Improving data systems
The WDR 2021 introduces the idea of a social contract for data. An Integrated National Data System (INDS) is a way of countries realizing the potential of data for development, using the principles of the social contract as a blueprint. The INDS framework allows a country to share data between national participants safely while maximizing the benefit equitably.
Collaboration is key
The integrated data system is built on an intentional collaborative approach. As a data governance framework, it is set up by the government. But it requires the cooperation of multiple stakeholders. As it grows and develops, it will integrate academic institutions, the private sector, international organizations, civil society, and individuals.
A system built on people
Fundamental to the INDS is building data production, exchange, and use into planning and decision making across government, to enhance service delivery and policy design. This requires demand for data, and the skills to use it.
All the participants of the INDS must work together to build a system around which data can flow to maximize its value. It’s not just about infrastructure or rules and regulations: the cornerstone of the INDS is people.
Analysis, exchange, and collection of data means skilled people are needed across the system. An ongoing investment in human capital is one of the fundamental pillars to making the INDS sustainable.
So how do you build an INDS? Out of five functions, five participants, four pillars, and five foundations.
How to build an Integrated National Data System
The INDS is built around data, which must be:
Produced in a relevant, timely, accurate manner, at a sufficient level of granularity to inform policy decisions.
Protected from misuse by regulations that prevent harm. Protection is a catalyst for trust and participation.
Open and able to flow between stakeholders. Common standards are key to the frictionless flow of data, while also facilitating international data transfer.
Quality controlled to safeguard the integrity of the data themselves. This requires sound methodological foundations in data production.
Used and reused beyond their original purpose by different stakeholders. This includes the routine use of data in planning and decision making across government entities.
As it develops, an INDS will integrate multiple participants. They are:
Government entities, which produce public intent data, and also use other data sources. They act as data stewards in setting the rules for the whole INDS.
Civil society organizations (CSOs), nongovernmental organizations (NGOs), and individuals, which all produce and use data to empower themselves and to hold the public and private sector to account.
Academia, including academic institutions, think tanks, and research organizations that produce and use data, generate public knowledge, and educate people on data use.
Private sector companies, which often produce data as part of their business operations. Some of this can be valuable to public policy and public interest.
International and regional organizations, which sometimes require members to report data, for example SDG data for the United Nations. They can help by setting standards to make data more comparable, and often act as donors to support data production.
The INDS is supported by four pillars:
Infrastructure policies, such as equal access to the internet, a vibrant competitive internet provider market, and internet exchange points.
Laws and regulations that protect individuals, ensure cybersecurity, and manage institutions. Regulation should be independent, but stewarded by the government.
Economic policies, such as government strategy for data governance. Policies are crucial to establishing the value of data, helping them flow across borders and between companies.
Institutions that are set up to govern and safeguard data and monitor compliance. Watchdogs monitor public and private sector compliance.
The INDS is sustained by five foundations:
Human capital, meaning talented people with the right skills to use data, safeguard them, design policies, and hold power to account.
Trust in each other and the system to uphold the social contract for data to maximize value and prevent misuse.
Funding for data production, exchange, and use. This includes competitive salaries for people working in data roles, and funding for technological infrastructure.
Incentives for institutions and individuals to produce, protect, and share data. Sometimes mandates are required for transparency.
Data demand and a culture of data use. Valuing data is crucial for the right data to be produced.
Implementing the INDS
The INDS isn’t just for countries that already have a strong culture of data use and the infrastructure to make it work. It’s for countries at every stage of their data journey.
Different levels of maturity
The framework defines three broad levels of maturity, in order to visualize the journey from outset to optimization.
At low levels of data maturity, countries should prioritize establishing the fundamentals for a national data system. This often involves ensuring that data producers have adequate resources, capacities, and infrastructure; putting in place data protection regulation; and recognizing the importance of data.
Once the fundamentals are in place, countries should focus on initiating data flows. This typically requires incentivizing data sharing, instituting common standards, and ensuring that data users have the data literacy to effectively work with data.
At advanced levels of data maturity, the goal is optimizing the system. Usually this implies coordinating the roles of the different participants and optimizing the flow of and insights emanating from data.
Not a cookie-cutter approach
However, an INDS will always need to be adjusted to reflect country-specific context. Some areas of one country’s data system may be ahead of other areas, and their priorities should reflect this. The maturity framework can be used as a guide to diagnose where the system is at the moment, and identify weaknesses to build on sequentially.
The WDR 2021 has an in-depth action plan for the five INDS participants at the three levels of maturity. A country might take action points from different levels for each participant to advance the national data system.
Integrating the participants of the INDS: Examples of priorities at different levels of data maturity
Building an INDS means integrating all the participants that create and use data. The priorities for each participant depend on the country’s stage of data maturity. It is not a one-size-fits-all approach. Local context will also have to be taken into account.
Level 1: Establishing the fundamentals
A good place to start is with a national data strategy or another high-level political document set by the government, to garner political commitment and resources. Such a document should include concrete policy steps to generate value from data, and it should be reflected in national development plans.
For example, Colombia’s National Development Plan 2014–18 was used as a vehicle to formally assign its National Statistical Administrative Department (DANE) to be the coordinator and regulator of the national statistical system.
Level 1: Establishing the fundamentals
At low levels of data maturity, it is also key to build the infrastructure necessary for the participants to work with data. Constructing national transmission networks for high-speed wireless broadband access is foundational to this end and is a prerequisite to integrating the private sector into the national data system. Fiber can be installed cost effectively in conjunction with new road construction.
This was the case in landlocked Mongolia, where the north-south fiber-optic backbone connecting it to China and the Russian Federation runs along the railway.
Level 2: Initiating data flows
As countries’ data maturity develops, it is essential that data flow between the participants and are used effectively. This requires data literacy. Lack of data literacy in civil society is a major barrier to the demand for high-quality, accessible data, and it limits the accountability role that civil society can play. Improving data literacy through project partnerships, training, and secondments can help address these skill gaps.
For example, the Ugandan Bureau of Statistics and Ministry of Education supported the civil society organization Twaweza in survey and sampling design for a numeracy and literacy survey. Twaweza independently carried out the data collection and processing, improving the quality of and trust in citizen-generated data. The data were later used by the Ministry of Education.
Level 3: Optimizing the system
Once data are flowing and being used, collaboration and coordination between the participants can be optimized. To integrate international organizations effectively into the national data system and to avoid overlapping and conflicting initiatives, domestic actors need to ensure that the data roles and responsibilities of international agencies within a country are coordinated
In India, this challenge was solved by creating committees in which the country offices of various United Nations organizations, line ministries, and research institutions participated. Through these committees, the SDG-related activities and technical support of the various international agencies were divided in a nonoverlapping manner.
Level 3: Optimizing the system
At high levels of maturity, it is also key that data insights are shared. Academia can play a role in transferring and applying global knowledge to local contexts. Innovations emanating from academia should be supported and, where relevant, adopted. For example, randomized experiments in international development research can be adopted as a decision-making tool by many governments.
A prominent example is Mexico’s National Council for the Evaluation of Social Development Policy (CONEVAL). The agency, endowed by the government of Mexico with budgetary, technical, and management autonomy, implements or commissions evaluations of the social policies developed by the Mexican government.
The INDS needs to evolve
Creating an INDS does not happen overnight. It’s an ever-evolving process that needs to grow and adapt over time. The intention is set from the outset by the government but the people are the driving force, keeping it active and extracting the value by sharing and using data openly and proactively.
Getting the fundamentals of security and data strategy in place from the outset sets the stage for safe use of data in the future. This can then be built on as the country’s requirements expand. As the amount of data in the system increases, the level of protection should increase with it.
The INDS is not just an aspirational model. It’s something that every country can work toward today in a scalable way that can fit to their individual context. With a functional INDS in place, we can safely share public and private data sources between users, learn from one another’s analysis, and gain insight that can help the world’s poorest communities. Through this cooperation, data can advance development goals.