,
analysis tools
and
data
governance
,
in an effort to avoid ambiguity in
the interpretation of the concepts set forth.
Chapter 1 also discusses data as public goods,
an aspect that is important when considering
improvements to capacity building, institutional
arrangements and coordination, which in
turn affect data governance arrangements. A
conceptual framework (
FIGURE 1
) is provided that
draws on previous work by the HLPE and others
(Bronfenbrenner, 1979; DFID, 1999; HLPE, 2017,
2020; UNICEF, 1990), linking food system policies
and actions to the food security and nutritional
status of individuals and the context in which
they live.
A key feature of the conceptual framework is
the distinction in levels based on the proximity
of the socioecological factors related to FSN
(and corresponding decision-makers) to the
individuals who are ultimately affected by FSN
5 This will be discussed more in later sections of the report, but
consider for example developments in the theory of measurement
that address the problem of quantification in behavioural and social
sciences (Bond, Yan and Heene, 2020; Mari
et al., 2017), or the
epistemological implications of big data for research (Kitchin, 2014b).
DATA COLLECTION AND ANALYSIS TOOLS FOR FOOD SECURITY AND NUTRITION
4
]
policies and actions. Furthermore, inspired
by and adapted from the data cycle presented
by Data for Decisions to Expand Nutrition
Transformation (DataDENT, n.d), the conceptual
framework identifies thematic areas for FSN
data collection and analysis and provides a
schematic representation of the main steps
to utilize FSN data for particular objectives.
These six steps along a
cycle for data-informed
decision-making
begin with identifying the
priority question and continue through using the
results, insights and conclusions (
SEE FIGURE 2
).
Effective data governance and inclusiveness are
described in depth, as highlighted at the centre
of Figure 2. This schematization complements
the conceptual framework as it highlights how
different actors use data to perform different
functions while illustrating how myriad roles
can coordinate efforts for FSN-related decision-
making. Together, the conceptual framework and
the data cycle help frame the discussion in the
subsequent chapters of the report.
The conceptual framework provides an effective
way to
guide the selection and analysis in an
organized manner
,
by completing a matrix
(
FIGURE 3
), where each step in the data cycle
is associated with the elements of the broad
system, from macro (or distal), to individual
outcomes.
Throughout the report, the conceptual
framework and the data-informed decision-
making cycle are used to highlight how data
and analysis tools relate to each of the six
dimensions of food security, as introduced by the
HLPE-FSN (HLPE, 2020).
It is important to note that the report adopts
a
broad definition of data
that includes all kinds of
information –
both quantitative and qualitative
– that can be
codified, stored and transmitted
in analogue or digital form
, and recognizes the
risks and limitations associated with exclusive
reliance on quantified variables in informing
decisions.
Chapter 2
reviews existing data and analysis
tools for FSN. Despite an abundance of FSN-
relevant data at all levels,
there is a lack of
broad, shared access to the disaggregated,
granular data, at subnational and local levels,
needed to inform action
. Existing data could
be better shared and analysed, by both public
and private agents at national and international
levels, in order to extract the wealth of useful
information it contains. This will require a
rethinking of the way FSN data are governed,
especially considering the rapidly changing
data ecosystem
, described later in the report.
The review of existing FSN data collection and
analysis initiatives provides various examples of
good practices that could be further enhanced
and used in developing similar initiatives.
The review also identifies the most important
remaining data gaps and challenges at each
step of the data cycle, such as: data on the
characteristics of agricultural holdings, such as
those produced by agricultural censuses; data
on the different characteristics of farms and
other operations across the agrifood system
at the local level, as provided by farm and
other industry surveys; data on household food
expenditure; and, most importantly, data on
individual dietary intakes. These kinds of data
are essential to guide targeted FSN intervention,
as they provide focused insights on local food
systems and on the extent of inequalities within
populations. While surveys and other sources
of household and individual level data exist,
the quality of the data they provide, and the
frequency with which they are generated, are still
largely insufficient to support effective decision-
making, especially in low- and middle-income
countries, and to conduct assessments during
emergencies and in other difficult contexts.
The second part of Chapter 2 discusses current
challenges and opportunities to improve data-
informed FSN decision-making at each step
of the data cycle. One finding is that
there
is a general lack of clarity and coordination
among decision-makers with regard to setting
priorities when deciding on data collection and
analysis
, and this stands in the way of filling in
current data gaps. Better coordination in setting
objectives for data use will contribute to creating
an enabling environment, where institutions at
various levels work together to gather, curate
and disseminate data. This will be instrumental
to favour increased access to existing data and
[
5
to prevent the unnecessary proliferation of
indicators, data-collection initiatives, and data
quality assurance which result in fragmented
data results that are difficult to reconcile and
that are inadequate for informing effective
action.
Of special note is the importance of qualitative
information for making decisions. A myriad
of personal, societal, cultural, religious and
other considerations may have direct relevance
for decision-making to improve FSN. Many of
these aspects may be difficult or impossible to
capture with quantitative data, and qualitative
data are less amenable to collection by simple,
standardized surveys, with the result that this
type of information may end up being excluded
from data consolidation and dissemination
efforts. A final aspect involves communication
and the importance of communicating data and
the results of data analysis in a way that it is
useful and effective for decision-making.
Chapter 3
discusses the major constraints and
bottlenecks that underpin many of the gaps
in FSN data collection and analysis identified
in previous chapters, with a special focus on
conditions prevailing in low– and middle–income
countries. The constraints are grouped into two
main categories: those related to insufficient
resources – financial, human capital and data/
research/analysis infrastructure; and those
related to inadequate institutional arrangements,
which lead to problems with data governance.
Timely allocation of sufficient financial
resources, in a predictable way, is a key
enabling element to sustain an effective FSN
data ecosystem in any country
. Notwithstanding,
this is a serious problem reported by many
countries, where National Statistics Offices
(NSOs) identify funding as one of their main
constraints, in particular in the agriculture
sector. Resource constraints continue to limit
data collection in agriculture (where sound
decision-making requires regular agricultural
censuses and surveys of operations along the
food supply chain
), and in food security and
nutrition outcomes (where up-to-date household
surveys and dietary intake information are
needed). Although it is recognized that these
are expensive initiatives, that demand adequate
levels of human capacity, they are essential as
they constitute the backbone of any FSN data
system.
Chapter 3 also highlights the trade-offs between
the financial and human resources needed to
secure adequate generation of quality data:
while the running costs of field operations, data
storage and dissemination might be reduced by
shifting from more traditional operations (as still
conducted by many National Statistics Offices
and other government statistics units in low-
income countries) to modern data-generating
technologies and digitalization, the process
must be accompanied by upfront investments
(infrastructure, machinery, etc.), but also by
the development of the necessary professional
capacity. Effective use of modern technologies
for FSN data generation and analysis requires
skills that are still in scarce supply.
The lack
of adequate investment in human capital,
namely, expanding education on data science
and statistics to all professionals involved
in the FSN data-informed decision-making
cycle, is the strongest binding constraint that
prevents FSN data systems from developing in
most low-income countries
. Thus, it is the area
where investments will certainly have the highest
returns.
In terms of institutional arrangements, we note
the
lack of coordination among the various
agencies that are involved in generating and
analysing FSN-relevant data
, which operate
often under different administrative and logistic
arrangements, for example, as units in different
ministries (agriculture, health, economy,
environment, etc.). This often results in costly
duplication of efforts, leading to redundancy and,
sometimes, inconsistency in the information
generated by different units. This problem is
not only present among government institutions
at country level, but also in academia, and
sometimes among international organizations,
including within the UN System. The review leads
to a strong call for increased coordination at all
levels, from local, to national, to international,
something to which we shall return to in
chapters 5 and 6.
INTRODUCTION
DATA COLLECTION AND ANALYSIS TOOLS FOR FOOD SECURITY AND NUTRITION
6
]
A discussion of data and analysis tools cannot
be complete without recognizing that we are in
the midst of a data revolution, including within
agriculture and FSN.
Chapter 4
reviews how
new and emerging technologies in digital data
can contribute in many ways to FSN data and
analysis, though perhaps requiring that the
traditional ways of thinking about and regulating
activities around FSN data collection and
analysis be challenged, including the roles and
responsibilities of public and private actors.
Several examples demonstrate how each of
these technologies can contribute to each
phase of the cycle for data-informed decision-
making, and how they may provide information
that is relevant for each of the six dimensions
of FSN. The review confirms that
these
technologies have the potential to make a huge
contribution, though their broad diffusion also
comes with risks
. These include uncontrolled
dissemination of digital data collected through
devices embedded in machines (from tractors to
personal phones), which can threaten privacy;
problems of accountability arising from reliance
on artificial intelligence, machine learning
and other automated or semi-automated
decision-making, which raises a number of
ethical considerations regarding the use of
these modern technologies; data quality and
interoperability issues which may be conditioned
by the specific technology used; and, finally,
the very important issues of equity, scalability
and inclusiveness that arise when considering
the differential capacity that exists both across
countries and between public and private actors/
institutions.
Many of the issues raised and discussed
in the previous chapters lead naturally to
considerations around
data governance
, to
which
Chapter 5
is devoted. The chapter begins
by addressing two somewhat controversial,
and strongly interlinked, issues around data
governance. One is the debate on the nature
of data: should data be considered public or
private goods, and what role can markets
play in this? Are market-based mechanisms
able to guarantee an adequate supply of and
access to data? The other issue is the question
of data ownership and the social value of
data. Especially when data contain personal
information, who should own it? And if the data
are considered to be owned by the people to
whom the information is linked, should they
have the right to sell it? With specific reference
to FSN, there are convincing arguments that
more disaggregated data are needed to better
guide FSN interventions, but that such data
might allow personal or group identification,
in which case the data would be considered
“personal data”. The question arises, then, as
to whether current mechanisms for personal
data protection, such as those based on
informed consent, are sufficient to protect
the rights of data owners, while ensuring that
the information can be accessed to express
its full potential for social benefits. One key
suggestion in this report is that, from a moral
standpoint,
personal data, like blood, are
something that individuals may choose to give
when that is necessary to obtain a personal
service
(for example, when blood is given for
medical testing),
but that people should also
be encouraged to donate, when there is a clear
indication that its use may contribute to a
greater good
(such as saving someone’s life).
What should be crystal clear is that any resale
of such data should be deemed immoral and
even prosecuted as illegal. The main conclusion
from the discussion in the first part of the
chapter is that, because modern data that are
recorded, stored and shared in digital forms,
can be used and re-used, even simultaneously
by many people, they must be conceived as
inherently
public goods
.
Access to such data
should be restricted only when necessary to
protect fundamental human rights, such the
privacy of the people involved. For this purpose,
innovative legal frameworks, such as those
based on the concept of
data trusts,
defined
as “legal structures that provide independent
stewardship of some data for the benefit of a
group of organizations or people” (Open Data
Initiative, 2018), are a promising option for
moving the data governance agenda forward,
including in the agriculture sector and with
regard to FSN data.
Fortunately, this is indeed a very active area of
research and debate, and the chapter presents
[
7
INTRODUCTION
examples of existing initiatives, which may serve
as models for yet more solutions.
Finally,
Chapter 6
summarizes the findings of the
report and advances the recommendations as a
call for action to all actors who play a role in the
data cycle. Recommendations are organized in
five areas based on the objectives of: (i) creating
greater demand for data in decision making;
(ii) optimizing and, if needed, repurposing
investments towards data collection, while
increasing collaboration among stakeholders
to harmonize and maximize the sharing of
existing FSN data; (iii) increasing and sustaining
investments in essential FSN data collection;
(iv) investing in human capital and infrastructure
to ensure sustainability of data processing and
analytic capacity; and (v) improving FSN data
governance and promoting inclusiveness and
agency among data users and generators. The
proposed actions, if followed, may prove useful
in moving towards more effective, evidence
informed, decisions that will make food systems
more sustainable and ensure food security and
better nutrition for all, particularly for the billions
of people throughout the world who suffer
from various forms of malnutrition, including
the seven hundred million or more who still
experience hunger (FAO
Dostları ilə paylaş: |