General introduction to the special issue: Governance Statistics and the UN Praia City Group

1.Introduction

Long considered a side-line of mainstream economics and public policy, the subject of governance has, over the last 30 years, gradually moved into the very centre of the discipline, and more broadly into the agenda of the international development community, in which it is today fully embedded. The focus on the concept of governance often raises questions due to the paradigm that it can convey, particularly when it is used to question the role of the State. However, despite the critical distance that should be kept when using this concept, the interest in the quality of institutions and issues of participation or accountability in development policy marks a turning point.

Of the 17 Sustainable Development Goals (SDGs) adopted by the United Nations in 2015, SDG 16 “Peace, Justice and Strong Institutions” relates directly to issues of governance. This constitutes significant progress, especially given that the SDGs are universal and represent commitments for all world nations, whereas the Millennium Development Goals only covered developing countries. In Africa, governance, peace and security are also some of the aspirations expressed in the Agenda 2063 (Aspirations 3 and 4) [1].

The adoption of the SDGs was also accompanied by a call for a data revolution by the Secretary-General of the United Nations in 2014 [2]. The report notes the growing need for statistics throughout the world and the ever-widening gap between developed and developing countries in terms of access to information. It gives several recommendations for mobilising technological innovation and financial resources to promote statistics as part of a partnership between all stakeholders [3]. However, despite this promising context, the situation in terms of statistics remains critical, especially in Africa, as evidenced by the 2019 Africa Governance Report [4]. This warning of a “data gap” across the continent shows that the alarmist statements made in the first half of the 2010s are still relevant [5, 6].

Despite the gaps, growing demand has created a significant boost in terms of governance statistics with the burgeoning emergence of statistical sources (household surveys, expert surveys, administrative data, big data, etc.). The official statistics system took up this opportunity at the international level with the creation of a UN City Group on governance statistics in 2015 (the Praia Group), which published a Handbook on the subject at the start of 2020 [7, 8]. This handbook has been complemented by a publication highlighting the importance of governance statistics during the COVID-19 pandemic and providing a number of practical recommendations for their use in this exceptional context and subsequent recovery [8]. UNODC also drew up a Handbook on corruption surveys at the same time [9].

Other institutions made considerable efforts to place governance at the centre of the international development agenda and to advance knowledge of governance and its policies. The OECD, for example, has made institutional trust a key subject of consideration, in terms of analysis and policy, and methodology. It drew up guidelines on measuring trust [10] and conducted an international survey, the results of which have been published in a dedicated report [11]. Africa, under the auspices of the African Union, has developed its own methodology, entitled Governance, Peace and Security (GPS) statistics, as part of its Strategy for the Harmonisation of Statistics in Africa (SHaSA) [12]. This initiative, which was launched despite an incomplete statistical information system (thereby attesting to the significant adaptability of the national statistical institutes (NSIs) on the continent), puts these NSIs ahead of the rest of the world in this area [13].11

This dossier on governance statistics is taken from a special edition of the Statéco journal, co-edited by the National Institute of Statistics and Economic Studies (INSEE), the French National Research Institute for Sustainable Development (IRD) and the Economic and Statistical Observatory of Sub-Saharan Africa (AFRISTAT). It includes articles updated and translated into English [14]. Its aim is to draw up a situation report on the topic, covering conceptualisation, experience, progress and prospects. It presents original contributions relating to both methodology (how can governance be “measured”?) and analysis (what outcomes can be drawn from the governance statistics?).

Composed of nine articles, this dossier covers a broad range of topics, approaches and perspectives and geographical scales. It includes authors from different disciplinary and continental backgrounds. In total, 26 contributors from Brazil, Cameroon, Canada, Cape Verde, Côte d’Ivoire, France, Mali, Madagascar, Norway and Peru and representing four continents (Africa, North and South America, and Europe) participated in this edition, with a clear emphasis on developing countries. In terms of background, professions and institutional membership, the vast majority are statisticians working in NSIs, as well as academics and researchers, economists and political scientists, executives of international organisations or even international consultants and governance specialists. While the articles focusing on official statistics form the core of the dossier, approaches driven by other stakeholders are also represented.

Three of the authors are also Directors-General of African NSIs (Cameroon, Cape Verde and Mali),22 which shows both the interest in the topic of governance within official statistics and the role that the continent plays in this emerging field of official statistics [13].

The first article presents the institutional framework and specific activities carried out by the Praia Group, the only UN City group on statistical methodologies based in Africa, as the sole global platform bringing together the NSIs around governance statistics. The Handbook on Governance Statistics drawn up by the group (see above) is the first official document bringing together the experiences of different countries and providing guidelines and advice on GPS data collection. The article’s four authors (João Cardoso, Arvinn Gadgil, Marie Laberge and François Roubaud) come from three institutions (INECV, IRD and OGC/UNDP), which have been the key stakeholders of the Praia Group since the beginning. They show how this collective has been able to push the boundaries and further highlight how crucial it is to monitor governance statistics in a pandemic context. Based on the experience gained, and in order to make the transition from recommendations to action, the Praia Group launched Task Teams to develop international standards and standardised tools for both surveys and administrative data collection. The first two teams to have been set up by late 2021 work on two major dimensions of governance: non-discrimination and equality; and political participation and participation in public affairs.

In Brazil, Leonardo Athias and Luanda Botelho focus on the measurement of governance and, particularly institutional transparency. The authors describe and analyse data collected at different administrative levels to measure various aspects of the quality of public or semi-public institutions.

The next article on France addresses methodological questions relating to the development, progress and challenges about statistical measurement of security and violence, including the levels of victimisation. Hélène Guedj, Tiaray Razafindranovona and Sandra Zilloniz describe the process that led to the redesign of the system used for the annual survey carried out by Official Statistics, which dates from the late 2000s. It took into consideration the emergence of multimodal data collection that combines online, phone and face-to-face methods. This new way of collecting information allows for a significant increase in the number of responses, which is now at over 100,000 individuals. It offers further benefits: greater accuracy of responses; a fine geographical level of representativeness of the results; and the possibility of adding targeted subject-specific questions regarding interactions with the police. This article discusses the methodological challenges posed by collecting these data in a high-income country, illustrating the universal nature of these questions.

Marc-Antoine Pérouse de Montclos focuses on measuring violence in countries in fragile contexts. He insists on the need to reflect on the categories chosen and the tools used to assess the change in violence. Their relevance and coherence may depend on the contexts, which are likely to vary across country and over time. In particular, he questions the quality of geocoded event databases. Among the various methods typically used to measure the number of victims, direct observation of the number of deaths resulting from violence from various causes is the method that has garnered the broadest agreement. However, establishing international and temporal comparisons is hindered by the difficulties in developing statistical measurement methods that enable the characteristics of the different affected countries to be taken into consideration and the changes over time to be accurately recorded.

Four articles in this dossier use data taken from the GPS-SHaSA modules, each from a different African country. Even though the articles’ thematic input is specific and limited to each precise field, they provide an illustration of the analytical potential of this type of survey, as they address different dimensions of governance to understand the phenomena observed.

In the case of Madagascar, Andriameva Yvan Assany and Faly Rakotomanana use GPS data use GPS data to examine the weak correlation between the perception weak correlation between the perception indicators (which relate to feelings and are therefore considered to be “subjective”) and experience indicators (which relate to personal experience and are therefore considered “objective”), a finding that is the subject of debate in the literature beyond Madagascar. They apply their analysis to three types of scourge the population may be a direct victim: corruption, discrimination and criminal violence. The authors show that the phenomena recorded by the perception and experience indicators reflect different and certainly complementary dimensions.

In Cameroon, the article written by Joseph Tedou, Rosalie Niekou and Guy Ndeffo also compares perceptions and experiences, but by focusing solely on discrimination. The authors benefit from using GPS data, added to the Cameroonian household survey (ECAM4), a survey that seeks to measure precisely the different dimensions of poverty to obtain accurate indicators of income poverty and to obtain accurate indicators of income poverty. The analysis highlights the forms of discrimination in Cameroon and the characteristics of the victims. The high level of discrimination (experienced and perceived) and the discovery of the significant role played by law enforcement agencies in the perpetration of discrimination are some of the major findings of the study.

In Côte d’Ivoire, Christian Boga also uses GPS data in a different way to analyse citizen participation, a criterion of social cohesion, by focusing specifically on young people in Côte d’Ivoire. By using different composite indices developed using political and social participation variables, he highlights the low engagement of the general population in (both political and social) citizen activities and that of young people in particular (which depends on multiple socio-economic characteristics). Among the factors that explain this low level of engagement, he identifies the impact of corruption, growing insecurity and problems of trust among the young and old.

The fourth article on Africa (by Siaka Cissé, Aminata Coulibaly, Daouda Aba Fane, Sékou Haïdara and Arouna Sougané) examines the factors that determine violence in Mali. This article uses data taken from the GPS survey modules to establish the profile of victims of crime in Mali since the start of the conflict. This work benefits from the possibility of using data on citizen experience of governance and security at local and national level as well as perception data. It contributes to a recent strand of the literature suggesting that perceptions, and in particular, the associated feelings of fear, have larger effects than those of experience itself.

The presentation of this Malian experience is of specific interest not only due to the current situation in the country and the urgency of governance issues at the time of writing this introduction, but, above all, to acknowledge the implementation of GPS surveys over many years by the NSI statisticians, despite a difficult political context. The analysis of Mali given here is unparalleled on several levels. Firstly, Mali plays a leading role in Africa in relation to GPS data collection. The authors have benefited from annual governance and security data being available in all regions nationwide since 2014. The series used is all the more unique given that the GPS surveys are included as modules in the permanent, modular survey, which covers a wide range of subjects associated with the SDGs. The analysis helps to record the phenomenon of crime, which has faded in importance since the explosion of violence as a result of the political and social crisis that has spread across the country since the first coup in March 2012. It makes it possible to bring to light a profile of victims of non-lethal crime in Mali.

We have chosen to conclude this dossier with the experience of Peru, unique due to its duration, institutionalisation and dissemination, as well as the methodological innovations implemented. Launched, coordinated and monitored over a long period (since the start of the 2000s) by the authors (Javier Herrera, Nancy Hidalgo, Mireille Razafindrakoto and François Roubaud) as part of a close partnership between the NSI and a research institution (IRD-DIAL), the Peruvian initiative demonstrated, well before the adoption of SDG 16, the relevance of using population surveys for the production and regular monitoring of governance indicators. The richness of available data from the survey system is the result of its sample size, its representative national coverage and the possibility of using regional and departmental inference, its panel component, and its integration of specific ad hoc modules on governance into a broader survey on standard of living. It allows information to be cross-referenced and, in particular, economic, social or regional disparities to be identified.

As can be seen, the majority of empirical data used in the articles in this dossier are taken from official household surveys, typified by the cases of Mali (historical leader in Africa) and Peru (leader at international level). This pre-eminence probably reflects the difficulty in using public registers for statistical purposes, with good representativeness and reliability, especially in developing countries, which are characterised by the fragility of their institutions (which means statistics are hardly a priority). For example, we could cite the reporting of (or sentencing for) cases of corruption before the justice system, which not only massively underestimates the real impact of the phenomenon, but also provides a biased image (grand corruption, media cases, etc.). The same phenomenon is at work in the case of victimisation (violence, harassment, incivility, etc.).

All the articles highlight the need to standardise the definitions and/or coordinate the bodies collecting the data, not only to limit collection costs but above all to prevent fragmented time frames and methodologies. Indeed, only data taken from systems that are consistent in their development phase can be used for comparisons and reveal dynamics that make sense. However, from one institution to the next, we see a lack of coordination and the use of different definitions and indicators (see Athias and Botelho) that are also sometimes inconsistent from one period to the next (see Pérouse de Montclos on measuring violence). This situation makes it difficult to analyse these indicators at a global level. Potential improvements are clearly identified for each of these data types, for which it is essential to highlight the fact that collection is still relatively recent. These databases, as new and diverse as they are, have benefited from scientific and methodological advances that have made it possible to remedy at least some of the potential reliability issues. Even though there is still room for improvement, they are already making it possible to answer central research questions to further the understanding of the challenges associated with GPS subject areas and inform the representatives of public institutes and civil society.

Overall, the articles in this dossier cover a broad range of subjects on governance statistics. They show the scope of the field statisticians, researchers and all stakeholders can seek to explain using data. A review of the content of this edition shows that it covers almost all of the eight dimensions of governance identified as part of the work carried out by the Praia Group [7]: non-discrimination and equality, participation, openness, access to and quality of justice, responsiveness, absence of corruption, trust, and lastly, safety and security.

Beyond merely observing the scope of the subject areas addressed, the articles show, above all, the extent to which statistical data can provide a useful insight into various phenomena that are traditionally considered difficult to understand or even totally unknown. The aim is not simply to provide indicators, the significance of which may remain obscure and difficult to translate into measures or policies. It is also, as can be seen from reading these articles, to provide relevant data established as part of a joint effort between the NSIs and social actors in the long term, based on local and context-specific knowledge, following a bottom-up approach. This is a condition that needs to be met in order to carry out a fine analysis of the actual situation, which really is of great value in informing public debate and policies.

In conclusion, generating knowledge on little-known or unknown phenomena, for which the data available were, in the past, deficient (such as corruption, discrimination, political violence, etc.), is crucial at a time when information lies at the heart of governance challenges. On the one hand, information can be hidden, filtered or even distorted by organisations with the power to do so for the purpose of manipulating opinion. On the other hand, taking the opposite approach, the availability of information is the precondition enabling citizens to exercise oversight and demand that governments or any other development stakeholder be accountable for their actions.

Finally, we would like to stress that, in general, the availability of statistical information makes it possible to put into effect the principles of inclusion, transparency, and public participation advocated by SDG 16. This makes statistics a direct governance issue. As statistical data on governance largely come from population surveys, thereby voicing public opinion to governments, they thus make a twofold contribution to achieving the principles of SDG 16.