ISI Academy - Ensuring the Professional Independence of Official Statistics

View the recording

Gerry O'Hanlon: Ensuring the Professional Independence of European Statistics - Much Done, More to Do?
The principle of Professional Independence (PI) is the first principle of the European Statistics Code of Practice (ESCOP) and is of fundamental importance in ensuring both the high quality of the statistics and the trust of users. The core objective of the PI principle is to ensure that statistics are developed, produced, and disseminated in a manner that is free from political and other external interference. This paper traces the development and articulation of the principle over the past thirty years in the context of the wider formulation and adoption of the European Union (EU) framework legislation for statistics and the complementary ESCOP. It is argued that the principle is of key importance in establishing an institutional environment for official statistics that meets the needs of a more integrated European Union, while also being consistent with the EU's espousal of liberal democratic values. Issues with achieving full compliance by the EU Member States with the principle are examined and some proposals are put forward for further improvements in this respect. View the slides

Nancy Potok: The Effectiveness of the US Approach to Statistical Agency Professional Autonomy
The United States has a long history of principles and practices that are designed to assure the professional autonomy of the federal agencies that produce official statistics.  Because the U.S. statistical system is decentralised, this approach has relied on cooperation from multiple cabinet-level agencies that house the 13 designated statistical bureaus and units.  Recent U.S. laws, such as the Foundations for Evidence-based Policymaking Act of 2018, were intended to strengthen the professional autonomy of the statistical units. The law requires the Executive Office of the President to issue regulations on how cabinet agencies should implement these statutory protections.  We will examine how well this approach to professional autonomy has worked, particularly in an administration that is willing to disrupt long-standing norms and practices and discuss the effectiveness of the US approach when compared to the EU regulatory approach. View the slides.

Walter J. Radermacher: Ensuring the Professional Independence of Official Statistics: the social and political foundations and conditions for a robust statistical governance
Official statistics should provide politically relevant indicators without, however, being policy-driven. Only by guaranteeing their independence can they be trusted to inform democratic public discourse and provide a basis for decision-making and political accountability. In this area of tension, three things are needed: Firstly, statistics must be produced to a high standard of quality. Secondly, relevant political actors must act with integrity. Thirdly, the general public must be data literate. This will help them to understand the differences in quality between facts and their sources. 

Strengthening the position of public statistics and safeguarding their independence requires a combination of approaches that target all three dimensions and their interaction. The most important of these is statistical governance, which ideally involves a legal constitution that guarantees the independence of the statistical institution with regard to its professional activities (with its affiliation to a public administration setting limits in terms of budget and other rules).

However, having a system of governance does not mean that it is universally respected, nor that the restrictions on political power it contains are necessarily accepted. What can be done in cases of political influence, whatever form it takes? How can the rules be enforced? Who can counteract the dangerous and harmful exercise of power, and how?

Robust statistical governance is only possible if the social and political foundations and conditions are in place for a separation of the spheres of politics and statistics. Empowerment measures serve this purpose and also improve data literacy. While they may not be helpful in an acute crisis, they do contribute to preventing crises in the future.

In the modern media landscape, statistics must establish a strong brand identity amid fierce competition and an overwhelming amount of news. Through its targeted evidence policy, it should promote quality and ethical information in the modern world, particularly with regard to the manipulation of official statistics and the spread of misinformation by AI.

In addition to institutional safeguards at the national and international levels, civil society and academic stakeholders are playing an increasingly important role. As an NGO representing the interests of statisticians and statistics, ISI is well placed to contribute to the independence of official statistics in this way. View the slides.