The Data Manifesto


Staging a Data Revolution

Accessible, useable, timely and complete data is core to sustainable development and social progress. Access to information provides people with a base to make better choices and have more control over their lives. Too often attempts to deliver sustainable economic, social and environmental results are hindered by the failure to get the right information, in the right format, to the right people, at the right time. Worse still, the most acute data deficits often affect the people and countries facing the most acute problems.

The Data Revolution should be about data grounded in real life. Data and information that gets to the people who need it at national and sub-national levels to help with the decisions they face – hospital directors, school managers, city councillors, parliamentarians. Data that goes beyond averages – that is disaggregated to show the different impacts of decisions, policies and investments on gender, social groups and people living in different places and over time.

We need a Data Revolution that sets a new political agenda, that puts existing data to work, that improves the way data is gathered and ensures that information can be used. To deliver this vision, we need the following steps.

12 steps to a Data Revolution

  1. Implement a national ‘Data Pledge’ to citizens that is supported by governments, private and non-governmental sectors
  2. Address real world questions with joined up and disaggregated data
  3. Empower and up-skill data users of the future through education
  4. Examine existing frameworks and publish existing data
  5. Build an information bank of data assets
  6. Allocate funding available for better data according to national and sub-national priorities
  7. Strengthen national statistical systems’ capacity to collect data
  8. Implement a policy that data is ‘open by default’
  9. Improve data quality by subjecting it to public scrutiny
  10. Put information users’ needs first
  11. Recognise technology cannot solve all barriers to information
  12. Invest in infomediaries’ capacity to translate data into information that policymakers, civil        society and the media can actually use

It is time for data to become a central building block for better choices.
It is time to put the power of information to work.

Setting a new agenda

A revolution in the production, access to and use of data demands a new agenda. Without clear political leadership and public commitments to a shared approach, the step-change in capacity for better data at every level will not be delivered. Three core political actions are necessary for the Data Revolution to release the power of information:

1.    Implement a national ‘Data Pledge’ to citizens that is supported by governments, private and non-governmental sectors

Commitments to better data must primarily be between a government and its people. All governments should make a public ‘Data Pledge’. Private and non-governmental sectors should follow suit and pledge to support national governments to fulfil the ‘Data Pledge’.

The national ‘Data Pledge’ should:

  • Publicly acknowledge the Data Revolution and the imperative to improve public data for decision-making, accountability and improved service delivery.
  • Commit to putting all existing data to work by publishing all data currently in government systems.
  • Produce a data investment plan to finance the systems and structures needed to achieve better data through to 2030.
  • Adopt a comprehensive and consistent policy for opening up all government data quickly.
  • Ensure the data does not become an extractive industry – extracting information from local populations – but protects the public’s right to information about their own society by publishing data in a way that is accessible not just to the few but to all.
  • Introduce regular consultations with civil society organisations (CSO), local governments and businesses to inform the release of data that meets their needs and enables better participation in local governing processes and decisions.

2.    Address real world questions with joined up and disaggregated data

Very few datasets tell a story in isolation. Answers to real world questions often need data from different sources to be joined up and disaggregated so that the impacts on different people and locations can be traced. Data can only be joined up if it can be compared – using common definitions and standards. We need urgent political will to establish basic pillars for the publication of data for all data publishers to adhere to. To get beyond the averages we need a major effort to gather disaggregated data on both the income and wellbeing (in health, education, water and sanitation) of the poorest 20% of people, ensuring no one is left behind.

3.    Empower and upskill data users of the future through education

It is not enough to put the data out there: the data needs to be accessible and useable by all sections of society to overcome of the current lack of trust in data: otherwise the Data Revolution risks merely reinforcing the existing power dynamic. We need a widespread education programme that informs people of their access to information rights and empowers them with knowledge and skills to act on those rights.

Putting existing data to work

Lots of data exists already. Across the world, data availability and quality has improved dramatically during the past two decades. Aggregated development, social and health indicators have been in the public domain for many years. Yet much of the data in most national statistics systems is still private, leaving its economic and social potential untapped.

4.    Examine existing frameworks and publish existing data

The Data Revolution should put to use the large amounts of data that already exist in national statistics and other information systems. Public data should be made public. All governments should start by examining the frameworks and publish data that already exists, in as detailed form as possible, with a particular focus on:

  • Administrative and geographic infrastructures
  • Budgets and censuses
  • Health, education and agricultural data from national management information systems
  • All anonymised microdata collected through publicly funded initiatives (such as household surveys)
  • Panel data measuring what has happened to the same people over time

5.    Build an information bank of data assets

The private sector has expertise in gathering and using data on its customers, supply chains and the provenance of goods and services. Mobile technologies, internet and social media all create opportunities for gathering and communicating data – including from people and places that are currently least well served with data. To put these existing data assets to work, they need to be visible. Governments need to provide incentives for open publication of all data from private and public sources. National audits of data assets should be developed, bringing together data that is often held in sectoral silos. The private sector should work in partnership with governments to show how the data they own, such as mobile phone data, could be provided as a public good to complement and improve public data.

Gathering better data

6.    Allocate funding available for better data according to national and sub-national priorities

There are limited resources to improve data and decision makers have different and sometimes conflicting needs. How the limited resources available to improve data are allocated should therefore be determined by national and sub-national priorities of where greatest impact on poverty can be felt. Allocations should not be driven by international commitments or donor requirements.

7.    Strengthen national statistical systems’ capacity to collect data

The challenges for national statistics systems and solutions are now well documented and must be addressed. They should be tackled through actions such as:[1]

  • Enhance the functional autonomy by recognising the national statistical system as an independent function and ensure national statistics offices receive greater independence from political influence.
  • Reduce donor dependency and fund national statistics offices more from national budgets, based on the business case for investing in better data.
  • Experiment with new institutional models such as public–private partnerships and crowd-sourcing to collect hard-to-get data or outsource data collection activities.
  • Build quality control mechanisms into data collection to improve accuracy.
  • Improve technical infrastructures for data collection, storage and publishing.

8.    Implement a policy that data is ‘open by default’

Where data is available, there are often limitations on how it can be accessed and used. Firstly, data sources are rarely openly licensed, so users are not clear whether they can use and redistribute the data. Secondly, the format is not always machine-readable. Much of the data lives in PDF files or websites. Users need different kinds of information for different purposes, so information needs to be published in a machine-readable format and under open licenses that allow users to access the data and re-purpose it to meet their individual needs.

9.    Improve data quality by subjecting it to public scrutiny

Good quality data – data that is accurate, complete and timely – is critical, and data providers should provide mechanisms for data users to feedback and suggest how it can be improved. Too often policy prescriptions fail to alert the reader of the underlying quality of the data. Hence policy makers are not alerted to the poor quality of the data, reducing their incentive to invest in improving it. Data users and policy makers need to adopt a code of practice and always ask, always tell on the provenance and quality of the data.

Ensuring better information

How data is produced and used matters for how beneficial it ultimately is. Despite the potential that ‘more data’ holds, publishing data on its own is not enough. More data does not always mean better information. Data needs combining, contextualising and explaining if it is to be turned into information that elected officials and civil society can act on.

10.  Put information users’ needs first

Data analysis and information delivery must be driven by the needs of national and sub-national decision-makers and those who need to hold them accountable. Too often published data meets the needs of producers rather than users of information, for whom it is inaccessible or too complex. The needs of data users and local decision-makers must be put first by asking themthrough regular consultations what information they need, how they need it, and when.

11.  Recognise technology cannot solve all barriers to information

Getting the right information, in the right format, to right people, at the right time is challenging. Despite legal frameworks for freedom of information or open data, significant technological, human and bureaucratic challenges must still be overcome. These include unstable power supplies, the lack of functioning websites, very limited numbers of information technology specialists or dedicated information officers working in government, cumbersome procedures to access information, and publication in formats that are difficult to disseminate. Addressing these obstacles requires an understanding of the whole data ecosystem – not just a focus on technological solutions.

12.  Invest in infomediaries’ capacity to translate data into information that policymakers, civil society and the media can actually use

Most ‘users’ don’t want to handle raw data; they need it translated into accessible information. Open data does not in itself provide usable information. In short, data is for machines; information is for people. We need investment to develop and support a cadre of data literate infomediaries – including government departments, information communication technology, research and non-governmental organisations, and journalists to translate data into a form that policymakers, civil society and the media can actually use.

[1] Centre for Global Development and African Population and Health Research Center, Delivering on the Data Revolution in Sub-Saharan Africa, July 2014 www.cgdev.org/sites/default/files/delivering-data-revolution-sub-saharan-africa-pdf.pdf