This study aimed to create a list of proposed policies towards healthier and more environmentally sustainable food systems, applicable to governments globally at any desired level of jurisdiction. We started by compiling existing international policy recommendations addressed at governments to identify potential food systems policies, and based on that we conducted a scoping review [4] to examine the effects and effectiveness of those internationally recommended policies on a total of five outcomes. The three primary outcomes were (i) undernutrition, (ii) obesity/NCDs, (iii) environmental sustainability. In addition, (iv) inequalities and (v) women’s empowerment were included as secondary outcomes, as they are not direct outcomes of the Global Syndemic and not considered when assessing the double- or triple-duty potential of policies, but in a non-linear way they are simultaneously drivers and outputs common for the three pandemics. The aspects considered for each outcome are available in Table 1.

Table 1 Inclusion criteria of the primary and secondary outcome areas analysed in this study

Due to the lack of evidence available for effectiveness of policies within some policy subdomains on different outcomes [4], additional insights on the list of proposed policies were gathered through an international expert consultation. This was done by conducting two online surveys and four regional workshops to identify the perceived effects (double- or triple-duty potential), the effectiveness, synergies and trade-offs of the list of proposed policies, regardless of their implementation level. All the inputs, changes in the number of proposed policies according to each step, and the complete process are depicted in Fig. 1.

Fig. 1

Flowchart of the steps undertaken in 2021–2022 to develop the proposed policies for governments towards healthier and more environmentally sustainable food systems, number of policies considered across each step, and input sources

Step 1 – Compilation of international recommendations

From March to July 2021, we conducted a desk review of international guidelines, reports and peer-reviewed scientific articles that recommended policy actions for governments to improve food systems concerning population nutrition, nutrition-related inequalities and/or environmental sustainability.

The list included documents found through a grey literature search conducted on the Internet and key documents already known to the co-authors involved in this research, mainly consisting on reports from international organisations [namely the Food and Agriculture Organization of the United Nations (FAO), the World Health Organization (WHO), the Global Alliance for Improved Nutrition (GAIN), the Organisation for Economic Co-operation and Development (OECD), the United Nations Children's Fund (UNICEF) and the International Food Policy Research Institute (IFPRI)]. To be eligible for inclusion, each report/guideline/paper was assessed against the three following criteria: (i) it contains specific, detailed recommendations for government policies or actions addressing food systems, and policy recommendations needed to be action-oriented and specific (ii) it contains information details about the funding sources; reports/guidelines produced by the agriculture or the food industry were not included (iii) it was published between 2011 and 2021. A total of 23 documents met these inclusion criteria and were used as sources for the compilation. The 23 documents were reviewed in detail, extracting a total of 291 policy recommendations for governments (Annex 1).

As foreseeable, overlaps were found across the policy recommendations. In such cases, they were merged keeping the content of the original recommendations as close as possible to how they were worded. This allowed both for the identification of potential policy areas to be addressed (that in this research I refer to as ‘domains’ and ‘subdomains’) and the generation of a shorter, merged list of policies for governments (Table 2).

Table 2 Recommendations extracted and merged, divided by policy domains and subdomains

All the recommendations were compiled and classified according to the food systems areas they were addressing in these ten subdomains, that were identified in two ways: for the food supply chains domain, an inductive approach was used based on the thematic areas of the policy recommendations; for the food environments domain, a deductive approach based on the Food-EPI tool was adopted (B. [33]). Once classified and merged per domain and subdomain, 46 proposed policies were retained, as they covered individual food systems aspects and proposed policy actions that could tackle one of more of the three outcomes studied. To ensure consistency across the policies recommended, the same approach and languages used for the Food-EPI were adopted in the phrasing of the recommended policies (B. [33]). The results of this compilation were used both to create the first survey to identify the perceived effect and effectiveness of the proposed policies (step 2).

Step 2 – First survey: perceived effects and effectiveness of the recommended policies

In November 2021, an online survey (LimeSurvey) was conducted among international agriculture, food and environmental sustainability experts. The objective of the survey was to get insights from experts on the perceived double- and triple-duty potential of the proposed policies, and to identify potential synergies and trade-offs across the outcomes (Table 1).

Experts were recruited within two networks: (1) the INFORMAS network and (2) the Food Sustainability Advisory Team (Food-SAT), which was established under the INFORMAS2.0 IDRC project.Footnote 1

Two surveys were created, one for the domain of food supply chains and one for food environments. Experts were asked to assess the potential effect of 46 recommended policies on the five outcomes. Experts could rate the effect of the policies as “positive”, “negative”, “neutral”, “non-applicable” or “unknown”. When experts selected “positive” as the perceived effect, they were asked to rate the effectiveness according to three levels: “very effective”, “effective” or “somewhat effective”. Therefore, each expert had to assess the effect and effectiveness level of each policy across the outcomes, to allow for the identification of policies with double- or triple-duty potential. At the end of the survey, we asked for additional feedback or suggestions for us to take into account during the analysis of the results or the further development of the list of proposed policies.

The survey responses were analysed per policy, considering their double- or triple-duty potential, and the potential synergies or trade-offs across the three primary outcomes. Based on the results of the survey, the policies were classified according to three categories: (1) essential to keep (those perceived as likely to have a strong positive impact on at least two outcomes, with no negative impacts perceived); (2) to be excluded (those considered likely to have a negative impact on at least one primary outcomes, with only low effectiveness in all others); and (3) policies with mixed results (those with perceived mixed effects, either likely to have strong effectiveness in two or three primary outcomes with negative impacts in others, or those identified as important single-duty policies which were likely to have strong effectiveness in just one outcome but without perceived effects in the others). The results of this survey were used to inform the experts participating in the regional workshops and to advise on the final selection of the policies.

Step 3 – Regional workshops

Between May and July 2022, four online workshops were organised. We invited agriculture, food, health and environmental sustainability experts from the regions included in the INFORMAS2.0 project (Europe, Latin America, East and West Africa). The objective of each of the workshops was to discuss the perceived effect (double- or triple-duty potential) of the proposed policies on the five outcomes. Based on their organisation, role, and relevant research field, experts from the four regions were identified and contacted by the INFORMAS2.0 partners in Belgium, Brazil, Kenya and Senegal. In order to be included, experts were assessed according to two main criteria: (i) he/she is directly involved in at least one of the ten food systems subdomains identified (either conducting research or through the design or implementation of policies); (ii) his/her country of origin belongs to the four regions analysed. A total of 235 experts were invited via email, with the request to reply if they were interested in participating. Their written confirmation was used as their consent. Prior to the workshop, the participating experts were divided into four different sub-groups (based on their field of expertise) to ensure optimal feedback within all subdomains. The distribution of the 96 experts who participated in the workshops, their field of expertise, type of organisation, country of origin and their assigned groups can be found in Annex 2.

The workshops were organised in English (for Europe and East Africa), Spanish (for Latin America) and French (for West Africa). A short introduction was given at the beginning of the workshop, after which participants were divided into breakout rooms according to their assigned sub-group and were invited to reflect on the following aspects of each proposed policy: (i) the content clarity and wording; (ii) the level of (dis)aggregation; (iii) the double-/triple-duty potential and its potential effect on inequalities and women’s empowerment; and (iv) any potential synergy or trade-off across the outcomes. The information on the policies’ double- or triple-duty potential, as well as the synergies/trade-offs identified through the scoping review [4] and the first survey was provided for each proposed policy. Experts were also asked to suggest additional double- or triple-duty policy options that were not covered in the proposed list.

The final list of proposed policies was created taking into account the findings from the scoping review, the survey and the feedback from the workshops. The results gathered through the scoping review were prioritised. When scientific evidence was not available, the input from the experts was considered. During the process of combining the results from the different steps, some policies were regrouped, others were disaggregated and others were reworded. Suggestions made by experts that were out of scope [as they did not directly impact the five outcomes, but had a more upstream focus (i.e. other social or economic determinants)] were not taken into account. All trade-offs identified were considered when creating the list of proposed policies, and for some policies experts identified potential solutions and proposed changes in the text to reduce or eliminate their negative impact. However, given the complexity of food systems and the differences across countries and contexts, it was not always possible to modify the policy wording to address all the potential trade-offs identified during the discussions. For those cases, the trade-offs were simply noted down.

Step 4 – Second survey: prioritisation

In October 2022, an additional meeting with the Food-SAT (n = 21) and INFORMAS2.0 (n = 10) experts was organised. In this last step, we sought to further verify the proposed changes and to identify which policies they considered should be prioritised. The 31 experts were invited via email, and to enable maximum participation, two meetings at different times/dates were organised. Before the meeting, the new list of proposed policies (resulting from step 3), the results from the prior conducted scoping review [4] and the feedback received from the workshops (step 3) were shared with them. During the meetings, experts were asked to share their feedback on the list of proposed policies, based on the scientific evidence available and their expertise in the field.

After the session, experts were asked to rank the proposed policies according to their perceived effectiveness to improve one (or more) primary and/or secondary outcome(s) of their choice. In order to rank them, experts had to select from the list of proposed policies the ones that had a positive effect on the chosen outcome(s). Once selected, experts had to order them according to their (perceived) level of effectiveness. Given the difficulty of comparing the importance of policies across the two core domains, the ranking was done separately for the 27 food supply chains policies and the 17 food environments ones. There was no limit to the number of policies that could be ranked. However, in order to ensure that a significant ranking among policies was conducted, experts had to select a minimum of 5 per domain. At the bottom of the survey, experts could give additional explanations or comments to be taken into consideration while analysing the data.

A numerical value was assigned to the ranking positions: the policy in the highest position in the ranking was assigned a value of 10, the second policy an 8, the third a 6, the fourth a 4, the policy in the fifth position a 2, and all the other policies that were selected and ranked in lower positions (below the fifth) were given a value of 1. This was done to ensure a distinction between policies that were considered relevant but in lower positions in the ranking, versus those that were not selected at all (therefore perceived as not effective for the chosen outcome). This way, we were able to identify the proposed policies considered to be prioritised, by domain and by outcome. The extent to which experts agreed on the level of priority of the proposed policies was then analysed by policy domain and outcome in Excel using Gwet AC2 inter-rater reliability coefficient with Agreestat360.