Data Collaboration and Stewardship for the Blue Economy

CONFIDENTIAL + WORKING DRAFT
TIDEPOOL DIGITAL
DATA SHARING STUDIO
Governance Models of Data
Collaboration
Stefaan Verhulst and Andrew J. Zahuranec
The GovLab

INTRODUCTION
01
2
GOVERNANCE MODELS OF DATA COLLABORATION

THE GOVLAB
3
The GovLab’s mission is to
transform how we make decisions,
leveraging new technologies and
science
● Its Data Program focuses on how to
access and leverage data responsibly to
generate new insights that can inform
decisions and problem solving .
https://thegovlab.org/

THE EMERGENCE OF
DATA COLLABORATIVES
4
A new form of collaboration, beyond the
public-private partnership model, in which
participants from different sectors exchange
their data to create public value.

CONFIDENTIAL + WORKING DRAFT 5
THE VARIABLES OF
DATA COLLABORATION
Aspects of Data Collaboration
● Data Accessibility:
○ Level of Openness
○ On-Off Line
● Data Attributes:
○ Raw/Pre Processed vs Insights
○ Single vs Multiple Data Providers
○ Single vs Multiple Data Sets
● Collaboration Dynamics:
○ Uni -vs Multi Directional
○ Directed vs Independent
● Scope
○ Purpose-bound vs Flexible
○ Time-bound vs Open Ended
○ Budget

THE MATRIX OF
CONDITIONALITY
Defining Variables
Y-axis: Level of Openness
● Open vs Restricted
X-axis: Level of Collaboration
● Collaborative vs Independent

TRUSTED
INTERMEDIARIES
TRUSTED DATA COLLABORATIONS
Tidepool digital enables sustainable ocean
innovation by supporting data sharing across
the public and private sector.
Third-party actor mediates collaboration
between (private sector) data providers and
data users from the public sector, civil society,
or academia.
● Data Holders: One or More
● Data Users: One or More
● Main Types:
○ Data Brokerage
○ Third-Party Analytics Projects
Example: POPGRID Data Collaborative
(Data Brokerage)

TRUSTED INTERMEDIARY: DATA BROKERAGE
POPGRID Data Collaborative
Planning Collecting Processing Sharing Analyzing Using
Columbia University noted
a lack of consolidated
georeferenced data on
population, human
settlements, and
infrastructure for research.
It creates POPGRID, which
provides a platform for 10
research centres and
government institutions to
exchange georeferenced
data for their benefit and
the public’s benefit.
Individual institutions clean
the data based on their
own methodologies in
ways to allow side-by-side
comparison. POPSTAT
ensures inputs are high
quality.
The data is exchanged and
made viewable through
the POPGRID Viewer, a
dashboard that allows
visitors to examine and
compare different inputs.
Researchers from various
institutions use this
collection to develop new
research on human
mobility and settlements.
Organizations like
the US Census
Bureau use the
insights to inform
their policies and
estimates.

TRUSTED INTERMEDIARIES:
OTHER TYPES OF PARTNERSHIPS
THIRD-PARTY
ANALYTICS
Third-party analytics projects
see trusted intermediaries
access private-sector data,
conduct targeted analysis, and
share insights, but not the
underlying data, with public or
civil sector partners. This
approach enables use of data
in the public interest while
retaining strict access
controls.
“We built an interactive tool that tracks mobile phone
top-up spending as a non-food household
expenditure, and market prices of crop staples across
the Greater Kampala Metropolitan Area.”
Dahlberg Analytics

DATA POOLING
TRUSTED DATA COLLABORATIONS
Data holders pool datasets as a collection
accessible by multiple parties.
● Data Holders: Some Data Holders
● Common Pool: Some Data Users
● Main Types:
○ Public Data Pools
○ Private Data Pools
Example: California Data Collaborative
(Private Data Pool)

DATA POOLING: CLOSED POOL
California Data Collaborative
Water supply in California is
strained but utilities lack
clean, reliable data outside
their jurisdiction that can
help them make decisions
and follow regulation.
Different water utilities
agree to form a nonprofit
that hosts a private pool of
their different water-meter
data collections. Only
collaborative members
have access to the raw
collection.
Members to the
collaborative maintain a
central staff who maintain
and clean data uploaded
for analysis and
visualization.
New water utilities can
apply to join, after paying a
fee. If they are accepted
by existing members, they
can, in turn, opt to share
data with verified research
partners.
Water managers use the
data compilation to answer
specific governance
questions or seek support
from CaDC staff.
Water managers act
to realize new water
efficiencies based
on the analysis
provided.

DATA POOLING:
OTHER TYPES OF DATA POOLS
PUBLIC DATA POOLS
Public data pools co-mingle
data assets from multiple data
holders and make those
shared assets available on the
web. Pools often limit
contributions to approved
partners, but access to the
shared assets is open,
enabling independent uses
“The Global Fishing Watch map is the first open-access
online platform for visualization and analysis of
vessel-based human activity at sea.”

RESEARCH AND
ANALYSIS PARTNERSHIP
Pairing between (private sector) data providers
and data analysts/users from the public sector,
civil society, or academia.
● Data Holders: One or More
● Data Users: One or More
● Main Types:
○ Data Transfer
○ Data Fellowship
Example: UN Food and Agriculture Organization’s
Building communities resilient to climatic extremes
(Data Transfer)

RESEARCH AND ANALYSIS PARTNERSHIP: DATA TRANSFER
UN Food and Agriculture Organization’s Building Communities Resilient to Climatic Extremes
UN FAO needed data to
better estimate climate
displacement in Latin
America, particularly
Columbia.
It contacted Telefonica, a
major telecom provider
with Call Detail Records
from 11+ million customers
in Columbia.
Telefonica aggregated and
anonymized its CDR data
and combined it with
relevant open and
government data for FAO
Telefonica provided a
platform for FAO and the
Government of Columbia
to monitor the data
regularly.
Using a mobile data
analysis platform,
researchers found 12,000
people left the La Guajira
region due to drought.
FAO will use this
insight to develop
support for IDPs and
better manage
natural resources

RESEARCH AND ANALYSIS PARTNERSHIP:
OTHER TYPES OF PARTNERSHIPS
DATA FELLOWSHIP
Companies provide
opportunities for their staff to
support another organization.
These opportunities often
involve one organization’s
researchers being embedded
in another organization to
analyze their data assets or
otherwise support their work.
“Through Google.org Impact Challenges, we award
nonprofits and social enterprises with support to help
bring their ideas to life.”

Data Stewardship:
Data Stewardship is about
making Data Collaboratives
- Systematic
- Sustainable
- Responsible
DATA COLLABORATIVES & OCEAN DATA

GOVERNING DATA
COLLABORATIVES
02
17

THE FOUR Ps:
● Purpose
● Principles
● Processes
● Practices

1a. PURPOSES:
ORGANIZATIONAL PURPOSES
19
● What is the purpose/mission of the
data collaborative (beyond providing
access to data) ?
● Who does the data collaborative seek
to inform/serve?
● What impact or action will be enabled
by the data collaborative that could not
be pursued by other means?
TOOLS:
- Problem Definition
- Theory of Change
- Logic Models
Theory of Change Model
Open North

1b. PURPOSES:
DATA ACCESS PURPOSES
● What’s the purpose/topic for which
access to data is needed?
● Who are the various stakeholders that
have or need data?
● What topics and/or questions need to be
prioritized?
TOOLS:
- Topping Mapping
- Question/Agenda Setting
- Voting/Prioritization
EXAMPLE: Topic Mapping

The 100 Questions Initiative

2a. PRINCIPLES
ORGANIZATIONAL PRINCIPLES
22
What are the principles that will inform decisions
at the organizational level (to meet the
purpose)?
Legitimacy
● Participatory, representative, transparent,
and accountable
Effectiveness
● Agile, cost-effective, impactful
Other
● Trusted
● Distributed
Example: PLACE Trust
Design Principle
LEGITIMATE TRANSPARENT: Openness and transparency, including procedural and financial
transparency
PARTICIPATORY: Strong participation by all affected stakeholders on recommending
criteria and policies that support the strategic goals
REPRESENTATIVE: Adequate engagement and representation of key groups
ACCOUNTABLE: Trustees, PLACE Board, and PLACE Members are held accountable
to their actions, and effective independent review procedures set in place
EFFECTIVE FOCUSED: A bounded and well-defined mission
COLLABORATIVE: Complementing existing data, standards and policy work and
collaborating closely with existing government departments, data communities, private
sector and other organizations
EFFECTIVE: Sufficient support and resources to implement governance principles
RECIPROCITY: Implement iterative and dynamic processes so the Trust can evolve
with its environment, while remaining agile and effective
TRUSTED TRUSTED: Strong strategic direction that can support PLACE Trust in a financially
sustainable way, coupled with public interest leadership.
ACCESSIBLE: Minimizing barriers to participate and accept new members
ETHICAL: Minimizing harmful impact on individuals, communities, organizations or the
environment

2a. PRINCIPLES
DATA HANDLING PRINCIPLES
● What are the principles that will inform
how data is handled?
● FIPPS (OECD)
(1) The Collection Limitation Principle.
(2) The Data Quality Principle.
(3) The Purpose Specification Principle.
(4) The Use Limitation Principle.
(5) The Security Safeguards Principle.
(6) The Openness Principle.
(7) The Individual Participation Principle.
(8) The Accountability Principle.
Example: Fair Information Practice Principles
Example: Health Data Governance Principles

3a. PROCESSES
ORGANIZATIONAL DECISION-MAKING
24
● What kinds of design making processes are
needed to ensure an organization can act on
its principles to meet the purpose?
○ Trustees/Governing Board
○ Working Groups/Committees
○ Data Assembly
Example: The Data Assembly

3a. PROCESSES
DATA LIFECYCLE DECISIONS
● What are the key decision processes and
makers across the data life cycle?
○ RACI assessment
○ Decision Provenance
Example: Decision Provenance
Example: Institutional Review Board

4a. PRACTICES
FRAMEWORKS

4a. PRACTICES
MECHANISMS
● Mechanisms to institutionalize or
operationalize principles such as
○ Contracts/Data Sharing Agreements
○ Codes of Conduct
○ Terms and Conditions
○ Technologies
Example: Code of Conduct
Example: Terms and Conditions

CONCLUSION
28
● Advancing reuse of data requires new
partnership models where data from many
sources—including proprietary business
data—is made accessible to generate insights
and address societal challenges. Data
collaboration is one such model.
● These data collaboratives can take a number
of forms but knowing which form is most
useful requires understanding the context and
problem statement.
● Implementing this model responsibly and
effectively requires aligning clear purpose and
principles with processes and practices.

LEARN MORE:
https://datacollaboratives.org/
29

Data Collaboration and Stewardship for the Blue Economy

Recommended

Recommended

More Related Content

Similar to Data Collaboration and Stewardship for the Blue Economy

Similar to Data Collaboration and Stewardship for the Blue Economy (20)

More from Stefaan Verhulst

More from Stefaan Verhulst (7)

Recently uploaded

Recently uploaded (20)

Data Collaboration and Stewardship for the Blue Economy