For the Record - On the redefinition of social indicators

This article seeks to clarify some of the points and misconceptions concerning two important social indicators – unemployment and poverty – mentioned in the article “Paranoid gov’t. seeing communists in media,” which came out in the Postscript column of Mr. Federico D. Pascual, Jr. in the 24 August 2006 issue of the Philippine Star

Mr. Pascual’s statement

“The NEDA under Neri has revised statistical methods and definitions to come up with better social indicators and make them affirm the claimed economic growth.

“By simply changing the definition of joblessness, and not by job creation, Neri was able to reduce the unemployment rate to 8.2 percent in April…

“Using its revised definition, government pegs the national average poverty threshold per day at only P33.72, which means that a Filipino who earns this much daily is already considered non-poor. This has allowed the Arroyo government to claim that it has lowered poverty incidence from 33 percent of the population in 2000 to 30 percent now.”

NSCB clarification

A. On the revision of official statistics

We would like to stress that the revision of official statistics, whether in terms of concepts, definitions, or methodologies, is a standard practice recognized and appreciated by the international statistical community. Dr. Romulo A, Virola, the NSCB Secretary General, has in fact provided at least five reasons why this is necessary in his 12 September 2005 “Statistically Speaking” article in the NSCB web site. These include the following:

(a) Official statistics are oftentimes released based on incomplete data from surveys and administrative reports and hence, have to be revised as soon as supplementary data become available.

(b) Official statistics is an evolving science that must adapt to emerging concepts, definitions, and classifications, and changes in existing ones.

(c) Improvements in statistical methodologies and the need of researchers and econometric modelers for comparable, consistent and long data series require the statistical offices to revise past estimates computed using old methodologies.

(d) Revisions are inherent in some statistical tools.

(e) Revisions are sometimes necessary to correct for human errors in official statistics.

B. On the revised definition of unemployment

The issue on the revised definition of unemployment has been discussed in detail in another “For the Record” article, which was released in response to a news feature that came out in the Daily Tribune.

C. On the revised methodology for poverty estimation

Based on Executive Order No. 232 on the System of Designated Statistics, the NSCB is mandated to generate the official poverty statistics for the country. For this purpose, the NSCB initially used a menu-based methodology following the basic needs approach to come up with national and regional poverty estimates with urban and rural disaggegation.

Since the release of the first set of official poverty statistics for 1985, the official poverty estimation methodology has so far undergone two major revisions. The first was in 1992 when the NSCB adopted a revised methodology as a result of the efforts of the Philippine Statistical System (PSS) to improve the measurement of poverty in the country. The main difference between the 1985 and the 1992 methodologies was on the non-food component of the poverty threshold. Under the 1992 methodology, non-food items previously included in the estimation of the threshold, but were later found to be non-basic, were excluded from the computation. These include alcoholic beverages, tobacco, recreation, durable furniture and equipment, and miscellaneous expenditures. The 1992 methodology was used to produce back estimates for 1985 and 1988, and to generate poverty statistics for 1991, 1994, and 1997.

The second major revision was in 2003 when the use of provincial prices was introduced to allow for the estimation of provincial poverty statistics. This new methodology, which adopted the old region-based menus but costed out using provincial prices, was approved by the NSCB Executive Board through NSCB Resolution No. 1, Series of 2003 issued on 15 January 2003. The NSCB used this methodology to come up with the first set of official provincial poverty estimates covering the years 1997 and 2000, and subsequently, for the 2003 preliminary and final poverty estimates, which were released on 24 January 2005 and on 6 June 2006, respectively.

While the 2003 basic methodology has been maintained, refinements in estimation procedures are introduced whenever necessary in order to improve the accuracy and reliability of the official poverty statistics being released by the NSCB and in response to improvement activities being done by the PSS. Following the usual practice for the revision of statistics, proposed refinements are rigorously discussed in a series of meetings of the NSCB Technical Staff and the Technical Committee on Poverty Statistics (TC-PovStat)¹ and in users and producers’ forums, and are presented to the NSCB Executive Board for approval.

The refinements adopted for the 2003 final poverty estimates and back estimates for 2000 were approved through NSCB Resolutions No. 05-2005 and 11-2006. The results of the computational exercises conducted by the NSCB Technical Staff concerning these refinements, as presented to the TC-PovStat on 5 May 2006 and to the Board on 22 May, indicated that they had minimal effects on the poverty incidence estimates.

Thus, the 3.0 percentage point reduction in poverty from the year 2000 to 2003 could not be attributed to the refinements on the poverty estimation procedures. Rather, this was actually due to the faster increase in the average per capita income of families in the lower income deciles relative to the poverty threshold. A table illustrating in more detail the changes in the per capita income distribution by income decile in 2000 and 2003, as derived from the 2000 and 2003 Family Income and Expenditure Survey (FIES) has been provided by the NSCB in an earlier “For the Record” article in response to a related issue raised by the Daily Tribune

D. On the revision of definitions in other sectors/sciences

In the compilation of the national accounts, the international guidelines on concepts, definitions, classifications, and accounting rules as contained in the 1968 United Nations System of National Accounts (UN SNA) have been superseded by the 1993 UN SNA. Necessarily, countries shifting from the 1968 UN SNA to the 1993 UN SNA have to undertake a comprehensive revision of their national accounts estimates. For the Philippines, initial revisions based on the 1993 UN SNA have recently been introduced in the input-output tables, which serve as the benchmark for the national accounts. Some of these revisions involve changes on concepts and definitions that include the following: (1) military expenditures, which used to be part of government consumption before, are now treated as investments; (2) the monetization/demonitization of gold, which was once measured under exports and imports, is now considered under other volume changes; and (3) imports, which were valued at cost, insurance, and freight (CIF) before, are now valued free-on-board (FOB).

Changes in concepts and definitions are not confined to statistics. In a totally different field, a new definition of the word “planet” has been approved by the International Astronomical Union, composed of respected experts in this field, to take into account new information and recent scientific discoveries enabled by technological advances. Under the new definition, Pluto, the smallest of the nine currently known planets and the farthest from the sun, will no longer be considered as a full-fledged planet. This means that the solar system will now be composed of only eight planets. Pluto has been classified as a “dwarf planet” along with Charon, its largest moon, the asteroid Ceres, and the so-called 2003 UB313, which was discovered only in 2003 and is now affectionately called Xena.

The NSCB would like to thank the Mr. Pascual and the Philippine Star for their interest in statistics. For inquiries, please contact Ms. Redencion M. Ignacio of the NSCB Social Sectors B Division at telephone numbered (632) 890-5390 or at e-mail address rm.ignacio@nscb.gov.ph.

¹The Committee has a multi-sectoral representation consisting of noted experts in the area of poverty measurement and is mandated to study and recommend improvements on the methodologies and estimation procedures used in generating official poverty statistics and related indicators