Data collection and analysis tools for food security and nutrition

, 
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