In this process, we also take into consideration the annual forecasts of the Capital Expenditure Survey (CES) for the current year (spending intentions released each February).

The detailed approach

9.66 The detailed approach consists of multiplying the most recent quarterly estimate of capital spending for each component by the percentage change in a related indicator, and then summing the results to produce an overall estimate. This method is used for building and engineering construction expenditures for both the government and business sectors.

Government sector

9.67 For building construction, the quarterly indicator used is the investment in non-residential institutional building construction, produced by Investment and Capital Stock Division. It is based on the Building Permits Survey of municipalities, which provides information on construction intentions for this type of building. Work put in place (WPIP) patterns are assigned to this category of structure. The patterns which are used to distribute the value of building permits according to project length, differ by type and value of construction project— a project worth millions of dollars will usually take longer to complete than a project worth $100,000.

9.68 Engineering construction expenditures are calculated for highways, bridges, railways and other engineering. Material costs (asphalt purchases in cubic metres multiplied by the Industrial Product Price Index (IPPI) price for asphalt) and labour costs based on earnings in the highway, streets and bridge industry from the Survey of Employment, Payrolls and Hours (SEPH) make up the highways' estimator. This same estimator is used to estimate expenditures on the other, less important, components.

Business sector

9.69 For the buildings category, the quarterly indicators are the investment in non-residential industrial and commercial buildings, produced by Investment and Capital Stock Division.

9.70 The business sector's engineering expenditures are made up of two categories: private roads and other engineering. The private roads estimator is the same one used for public highways described in paragraph 9.68. One of the estimators used for other engineering is based on capital spending (reported in the Quarterly Survey of Financial Statistics for Enterprises) by large companies, in mining, telecommunications, railways, etc., which invest mainly in engineering. Other information used includes such items as capital spending by provincial electric utilities and the value added in drilling and rigging services4 in the oil and gas industry.

9.71 Finally, qualitative information on major construction projects is extracted from budget documents, newspapers articles, specialized publications and project inventories maintained by private consulting firms, etc.

The global approach

9.72 The global approach consists of constructing an overall indicator of capital spending on construction. This indicator or estimator is also used to derive quarterly estimates of investment in non-residential construction. Total construction spending5 is estimated primarily by combining data on employment, wages and shipments of various materials used in construction. Other variables, such as profits in the construction industry and inventories of building materials are also taken into account.

9.73 Employment in construction is taken from the Labour Force Survey (LFS). Average hourly earnings are pulled from the monthly Survey of Employment, Payroll and Hours (SEPH). Another key monthly indicator, the value added in drilling and rigging services, is obtained from provincial ministries in oil and gas producing provinces. These indicators are combined to obtain an overall level of construction activity for each quarter. The quarterly estimate for residential construction (new construction, alterations and improvements) is then subtracted from the total, yielding an estimate of spending on non-residential construction. The resulting estimates are then compared with the annual estimates from the Capital Expenditure Survey. The allocation of expenditures between the government and business sectors6 is based on Private and Public Investment Intentions.7

9.74 Little information is available on the supply side, other than data on building materials from various monthly surveys, such as the production of cement and asphalt, sales of lumber, and imports and exports of other building materials. Supply based estimates for the construction industry tend to be lower than the demand-side estimates incorporated into the National Accounts.

9.75 Finally, the quarterly data are benchmarked to the annual levels at the time of the production of the annual estimates. These annual estimates reflect the most recent aggregates from the Input-Output Tables, data from the Capital Expenditure Survey by Type of Asset8 and the latest Capital Expenditure Survey.

9.76 For the quarters of the current year, the series are built on the basis of the movement of the specially constructed indicators discussed previously, of which inputs are all available up-to-date on a sub-annual basis.

Provincial and territorial estimation methods and data sources

9.77 The provincial distribution of non-residential construction investment comes from the Capital Expenditure Survey. Note that the provincial estimates are benchmarked to the most recent provincial Input-Output Tables (for t-4 and t-3 years).

Deflation – Estimates in real terms

9.78 Estimates are produced for each component at current prices and then deflated at the following levels of detail:

  • Non-residential buildings
  • Engineering
    • Highways and roads
    • Railways
    • Other engineering

9.79 Both output and input price indexes are used. Output price indexes produced by the Prices Division are used to deflate building investment, while input prices indexes, most of them produced within the IEA, are used to deflate engineering construction. The input price indexes are based on:

  • average earnings for the labour component;
  • IPPI for the materials; and
  • a mix of average weekly earnings indexes and consumer price indexes for overhead costs.

9.80 These price indexes are then weighted together on the basis of detailed annual estimates from the Input-Output Tables.

9.81 The non-residential buildings deflator is based on non-residential building construction price indexes, with contracted investment by industry having a weight of 90% and own-account work having a weight of 10%.

9.82 Output price indexes measure the variation in contractors' selling prices for non-residential structures (i.e., commercial, industrial and institutional buildings). These indexes relate to both general and trade contractors' work and exclude land improvement costs and real estate fees. Prices for work put in place (WPIP) are provided by the Prices Division, which surveys subcontractors and general contractors whose bids are based on specifications and quantities set by real estate markets. The latter survey covers such items as overhead and profits that reflect market conditions. Prices for certain materials, labour rates, equipment rental costs, municipal charges and sales taxes are obtained from a variety of secondary sources.

9.83 Prices for own-account work are obtained from a fixed-weighted composite index of average hourly earnings for the construction sector, based on the Survey of Employment, Payrolls and Hours (SEPH) data; materials prices, based on the Industrial Product Price Index (IPPI); and overhead costs, based on various price indexes, such as average weekly earnings indexes and various consumer price indexes.

9.84 The highways and roads, railways and other engineering deflators are all input price indexes based on a fixed-weighted composite of wages, materials and overhead costs. Table 9.6 shows the relative weights of the deflators, derived from the 1997 Input-Output Tables. Indexed average weekly earnings from Survey of Employment, Payroll and Hours are used for the wages portion. The materials component of the deflator is based on IPPIs. The overhead costs share is based on a mix of average weekly earnings indexes and consumer price indexes.

Table 9.6 Relative weights of engineering deflators. Opens in a new browser window.

Table 9.6
Relative weights of engineering deflators


Back to


1. Survey no. 2803. The data are published in Private and Public Investment in Canada, Intentions, catalogue no. 61-205. This survey is often referred to as Private and Public Investment or PPI.

2. Survey no. 2501. The data are published in Quarterly Survey of Financial Statistics for Enterprises, catalogue no. 61-008.

3. For more information on the Capital Expenditure Survey concepts of intentions, preliminary actual and actual expenditures, see Appendix 1: A Brief Description of the Capital Expenditure Survey.

4. Drilling activity is measured in metres, while rigging is measured in rig operating days by depth (in metres).

5. Includes both residential and non-residential expenditures.

6. Some government enterprises are transferred from the government to the business sector to maintain consistency with the CSNA.

7. Survey no. 2803. The data are published in Private and Public Investment in Canada, Intentions, catalogue no. 61-205. This survey is often referred to as Private and Public Investment or PPI.

8. This information is also produced by the Capital Expenditure Survey (CES). Previously published in Capital Expenditures by Type of Asset, catalogue no. 61-223, now available only on CANSIM.