Organisation for Economic Co-operation and Development Organisation de Coopération et de Développement Economiques

Transcription

1 OECD OCDE Organisation for Economic Co-operation and Development Organisation de Coopération et de Développement Economiques STATISTICS DIRECTORATE National Accounts Division SECOND MEETING OF THE CANBERRA GROUP ON CAPITAL STOCK STATISTICS Château de la Muette, Paris 29 September - 1 October 1998 Beginning at 10 a.m. on the first day Agenda item : 10a Document number : 21 Title : Determining Asset Lives by Direct Survey Author(s) : Jeff Cope - Statistics New Zealand, National Accounts Division September 1998

2

3 Determining Asset Lives by Direct Survey Preliminary Results from the 1992 Inland Revenue Department Depreciation Survey Introduction 1. Statistics New Zealand (SNZ) is currently developing estimates of capital stock based on a perpetual inventory model. The capital stock series are being developed at the finest level of detail (industry and asset type) that the available data permits. Critical to the estimation method is information on the average asset lives of different asset types used within different industry groups. 2. One possible source of this information is a depreciation survey conducted by SNZ on behalf of the Inland Revenue Department (IRD). This paper briefly summarises the design of the survey and presents the results, comparing them with (a) the official taxation depreciation rates that were subsequently issued by the IRD, and (b) a range of asset depreciation rates currently used in a number of OECD member countries. 3. The depreciation survey also collected information on the purchase price and scrap / resale values at the end of the asset s useful life. This information is still being analysed and is not discussed here. 4. No decisions have yet been taken on the asset lives to be used in the Perpetual Inventory Model (PIM). The IRD Depreciation Survey 5. The survey was conducted by SNZ on behalf of IRD in May The survey was expected to provide some vital information for a comprehensive tax review, including a new depreciation regime to be introduced by the IRD in the following year. In New Zealand, businesses must use approved IRD depreciation rates for specific assets within their industry group and one of the aims of the review was to introduce depreciation rates that better reflected true economic asset lives, free of investment and other policy distortions. The objective of the survey was to provide estimates of the mean depreciation rates, useful life, and residual value for a specified list of assets held by a selected group of industries. It was expected that the depreciation rates to be derived from the survey would be close to the true economic life of the assets. 6. Methodology. a) The population was defined as all New Zealand tax paying businesses. The population was identified by using the SNZ Business Directory which is a register of all businesses in New Zealand. As the survey was focused on tax payers the following institutions were excluded from the survey. Private non-profit organisations Central Government Local Government (excluding trading subsidiaries) b) A list of 250 asset types was defined. These were classified using the Standard Industrial Trade Classification (SITC). For each asset type a target group of businesses was chosen using the New Zealand Standard Industrial Classification (NZSIC). The sample was randomly selected within the 1

4 target NZSIC group. Data was collected through a structured questionnaire posted to the identified enterprises in the sample. c) There were 10,000 enterprises in the sample and all identified enterprises were asked to report on a specified asset type. For larger and more complex enterprises, different parts of the enterprise (equivalent to establishments) were further selected with the result that the final post out was 12,302 questionnaires. A target response population of 3700 was set (i.e. the aim was to get 3700 businesses replying with a positive response and information on the specified asset.) This was achieved. d) Questionnaire: The following were the key questions asked for the asset specified: Year of purchase Purchase price (capitalised cost if manufactured in-house) If purchased second hand number of years owned by the previous owner Estimated usage period in the current location (number of years) Estimated disposal price (when sold for further usage) Post disposal usage (estimated number of years use the purchaser would obtain) Estimated scrap value (if sold for scrap) Estimated total useful life (from new to the end of its useful life) 7. Survey results: The IRD was provided with summarised data for each of the 250 asset types. This included: Number of questionnaires posted for each asset category. Number responded. Mean Useful life (average of all ). Mean useful life when bought new. Mean depreciation rate. The depreciation rate was calculated as: Purchase price - residual value = annual depreciable amount Useful life(years) Rate of Depreciation = Annually depreciable amount *100 Purchase price 8. The survey results were just one information source that the IRD used when revising the depreciation regime. Other sources such as information from registered valuers and from interviews with key businesses were also used, in order to arrive at these schedules. When comparing results from the survey with the final revised IRD depreciation schedules, one notes that the revised depreciation rates are higher than those suggested by the survey (see Table 3 below). Also, the official diminishing value rates are based on a residual value equal to 15% of the purchase price. To what extent this rules out using the IRD depreciation rates in the PIM is still being discussed. Using the Survey Results for the PIM 9. The main objectives of further analysing the survey database are: a) to derive average asset lives at a more aggregated grouping of the asset types, more in line with other data on capital expenditure and capital price indexes. b) to sort the asset groups by industry of use to examine whether lives differed across industries. c) to determine whether a crude scrap value : purchase price ratio could be obtained, for each asset type, which would influence the depreciation rate ultimately chosen. 2

5 10. The tables below provide information on the mean asset lives by industry group for eight categories of plant, machinery and equipment (PME) and five categories of transport equipment (TE). Table 1 - Mean asset life of PME by Industry group Industry group M. Gen PME Metal working power generating data processing telecom and sound Electrical Equipment Furn & fitting Prof apparatus AGRICULTURE FISHING & HUNTING FORESTRY & LOG MINING & QUARRYING (1) FOOD, BEV. & TOBACCO (1) TEXTILES, APP. & F.WEAR WOOD PRODUCTS (1) PAPER, PRINTING AND (1) (1) PUBLISHING CHEMICALS ETC (1) NON METALLIC MINERALS BASIC METALS (1) FABR METALS, (1) 17 6 (1) OTHER MANUFACTURING (1) (1) ELECTRICITY, GAS ETC BUILD & CONSTRUCTION (1) DISTRIBUTION (1) TRANSPORT (1) (1) COMMUNICATION FINANCE (1) (1) COMM Y & SOC. SERV OTHER (2) (1) TOTAL Total Obs Table 2 - Mean asset life of TE by Industry group Buses Small passenger Goods vehicles Ships Aircraft Total Industry group Mea obs Mean obs Mean obs Mean obs Mean obs Mean obs n life life life life life life AGRICULTURE FISHING & Hunting FOOD, BEVERAGES AND TOBACCO WOOD PRODUCTS PAPER, PRINTING AND PUBL NON METALLIC MINERALS FABR METALS, MACHINERY ETC BUILD & CONSTRUCTION DISTRIBUTION TRANSPORT COMMUNICATION FINANCE COMMUNITY AND SOCIAL SERVICE OTHER (2) TOTAL

6 11. A comparison of the survey asset lives with the final IRD lives included in the revised depreciation regime is given below in Table 3. The table also provides a few examples of assets specified by the IRD under each industry and their lives. Although it is possible to extract PME asset life information by industry from tax data, it is not possible for TE, as these are the same across all industries. Table 3: Comparison of PME asset lives (derived from the survey) with the IRD tax lives INDUSTRY SURVEY IRD Some specific asset lives from IRD AGRICULTURE Harvester 15.5, Tractors 15.5, Milking Machine 8. FISHING & HUNTING Vessels non steel 20/steel 15.5, Refrigerators 10, FORESTRY & LOG Bandsaws 20, Bull chains 20, Log carriages 20. MINING & QUARRYING Bulldozers 10, Draglines 10, Cutting machines 10 FOOD, BEV. AND TOBACCO Butter system 15.5, Conveyers 15.5, Packer 15.5 TEXTILES, APPAREL, F.TWEAR Sewing machine 15.5, Zip maker 12.5, Bleaching 20 WOOD PRODUCTS Jointing machines 15.5, Drilling 15.5, Sawing 15.5 PAPER, PRINTING & PUBSH Printer(news paper) 20, Screen printer 12.5, cutter 20 CHEMICALS ETC Mixers 15.5, Fertiliser plant 12.5, Feeders 10 NON MET MINERALS 15 N/A N/A BASIC METALS Furnace 20, Die casting 15.5, Rolling mill 25 FABR METALS Forging press 20, Maulders 20, Fettling Machines 15.5 OTHER MANUFACTURING 8 N/A N/A ELECTRICITY, GAS ETC Turbines 20, Poles 25, Generator 20 BUILD & CONSTRUCTION Concrete mixer 12.5, Brush cutter 5, Diggers 12.5 DISTRIBUTION 12 N/A N/A TRANSPORT Radar navigator 10, Containers 20, Ferries 20 COMMUNICATION Broadcasting masts 25, Radio equipment 10 FINANCE 11 N/A N/A COMMU Y & SOCIAL SERV 11 N/A N/A OTHER (1) 12 N/A N/A 12. Some observations can be made: Derived PME lives are very similar for the manufacturing industries, but not for a number of the other industries, such as Agriculture, Fishing & Hunting, Electricity Gas & Water and Building & Construction. This may be due to a more varied asset mix in the two sources for these industries. Tax lives are generally lower than the derived lives (some times marginally). The IRD depreciation rates assume a standard 15% scrap / resale value. The IRD appears to have largely standardised asset lives across industries, i.e. asset lives do not differ significantly from industry to industry. There are only four different PME lives (i.e. 6.5, 10, 15.5 and 20) applied to the different industries. However, most industries have an asset life of 15.5 for their PMEs. In Electricity, Gas and Water asset lives are quite different. However, the survey results are similar to those used by other countries for this industry (refer to table 4 below). 4

7 The IRD has revised the depreciation schedules ( which includes asset lives revisions) several times since the 1994 revision. These revisions usually resulted in an increase in the length of the tax life. 13. Comparing the survey asset lives with those currently used by a selection of OECD countries. These are shown in Table 4. Table 4: Comparison of derived PME lives with those for selected countries INDUSTRY SURVEY IRD New Zealand AUSTRALIA UK CANADA AGRICULTURE FISHING & HUNTING FORESTRY & LOGGING N/A 10 MINING & QUARRYING /15 20 FOOD BEV. & TOBACCO /15 TEXTILE APP: & F.WEAR /21/15 WOOD PRODUCTS PAPER PRINT. & PUBLISH: /30 CHEMICALS NON METALLIC MIN. 15 N/A BASIC METAL FABRICATED METAL OTHER MANUFACTURING 8 N/A ELECTRICITY GAS WATER /20/27 35 BUILDING & CONSTR DISTRIBUTION 12 N/A TRANSPORTATION N/A 25/25/8 28/10 COMMUNICATION FINANCE 11 N/A COMM Y & SOC SERV 11 N/A N/A N/A N/A ARITHMETIC AVERAGE Source; Methods used by OECD Countries To Measure Stocks of Fixed Capital 14. Points to note are: Asset lives used by Australia do not differ significantly from industry to industry. However, they do not vary significantly from the asset lives derived from the survey in relation to many industries. Significantly different values are observed in Agriculture and Electricity, gas & water industries. The U.K has varying lives for PME in different industries but they are significantly different from the lives estimated from the NZ IRD survey. PME in most industries in UK have a very long life time when compared to New Zealand. It is also observed that the UK PME lives are generally higher than that of all other countries considered in this comparison. Estimates of PME lives by industry in Canada are also higher than the values used in Australia and the values derived from the survey. 5

8 15. Average Useful for Computers. The survey results for computers show what might be expected, viz. that the useful lives will not vary across industries. The tax life assigned to computers is much lower than indicated by the survey (tax life = 4 years, survey = 8 years). The derived value was similar to the values used in capital stock estimations of some other countries (refer Table 4) Table 4: Average Computer Lives for Selected Countries NEW ZEALAND AUSTRALIA GERMANY UNITED JAPAN SURVEY IRD STATES 8 4 N/A Shortcomings of the data 16. The survey results are still being analysed and compared with other sources. Possible shortcomings include: The number of observations within some industry groups is too small to draw firm conclusions on asset lives. There is considerable - and unexplained - variation in respondents estimates of key variables such as expected life and residual value. Similar asset types with similar purchasing price and similar capacity were observed to have been estimated with largely varying life by respondents in the same industry category. For example, milking machines claimed to have been purchased brand new were observed to have an estimated life of between 5 to 30 years depending on the respondent. There may be sound explanations for this, such as quite different models, but this requires further investigation. There is an element of guesswork in the respondents asset life answers where the assets have been purchased second hand or are sold for further use. In these cases, the use of the assets is outside the respondents direct experience and they are asked to estimate for the full life of the asset. 6