E Africa Data Revolution Report 2016 - Highlighting developments in African data ecosystems - United Nations ...
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
The Africa Data
Revolution Report
2016
Highlighting developments in African data ecosystems
The Africa Data
Revolution Report
2016
Highlighting developments in African data ecosystems
Africa Data Revolution Report 2016
Contents
Foreword v
Acknowledgements vii
Acronyms viii
Executive Summary ix
1. Introduction: The African data revolution and sustainable development 1
1.1 Defining the data revolution in Africa 2
1.2 What is revolutionary about the African data revolution? 4
1.3 Purpose of this report 7
1.4 Methodology 8
1.5 Outline 8
2. Data ecosystems in Africa 9
2.1 Transformations in African national data ecosystems in recent decades 10
2.2 Key data communities within national data ecosystems in Africa 15
3. Data innovation in Africa: Selected case studies 42
3.1 Rwanda: Innovations within the official statistics system 42
3.2 United Republic of Tanzania: Innovations in technology, systems and tools in the
national statistics system 44
3.3 Kenya: Citizen-generated data: Ma3Route 45
3.4 South Africa: The Victims of Crime Survey: An example of a national statistical
office going beyond its traditional domain 48
4. Conclusions and recommendations 52
4.1 Conclusions 52
4.2 Recommendations 55
References 57
iii
Africa Data Revolution Report 2016
Foreword
African countries pledged to achieve sustain- ficient. Infrastructure would be useless if it is
able development and inclusive growth for all not adapted to local context, and if communi-
following the adoption of the 2030 Agenda for ties cannot access or afford it. Data communities
Sustainable Development and the Agenda 2063. would not fully grow and prosper if they are not
The need for timely and quality data to inform empowered to voice their particular needs and
the two agendas exerts pressure on the national capacitated to engage in partnerships and design
statistical systems more than ever before. While their own innovative solutions.
much progress has been made in past decades to
strengthen national statistical systems for evi- To fully harness the data revolution, a holistic
dence-based decision making, countries still face strategic approach is therefore required. Each and
challenges in many areas. The data required for every component of the data ecosystem should
monitoring progress toward the ambitious and be steered toward the shared goal of establishing
aspirational global and continental development a conducive environment for leveraging data and
agendas is unprecedented both in scope and gran- partnerships to respond to development priori-
ularity with which the data should be collected. ties while leaving no one behind.
Technological advancement, growing recogni- This first edition of the biennial Africa Data
tion of individual rights for information and Revolution Report aims to address issues on the
knowledge, the spread of open data initiatives, process of transforming data ecosystems, from
and increasing involvement of various non-tra- the status quo to an aspirational state where data
ditional data stakeholders offer opportunities from both conventional and new sources are be-
to meet data challenges for the implementation ing harnessed to better inform decision-making
and monitoring of the development priorities at and enable sustainable development with contri-
global, regional and national levels. butions from diverse data communities. The re-
port is envisioned to assist countries on the con-
Indeed, the potential of data goes beyond report- tinent in charting their way towards strengthened
ing of development progress. Empirical evidence national data ecosystems by identifying common
is mounting that data enables citizens to make problems, sharing home-grown examples and
more effective decisions in their daily lives, en- lessons learned and coming up with actionable
trepreneurs to create new business opportunities, recommendations applicable in the regional,
and institutions to make the governing process national and local contexts. Further, it reviews
more efficient, responsive, inclusive and trans- the current state of data ecosystems in Africa in
parent. Nevertheless, data alone cannot unlock terms of the diversity of data actors and their ca-
the above-mentioned opportunities if laws and pacity needs, legislative and policy frameworks,
policies allowing people free access, use and re- technological infrastructure, tools and platforms,
use of data are not in place. And such laws and and the dynamic interactions between them.
policies mean little if technical capacity and in-
frastructure to make use of the data are not suf-
vAfrica Data Revolution Report 2016
We hope that this inaugural report will serve as nize African-led regional efforts and feed into the
a baseline from which countries will assess their discourse on data for sustainable development in
progress toward harnessing the data revolution at global fora.
the service of their development priorities, galva-
Oliver J.M. Chinganya Mansour Ndiaye Fernando Perini Anne Jellema
Director, Team Leader, Coordinator, Chief Executive Officer,
African Centre for Inclusive Growth Open Data for World Wide Web
Statistics and Sustainable Development Network Foundation
United Nations Development International
Economic Commission Regional Service Development Research
for Africa Center for Africa Centre
United Nations
Development
Programme
viAfrica Data Revolution Report 2016
Acknowledgements
This report is the joint product of four collaborating nana Eric (Independent Consultant, Madagascar);
institutions: the United Nations Economic Com- Suwadu Sakho-Jimbira (Initiative Prospective Ag-
mission for Africa (ECA), the United Nations De- ricole et Rurale, Senegal); and Charles Doukouré
velopment Programme (UNDP), the International Fe (Independent Consultant, Côte d’Ivoire). The
Development Research Centre of Canada and the country researchers produced 10 background papers
World Wide Web Foundation. These institutions for this report. They received support from national
were represented by a Steering Committee com- statistical offices and UNDP country offices includ-
posed of Chukwudozie Ezigbalike and Inkyung ing those in Côte d’Ivoire, Ethiopia, Madagascar,
Choi (ECA), Serge Kapto and Alessandra Casazza Nigeria, Rwanda, South Africa and the United Re-
(UNDP), Fernando Perini (International Devel- public of Tanzania. The country research in Nigeria
opment Research Centre of Canada) and Nnen- was supplemented by a report on “Unpacking the
na Nwakanma (World Wide Web Foundation), data revolution in Nigeria” prepared by the Feder-
together with the lead author of the report, Cos- al Public Administration Reform Programme, with
mas Ochieng of the African Centre for Technolo- permission from the United Kingdom Department
gy Studies. The contribution made by UNDP was for International Development. Country research in
made possible thanks to the generous support of the Senegal and Swaziland was reused from the UNDP
Government of the Republic of Korea. The Open Ecosystems Mapping Project. Solicited contribu-
Data for Development is funded by the Internation- tions were also received from Davis Adieno (Civi-
al Development Research Centre, the Government cus, the World Alliance for Citizen Participation),
of Canada and the World Bank, all of which were Danielle Doughman and Donatien Beguy (African
indispensable for the production of this report. Population and Health Research Center), Paula
Hidalgo–Sanchis (United Nations Global Pulse,
Under the supervision of the Steering Committee, Uganda) and Nnenna Nwakanma (World Wide
the report was put together by a team of country re- Web Foundation).
searchers headed by the lead author and comprising
Jean-Paul van Belle (University of Cape Town, South This report benefited from an earlier consultative
Africa); Canisius Nzayisenga (Independent Con- workshop in Nairobi, Kenya in August, 2016. Pe-
sultant, Rwanda); Olusegun Akanbi (Independent ter da Costa (Kenya), Inkyung Choi (ECA), Serge
Consultant, Nigeria); Solomon Selassie (Independ- Kapto and Bouazza Bouchkhar (UNDP), Yosr Ab-
ent Consultant, Ethiopia); Kenneth Mdadila (Unit- del Faatah Ahmed Ibrahim (CAPMAS, Egypt) and
ed Republic of Tanzania); Paul Mungai (University Steve Kenei (Development Initiatives) provided
of Cape Town, South Africa); Godwin Hlatshwayo, comments on an early draft of the report.
(Independent Consultant, Swaziland); Rakotoma-
viiAfrica Data Revolution Report 2016 Acronyms AfDB African Development Bank ECA United Nations Economic Commission for Africa ICT Information and communication technologies STEM Science, technology, engineering and mathematics UNDP United Nations Development Programme UNESCO United Nations Educational, Scientific and Cultural Organization viii
Africa Data Revolution Report 2016
Executive Summary
The data revolution can be harnessed to catalyse the Strategy for the Harmonization of Statistics in
positive social, economic and environmental trans- Africa. A data ecosystem which is capable of har-
formation in Africa. Considerable innovation and nessing the full transformative potential of the data
experimentation is currently under way within mul- revolution also requires appropriate technological
tiple data communities and ecosystems in many Af- infrastructure and architecture and human and fi-
rican countries. However, for the most part, these nancial resources. The rise of information, commu-
are small-scale, pilot, isolated or ad hoc initiatives. nication and geospatial technologies, among others,
If Africa is to benefit from the full transformative together with demographic changes, Africa’s recent
potential of the data revolution, more systematic, economic growth rates and increasing democratiza-
large-scale, integrated and sustainable efforts are go- tion provide foundational structures for data ecosys-
ing to be needed. tems capable of harnessing the data revolution.
African governments and other stakeholders recog- In spite of this, significant deficits, gaps and weak-
nize the importance of the data revolution for the nesses remain within the current data ecosystems
continent’s development. At the continental level, in many African countries. If Africa is to harness
this recognition is embodied in the Africa Data the full transformative power of the data revolution,
Consensus1 and other statistical and development significant investment will need to made in hu-
initiatives, including the African Charter on Statis- man resources, technological capabilities, platforms
tics and the African Union’s Agenda 2063. At the and tools, and in establishing effective governance
national level, this can be seen in long-term national frameworks related to the production, processing,
development plans and numerous legal, legislative protection, ownership, quality, openness, timeli-
and policy reforms aimed at improving the quality, ness, relevance, accessibility, harmonization, inter-
timeliness, relevance, availability and accessibility of operability and use of different types of data, regard-
data. In many countries, these national and conti- less of who produces or owns them.
nental development goals are in the process of be-
ing aligned with the global development priorities This report examines the current state of the data
to which most African countries have subscribed, ecosystem in Africa, its desired end state, and the
including most notably the United Nations Sustain- gaps in between. It also provides recommendations
able Development Goals. This alignment entails ef- on how to bridge these gaps. It maps the current
forts to harmonize, coordinate or integrate data re- data ecosystem in Africa in terms of purpose, actors,
quirements for the Sustainable Development Goals principles and protocols; legal, legislative and policy
and national development priorities. frameworks; technological infrastructure, tools and
platforms; and the dynamic interactions between
The building blocks for a data ecosystem which is purposes, actors, frameworks, technologies and sys-
capable of harnessing the data revolution for acceler- tems.
ated sustainable development already exist in Africa.
These include multiple dynamic and growing data This report conceptualizes the “African data revolu-
communities which range from official statistics tion” and identifies what is revolutionary about it –
and private-sector, civil society and citizen-based notably (a) the exponential increase in the volume,
data groups to scientific, open and big data com- types and speed of data available in African coun-
munities. They also include legal, legislative and tries; (b) the increased availability and use of new
policy frameworks aimed at creating enabling envi- types of data as well as new uses of old or conven-
ronments and governance frameworks for harness- tional types of data; (c) new principles of data gov-
ing the data revolution at the national and regional ernance, including the principles of inclusion and
levels, such as the African Charter on Statistics and openness; and (d) new data-related rights and free-
1 United Nations, Economic Commission for Africa, 2015.
ixAfrica Data Revolution Report 2016
doms. The report argues that underlying the con- data ecosystems which are capable of harnessing the
cept of the African data revolution is a fundamental data revolution, it would be helpful if policymak-
conceptual and paradigmatic shift on the question ers and other decision makers had a clear sense of
of who and what officially counts, is counted, how, the full range of impacts of different types of data,
by whom, for whom and for what purposes. data platforms, technologies and tools. The scientif-
ic community and the international and continen-
The question of what and who officially counts and tal development bodies such as ECA, the African
is counted, by whom, how and for whom is more Development Bank (AfDB) and UNDP can help
than simply a question of available or feasible sta- in commissioning, financing or undertaking these
tistical techniques and methodologies. At its core, scenario studies and impact studies. If data, includ-
this question embodies political economy and ideo- ing new and novel types of data, are to be used in
logical considerations of what and who constitutes a decision-making, the case for many of the claims
national priority, and is worthy of counting, and by made on behalf of such data must be backed up by
extension, planning public, social, development and solid evidence. This is likely to enhance public in-
other services for. Official statistics can be and have vestment in national data ecosystems, but also the
been used as tools for social, economic and political use of data produced by such systems. Africa must
exclusion and segregation in Africa, especially under create a virtuous cycle of good-quality data, good
colonial and apartheid rule. However, statistics and decision-making, good policy outcomes and suffi-
data in general can also be used as tools for pro- cient investment in national data ecosystems if it is
moting social, economic and political inclusion and to harness the data revolution on a large scale and
integration, for example by including marginalized sustainably.
communities, regions and peoples in national cen-
suses, household surveys and administrative, civil The importance of capacity needs
and other vital data registration exercises. The report assessments early in the cycles covered by
concludes that the African data revolution embraces
the Sustainable Development Goals and
a broader concept of who and what officially counts,
is counted, by whom and for whom. This derives in Agenda 2063
part from the broader development goals for which
African countries seek to harness the data revolu- It is almost taken for granted that every African
tion: the Sustainable Development Goals adopted country is ready for different types of data, technol-
by the United Nations and the African Union’s ogies, tools and platforms. Yet there is widespread
Agenda 2063. recognition that despite recent improvements, the
national statistical system in many African countries
The report makes the following recommendations: is capacity-constrained. Different data communi-
ties might have arisen in different African coun-
Proof of premise: What is the full range of social, tries, but the assumption that every African country
economic, environmental and other impacts – pos- is ready to harness different types of data ignores
itive and negative – of the data revolution? What fundamental differences in political economies and
might African countries do to maximize the positive institutional, technological, financial and human
and mitigate the potentially negative impacts of the resource capabilities. Whether a country is in a po-
data revolution? sition to effectively harness geospatial and big data,
for example, depends on a whole range of factors
The full range of socioeconomic and other impacts that cannot be taken for granted. If African coun-
of the data revolution and the various types of data tries are going to establish data ecosystems which are
such as big data and open data remains largely un- capable of harnessing the data revolution, national
known. Without this knowledge, what interven- assessments of readiness to handle different types
tions governments and other stakeholders might and sources of data, technologies and platforms are
undertake to maximize the benefits of the data rev- going to be critical in helping them to identify gaps
olution and minimize its negative potential remains or areas where key interventions might be directed.
limited. In order to accelerate the establishment of
xAfrica Data Revolution Report 2016
Legal, legislative and policy reforms to in Africa. These include private-sector, civil society
anticipate and allow for the harnessing and citizen-based data communities and scientific,
of the data revolution for accelerated open and big data communities. Greater collabora-
sustainable development in Africa tion, coordination, harmonization and integration
of data initiatives among these communities have
A number of reforms have been initiated in many the potential to significantly reduce the costs of data
African countries, but in many cases they are either collection and to fill key gaps in national official sta-
incomplete or insufficient. The right legal, legisla- tistics. They also have the potential to enhance data
tive and policy environments will create conditions accessibility, dissemination and use. Public-private
that are conducive to harnessing the data revolution and other partnerships, and other business models
for sustainable development in Africa. for facilitating collaboration in data collection, pro-
cessing, analysis, dissemination and storage, need to
Significant investment in data be explored.
technologies, platforms and tools and
human and financial resources is going The importance of strengthening
to be needed if Africa is to benefit from investment in administrative data
the full transformative potential of the collection and use
data revolution
Administrative data, including civil registration
This includes general investment in information and data, are critical in national development and plan-
communication technologies (ICT) and infrastruc- ning. Civil registration is the means by which coun-
ture, including most notably Internet, mobile and tries keep continuous and complete records of births
digital technologies. It also includes investment in and deaths. The issue and possession of birth and
data science, computing, STEM (science, technol- death certificates dictate several civic, professional,
ogy, engineering and mathematics), statistics, social business and personal activities and transactions. In
sciences and economics skills at a broader tertiary many countries, it is through civil registration docu-
level and with specific reference to the national data ments that individuals are admitted into schools and
ecosystem. hospitals, gain nationality and formal employment,
vote or present themselves for electoral offices, buy
Greater collaboration and coordination and transfer properties and access financial and le-
gal services. To be excluded from civil registration,
between data communities can
then, is almost tantamount to being excluded from
significantly reduce the costs of data public services. In spite of this, 46 African countries
collection for use in pursuing the do not have complete civil registration systems to
Sustainable Development Goals, register births. If the continent is going to harness
Agenda 2063 and long-term national the data revolution to accelerate sustainable devel-
opment, a majority of African countries are going
development plans by helping fill
to need to strengthen their capacities to collect and
gaps in official statistics and enhance use high-quality administrative data.
data quality, timeliness, relevance,
accessibility, dissemination and use. National statistical offices are and will
remain the cornerstone of national data
Data collection is costly. National statistical offic-
ecosystems
es are unlikely to have the necessary and sufficient
technological, financial and human resource capac- In spite of - in fact because of - the rise of multiple
ities to collect, process and disseminate the data data communities, the national statistical system
required for the pursuit of the Sustainable Devel- will continue to play a central role in national data
opment Goals, Agenda 2063 and national develop- ecosystems in many African countries. It can be a
ment plans. However, significant capabilities and re- key facilitator and leader in fostering collaboration,
sources are available within other data communities harmonization and coordination within national
xiAfrica Data Revolution Report 2016
data ecosystems. The other data communities will data actors and communities. To be sure, many con-
continue to complement rather than supplant the straints to participation by non-State actors in na-
national statistical system. How the official statistics tional data ecosystems remain. However, the policy,
community relates to the other data communities legislative and legal openings are there. Non-State-
will to a large extent determine the degree to which based data communities can seize these opportu-
African countries harness the data revolution. nities not only to hold governments accountable
through the open data movement, but also to share
Functional autonomy for the national data, improve data quality and use, and proactively
statistical system seek collaboration with the official statistics com-
munity.
Owing to its continuing centrality within the na-
tional data ecosystem, the national statistical system Learning from experimentation and
requires functional autonomy, or a lack of politi- innovation
cization of its work, coupled with sufficient finan-
cial and human resources. It cannot be expected to Realizing an effective, responsible, coordinated, har-
collaborate with the other data communities or to monized and integrated national data ecosystem will
facilitate collaboration, harmonization and coordi- require significant experimentation and innovation,
nation if its technical and technological capabilities and efforts to learn from them. There is nothing au-
are substantially weaker than those of the other data tomatic about the data revolution. It is the result
communities, or if it does not have sufficient finan- of deliberate legal, policy and technological choices
cial resources to undertake its core mandate, which and actions. Successes or failures of experimentation
would form the basis of any collaboration, harmo- and innovation will be constrained by these choices
nization and coordination. and structural and external factors, among others.
The ability to identify the reasons behind success-
Non-State-based data communities have es and failures within different data communities,
a responsibility to act proactively and and to apply these lessons in future interventions,
responsibly to contribute to the creation will determine the type of data communities and
ecosystems that emerge in Africa in the future, and
of functional, vibrant, harmonized and
the degree to which the continent harnesses the full
integrated national data ecosystems transformative power of the data revolution. The re-
ality of experimentation and innovation will need to
Legal, legislative and policy reforms coupled with be appreciated especially by the public sector, where
advances in ICT have opened up and or are opening failure is not always tolerated. Innovative models for
up the data ecosystem in many African countries to conducting experimentation and innovation such
non-State-based data communities. The opportuni- as incubator labs or pilot projects may need to be
ty to seize this growing policy, legal and legislative explored before large-scale projects are rolled out.
space is increasingly available to non-State-based
xii1. Introduction: The African data revolution
and sustainable development
The data revolution can be harnessed to acceler- number of recent scholarly, policy and commercial
ate sustainable development in Africa. Data are reports reinforce this characterization. Goldman
available ever faster, in greater volume and scope. Sachs observes that while economic productivity
Globally, the quantity and quality of data on social, doubled in 150-200 years during the industrial rev-
economic, political and ecological activities are ex- olution, it “skyrocketed in the computer revolution,
panding exponentially. This is driven by growth in with a three-fold gain in the half-century starting in
the numbers and analytical power of computers, the 1960s”. The firm believes that with the “fourth
machines and mobile devices; the spread of the In- revolution” (i.e. the data revolution), the world may
ternet; the falling cost of data storage; and the surge have entered a period of even more rapid productiv-
in remote sensing, satellite and other technologies ity gains.3 Reports by the World Economic Forum,
that are creating new data and tools on an unprece- the McKinsey Global Institute, the EMC Corpo-
dented scale. It is also driven by a surge in demand ration4 and a number of academics describe huge
for data from the public and private sectors and civil potential benefits of the data revolution for entire
society, and by greater experimentation, innovation economies and sectors such as health, agriculture,
and adaptation in data production, dissemination industry, transport, retail and public services. There
and governance across multiple data communities. is growing consensus that while the data revolution
does not constitute the first time that technology
In its Data Manifesto,2 the Royal Statistical Society has promised to transform the world, this process
in the United Kingdom argues that what steam was might be the fastest.
to the nineteenth century, and oil to the twentieth,
data are to the twenty-first. According to the So- Big data has been variously described as the oil of
ciety, data are the “driver of prosperity, the revolu- the digital economy, the next big thing in medical
tionary resource that is transforming the nature of care and an integral tool for smart city planning and
social and economic activity, the capability that dif- development.5 The global open data community
ferentiates successful from unsuccessful societies”. A makes similarly strong arguments for the potential
2 Royal Statistical Society, 2014.
3 Goldman Sachs Asset Management, 2015.
4 EMC Corporation, 2014. See also Connelly and others, 2016; Einav and Levin, 2013; Einav and Levin, 2014.
5 de Montjoye, Y.-A., 2014.
1Africa Data Revolution Report 2016
applications of open data in enabling innovations to develop new insights and innovations that can
that are catalysing social, economic and political improve the lives of others and help to improve the
transformations by changing sectors as diverse as flow of information, spur development, promote
health, education, urban development and fiscal transparency and build capacities”. Similar argu-
governance. Openness is seen as foundational in ments for the potential benefits of the data revo-
harnessing the opportunities for data innovation lution for sustainable development in Africa have
that are unlocked when data cross geographical, been made by various African governments in their
territorial, sectoral, disciplinary, social and other national development strategies and by African
boundaries.6 Data generated by citizen groups have scholars and development agencies in various books
also been shown to hold great promise, while the and papers, including Ben Kiregyera’s The Emerging
power of high-quality official statistics and scientific Data Revolution in Africa; the Mo Ibrahim Founda-
data in helping inform policymaking has long been tion’s Strength in Numbers: Africa’s Data Revolution;
recognized. and Johan Fourie’s The data revolution in African
economic history.
African governments, civil societies, academics
and private-sector organizations have recognized 1.1 Defining the data
the potential applications of the data revolution to revolution in Africa
the continent’s sustainable development efforts. Ja-
kaya Kikwete, the former President of the United The United Nations Independent Expert Advisory
Republic of Tanzania, spoke for many African gov- Group on the Data Revolution for Sustainable De-
ernments when he opened the first ever Open Data velopment defines the data revolution as an “explo-
Conference in Africa in Dar es Salaam on 4 Septem- sion in the volume of data, the speed with which
ber 2015: “I totally subscribe to the idea that data data are produced, the number of producers of data,
is an important tool for development. No successful the dissemination of data, and the range of things on
and meaningful development will take place with- which there is data, coming from new technologies
out the use of data … We cannot honestly speak such as mobile phones and the ‘Internet of things’,
of developing Africa today without improving on and from other sources, such as qualitative data,
ways we source, process and make use of data. More citizen-generated data and perceptions data”.8 The
importantly, we must transform ways we manage Group considers that the data revolution includes a
data by increasingly making public data open and “growing demand for data from all parts of society”.
shared. In that way we will enable decision makers The data revolution for sustainable development, in
to make informed decisions. We will also empower the Group’s view, involves “the integration of … new
the public to use data to hold the public officials data with traditional data to produce high-quality
accountable. In that way we will enhance good gov- information that is more detailed, timely and rel-
ernance and take Africa to the next level.”7 evant for many purposes and users, especially to
foster and monitor sustainable development; the
In its Data Revolution Policy and SMART Rwanda increase in the usefulness of data through a much
Master Plan 2015~2020, the Government of Rwan- greater degree of openness and transparency, avoid-
da makes the case that the explosion in the world ing invasion of privacy and abuse of human rights
of data has “enormous potential to spur innovation from misuse of data on individuals and groups, and
and creativity, increase the proliferation of highly minimising inequality in production, access to and
skilled jobs, contribute to economic growth, enable use of data; … more empowered people, better pol-
better decision-making and create a more accounta- icies, better decisions and greater participation and
ble, efficient, responsive and effective government”. accountability, leading to better outcomes for peo-
The master plan commits the Government to devel- ple and the planet”.
op a policy on Open Data, on the basis that “easy
access to data allows individuals and organizations
6 Third International Open Data Conference, 2015.
7 United Republic of Tanzania, 2015.
8 United Nations, 2014.
2Africa Data Revolution Report 2016
The Africa Data Consensus, developed at the High providing timely, user-driven and disaggregated
Level Conference on Data Revolution, held in Ad- data for public good and inclusive development”.9
dis Ababa, Ethiopia, from 27 to 29 March 2015,
builds on the Group’s definition. It defines the Afri- The Consensus is a strategy for nurturing the data
can data revolution as: “A profound shift in the way revolution in Africa. It aims to create a new data
that data is harnessed to impact on development de- landscape or ecosystem by opening up the field of
cision-making, with a particular emphasis on build- data production and dissemination to State and
ing a culture of usage. The process of embracing a non-State actors. It was developed in response to
wide range of data communities and diverse range calls for a framework on the data revolution in Af-
of data sources, tools, and innovative technologies, rica and its implications for the African Union’s
to provide disaggregated data for decision-making, Agenda 2063 and the Sustainable Development
service delivery and citizen engagement; and infor- Goals. The plan of action for the application of the
mation for Africa to own its narrative”. The Con- Consensus is being spearheaded by ECA, the Afri-
sensus views the data revolution as a “partnership of can Union Commission and AfDB, with support
all data communities that upholds the principles of from UNDP and the United Nations Population
official statistics as well as openness across the data Fund, and in collaboration with partner institutions
value chain, which creates a vibrant data ecosystem from the public and private sectors as well as civil
society organizations.
The core principles of the Consensus include:
Box 1. Principles of the Africa data revolution
Data must be disaggregated to the lowest levels of administration by gender, age, income, disability and other
categories.
People must be counted to make them count. Civil registration should be accessible and provided at no cost.
Official data belong to the people and should be open to all. They should be open by default.
The data community should embrace the United Nations Fundamental Principles of Official Statistics as a starting
point.
There is a need for governance and coordination of the data ecosystem.
African governments should acknowledge open data provided by credentialled data communities as acceptable
sources of country statistical information.
Technology, new forms of data and other innovations should be actively embraced.
Data communities should promote a demand-driven data user culture spanning the entire ecosystem.
Privacy and intellectual property rights should be respected.
Data should be translated into information that is simple, understandable and relevant. Information must be timely,
accurate, relevant and accessible.
Data must be driven by needs rather than for their own sake.
The data revolution in all its facets should be gender-sensitive.
Source: United Nations, Economic Commission for Africa, 2015.
Analysis of continental and national policy, legal • The considerable changes already under way
and legislative frameworks, principles and protocols within data ecosystems in Africa: the explosion
shows that the African data revolution can be un- in the volume, velocity and variety of data pro-
derstood in a two-dimensional sense:
9 United Nations, Economic Commission for Africa, 2015.
3Africa Data Revolution Report 2016
duced within the continent, their dissemina- statistical offices and other public agencies, but also
tion, accessibility and use from non-public-sector actors such as academic in-
• The data ecosystems to which the continent as- stitutions, private-sector companies, development
pires: inclusive data ecosystems involving gov- agencies, civil society organizations and citizen
ernment, private sector, academia, civil society, groups. For example, almost 9 out of 10 people in
local communities and development partners Africa now live in a country which has conducted a
that tackle the informational aspects of devel- population census in the last 10 years. A third of all
opment decision-making in coordinated ways; Africans live in a country where a census has been
ones in which data are produced and used by a conducted since 2010, and almost all Africans (99
wide range of data communities and harnessed per cent) live in a country which has conducted a
to impact development policymaking; and ones household survey in the last 10 years. Since 2005,
which embrace the core principles highlighted 80 per cent of African countries have published a
above. For example, the Government of Rwan- household survey which includes a health compo-
da, which is one of the few African countries nent.10 At least 12 African countries have signed up
with an explicit Data Revolution Policy, follows to the Open Government Partnership, and at least
the United Nations Expert Advisory Group and 10 have established open data portals. In a number
defines the data revolution as the “transforma- of countries, citizen groups and civil society actors
tive actions needed to respond to the demands are not only producing data that sometimes com-
of a complex development agenda and pri- plement those produced by national statistical of-
vate-sector needs, improvements in how data is fices or fill in the gaps left by such data, but many
produced and used; closing data gaps to prevent of them are also promoting data-sharing and use
discrimination; building capacity and data liter- among different stakeholder groups.
acy in “small data” and big data analytics; mod-
ernizing systems of data collection; liberating A number of data communities in Africa have ei-
data to promote transparency and accountabili- ther piloted or are experimenting with new sources
ty; and developing new targets and indicators”. of data collection or platforms for data accessibili-
ty and sharing. By tracking population movements
From the foregoing, it is apparent that the “African through mobile phone data and by running ana-
data revolution” implies something revolutionary lytics on social media, opt-in services such as the
both in what is already taking place within African Kenya-based Ushahidi have made it possible for hu-
data ecosystems and in what is aspired for or desired manitarian organizations to better understand and
in ideal African data ecosystems. respond more quickly and effectively to human-
itarian crises and natural disasters. By looking at
1.2 What is revolutionary about population flows and conducting real-time CCTV
the African data revolution? analytics to better understand traffic issues, some
African cities, including Abidjan, Cape Town and
From the general definitions of the data revolution Nairobi, seek to better plan for and accommodate
to definitions that are specific to the African data their growing urban populations by building “smart
revolution, the revolutionary aspect centres on at cities” that have adequate health, water, electricity,
least five key factors: education and public transport infrastructure. For
example, Harvard University’s Engineering Social
(a) The exponential increase in the volume, Systems programme is overlapping mobile phone
types and speed of data available in data with census data in Kenya to model the growth
African countries of slums, enabling city planners to better plan for
and build public services such as water pumps and
While there are significant variations among them, public toilets.11 Through the Outthink Urban Plan-
many African countries are producing more, reg- ning Initiative, the city of Nairobi has partnered
ular, and better-quality data, both from national
10 Mo Ibrahim Foundation, 2016.
11 https://www.hsph.harvard.edu/ess/. Accessed 13 November 2016.
4Africa Data Revolution Report 2016
with IBM’s Smarter Cities project to transform and sources of data as well as new applications of
waste management.12 old or conventional types of data. Data are increas-
ingly available on novel types of variables and on
Work to harness the substantial increase in the vol- new activities, peoples or regions that were previ-
ume, types and speed of data available is beginning ously excluded from many conventional types and
to yield significant social, economic and scientific sources of data. Some of the new types or sources
results. Over the last few years, a number of African of data have already been highlighted (such as cit-
countries have revised the methods and base year izen-generated data and satellite or remote sensing
data used to calculate gross domestic product in or- data). Others include “big data” or large quanti-
der to reflect the structure and size of their econo- ties of complex data (sets) which can be curated and
mies more accurately. This has entailed accounting analysed to derive insights and meaning. In popu-
for economic activities previously omitted from na- lar balance, big data is often associated with either
tional income accounting either because they were social media or business data, or both. However,
performed in the informal sector or because they oc- this is misleading, since not only are there a range
curred in sectors that were previously non-existent or of different types of big data, but different types of
underdeveloped such as information and communi- big data sometimes require different analytical ap-
cations technologies. The results of these exercises proaches. For example, administrative data or data
have been dramatic, and reflect the importance of collected by government agencies for the purpos-
high-quality data to national development. Nigeria’s es of registration, transactions and record-keeping
gross domestic product nearly doubled from $270 (such as data for the administration of tax systems,
billion to $510 billion between 2013 and 2014, social programmes and regulations) can be a type
making it the largest economy in Africa. Kenya was of big data. Especially in developed countries, some
catapulted from a low-income to lower-middle-in- of these administrative data now come in large data
come country, while Zambia’s economy turned out sets with tens of millions of distinct observations
to be 25 per cent larger than previously thought.13 and huge numbers of covariates. The increased scale,
The rebased or updated figures not only provided dimensionality and lack of structure in some of the
better assessments of the size of the economy and administrative data in developed countries have led
the composition of and sectoral contributions to some scholars to conclude that administrative data
gross domestic product in a number of countries; not only constitute a distinctive form of big data
the more accurate measures of these countries’ econ- but are significantly underutilized because, among
omies provide an opportunity for policymakers and other things, they are not easily shared (among State
citizens to assess their fiscal circumstances better and federal authorities, for example, or ministries
and prioritize their national development initiatives and national statistical offices).14 While the issues
better. Besides the improved quality of official sta- of scale, multidimensionality, lack of integration
tistics, non-traditional sources of data collection, and coordination, and inadequate sharing and un-
processing, dissemination and sharing such as opt- derutilization are all relevant to administrative data
in and mobile surveys and open data platforms are in many African countries, a general lack of admin-
generating constructive debates about data quality, istrative data is a bigger and more urgent problem
timeliness and governance frameworks across the on the continent.
continent.
The growing availability of large data sets that cap-
(b) Increased availability and use of new ture the behaviour of millions of individuals, or
types of data as well as new uses of old even the complexities of the galaxy, has yielded ex-
or conventional types of data traordinary scientific results, including in Africa.
As Johan Fourie has argued, the applicability of the
The “revolution” in the African data revolution also data revolution to the past is especially valuable in
centres on the increased availability of new types areas with limited written records such as sub-Sa-
12 IBM, 2016.
13 Mo Ibrahim Foundation, 2016.
14 Connelly and others, 2016.; Einay and Levin, 2013; Einay and Levin, 2014.
5Africa Data Revolution Report 2016
haran Africa. Fourie documents how a new gener- (c) Principles of data governance,
ation of economists, geographers and historians are especially the principle of inclusion or
rewriting African history using not only thousands “Leaving no one behind”
of individual archival records but also geographical,
climatic and demographic projections into the dis- What is “revolutionary” in the African data revo-
tant past. A new African economic history has come lution also centres on the fundamental principles
of age, in large part thanks to the data revolution. for governance of the production, dissemination,
use and ownership of data. Many of these have
Scholars such as Fourie have used analysis of the been highlighted in box 1 above, and they include
height of African military recruits during colonial the principles of inclusion and openness. One of
rule (as a proxy for living conditions) to document the most “revolutionary” principles of the African
the evolution of living standards of African peoples data revolution is the principle of inclusion or of
during an era of unreliable data. These data have leaving no one behind. At the heart of the princi-
turned out to be useful not only as a tool to meas- ple is the idea that everyone matters or counts and
ure the level of living standards over time, but also that people must be counted or captured in formal
to test the effects of different colonial policies on data processes such as civil registration to allow the
African populations. “The data revolution has not government to plan for their basic social, economic
only offered new variables to measure past events, and other needs. In the context of the Sustainable
but it has been especially useful as a contemporary Development Goals, this principle amounts to the
outcome variable. In the absence of data on region- idea that no goal, target or indicator should be con-
al African economies, for example, Michalopoulos sidered met unless it has been met for every person
and Papaioannou use light density at night ob- around the globe.
tained from satellite imaging to test the impact of
precolonial ethnic institutions and the impact of the Civil registration is particularly important as far
borders drawn during the Scramble for Africa on this principle is concerned. Civil registration is the
current economic performance”.15 In this case, the means by which countries keep continuous and
data revolution is not only helping to solve a fun- complete records of births and deaths and the issue
damental problem in the study of African history of birth and death certificates, which then dictate
and economics (i.e. a lack of data); it is also leading several formal civic, personal, professional, business
to innovative uses for old or conventional types of and political activities and transactions. In many
data. The decision by a growing number of African countries, it is through civil registration that indi-
governments to make their traditional data (i.e. data viduals are admitted into schools and hospitals, gain
produced by national statistical offices) open by de- nationality and formal employment, vote or pres-
fault or accessible through open data portals is also ent themselves for electoral office, buy and transfer
leading to increasing accessibility of this type of data properties and access financial and legal services. To
and in some cases, new uses of it by various stake- be excluded from civil registration is in many cases
holder groups. synonymous with exclusion from public services. In
spite of this, 46 African countries do not have com-
plete civil registration systems to register births. By
one estimate, 83 per cent of Africans live in a coun-
15 Fourie, 2015.
6Africa Data Revolution Report 2016
try without a complete and well-functioning birth ness, timeliness, relevance and ownership all stand
registration system. Less than one in five births oc- in contrast to the principles and paradigms that
cur in a country with a complete birth registration have governed the production, dissemination and
system, while 87 per cent of deaths occur in coun- usage of official statistics at some points in the Afri-
tries without complete death registration systems.16 can past, notably during the colonial period. What
and who counts and is counted, by whom, how and
(d) Data-related rights and freedoms for whom is more than just a matter of technical
considerations or the applicability and utility of var-
What is “revolutionary” in the African data revolu- ious statistical methods and techniques. At its heart,
tion also centres on data rights and freedoms. The this question embodies political economy and ide-
Consensus is not explicit on this point, but underly- ological considerations of what are considered na-
ing its many fundamental principles are data-related tional priorities by governments in power or dom-
rights and freedoms. A number of African govern- inant interest groups in society. Accordingly, in the
ments have policy, legal and legislative frameworks past official statistics in Africa have been used either
that either explicitly or implicitly provide for da- as tools for social, economic and political exclusion
ta-related rights and freedoms. These include: and exploitation or in order to further inclusion,
conflict resolution and integration, depending on
• The right to be counted, as an integral part of prevailing dominant ideologies, political economies
the principle of leaving no one behind and development paradigms. The broader defini-
• The right of access to information (for example, tions of development and the principles of inclu-
information held by the State – a right embod- sion, participation and empowerment embedded in
ied in article 35 of the Constitution of Kenya both the Sustainable Development Goals and the
and article 34 of the Constitution of Rwanda, African Union’s Agenda 2063 mark a fundamental
as well as Rwanda’s Access to Information Act) shift in the conception of what and who counts and
• The right to participation (for example in data is counted, by whom and for whom. This report
production, dissemination and use) emphasizes that the question of who and what offi-
• The right to non-discrimination and equality cially counts and or is counted, how, by whom and
• The right to privacy and to ownership of per- for whom is determined not just by technical con-
sonal data siderations but by political, ideological, cultural and
• The right to freedom of expression (e.g. citi- historical priorities or factors. That the African data
zen-generated data) revolution answers this question in fundamentally
different ways from past behaviour and practices is
The African data revolution, viewed in its aspira- part of what is revolutionary in the African data rev-
tional dimension, is inclusive of all these rights and olution.
freedoms.
1.3 Purpose of this report
(e) A conceptual, ideological and This is the inaugural Africa Data Revolution Re-
developmental paradigmatic shift on port, the first in a series of biennial reports that will
the question of who and what officially focus on developments in national data ecosystems
counts, is counted, how, by whom and in Africa. It highlights the current state of the Af-
for whom rican data ecosystem, its desired end state, and the
gaps in between. It also provides recommendations
Underlying both the generic concept of the data for creating data ecosystems capable of harnessing
revolution and the more specific notion of the Af- the full transformative potential of the data revolu-
rican data revolution is a fundamental conceptual, tion in Africa. The specific objectives of this report
ideological and developmental paradigmatic shift are threefold:
on the question of who and what officially counts,
is counted, how, by whom, for whom and for what • To review and map components of national
purposes. The principles of inclusion, data open- data ecosystems in Africa, with a view to assess-
16 Mo Ibrahim Foundation, 2016.
7Africa Data Revolution Report 2016
ing their ability to help advance the attainment • Review of scholarly, policy and grey literature
of national and regional development plans, as on the data revolution globally and the African
well as the Sustainable Development Goals data revolution in particular. The literature re-
• To identify common systematic challenges that views were conducted by the lead author and
hamper the full realization of the African data country researchers, with the latter taking the
revolution lead on each of their case study countries: Côte
• To put forward recommendations for accelerat- d’Ivoire, Ethiopia, Kenya, Madagascar, Nigeria,
ing progress towards the desired end state of the Rwanda, Senegal, South Africa, Swaziland and
data ecosystem in Africa, in keeping with the the United Republic of Tanzania.
aspirations expressed in the African Data Con- • Interviews with key informants from represent-
sensus. ative data communities in each case study coun-
try (such as national statistical offices and other
1.4 Methodology members of official statistics communities, rep-
Data for this report were compiled using a combi- resentatives of the private sector, civil society,
nation of methods, including the following: scientific and citizen-based data communities,
as well as key players and stakeholders in open
• In-depth country case studies in 10 African and big data communities).
countries, aimed at understanding their nation-
al data ecosystems: key actors; legal, policy and 1.5 Outline
legislative frameworks; infrastructure, technol- The rest of this report is organized as follows. Sec-
ogies and tools; and the dynamic interactions tion 2 maps the current data ecosystem in Africa,
among actors, subject to the limits imposed highlighting its strengths and limitations with spe-
by legal, legislative and policy frameworks and cific reference to the requirements of the Sustaina-
infrastructural and technological limitations. ble Development Goals, Agenda 2063 and various
In a majority of countries, country researchers national long-term development plans. Section 3
worked closely with national statistical offices presents case studies of some innovations within
and with UNDP country offices. various data communities in Africa, while section 4
concludes the report.
82. Data ecosystems in Africa
A data community is a group of people who share data communities and citizen-based data communi-
social, economic, political and/or professional inter- ties. If classified by types of data, in addition to the
ests in data across the entire data value chain – pro- sectorally based classification, one might have open
duction, analysis, management, dissemination, use and big data communities. Each African country
and storage. A data ecosystem is the complex system has a unique national data ecosystem depending on
of relationships between individuals, organizations, the distinctiveness of its data communities; types of
data sets, standards, resources, platforms and other data; institutional frameworks; data infrastructure,
elements that define the environment in which each technologies, platforms and tools; and the dynamic
particular data resource exists.17 A data ecosystem interactions among the actors, which are subject to
encompasses multiple data communities (e.g. pub- different technological, infrastructural, legal, soci-
lic, private and civil society actors); different types opolitical and other constraints. Figure 2 below is
of data (old and new); institutions, laws and policy a representation of the national data ecosystem in
frameworks; technologies, platforms and tools; and Swaziland.
the dynamic interactions among the actors within
prevailing technological, infrastructural, legal, pol- This report focuses on data communities as classi-
icy and other constraints. In this report, the term fied by sectors and types of data, as follows:
“data ecosystem” in Africa refers to the state of na-
tional data ecosystems in individual African States. • Official statistics data communities
The rest of this chapter examines the various data • Private-sector data communities
communities, institutions and frameworks, and • Civil society sector data communities
technologies, platforms and tools, that constitute • Scientific data communities
national data ecosystems in African countries. • Open data communities
• Big data communities
Data communities can be categorized variously - for • Citizen-based data communities.
example, by sector, types of data, technological plat-
forms, tools and methods. If classified by sector, for This classification is also adopted partly for ease of
example, one might have public-sector-based data presentation. There can be significant overlaps be-
communities such as official statistical systems or tween data communities. For example, open data
national statistical systems, private-sector data com- communities can be inclusive of or interact with
munities, civil society data communities, scientific official statistics communities, citizen-based data
17 This definition draws from both the Open Data Charter (opendatacharter.net) and the Africa Data Consensus.
9You can also read
