Verified facts for people in the 21st century: How can professional ethics contribute?

Abstract

Ethics play an important role in the practice of all professions. The best known of all is probably the Hippocratic oath in medicine. But what if a doctor relies on data and statistics to diagnose and treat his patients? If his work involves collecting, analysing, interpreting and drawing conclusions from data, the doctor becomes a statistical practitioner and is expected to follow the values and principles that apply to this profession. Professional ethics for statistics is application-oriented and committed to the goal of guiding statistical practitioners to act responsibly and to ensure the trustworthiness of statisticians and thus also of the statistics they produce. Alongside integrity and professionalism, the third value of professional statistics is respect, both in dealing with confidential data and with the impact of statistics on the communities affected. Respect, professionalism and integrity are the three components of the ‘statistical oath’. However, there is a long way to go from theory to practice. In this article, we will explore different paths we can take towards improving the knowledge and application of these professional ethics in areas of applications and forms of statistical practice. We will focus on one specific area, official statistics.

Keywords

Professional ethics statistics good governance data culture statistical literacy sociology of quantification

1 Introduction

This article is not about statistical methods, neither about their development nor about their application in statistical practice. In this sense, it does not deal with the scientific topics familiar to statisticians. Rather, this article deals with questions relating to the science of statistics as a whole, namely its possibilities and limitations as well as its impacts, risks and side effects. This does not mean that it is about less scientific aspects, quite the opposite. “In the research literature, there is widespread agreement that the theory of science “operates in the spectrum of sociology of science and history of science” and that these three approaches “complement” and “contribute together to a better understanding of science.”¹ It is therefore a question of the extent to which statisticians can succeed in depicting reality (defined as?) objectively and without distortion using statistical methods, i.e., in tracking down the ‘truth’. In addition to the question of scientific knowledge and how it is gained, it is about the social integration of science: What impact do facts have on society and what influence does society and its processes have on the processes and results of statistics? These topics form the background against which the questions of ethics are dealt with, which are primarily concerned with the responsibility of science, scientists and practitioners for procedures, their results and the application of the results.

Like engineers, mathematicians, computer scientists and economists, statisticians have been trained in a professional world that is strongly determined by a positivist mindset. The discussion of questions of truth, knowledge and responsibility tend to be treated as something less important and less scientific. In the worst case, these are (mis)understood as relativisation, a threat to the statistical profession or even heresy. With this narrow understanding of science, the acquisition of knowledge and the role of evidence in today's society, there is a risk of being insufficiently prepared for significant problems. This initially concerns individual research and statistical production when, for whatever reason, these reach their limits and, for example, questions arise about the robustness and scope of statistical results (One illustration of this is the discussion about the so-called ‘reproducibility crisis’.²) On another scale, it is about the role of science and experts (here in statistics) as part of the public discourse in the (late) modern society. “[Anthony] Giddens asserts that with the late modern reliance on generalized expert systems over local knowledges, and upon symbolic tokens such as money, trust remains a necessary part of life: ‘The disembedded characteristics of abstract systems mean constant interaction with “absent others”– people one never sees or meets but whose actions directly affect features of one's own life“³”.⁴ Citizens are dependent on expert knowledge when it comes to the complex issues of our time (such as Covid, globalisation, climate), which in turn requires experts to handle their role and power responsibly. At the core of relations between the producers and consumers of evidence is the need for trust, or more precisely trustworthiness, which experts have to earn by demonstrating and proving their competence, reliability and honesty.^5–7 A lack of trust in experts, science and statistics is grist to the mill of those who operate their political business models on that ground. Because of a “cognitive dissonance between possibly adverse impact of quantiﬁcation, and their purported function of universal certainty, neutrality, and control”⁸ it is necessary to promote receptiveness and understanding of the topics of sociology and the ethics of quantification, and to ensure that this begins as early as possible, i.e., by integrating such topics into the educational programmes.⁹

The article is intended to contribute to achieving this goal. We will first introduce the topic and present the internal structures and sub-areas of statistical ethics. We will then look at two of these structural elements, firstly the professional ethics for applications in practice by individual statistical practitioners and secondly the rules for good practice in public statistics. Finally, different ways of promoting and implementing such professional ethics will be discussed.

2 Ethics in statistics

2.1 Professional ethics as a compass

Why are ethics for statistics necessary at all? Isn't it precisely one of the great hopes and expectations we have of statistics that its methods and data will provide us with the truths we are longing for? Isn't there a deep desire in phrases such as ‘data-driven decision making’ for a quasi-automatable, programmable production of facts that is independent of humans (and their faults)? Will this (erroneous) belief be further reinforced by artificial intelligence, statistics without pilots, self-driving algorithms? Is this trend eventually turning augmented into automated decision-making, and what would this mean for the quality of the decision, but also for public discourse and opinion-forming in civil society? The disappointment when this wish is not fulfilled or when robots will dominate the production of ‘facts’, is therefore great and turns into cynicism and whimsical jokes that put statistics entirely on the level of lies and manipulation. That poses a great danger of relativising and questioning fundamental differences between solid facts and their opposite, fake news.

“The ideal subject of totalitarian rule is not the convinced Nazi or the convinced Communist, but people for whom the distinction between fact and fiction (i.e., the reality of experience) and the distinction between true and false (i.e., the standards of thought) no longer exist.”¹⁰

Conclusions were drawn from the terrible experiences of the dictatorships of the twentieth century: the Declaration of Human Rights in 1948,¹¹ the ISI Declaration on Professional Ethics in 1985,^12,13 the UN Fundamental Principles of Official Statistics in 1994.^14,15 The ethical principles in statistics, which have been attributed universal significance, are concerned with the rights and interests of citizens and with the quality of statistics for the common good: privacy, freedom of information (access to open and transparent information) and statistics (access to and being involved in the provision of good quality statistical information intended to govern a society).

Ethics become important when statistical methods are applied to real life issues, health, income, inflation, poverty, education and much more. If it is challenging to produce accurate statistics, it requires a great deal of expertise and skill to design statistics in such a way that they are timely, comparable, etc. and at the same time relevant to answering information needs. Only then do they fulfil the ‘fitness for purpose’ quality standard. And as the rule named after Charles Goodhart (Goodhart's Law, “that any observed statistical regularity will tend to collapse once pressure is placed upon it for control purposes”.¹⁶) puts it, the risk of feedback and the temptation for decision-makers to influence the facts to their liking increases with their (political) relevance, resulting in policy-based evidence making.

2.2 Mapping a complex topic: Ethics

As far as it is referred to the importance of the Enlightenment for statistics in modern societies, it is necessary from today's perspective to point out, at least very briefly, that an ethical orientation necessarily requires an epistemological positioning and in particular an answer to the question of how objective, neutral and ‘true’ statistical representations of reality can be. The sociology of quantification deals with this topic and the back and forth between the spheres of influence of society and statistics.^3,17–20 Ethics of quantification is one of the requests that follow from “the spread of quantiﬁcation and the signiﬁcance of new regimes of measurement”.⁸

Figure 1.

Hierarchical levels of statistical ethics.

Luciano Floridi / Mariarosaria Taddeo²¹ and David Hand²² have developed convincing thoughts on this subject matter and have provided useful pointers, such as

“Navigating between the Scylla of social rejection and the Charybdis of legal prohibition in order to reach solutions that maximize the ethical value of data science to beneﬁt our societies, all of us and our environments is the demanding task of data ethics.”²¹

“As far as ethical codes for data collection, manipulation, and use are concerned, these have various functions, including things such as the following:

providing guidance on how to behave in difﬁcult circumstances;

preserving privacy in a way that users and the public will ﬁnd acceptable;

ensuring that data are used in such a way as to beneﬁt the public;

reassuring customers, the public, and others about an organization's integrity; and

reassuring employees that they work for a trust-worthy organization.” ²²

And

“However, the context of data science is so vast and diverse, and is changing so rapidly over time, that we cannot hope to put in place precise regulations. There cannot be a single and simple universal set of rules, and unexpected and unforeseen circumstances are certain to arise. The best we can hope for are some ethical principles that have to be interpreted or instantiated in particular applications. That is, the principles must be mapped to low-level guidance, and this is likely to be application speciﬁc.” … “At the highest level, the principles include such things as integrity, honesty, objectivity, responsibility, trustworthiness, impartiality, nondiscrimination, transparency, accountability, fairness, robustness, resilience, usability, efﬁciency, and independence. All good and desirable characteristics. These are then reﬁned into lower, but still high-level principles.”²²

Obviously, ethical values and principles are to be understood hierarchically (Figure 1), in such a way that the fundamental ones (i.e., values) should be as general as possible and thus robust against changes (e.g., through technologies) over time, while the guidelines derived from them are more precise, but also more in need of evolution/adaptation, similar to the situation in legal systems.

Depending on the area of application in which statisticians work, they will also encounter the ethical guidelines and specifications that are relevant there, be they general guidelines for scientific research,²³ rules for professional conduct in the neurosciences²⁴ or official statistics,¹⁴ to name but a few.

Which level of the ethical hierarchy and which characteristics are suitable depends on the intended purpose: Whether a relatively open conveying of values is intended (e.g., in statistics training) or whether, as in the Code of Practice of the European Statistical System,²⁵ binding rules are required with regard to compliance with which reviews and certifications are to be carried out.

It is also necessary to distinguish for whom these ethical principles and guidelines are meant and relevant, for the individual practitioner of statistics (be it as a researcher or as a manager), for an institution (statistical authority, bank, etc.) or for those who determine the framework for statistics as ordering parties, financiers or users (Figure 2).

Figure 2.

Forms and areas of application of ethics.

Following this distinction, questions will also arise regarding the tools for implementation (and control). On the one hand, this concerns the possibilities of verification through e.g., accreditation, certification or compliance monitoring and reporting. Finally, it is important to consider which measures will achieve the greatest success in promoting ethics. Various tools are available for this (Figure 3), and when using them it is important to consider what effect, reaction and possibly also counter-reaction they will have.

Figure 3.

Toolkit for implementation of ethics in practice.

Furthermore, especially in a discourse on ethics from a global perspective, it should not be overlooked that it cannot be assumed that its derivation from universal human rights is not comprehended and shared without reservations. Rather, representatives of so-called critical theory and post-colonialism question the Enlightenment and articulate the tension between values and power,^26–30 for example in connection with the demand for data sovereignty of indigenous peoples.^31–34 Under these circumstances, an imbalance in favour of enforcement tools can potentially be counterproductive in communities with which a trust-based relationship is to be established.

3 Professional ethics: Guidelines for practitioners

Professional ethics in the narrower sense is aimed at the professionals in statistics, i.e., the individuals who work with statistical methods and thus generate statistical results. What matters here is not so much that these individuals possess training and a degree in the discipline of statistics. Rather, these ethical principles fundamentally address all those who practise statistics, regardless of the specialist discipline involved. When doctors are treating patients, they are bound by their Hippocratic Oath, but also by statistical ethics if they are working with data and performing statistics. Of course, this also applies vice versa, so that statisticians must observe the ethics of the specialised fields in which they work.

The International Statistical Institute ISI sees itself as the global statistical institution whose task it is to define, publicise and promote ethics. Due to the considerable breadth of the areas of application of statistics, the aim can only be to cover the top level of the aforementioned hierarchy of ethical rules, namely the generally shared professional values and the essential principles that derive from these values.

The ISI Declaration on Professional Ethics introduces the topic as follows:

“The aim of this declaration is to enable the statistician's individual ethical judgments and decisions to be informed by shared values and experience, rather than by rigid rules imposed by the profession. The declaration seeks to document widely held principles of the statistics profession and to identify the factors that obstruct their implementation. It recognises that, the operation of one principle may impede the operation of another, that statisticians – in common with other occupational groups – have competing obligations not all of which can be fulfilled simultaneously. Thus, statisticians will sometimes have to make choices between principles. The declaration does not attempt to resolve these choices or to establish priorities among the principles. Instead, it offers a framework within which the conscientious statistician should be able to work comfortably. It is urged that departures from the framework of principles be the result of deliberation rather than of ignorance.”³⁵

Responsibility is the leitmotif of this ethics, which manifests itself in three core values: Respect, Professionalism as well as Truthfulness and Integrity. These values are used to derive a total of twelve principles (Figure 4), which are explained in more detail in the Declaration.

Figure 4.

Principles of the ISI declaration on professional ethics.

What is important is, firstly, the strategic importance that the ISI attaches to ethics and, secondly, the existence of a board (https://isi-web.org/ethics) whose tasks are to advise the management level on ethical issues and to undertake activities for promoting observance of ethical principles in statistics.

4 Codes of conduct in public statistics

4.1 Politics of numbers: Public statistics and the state

Considering that (official) statistics as a child of the Enlightenment, especially in its application as official statistics, is already more than two hundred years old, the fact that the first version of a corresponding global ethics document (the ISI Declaration of Professional Ethics) dates back to 1985 is surprising.³⁶ Developments in the area of application of official statistics fall within the same period:

“The need for a set of principles governing official statistics became apparent at the end of the 1980s when countries in Central Europe began to change from centrally planned economies to market-oriented democracies. It was essential to ensure that national statistical systems in such countries would be able to produce appropriate and reliable data that adhered to certain professional and scientific standards.”¹⁴

If we go even further back in history and read the Declaration of Human Rights of 1948,¹¹ we will search in vain for a direct reference to statistical information, although it is self-evident (from today's perspective) how important solid statistics are for the realisation of such human rights. It was not until the outcome of the Rio Conference in 1992, the ‘Agenda 21’, that information for decision-making was explicitly and in detail addressed (Chapter 40) as well as the role of science (Chapter 35). In this context, i.e., also at the beginning of the 1990s, the promotion of codes of practice and guidelines for science and technology is included in the strategy (para 31.9):

“The objective should be to develop, improve and promote international acceptance of codes of practice and guidelines relating to science and technology in which the integrity of life-support systems is comprehensively accounted for and where the important role of science and technology in reconciling the needs of environment and development is accepted. To be effective in the decision-making process, such principles, codes of practice and guidelines must not only be agreed upon by the scientific and technological community, but also recognized by the society as a whole.”³⁷

The era between the 1980s and the first decade of the twenty-first century was characterised by major political changes, especially with the fall of the Berlin Wall and the hegemony of neo-liberal politics.

“Neo-liberal governance, which seeks to achieve self-regulation of actuaries using measures and indicators rather than instructing them with regulations. This approach promises no less than modernity, democratisation and transparency. Furthermore, ﬂattening of hierarchies and auditability are to be ensured as well as a general quality improvement. Improving efﬁciency is the overall goal of this approach.”³⁸

In view of the importance of data for neo-liberal management and policy control, one would at first expect that the provision of such information would be valued accordingly (in terms of willingness to pay) and that information competence should be available on the user side. Paradoxically, this view has not materialised, at least not for official statistics. Rather, starting with the reforms of the civil service, marked above all by those in the British Thatcher Government (See discussion of the Rayner Revie.³⁹), official statistics suffer from their odour of superfluous bureaucracy (red tape), which were the first to fall victim to budget cuts in the waves of streamlining the state apparatus. At the same time, by utilising modern IT technology and restructuring its staff, through increased use of new data sources, etc., official statistics has succeeded in significantly increasing its productivity and thus meeting budgetary challenges. At the same time, global methodological (e.g., SNA, SEEA) and quality standards have emerged during this time, which are of fundamental importance for the comparability of statistics and for global development strategies (e.g., the Sustainable Development Goals).

Since (official) statistics is a result of the Enlightenment as much as the nation state, it is closely tied to it (in both good and bad periods). Not least as a result of digitalisation and globalisation and as a response to emerging new information needs in times of crises, these statistical principles require continuous practical application and interpretation, and, to a limited extent, adaptation to new data sources, circumstances and methods.⁴⁰

With the Sustainable Development Agenda, a global strategy has been adopted for the individual goals and overall success of which reliable, verified facts are of decisive importance. Two goals should be explicitly cited here:

“Goal 16 is about promoting peaceful and inclusive societies, providing access to justice for all and building effective, accountable and inclusive institutions at all levels. Goal 17 is about revitalizing the global partnership for sustainable development. … It requires partnerships between governments, the private sector, and civil society”; in particular Target 17.18 …, enhance capacity-building support to developing countries, …, to increase significantly the availability of high-quality, timely and reliable data …”⁴¹

In this global effort, common ethics for the statistical profession, good governance rules for statistical institutions and support for their independence, strength and innovativeness are critical. The policy objective should be to create a strong data culture in which value creation on the part of statistics producers and value appreciation on the part of users are mutually supportive and reinforcing.

4.2 Trust in numbers in the time of datafication

The conditions under which public statistics operate are marked by the difficulties of their time. In the dictatorships of the twentieth century, statistics were put on a short leash by state interests; in the neo-liberal era, they were shrunk as superfluous ballast. Both had a devastating effect on the quality of evidence and the public trust based on it. Tim Holt, then President of the Royal Statistical Society and Director of the Office for National Statistics, commented on the establishment of modern statistical governance through the Statistics and Registration Service Act 2007:

“Statistics, their production and interpretation, really do matter. It is uncommon to ﬁnd some area of public debate that does not draw on a body of statistical information. As a result, public conﬁdence in the statistical system and its outputs is vital. In particular this is essential if people are to have conﬁdence in the decisions that are made on their behalf by elected leaders. For example, people will not accept the case for closing a school or hospital if they do not trust the statistics on which the case is made. More generally, no one will trust government if it is seen to rely on statistical information which is perceived as untrustworthy.”⁴²

Today, a new threat is that the statistical logic of the past two centuries is being replaced by a data logic:

“The new apparatus of number-crunching is well suited to detecting trends, sensing the mood and spotting things as they bubble up. It serves campaign managers and marketers very well. It is less well suited to making the kinds of unambiguous, objective, potentially consensus-forming claims about society that statisticians and economists are paid for.”⁴³

Statistics and the significance of statistics with the function of authority are about to fade in the public consciousness and to take a back seat in the political agenda. If nothing assumes this role of statistics with the stamp of officiality, the door is opened for any form of arbitrariness, not least also and especially on the basis of data sciences and a new generation of data.^43,44 Imagine that the GDP in its current form would be up for disposal. Instead of an extremely elaborate accounting based on international standards, it could be replaced by an estimate based, for example, on night-time light intensity from remote sensing measurements. Instead of a single official indicator with possibilities for comparison over time and space, we would find ourselves in a maze of parallel solutions, all claiming to be based on statistical and data science methods.

In 2019, the Royal Statistical Society has set out the following comprehensive and well-balanced recommendations in a Data Manifesto:

“Ensure official statistics are at the heart of policy debate

Commit to greater data sharing between government departments for statistics and research purposes

Champion basic training in data handling and statistics for politicians, policymakers and other professionals working in public services

Maintain the commitment to keep pace with other leading scientific nations on investment in research and development

Give the Office for National Statistics and the wider Government Statistical Service adequate resources

Prepare for the data economy by skilling up the nation

Involve the public in shaping the conversation about how data is used

Misinformation needs countering but without undermining free speech

Move beyond averages and break down data to a much more granular level

Keep data regulation updated to protect the public“ ⁴⁵

What is (positively) striking about this Manifesto and the list of recommendations is the close interlocking between statistics / data sciences and their applications. A holistic view of the interaction between the three historically important drivers of development in statistics, namely science, statistics and society,³⁸ ensures an adequate response to the challenges of our time. It is interesting to note that it is not so much any methodological-conceptual differences between statistics and data sciences that need to be bridged. Rather, they are two concepts and approaches to tasks of quantifying aspects of reality that are as complementary to each other as individual transport and rail transport are responses to questions of mobility. So, what essentially characterises the two is a decentralised organisation of statistical processing for the Big Data world on the one side and a centralised, standardised logic embedded in quasi-industrialised processes for the Official Statistics world on the other. It should be noted that the point here is not to evaluate one approach against the other, but to emphasise their complementarity (Source⁴⁶).

4.3 UN Fundamental Principles and other codes of conduct

Professional ethics is an important basis for the formulation of rules for statistical institutions. In addition, however, there are two others:

Firstly, as statistical offices generally belong to the public sector, the respective framework conditions and rules of the particular administration apply to them, with the following principles emphasised as generally valid by the International Federation of Accountants:

A.
“Behaving with integrity, demonstrating strong commitment to ethical values, and respecting the rule of law.
B.
Ensuring openness and comprehensive stakeholder engagement.” ⁴⁷
Secondly, there is a close link to the development of quality management and quality codes in official statistics, as can be seen in the European Statistics Code of Practice, which has largely emerged from proposals from the joint working on quality proposals.⁴⁸
“The Fundamental Principles of Official Statistics (FPOS) is a document often referred to as the Statistical Decalogue. It was developed in the years 1989–1991 under the auspices of the United Nations Economic Commission for Europe (UNECE) Conference of European Statisticians as a set of recommendations for governments and statistical offices of countries belonging to the UN European region. The direct and main purpose of this initiative was to create a document that would provide guidance and form the basis for statistical offices in post-communist countries.”⁴⁹
The Decalogue of Principles has remained stable over the past decades and has been able to retain its function as a guideline thanks to this continuity.⁵⁰ Again, it is important that these rules are high on the agenda of international statisticsSee for example the UN Statistical Commission 2024 (https://unstats.un.org/UNSDWebsite/statcom/55) and, secondly, that there is a UN manual⁵¹ for their implementation that can be consulted, especially in capacity building of statistical systems for better capacity and better results.

Meanwhile, it belongs to good practice in public statistics for institutions to specify quality codes tailored to their circumstances and framework conditions. Whether and to what extent international standards are adequately translated at national level and their compliance (externally) verified in this way are, however, further questions, the answers to which ultimately come up against the possibilities and limits of enforcing international standards in general. The European Union is a special case in this regard, where the statistical system, consisting of institutions at national and European level, are all subject to the same code, which is embedded in Community law. This makes it possible to enforce compliance with the quality code by means of joint governance, which includes an Advisory Board (ESGAB https://ec.europa.eu/eurostat/web/esgab/introduction) and regular Peer Reviews (Peer Reviews https://ec.europa.eu/eurostat/web/quality/peer-reviews).

To summarise, it can be observed that remarkable progress has been made in public statistics in recent decades, which has been driven not least by the dilemmas, conflicts and necessities of application in politics and public discourse. As the UNECE Data Ethics Review⁵² has outlined, the discussion on appropriate quality codes will continue rapidly and intensively in the coming years.
5 Professional ethics in data literacy and capacity building

The following sections focus on measures and activities dedicated to empowerment. The aim will be to create the prerequisites and framework conditions that prepare fertile ground for facts in the form of high-quality statistics to actually fulfil their purpose of serving public discourse, objectifying conflicts and providing trustworthy evidence for political decision-making. To this end, it is essential to keep an eye on the interaction between the producers of statistics on the one hand and their users on the other: “The pragmatic intent is completed by retroaction and feedback loops so that the indicators resulting from the model influence the behaviour of the various agents involved (citizens, consumers, firms, Government).”⁵³

5.1 Competences, competence levels

“Data literacy is the cluster of all efficient behaviours and attitudes for the effective execution of all process steps for the creation of value and/or decision making from data.”⁵⁴ With this broad concept for data literacy, it becomes clear that it is about much more than just being familiar with techniques and methods. Rather, competence in a comprehensive sense involves a combination of knowledge, skills and attitudes, which includes an awareness of and commitment to the values and principles of professional ethics. Furthermore, it is important at what level such competences must be available, which in turn - comparable to language competence - depends on the context in which and the purpose for which such competences are to be applied.

5.2 The need for professionalisation: Accreditation

A broad concept of skills, knowledge and behaviour should be applied in the training of professional statisticians and data scientists, including during their first years of professional practice. With this objective in mind, scientific statistics associations have developed accreditation procedures and formalised final certificates. An example is the certification as ‘European Accredited Statistician’:

“The overall goal is to increase the quality of statistical work and information and thereby supporting mankind with statistical information in any field where statistics can give useful guidance. The purpose of the accreditation system is to provide a common European standard for defining the statistical profession, to provide steps for the applied statisticians and simplify the description of merits.”⁵⁵

The objectives of the Alliance for Data Science Professionals are similar:

“We are defining the standards needed to ensure an ethical and well-governed approach so the public, organisations and governments can have confidence in how their data is used.”⁵⁶

5.3 Critical thinking skills

In the interest of empowering civil society with the aim of broad participation in evidence-based political discourse, it is not least a matter of being able to deal critically with statistics in addition to technical and methodological skills. This task can be approached in relation to statistics in general by analysing the characteristics of statistical literacy from the perspective of critical thinking.⁵⁷ With the eight critical thinking skills distinguished here, it should be possible to successfully develop a soundly based culture of data and to counter ‘crank’ science using sophisticated pseudo-statistics.⁵⁸ In the specific domain of civic, public and official data, the task of trustworthiness, transparency and institutionally based authority from the perspective of the users of statistics arises with particular urgency, far-reaching consequences for the democratic order and often without the existence of alternative data sources. In this context, it is worth emphasising that an early and consistent introduction to the assessment of statistical facts (quality, trustworthiness, etc.) is important for everyone in schools, as well as learning how to use such facts for one's own information needs.

mention the importance early childhood education and its high return What this means in concrete terms for statistics in the public sphere is explained in detail in the article “Literacy in statistics for the public discourse.”⁵⁹

5.4 Statistical and data literacy skills required in policy-making

As briefly described in the introduction, policy has several roles and tasks in the processes of evidence-based decision-making. Just as there are data and facts for policy, it is conversely important to emphasise the different policies regarding the framework for data on the one hand and for statistics on the other.⁶⁰ Accordingly, different competences are also required in the political sphere, depending on whether someone is preparing decisions on the analytical side of policy advice with the help of evidence, whether it is a matter of communication or whether the legal-political determinants for actors and action in informational value creation are to be designed.

For the one side, namely the use of evidence for policy-making, the competences and literacy required for doing so can largely be referred to what has already been elaborated. More interesting, however, is the opposite direction of activities, namely the politics for data and the politics for facts. In this respect, qualities and competencies that generally apply to good management are (additionally) important. W. E. Deming has established his criteria for profound knowledge, according to which successful management is based on the following four competences: “Knowledge about variation, Psychology (psychology of individuals, groups, society and change), Theory of knowledge, Appreciation for a system.”^61,62 In the field considered here, the system for which ‘Appreciation’ is required is about a comprehensive and deep understanding of the processes and interrelationships related to statistics as

a science of variation, data and uncertainty, with the extraction and interpretation of knowledge involving constant querying of all aspects, including issues, data, models, assumptions and analyses,

a (modern) technology, reliable data sources, efficient processing methods,

an infrastructure that needs to be regularly maintained and modernised,

a common language between producers and users,

all framed by and based on values, ethics, governance.

5.5 Governance, institutions and trust: Data culture

Trust is built on knowledge and experience as opposed to blind faith.

Trust in democratic institutions, such as public statistics, is characterised by a mutually reinforcing relationship between Wertschöpfung (value generation) and Wertschätzung (value appreciation). Creating Wert (value) in the form of informational products and related services is the task of public statistics. Whether and how well this can be achieved depends not least on structural preconditions such as governance, budget, competencies, etc. Wertschätzung (appreciation) on the part of the users of these products and services is influenced by concrete experiences, but also by opinions, attitudes, values and not least by (statistical) literacy.

For public statistics, danger arises when either the conditions for Wertschöpfung (value generation) or the factors influencing Wertschätzung (value appreciation) are unfavourable. Recent experience from the Covid pandemic has shown weaknesses and risks on both sides: On the one hand, weaknesses in the coordinated and qualified provision of relevant indicators, on the other hand, considerable gaps in statistical literacy. As a result, official statistics have come under pressure, their relevance to the breadth and urgency of politics is fading, their competence is suffering from a lack of investment, and their authority is being weakened by ‘alternative’ ways of producing quantitative world views, regardless of how they are obtained.^63,64 This conceals an essential implication: Of course, an appreciation of public statistics in the policy-making circles is a prerequisite for them to initiate appropriate programmes and measures. In the absence of statistical literacy and an understanding of the role of public statistics in the democratic process, greater trouble is inevitable. Winning trust back is an objective whose importance can hardly be overstated. This can only be achieved through political initiatives and investments on both sides.

Let us summarise: There is a social and political dimension to literacy in data and statistics that we might call data culture.⁶⁵ If this important prerequisite is missing, then political programmes will be one-sidedly focused on the elements to which the public's attention is devoted in the short term. Comprehensive knowledge of structures and forward-looking investments in public infrastructure will be lacking, value creation and appreciation will suffer, trustworthiness will be (further) weakened.

6 Concluding remarks

This article takes an approach that firstly develops ethical guidelines and good practice codification for statistics, and official statistics in particular, and secondly examines the various ways in which such guidelines can be implemented in practice. It turns out that ethics is a complex and multi-layered subject, even at the theoretical level, for the realisation of which an apparatus of tools is available. The essential lesson from these discussions seems to be that ethics is a topic that can and should guide us in a social learning process to deal knowledgeably and wisely with the opportunities, risks and side effects of the comprehensive quantification of all areas of life.

Footnotes

ORCID iD

Walter J Radermacher

Funding

The author received no financial support for the research, authorship, and/or publication of this article.

Declaration of conflicting interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

References

Kühne

Berr

. Philosophy of science—philosophical foundations and positions. In: Kühne

Berr

(eds) Science, space, society: an overview of the social production of knowledge. Wiesbaden: Springer Fachmedien Wiesbaden, 2022, pp.47–110.

Wasserstein

Schirm

Lazar

. Moving to a world beyond “ p < 0.05”. Am Stat 2019; 73: 1–19.

Beck

Giddens

Lash

. Reflexive modernization: Politics, tradition and aesthetics in the modern social order. Stanford: Stanford University Press, 1994.

Lupton

. Risk. 3rd ed. London: Routledge, 2023.

O'Neill

. A question of trust: The BBC reith lectures 2002. Cambridge: Cambridge University Press, 2002.

O’Neill

. What we don't understand about trust. TED: TED, 2013.

UNECE Conference of European Statisticians. Core values of official statistics: defining the behaviours that show how we live the values. Geneva: UNECE/CES, 2023, Contract No.: ECE/CES/2023/1.

Saltelli

Di Fiore

. From sociology of quantification to ethics of quantification. Humanit Soc Sci Commun 2020; 7: 1–8.

Rogliano

. La formation à l’éthique du numérique dans l’enseignement supérieur (ingénierie et filières scientifiques) entre injonctions politiques et pratiques de terrain à l’heure de la crise écologique et de l’essor de l’intelligence artificielle. Quelles pistes d’évolution pour les formations ? Bruxelles, BE: Université Libre de Bruxelles (ULB), 2022.

10.

Arendt

. The Origins of Totalitarianism. New York: Harcourt Brace Jovanovich, 1973.

11.

Universal Declaration of Human Rights, (1948).

12.

International Statistical Institute (ISI). ISI declaration on professional ethics - adopted in August 1985 - Background Note The Hague: ISI; 2018 [cited 1985]. Available from: https://www.isi-web.org/isi-declaration-professional-ethics-0.

13.

Morganstein

. Revising the ISI declaration on professional ethics: background and challenges. Stat J IAOS 2011; 27: 71–75.

14.

United Nations. Fundamental principles of national official statistics New York: UN; 2023. Available from: https://unstats.un.org/unsd/dnss/gp/fundprinciples.aspx.

15.

United Nations Statistics Division. Fundamental principles of national official statistics New York: UN; 2024. Available from: https://unstats.un.org/fpos/.

16.

Chrystal

Mizen

. Goodhart's law: Its origins, meaning and implications for monetary policy. Cent Bank Monet Theory Pract Essays Honour Charles Goodhart. 2003;1.

17.

Ravetz

. Scientific Knowledge and Its Social Problems. New Brunswick, USA and London, UK: Transaction Publishers, 1971.

18.

Espeland

Stevens

. A sociology of quantification. Eur J Sociol/Archives Européennes de Sociologie/Europäisches Archiv für Soziologie 2008; 49: 401–436.

19.

Henneguelle

. Socio-economics of quantification and value: The perspective of convention theory. In: Handbook of Economics and Sociology of Conventions. Cham: Springer International Publishing, 2022, pp. 1–20.

20.

Roy

. Diverse ideas about the value of official statistics systems. Stat J IAOS 2024; Preprint: 1–6.

21.

Floridi

Taddeo

. What is data ethics? Philos Trans R Soc : Math Phys Eng Sci 2016; 374: 20160360.

22.

Hand

. Aspects of data ethics in a changing world: where are we now? Big Data 2018; 6: 176–190.

23.

Tuana

. Leading with ethics, aiming for policy: new opportunities for philosophy of science. Synthese 2010; 177: 471–492.

24.

Society for Neurosciences. Professional conduct: SfN; 2024. Available from: https://www.sfn.org/about/professional-conduct.

25.

Eurostat. European statistics code of practice - For the national statistical authorities and Eurostat. Luxembourg: Publications Office of the European Union, 2018. p. 20.

26.

Hacking

. How should we do the history of statistics? In: Burchell

Gordon

Miller

(eds) The Foucault effect - studies in governmnentality. Chicago: University of Chicago Press, 1991, pp.181–195.

27.

Feder

. Power/knowledge. In: Taylor

(eds) Michel Foucault: key concepts. Key Concepts: Acumen Publishing, 2010, pp.55–68.

28.

d’Entrèves

. Critique and enlightenment: Michel Foucault on ‘Was ist Aufklärung?’. In: Geras

Wokler

(eds) The enlightenment and modernity. London: Palgrave Macmillan UK, 2000, pp.184–203.

29.

Mounk

. The identity trap: A story of ideas and power in our time. London: Penguin Books Limited, 2023.

30.

Neiman

. Left is not woke. New York: Polity Press, 2023.

31.

Hudson

, et al. Indigenous peoples’ rights in data: a contribution toward indigenous research sovereignty. Front Res Metr Anal 2023: 1–8.

32.

Denzin

Lincoln

Smith

. Handbook of critical and indigenous methodologies. Los Angeles: Sage, 2008, Available from: https://doi.org/10.4135/9781483385686.

33.

Smith

PLT

. Decolonizing Methodologies: Research and Indigenous. Peoples: Zed Books, 2021.

34.

Carroll

Herczog

Hudson

, et al. Operationalizing the CARE and FAIR principles for indigenous data futures. Sci Data 2021; 8: 108.

35.

International Statistical Institute (ISI). Declaration on professional ethics The Hague: ISI; 2023. Available from: https://isi-web.org/declaration-professional-ethics.

36.

International Statistical Institute (ISI). Background documentation and bibliography: ISI declaration on professional ethics the Hague: ISI; 2023. Available from: https://isi-web.org/background-documentation-and-bibliography-isi-declaration-professional-ethics.

37.

United Nations. Agenda 21 - United Nations conference on environment & development. In: Development UNS, editor. New York: UN; 1992.

38.

Radermacher

. Official Statistics 4.0 - Verified Facts for People in the 21st Century. Heidelberg: Springer Nature Switzerland AG; imprint Springer, 2020.

39.

Hoinville

Smith

TMF

. The rayner review of government statistical services. J R Stat Soc Ser A (Gen) 1982; 145: 195–207.

40.

Bidarbakhtnia

. Rethinking official statistics: a sociological perspective. Stat J IAOS 2024; 40: 521–532.

41.

United Nations. The Sustainable Development Agenda New York: UN; 2017 [Available from: http://www.un.org/sustainabledevelopment/development-agenda/.

42.

Holt

. Official statistics, public policy and public trust. J R Stat Soc: Ser A (Stat Soc) 2008; 171: 323–346.

43.

Davies

How statistics lost their power – and why we should fear what comes next. The Guardian. 2017.

44.

Wise

. Educating data scientists and data literate citizens for a new generation of data. J Learn Sci 2020; 29: 165–181.

45.

Royal Statistical Society. The data manifesto London: Royal statistical society RSS; 2019. Available from: https://rss.org.uk/policy-campaigns/policy/our-data-manifesto/.

46.

Radermacher

. Standardisation and statistics. Surv Stat 2021; 84: 24–31.

47.

IFAC. International framework: Good governance in the public sector - executive summary. London: The International Federation of Accountants (IFAC) and the Chartered Institute of Public Finance and Accountancy (CIPFA), July 2014.

48.

Leadership Expert Group (LEG) on Quality. Quality in the European statistical system - The way forward. Luxembourg: Office for Official Publications of the European Communities, 2002.

49.

Oleński

. The fundamental principles of official statistics as the basis for a social information environment in the globalised world. Wiadomości Statystyczne The Polish Statistician 2024; 69: 30–37.

50.

Willis-Núñez

Ćwiek

. Thirty years on, the fundamental principles of official statistics remain the shared foundation of an informed society. Stat J IAOS 2022; 38: 711–723.

51.

United Nations. National quality assurance frameworks manual for official statistics. New York: UN, 2019.

52.

UNECE. In-depth review of data ethics. Geneva: UNECE, 2022 27 September 2022, Contract No.: ECE/CES/BUR/2022/OCT/2.

53.

Barbieri

, editor Statistics, reality, truth. 10th ICOTS Conference; 2018 July 2018; Kyoto

54.

Schüller

. Future skills: a framework for data literacy. Working Paper No. 53. Berlin: Hochschulforum Digitalisierung; 2020 July 2020.

55.

Federation of European National Statistical Societies (FENStatS). Accreditation of European statisticians 2024. Available from: https://www.fenstats.eu/acc_by_law.

56.

Alliance for Data Science Professionals (AfDSP). Alliance for data science professionals 2024. Available from: https://alliancefordatascienceprofessionals.com/.

57.

Koga

. Characteristics of statistical literacy skills from the perspective of critical thinking. Teach Stat 2022: 59–67.

58.

Lee

Yang

Inchoco

, et al. Viral visualizations: how coronavirus skeptics use orthodox data practices to promote unorthodox science online. In: Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems; Yokohama, Japan: Association for Computing Machinery, 2021. p. Article 607.

59.

Radermacher

. Literacy in statistics for the public discourse. Stat J IAOS 2021; 37: 747–752.

60.

Radermacher

. Governing-by-the-numbers - reflections on the future of official statistics in a digital and globalised society. Stat J IAOS 2019; 35: 519–537.

61.

Moen

Norman

. Always applicable - Deming’s system of profound knowledge remains relevant for management and quality professionals today. Qual Prog 2016; 2016: 47–53.

62.

Deming

. Out of the crisis. Cambridge, MA: Massachusetts Institute of Technology, Center for Advanced Engineering Study, 1994.

63.

Radermacher

. Medium [Internet]. Policy D, editor2021. [cited 2021]. Available from: https://medium.com/data-policy/official-statistics-language-for-public-discourse-part-1-294d31e56b6e.

64.

Radermacher

. Medium [Internet]. Policy D, editor2021. [cited 2021]. Available from: https://medium.com/data-policy/official-statistics-language-for-public-discourse-part-2-b9e912621523.

65.

Radermacher

. Statistical awareness promoting a data culture. Stat J IAOS 2022; 38: 453–461.