Interview with Reimund Mink, about his book ‘Official Statistics – A Plaything of Politics? On the interaction of Politics, Official Statistics, and Ethical Principles’¹

1 1

Ivo Havinga	Reimund Mink

Ivo Havinga: Dear Reimund, could you tell us more about yourself, your career, and the relation between your career and the writing of this book?

Reimund Mink: I have had the good fortune to deal with questions of statistics and economics throughout my professional career. Important stations in my life were my childhood and school years near Limburg, a city between Frankfurt and Cologne, my studies in economics and my time as an assistant at the University of Giessen as well as professional activities at the Federal Statistical Office of Germany (Destatis), the Deutsche Bundesbank, the International Monetary Fund as well as the European Monetary Institute and the European Central Bank.

So, what contributed to writing this book? On the one hand, it was certainly the many positive experiences of cooperation between central banks, government institutions and statistical offices in Europe and at the international level. The many discussions in Europe surrounding the accession negotiations to the European Economic and Monetary Union were certainly formative. I can still remember the visits to Greece, where I, as a representative of the European Central Bank, discussed the many inadequacies in the country’s statistics together with the European Commission and helped with the development of new statistics. But there were also the heated debates about France Telecom, the correct treatment of gold reserves in the accounting system of the Deutsche Bundesbank or the ambitious attempts of Vienna to show a surplus of a positive zero in the Austrian government accounts despite many objections.22

There were the often-impressive meetings and laborious discussions in Greece about the adequate approach to certain reform projects. The tendency not to implement legal requirements in practice was clearly noticeable. Proposals to improve cooperation between the Central Bank, the statistical office (Elstat) and the Ministry of Finance through a well-intentioned Memorandum of Understanding were ultimately not successful.33 The tensions and animosities between the individual institutions were too great. Andreas Georgiou’s difficult professional situation played a special role in this.44

My visits to the Statistical Office of Kazakhstan (Kazstat) were similarly fraught with tension as the work in Greece. Here I felt particularly acutely the “pressure to succeed” that the middle management staff was under in implementing the project goal of “Strengthening the National Statistical System of Kazakhstan”.55 Here, too, the attempts to improve cooperation within the Office and between the Office, the ministries and the Central Bank could hardly be realized.

To conclude I had the opportunity to work in so many European and international projects having lived up to the claim of the United Nations’ Fundamental Principles of Official Statistics (UNFPOS) to create statistics as an “indispensable element in the information system of a democratic society”. Work did not always have a positive outcome. Sisyphus was often the inspiration. The book, therefore, provides the many facets that make statistical work particularly appealing. Often there are limits to be fathomed. But the book also shows the difficulties of why this work is rarely crowned with success.

Ivo: With what intention did you write your book on “Statistics – a plaything of politics”?

Reimund: That is a complex question. It is mainly the experiences I had while working in Greece during the sovereign debt crisis that led to the writing of the book. I have often asked myself: How is it possible that in a country that is repeatedly described as the cradle of the European idea of freedom and democracy, conditions prevailed that were incompatible with a democratic social order and with the basic rules of freedom? Therefore, lessons should be learned from this and other documented cases of political influence in the professional ethics of statisticians to identify unacceptable and unethical political interference.

The fate of Andreas Georgiou, the first head of an independent statistical office in Greece, is still in my mind: I am saddened by the accusations and the unfair treatment of a fundamentally decent person who carried out his professional duties with the utmost care

and without scruples. Petros Markaris, the well-known Greek writer, fictionalized this situation in one of his crime novels, “Times of Hypocrisy”. There, a high official of the Greek Statistical Office falls victim to a murder because he allegedly spreads “falsified” data.66

It was only through a deeper understanding of “Greek conditions” that I gradually became aware of the many shortcomings in the government system, which includes official statistics. There, all criteria for “good” statistics were violated. In business accounting, we speak of the principles of proper bookkeeping and accounting, and by this we mean the principles of accuracy, clarity, and continuity. In fact, these principles should apply to official statistics – and beyond that to all areas of statistical work.

I had similar, but less dramatic experiences in other countries. Statistical work was often subject to strong political influences. Its organizational incorporation is usually of secondary importance. The ideas of the “Meltzer Commission” of more than 20 years ago to regard statistics and thus the provision of data of the member countries within the framework of the Special Data Dissemination Standard (SDDS) as one of the central tasks of the propagated “new IMF” was often only accepted in the industrialized countries.77 Elsewhere this could hardly be implemented in the desired form. This is because politicians are often not prepared to give statistics an appropriate and independent place as a central information system of government agencies. Often the financial resources are not particularly lavish. Projects by international institutions are gladly accepted but are not followed up with appropriate care after completion. Intra- and inter-agency cooperation, which is indispensable for statistical work, often leaves much to be desired. Under conditions in which unfavorable political conditions arise, basic democratic rules are no longer observed and ethical principles are undermined, situations arise as I have described them drastically in the book for different countries and in different time periods.

Ivo: It was first published in German and only later in English, why this?

Reimund: This is because I, together with various enthusiasts of national accounts, have been in very close contact with Hubert Hoffmann, the managing director of Metropolis Verlag in Marburg, for many years. This publishing house has been issuing the papers for the Berlin Colloquia on National Accounts since 1996. It was only a few years after reunification when the first colloquium took place in May 1995 in Berlin-Köpenick, in the “Wendenschloss”. However, the name “Wendenschloss” has nothing to do with the “Wende” in 1989 (the dissolution of the GDR) but refers to Köpenick’s Slavic (Wendish) past. Theodor Fontane described this area wonderfully in his popular “Wanderungen durch die Mark Brandenburg” (“Walks through the Mark Brandenburg”).88 So far, by my count, nine volumes have been published on it, the last on “Preisbereinigung und Realwirtschaft” (“Price adjustment and real economy”), edited by Klaus Voy and myself in 2019 – still shortly before Corona broke out. Now, after a break of more than three years, the 13th Berlin VGR Colloquium took place again in June 2023.

Through these contacts, I approached Mr Hoffmann about publishing my book. It then appeared in 2020 – two years after the release of “A Greek Tragedy – Sovereign Debt Crisis and No End?” by the Metropolis Verlag.99 Several colleagues from my professional environment found the treatise on ethical questions in relation to official statistics and politics interesting and exciting. They saw the broad treatment of the topic by describing episodes beyond Europe as quite interesting for a wider readership. Finally, they encouraged me – not least my former colleague Werner Bier – to disseminate the book also in English. I found an open ear at Springer Verlag and Hubert Hoffmann also had no objections to an English edition. However, the conversion of the German text into the format specified by the publisher proved to be quite time-consuming. Corona was also a stumbling block, despite the progress of digitalization, so the English edition did not come onto the market until the beginning of 2023.

Ivo: The book not only covers a wide range of topics related to the organisation of official statistics, but also has a strong historical character. Which target group is likely to read/use your book?

Reimund: Indeed, the reader might think that much of what is described in the book is history, has already been said and no longer has any relevance for today. But don’t the events of recent years teach us that history repeats itself – after the pandemic, the war, inflation? My conclusion from this is that the historical perspective does not really play the decisive role, because the ethical questions raised in the book in the interplay between statistical work and political will always arise in all places. Thus, I assume that through the episodes that took place in different countries, under different political conditions and at different times, a better understanding of how to approach today’s issues in statistics, ethics and politics can be developed.

Of course, it is important to keep the potential readership in mind and to confront them with the issues raised in the book. These readers are first and foremost clients who deal regularly with statistical data – be it as producers or as consumers of such data. Ethical questions may be tackled by mutual agreement based on democratic principles. As Horst Siebert once described it, ethical norms represent for everybody an inner gyroscope according to which decisions are taken.

But how do I behave in situations where ethically sound decisions are hindered or, in the worst case, prevented? In other words, have rules been established that allow staff to appeal, protest and possibly reverse decisions that have already been made in the case of obvious breaches of the rules? Following on from my experiences, I became increasingly aware of the tensions that can arise as a statistician. On the one hand, statistical action should be guided by the UNFPOS. I refer here primarily to the rule of professional ethics and independence to withstand political pressure and to the rule of using scientific statistical standards to guide one’s statistical work. On the other hand, there is often resistance in statistical practice. Occasionally, one finds oneself in a defensive position in which one must justify one’s own behavior. Pressure is also built up by users of the data. Thus, it is often heard that statistics fail to provide the data "urgently needed" for theory and policy with the desired precision and speed. This was evident, for example, in the case of Ireland’s miraculous economic growth a few years ago as described in Chapter 10 of my book. The restructuring of internal processes in some large multinationals led to a jump in the level of real GDP that had little to do with the ’physical’ growth of the Irish economy. Politicians and some academics ridiculed the statistics, claiming that they were incapable of adequately assessing these effects.

So, to return to the question of readership, I would think that the book could be useful both to people interested in statistical and ethical issues, and to people who are more critical of official statistics.

Ivo: In Chapter 4 of your book, you emphasize three main principles that characterize the professional ethics of a statistician: A statistician should (1) apply sound technical practices and internationally agreed standards in the collection, processing, and presentation of statistical data, (2) ensure the confidentiality of data, and (3) maintain the integrity of the national statistical system. Can you elaborate a little on the content of these principles?

Reimund: At first glance, the application of sound technical procedures and internationally agreed standards may not seem to have anything to do with ethical considerations. Nevertheless, their non-application over long periods of time can lead to situations where staff of a statistical office stick to a weak or outdated method, although its shortcomings are obvious and alternative approaches are already available. However, these new approaches are not applied for various reasons (e.g., political pressure on an office). Old and therefore often flawed methods thus remain in use for decades. This affects the quality of the calculation of important macroeconomic variables and is an indication of a possible ethical challenge. However, a well-organized statistical system benefits from technical procedures and statistical standards being constantly reviewed and renewed over time.

The origins of the concept of statistical confidentiality and protection from harm that may arise in statistical work can be compared to the Hippocratic Oath. The modern concept of statistical secrecy was developed in many countries in the post-war period. The tradition of protecting individual data sets and publishing only aggregated data is now reflected in the laws and practices of most countries, as well as in several international guidelines.

The third important cornerstone of ethical statistical work is to maintain and improve the integrity of statistical offices and the national statistical system. Attacks on this integrity can occur in a variety of ways, including arbitrary political manipulation of concepts, definitions, and the scope and timing of data publication. Certainly, the public, most politicians and scientists have a vital interest in preventing such manipulations. For the management and staff of a statistical office, an international organization and for national and international statistical societies, the effort to resist such threats becomes an ethical obligation.

Ivo: How would you formulate in five lines a piece of advice to IAOS members to read this book?

Reimund: This is not as easy as asking. If I am to present the whole thing for the IAOS members in five lines, the following issues seem important to me:

Ivo: Ethics seems to be an important binding element in the book. Can you explain how (ethically sound) statistical work can best be communicated to the public in times of fake news or distorted news? What is the role of theoretical and practical education and training in statistics?

Reimund: I think we must ask ourselves these questions more and more often because content generated by human intelligence and increasingly by artificial intelligence in the form of fake news is becoming part of everyday life in the media and on the net: This is sometimes funny and entertaining, but most alarming, for example when right-wing radicals create and publish pictures of aggressive “foreigners”. Currently, there is a flood of such computer-generated images, videos, and texts. Opinion polls are published that are based on statistical data. Hardly anyone asks who is behind the survey results, how high the number of respondents is or how large the margins of error might be. The question is: How can we detect disinformation better and faster? How do we protect ourselves, our peaceful coexistence, and our democracy in the digital age?

A saying by Seneca occasionally comes to mind. What he means is: “If a man knows not to which port he sails, no wind is favorable.” So I believe that the key to solving the problem is in the schools and universities, in learning and training, and in enlightening people. The key word is education to become responsible citizens in a democracy. There is often talk about improving the economic and the financial literacy among the population. Shouldn’t we also talk about improving “statistical literacy”? It is important that not only statisticians understand their statistics, but also the “interested layman” – despite the often highly scientific methods. Case studies could be used to show how, for example, a price index comes about, what methods are used to calculate and what problems arise over time due to changes in the basket of goods. The foundations for this should already be laid in schools to make it easier to recognize the many pitfalls in the dissemination of falsified statistical data.

Ivo: The combination of topics (from happiness to national accounts and globalization to national egoism) you deal with is astonishing. Is there for you a binding factor between all these topics, or are they just random examples of historical/development and governance/ethics issues relevant to official statistics?

Reimund: “The true is the whole”, writes the German philosopher Georg Wilhelm Friedrich Hegel in his “Phenomenology of Spirit” from 1807. According to this, we should, if possible, also concern ourselves with things that are somewhat peripheral. In fact, we are dealing here with things that at first glance seem to have nothing to do with each other. In my opinion, however, these are topics where we are exploring the limits of official statistics or are already beginning to exceed them. Ethical questions are gaining importance.

Take the happiness or welfare of the human being or of a society and its measurability. As in economic theory, one concludes that happiness is not measurable. In Germany there is the happiness atlas, and in Bhutan gross national happiness. The index in Bhutan covers various areas such as health, education, mental well-being, cultural diversity, environment, and the availability of time. Is it because of the construction of the index that the Kingdom of Bhutan calls itself the happiest country in the world?

At international level, the OECD has been publishing a Better Life Index for years and the United Nations a World Happiness Report.1010 The question that arises for me is whether a structural connection can be recognized for the indicator systems proposed for this purpose so that they can then be combined into one variable. For there will not be an indicator system that can be consistently aggregated. This is not the case because the indicators are of different types – partly based on qualitative characteristics. This carries the risk that only the development of partial indicators are discussed in public, such as the assessment of life satisfaction. The dream of measuring the happiness of an individual or even an entire nation with the help of a single indicator always fails miserably.

Nevertheless, statistics should meet these challenges with new approaches and solutions. And wouldn’t much of the criticism of GDP fall away if it were possible to internalize “externalities” as much as possible? On the other hand, I think it is a big step forward that since the financial crisis ten years ago, efforts have been made to include data on income and wealth distribution more strongly in the analysis of national economies.

Another challenge is to capture globalization in an (national) accounting system that has set itself the task of measuring the economic activity of national economies – a contradiction in terms. If one wants to measure the effect of globalization, however, this requires intensive cooperation between statisticians – at the national as well as the international level. However, I see the danger that international cooperation will fall by the wayside due to the resurgence of national egoism instead of multilateralism in the economy and society worldwide – and statistics would be particularly affected here. This should also make it more difficult to solve the “problem” of globalization statistically. Added to this are the devastating effects on international relations of Russia’s invasion of Ukraine, the geostrategic decoupling of dependencies in supply chains of critical technology and materials. Not a particularly rosy outlook for international cooperation.

Ivo: What was the (your personal) reason for discussing specifically the cases of unethical behavior towards statisticians in Greece and Kazakhstan?

Reimund: The reason was my personal work experience. I have already spoken at length about Greece in a few places. In Greece, working relations between the Ministry of Finance, Elstat and the Bank of Greece were poisoned. Even a memorandum of cooperation did not lead to any significant improvement in working together. Ultimately, it was the initiatives of international organizations such as the IMF, the European Commission (Eurostat) and the European Central Bank that brought progress in cooperation and structural improvements – also in official statistics. I am particularly proud of the monthly government finance statistics, which are published in a timely manner by the Ministry of Finance and constitute one of the cornerstones of the Greek national accounts.

The situation in Kazakhstan was different. I had already advised the central bank in Almaty in the 1990s as a member of an IMF delegation. At that time, the aim was to establish an independent central bank – independent of the central office in Moscow and of the Kazakh government and parliament. It was quite a difficult time, as the know-how was lacking in many respects.

My visits 25 years later to Astana, renamed Nur-Sultan in 2019 in honour of the long-time president, showed an outwardly modern capital with imposing administrative buildings. Hidden from my view, however, was the poor cooperation of the statistical office with the ministries. The National Bank of Kazakhstan resided far away in Almaty. A meeting with the various institutions arranged during one of my visits only brought about a selective improvement in cooperation. Another striking feature was the hierarchical structure of the institution, which offered individual staff members hardly any scope for action. Mostly women were in middle management, while the leadership was largely in male hands.

In general, cooperation always left much to be desired – and apparently also the understanding to act according to ethical principles, as they have been recommended internationally for years. In such a climate of working together (or not working together), fates like those of Anar Meshimbayeva are possible.

Ivo: Why do you describe human dignity and human rights in detail in your book and how do they relate to the UNFPOS, professional independence by law and self-commitment through code of practice?

Reimund: Due to the catastrophic political developments in the first half of the 20th century, questions of human rights and human dignity came to the forefront of political interest after the Second World War. Just as the theory of human rights was already the answer to political absolutism in the 18th century, the concept of human dignity must be seen as the answer to the experience of absolute terror in the 20th century.

This was first effectively expressed in the Universal Declaration of Human Rights adopted by the UN General Assembly on 10 December 1948, which speaks of the “recognition of the inherent dignity and of the equal and inalienable rights of all members of the human family” and that “all human beings are born free and equal in dignity and rights”.1111 Even before that, in the UN Charter of 1945, the belief in the (originally: sanctity, but later) dignity of the human person had been affirmed. Human rights were “discovered” (and have applied ever since) to protect human dignity. And since human dignity consists in personal freedom, human rights represent the globally codified form of protection of this freedom.

It is indeed difficult to understand why it has taken so long to establish the UNFPOS and other codes of conduct on official statistics and then link them to the idea of human rights. The setting up of these principles and codes in early 1990s followed the fall of the Berlin wall and the creation of new East-European democratic countries and of the supporting systems for official statistics. In 1992, the United Nations Economic Commission for Europe (UNECE) adopted the fundamental principles of official statistics in the UNECE region. The UN Statistical Commission adopted these principles in 1994 at the global level. Finally, the Economic and Social Council endorsed the UNFPOS in 2013; and in January 2014, they were adopted by General Assembly.

The development ten years ago seemed to be triggered by the discussion in Europe about the professional independence of official statistics around the events of the sovereign debt crisis in Greece. There, the discrepancy between the obligation of government agencies to provide reliable statistical data and the observed practice of manipulating data prior to their dissemination and publication was particularly serious. These problems in Greek government finance statistics have existed for a long time, but only came to the attention of the public in 2009, when it became clear that Greece’s government deficits and debt levels had been miscalculated for years. In fact, the means of the statistical office, which was under strong political influence, were rather limited to avoid certain manipulations of the data.

Now to the relationship between the UNFPOS and human rights. Amartya K. Sen postulates basic human rights to the development of basic human functions such as health, creativity, personal responsibility or social integration.1212 They are an expression of freedom and are the right of every human being. Self-determination is conceived as a decision-making space to which every human being is entitled. Accordingly, freedom is an individual option space of responsible decision-making, which allows appropriate choices between alternatives. All these features are included in the UNFPOS and other codes of conduct.

The UNFPOS are often compared to the codes of ethics that exist for various disciplines and subjects, including statistics. One example is the Declaration of Professional Ethics issued by the International Statistical Institute (ISI), which sets out ethical principles for professional statisticians. In contrast, UNFPOS formulates principles and guidelines for the conduct of governments, data providers, users of statistics and professionals working with official statistics. I don’t see any disagreement between the approaches.

Some days ago, I was pleased to read that the Greek government accepted the European Court of Human Rights ruling in favor of Andreas Georgiou. Why a trial before the European Court of Human Rights in Luxembourg? The case is about trust, transparency, accountability, and relevance – key elements of the UNFPOS and human rights. Specifically, it dealt with the question of whether Georgiou should have had a vote on the higher deficit figures while in office. He is convinced that this would have been illegal because according to the European Code of Practice for Statisticians, only the head of the statistical office is responsible. Also, in his opinion, one does not vote on statistics as a matter of principle because they are unambiguous if the rules are followed uncompromisingly – facts that are self-evident in a democracy, under the conditions of an independent judiciary and an independently acting statistical office. The European Court of Human Rights condemned Greece in March this year because the procedure was not fair. Thus, the highest court in Athens should have asked the European Court about the sole responsibility of the head of the statistical office. No reasons were given for this refusal, and no mention was made of a corresponding application by Georgiou. The Greek government acted against this European legal ruling and lost.

It is worth noting that in the second basic principle of UNFPOS, there is a reference to the fact that the statistical authorities decide on the methods and procedures for the collection, processing, storage, and presentation of statistical data on the basis of strictly professional considerations – including scientific principles and professional ethics – in order to maintain confidence in official statistics. However, it is not explained what is meant by the term professional ethics. There is certainly room for improvement here. I am convinced that this wording needs to be clarified. Elements belonging to the requirements profile of professional ethics might be such as honesty, objectivity, and professional quality. In addition, a great deal of importance needs to be attached to fairness and respectful mutual recognition in cooperation.

Ivo: Has the adoption of UNFPOS tended to improve the participation rights as well as the professional participation opportunities of statisticians in national statistical offices? Can political influence on statistical work be limited in the long term, or are statisticians always just pawns in cases of doubt?

Reimund: Quite interesting questions you raise here: As already said before, UNFPOS has certainly led to increasing awareness of ethical principles in official statistics. However, it is also necessary to face ethical questions in collegial cooperation. But it often seems to take a lot of courage to recognize grievances and bring them to public attention. An example of this is the investigative report by the law firm WilmerHale described in Chapter 12 of my book. It accuses the IMF’s Managing Director, Kristalina Georgieva, of manipulating the influential 2018 “Doing Business” report in favor of China.1313 Georgieva was managing director of the World Bank Group from 2017 to 2019 and served as acting president of that institution from 1 February to 8 April 2019, following the resignation of Jim Yong Kim.

The circumstance that diplomacy and faithfulness to facts sometimes collide at the World Bank was also criticized by the institution’s former chief economist and Nobel laureate Paul Romer. He left the World Bank in January 2018 after expressing concern about arbitrary adjustments in the preparation of the “Doing Business” report. According to a media report, Romer said that Georgieva had glossed over these facts when he pointed them out to her. There had been a “lack of integrity”, Romer said. Subsequently, the economist suggested that the World Bank’s research reports be prepared externally. He concludes that “all that matter in science are the facts. When complex political sensitivities are allowed to influence research by stifling open disagreement, it ceases to be scientific. … Diplomacy and science cannot thrive under the same roof: The delicate task of keeping the member countries of the World Bank and IMF happy is not compatible with independent research that also reveals unpleasant things for members.”1414

It is interesting to read, at the end of the investigative report, the various recommendations made to ensure that similar irregularities do not occur again. One of the recommendations for the Board’s consideration refers to the Bank’s culture. It says: “Establish firm procedures to protect the “Doing Business” team from pressure or influence from Bank senior management, regional vice presidencies, advisory services colleagues, and member countries.”1515

As already mentioned in my answer to the previous question, it would certainly be advantageous if national and international institutions worked on clarifying the concept of professional ethics for statisticians – not only in the sense of a producer of statistical data (collection, processing, publication with confidentiality of the data) but also in the personal area vis-à-vis superiors and colleagues.

The IMF Executive Board reportedly held eight meetings on the issue, including interviews with WilmerHale lawyers and Georgieva, who denied any wrongdoing. The board was divided on whether Georgieva should retain her position at the IMF, with the US and Japan (the two largest shareholders) favoring removal, while several European countries (France, Germany, Italy, and the UK), China, and Russia continued to support Georgieva. Ultimately, the Executive Board issued a statement in October 2021, supporting her continued leadership at the IMF.

Ivo: Are the ethical principles developed for official statistics, such as the UNFPOS, not already being called into question again today in the face of crises and wars, and in the context of new data sources and new methodologies (like AI, machine learning etc.)?

Reimund: I am sure that this is already the case. Official statistics are being questioned again in view of crises, wars, artificial intelligence (AI), and machine learning. This development correlates very strongly with the dwindling independence of the media worldwide, which has been observed for some time. If we look, for example, at the new press freedom ranking recently published by Reporters Without Borders, organized disinformation is a growing problem in many countries: in 118 countries, i.e., in two-thirds of all countries, most respondents stated that political actors in their country are involved in massive disinformation or propaganda campaigns. It can be assumed that this has also tended to compromise the integrity of official statistics.

For all the justified objections to the validity of such rankings, the fact that Greece, for example, has fallen sharply on the press freedom index is based on verifiable facts and solid evidence. The country, which already has a poor record on media freedom, ranks last in the EU, even behind Orbán’s Hungary, at 108th place, between Burundi and Zambia. Countries that I also mention in my book are even further behind: Kazakhstan, for example, ranks 134th and has plummeted because of the bloody protests in January 2022 and a never-ending wave of violence against journalists. India has slipped further and is now ranked 161 (-11). Not surprisingly, Russia’s position (164, -9) in the ranking has worsened. Since the start of Russia’s war of aggression against Ukraine, almost all independent media have been banned, blocked, and classified as so-called foreign agents. China’s rank (179, -4) is even lower.1616

Ivo: What are the interrelationships and specific political, statistical, and ethical challenges arising from the Irish Case (globalization) and the Covid pandemic (digitalization)?

Reimund: First, the Irish case: What happened when the Central Statistical Office (CSO) in Dublin calculated and published an extremely high GDP growth rate? Everything cried out: that can’t be right, the statistics are wrong, these are Chinese conditions, scientists spoke out, Paul Krugman spoke of Ireland with its leprechaun economy1717 the finance minister rejoiced because Ireland’s national debt and deficits were significantly reduced as a percentage of GDP, the statisticians went on the defensive, doubts arose about the credibility of the data.

From an ethical point of view, it was gratifying that the statistics were not immediately taken to task for this. Instead, the CSO set about improving the “quality” of the data. Procedures were developed relatively quickly to identify the impact of globalization on the Irish economy. The publication of sector accounts for Ireland in particular, considering the different influences, shows that statistics must not resign itself to serving as a plaything of politics (but also of economists and the media), but must provide an answer to current questions. It would only be helpful if the media would also think in this direction and do more educational work.

Things are probably quite different when it comes to reporting on Covid. On the one hand, interest in statistical data grew. At times, the daily statistics on infections, cases of illness and deaths, on vaccination rates and herd immunity played a special role. The extent to which knowledge of statistics increased, as a result can hardly be determined. On the other hand, there were intense discussions about questions of secrecy, and data protection – especially in the context of the corona warning apps. But the astonishing thing is that Apple Watches were accepted without any problems and were and are enormously popular – despite the cautions about data misuse. The question arises: Who protects whom from data misuse in view of largely unlimited data access?

For Germany, the problems of insufficient digitalization, but also a lack of cooperation between Destatis, the regional authorities, and the Robert Koch Institute,1818 became particularly apparent. Ultimately, it was not possible in Germany to set up central registers for infections, vaccinations, and cases of illness – as had happened in Austria, for example.1919 Thus, for long phases of the pandemic, one was poking around in the fog of uncertain and inaccurate data.

Irrespective of these questions, both issues – globalization and digitalization – are the subject of the SNA revision currently underway – largely from the point of view of statistical recording, of course.

Ivo: Are there differences in political influence on statistical work in institutions such as the Federal Statistical Office, the German Bundesbank, the IMF, and the European Central Bank? What might a general code for statistical work in these authorities look like? Is it possible to implement ethical principles in statistical practice at the national level due to the increasing international cooperation and coordination in the field of statistics that can be observed?

Reimund: In general, I can say that statistics in central banks, as political institutions, tend to be more exposed to political pressure. This was particularly the case of the Deutsche Bundesbank. Especially when important financial indicators were published, such as monetary aggregates, external values, or interest rates. In the case of monthly data, the appropriate seasonal adjustment procedure was an important aspect because it also offered possibilities for data manipulation. Fair cooperation with the Federal Statistical Office, the Ministry of Economics, and the Ministry of Finance in economic statistics and the national accounts played an important role.

When working at the European level, the aspect of cooperation and coordination was particularly important. Pioneering work was required, it took a long time before the data for the euro area was ready for publication. There was no political pressure – it was just that countries were not always willing and able to provide the data the ECB wanted. What I found particularly interesting was the competition between national data compilers. The goal of producing quarterly national accounts for the Eurozone meant that we requested national data on banking, balance of payments, and securities statistics as well as on financial accounts. These data were not always provided with the desired periodicity and level of detail but were gradually made available.

I felt intense political pressure especially in the discussions on the fiscal rules and the convergence reports.

Ivo: Are there issues that have not been covered in this interview that you would like to emphasize as a closing of this interview?

Reimund: Two topics have been on my mind for some time: both are only dealt with in passing in the book but are increasingly forcing their way into my consciousness: AI and the further development and consequences of the war in Ukraine.

AI starts with an unfortunate name. “Artificial intelligence” is a term that conjures up the Frankenstein myth. Our image in a machine. You could also call them learning machines, or statistical models for language generation. But that doesn’t sound nearly as exciting. AI, statistics and ethical issues are probably only at the beginning of the discussion. Geoffrey Hinton, former head of Google, which helped AI make its breakthrough, now fears that chatbots like GPT-4 could do serious harm.2020 There is a lot of concern about fake photos, videos, and texts. Recently, Spotify reportedly deleted tens of thousands of AI songs from its platform because of alleged “fake streaming”. There is also talk of falsified data. Particularly dramatic are the statements by Sam Altman, head of the chat GPT provider Open AI, and the historian Yuval Harari, who warn of the extinction of humanity through computer systems.

Now to my second topic: For 2023, the IMF predicts that the Russian economy will grow by 0.3 percent – more than the German and British economies. The Russian Central Bank fluctuates in its forecast between -1 and 1 percent growth. On the other hand, the sanctions against Russia are unprecedented: never have attempts been made to isolate such a large economy, which is also one of the world’s main energy exporters. At the beginning, hundreds of individuals and companies were sanctioned, and half of the currency reserves were frozen. Besides the central bank, most Russian banks are still excluded from the international payment system today, and a considerable part of imports to Russia are boycotted and prohibited. In addition, there was another effect: at first there were only sanctions against Russian imports, but not against exports. At the same time, the prices for oil and natural gas, Russia’s main export products, skyrocketed. This led, as a paradoxical consequence of the import sanctions, to a huge surplus in the trade balance, and the rouble stabilised.

I question whether gross domestic product is a good yardstick at all in this situation. In Russia, for example, the production of tanks, missiles and uniforms strengthens this indicator. Whether this state-financed war expenditure represents economic prosperity for Russia is more than questionable. There is also the fact that since the outbreak of the war, more and more economic data has been classified, which makes analysis difficult. The costs of the war in Ukraine already appear enormous – for both sides. The number of military casualties is not certain, as both sides downplay their own losses and exaggerate those of the enemy. But this conflict is certainly one of the bloodiest in modern history. Estimates of the cost of rebuilding Ukraine after the war vary from $400 billion to $1 trillion. The West is expected to bear the lion’s share of the reconstruction costs. Spurred by the war, global defense spending is also expected to soar. The benefits from higher gas and arms sales are likely to be limited to a small number of energy and arms producers, while taxpayers are and will be hit by inflation and higher government spending on transfers to Ukraine.

Bad cards for credible statistics – and international

cooperation in the field of statistics will certainly haunt us for a long time.