Chair: Jean-Pierre Poncelet
Room: S3A Barbakan
Time: 13:15 - 14:00
Date: 27 June
|Title: <<< A common quality framework for Swiss federal statistical offices >>>
The presentation proposed by Switzerland concerns the creation of common quality criteria that can be used by all swiss federal statistical authorities. It aims to present the various stages and difficulties encountered so far, the defined quality criteria and the possible mechanisms for their implementation and evaluation in the future. Having a coherent and effective national statistical system has been a priority for the members of the European Statistical System for many years. Switzerland is no exception to this reality and the Federal Statistical Office is committed to intensifying the dialogue with its federal partners on quality in public statistics. The notion of "quality" is undoubtedly complex and multidimensional. A working group has been set up bringing together several actors from the swiss federal statistical system to define a set of concrete and comprehensible quality criteria. In order to do so, the working group relied on the European Statistics Code of Practice (CoP), the recommendations of the Peer Review 2014/2015 of the swiss statistical system and the “ESS Quality Guidelines and Performance Indicators" but has limited its reflections on the elements that have emerged as priorities. This made it possible to have a narrower reference framework than that of the CoP, while ensuring an acceptable level of compliance for users and encouraging their implementation and continuous improvement.
|Title: <<< The implementation of the OECD Recommendation on Good Statistical Practice: Professional independence and coordination >>>
In 2015, the Organisation for Economic Development and Cooperation (OECD) adopted its first legal instrument for statistics, i.e. the Recommendation on Good Statistical Practice (hereafter the Recommendation) intended to provide a common reference to assess the quality of national statistical system and official statistics, which are fundamental for OECD statistical and analytical works. The Recommendation complements existing codes of practice and international standards currently applied by OECD Member countries, but is also more specifically relevant for OECD statistical activities, for example in including good practices on the use of new sources of statistical information, or on the coordination of the statistical system.
The Recommendation includes twelve recommendations structured in five main areas: (I)institutional, legal and resource requirements that enable statistical systems to function; (II)methods, quality and processes of statistical production; (III) dissemination; (IV) co-ordination and co-operation; and (V) statistical innovation. Each of the twelve recommendations is presented with a set of indicative good statistical practices based on the OECD’s experience in statistical reviews.
This paper presents the Recommendation and provides an overview on the importance of quality for an international organisation as the OECD. While the professional independence of national statistical authorities was questioned in several recent cases in OECD member and non-member countries, the paper also sheds some light on specific questions related to this principle as well as to the coordination of the National Statistical System. To this end, preliminary results from the activities currently carried out by the OECD in order to assist adherents in implementing the Recommendation.
|Title: <<< How to turn quality into a habit in the statistical production? >>>
One of the main purposes of the Statistics Department of Banco de Portugal is to ensure a statistical production with high quality standards aiming at fully meeting users’ needs, aligned with the best practices and procedures recommended by the international organizations. Following its commitment to quality, one of the Bank’s priorities is to develop a wide set of quality control procedures that ensure high levels of regular and thorough review of the key statistical outputs. Statistical quality control is based on different procedures and working arrangements that make sure that processes are effective and efficient and the risks are mitigated. In order to achieve higher quality statistics, there are several quality indicators performed by the primary statistics’ compilers. This paper will present the main quality indicators used and the ongoing process to improve the model of regular and systematic quality controls.
|Title: <<< A Framework for Assessing the Quality of Banking Supervision Data >>>
Since the start of the economic and monetary Union, the European Central Bank (ECB) has placed strong emphasis on key aspects of statistical quality, as described in its Statistic Quality Framework. The introduction of a harmonised framework for supervisory reporting in 2014 has raised the awareness of competent authorities on the need to apply similar high standards of statistical quality to data collected for the purpose of banking supervision. Accordingly, the ECB has established a process for the quality assessment of supervisory data, with a twofold purpose. Firstly, checking whether data constitute a suitable basis for informing supervisory decision; secondly, recognising data quality as an integral part of banks’ supervisory evaluations. Following these two objectives, a number of data quality dimensions have been identified and implemented as part of a general framework for assessing the quality of banking supervision data. This paper presents some reflections on the framework, based on the experience gained at the ECB over the past four years.
|Gloria Martha Rubio Soto
|Title: <<< Quality Assurance and the GSBPM adoption at INEGI >>>
At the end of 2014, INEGI introduced new institutional and technical measures for strengthening data quality, as part of the Quality Assurance Norm (QAN). The Norm outlines a general framework for quality assurance, establishes assessment requirements and defines institutional arrangements. The QAN implementation will involve several phases over the medium and long term. The initial implementation phase comprised the quality framework definition, a pilot self-assessment exercise, and priority setting. More recently, the quality assurance reform has been intertwined with the gradual adoption of the General Statistical Business Process Model (GSBPM). The document will review the progress to date, discuss the lessons learnt from the early implementation experience and examine the challenges ahead.
|Title: <<< Cost Accounting of Products in Turkish Official Statistical Program >>>
Cost accounting is a relatively new concept in the public administration. Many countries have adopted this approach since the 1990s. It helps increase productivity by providing essential inputs to decision makers. But it is not a common practice among National Statistics Offices to measure the cost by statistical product. International organizations are supporting cost analysis of statistical products. United Nations’ Sustainable Development Goal indicator 17.19.1 is about the measurement of “dollar value of all resources made available to strengthen statistical capacity in developing countries”. In 2016 and 2017 TurkStat conducted two studies to calculate the cost of statistical products in Turkish Official Statistical Program. One was done to calculate the cost of statistics produced by TurkStat and the other to calculate the cost of statistics produced by other national authorities (ONAs). First, we measured costs of all statistics produced by TurkStat for 2015 and 2016. In this study both direct and indirect costs were calculated by product. Indirect costs are allocated to statistics by cost drivers. Cost driver must be a kind of factor which has the highest impact on all cost components. The number of personnel and their salaries are used as cost driver; since personnel expenditure makes up the largest part of total cost. Second, a survey was undertaken with ONAs to collect data on the number of personnel working on the production of statistics and their full time equivalent in 2015. Personnel expenditure was calculated by statistics. However indirect costs were not included in the cost calculation for ONAs. Methodology and results of the cost analysis study in Turkish Official Statistical Program are presented in this paper.
|Title: <<< The NSO, the NSS and Beyond! >>>
This paper takes a brief look at some early ideas and concepts about quality management at a National Statistical Office (NSO), and reflects on the breadth and depth of content included through extensive collaboration in the first generic National Quality Assurance Framework (NQAF). Since the deployment of the generic NQAF in 2013, many NSOs have adapted it to their own circumstances and adopted either a regional or their own national version. The journey is long yet fruitful. The exercise of reviewing priorities, challenges, bottlenecks and inefficient practices provides an opportunity to develop and document good quality practices, and goals to work towards. While the expected audience of an NQAF is the NSO, in few countries does the NSO produce all or even most of the official statistics. Hence the intended audience for an NQAF should really be all federal ministries, departments and agencies (MDAs) participating in the National Statistical System (NSS). The latter half of this paper looks at the experiences of the Statistical Institute of Jamaica providing data quality workshops to introduce other participants of the Jamaican NSS to quality assurance, and the experiences of Statistics Canada in producing a data quality toolkit intended for data producers and users outside of the NSO.
|Raquel Rose Silva Correia
|Title: <<< Improvements on the Brazilian Statistics Code of Practice >>>
In 2013, the Brazilian Institute of Geography and Statistics (IBGE) made available its Statistics Code of Practice, disseminating a set of guidelines, principles and practices that the Institute is committed to uphold in the statistical production process, taking as reference the Statistics Code of Practice for Latin America and the Caribbean. The Code aims at standardizing professional procedures to foster best practices in statistics, which are crucial to establish institutional credibility and, consequently, trust in the information produced by the Institution. The Code sets out 17 key principles and 80 good practices, concerning the institutional environment and coordination, and the statistical processes and products. In 2016, the Institute underwent an external auditing to assess its compliance with the principles and practices of the Code of Practice, which resulted in a set of recommendations that led to its revision and improvement. One recommendation was to develop and incorporate explanatory notes that clearly outline the context and objective of each good practice, in order to reduce the possibilities of different interpretations. Furthermore, a set of measurable criteria (quantitative or qualitative indicators) should be developed and incorporated to the Code as a means of assessing compliance with each practice. This paper presents the steps developed, the external references and frameworks used, and the choices that were made during the revision process. The IBGE Statistics Code of Practice is also seen as a mechanism to introduce a common understanding of quality across the producers of official statistics in Brazil and align national practices with international standards. The new edition of the Code will provide the basis for a more comprehensive version to be discussed with other institutions responsible for the production of official statistics at the National Statistical System.