Skip to main content Skip to navigation

Warwick's Applied Statistics & Risk Unit supporting Warwickshire Food Strategy

Dr Martine Barons has a passion for using mathematical modelling to improve people's lives. Through developing mathematical support tools, she is working with a variety of businesses, industry, and government agencies to help solve complex social challenges.

Working with local government to reduce food insecurity

Martine is collaborating with the Warwick Food Forum (WFF) to reduce food insecurity through creating an Integrated Decision-making Support System (IDSS) that contrasts and filters different types of complex information. This mathematical tool helps decision-makers compare different policy approaches and select the best combination of options to reduce as much as possible the burden of food insecurity on families in Warwickshire, and to identify where further resource would make the most impact.

In the UK food insecurity has been increasing as the cost-of-living crisis, climate change and geo-political upheavals have affected the ability of households to consistently afford or access sufficient healthy and nutritious food. In 2023 a UK government survey of 6,000 adults revealed that nearly a quarter (23.4%) of households with children and 14.8% of households without children reported experiencing food insecurity.

Managing multiple elements and relationships

The difficulty in making decisions within complex systems is predicting which combination of actions will give the ‘best’ results. Levels of food insecurity in part are predicted by the cost of living, which in turn is dependent on weather, farming subsidies, climate-change, social benefits, charity provision and geo-political stability. By bringing together expert judgements, data and models, and the relationship between these distinct elements, the Food Insecurity IDSS can score and rank different combinations within parameters set by the decision-maker. As the IDSS is a ‘white box’ system, decision makers can interrogate inputs and adjust the underlying information to better reflect real world outcomes. It also makes scrutiny straightforward, so that decisions can be audited.

Martine has already published a proof of concept for the IDSS approach thanks to a Turing Institute mediated collaboration with the The National Archives. This collaboration demonstrated that the IDSS could be embedded as an archivist's software tool to help safeguard the nation’s digital memory reducing risk. The IDSS tool, called DiaGRAM has been engineered to the government's high standards for accessibility and user requirements.

Supporting the Warwickshire Food Strategy

To configure the IDSS to Warwickshire’s food insecurity decisions, Martine has worked with stakeholders in the WFF to identify pertinent sources of data, models, and information, and pick out the relationships between these different elements. Where data is indirect other sources have been translated - for instance agricultural run-off can be translated into approximate fertiliser application rate - and she has included wider concepts of financial resilience and social capital to provide a fuller model of food security.

Her persistence and commitment within the Forum has familiarised the food insecurity specialists with the underlying principles of the IDSS paradigm for modelling data, and through knowledge transfer, when personnel change, the value of the model continues to be recognised within the Warwickshire Food Strategy programme. Martine and the network of food policy experts are now looking for further resources to develop the bespoke IDSS for Warwickshire, and potentially add value to other authorities through the local government analyst network LARIA.

Engineering a sustainable robust and accessible policy tool

Developing Martine’s analytical engine into an operational platform in close collaboration with the experts in Warwickshire requires a Post-Doctoral Researcher to run final tests of the model and run back-casting to see whether it predicts correctly. Once fully coded, the model must be software engineered, so it is a sustainable, robust and accessible tool for policy officers to use in assessing the impact of different policy choices.

Interested in partnering with the Applied Statistics and Risk Unit?