Logo International Association for Official Statistics

Improving Official Statistics

IAOS Special Meeting on Fundamental Principles, Population Censuses and Misuse of Statistics

Monday, April 25, 2022

Krakow, Poland

On April 25, a day before the 18th IAOS Conference that took place on April 26-28, 2022 in Krakow, Poland, the International Association for Official Statistics (IAOS) organized a special meeting on Fundamental Principles, Population Censuses and Misuse of Statistics. The meeting was held in the Krakow office of the National Bank of Poland. Participation at the special meeting was by IAOS President’s invitation only and gathered a group of about 80 colleagues who actively participated in the discussions.

Opening of the meeting

Dr. Misha Belkindas, IAOS President, opened the meeting. Dr. Belkindas thanked participants for coming to the meeting and noted that it was good to see many friendly faces. He expressed the hope that the three topics: (i) Fundamental Principles of Official Statistics (FPOS), (ii) census in uncertain times, and (iii) use and misuse of statistics, are important and of interest. He noted that as many changes are seen in the statistical systems and in the surrounding environment, these topics become even more and more important and relevant. That is why together with the co-organizer of this event Dr. Hermann Habermann it was decided to include these topics in the program. Furthermore, Dr. Belkindas shared that he had invited “who is who” in official statistics to this meeting to get wisdom of the colleagues on these important topics.

Dr. Belkindas said that it was his pleasure to introduce Dr. Dominik Rozkrut, President of Statistics Poland and his successor as President of IAOS. He was also happy to introduce Ms. Agnieszka Szlubowska, Director of Statistical Office in Krakow as the host of this special meeting.

The meeting included three sessions:

  1. “Fundamental Principles – history, relevance, and the future”
  2. “Conducting Census in uncertain times and usage of results”
  3. “Use, misuse, and trust in statistics”

Opening of the first session “Fundamental Principles – history, relevance, and the future”

The session was chaired by Dr. Misha Belkindas. In his opening remarks for the session Dr. Belkindas provided an overview of the history of the FPOS and outlined each of the ten FPOS.

In the late 1980s, countries transitioning from centrally planned economies to market economies found that their statistical systems needed reform.

In 1990, the Conferences of European Statisticians (CES) determined to prepare a document on fundamental principles to guide statisticians, particularly in the transition.

These conversations were Influenced by the ISI Declaration on Professional Ethics (1985), the American Statistical Association’s Ethical Guidelines for Statistical Practice (1983), and Association des Administrateurs de l’INSEE Code de Déontologie Statistique (1984).

Poland led a working group for the CES, with the participation of Bulgaria, France, Romania, Spain, Switzerland, Turkey, Eurostat and the International Statistical Institute (ISI) that held three meetings in 1990-1991 to draft principles.

In 1991, the Conference of European Statisticians drafted the Fundamental Principles of Official Statistics (FPOS).

These were approved at the Ministerial level of the UN Economic Commission for Europe (UNECE) in 1992.

These principles were circulated among other UN regional agencies where they were generally embraced.

The Fundamental Principles were adopted (with a revised preamble) by the UN Statistical Commission in 1994.

In 2011, after revising the preamble, the UN Statistical Commission endorsed the UNFPOS.

In 2013, the UNFPOS were adopted by the UNECOSOC.

In 2014, the UNFPOS were adopted by the UN General Assembly.

The Fundamental Principles of Official Statistics were adopted by the United Nations Statistical Commission at its special session held from April 11 to 15, 1994.

The program for the first session included:

History of the FPOS


  • Jozef Olenski – former President GUS, professor Lazarski University (regrets)
  • Jan Louis Bodin – former President of ISI and IAOS
  • Hermann Haberman – former Chief statistician of USA, former Director UNSD

Mr. Jan-Louis Bodin delivered a presentation on the history of the FPOS.

Dr. Hermann Haberman gave a talk, which had two main themes. The first theme discussed the position of the United States statistical system during the creation of the FPOS and how that position changed over time. Dr. Habermann noted that Initially while the United States of America was in favor of the fundamental principles it did not see them as critical to the integrity of its own official statistical system. The current socioeconomic and political environment has changed that position and their critical nature is now apparent. The second theme discussed two changes that need to be made in the fundamental principles. These include recognizing the importance of the private sector in producing statistical information and, therefore, the need to incorporate the ethical policies of the private sector into the fundamental principles. The second change was the need of the fundamental principles to recognize that official statisticians were being treated in a hostile manner similar to investigative journalists by political administrations in an increasing number of countries and the need to adapt the fundamental principles to reflect this.

View from UNSD


  • Stefan Schweinfest – Director, UNSD

What future holds


  • Gabriella Vukovich – President, Hungarian Central Statistical Office
  • Gero Carletto – Manager, Development Data Group, World Bank
  • Jon Clifton – Global Managing Partner, Gallup
  • Ola Awad – President, Palestinian Bureau of Statistics
  • Rafael Diez de Medina – Director, Statistics Department, International Labor Organization (ILO)
  • Gerry Clancy – GIS Executive Consultant, Esri

Dr. Gero Carletto gave a presentation on the future of FPOS, which is summarized below.

The development of the Fundamental Principles of Official Statistics has been a powerful instrument for grounding and influencing discourse around official statistics and guiding the international statistical community. Moreover, its evolution and uptake has been in alignment with changes in national statistical systems over the years. While most of the fundamental principles still apply today, as we move towards integrated national data systems, we may require complementary sets of universal principles that cater to the current spectrum of instruments and actors in the data landscape.

The rise of data from private actors

One of the new elements of today’s data landscape is the profusion of data from private actors, including mobile phone data, social media data, satellite data, and much more. As these new data sources and producers play an increasingly important role in data systems, we must consider their differing purposes and incentives, and whether they can be governed by the same fundamental principles. Ultimately, a common denominator and the right incentives are needed in order for private actors to be bound by similar rules and principles.

Ensuring compliance through better documentation

The Fundamental Principles of Official Statistics are not legally binding; instead, they represent a mindset and approach to statistics. As such, it can take time for data producers to adjust. Moreover, compliance is voluntary, raising questions on how compliance can be increased. One option is to be more systematic at documenting examples of country alignment and/or non-compliance with the fundamental principles. In some cases, for example, data privacy has been unduly overexploited by some countries to limit data access. Documenting the lack of egregious case of abuse of data access policies and their consequences could help persuade countries towards more open data policies. Conversely, documenting how countries have dealt with cases of misuse of data could be helpful. International professional associations like the ISI can play a role in these documentation efforts. At the same time, we must also recognize and resolve potential frictions between the global fundamental principles and domestic priorities and laws.

Turning unofficial data into official statistics

Given the plethora of new types of data sources, many countries are looking to adopt unofficial data into their official statistics, which will require new quality assurance frameworks which are currently lacking. Systematic validation based on rigorous testing is critical for addressing several sources of bias which are a key concern with most new data sources. Several national statistical offices (NSOs) in high-income countries have created departments of experimental statistics, but scalability within and across countries has been difficult to achieve. Meanwhile, testing and applications of new data sources are much more limited in low-income and lower-middle-income settings. This presents a risk of exacerbating existing inequities, as low-income countries may find it more difficult to adopt and meet the quality standards of new data sources in their integrated national data systems. Ensuring an equitable transformation will require building capacity for national statistical offices to mainstream new data sources into statistical production, supported by international support and financing (such as the Global Data Facility and other funding mechanisms).

Dr. Ola Awad’s presentation is also summarized below.

Since over 25 years, the Fundamental Principles for Official Statistics serve as the most important foundation for the quality and integrity of official statistics. The UNFPOS provides a reference frame for worldwide, national, and international statistical offices to deliver independent, authoritative, and high-quality statistics. The national statistical systems of almost all countries have adopted these principles, translated them into the national language. Those principles are commonly used as the framework for questioning the quality of the national and international statistical systems and formulating improvement suggestions.

In the past few years, NSOs have succeeded in making national governments, national politicians, and other institutions realize the importance of UNFPOS in producing high-quality official statistics and over time, the UNFPOS have grown to be a well-recognized framework within most traditional worlds of official statistics. They were originally developed by the official statistical community as a self-regulatory system.

In the changing world of big data and in a post-pandemic world, adjustments are vital for official statistics to avoid becoming obsolete with the UNFPOS, since the context of the institutional framework of official statistics, the policy-making environment of ministries and central government on a national scale, has changed where they might have taken note of these principles but seems not to have committed to it.

NSOs will need a stronger communication role, partnerships with private sector mainly, and finding new partners. NSOs will require capacity-building skills in the domain of leadership and data systems management, rather than statistical topics.

The core methodologies which work today will not work in the future, NSOs must expand data sources to include big data, open data, and data science tools to utilize statistical projects (data mining, artificial intelligence, etc.). NSOs must ensure data interoperability to make the best use of administrative data, especially in developing countries where admin data sources are still fragile and in need of capacity building.

These changes will require an update to the fundamental principles while ensuring their implementation; the principles must keep up with recent developments in the world of data. An update in fundamental principles would mean an update in the code of practice and eventually the statistical law.

NSOs should be conscious of adopting big data into official statistics to take advantage of these innovative sources, including their application to the monitoring and reporting on the SDGs. However, we should be careful and we need to adequately address issues pertaining to methodology, representativeness, quality, technology, data access, legislation, privacy, management, and finance, and provide adequate cost-benefit analyses.

A summary of the presentation by Mr. Rafael Diez de Medina is noted below.

After so many years of having the Fundamental Principles of Official Statistics guiding our daily work at the national and international levels and having reached consensus around the 10 pillars which the international community decided to embrace to move forward, we can take a break and see how real the compliance of these principles have been, how fit they have been to adapt to the new shocks to the new challenges, which hectically have been popping up.

We surely will conclude that the FPOS have been quite instrumental and have set the basic rules under which the so-called “data revolution” is functioning. An era of dramatic changes: irruption of new statistical sources, new blurred boundaries between official and non-official statistics, dramatic challenges to the notion of State itself, an unexpected re-definition of what truth is (many observers refer to a new era of “post-truth”, of “fake news”, of “alternative facts”), the irruption of social networks as main sources of data, often challenging even science or academic knowledge. Who could have envisaged all this 30 years ago?

In this storm we are now navigating all of us, the FPOS are still guiding us for the future and their content is still valid, as it was even before they were formally approved.

Mr. Rafael Diez de Medina shared an interesting story from the origin of the organization he is coming from, the International Labour Organization (ILO). It was created by the Treaty of Versailles in 1919, the Peace Treaty which ended WWI. In the same Peace Treaty, in its article 396, the international community agreed on the need to establish cooperation to move in key international statistical standards on a set of variables which were judged key for peace and democracy during those very turbulent times: labour issues. There was an acknowledgement that the world would need to measure key aspects on employment, unemployment, wages, occupations, household budgets, prices, purchasing power, accidents at work, hours of work, labour rights, among many others. It was openly agreed that this would require consensus, discussions, harmonization, and cooperation to avoid misuse and lack of trust between social partners and promote social justice.

Underlying this political decision which gave birth to the International Conference of Labour Statisticians in 1923, it was implicit the strong link between democracy, trust, and social peace. It was acknowledged the need to be transparent in how these key variables would be measured to allow fair play in collective bargaining and wage negotiations. It was clear the need to avoid misuse and look for harmonization between countries based on solid international and national cooperation between producers and users of statistical data. This was a decision made 100 years ago and we can see that it contains many of the 10 FPOS. With a distinct feature: while it started by referring to official statistics, since the ILO is an organization which is not only driven by member states, but also by employers’ and workers’ organizations, it also called for these actors to be included in this new “code of conduct” which was being created. The ICLS, one of the oldest standard-setting mechanisms in statistics, also convenes other stakeholders beyond the State: social partners who are actively involved in the definition of the variables in labour markets. This is happening still now, after 100 years.

This external link, beyond the official spheres, is topical today where the boundaries of official statistical sources and non-official statistics are not so clear as it used to be in the past, and the irruption of new sources of data, like big data, calls for other actors to be sheltered under the same umbrella of the FPOS: mass media, social media, platforms, technology companies, private stakeholders, NGOs, citizens in general. Many of the challenges we are now discussing for the future of the FPOS were confronted by us in the ILO precisely because of our intrinsic mixed composition. It has not been easy, but the final output turns out to be extremely useful and rich. We must see this openness to other stakeholders, beyond official statisticians, as extremely rich and powerful.

It is true that both official and non-official statisticians have been equally confronted with uneasy situations lately. We all have witnessed serious attacks to the FPOS, attacks to autonomy of NSOs by politicians, challenges to professional independence, disrespect for quality and misuse of statistics by policy makers or important social actors. We also have seen failures in how the international system supports the national statistical offices, through donors whose goals are often independent of national interests, we have witnessed even personal attacks to serious Chief statisticians because they were fully supporting the FPOS and often, we have witnessed silent resignations or “early” retirements which were based on these attacks.

In France, last week [the week of April 18, 2022] during the last debate of main candidates to the Presidency, statistical standards were explicitly challenged attacking a solid system of measuring unemployment only for very immediate political goals. With a couple of sentences, in just a second, a consolidated solid system was challenged, and doubts were cast to the citizen, creating tension with strong democratic values… We heard this from France last week, but it was the same in the USA, during campaign in Brazil, just to mention the more recent ones. By attacking the FPOS we see how democratic values are eroded and this should drive us all to defend tirelessly their compliance and consolidate international mechanisms to watch for them.

This “code of conduct” which guides us has provided us with many tools and support to face these challenges. We have promoted open data, we have increased the institutional capacity of NSOs and NSSs through training and institutional support, we have collectively managed to define an impossible task: the SDG Monitoring system. We are constantly reviewing statistical methods to cope with urgent matters as it was shown during the COVID-19 pandemic but also many dramatic changes in economic and social behaviours which require new ways of thinking.

In the case of the ILO, our tripartite constituents are actively promoting new forms of measurement for the emerging new forms of work, we have jointly worked on how to deal with platform workers and their special nature, how to define informal work, how to deal with dramatic changes in the occupations and how to implement new ways of defining classifications… However, the FPOS are still completely valid. Now we must extend them de facto to other concepts: how to deal with AI, big data, algorithms, the role of media, the need to move strongly with statistical literacy to avoid attacks to democratic values, promote trust and rely on truth.

We all need to adapt our different domains in statistics to changing needs from users.

Mr. Rafael Diez de Medina wanted to finish with an optimistic idea. Now the scope of official statistics is broadening, and many are embracing this. The richer the emerging data ecosystem is, the better will be for us all, and for the citizens. More checks and balances will be in place. Strong legal frameworks are still key for this emerging trend, but the need to intrude in statistical domains which call for many different stakeholders to provide their opinion (like the case of the environment) will surely assure more transparency and stronger resilience against future attacks to the FPOS.

Mr. Rafael Diez de Medina concluded his presentation by quoting Prof. Corrado Gini, famous professor for his Index, but least known as Vice-President of the 3rd International Conference of Labour Statisticians (ICLS) in 1926, which in Mr. Rafael Diez de Medina’s view summarizes well how the FPOS are to be used: “Statistics is not pessimist or optimistic, it should not serve to support preconceived thesis. Its guiding principle should seek the truth, whether it is to be liked or not. Only if this is accomplished, statistics will have a highly-appreciated educational function and be moralising for individuals and for the people”.

Ms. Gabriella Vukovich and Mr. Jon Clifton also shared their views on the future of the FPOS.

Ms. Gerry Clancy’s presentation concluded the list of presentation for the first session.

Discussion and summing up

Ms. Ada van Krimpen, ISI Director1, noted the fact that this special meeting is attended by those who were involved in the birth of the FPOS; however, it is important to think about who will take over this important project in the future, i.e. the next generation. It is also important to find the means of communicating the FPOS to users of official statistics. The communication which is currently being done by UNECE, by having every week another NSO promoting one of the principles on Twitter, is a good example. She noted that after inquiring among the special meeting attendees, she found out that only six people or so used Twitter. She concluded that if the statistical community wants to keep the FPOS alive and kicking it is necessary to create the means to reach out to the younger generation.

Ms. Angela Me, Chief of Research and Trend Analysis Branch at UNODC, elaborated on the question related to the meaning of official statistics as with the proliferation of sources and actors engaged in government statistics. She noted that “official” may have lost its clear meaning. With the growing mis-trust from the public on some government institutions, the term “official” may actually loose traction as synonym of impartiality and high quality. Based on Ms. Me’s intervention, alternative terminology such as public good statistics were discussed to emphasize that it is not who produce statistics that makes it a good output but it is ultimately the adherence to the Fundamental Principles of Official Statistics for the benefits of all that makes the high standards. Ms. Me added that the notion of public good statistics goes beyond a simple change in terminology. It shifts the ownership from NSO and government institutions to a shared ownership of a common good that encompasses a wider multitude of actors as contributors to public good statistical systems with the NSO taking the role of coordinator and guarantee of the quality and ethical standards of what constitute the shared public good statistics (that is larger than statistics for public good).       

Participants in the discussion emphasized that statistics has today become transnational in its nature and that protecting the quality of national statistics is not enough for a single country to provide quality data to its citizens. The declaration and adoption of the Fundamental Principles of Official Statistics alone is not anymore sufficient to ensure impartial statistics and the international community could think about stronger systems to protect national and international statistics. As done in other fields, an international data convention could provide a system of incentives for complying to the principles and disincentives for non-compliance.   

Dr. Denise Lievesley, Honorary Fellow of Green Templeton College, University of Oxford, expressed satisfaction that attention was being paid to the application of the Fundamental Principles to international agencies as well as NSOs as statisticians in such international agencies can be put under pressure to distort data, particularly if their employer is concerned with advocacy and wishes to maximise the impact of the statistics.  She recognised that inappropriate use of data at the international level, including of course its manipulation, can undermine the integrity of national data. 

Ms. Irena Krizman, former Vice-President of ISI, former Director –General of the Statistical Office of Republic of Slovenia (2003-2013), commented concerning the need to revise or to update the current FPOS. She emphasized the importance the FPOS has had in the development of Slovenian national statistics. In the nineties it served as a strong reference point when preparing and implementing the statistical legislation and the organisation of the National Statistical System. In Ms. Krizman’s opinion it will not be helpful to start changing ten principles but rather leave them fixed as “ten commandments”. But what the statistical community needs to keep doing is to better promote FPOS and DPE in the wider data ecosystem including the private sector, making stronger link between the ISI Code on the Professional Ethics (principles related to individuals) and FPOS (principles related to organisations), agree on the roles of different players (e.g., UN, ISI, NSOs would have in implementing the principles but also about the process on who and how to react when the principles are not respected).

Mr. Mario Palma, former IAOS President and former Governing Board of INEGI, stressed the importance of the legal status of the FPOS in their future implementation and improvement. The legal status can be considered from local, regional and international perspectives. At local and regional levels, NSOs should promote their adoption by their legal systems and in regional agreements. At the international level, the official statistics community – notwithstanding the inherent difficulties of attaining the formal status of international law – should promote and emphasize the fact of their unanimous adoption both by the Economic and Social Council and by the General Assembly of the UN, in particular, that they prescribe principles and their legal adoption not only to statisticians and NSOs but that they are also addressed to governments at all political levels.

Mr. Qasem Alzoubi, IAOS Executive Committee member, stressed the importance of the independence of the National Statistical System (NSS) according to the FPOS especially in developing counties. As NSSs in many counties lack of independence, low quality of information, inadequate response to user needs and weak adherence to ethical standards of transparency and clear methodologies for collecting and analysing data, especially in the time of COVID-19. He appreciated the role of the IAOS and the Executive Committee for sending a strong statement to Turkey and Russian governments. 

Mr. Stephen Penneck, ISI President, noted that the Fundamental Principles, as principle applying to official statistics, do not, in his view, need much revision. They are high level and generic enough to still be relevant. He emphasized that the priority now should be on implementation and assessment.

But as Dr. Habermann said at the beginning, Mr. Penneck concurred that the context has changed, and we have concerns about other data and statistics (alterative facts) which are part of the landscape and for part of the public debate, Stephen added. He stressed that the FPOS have already changed once, from principles owned by statisticians to principles owned by governments when they were adopted by the General Assembly in 2014. Do they need to change again, to become Fundamental Principles of statistics and data used in public debate (or to inform public decisions)? What might these look like? And how do we get societal consensus on these to get agreement. There is merit in exploring a wider set of principles that are more relevant to current concerns. These principles might develop by using the existing FPOS as a starting point, but Stephen thinks it might be more helpful to do some initial work unconstrained by these.

Mr. Vilas Mandlekar, Consultant, World Bank, made the following two points on the FPOS:

First, Access to official statistics should be made as obligatory as may be possible for the government. Perhaps, the FPOS includes this point either explicitly or implicitly. However, given the technological advances and moves toward democratization of data and metadata, the younger generation of leaders in research, academia, private sector, international organizations, and yes the government officialdom, would need, as well as, insist on access to data in different forms and formats. Technology would enable such access at the press of a button, now and more so in the future.

Second, make more and more data “Open” that falls under the “Official Statistics” banner. New categories of data, particularly geo-spatial and social media to name a couple, provide far more insight than the prevalent forms in which data is disseminated. These two types of data have a tremendous impact in understanding a few of the major challenges of our times – climate change, environment, and understanding public sentiment. Needless to say that making data “Open” would have to be balanced reasonably, by governments, within the constraints of data confidentiality and privacy.

Mr. Mandlekar summarized that his message here is to make official statistics more open and easily accessible using technology.

Dr. Misha Belkindas shared his thoughts and summed up the session “Fundamental Principles – history, relevance, and the future” as follows:

After the fall of the Berlin Wall, to meet the challenges created by the transition to market economies and democracy, Eastern and Central Europe Countries needed new references and new landmarks for official statistics, not so different from a strictly technical point of view, but totally different where the concept itself of the role of statistics in society was concerned. The result was the adoption in June 1991, by the Conference of European Statisticians, of the FPOS which later proved to be of universal value.
The UNFOP for NSOs encompasses: Ethical Spirit, Credibility, Identity and Trust. It is our protection especially under difficult circumstances. Nevertheless, and to keep up with the transformation of data and statistics eco-system, we need to revise and improve the UNFOP to introduce the new areas of work such as: big data, modernization, data stewardship, digitization, etc.

Broadening the scope of the FPOS to other actors beyond official statistics can be an enriching experience and can provide us all with more tools to react to many attacks to them by some political systems, creating the basis for their full compliance. For example, the ILO has passed through an early process of engaging social partners in the actual statistical definitions and, though difficult, it proved to be resilient and fit for purpose for new emerging challenges in the world of work, but this could also be extended to comply with the 10 FPs in other domains.

Adding a “geo” component to your official statistical data is critical to achieving broader efficiencies in data collection and wider use of the data. Achieving a Geospatial Infrastructure for Official Statistics is possible and achievable

The Fundamental Principles are (1) more universal (2) more needed and (3) more fundamental than ever. The best 30-year birthday present for the principles would be a collective and broadened understanding of just HOW fundamental and foundational they are and how the grand the edifice of statistics, data and information is that they are supporting.

1 Ms. Ada van Krimpen retired from the position of ISI Director in July 2022.

Opening of the second session “Conducting Census in uncertain times and usage of results”

The session was chaired by Dr. Hermann Habermann.

The program for the second session included:

Overview of the 2022 Census round


  • Jean Michel Durr – Consultant, former Project Director, INSEE

Mr. Jean Michel Durr’s presentation focused on the global situation of the 2020 round of Population and Housing Censuses.

Views from Countries


  • INSEE France – Jean Michel Durr, former Project Director
  • Statistics Poland – Dominik Rozkrut, President
  • CBS Israel – Danny Pfeffermann, National Statistician and CBS Director

Mr. Jean Michel Durr, Mr. Dominik Rozkrut, and Dr. Danny Pfeffermann shared France, Poland and Israel’s experience in census taking.

Discussion and summing up

Dr. Hermann Habermann summed up the discussion and concluded the second session “Conducting Census in uncertain times and usage of results”.

Opening of the third session “Use, misuse, and trust in statistics”

The session was chaired by Dr. Misha Belkindas.

The program for the third session included:

Framework for IAOS work


  • Martine Durand – former OECD Chief Statistician and Member of the French Statistical Authority
  • and Jan Robert Suesser – Member of IAOS Executive Committee and Member of ISI Advisory Board on Ethics (ISI ABE), IAOS Reference group co-leads

Ms. Martine Durand and Mr. Jan Robert Suesser’s presentation focused on misuse, governance, trust and ethics: enhancing official statistics’ capacity to address these issues, and next steps on this matter.

Links with work of the ISI Advisory Board on Ethics


  • Walter Radermacher – Retired Director-General, Eurostat, President of Federation of European National Statistical Societies, Chair of ISI Advisory Board on Ethics

In his presentation, Dr. Walter Radermacher elaborated the issue of integrity of official statistics.

Discussion and suggestions for the future IAOS work in this area

Dr. Denise Lievesley noted that statisticians and responsible journalists share many common aims as they seek to establish the ‘truth’ and to empower through the use of information.  They can be exposed as they may uncover information which is uncomfortable for those in power.  Thus it would be good to establish closer links between statisticians and journalists. Dr. Lievesley added that Incidentally this article which she has just come across https://blogs.scientificamerican.com/observations/what-journalists-and-scientists-have-in-common/  is on this theme. Furthermore, she mentioned that having said we seek the truth we need to acknowledge that we rarely are confident enough to produce one answer, and so need to think about how we can communicate uncertainty without undermining trust in the data.  We also recognise that our data is an artefact of the data collection methodologies, and that assumptions are built into the data when we apply models or forecasts (as many statisticians have done in relation to COVID-19 predictions).

Stephen Penneck, ISI President, shared a follow-up intervention to the discussion. He noted that Denise’s point about ethical issues in academia reminds him of a discussion he had with the academic members of the ISI ABE a few years back as to why the ISI did not get cases from academics/ students referred to us.  All ISI activity seemed to be in official statistics, which could not be right.  Their answer was that the universities have very good ethics procedures, committees, etc., and are well able to cope with all of this.  Stephen doesn’t think this means that there are not questions that the ISI should discuss, and it might be good if the academic members of the ABE could be charged with proposing a session in the ISI Congress in July 2023 in Ottawa.

The main point Stephen would like to raise is the narrowness of the remit of what we are now calling the Krakow Group [the reference group], in two respects.

Firstly, Stephen doesn’t recall the exact words on the slide that Jan-Robert proposed at the end of the session, but his recollection was that it was too official statistics focused. From the first session discussion it seemed to him that the main issue on FPOS for official statisticians is compliance.  He added that it needs to be ensured that there is the best possible strategy for this before the next round of assessments in 2024.  And this activity should build on the excellent work done at the UNSC 2020, where an additional chapter on the Implementation Guidelines was agreed.  Stephen concluded that in his view this would be useful work for the Krakow Group.

Secondly, he thinks that the first session established that our main concern was the impact of other actors who produce data/ statistics relevant to public policy which has questionable provenance and hence quality.  Meeting participants talked about the possibility of generating a new set of Fundamental Principles which would apply to all statistics used for public decision making, and the attendant problems of how this could be effective.  Stephen thinks this could also be useful work for the Krakow Group.  It would be good if media experts and fact checking experts are added in the discussion.

Dr. Misha Belkindas outlined the suggested way forward for the IAOS in the area of use, misuse, and trust in statistics. He noted that the IAOS will create a reference group headed by Martine Durand and Jan Robert Suesser. The Terms of Reference is of a different scope than the ISI ABE. Dr. Belkindas added that he was happy that Dr. Walter Radermacher agreed to join the group. He concluded that in the next few months the IAOS will approach colleagues asking to join the reference group.

Closing the session and closing the day

Dr. Misha Belkindas and Dr. Hermann Habermann thanked the meeting participants and wished all a pleasant reception and a successful conference.