Monday, November 28, 2016

Statistics Bill 2016 - I


The Statistics Bill for Myanmar appeared in Kye-Mon (The Mirror) daily in three parts on November 19-21, 2016. As a statistician of sorts in a government agency (actually we dealt mostly with administrative records) a long time ago, I was interested in statistical systems and had read a bit of works in the Journal of Official Statistics and absorbed some knowledge about official statistics and its organization through the earlier version of the Handbook of Statistical Organization by the United Nations. By the time I had left my job as a government employee and became a dabbler in statistics in the Pacific and then in Jakarta after that, Fundamental Principles of Official Statistics by the United Nations Statistical Commission had came out and by the nature of my work I got a bit familiar with it.

With this background, when I started reading the Statistics Bill, I've taken for granted that the subject matter of the Bill will be the domain of “Official Statistics” and its organization as envisaged in those two UN documents and those which grown out of them. After-all, has not the opening paragraphs of the Handbook stated its conviction as:

The two main intended audiences for the Handbook of Statistical Organization, Third Edition: The Operation and Organization of a Statistical Agency are: (a) the chief statistician (or soon to be chief statistician of a statistical agency) and his or her colleagues; and (b) those charged with oversight of the official statistics function.

As the heads of agencies, those who hold these positions are both formally and emotionally committed to continuity of a tradition embodied by the Fundamental Principles of Official Statistics.

Nevertheless I was puzzled by the choice of wording used in the Statistics Bill:


Within the context of that paragraph, I felt that the marked text above would stand for “official statistics”. In Myanmar language the wording "တရားဝင်" is taken to mean lawful or legal or legitimate in common usage. For example this is from English-Myanmar Dictionary (Third Printing) by Myanmar Language Commission:


So it looks as if the Bill has adopted a common usage translation of “official” with a connotation very different from “official statistics” taken as a whole. Unfortunately, the dictionary doesn't include “Official Statistics” which is a technical term having a specific meaning and as such could be overlooked even in such a big volume of 1,621 pages or other standard dictionaries. However we could find Wikipedia defining Official Statistics as:

Official statistics are statistics published by government agencies or other public bodies such as international organizations as a public good. ... This bulk of data is usually called official statistics.

Anyway, to please those who are skeptical of any citation involving Wikipedia, here is the definition from an authoritative source (Committee for the Coordination of Statistical Activities SA/2012/3, Nineteenth Session 21 February 2012):

What is official statistics?

Official statistics is defined in the SDMX Content‐oriented Guidelines as “any statistical activity
carried out within a national statistical system, or under the statistical programme of an
intergovernmental organization”. However, this is a rather circular definition and it is not widely
recognized outside the statistical community. Different interpretations may occur for statistics
produced by governmental and public agencies outside national statistical systems (which may or
may not follow standard statistical practices and principles), or for statistics produced as
collaborative efforts between statistical offices and other institutes (NGOs, research institutes,
public sector).

The key question is what makes statistics as “official”. The label “official” can be based on the
characteristics of the actors and institutions (the source) or on the characteristics of the process.
Few countries clearly define official statistics by law. For example the UK Statistics and Registration
Service Act (2007) defines official statistics as “all those statistical outputs produced by the UK
Statistics Authority’s executive office (the Office for National Statistics), by central Government
departments and agencies, by the devolved administrations in Northern Ireland, Scotland and
Wales, and by other Crown bodies (over 200 bodies in total)”1. At international level, there is not
an universally accepted definition of official statistics. Outside of the statistical system, agencies
which focus on thematic areas, may have a different connotation of official statistics from the one
included in the SDMX Guidelines identifying official statistics as any data provided by
representatives of member states (which may not have a close relationship with national statistical
systems1).

Another key question is "do available official statistics exhaust the domain of relevant economic
and social information, or does it exist additional sources of good quality data that complement
official ones, or provide information not produced by national statistical offices? In modern
societies, an increasing amount of information is produced by non‐official organizations, which
may even include in their routine activities the production of specialized statistics, produced by
professional statisticians or derived from their own administrative registers.

Very often the paradigm to elevate official statistics at a higher level is based on the assumption
that official statistics is most solid, complete, and independent. This can be true for many countries
and several areas, but it is not universally true. In countries where the statistical process is not
transparent, statistics on areas that are highly politically sensitive and relatively “young” in terms
of existing statistical standards (for example environment, food security, poverty, crime, drug
trafficking, ….) official or government statistics may not always meet the highest quality standards.

Development of guidelines/best practices on the use of nonofficial data
Follow up note
from the Task Team composed by: UNODC (chair), Eurostat, IMF, ITU, The World Bank,UNECE1, UNESCAP, UNHabitat, UNIDO, WTO
Available here.

Then a drastically simplified definition of official statistics may be given as: “statistics produced by government and related bodies”. So shall we not call it simply “ အစိုးရစာရင်းအင်းor “အစိုးရနှင့်၊အစိုးရမိတ်ဖက် စာရင်းအင်းor anything to that effect? I don't know if the wording “တရားဝင်is meant to be value laden as the law/legal/legitimate qualifier for the word “statistics”. If so the third and fourth paragraphs of the preceding excerpt legitimately questions this notion. Then shouldn't we go non-committal in the wording and let users and public decide the merits of our work afterwards?

Additionally, here is how a National Statistical Authority defines National Statistics accommodating the needs of changing times:

What are ‘National Statistics’?

National Statistics’ are a subset of official statistics which have been certified by the UK Statistics Authority as compliant with its Code of Practice for Official Statistics. The label currently comprise three basic types:
  • legacy ‘National Statistics’ – those statistical products which obtained their designation as ‘National Statistics’ before April 2008, in accordance with the arrangements set out in the ‘Framework for National Statistics’ (2000) and which have retained their status, but which have not yet been formally re-assessed by the UK Statistics Authority for compliance with the Code of Practice for Official Statistics. The Assessment Programme began in November 2008.
  • re-assessed ‘National Statistics’ – any statistical product which has retained its National Statistics status after a formal re-assessment by the Statistics Authority of compliance with the Code of Practice for Official Statistics.
  • new ‘National Statistics’ – any statistical product which has been proposed by ministers as a candidate ‘National Statistics’ which has been subject to its first formal assessment by the UK Statistics Authority and which, as a result, has been granted accreditation by the Statistics Authority.


No comments:

Post a Comment