Data collection and analysis tools for food security and nutrition

Nutrition report (FAO et al., 2017). Despite the coordination and standardization achieved through the SDG monitoring 
framework, its implementation is incomplete in 2022, more than halfway through the SDG time frame. For example, 
the reporting rate for the 21 SDG indicators under FAO custodianship, in 2020, was only 51 percent (FAO, 2020a). 
Moreover, in many cases, the indicators reported by countries do not adhere to the standard definitions and have 
been replaced by proxy indicators. This hinders cross-country comparisons and may lead to misinterpretation of 
results in terms of progress made.
[
37
Despite the growing recognition of the costs of 
not basing policy decisions on data, this is still 
not a widespread practice with respect to FSN. 
Drawing on examples from business, Gartner, 
a business support company, estimated in 2018 
that poor quality data costs businesses an 
average of USD 15 million per year in losses. 
When profits are the bottom line, as is the case 
in business, these figures make a compelling 
case for better data. Gartner proposes a 5-step 
process to develop the business case for better 
quality data: understand business priorities; 
carefully select the right metrics; develop the 
approach to consolidating and using the data 
from the outset, including benchmarks; set 
targets; estimate the financials – both cost of 
data quality improvements and the quantified 
benefits of using it.
10
The first four steps are 
similar to those of our data cycle. The FSN data 
10 For more information, see 
https://www.gartner.com/
smarterwithgartner/how-to-create-a-business-case-for-data-
quality-improvement
.

---

DATA COLLECTION AND ANALYSIS TOOLS FOR FOOD SECURITY AND NUTRITION
community has been remiss not to quantify the 
cost of not using data and the potential cost 
savings of doing so. We did find one example 
of this in the literature relevant for FSN data. 
In Cameroon, researchers used optimization 
modelling to illustrate the potential impacts 
and cost savings of alternative approaches 
to addressing vitamin A deficiency in the 
population. The data collection and related 
expenses required for the modelling cost 
approximately USD 900 000. The authors 
estimate, however, that this reflects only 5 
percent of the cost savings if the government 
were to implement the modified programmatic 
approach resulting from the modelling over 
a period of 10 years (Vosti 
et al., 2015). Such 
estimates could substantially mitigate the 
concerns related to the cost of data collection.
These examples demonstrate that the existence 
of data, the cost of not using data and even the 
linkage of data to agreed targets are insufficient 
for that data to be brought to bear on FSN 
decision-making. In order to address this, it 
is necessary to examine the challenges and 
opportunities across the data cycle. These are 
set forth in the following sections.
SET PRIORITIES FOR DATA
•
Lack of clarity on how to prioritize:
The various 
sources of data highlighted in Annex Table 1 
illustrate the abundance of topics that hold 
relevance for FSN. The criteria used to set priorities 
for data gathering or compilation are often allusive. 
This is due, at least in part, to the multitude of 
reasons for which existing data have been collected 
and the multitude of purposes for which data 
systems are compiled and used. Articulating clear 
objectives for data utilization and being explicit in 
the types of data-informed decisions to be made, 
and by whom, can help navigate the abundant data 
and identify gaps. For example, the Countdown to 
2030 initiative (
BOX 10
) was developed to prioritize 
data for advocacy and accountability for women’s 
and children’s health and to enhance the capacity 
for its utilization.
38 
]

---

2 
A REVIEW OF EXISTING FSN DATA COLLECTION AND ANALYSIS INITIATIVES
BOX 10:
COUNTDOWN TO 2030
The 
Countdown to 2030 initiative
aims to improve coverage, measurement and monitoring of health interventions for 
women, children and adolescents, and to strengthen¬¬ the regional and country capacity for evidence generation in 
this regard. It builds on the Countdown 2015 initiative that was set up to enhance accountability for the related 2015 
Millennium Development Goals. Countdown seeks to strengthen in-country evidence and analytical capacity, creating 
partnerships among global, regional and country analysts from public health institutions, research institutions and 
ministries of health.
As of the writing of this report, Countdown 2030 has established data collaborations in 19 countries in Africa and 
Asia (see 
https://www.countdown2030.org/country-collaborations
). They have held numerous workshops with 
participants from over 150 countries, published many documents, reports, technical notes and other medium on 
good practices related to data and on the results of the data themselves. The latter have been included in global 
monitoring and accountability reports, such as 
Leaving No One Behind
, the 
UN Every Woman Every Child progress 
reports
, the 
UNFPA State of World Population Report
 and the 
Global Nutrition Report
.
Prioritization of data for inclusion in the Countdown efforts is enhanced by the establishment and following of a clear 
set of guiding principles. Coverage, that is, the proportion of individuals needing a service or intervention who receive 
it, is the central focus of Countdown. Data are tracked only for interventions that have been scientifically proven to 
reduce mortality among women and children and are feasible for delivery in low- and middle-income countries. 
Data are also collected for coverage of services that serve as delivery platforms for interventions such as antenatal 
care and family planning, among others. Included interventions must have coverage indicators that are reliable and 
validated across multiple country contexts and over time. Countdown does not collect primary data, so sources must 
be from nationally representative surveys and must be regularly available for inclusion.
Fanzo 
et al. (2021) call for the establishment of a similar rigorous, science-based monitoring framework that can 
provide a countdown on advances to transform food systems for nutrition. The authors propose the adaption of the 
HLPE-FSN food systems framework to guide priority setting for data inclusion, focused around five thematic areas: 
diets, nutrition and health; environment and climate; livelihoods, poverty and equity; governance; and resilience and 
sustainability. Setting such priorities for data that are both relevant for policymaking and feasible to collect rigorously 
across settings is an important first step to establishing a data system that can support accountability and inform 
decision-making.
[
39

---

DATA COLLECTION AND ANALYSIS TOOLS FOR FOOD SECURITY AND NUTRITION
GATHER, CURATE AND DISSEMINATE 
DATA
•
Lack of availability and access to data:
As 
highlighted previously, both availability of and 
access to data continue to be important constraints 
for some of the domains relevant for FSN. Some 
of the data sources listed in Annex Table 1 are 
proprietary, generated and held by private-sector 
data firms. Even public data (e.g. some national 
surveys, or information on the extent of food 
reserves) may be held behind firewalls that restrict 
access to authorized users, or there may be lengthy 
delays before such data are made publicly available. 
For several topics in the table, we were able to 
identify reports that consolidate relevant data, 
but the data themselves may not be available and 
accessible in the public domain. To improve data 
sharing and accessibility, having clearer objectives 
and setting priorities could help adapt existing 
data systems, focusing on the most important 
gaps and exploring feasible solutions. However, 
as discussed in Chapter 5, some of the problems 
in data sharing derive from unresolved issues 
regarding data governance and the associated 
legal and ethical aspects of 
open data
. Such 
issues are well recognized, and several initiatives 
are already underway to address them, including 
the previously mentioned 
Agricultural Market 
Information System
(AMIS) (
BOX 2
) and the 
Global 
Open Data for Agriculture & Nutrition
 (GODAN) 
(
BOX 11
). Another relevant example is the 
Rural 
Livelihood Information System
, a joint initiative 
of the FAO Statistics Division, the World Bank and 
IFAD, to support policies for reducing rural poverty. 
This system provides open access to standardized 
indicators produced from household surveys and 
time series data from official national statistics. 
Additionally, a very promising initiative has been 
recently established as a collaboration between the 
nutrition, fisheries and statistics divisions at FAO, 
aimed at creating a new food and diets domain in 
FAOSTAT that will disseminate harmonized food, 
diet and nutrient statistics from different data 
sources (food balance sheets, household surveys 
and dietary intake surveys).
DATA ANALYSIS
•
Poorly conceived or inappropriate measures, 
indicators or scales:
As discussed in Chapter 1, 
it is common that insufficient attention is given to 
defining FSN-relevant constructs and to selecting 
appropriate measures, indicators and scales. 
This is particularly pertinent as there is often 
a lack of transparency about the cross-context 
validity of measures, indicators and scales. Lack 
of agreed-upon standard vocabulary, measures 
and constructs makes data collection, analysis and 
interpretation very problematic. How to effectively 
measure household food security in surveys, 
for example, has been long debated. Existing 
BOX 11:
GLOBAL OPEN DATA FOR AGRICULTURE AND NUTRITION (GODAN)
Few would question the importance of improving data access, but insufficient attention is often paid to why data are 
not accessible and to the policies, procedures and institutional arrangements that constrain or act as disincentives to 
make data accessible. The 
Global Open Data for Agriculture and Nutrition
 initiative seeks address these challenges 
by building high-level policy and public and private institutional support for open data. GODAN is an innovative 
voluntary alliance of over 1 000 national governments, non-governmental and international organizations, and private 
sector companies. Members contribute directly to GODAN activities, which include guiding and assisting organizations 
and companies to develop open data policies, advocating for access to data and linking partners to required technical 
expertise. The GODAN website provides several tools useful for organizations interested in developing open data 
policies, and holds a repository of those that have such policies, providing convenient links to access them. Additional 
resources include training courses, webinars and certification of open data policies and procedures.
40 
]

---

[
41
2 
A REVIEW OF EXISTING FSN DATA COLLECTION AND ANALYSIS INITIATIVES
approaches to food security measurement were 
found to be incapable of gauging shocks and had 
poor nutritional relevance a decade ago (see for 
example Headey and Ecker, 2012) and some of 
the challenges identified remain (Cafiero, 2020; 
Cafiero 
et al., 2014). Similar problems exist with 
respect to other variables targeted in surveys. This 
shows how important it is to develop standards 
and to use standardised methodologies in surveys 
to ensure comparability. Since it was introduced 
by FAO in 2014, through the 
Voices of the Hungry
project, the Food Insecurity Experience Scale 
(FIES) (Cafiero, Viviani and Nord, 2018) is rapidly 
gaining acceptance as a relatively inexpensive, yet 
theoretically and empirically valid, survey tool to 
be included, with proper adaptation, in virtually any 
household or individual survey.
•
Inadequate data-collection designs:
The usability 
of data may depend heavily on decisions made at 
the design stage. Cross-sectional designs for FSN 
surveys, for example, may limit the usefulness 
of data collected for evaluating causal drivers of 
food security and nutritional status and changes 
over time. Farm and population surveys are often 
designed in ways that make them less inclusive of 
what would be needed to represent the diversity 
of the reality they intend to represent. Household 
survey frames based on lists of residents, for 
example, exclude the homeless and incorrectly 
represent transient populations in FSN surveys. 
Also, special survey design may be needed to 
adequately represent Indigenous Peoples who 
live in remote areas or do not have formal titles 
to housing or land. Similarly, agricultural data 
mainly obtained through interviews with farmers or 
through surveys, may lack sufficient representation 
of small farmers (Lowder, Skoet and Raney, 2016). 
Surveys designed to collect information by only 
interviewing the head of household, predominantly 
an adult man, fail to adequately record information 
on the relevance of women’s activities with regard 
to crops, income, needs (including childcare) and 
decision-making. As a result of these various 
issues in data collection, some FSN data may not 
represent the reality of small farmers, women, 
migrants and Indigenous Peoples. Furthermore, 
available data may be insufficient because they 
lack the granularity needed to make group-specific 
decisions. For example, data on the nutritional 
status of specific groups over a long enough period 
to track trends may be unavailable, although 
overall estimates may exist. The level of granularity 
required to evaluate disparities in nutritional 
status by gender, for instance, is currently absent 
in many countries (MTR Foresight report 2020; 
UNSCN, 2018). Individual level data are needed 
to track progress not only on food security and 
nutrition, but also on gender equality and women’s 
empowerment in food and nutrition spheres.

---

DATA COLLECTION AND ANALYSIS TOOLS FOR FOOD SECURITY AND NUTRITION
BOX 12:
AN EXAMPLE OF AN AFFORDABLE, GLOBAL DATA MANAGEMENT PLATFORM: REDCAP
REDCap stands for Research Electronic Data Capture. REDCap was created by the University of Vanderbilt in the 
United States. The secure REDCap web application permits users to create and manage surveys and the associated 
databases quickly and securely, including by collecting data offline. REDCap is used all over the world. At the time of 
this publication, it is used in 145 countries, by 5 961 institutions, working on 1.5 million individual projects, with a user 
base of 2.1 million users. REDCap is designed for use by institutions. Installation is free and the only costs involved 
are for server space and for certificates of security. However, licenses are only issued to non-profit organizations 
having sufficient internal IT infrastructure to self-host. According to the REDCap website, it is not permissible for a 
business, company or other for-profit organization to hold a license or self-host. However, REDCap Cloud (
https://
www.redcapcloud.com/
) is a third-party company which offers fee-based hosting in their custom version of REDCap. 
Because a license is always required to gain codebase access, REDCap is not considered open source. However, 
the license or institutional agreement, codebase and all consortium support are provided at no cost to any non-
profit organization. REDCap works as a programme for data entry within data (survey) forms previously created by a 
REDCap user. There are multiple benefits of REDCap over spreadsheets used for data entry. Primarily, REDCap allows 
for simultaneous data collection, online or offline, and data management. REDCap offers many advantages over 
spreadsheets as all the following features are much more difficult to execute in spreadsheets and related programmes:
• 
Allows for multi-language management: “Create and configure multiple display languages for projects, surveys, 
data entry forms, alerts, survey invitations, etc. Data collection instruments may be designed to display in any 
language that has been defined and translated, so that data entry persons can view the text in their preferred 
language. This eliminates the need to create multiple instruments or projects to handle multiple languages. When 
entering data on a data entry form or survey, users and participants will be able to choose their language from 
a drop-down list to easily switch to their preferred language for the text displayed on the page. All text related to 
the data entry process (both for surveys and for data entry forms), various survey settings and email text can be 
translated.” (See 
https://www.project-redcap.org/software/
).
• 
Validating ranges for dates and numbers (for instance, a mother in a study cannot be born the day of the survey or 
the number of portions in a food package is probably not more than 250).
• 
Standardized variable names (one institution could call question 1-date on the same form “q1” and another 
institution “date”).
• 
Allowing for double data entry (for example, each of these birthdates is plausible for a mother in a study 
(15/10/1976 and 15/12/1976), but without double data entry, it would not be possible to determine the correct 
birthdate that was written down on the paper survey form.
• 
Designing data entry forms that are nearly identical to those on paper that facilitates the speed and the precision of 
data entry, for instance, through the easy incorporation of skip patterns.
• 
REDCap offers a long-lasting data storage, prevents potential errors in handwriting information and minimizes 
potential errors arising during data entry by typing, for instance, by having standardized codes for questions with 
more than one response (such as categorical variables: 0 = No health claim, 1 = Health-related ingredient claim, 
2 = nutrient content claim, etc.). Such codes are often not standardized within and between countries, which often 
delays or inhibits harmonized data analysis because there are not enough human resources available to both clean 
and standardize databases within or between countries.
42 
]

---

2 
A REVIEW OF EXISTING FSN DATA COLLECTION AND ANALYSIS INITIATIVES
• 
Export of data ready for analysis within a statistical programme (that is, little or no clean-up of dataset prior to 
analysis).
All of these characteristics greatly facilitate pooled analysis of data from multi-country studies, especially 
longitudinal studies and those with mother-child dyads, because all data entry and related data management has 
been standardized when the data was collected.
Confidential data stored on REDCap are secure. REDCap’s webpage – while based on an open-source platform 
– is hosted through an institution that must have certificates of security to enter into an installation agreement 
with REDCap. Therefore, all the data entered into REDCap are saved within a secure server with extra antivirus 
protections on the REDCap server hosted by the given institution. Only institutional administrators can access all 
the data and study-specific data can only be accessed by those involved in the study or project who have the required 
rights and permission.
See 
http://project-redcap.org/
for the installation guide for institutions and 
https://projectredcap.org/about/faq/
for 
more information about REDCap.
•
Lack of harmonization and poor data quality:
Data collection, processing and storage protocols 
often vary considerably by context and over time, 
limiting the utility of the data to analyse trends and 
to identify specific areas of risk and vulnerability. 
Even simple common data types (such as dates) 
are often collected in non-standard ways, creating 
issues for merging or comparing data sources. 
Data cleaning protocols (including data range 
checks, treatment of out-of-range data and many 
other considerations) are not always applied or 
vary substantially in their approach across data 
sources. One relevant example of this is the 
lack of harmonization in the way in which food 
consumption data are captured in household 
consumption and expenditure surveys. To address 
this, the Inter-Agency and Expert Group on food 
security, agricultural and rural statistics (now 
known as the United Nations committee of Experts 
on Food Security, Agricultural and Rural Statistics 
[UN-CEAG]) convened a series of technical 
workshops, between 2014 and 2016, involving 
professionals and decision-makers from national 
statistical and international agencies to discuss 
solutions. The process led to the publication of a 
set of guidelines
on collecting food -consumption 
data in household consumption and expenditure 
surveys for low and middle income countries, which 
was endorsed by the forty-ninth session of the 
United Nations Statistical Commission in 2018 (FAO 
and The World Bank, 2018).
•
Timeliness:
Where primary data are needed, 
data collection and analysis can be a slow process 
and data may not be available in a timely manner 
for decision-making. This may be particularly 
problematic in emergency and crisis situations 
where analyses are needed to inform immediate 
humanitarian action. The 
Integrated Food Security 
Phase Classification (IPC)
initiative is a multipartner 
initiative designed to provide timely data to inform 
emergency response assistance for people exposed 
to acute severe food insecurity (
BOX 13
).
[
43

---

DATA COLLECTION AND ANALYSIS TOOLS FOR FOOD SECURITY AND NUTRITION
•
Data protection:
There have been substantial 
improvements in standard approaches and 
governance processes to ensure research and 
data collection ethics are upheld. However, the 
expanding use of mobile and electronic methods 
for data collection and sharing, including crowd 
sourcing and the ever-growing cloud storage of 
data, can present particular challenges for FSN-
related data protection (see further discussion in 
Chapter 4 and Chapter 5).
•
Heavy reliance on quantitative data:
Reliance 
on quantitative data has important advantages as 
quantitative data are amenable to harmonization, 
BOX 13:
THE INTEGRATED FOOD SECURITY PHASE CLASSIFICATION (IPC) INITIATIVE
The 
Integrated Food Security Phase Classification
 (IPC) initiative is a formal partnership of UN, non-governmental 
(NGO), intergovernmental and other organizations at global, regional and country levels. The IPC is used to assess 
the extent and severity of food insecurity and malnutrition in emergency situations to inform the rapid mobilization of 
humanitarian assistance. The IPC was originally developed in 2004 by the FAO Food Security and Nutrition Analysis 
Unit (FSNAU) in Somalia, in response to the growing need for rigorous, neutral and objective actionable information 
to facilitate evidence-based, effective and coordinated humanitarian response in the context of a country that had 
been undergoing repeated crises. IPC has since grown to a partnership of 15 organizations and intergovernmental 
institutions active across all IPC activities. IPC is now implemented in over 30 countries, with findings used to make 
decisions on allocation of food and other forms of assistance.
One of IPC’s distinct features is the high degree of ownership by governmental institutions, whose representatives 
participate in the country teams that make the assessments. From a methodological standpoint, the IPC is 
predicated on consensus, creating a space for rapid, objective analysis of the relevant available data and evidence 
(which is often scarce and of less-than-ideal quality). Experts from the various agencies that share the responsibility 
for a humanitarian response openly consult on the available data, analysing them according to established protocols 
organized according to four functions: 1. Consensus building, 2. Analyses, 3. Communication, and 4. Quality 
assurance. Assessments provide estimates of current and projected food insecurity and malnutrition in the areas 
analysed, which are typically subnational areas, including refugee camps and the local communities that host 
refugees, when appropriate.
To be useful, IPC assessments must be very rapid, yet reliable. Several features facilitate that timeliness and 
relevance. First, analysis is guided by a formal set of tools and procedures designed to formulate simple, actionable 
statements regarding the classification of the areas at risk, including providing rough estimates of the number 
of people potentially affected. IPC Global Reference Tables provide analysts with benchmarks for three different 
kinds of assessments, one for acute food insecurity, one for acute malnutrition and one for chronic food insecurity. 
Each reference table is designed to define four or five potential phases or levels of severity of the situation, which 
are described in qualitative terms, and then provide guidance on how the evidence conveyed by various indicators 
can be used to classify the areas by level of risk. For example, an area is classified under IPC Phase 4 of acute 
food insecurity (labelled as “Emergency”) when evidence points towards a situation where at least 20 percent 
of households in the area either likely have large food consumption gaps (which are reflected, for example, in 
remarkably high levels of acute malnutrition in children or excess mortality), or are able to mitigate large food 
consumption gaps but only by employing costly livelihood coping strategies. To ensure timeliness, all relevant 
available data are considered, even when less than ideal or incomplete. All available evidence is assessed for 
reliability, considering the conditions under which the data have been collected, and for time and spatial relevance. 
Data found to be sufficiently sound and relevant are then used in the analysis, and results are critically reviewed in 
relation to the specific area’s context and the typical local livelihoods, as well as in relation to other indirect evidence 
and past trends.
44 
]

---

2 
A REVIEW OF EXISTING FSN DATA COLLECTION AND ANALYSIS INITIATIVES
systematization and reproducibility (when issues 
of cross-context validity are addressed). But 
this quantitative data bias also has important 
disadvantages for the utilization of FSN data. 
Moving from the outer to inner circles (macro 
to individual) of our conceptual framework, it is 
necessary to understand the nuance and unique 
contexts in which communities, households 
and individuals interact to produce, procure, 
prepare, share and consume food. A myriad of 
individual, societal, cultural, religious and other 
considerations may have direct relevance for 
decision-making to improve FSN. For example, 
understanding policymakers’ motivations and how 
they perceive and balance the many trade-offs 
that each of their decisions inevitably entail would 
be enormously important. Many of these aspects 
may be difficult or impossible to capture with 
quantitative data. As a result, these considerations 
are often omitted from decision-making processes. 
As qualitative data are less amenable to collection 
by simple, standardized surveys, they may end 
up being excluded from data consolidation and 
dissemination efforts. The 
Exemplars in Global 
Health
 programme (
BOX 14
) is an interesting example 
of how such data can be included in data platforms 
and initiatives.
BOX 14:
EXEMPLARS IN GLOBAL HEALTH
The 
Exemplars
 programme seeks to highlight success stories and the factors that have contributed to them by 
conducting in-depth case studies in public health. Rigorous methods are applied to prioritize topics for analysis, 
identify exemplar countries and consolidate vast amounts of quantitative data on the topic, from published literature, 
websites and national resources, among many other sources. This is complemented by qualitative analysis, including 
dozens of in-person interviews with in-country experts who designed, implemented, or have deep first-hand 
knowledge of the most impactful policies and programmes. In this manner, Exemplars is still resource-intensive in 
that it requires primary (qualitative) data collection. However, its uniqueness and relevance for this report lies in the 
utilization and combination of this information into a public data portal that allows the comparison and contrasting of 
situations across countries and regions.
TRANSLATE DATA AND USE FOR 
DECISION-MAKING
•
Translating data into results, insights, 
conclusions and recommendations:
Data are often 
presented in long reports with complex graphics, 
tables and considerable detail. This is insufficient 
to glean decision-focused results, insights, 
conclusions and recommendations for action to 
improve FSN. Busy policymakers do not have time 
to review multiple data sources nor the necessary 
technical skills to consolidate the information 
from those sources, highlight the gaps and identify 
specific actions. This requires a purposeful and 
complementary set of activities. The 
Food Systems 
Dashboard
 (
BOX 15
), with its 
diagnose and decide 
functions, seeks to address this issue.
[
45

---

46 
]
DATA COLLECTION AND ANALYSIS TOOLS FOR FOOD SECURITY AND NUTRITION
BOX 15:
THE FOOD SYSTEMS DASHBOARD
The 
Food Systems Dashboard
 was developed in 2020 by Johns Hopkins University and the Global Alliance for 
Improved Nutrition (GAIN), along with several other partners. The dashboard combines data from diverse sources 
to give users an overview of a food system, using the HLPE-FSN 2017 food systems framework (HLPE, 2017) as 
the basis for data organization. With over 150 indicators, users can review the current status across all domains 
for a particular country or compare components of food systems across countries within a region or globally, or by 
other variables, such as food system type or national income classification. The developers of the dashboard have 
prioritized 41 key indicators that can also be used to provide more in-depth insights into food system issues and 
opportunities in individual countries.
The dashboard also contains a compendium of 42 actions that have been identified to have potential (through clear 
pathways to impact) to enhance the availability, affordability, acceptability or safety of food. These are organized 
according to their primary sector of action: agriculture, international trade, research processing and technology, 
supply chain infrastructure, financial, public institutions, business initiatives, regulation and law, education and 
public awareness, and national guidelines.
Ultimately, the dashboard aims to go beyond describing food systems to providing the basis for diagnosing food 
systems in a given context and for deciding on specific actions to address the gaps and issues identified. As such, 
there is a clear intent on the part of the developers to contribute to translating knowledge into action. To this end, 
several workshops have been held in countries building on the description and data diagnosis to explore policy and 
programmatic options.
Efforts are now underway to develop subnational Food Systems Dashboards in several countries. This is an 
important step, in view of the diversity of food system issues and opportunities that exists at subnational level. In 
many contexts, these local or regional adjustments to policy and action may be critical to adapt food systems. This 
may be particularly important to ensure that the unique needs of those most vulnerable to food insecurity and 
malnutrition are not missed in general, national-level efforts.
•
Using data for decision-making requires buy-
in and involvement on the part of those with the 
responsibility to make decisions, and clarity on 
the decisions to be made:
As noted in Chapter 
1, and illustrated again in this review, multiple 
stakeholders and sectors are relevant to FSN. 
Although FSN data relevant to these different 
stakeholders and sectors often overlap and are 
complementary, there are also gaps in the data, 
and too often, the intended users of the data are 
not engaged in data-related activities. The POSHAN 
Network in India (
BOX 16
) was specifically designed 
to address these challenges and enhance the 
effectiveness of data-informed decision-making 
for FSN. The network draws on different sources of 
data and brings together a variety of stakeholders 
who work together to apply the data in initiatives 
aimed at improving child nutrition in the country. 
It is a clear that programmes of this type should 
be strengthened – increasing spending on them, 
integrating them with local food systems and 
expanding their reach. In the current economic 
situation, however, such programmes have come 
under considerable strain and the budgetary 
allocations for them have contracted. Institutions 
and governance (the centre of the data cycle in 
FIGURE 2
) have a critical role to play in this regard, as 
is discussed in detail in Chapter 5.

---

[
47
2 
A REVIEW OF EXISTING FSN DATA COLLECTION AND ANALYSIS INITIATIVES
BOX 16:
THE POSHAN NETWORK
In 2017, the Ministry of Women and Children of the Government of India launched the POSHAN Abhiyaan 
programme, aiming to substantially reduce the prevalence of all forms of child undernutrition, particularly stunting, 
wasting and low birthweight, by reducing the evidence gap in Indian nutrition and supporting efforts to generate, 
synthesize and mobilize diverse types of nutrition data and evidence to support policy decisions. POSHAN is led by 
IFPRI (International Food Policy Research Institute) Delhi and funded by the Bill & Melinda Gates Foundation.
The programme has brought together many different schemes that have played a very important role over the last 
few decades in extending nutrition services to children and women in India. The need for a mechanism to coordinate 
and support evidence-informed dialogues and decision-making at national and state levels in order to inform the 
needed actions was identified early on as a critical element for success. Thus, the 
POSHAN Network
 (Partnerships 
and Opportunities to Strengthen and Harmonize Actions for Nutrition in India) has the objective of “[…] generating, 
synthesizing, and mobilizing nutrition data and evidence, by engaging a variety of stakeholders, to support strategic 
nutrition policy and programme actions in India.”
POSHAN works across all six steps in the data cycle, working with counterparts to identify and prioritize evidence 
and knowledge needs; consolidating and analysing data, including qualitative data in the form of success stories 
of change; translating data into policy briefs and similar media; and disseminating results through workshops and 
similar activities.

---

48 
]
Chapter 3
CONSTRAINTS, 
BOTTLENECKS (AND SOME 
SOLUTIONS) FOR EFFECTIVE 
USE OF FSN DATA
48 
]
Kyrgyzstan, 13 May 2019, Head of Laboratory Gulay Abdymambetova checks vegetables for nitrates in Logistic food center in 
Kemin some 80 km from Bishkek.
© FAO/Vyacheslav Oseledko

---

[
49
3 
CONSTRAINTS, BOTTLENECKS (AND SOME SOLUTIONS) FOR EFFECTIVE USE OF FSN DATA
T
he discussion in Chapter 2 points to the 
wealth of existing FSN-relevant data and 
information. It also suggests, however, that 
significant data and information gaps still exist, 
especially for low-income countries. This chapter 
examines the most relevant constraints and 
bottlenecks that underlie those gaps and hinder 
the effective collection, analysis and utilization 
of FSN data. The intent in doing so is to derive 
recommendations that may lead to feasible 
solutions.
The identified constraints and bottlenecks are 
broadly categorised as relating to insufficient 
resources for data collection and analysis, and to 
inadequate institutional capacity and arrangements 
and problems with data governance.
One area of special interest in this chapter is the 
human capital needed to achieve effective use 
of data in all areas that contribute to FSN, from 
policymaking and the actions of food system actors, 
all the way through citizens’ food choices. Data 
are crucial to inform all these levels of the food 
system, yet, despite the abundance of data (but 
perhaps, in part, also 
because of it), 
there is still 
very limited ability across the board to make 
full sense of the continued flow of data
. Only a 
small minority of people in the world possesses the 
necessary skills to properly interpret, process and 
distil information from data in all the various forms 
– numbers, images, texts, words – in which it is 
continuously generated, stored and distributed. Of