RE: What can we do to improve food security data? | Eval Forward

I sincerely thank the participants and the various contributions to the debate launched since June 4, 2019 on "What can we do to improve food security data? ". The harvest was good because a number of deplorable national situations were reported and, fortunately, an interesting solution was revealed.

Indeed, from the Nema program in the Gambia, Paul Mendy informs that the ability of evaluation staff to collect and report on food and nutrition security is not up to the task. The same is true in India where Archana Sharma reports that most enumerators, surveyors and field workers not only work in poor conditions, but also live in adverse conditions, are not sufficiently problem-oriented and do not have access to any type of training in tools, techniques, methodologies, approaches and processes involved in data collection. These are poorly paid workers. The consequence is that investigators complete surveys based on their bias or expected investigation bias, as pointed out by Richard Tinsley. The result is that the data are biased and unreliable, but consistent with the country's financial situation. Tinsley suggested extrapolating from projects run by donor-assisted NGOs with a sufficient budget to manage a reliable survey, while Sharma found that the research or evaluation agency should invest proportionately in high-end surveyors and field staff for quality data collection.

There is therefore a need for our agriculture and food distribution policies to be based on common sense and for nutrition standards to be founded on the food needs of local people in line with their food culture (Lal Manavado). In the same vein, Kebba Ngumbo Sima warns that it is high time that attention be given to the context of local communities or indigenous peoples and their perceptions and understandings related to food security issues. However, because of the volume of work that the 2030 Agenda imposes on countries, the outcome will really depend on (a) the willingness of governments to invest in data collection, (b) the financial assistance that national statistical offices will be able to obtain from regional and international organizations (Filippo Gheri). Many factors must be taken into account when choosing the indicator to use to monitor food insecurity, as well as the type of data to be collected to obtain the indicator. An indicator should be easy to use, provide timely information, and be informed by data that is easy to collect (cost effective). It should also provide valid and reliable information. These characteristics are very difficult to find in indicators aimed at "measuring" food insecurity. For this reason, the Food Security and Nutrition Statistics Team of FAO launched the Voices of the Hungry project in 2013 (http://www.fao.org/in-action/voices-of-the-hungry/en/), which led to the development of a new tool called “Food Insecurity Experience Scale (FIES)", which has become indicator 2.1.2 of the Sustainable Development Goals. This new tool is according to Filippo Gheri direct, easy to use, low cost and statistically valid. It also helps to distinguish levels of severity, subdivide results and compare results across countries and over time. The FIES-based survey module has already been included in more than 50 nationally representative surveys around the world and another 60 countries have already planned to include it in their national surveys.

In short, the problems of reliability of food security data are everywhere in underdeveloped countries. This is due to the low investment of the states in this activity; which results in the use of agents of low quality. But, the FIES tool is an interesting solution to correct the situation, because it allows to significantly reduce the collection bias. The FIES tool is applied to the month, quarter or last 12 months; this allows to correct biases raised about changes in the food situation of households and individuals between seasons of the year. It is therefore imperative and urgent that many food security and development assessment specialists are trained in the effective use of this important tool for improving indicator data collection practices and improving the quality of the data for the indicators, and for the achievement by 2030 of SDG 2 in particular.

Dr Emile N. HOUNGBO