Joint
OECD-ESCAP meeting on national accounts:
the 1993 SNA five years on
The Joint OECD-ESCAP meeting, the first
ever between the two organizations in the field
of statistics, was held in Bangkok, 4-8 May
1998. It saw participation by as many as 24
countries/areas. It was attended by Australia;
Bangladesh; Brunei Darussalam; China; Denmark;
Fiji; Hong Kong, China; India; Italy; Japan;
Macau; Malaysia; Nepal; the Netherlands; New
Zealand; Philippines; Republic of Korea; Russian
Federation; Singapore; Thailand; Turkey; United
Kingdom; United States of America, and Viet
Nam. In addition representatives from
the following organizations attended: Economic
and Social Commission for Western Asia (ESCWA);
Food and Agriculture Organization of the United
Nations (FAO); International Labour Organization
(ILO); Asian Development Bank (ADB); and the
Statistical Institute For Asia and the Pacific
(SIAP).
Mr. Mooy, Executive Secretary of ESCAP, opened
the meeting by thanking the various governments
for sending their staff, the Asian Development
Bank (ADB) for having supported the attendance
of some participants, and the Organization for
Economic Cooperation and Development (OECD)
for its technical input. He commented that five
years after its release the 1993 SNA should
be critically reviewed to evaluate practical
difficulties in its implementation.
Mr. Derek Blades, Head of the National Accounts
Division in the OECD's Statistics Directorate,
explained the OECD's role and its extensive
links with various United Nations organizations,
especially the Economic Commission for Europe
(ECE). Mr. Blades thanked all those who contributed
papers and encouraged all participants to join
in the discussion.
After the adoption of the agenda the meeting
elected Mr. J. Steven Landfeld (USA) as Chairperson
and Ms Estrella V. Domingo (Philippines) as
Vice-Chairperson.
The following substantive topics were discussed
at the meeting on the basis of papers prepared
by several participants:
The recommendations of
the 1993 SNA on FISIM were discussed. The
concept of bank output includes the
value of services implicitly charged via the
interest rates granted to depositors and those
charged to borrowers. The approaches of estimating
this imputed output in the 1953 and 1968 versions
of SNA were briefly outlined showing the similarities
and differences. The 1953 SNA allocates the
output to both producers and final users while
the 1968 SNA attributes the use only to producers
as intermediate input. The latter reduced
GDP by the amount that would have been paid
for by the final demand users.
The Australian Bureau
of Statistics (ABS) has developed a new approach
of calculating FISIM, in consultation with
Eurostat (which has the responsibility for
research on FISIM) through a bottom-up
approach which also allocates the implicit
output to the users. The estimation procedure
is now being adopted by Eurostat and countries
of the European Union are experimenting with
putting FISIM into practice.
Many countries presented
their studies on estimation of FISIM with
supporting examples. The divergence in opinion
on the merits/ demerits of the different methodologies
points to the need for further studies as
the estimation could be determined by specific
country considerations.
Institutional
sectoring
The discussion dealt with
a wide range of issues in classifying institutions
into market or non market producers, quasi-corporations
or household unincorporated enterprises,
non-profit institutions or non-financial corporations,
government or non-profit institutions
serving households, government or non-financial
corporations, social security or pension funds;
rest of the world or households (in
the case of cross border workers); and rest
of the world or non-financial corporations
(in the case of mobile equipment, construction
abroad, etc.).
Several suggestions were
offered in resolving the issues, although
some which were country specific would best
be solved within a conceptual basis consistent
with the particular circumstances of the countries.
Informal/hidden
activities
The activities of the
Expert Group on Informal Sector Statistics
(Delhi Group) were reported in the meeting.
The operationalization of the definition of
informal sector is being discussed to cover
both social and economic dimensions. Of particular
interest to the group are the characteristics
of informal sector participants (employees,
etc), coverage of the sector in national accounts
compilation, and the most effective approaches
of collecting data. Subgroups have been
set up among the members of the Delhi group
to pursue specific areas of the informal sector:
operationalizing the definition, developing
strategies to address sampling frame and weighting
issues, quality assurance methods, and appropriate
methodology to produce accurate measures of
value added.
Countries presented their
experiences in setting up an operational definition
of the informal sector, measuring its activities
and estimating value added. The estimation
of informal sector value added is based on
labour input and is premised on the assumption
that characteristics of labour input in the
recorded sector can be imputed to the labour
force in the unrecorded sector. Inclusion
of agriculture in the informal sector and
how to handle illegal activities were also
raised. Because of the mixed nature
of the informal sector, households and business
surveys are valuable inquiries for collecting
data on informal activities.
New
kinds of capital assets
Valuation of film originals
and software were extensively discussed in
the meeting. For film originals, value of
production or net present value are possible
alternatives for estimation. The former is
simple and direct but accounting practices
might not truly reflect the actual cost of
production; the latter is a preferred method
although it is more demanding in terms
of data and would entail problems of
revisions.
For software, valuation
should be consistent with the two-stage production
nature of software: production of originals
and production of copies for sale. In the
1993 SNA, software is capitalized and some
concern was raised on whether it should be
classified under artistic originals or investment
in computer software. Acquisition of software
as investment could be measured either from
the supply side (estimate by commodity flow)
or demand side (purchaser). Recording the
investment on the demand side is preferred
but business practices often treat this
as business expense rather than capital expenditure.
A corollary concern is the use of prices of
computer software. Because of the qualitative
differences of the software, pricing would
be quite complicated. USA and Germany are
developing price indexes (hedonic approach)
for software which would be available sometime
in 1999.
A discussion on the "year
2000 problem" debated whether the costs of
solving the problem should be treated as intermediate
consumption or as capital formation. The conclusion
was that, if the opportunity is taken to redevelop
the associated systems significantly, then
it would definitely be investment. However,
opinions were divided on the treatment of
less extensive re-engineering efforts to fix
the year 2000 problem. Data problems
are likely to be severe because of the differing
accounting approaches adopted by businesses
and whether or not they are able to write
off the costs as current costs for tax purposes.
Functional
classifications
This item had been put
on the agenda mainly for information. Participants
were informed about efforts of revision on
three of the four functional classifications
contained in the 1993 SNA. Proposed revisions
made by OECD, Eurostat and UNSD (United Nations
Statistics Division) had previously been circulated
widely.
Chain
versus fixed-weight indices
Many methodological issues
regarding estimation of GDP using constant
prices were presented. A distinction was made
between base year (basis of weights) and reference
year (index=100). The "fixed-base" constant
price, the standard practice of most countries,
often caused distortion in measuring growth
rates as relative prices and the underlying
weights change over time. Recent experiences
showed that computer prices have been a major
factor in producing large revisions to growth
rates in the revision of base year.
Chain volume series provides
a solution to the revision problem and is
more appropriate when relative prices are
changing fast. By changing the base year every
year, the weights are adjusted accordingly
and therefore are more reflective of the real
situation and more accurate in measuring volume
changes. However, the elements of additivity
are lost in the process. In view of
this the use of Fisher's index was viewed
as the best approach although its data implications
are beyond the capability of many countries
of the region. In this case, adopting
a more recent base year would constitute a
substantial improvement for countries in the
region with an out-of-date base year.
Participants agreed that
gaining the support of users before introducing
chain indices is essential. This can
be done by showing users the differences in
measured growth rates using chain as opposed
to fixed-weight indices.
The meeting noted that
sometimes the adoption of chain indices could
overcome measurement problems specific to
some sectors like investment in computers.
In the USA the growth of investment in computers
led to particularly large substitution bias
in the fixed-base indices used for the US
National Income and Product Accounts; the
BEA (Bureau of Economic Analysis) initiated
a debate on the relative merits of chain versus
fixed-weight indices in the late 1980s which
culminated in a major users' conference in
1995.
Real
output of non-market services
Participants discussed
their experiences in estimating the real output
of non-market services with particular attention
to:
education
health services
public administration
justice
social security
Several new approaches
were presented to improve estimates
in respect of the traditional approach
based on the use of the deflated values of
the cost components.
In the discussion, a
distinction was drawn between outputs and
outcomes. It is the latter, e.g., health
of the nation or national education level,
which is of principal interest in many countries.
Usually, national accountants take the view
that their job is only to measure outputs
although in some countries there is increasing
pressure from the academic world to introduce
outcome measures.
Other
problems in implementing the 1993 SNA in the
ESCAP region
Participants informed
the meeting about their current practices
and problems encountered in implementing the
1993 SNA, addressing many substantive topics
covered by the meeting. Reference should be
made to the papers of the countries, available
at the OECD's Web site at the following URL:
http://www.oecd.org/std/escap98/index.htm.
User
education
The meeting was informed
about programmes that had introduced the 1993
SNA to the users. The process of information/consultation
took place in different forms, mainly through
seminars, publications, and regular meetings.
Particular users (econometricians) needed
more sophisticated answers on how to adapt
to the changes introduced by the 1993 SNA
(e.g., adapting forecasting models developed
using fixed-base indices to the new volume
measures based on chain indices).
Details of the discussions and the conclusions
that the participants reached at the meeting
on each of the above topics would take too much
space here. The full report of the meeting can
be found at theOECD's
Web site.
Technical
workshop on the Year 2000 (Y2K) problem in computers
and strategic issues for national statistical
offices.
Bangkok 18-19 June 1998 The Statistical
Institute for Asia and Pacific (SIAP) and ESCAP
organized a technical workshop on the Year 2000
problem. The workshop was held in Bangkok during
18-19 June, and saw participation of several
countries from the region, represented mainly
by their national statistical offices with additional
participants from other government offices.
In an attempt to assist countries to accelerate
Y2K problem remediation, the Workshop prepared
a"Y2K action list" (below).
Y2K action list
Stop waiting for somebody
to come to study and fix Y2K problems. That
is not going to happen. Accept ownership of
systems and full responsibility for them.
Lack of technological expertise is not an
acceptable excuse for inaction.
Appoint immediately a
full-time Y2K coordinator, with managerial
skills and necessary authority to initiate
actions and delegate responsibilities.
Backup data and systems
securely before doing any Y2K compliance testing.
Ensure that the backup
data can be read. Document backups properly.
Use all talents within
the organization to create multi-disciplinary
teams to undertake step-wise rectification.
Include a mix of management staff, IT staff,
and substantive experts. Locate them together
and relieve them of other responsibilities
to the extent possible.
Ensure that anything
new, developed or purchased, is Y2K compliant.
Use industry standard
steps in achieving Y2K compliance, but simplify
where possible. Information is available on
the Internet, literature, and IT magazines.
Include embedded chips
in the Y2K inventory and seek out compliance
information for them.
Start conversion of mission-critical
systems in priority order.
Remember that, although
necessary, awareness campaigns or lengthy
planning processes will not resolve Y2K problems.
Start practical work immediately.
Do not wait for funding
before starting. A lot of preparatory work
can be done without separate funding, especially
in awareness creation, in preparation of an
inventory of affected items, and in seeking
compliance information from vendors.
Do not rely on general
compliance statements. Be specific when requesting
compliance information from equipment and
software suppliers. Ask when compliant replacements
will become available, and how they will be
installed and operationalized.
Demonstrate the impact
of non-compliant components and systems to
management by writing down what would happen
if each affected system was not available.
Transmit those technology
and business assessments to top management
and to the authority providing the budget.
Enlist the support of
important clients to strengthen Y2K funding
requests.
Concentrate on resolving
only the Y2K problem itself. Do not attempt
to simultaneously improve the functionality
or other features of existing systems.
Slow down or postpone
new IT development to conserve resources.
Remember that Y2K projects
carry pitfalls of typical IT projects, including
overly optimistic scheduling, poor documentation
and incomplete debugging. Projects have a
tendency to drag on longer than initially
planned.
Make a contingency plan
at an early stage.
Include a worst-case
scenario in the contingency plan.
Document all Y2K efforts
from the beginning. That written evidence
may be invaluable later.
Do not let top executives
delegate or outsource their responsibility
(it is not possible). Y2K compliance is a
major business issue, the alternative is often
a closure of operations.
Disseminate these recommendations
to all staff in the organization.
Participants of the
workshop, Bangkok 18-19 June 1998
Here are some comments
from Mr Ilpo Survo, Programme Officer
in charge of the secretariat's Public Sector
Computerization activities.
Q. Mr Survo, the workshop you were organizing
with SIAP on the year 2000 problem saw the participation
of several representatives of Asian and Pacific
NSOs. The Y2K problem is a global one, however
many of the Asia and Pacific countries were
recognized as reacting too late for tackling
it. Did the meeting identify risks that seem
specific to this region?
A. The region is exposed to the Y2K problem
as much as any other region: there is a broad
range of computer hardware, software, and devices
with embedded chips installed -- that is in
terms or age, technological sophistication and
origin. The penetration rate of date dependent
electronic systems is of course lower in developing
countries, but that does not remove the risk,
or rather a certainty, that they are affected.
If I have to mention a specific risk, it would
arise from the fact that business partners abroad
are losing their confidence in the region because
it is reacting so late and will not be prepared
in time. The world is wired and interdependent
and great volumes of data are transferred in
international banking networks, enterprise networks,
on the Internet etc. Even if your own
system is internally fully compliant, it might
still go down or require manual involvement
if incoming data do not comply with the new
millennium's date expression.
Q. We can see an increasing coverage of
the Y2K problem by national and international
newspapers and magazines.How would you
evaluate the quality of the information usually
provided through these media?
A. It is difficult to monitor what the media
is telling the public in all countries.
It seems, however, that it ignored the warnings
by experts until the latter half of last year.
Something that was so far in the future did
not make good news. The situation has
fortunately improved remarkably, at least in
Thailand. The English language press,
TV and radio, which I am able to monitor,
are all having Y2K stories almost on a daily
basis. Nothing wrong with the quality
in general, and the public awareness for which
they are "responsible" has improved a lot.
They do have limitations in terms of how deep
into practical Y2K problem solutions they can
get into. People in charge need
practical advice on how to organize the fixing
of their systems. In that, the Internet
and its Web are invaluable. Using simple
Web searches, say with Y2K + the specific product
name, you may be able to find specific information
about Y2K compliance and how to reach it in
your environment. If that does not work,
use more generic key words and follow the leads
until you find the information.
Q. Attribution of responsibility to the
top executives of public and private organizations
is one of the cornerstones of the awareness
campaign that ESCAP is developing. The temptation
to attribute responsibility to technical/IT
units of enterprises and organizations has been
underlined many times. Could you explain to
us again why the Y2K problem has to be managed
by high level executives?
A. Involvement of top management is necessary
because Y2K failures can put private and public
sector organizations out of business for extended
periods. It is hard to find a more important
concern for executives. Their involvement
is necessary also because the resolution of
the Y2K problem requires resource (re)allocation,
which in turn can postpone other work and strategic
new IT development. Y2K projects are also
challenging to manage as they require cooperation
among various units and people with different
backgrounds.
Q. Technical/IT units will be anyway deeply
involved in resolving the problem, in some organizations
the level of know-how could be sufficient, while
in other ones it might not be. What are the
additional or specialized IT skills that, in
your opinion, would be most useful to tackle
the problem?
A. Beyond project management skills, they need
to be "detectives" and "doctors". The
problem resolution is easier if organizations
have good housekeeping, i.e. detailed records
of existing equipment and software and their
vendors, so that the compliance information
can be requested easily. Detective skills are
required particularly in locating in-house program
code pieces that carry the bug. Doctoring
of the programs must be done carefully so that
compliance is achieved with unchanged functionality.
Plenty of time is needed for testing and debugging.
Q. Another main point of the meeting guidelines
is the advice to handle the problem internally
whithin the concerned organization. Nevertheless
some private enterprises as well as consultancy
firms provide services, at market prices, to
help organizations to solve the problem. Can
all the work be delegated to external resources?
A. We heard in the Workshop that while the
"reactive" NSOs had chosen to rely on internal
expertise for their Y2K problem resolution,
they had contracted some clearly identifiable
pieces of work outside. But a completely
turn-key solution by outsourcing is hardly feasible
as the contractor would need a lot of cooperation
from the organization anyway. The Workshop
was very clear on one issue, namely that top
management cannot escape responsibility by delegating
it to the organization's own IT personnel.
Missions
of ESCAP Statistics Division Staff
Mr Andrew J. Flatt, Director, Statistics
Division, undertook missions as follows:
Rio de Janeiro
(13-15 May): To represent ESCAP and present
a paper at the second session of the Expert
Group on Poverty Statistics (The Rio Group).
Paris (18-20 May):
To represent ESCAP at the 46th session of
the Conference of European Statisticians held
at OECD Headquarters, Paris, and the Eighteenth
Meeting of the Steering Committee on the Coordination
of Technical Assistance in Statistics to the
Countries of the Former Soviet Union.
Manila (25-27 May):
To represent ESCAP and present a paper at
the Concluding Workshop of the Asian Development
Bank's Project (RETA 5555) on Institutional
Strengthening and Collection of Environment
Statistics in Selected Developing Member Countries,
and to coordinate with ADB on possible future
work in this field.
New York (15-18
June): To represent ESCAP at the 32nd session
of the ACC Subcommittee on Statistical Activities,
held in New York from 16 to 18 June 1998,
and to attend related meetings and discussions
with the United Nations Statistics Division.
Mr Loh, Meng Kow, Chief, Statistical
Information and Services Section, undertook
a mission to:
Ankara (25 April-1
May): To participate in the Second Meeting
of the Expert Group on the Informal Sector
Statistics (Delhi Group), held in Ankara,
from 28 to 30 April 1998.
Ms Heidi R. Arboleda, Regional Adviser
on National Accounts, provided advisory services
as follows:
Thailand (18-29
May and 3 June): to the National Economic
and Social Development Board (NESDB) on national
accounts compilation by conducting the training
workshop for national accounts compilers,
data producers and users from 18 to 29 May
1998 in Bangkok. As a follow up, another meeting
was held with the NESDB to identify the priority
activities based on the output of the workshop
in the short and medium term (3 June).
Myanmar (21 June-4
July): To conduct training on the 1993 SNA
and to advise on its implementation.
Mr Nuri M.Ozsever, ESCAP Adviser on
Population Data Processing and Database Management
with the UNFPA/CST in Bangkok, visited the following
countries:
Viet Nam (3-9 May):
To participate as a resource person in the
workshop on a population project organized
by the General Statistical Office (GSO), and
to discuss with GSO projection methodology
and techniques as well as population projection
software packages.
Cambodia
(17-20 May): To participate in the TPR meeting
for the project "Support to the National Population
Census Phase II" CMB/97/PO2.
Ms Luisa T. Engracia, ESCAP Adviser
on Population Statistics with the UNFPA/CST
in Almaty, undertook the following mission:
Maldives (10-26
June): To assist in planning 2000 population
census.
Mr L.H. Lewis, ESCAP Adviser on Population
Statistics with the UNFPA/CST in Suva, undertook
the following missions:
Papua New Guinea
(10-21 May): To review findings of Mile Bay
census pre-test, to take part on the Enga
pre-tests, and to assist in the preparations
for 2000 population and housing censuses.
Suva (22-31 May):
To design model population and housing census
questionnaires for the Pacific.
Canberra (8-12
June): Together with the Australian Bureau
of Statistics, to design and test model
population and housing census questionnaires
for the Pacific island countries, and to assist
in development of census training manual.
Papua New Guinea (15-19
June): To assist with (i) evaluation of pre-test
findings, reference documentation, and outputs
for provincial planning, (ii) the first provincial
census coordinators' workshop, (iii) finalization
of 2000 census information paper, and (iv)
2000 census questionnaire design.
Suva (29 June-1
July): To participate in a UNDP-sponsored
training workshop on IMPS software for Pacific
island countries.
Publications
released by the Statistics Division
Statistical Newsletters
Nos. 108 and 109
Statistical Indicators
for Asia and the Pacific, Volume XXVIII, No.2,
June 1998
Women in Indonesia -
A Country Profile, Statistical Profiles, No.
14, 1998
Women in the Islamic
Republic of Iran - A Country Profile, Statistical
Profiles, No. 15, 1998
Visitors
to the Statistics Division
Ms Rosanita A. Serrano,
Programme Coordinator, United Nations Development
Programme (UNDP), Makati City, Manila.
Mr Dominic Leung Kam-to,
Assistant Commissioner, Census and Statistics
Department, Hong Kong.
Ms Nyunt Nyunt Win,
National Consultant, UNDP, Yangon.
Mr Chee Kim Loy,
Department of Public Policy and Administration,
Universiti Brunei Darussalem, Banda Seri Begawan.
Ms Suvannee Vathanachit,
Assistant Professor, Faculty of Economics,
Thammasart University, Bangkok.
The International Association
of Survey Statisticians: training opportunities
for young statisticians in developing and transition
countries
The International Association of Survey Statisticians
(IASS) was created 25 years ago, as a section
of the International Statistical Institute (ISI).
Its objectives are to promote the study and
development of the theory and practice of statistical
surveys and censuses.
The IASS secretariat is located at Libourne,
France but there is also a network of country
representatives, who report on events and projects
in their own countries and disseminate information
from the secretariat: the IASS is currently
looking for new representatives from countries
not yet covered.
Activities of the IASS vary from the organization
of
international and regional
meetings, seminars and conferences,
short courses, workshops
and training programmes,
theoretical or practical
investigations,
publishing and distributing
periodicals, pamphlets, books etc.
providing services to
its members.
The IASS is sponsoring a programme of short courses
to take place before and after the 52nd session
of the International Statistical Institute, which
will be held in Helsinki, Finland, August 11-18,
1999:
Workshop on Survey
Sampling in Developing Countries; 7-8
August, 1999.
Variance Estimation
in Complex Surveys; 9-10 August, 1999.
Methods for Designing
Business Surveys; 7-8 August, 1999.
Introduction to Small
Area Estimation; 19 August, 1999.
Continuous Quality Improvement
(CQI) in Statistical Agencies; 22-23
August, 1999.
Analysis of Complex
Survey Data; 22-23 August, 1999.
Courses 1, 2 and 3 will be held at Jyväskylä
University, Finland while courses 4, 5 and 6
will be held in Riga, Latvia.
All courses will be presented in English and
participants should have the ability to work
in this language, however for those participants
registering for the courses before May 1999,
the course introductory materials will be translated
into French, the other official language of
the ISI.
Other interesting materials, like the electronic
version of the IASS's journal The Survey Statistician,
are available at the IASS Web site: http://www.cbs.nl/isi/iass.htm
For more information about IASS please contact:
Benoit Riandey
IASS Executive Director
INED
133 Boulevard Davout
75980 PARIS Cedex 20
France
fax: 33 1 56 06 21 92
e-mail: riandey@ined.fr
Contact for membership:
Mrs. Claude Olivier
IASS Secretariat
c/o INSEE - CEFIL
3, Rue de la Cite
33 500 LIBOURNE
France
fax: 33-5 57 55 56 20
e-mail: claude.olivier@dg33-c290.insee.atlas.fr
