RDP 9612: External Influences on Output: An Industry Analysis Appendix A: Description and Sources of Data

A.1 Production Data

The data are constant-price series. For the United States, the output measure is gross product by industry, estimated as gross output (for example, sales plus the change in inventories) less intermediate inputs (consumption of goods and services purchased from other industries or from overseas). In most cases, constant dollar gross product is estimated by the double deflation technique (that is, constant-dollar gross output less constant-dollar purchased input by industry). US production data are based on the (American) Standard Industrial Classification (SIC). Estimates from 1977 to 1987 are based on the 1972 SIC, while those from 1988 onwards are based on the 1978 SIC. The two standards overlap in 1987. The changes appear minor, in the sense that the changes only occur at a highly disaggregated level (level 5) and only for a few categories (and here it seems that it is largely a name change). The exception is instruments (a level 5 category). The source of the data is the Bureau of Economic Analysis, Department of Commerce, United States Government.

The Australian data are constant-price output by industry (or industry-revalued estimates, as the ABS call them) from the ABS release, National Income, Expenditure and Product, Cat. No. 5206.0. The production data are in 1989/90 prices and the annual figures are the calendar year sum of the seasonally adjusted quarterly series. The industry classification of the national accounts changed in September 1994 from the ASIC to the ANZSIC system. The data used are all estimated on the ASIC basis. The latest data were drawn from the September quarter 1994 release. These data go back to September 1984. Data before this date were spliced onto these series using growth rates from the earlier periods. This way of backcasting avoids level shifts.

The ABS estimate constant-price industry output by one of three methods. The standard is the gross output method by which base year industrial output is extrapolated by movements in constant price estimates of output. The second, applied to agriculture, mining and gas, is double deflation, by which constant price estimates of intermediate input are subtracted from constant price estimates of gross output. The third, applied to public administration, finance, and community services, is to use hours worked to extrapolate base year gross product.

A.2 Capital Stock Data

Capital stock estimates for one-digit sectors are from ABS Cat. No. 5221.0. Capital stock estimates for the agriculture sector are not available. Capital stock estimates for the eight manufacturing sub-divisions were estimated using the perpetual inventory method (PIM). Quarterly gross fixed capital expenditure (GFCE) data were sourced from ABS Cat. No. 5626.0. Deflators for GFCE equipment and GFCE non-dwelling construction were sourced from Table 23 of ABS Cat. No. 5206.0. The mean asset life for non-dwelling construction was assumed to be 39 years. The mean asset life for equipment was assumed to be 20 years in the 1950s, declining by 5 per cent for each successive decade. Assets of the same class and vintage were assumed to exit simultaneously when they reached their mean asset life. Straight-line depreciation was assumed. Starting point estimates of the stock of non-dwelling constructions and equipment in June 1959 were sourced from Table 63 of the CBCS Secondary Industry Bulletin for 1958/59 (50 per cent of the value of land and buildings was assumed to be non-dwelling construction). The figures used in Table 3 are estimates for 1994.

A.3 Trade Concentration Data

The ratios listed in Table 3 are derived from the input-output tables and are estimated as the ratio of exports and imports to industry nominal value added. The input-output table for 1990/91 was used.