What makes official statistics trusted by users?

1
What makes official statistics trusted by users?

• Heinrich Brüngger
• Director, Statistical Division
• United Nations Economic Commission for Europe,
  Geneva

2
Economic Globalisation A Challenge for Official
Statistics

• Joint EFTA/ECE/SSCU Seminar
• Kiev, 3-6 July 2007

3
Whose trust?

• FP Reference to UN Fundamental Principles of
  Official Statistics
• Trust by the various user categories is only one
  element of trust
• Other elements
• Trust by respondents (in exclusively statistical
  use and non-disclosure of unit-level information)
  FP6
• Trust by budgetary authorities, esp. Parliament,
  that official statistics, which uses tax-payers
  money, provide good value for the society as a
  whole, and has a sound legitimacy FP1/FP5

4
User categories

• Public at large (and mass media as their
  intermediaries) citizens right for information
  in a democracy FP1
• Governmental and related public sector users (at
  national, regional and local levels)
  evidence-based decision making
• Business and financial community (national and
  international) NGOs
• Research community
• Educational institutions
• International organisations

5
Trust by users

• Trust by users can be subdivided
• Trust in results and products of official
  statistics
• Trust in the institutions that act as producers
  of official statistics and in their senior
  management
• Trust in the specific institutional set-up for
  official statistics and the respect by everybody
  of the legal framework for official statistics

6
Trust in results

• Relevance of outputs (including adaptation of the
  system to evolve and address new phenomena such
  as globalisation) FP 1
• Authoritativeness of results (derived from
  processes and institutions, plus from
  coordination of dissemination process across the
  system) FP 8
• Quality (timeliness, accuracy, comparability over
  time, space and between groups etc.) FP 5
• Accessibility of data (including comments for
  non-expert users) and of metadata FP1/FP3

7
Trust in producers of official statistics,
notably NSO

• Integrity (all FPs that address ethical behaviour
  and professionalism depends on institutional
  framework and senior management)
• Professional expertise of staff
• Proven ability to meet ad hoc requests for
  additional tabulation
• Transparency about methods and sources and
  processes of decision-making FP 3/FP 7

8
Trust in producers (ctd.)

• Networks for regular contacts with user groups
  FP1
• Impartiality (perceived as being outside partisan
  political games) FP1
• Adherance to international standards and
  internationally recognised good practices FP 9
• NSO not being in charge of non-statistical tasks
  that may create conflicts of interests with
  impartiality or confidentiality

9
Trust in producers (ctd.)

• System of quality management and monitoring
• Comments by the NSO that transform released
  figures into information
• Regular analytical activities by the NSO that add
  insights without being policy-prescriptive
• Way to react to critics
• Way to address and communicate errors
• Limited but effective use of the right to react
  to erroneous interpretations and misuse of
  statistics FP4

10
Chief Statistician (CS)

• For the trust in the NSO and the statistical
  system as a whole, personality and status of the
  CS is key
• He/she personifies professionalism and integrity,
  especially impartiality, to users, the public,
  respondents and staff
• Selection and appointment process has to reflect
  these particular roles and can therefore differ
  in certain aspects from the general process that
  is applicable in the government for this level

11
Trust in institutional set-up

• Specific legal framework for official statistics
  (statistical law)
• Statistical law has to ensure professional
  independence for NSO and all other producers FP2
• Government has to refrain from any interference
  into professional issues of official statistics
  (the  how ), and back the NSO (and other
  producers) against attempts of interference into
  professional decisions from other sides

12
Trust in institutional set-up (ctd.)

• Dissemination and release process is key for
  independence
• No clearance procedure by any authority that is
  outside the statistical system
• Simultaneous release for all users no advance
  information for governmental users
• Advance communication of release dates fixed by
  CS
• Strict separation between comments made by the
  NSO and comments made by users, especially
  government users

13
Trust in institutional set-up (ctd.)

• Once appointed, CS has to have legal protection
  against threats of being fired or removed to
  another position during his term
• For this reason, CS should be appointed for a
  fixed term of at least 4 years, irrespective of
  changes in governments
• These rules have be enshrined at the level of law
• Government has to refrain from attempts to
  sidestep or circumvent these protections

14
Trust in institutional set-up (ctd.)

• NSO should have the legal rigth to accept funding
  from other budgets than its own for certain
  statistical activities, including from
  international donors, but not in exchange of
  compromising on the FPs
• As a consequence, the results of such activities
  have to be in the public domain
• A Statistical Council as advisory body, composed
  of main users, can add credibility in monitoring
  compliance with FPs and ensuring relevance and
  adaptation to user needs

15
Trust in official statistics 3 asymmetries

• timewise it takes a long time to build up traust
  and credibility, but it can be destroyed rapidly
• throughout the national system of official
  statistics the credibility of the NSO is not
  easily extended to other producers, but any
  serious misconduct in the system can undermine
  the trust into the system as a whole

16
3 asymmetries (ctd.)

• international trust and credibility cannot be
  imported from neighbours, but in a closely
  integrated system like the ESS, any scandal
  affecting one member sends shock waves across the
  whole system (example Greek scandal)
• Regular reviews and global assessments of the
  statistical system, or of key statistics, by
  international peers are important for
  credibility, if the results are public and
  communicated to the government, and if gaps
  identified are addressed subsequently

17
Summary

• Take care of your capital of trust and
  credibility, use all opportunities to add to this
  capital, but have a continuous policy of
  prevention against anything that might undermine
  it concerning all elements listed (watchdog
  function)
• Have regular contacts with all user groups, and
  repeat explaining the role of modern official
  statistics to them

18
Summary (ctd.)

• For users
• Do not judge the NSO by whether a new release
  supports your policy and advocacy positions, but
  rather by whether you would trust the results as
  being unbiased irrespective of your present
  political and other preferences, on the ground
  that they are produced and disseminated by
  institutions which credibly adhere to the
  international standards of official statistics

19
Thank you for your attention