The capital stock survey project.
In this paper, capital is defined as a factor input that consists of commodities produced by the economic system for use in the production of other commodities. Capital stocks therefore consist of reproducible tangible assets and exclude tangible assets such as land, mineral deposits, and natural forests.
Reproducible tangible assets vary in age, in their degree of obsolescence, and in the intensity of their utilization. Consequently, although there is a substantial theoretical and technical literature about the measurement of capital, the diverse nature of capital goods and their technological evolution make the measurement of capital a difficult task.
Although there are a number of methods-such as surveys of physical assets, of book values, and of insured values-for measuring capital stocks, the "perpetual inventory" method is the one that is the most often used, prmarily because of its flexibility in creating time series of capital stocks. This method requires an investment series, the relevant price indexes, service life information, and a method of depreciation. The reliability of the capital stock time series generated using this method depends on the quality of the investment data that are cumulated, the accuracy of the price indexes that are used, the realism of the service life assumptions that are made, and the method that is used to depreciate the capital assets over their service lives.
The investment series for various vintages must be combined into overall capital stock estimates that are relevant for a particular analysis. The information needed to aggregate vintages can be summarized in terms of an "efficiency function" that gives the relative efficiency of capital stocks of different ages and that is based, for example, on the data for prices from transactions in used assets. Given the efficiency functions for existing assets and the discard rates for assets by age, it is possible to employ the perpetual inventory method to obtain annual, or more frequent, time series estimates of capital stocks.
Because the perpetual inventory method requires so many assumptions and so much data, not all of which are equal to the uses made of them, it would be useful to have periodic benchmark surveys of assets actually in place at a particular time. This paper discusses the development of such a survey.
The Feasibility Study
Because of concern about the accuracy and the adequacy of the existing data for capital stocks, Statistics Canada, the Bank of Canada, the Economic Council of Canada, and the Department of Finance established the Capital Stock Survey Project in early 1982. In 1983, 125 companies were interviewed about the availability of information on economic lives of fixed assets, sales of used assets, levels of assets, and retirement policies for assets.
In these interviews, five aspects of capital data in which information was said to be unavailable or incomplete were identified. These five aspects are (1) economic service life of fixed capital assets, (2) transactions in used assets, (3) premature retirement, or discarding, of fixed capital assets in response to changes in relative factor prices or changes in technology, (4) leased fixed capital assets, and (5) the investment in fixed capital assets for pollution abatement and control. As the final phase of the feasibility study, a pilot study was sent.
The Pilot Survey of Fixed Assets
The Pilot Survey of Fixed Assets was sent to more than 1,000 establishments. The survey questionnaire asked these establishments to report information about existing capital stocks and the flow of new investment.
The questions about capital stocks requested the following information:
* The value of fixed capital assets-specifically, the total
original cost for assets still in use, the total estimated
current (or replacement) value of assets, the average
age of assets, the estimated remaining useful life
of fixed assets, and the percentage of assets leased to
* The percentage of structures and of machinery and
equipment acquired for pollution abatement and control;
* A disaggregation of assets into categories of building
and engineering structures and into categories of machinery
The questions about the flow of new investment requested the following information:
* The expected useful life of new assets;
* The expenditures on renovation, retrofit, refurbishing,
overhaul, or rehabilitation of existing assets;
* The floor space of buildings and the number of units of
* The purpose of capital expenditures;
* The expenditures on assets for lease to and from others;
* The reasons for discards; and
* The purchases and sales of used assets.
The Results of the Pilot Survey'
Capital stock questions
The overall rate of response to the questions about capital stocks was 77 percent, and most industries had response rates near the average. The rates of response were particularly high from the petroleum and natural gas industries, petroleum and coal products manufacturers, retail trade, and services. The rates of response were somewhat lower from the electric power utilities, primary metal and nonmetallic mineral manufacturers, and finance, insurance, and real estate industries. In general, most of the companies took the time to complete the survey.
Nearly all of the respondents to the survey reported the value of fixed capital assets in terms of total original cost for assets still in use. This response is consistent with the responses of the companies during the earlier interviews. Although some companies had expressed reservations in the interviews about the availability of the data for the total original cost of fully depreciated assets still in use, their availability does not appear to have been a serious problem for the survey respondents.
Approximately 90 percent of the respondents reported the average age of assets. This high rate of response is particularly gratifying; in the earlier interviews, the initial industry response indicated that the average age of assets would be very difficult to report.
About 75 percent of the respondents reported estimates for the remaining useful life of fixed assets. However, this question was clearly somewhat difficult for companies to answer; there was considerable interindustry variation in the rates of response to this question. The rates of response were higher from the manufacturers of paper and allied products, of primary metal, and of electrical products than from the manufacturers of transportation equipment and of chemical products. It would be worthwhile to investigate the reasons for this variation in responses among industries. This overall rate of response rate should be considered positive because the initial industry response indicated that projected service lives would be very difficult to report. But these estimates must be checked to ensure that industries have not simply used tax lives to make this calculation. Flow questions
The overall rate of response to the flow questions was nearly 100 percent. About 82 percent of the respondents reported estimates for the expected useful service life of new assets. This rate of response should be considered high, because this question asks companies to predict the future and is, therefore, inherently speculative. Moreover, although it may be relatively easy to make this estimate for well-known types of capital assets, it is more difficult to make it for new types of capital assets. For example, the rate of response from retail trade was 54 percent; this low rate was, perhaps, partly attributable to the difficulty in estimating the service life of new automated machinery, such as the integrated point-of-sale systems that are designed to facilitate inventory control.
It is clear from the survey responses that firms can report the reasons for discards and the purchases and sales of used assets, though it is difficult to assess the validity of these data. The data for discards are particularly interesting because these data from the survey may be used to cross-check the estimated level of discards in each industry with comparable data from the perpetual inventory. Virtually all of the respondents who reported data for discards also gave some reason for the discards; this response is interpreted as a positive result for the feasibility of capital stock surveys. The data for discards should be especially useful when they are regularly available. In particular, the data provided in response to this question would provide direct evidence that indicates when assets are discarded prematurely in response to changes in relative factor prices.
The expenditures on assets for lease were reported by two-thirds of the respondents in services and by one-third of the respondents in finance, insurance and real estate; additionally, a significant number of respondents in retail trade reported some leasing activity. These results suggest that assets for lease are an important part of the capital stock owned by the services industry. Moreover, the survey responses clearly indicate that companies are capable of reporting this information.
The Capital and Repair Expenditures
Based on the results of the Pilot Survey of Fixed Assets, Statistics Canada revised and expanded the capital and repair expenditures survey. The revised survey collected data for the following:
* Capital expenditures-specifically, the acquisition of
new assets and the renovation, retrofit, refurbishing,
overhauling, and rehabilitation of existing assets;
* Expected useful service life of new assets;
* Original cost and age of assets that are discarded, retired,
* Sales and purchases of used assets;
* Reasons for capital expenditures and for discards;
* Expenditures for automation-for example, for robots,
automated material-handling equipment, automated
production machinery and equipment, and computerized,
numerically controlled machine tool equipment;
* Value of work in progress at yearend; and
* Value of fixed assets-that is, the cost of accumulated
capital (gross book value) and accumulated depreciation,
beginning with 1987 data.
The revised survey provides detail that has not been available; for example, 30 different types of equipment have been reported and classified according to the 45 industries that buy the equipment. (See table 1.) Moreover, an analysis of the data from this survey will now yield the amount that an industry invests, the kind of machinery being purchased, and the technological and structural changes taking place.
Studies Using the Survey Data
The data from the expanded capital and repair expenditures survey will help us to improve our understanding of the nature of investment expenditures. The survey provides new information on modernization, on overhauls, and on types of equipment. Using the survey data, researchers can study various topics, such as the importance of the investment in fixed capital assets for pollution abatement and control.
For example, based on the information gathered during the pilot survey, the weighted, average age estimates for capital equipment used in some manufacturing and non-manufacturing industries were calculated. Of the industries tabulated, the oldest equipment, on average, was found in the electrical products industry, the newest in the chemicals industry. Similar estimates can be made for subsequent years from data collected from the ongoing survey.
Based on the data from the 1985 and 1986 Capital and Repair Expenditures Surveys, the new estimates of the service lives of equipment can now be compared with the existing estimates of service lives that have been used in the perpetual inventory method. The existing estimates of service lives rely primarily on the estimates made from a variety of sources, such as corporation tax returns, the relationship between capital cost allowances and investment expenditures, and tables from Bulletin F'.2 In most cases, the new estimates of service lives are shorter than the existing ones.21
Using the new estimates of service lives collected in the new investment survey, a number of simulations have been performed to show the impact of these data on capital stock estimates. The difference between the new and the existing estimates of the service lives of gross capital stock lies in how the new estimates are introduced; that is, the earlier the new estimates are introduced and the faster one moves from the existing estimates to the new ones, the sooner the differences will show up in the capital stock estimates. Consequently, this difference will affect the growth rate of the gross stock estimates, but the differences in the annual growth rates will be small.
In our paper, J.C. Hwang and I used the price-age profiles of used assets that were reported in the revised capital expenditures survey to estimate the form and the rates of economic depreciation. We concluded that the patterns of depreciation are close to the geometric form for the manufacturing sector and are accelerated (vis-a-vis straight line) for the nonmanufacturing sector.
Other studies using the survey data are under way. One study is comparing the new estimates of service lives from Canada with those from other Organization for Economic Cooperation and Development countries. Another study is comparing the estimates of fixed assets from the survey with those from administrative data.
Based on the work that we have done on the capital stock survey project, we are planning to use the capital and repair expenditures survey and the data collected from it to produce an integrated system for measuring capital stocks and flows in 1989. Producing this system will require the following activities:
* Tabulating new industry benchmarks for construction
and for machinery and equipment;
* Incorporating the new estimates of service lives into our
calculation of fixed assets;
* Introducing new forms and new rates of depreciation
that were suggested by our research; and
* Introducing the normal distribution model as our new
We hope that this system will provide the accurate data that are needed for economic analysis.
|Printer friendly Cite/link Email Feedback|
|Publication:||Survey of Current Business|
|Date:||May 1, 1989|
|Previous Article:||U.S. business enterprises acquired or established by foreign direct investors in 1988.|
|Next Article:||The social policy simulation database and model: an example of survey and administrative data integration.|