OCHA AND PIM TEAM CO-HOSTED DATA SHARING MEETING

A FRAMEWORK FOR
DATA SHARING IN PRACTICE:
PART ONE
New York, May 2017

I. Introduction
The meeting was held on 16 May at the World Economic Forum offices in New York. The purpose of the gathering was to
explore ways in which as a community we can facilitate the safe, responsible and purposeful sharing of data,
information and analysis for stronger humanitarian response and protection outcomes. It brought together
practitioners and thinkers from Global Pulse, World Economic Forum, GovLab, University of Leiden, UNDPKO, Harvard
Humanitarian Initiative, Data & Society, UNICEF, OHCHR, InterAction, DRC, ICRC, WFP, OCHA and UNHCR.

II. Background
Data, information and analysis are prerequisites for improved humanitarian response. Yet the push for quality data and
evidence has not been matched by a similar increase in the sharing and use of the data collected. There is currently no
common framework for how and when data is shared in support of humanitarian action and protection response. The
absence of a common framework results in a) less or no sharing b) irresponsible sharing or c) confusion among partners
about what can or should be shared. All of which may result in a loss of knowledge and evidence needed for decision-
making and response, both internally and with operational stakeholders and partners.

Several data and information management initiatives have been launched in recent years. Through these, various
streams of work have emerged seeking to facilitate, encourage, and ensure safe and responsible data sharing for
stronger humanitarian action and quality outcomes. The objective of the 16 May meeting was not to replace or dismiss
this work but rather to support and build upon it, with a view to explore how different institutional perspectives and
various data and information management initiatives could be brought together for a more successful result as a
community.

Data combined with the changing nature of technology poses complications for understanding the opportunities and
challenges around data sharing. For example, discussions often combine issues of process, risk, data and analysis,
making it difficult to separate and analyze the various components.

Part of the motivation for the creation of a Framework for Data Sharing in Practice is to provide a conceptual starting
point from which to separate and begin to address specific issues relating to data in support of humanitarian and
protection response in order to develop them collectively. Further, there is a recognition and shared value among
stakeholders for the need to create a Framework for Data Sharing in Practice.

III. Objective
The purpose of the meeting was to discuss ways to facilitate and ensure the safe, responsible and purposeful sharing of
data, information, and analysis for stronger humanitarian response and protection outcomes. To meet this objective,
the agenda for the Framework for Data Sharing in Practice meeting covered the following elements:

1
OCHA AND PIM TEAM CO-HOSTED DATA SHARING MEETING

A FRAMEWORK FOR
DATA SHARING IN PRACTICE:
PART ONE
New York , May 2017

• Sharing the latest developments from partners aimed at addressing challenges with sharing data, information and
analysis;
• Identifying critical questions not being addressed with regard to sharing of data, information and analysis;
• Exploring elements of a common framework for the sharing of data, information and analysis;
• Identifying practical outcomes that could be applied to improve current sharing practices; and
• Exploring next steps for continued collaboration in the year ahead.

IV. Elements of a Framework for Data Sharing in Practice

What follows is a distillation of those talks into an initial outline of a Framework for Data Sharing in Practice. Note that
some of what is presented below incorporates ideas that were shared in the 16 May meeting, while other elements build
upon thinking from the various initiatives represented at the meeting.

This report is the first in a two-part document. In the coming months, a second document / Part II, based on further
consultation and drafting, will put forward concrete proposals for the framework, as outlined below. Part II will identify
prerequisites for sharing of data, information and analysis across organizations and between actors. Its objective will be
to define a minimum structure and components for a framework for sharing of data, information and analysis, which
can remain flexible enough to adapt and guide colleagues through an often complex decision-making process in this
area.

The current humanitarian data sharing environment makes it difficult for people to communicate or organize around
shared principles, concepts, processes or skills. For example, there is no typology for sharing to which people can refer;
nor is there an agreed ‘decision tree’ on how to move forward when considering data sharing risks or benefits. As such,
the Framework will promote an environment of common understanding and collaboration, conducive to sharing.

The framework will comprise two sections. The first will offer a series of shared principles along with a common ‘trust
statement’, shared concepts and core competencies in this area (below, section 1 - 4). The second section will define a
framework for data, information and analysis sharing for operationalizing those shared principles (below, section 5 -
5.5).

1. Shared Principles1: Although there exist a range of principles for sharing data, there is no agreed upon set of
principles to which humanitarian and other collectors have signed up. These principles, yet to be proposed, will build
on already agreed upon principles within humanitarian action such as those outlined in ICRC’s Professional Standards
for Protection in Humanitarian Action, the Protection Information Management Principles and the draft Privacy
and Data Protection Principles under development with Global Pulse, among others. They will cover issues of
data responsibility, privacy, protection and security.

1 Gwi-Yeop Son, Director Corporate Programmes for OCHA also requested colleagues to work together on a communication strategy which may
fit under a Trust Framework , articulating why data is integral to decision-making and critical to achieving results.

2
OCHA AND PIM TEAM CO-HOSTED DATA SHARING MEETING

A FRAMEWORK FOR
DATA SHARING IN PRACTICE:
PART ONE
New York, May 2017

2. Trust Statement: The ‘trust statement’ will be a high-level statement committing to work and share according to
these principles, with parameters guided by the UN Secretary-General. Such a statement may also extend to donors,
who have the responsibility and leverage to enable data sharing and cooperation among stakeholders.

The goal of the trust statement is that colleagues agree to work in a trusted fashion through the elements required to
establish a trusted environment, i.e. through a shared minimum approach as outlined in the elements of the
Framework below.

In this manner the Framework articulates a minimum level for shared processes, principles and standards which would
also be open to participation form both internal and external actors outside of the humanitarian community (e.g.
development, peacebuilding, human rights, academics, private sector, media, affected populations), based on the
needs of a given data sharing situation and the application of the Trust Framework.

If colleagues are not operating in a trusted environment, than it becomes more difficult to assess the risks and/or
benefits of data sharing.

3. Share d Conce pts: We need commonly de fine d conce pts to facilitate a discussion around data sharing. The
frame work propose s the de finition of a se t of conce pts. The following que stions ide ntify some conce pts that need
further clarification:
• What responsibility do data and information providers have for secondary uses made of data? How can the
purpose for data collection, sharing and secondary uses of data be defined.
• What are the risks as it relates to data, information and analysis sharing in a humanitarian context? What does a
taxonomy of ‘risk’ look like?
• What is the implication / what does ‘informed consent’ entail with respect to data and information sharing?
• When is data considered anonymous / anonymised?
• What is considered sensitive data?
• What is considered personal data?
• What is considered personally identifiable data/information?
What is considered demographically or community identifiable data/information?
4. Core Competencies: This re fe rs to the skills, knowledge and attitude re quire d to apply this trust framework to
facilitate the safe and responsible sharing and use of data. Not having the required skills to share data in a safe manner,
or undertake proceeding steps before and after the actual act of sharing presents another form of risk(s).

5. Process for Risk and Benefit Assessment: The ways through which we will work toward an overall reduction of
risk through share d conce pts, me thods and processe s will be propose d in this se ction, through a framework for
data, information and analysis sharing. Operational processes and approach for asse ssing, mapping and reflecting
on the risks and benefits of sharing data, including a continuous approach to assessing risk through a ‘shared minimum’

3
OCHA AND PIM TEAM CO-HOSTED DATA SHARING MEETING

A FRAMEWORK FOR
DATA SHARING IN PRACTICE:
PART ONE
New York, May 2017

approach, based on the trust statement and framework for data sharing and accountability. The specific components of
this structure are illustrated below in sections 5.1 - 5.6.

5.1 Purpose: The first step will be a consideration of how to define the protection outcome (value proposition) of
sharing data, based on a list of shared questions on motivations, risks and benefits, for partners and stakeholders. A
high-level decision tree will be developed to support this process. Shared questions may focus on the original purpose
for which the data was collected, the purpose for sharing and the purpose for which it will be used.

This section may also include questions such as: How does the core principle “on a need to know basis” translate into
practice? What protection standards and safeguards are needed to ensure this is understood and upheld?

5.2 Nature of Data - Typology: The second section will propose a method for mapping the data and information
ecosystem from a data systems perspective (i.e. who has what; what systems are in operation where; and what levels
and types of data, are these systems producing or capable of producing when cross-analyzed), in order to understand
what is needed and measure results that can support the achievement of a protection outcome. Neither risk nor benefit
can be assessed without understanding the ecosystem, which is wider than just the data sharing proposed.

This section will also propose levels and types of data required for critical decision-making, and will suggest a method
for how to support the sorting of data into actionable levels in support of this.

Mapping the data and information landscape at a high level also includes defining the different types of data. For
example, is the data personal, non-personal, sensitive and so on.

5.3 Temporality, Scenarios: This section contains scenarios through which to formulate discussions around purpose,
supporting the identification of what may be unique or common about a given situation, such as an acute onset
emergency, protracted emergency etc.

5.4 Sharing Stakeholders and Relations: This section will propose a typology of sharing and a method for the mapping
of intended use and the categorization of data collaborators in humanitarian contexts.

5.5 Sharing Methods: This section will look at the different mechanisms for sharing data, as the sharing mechanism is
also part of how you address and mitigate risk. This includes the different mechanisms that allow sharing of data in a
more closed or open setting, and other variables.

5.6 Tools and Practices: The component contains a selection of tools and practices (here ‘practice’ refers to the actual
application of a tool or method) that can be used to implement the data sharing arrangement as designed. This section
will also contain a description of roles that can be useful in setting up data sharing arrangements.

4
OCHA AND PIM TEAM CO-HOSTED DATA SHARING MEETING

A FRAMEWORK FOR
DATA SHARING IN PRACTICE:
PART ONE
New York, Ma y 2017

V. Next Steps
Part II of this document will be shared with colleagues who participated in the NYC meeting in the next month or two for
them to further refine and build upon. Thereafter, another meeting will be held in Europe, where we will continue to
build upon the work and conversations that were started in New York. The objective is to have a draft Framework for
Data Sharing in Practice by the end of the year.