| The ninth session of the Working Group of
Statistical Experts, organized by the secretariat
of the United Nations Economic and Social Commission
for Asia and the Pacific (ESCAP), was held at
Bangkok from 30 January to 2 February 1996.
The meeting was attended by representatives
of the following 23 members and associated members
of ESCAP: Australia, Bangladesh, Brunei Darussalam,
China, Fiji, Hong kong, India, Indonesia, Islamic
Republic of Iran, Japan, Kiribati, Lao People's
Democratic Republic, Macau, Malaysia, Maldives,
Nepal, New Zealand, Philipines, Republic of
Korea, Russian Federation, Singapore, Thailand,
and Viet Nam.
Participants of the
9th session of the Working Group of Statistical
Experts The executive secretary of
ESCAP in his opening address welcomed participants
and apologized for the delay in convening the
Working Group, caused by the financial crisis
facing the United Nations.
The Bureau of the ninth session of the Committee
on Statistics officiated at the meeting of the
Working Group. It consisted of Mr Fredrick W.H.
HO (Hong Kong) serving as Chairperson, Mr Timoci
Bainimarama (Fiji) and Mr Ataul Haq (Bangladesh)
as Vice-Chairpersons, and Mr Romulo Virola (Phillipines)
as Rapporteur.
The bureau met prior to the session of the
Working Group on 29 January 1996 and dicussed
the following topics:
- Terms of reference
and functioning of the Bureau, including financial
arrangements
- Terms of reference of
the Committee on Statistics, with special
reference to its designation as the focus
of regional statistical development
- Mechanisms for mutual
cooperation in statistical development among
member countries
- Programme matters
- Tentative provisional
agenda for the tenth session of the Committee
on Statistics.
The Working Group agreed with the Bureau recommendation
that in any future revision of the conference
structure of ESCAP, the Committee on Statistics
should be retained.
The Working Group was advised of the high rate
of implementation achieved by the secretariat
in its 1994-1995 statistics subprogramme. It
also acknowledged the intensive efforts of the
secretariat in collecting and compiling information
on the regional statistical activities of other
international organizations and agencies, in
the context of the experimental presentation
of statistical work programmes in the region.
In discussing implementation of the 1993 System
of National Accounts(SNA), the Working Group
endorsed the proposal of the secretariat to
organize subregional workshops focusing on the
procedures for switching over to the updated
concepts and classifications of the 1993 SNA.
The topics covered would be based on the stage
of national accounts development amongst the
member countries in the group. The Working Group
considered that given the importance of the
development of national accounts statistics
and the emphasis given to implementation of
the 1993 SNA, ESCAP should accord priority to
the delivery of advisory and training sevices
and other technical assistance in that field.
The Working Group also considered the topic
of statistics on trade in services based on
the report of seminar.
The Working Group dicussed the use of information
technology(IT) in statistics focussing on geographic
information systems(GIS), remote sensing and
the internet. It noted that the region's statistical
offices were clearly at different stages of
IT development. Some advanced national statistical
offices (NSOs) had a personal computer on almost
every employee's desk, whereas in others they
were shared by several workers.
Regarding statistical database development
in general, more advanced countries were already
redesigning their old applications and converting
from mainframe systems to client-server environments.
In that conversion hardware was a relatively
small problem compared to the challenges of
designing an integerated system that catered
to all needs in data collection, processing,
analysis and dissemination, and that was capable
of reaching customers conveniently at their
sites.
Countries that had used geographic information
systems in statistics reported a high demand
for location-specific products and services.
The Working Group noted the experience of several
NSOs on the benefits derived fron GIS techology
when planning censuses and surveys, with a view
to presenting their results in a more user-friendly
format.
During the past few years, local area networks(LANs)
had been introduced in many statistical offices.
In the more advanced countries, goverments had
developed wide area networks (WANs) connecting
various goverment departments including statistical
agencies. The Internet, which has experinced
a phenomenol growth during the past couple of
years, had opened new opportunities in all areas
of statistics from data collection to data dissemination
and marketing. E-mail was used in several offices,
although only four NSOs had so far gone on-line
to use the full internet connection and establish
World Wide Web (WWW) pages on the Internet,
several others were about to create their home
pages soon.
The Working Group heard that the recent technological
deveopments in micro computing, networking and
telecommunications had offered great opportunities
for the statistical offices to improve dissemination
of their statistics. The enormous potential
of the Internet had added a new dimension to
statistical data dissemination. Some countries
had already initiated the use of the internet
for statistical dissemination on a limited scale
whereas others were still exploring the possibility
of using it in the future. The Working Group
therefore felt that every country should adopt
a suitable strategy of data dissemination according
to their infra structural development and the
requirements of the users.
The Working Group stressed that without effective
coordination, it would be difficult for national
statistical agencies to achieve a number of
their important objectives, many of which were
listed in the secretariat document. Since coordination
of statistical activities was particularly important
for avoiding duplication and dissemination of
conflicting figures, and for reducing the reporting
burden, explicit mention of those objectives
should also be made in the list. Some countries
described steps that had been introduced to
reduce the reporting burden through comprehensive
coordination, such as simplification of the
process of clearance for statistical activities
of other agencies and introduction of simple
and easy reporting methods. Improvement of international
comparability of data and promotion of technical
cooperation in statistics were also considered
important.
In reviewing the work programme of the secretariat,
the Working Group recommended that the work
on statistics on gender issues in the region
be given a higher priority in the statistical
work programme of ESCAP. Another area requiring
immediate attention was the estimation of poverty.
It was agreed that there should also be significant
commitment in the work programme to address
the statistical issues arising from the recent
United Nations summit meetings in Cairo (the
International Conference on Population and Development,
1994), Copenhagen (the World Summit for Social
Development, 1995) and Beijing (the Fourth World
Conference on Women, 1995). |
| Introduction
1. Poverty is a commonly known concept often
identified with such human deprivation as starvation,
malnutrition and homelessness. While it may
not be so difficult to comprehend the notion
of extreme poverty, the conceptualization of
poverty for the purposes of production and analysis
of needed quantitative information may not be
as straightforward. The systematic prevention
and alleviation of poverty require reliable
and timely data not only to estimate the number
of poor people, but also to identify target
groups, construct profiles of the poor, understand
the causes and consequences of poverty, design
policies and programmes and monitor the progress
made. The tasks of data collection and analysis
are the more challenging when one is dealing
with the entire spectrum of poverty rather than
just extreme poverty.
Concepts and Measures
of Poverty
2. A quick review of the literature on poverty
measurement2
suggests that the concepts and measures which
were explicitly or implicitly introduced by
various western social reformers during the
last century, in particular Le Play and Charles
Booth, and those by Rowntree3
at the turn of the century, still dominate the
measurement of poverty in developing countries.
The most common measure of the incidence of
poverty evolving from those studies is the so-called
'headcount index'. It is the simple ratio of
the number of the poor (headcount) to the total
number of people in a community, where the poor
are defined as those whose consumption (or income)
standards fall short of certain 'norms'. The
consumption or income norms, specified on a
certain basis, are known as the 'poverty line'.
In the discussions and analysis of poverty,
the index is often accompanied by the absolute
number of poor.
3. Another measure which utilizes the concept
of poverty line is known as the 'poverty gap',
which is the aggregate shortfall of income of
all the poor from the specified poverty line.
The shortfall of the average income of the poor
from the poverty line expressed as a percentage
of the latter is called the 'income-gap ratio'
(Sen, 1982). Both the poverty gap and headcount
measures are insensitive to the extent of differences
among the poor and thus do not provide any information
on the severity of poverty. The headcount measure
provides information on the number of poor,
which is not even available from the poverty
gap, but it does not take into account whether
someone is just below the poverty line or at
the other extreme, in acute misery. The use
of the poverty gap measure in developing countries
may not be feasible in most cases due to practical
problems of collecting income data and their
poor quality.
4. Since the headcount measure is widely used
in the ESCAP region, it would be useful to discuss
its merits and limitations. The poverty line
used is mostly based on biological considerations,
i.e., nutritional or subsistence requirements.
It has been argued that it would be virtually
impossible to arrive at an objective definition
of a poverty line. For example, scientists are
by no means in agreement on minimum food requirements,
which also vary according to age, sex, height,
weight, occupation, activity and climatic conditions
of the place of residence. Thus the common approach
of using one or two minimum nutritional standards
would not suffice. Also, for sustenance, an
individual needs in addition to calories a certain
amount of other nutrients such as proteins,
vitamins, calcium and iron. Even where the same
nutritional standard is used over time, distortions
are introduced due to variations in costs of
commodities across localities and vulnerable
groups of people, and uneven distribution of
inflation across commodities. Moreover, there
are issues concerning the choice of commodities
and their costs, various combinations of which
can deliver the same calorie count. For subsistence,
a certain amount of non-food items are also
required, which must be taken into account.
These are complex issues; some of them may be
overcome, for example, with the availability
of relevant information, while for others plausible
compromises may or may not be worked out.
5. The end result of these conceptual and measurement
difficulties is confusion in estimating the
numbers of poor people. Examples abound where
for the same year the national estimates on
poverty incidence vary widely, such as from
26 to 85 per cent4.
The same is true at the international level,
as various estimates of the total number of
the poor at the global, regional and subregional
levels show wide margins of variations.
6. To know the number of poor people is without
doubt of political and administrative interest;
it may also help to mobilize public awareness
in the more affluent countries, with resultant
resource flows particularly through non-governmental
channels. It is also useful for research and
intercountry comparisons for assessing the situation
in the region (and globally) and for understanding
the overall influences of various development
strategies on poverty alleviation. However,
mere knowledge of the number of the poor is
a blunt tool in the fight against poverty. For
the policy point of view it is important to
know where the poor are and what are the factors
contributing to their poverty. In fact, poverty
is a multidimensional phenomenon which can be
caused by various combinations of several factors.
Therefore, it is required to study the characteristics
of people below the poverty line in terms of
characteristics such as income distribution,
occupation, unemployment, widowhood, female-headed
households, family size, and on identifying
their specific deprivations in terms of various
factors, including income, employment, housing,
land ownership, health, education, water and
food. Such an extensive knowledge of the characteristics
and profiles of the poor is important for implementing
sustainable strategies for poverty alleviation.
The causes of poverty and its consequences -
both to the individual and the society - must
also be studied.
Absolute and relative
poverty
7. The concept underlying headcount measures
is characterized as "absolute poverty" which
considers an individual as poor whose standard
of living is below the minimum acceptable level
of subsistence. Since the concept of absolute
poverty is based on the notion of ability to
meet basic minimum needs for sustaining life,
the content of its underlying poverty standard
is taken to be fixed across time and space5.
It can be argued that the notion of subsistence
is never fixed at least across space as 'needs'
vary from one place to another according to
such factors as climate and customs. What has
been termed as 'absolute poverty' is also relative
to the extent that standards change over time.
For various reasons, including the lack of data,
the headcount measures now used in developing
countries utilize a notion of subsistence which
is very close to mere existence. Otherwise,
the modern notion of subsistence includes not
only physiological necessities, but also social
and legal needs6.
8. The concept of 'relative poverty' is mostly
used in developed countries, where poverty is
defined as the inability to attain a given contemporary
standard of living. It goes beyond the biological
approach and defines the poverty line based
on what is required for leading a full life.
Here, the poor households are compared with
an 'average' household in terms of their power
over resources (or income). The standard for
'full life' may be ascertained in a variety
of ways, such as from the actual living conditions
of the people or from their perceptions about
an adequate standard of living. The notion of
a decent standard may also be found in development
plans and national policy statements. However,
such statements may not be directly relevant
for determining standards of living as the former
may reflect the commitment of the state according
to the resources available to it. The measurement
of poverty according to a relative approach
demands much wider variety of statistical information
than that utilized by developing countries under
the absolutist approach.
Statistics and Indicators
of Poverty
9. Any meaningful measurement of a social phenomenon
ideally requires some sort of conceptual framework,
the absence of which often leads to the collection
of irrelevant data, misguided utilization of
existing information and misleading interpretation.
The statistical development suffers, as national
statisticians do not have the capabilities for,
nor should they engage in, the theoretical work
necessary for producing concepts, definitions
and guidelines for operationalizing statistical
measurement. The situation can be improved if
concerted efforts are made in the countries
to review existing concepts, definitions and
measurement processes and if adequate resources
are devoted to improve them through a systematic
process of studies and research.
10. There is no doubt that for understanding,
assessing and monitoring the poverty situation,
the availability of quality and reliable information
is important. However, in developing countries
the required data are scanty and whatever information
is available is obtained as a by-product from
data sources intended mainly for other purposes.
That situation leads to the choice of statistical
indicators which sometimes could be misleading.
Very frequently, averages and national figures
are used which reveal very little, if anything,
of the situation of the poor. For example, the
national data on population per hospital bed,
per physician or per nurse do not tell us much
about the access of the poor to health services.
Similarly, the use of the national estimates
of ratio of food to population could play an
obscuring role.
11. A study by the United Nations Statistical
Division (UNSTAT) has identified a range of
problems that developing countries face in measuring
and monitoring the social and human development
process, including poor data quality, outdated
information, lack of awareness of current data
availability among users, and lack of disaggregated
data, such as on the urban poor7.
The main reasons for the paucity of reliable
data in most developing countries of the region
are lack of adequate human and financial resources,
lack of coordination among statistical information
gathering agencies and poor management of existing
resources. However, in the case of poverty statistics,
the undue emphasis placed on headcount measures
has also hampered any development and improvement
of data which might otherwise could have evolved.
While borrowing the idea from Rowntree of the
poverty line, later researchers have rather
ignored the fact that the brilliance of his
study lay in the use of a multiplicity of data
sources, methods of data collection and analysis
to obtain detailed evidence on the factors associated
with poverty, as well as the replication of
the study. These are areas which should still
be of concern in any serious attempt to improve
poverty statistics in the countries of the region.
12. The data sources which may potentially
be used for deriving poverty-related statistics
and indicators include population censuses,
household surveys and administrative records.
However, as mentioned earlier, the data collection
processes often do not focus on the concerns
of poverty as such. The population and housing
censuses could serve as an important source
of disaggregated data on housing conditions,
including access to safe drinking water, occupation
and unemployment. However, the information is
not useful for frequent assessment and monitoring
as censuses are conducted quinquennially or
decennially.
13. The potentially most important sources
of poverty-related information are household
surveys, which can provide periodic data for
monitoring. Since administrative records, such
as civil registration -- which can provide a
continuous flow of data for monitoring -- are
mostly inadequate in the developing countries
of the region, greater reliance will have to
be placed on household surveys to obtain data.
Currently very few countries collect data in
household surveys on such important aspects
as housing, health, education, water and utilities.
Consequently census data are cited, which are
useful only as background information. Landlessness
is another aspect of poverty which hardly figures
in survey data collection. The sole source of
the limited poverty-related statistics in developing
countries has frequently been surveys of income
and expenditure, concerning primarily consumer
expenditure. Income data are unreliable due,
for example, to under-reporting by higher income
groups. More importantly, in developing countries
it is difficult to resolve issues concerning
the treatment of production for sale and home
consumption. Even for expenditure data, there
is a need for improvement as questions have
been raised concerning the consistency of definitions
for comparability over time, the quality of
expenditure data and the representativeness
of certain population groups. Moreover, the
absence of data on some important aspects, such
as wage rates for different categories of the
population, makes it difficult to undertake
any serious study.
14. The above issues identified with the use
of household survey data for poverty studies
clearly point to the need for conducting special-purpose
enquiries. The nature of enquiries would depend
on the state of current knowledge in the country,
e.g., when there is sufficient knowledge for
identifying the poor, the resources may be used
for some other purposes such as to identify
specific deprivations of the poor. Depending
on the availability of resources, large-scale
surveys may be required to obtain reliable data
on subcategories. Another alternative is to
conduct separate (and preferably frequent) surveys,
focusing on specific areas and population groups.
Conclusions
15. In developing countries the work on the
measurement of poverty has been overly concerned
with the estimation of poverty incidence, while
several other important aspects, such as the
understanding of the deprivation of the poor,
the causes and consequences of poverty and the
monitoring of progress in poverty alleviation
have received inadequate attention. The disaggregated
statistical data on poverty-related issues are
also scanty as developing countries lack the
resources and capacities to meet that demand.
However, the situation can be improved if attention
is paid to clarifying concepts and refining
definitions, without which statistical data
collection and analysis will continue to progress,
if at all, in a vacuum. There is also a need
to develop and implement innovative ways of
data collection, such as from household survey
enquiries and administrative records, rather
than depend heavily on "secondary" data. The
role of international organizations should be
strengthened and their efforts should be coordinated
to improve the measurement of poverty in the
countries of the region. There is also a need
to improve the methodology for developing a
set of standard and comparable indicators which
way be used for monitoring the progress made
by the region in alleviating poverty. Collection
of a minimum set of data at the subnational
level - such as district, province or state
- should be seriously pursued to obtain disaggregated
estimates. Since many national statistical offices
already have field offices or have some sort
of influence on local statistical offices, of
a higher priority attached to this matter could
improve the situation considerably.
-----------------------------------------
* The author is a staff member
of Statistics Division of ESCAP. The views expressed
are those of the author and do not necessarily
reflect those of the United Nations.
2
See for example, Amartya Sen, Poverty and Famines:
An Essay on Entitlement and Deprivation (NewYork,
Oxford University Press, 1982); and Wolf Scott,
Concepts and Measurement of Poverty (Geneva,
United Nations Institute for Social Development,
1981).
3
B.S. Rowntree, Poverty: A study of Town Life,
Macmillan, London, 1901.
4
Scott, 1981, p.24. Also see, for example, Henk-jan
Brinkman, "Why estimates of the incidence of
poverty differ", Working paper No. 14, DIESA
Working Paper Series, October, 1990.
5
United Nations Development Programme, Human
Development Report 1990(New York, OxfordUniversity
Press, 1990).
6
"Every family at the poverty level should be
able, first of all, to satisfy its physiological
needs for food, shelter and clothing, sufficiently
for physiological survival; that is, it should
have sufficient income to be reasonably free
from the disease or death caused by malnutrition
or deficient shelter or clothing. In addition
family income should be sufficient for the family
to be able to conform with the laws and regulations
of society.... Finally, a family living at the
poverty level should be capable of achieving
a minimum of social acceptance and personal
development. These three aims (the physiological,
the legal and the social) are roughly hierarchical:
each implies higher quality or additional goods
and services than the preceding one." (United
Nations, Social Policy and the Distribution
of Income in the Nations, Sales No.E.69.IV.7,
New York, 1969).
7
"Programme to monitor the achievement of social
goals in the 1990s and related methodological
work" (E/CN.3/1993/20).
----------------------------------------- |
Indonesia
Central Bureau of Statistics of Indonesia (BPS-WEB)
has an online service targeted for the Internet
community since October 1995. The BPS-Web provides
public with sample tables, organized in entries
such as agriculture, population, employment,
foreign trade, prices, national and regional
income, social welfare, manufacturing, mining,
energy, construction etc. Futher an explanation
is included pertaining the data collection methods,
concepts and definitions, coverage. The user
can browse Indonesia's statistical profile.
The homepage of BPS-web is URL: http://www.bps.go.id/.
For more information please contact the BPS-Web
of Bureau SIS attention B. Isworo E-mail: Webmaster@mailhost.bps.go.id
United Kingdom
New Statistical Organization
Emerges in the United Kingdom Office for National
Statistics
On 1 April 1996 the Central Statistical Office
and the Office of Population Censuses & Surveys
will be merged to form the office for National
Statistics (ONS) in the United Kingdom. Dr Tim
Holt, Director and chief executive of the newly
created ONS noted that for the first time in
the country, a single government agency would
provide a comprehensive picture of the economic
and social life of the nation. "This means that
all the important data about the lives of everyone
in Britain, together with information about
business and high-level economic and social
statistics, will be brought together in a single
organisation. An important part of our mission
is to make that information more accessible
and easier to understand."
The merger aims to meet a widely perceived
need for greater coherence and compatibility
in government statistics through improved presentation
and easier public access. The merger proposals,
published on 25 April 1995, formed the basis
of public consultation until the end of July
last year. They were widely welcomed and the
need for better intergeration and availability
of statistics confirmed. The ONS will have about
3400 staff.
The ONS Director, who will also serve as Registrar
General and Head of the Government Statistical
Service, said: "Reliable and professionally
produced official statistics are a vital and
important part of any democracy. They make possible
an independent assessment of economic and social
developments and give people the opportunity
to make up their own minds. I want official
statistics to be more widely available and more
widely used not just by businesses and analysts
but by ordinary people. I want to help overcome
the average Britain's fear of figures by making
them more user-friendly, more relevant and more
easily understood. Statistics that are collected
or created and then never referred to or used
again are of no value. The policy of the office
for National Statistics will be to design, collect,
process and publish statistics that are wanted."
In launching the new organisation,
Dr Holt gave four pledges:
To the government: He pledged
that ONS statistics would be produced to the
highest professional and internationally recognized
standards, to a mutually agreed timetable, and
using the most cost-effective methods: that
the organisation would be responsive to changing
needs; and that it would try to provide the
statistical base for policy development.
To the public: He pledged
to make ONS statistics more user-friendly and
more readily available, and to encourage statisticians
in other parts of government to follow suit.
To business: He pledged to
make available to business and commerce the
information they needed to maintain competitiveness;
to keep demands for data to the minimum consistent
with ensuring the information gathered was of
good and usable quality; and to ensure the forms
businesses were required to complete were kept
as simple as possible.
To the individual: He pledged
to protect the privacy of all personal data,
and to provide an effective serice for civil
registration.
Dr Holt acknowledged that his aim of making
the ONS a world leader in every area of its
business could not be achieved without the commitment,
professionalism and dedication of the staff.
In a fifth pledge, Dr Holt promised his staff
that he would work to provide the resources
necessary to carry out their duties; and that
he would ensure they were properly trained to
the highest standards.
The current address of
the ONS is as follows:
- Government Offices, Great
George Street, London SW1P3AQ
- Tel: 0171 270-3000; Fax:
0171 270-6019
|
About Us | Media Centre | Members | Programmes | Documents | Publications | Jobs | 
Bangkok, Thailand