Summary of Comments on Draft Manual
Canberra Group on Capital Stock Statistics
Washington DC, 8 to 10 November 1999
(Prepared by Derek Blades, 15 October 1999)

This note summarises the comments that have been received on the draft Manual up to 15 October 1999. Part 1 summarises general points that are not specific to particular chapters; Part 2 summarises points relating to chapters of the draft Manual.
Note that comments received on two topics are not summarised below. These relate to:
  1. Capital services. Comments received on this topic refer to the original Chapter 7; Capital Services, that was circulated as the last chapter of the draft manual. This chapter has been radically revised to meet the criticisms contained in the comments from Jack Triplett, Australian Bureau of Statistics (ABS) and Barbara Fraumeni. The revised version of this chapter was posted on the EDG on 15 October.
  2. Obsolescence. Peter Hill, Barbara Fraumeni and ABS have suggested that a more extended discussion of obsolescence is required. There are separate discussion papers on this topic (see Doc. No. 3 and Doc. No. 4).

A selection of comments from the Canberra electronic discussion group relating to the Draft Manual are available below. Please click on the relevant name to download a PDF file containing individual comments from the eight participants:
- Gwennaelle Brilhault (INSEE, France)
- Barbara Fraumeni (BEA)
- Peter Hill (Consultant, Australian Bureau of Statistics)
- Dagmar Hanzlova (Czech Republic)
- Peter Harper (Australian Bureau of Statistics)
- Lianne Ritter (Germany)
- Paul Schreyer(OECD)
- Jack Triplett (Brookings Institute)
The summary and selection of comments does not include the many other helpful comments received on points of detail from, in particular, France, Germany, Australia, Gerhard Meiner, Peter Hill, Anne Harrison and Paul Schreyer.

Part 1. General points

Structure of the report

Peter Hill

- The last section of Chapter 2 and Chapter 7 should be combined into a new Chapter 2 which will provide a theoretical framework explaining the links between the different stock and flow measures. The rest of the material in Chapter 2 will follow this new chapter.

Jack Triplett

- Add an Overview chapter. This will be an overarching theoretical section to set out the rudiments of capital theory.

Australia

- There should be a new Chapter 2 presenting the basic theoretical framework.

Czech Republic

- Separate chapters for concepts and uses.

Germany

- The present structure is appropriate for national accountants and it should not be changed.
- Add a chapter 8 on uses of capital stock and related statistics.

New topics to be included

Australia

- Manual needs a discussion of "second-best" solutions; e.g. depreciation methods that are not consistent with assumptions about efficiency schedules and use of gross capital stock for productivity analysis.
- Manual should include detailed descriptions of procedures used by countries to estimate capital stock, consumption of fixed capital and capital services. Examples should be taken from "advanced" and "less advanced" countries.
- Chain volume indices should be discussed and recommended.

Czech Republic

- Measurement of new types of assets
- Formula for the calculation of accumulated depreciation.

France

- Country practices in measuring stocks of dwellings and infrastructure assets of general government.
- Methods used by different countries to measure capital stocks on a quarterly basis.
- Compilation and use of capital stocks at a micro-level

Germany

- Stocks and flows of intangible fixed assets.
- Costs of ownership transfer.
- Unforeseen obsolescence.
- High inflation.

Terminology

Jack Triplett

- The manual should use terms in common use and not use only SNA terms.
- Depreciation should be used instead of consumption of fixed capital.
- Replacement" prices could be misleading.

Australia

- Favour use of SNA terms but other terms can be used if strong opposition to them. Consumption of fixed capital may be a case in point.

Czech Republic

- Consumption of fixed capital should be retained. Depreciation is a firm-level concept and is measured differently.

Peter Hill

- Problem terms include: rental value and user cost; discards, disposals, retirements, scrapping. Replacement should be dropped.
- Depreciation should be used in the manual instead of consumption of fixed capital.

Germany

- SNA terminology should be used for all aggregates that are included in the SNA.

Glossary

Peter Hill

- A glossary of basic concepts and terms is needed. This will cover many terms not presently included in the SNA.

Australia

- Both SNA terms and terms for aggregates and concepts not found in the SNA should be included. The glossary should not be divided into separate parts.
- Capital services need clear definition.

Czech Republic

- SNA terms and other terms in common use are needed in the glossary.
- Define second-hand asset.

Barbara Fraumeni

- Glossary needs both SNA and other commonly used terms.
- Concept of the productive capital stock (for homogeneous asset groups) is a useful educational tool.

Germany

- The glossary should show the non-SNA terms separately from SNA terms.
- Synonyms for SNA terms should be identified.

Jack Triplett

- Glossary needs both SNA and other commonly-used terms

References/bibliography

Australia

- Extensive bibliography needed. All material on capital stocks and related material should be brought together in one place.

France

- Bibliography should include papers from all three Canberra Group conferences.

Jack Triplett

- Manual needs a comprehensive bibliography of recommended readings on capital measurement

Barbara Fraumeni

- Extensive bibliography needed.

Peter Hill

- Extended bibliography required including articles from academic journals

Part 2. Comments on Chapters

Introduction

Peter Hill

- Should be more explicit about the aims of the Manual. Should capture the flavour of the challenge which measuring capital presents.

Chapter 1. Coverage and classification of capital stocks

France

- Even though R&D is excluded from GFCF, the Manual should discuss measurement of R&D stocks because this is being done in several countries.
- More discussion needed on rented assets. Does attribution of assets to owners distort/underestimate capital stocks?

Jack Triplett

- Discussion of classifications should not be confined to those in the SNA.

Chapter 2.Capital stock and related flows: concepts and uses

France

- More discussion needed on methods of deflation, including chaining

Jack Triplett

- Discussion needed on alternative production account for productivity analysis
- The productive capital stock should be included. The parallel between deterioration and depreciation should be introduced.

Germany

- More explanation needed on uses and calculation of capital-output ratios.
- More explanation needed on the valuation of old assets that are no longer being produced. Seen as a special problem for transition countries.
- Border-line between current repairs and those which count as GFCF.

Chapter 4. Perpetual inventory method

Australia

- Stocks at current year prices should be valued at year-end prices but stocks at constant prices are valued at average prices of the base year. The Manual is wrong in requiring average year prices to be used for both current and constant price estimates.

Germany

- There should be tables showing the service lives used by countries. Average lives from such tables are useful as a crosscheck for country compilers.
- Discussion of retirement distributions should not be restricted to Winfrey and Weibul. Better to give a summary of the distributions used by a number of countries. (including gamma)

France

- Too much discussion of Winfrey. Other retirement distributions should be mentioned. (including log-normal)
- Gross fixed capital formation needs to be defined. Is it "gross gross" or net of discards?
- Should the manual recommend standard service lives, at least for countries with similar economic structures?

Chapter 5.Measuring consumption of fixed capital

Barbara Fraumeni

- More discussion of obsolescence is required. Is it or is it not a part of depreciation?
- The efficiency pattern of one asset cannot be generalised to a group of assets. The way in which an individual asset loses its efficiency over time is different from the way a group of those same assets lose their efficiency

Peter Hill

- Obsolescence should be reflected in the valuation of assets in the gross capital stock. The depreciation factors applied to the gross capital stock should measure only the effects of aging (wear and tear and accidental damage).

Germany

- To improve international comparability, the manual should propose a single depreciation method. Straight-line preferred for this purpose.