We are pleased to provide the following comments and feedback on V0 Draft report on Data Collection and Analysis Tools for food security and nutrition. We thank the CFS HLPE and the report authors for providing this opportunity for review and we trust that our comments will help support the development of this important report.
General Australian comments
As an overarching comment, we consider that the report has only partially met its stated objectives. A primary issue in our view, relates to the generic nature of the included examples, however we hope that this consultation process will support the authors to provide more specificity when identifying and highlighting barriers, gaps and initiatives. We consider that the report may benefit from adopting a geographical lens, providing a summary of the main food and nutrition security data strengths and weaknesses from each region and scale. There is opportunity also for the authors to include thematic areas relevant to this topic – e.g., availability vs. consumption data.
The report would benefit from a review of the structure and may find more cohesion by taking a systematic approach as currently we consider there are challenges with the way the information is presented. One option may be to follow the conceptual framework and data cycle of food security and nutrition. We found that Section 3 was particularly challenging with the same topics addressed under different subsections. This could be remedied by breaking the section into smaller groupings to help focus and target the narrative more clearly. While both Sections 3 and 4 cover innovative methods, there is some overlap that should be addressed. The generic nature of Section 4 should also be addressed, along with additional specificity that will help the reader understand food security and nutrition data throughout the data cycle. This could be alleviated by additional development of sub-section 4.3 which is helpfully presented in the framework of food and nutrition dimensions – though the nutrition information is missing.
Section 3 would also benefit from a greater refocusing as the authors risk a report that covers all aspects of food system monitoring rather than drilling into food and nutrition security outcomes. For example, the discussion around SIS does not seem relevant to FSN data collection and analysis but on how AI can be used for food systems intervention. It is suggested the report provide greater clarity around whether the report is focusing also on food system data collection and analysis or, as the report title suggests, food and nutrition security data more specifically.
Additionally, in relation to the introductory statement that “food systems have failed us,” we believe that while recognising that our food systems must continue to adapt and do better, the report should also recognise the strengths of current food systems and what has been working well to date and can be built upon. We note that the report could benefit from further exploration of the considerable role that international trade (and related datasets) has for food security and nutrition outcomes.
As an editorial statement, we suggest that descriptors, tables, boxes, figures and table numbers be consistent throughout the report.
Section Specific Feedback
Section 1
The conceptual framework put forward by the authors is sound. However, the matrix (figure 3) appears somewhat impractical and would benefit from harmonisation with the technology used in the conceptual framework. The example matrix (p.16) is troublingly generic with the potential for the content to be used to apply to almost any nutrition ‘problem’.
Detailed commentary:
The socio-ecosystem framework of food security is a good initiative as it captures the interrelated scales well.
The data driven decision making cycle (figure 2) for FS though arguably best practice for any data collation exercise is quite generic. We would like to this better adapted to the FSN context.
We consider there to be some challenges with figure 3 that should be addressed:
- The first column should correspond to the conceptual framework however, and unless this is referring to Meso or Proximal determinants (how the conceptual framework refers) ‘systems level’ is missing from the framework, otherwise we suggest that terminology should be harmonized. Additionally, the conceptual framework is unclear on whether individual level is the same as the micro level, or is the Personal, HH, Community decision making? Either way, these should be reviewed and consolidated.
Referencing all definition of evidence priorities in relation to ‘the identified problem’ is problematic and unclear. The primary aim of the template is defined as ‘to identify problems that require data’ (p.15) yet the approach of the matrix is unclear – whether it identify the problem or will the problem be identified upon completing the matrix? As a further complicating factor, in real-life situations there are often multiple problems which makes a single identifier challenging. Suggest re-working the language in the column defining evidence priorities.
We suggest allowing more space on p.15 to highlight the steps required prior to data collection. It is probable many people will not have this knowledge – particularly how to identify indicators that are known to be measurable and meaningful. This section may be strengthened with the inclusion of a figure/image to demonstrate.
The first example (p.16) contains relatively generic information that could be used for most diet and/or nutrition problems. It appears to lack some specificity with the identified problem or the context. This is one example of the limited practical use of the matrix as proposed.
Section 2
The authors may wish to revisit the readability of this section which, in terms of flow, was difficult to follow. To counter this, suggest the table be split up, and also summarized in the text as the text currently focuses on general challenges related to FSN data relevant to the data cycle concept. The section would benefit from comment on the methodology or inputs used to make this type of assessment which alleviate some of the issues with the generalised nature of the section.
While there is absolute value in detailing current initiatives, titling this section ‘Review’ ensures an expectation from the reader that these initiatives will have some critical commentary also. However, these initiatives seem to focus mainly on global data sets – suggest the section could be strengthened with the inclusion of initiatives that collect national or subnational data. If the report is interested only in global FSN data initiatives this should be defined somewhere for the reader. Otherwise, recommend including the opportunity for other initiatives to be suggested for inclusion and review – for example, the FAO GIFT initiative and the global burden of disease study.
Section 3
As with feedback for Section 2, the authors may wish to revisit the readability of this section and the structure to avoid the lack of current cohesion and, for this section, the duplicative nature of the information provided. Additionally, we recommend the report ensure consistency of terminology – in this section data cycle and data value chain are used interchangeably.
We recommend broadening the section on policy decisions references seem to focus on agriculture and trade along with the current emphasis on diets/health.
Given the current environment, suggest authors acknowledge the impact Covid-19 has had on data collection – i.e., the reduction in face-to-face modalities, the increased pressure on already limited capacity, the need to shift to more mobile surveys and what that means for data reliability especially when trying to target the most food and nutrition security vulnerable.
We recommend the text in box 3 be either reincorporated back into the main text or the focus of the narrative sharpened. At the moment, it is unclear whether the intent is to highlight the high cost of FSN data or the impact of these high costs. We suggest the same for box 4 – there is too much text and the box does not adequately highlight how critical this information is to the report.
The section could be improved by identifying the constraint per section of the data cycle – i.e., collection, analysis, interpretation, dissemination and use. While the challenges are similar (i.e., financial, human resource) they have different implications at the different stages. It would be good to see this incorporated into the section. Additionally, the section appears to focus on constraints of household/individual data collection at the expense of exploring data collection at the other scales presented in the conceptual framework. Perhaps restructuring this particular section to identify the financial and human resources costs as the main constraints of the data cycle before exploring the how this impacts the quality and usability of the data produced would be of benefit.
We recommend that insufficient data and data quality issues have its own sub-heading in the section. Currently, this flows with the previous subsection which is focused on the resource constraints to data collection.
Suggest expanding the section on advanced technologies to include those specific to measuring food and nutrition security rather than limiting the section to the production aspect of food systems.
Section 4
While subsection 4.3 addresses new technologies for data, it lacks specificity with regards to FSN data and would benefit from further development. As with above, nutrition is also missing from the food and nutrition dimensions of the framework.
It is unclear how tables 1 and 2 differ as they both aim to present initiatives that address some aspect of FSN data through the data cycle. The way table 1 is titled, it could easily include the information in table 2. There is potential to consider consolidation – for instance with tables 2 and 3 being combined.
Section 5
We recommend further development of this section.
Lynda Hayden