Submission: Inquiry on the Future of UK aid and development assistance              Leonard et al.

Evidence submission: Future of UK aid and development assistance

Alycia Leonard, Miguel Sánchez-Lopez, Stephanie Hirmer

Department of Engineering Science, University of Oxford

This submission responds to the UK Parliament International Development Committee inquiry on the Future of UK aid and development assistance. It draws on our experience as sustainable development researchers UK aid-funded programmes such as the Climate Compatible Growth (CCG) and UK Partnering for Accelerated Climate Transitions (UK PACT) programmes. We address three questions of the inquiry in turn below.

How should the Government be prioritising UK Official Development Assistance (ODA) in the coming years?

In light of budget cuts, UK ODA should be better targeted to maximise its benefit. This can be achieved by aligning aid-funded interventions with local, area-specific needs and values. Understanding where interventions will have the greatest impact can help to avoid low uptake caused by a mismatch between intervention design and local context.

Clean cooking provides a helpful illustration of this potential. While clean cookstoves have historically been underused when deployed indiscriminately[1], they could be more carefully targeted – for instance, to areas where charcoal is scarce/expensive or where firewood collection is restricted – to increase uptake. If such patterns could be quantified, ODA-funded interventions could be targeted to increase spending efficacy, enabling UK aid to “do more with less”, as per its stated targets.

Predictive aid targeting using non-traditional data sources has gained momentum in recent years. During the COVID-19 pandemic, for instance, innovative approaches were implemented to enhance humanitarian aid allocation using non-traditional data. For example, in Togo, mobile phone data were employed as a proxy to identify vulnerable areas in which to concentrate aid distribution[2]. Similar approaches can be applied to UK ODA, leveraging the UK’s exceptional analytic talent.

Learning from experience, the UK should prioritise interventions where they are most likely to be adopted and used to improve spending efficiency and development outcomes. Statistical and predictive approaches can be used to identify demographic and geographic drivers of success and failure: that is, what projects are most likely to have uptake, where, and with who. Geospatial modelling can rapidly and efficiently map locations where these area-specific enabling conditions coincide, allowing policymakers to visualise and prioritise areas with the highest likelihood of intervention uptake and success. Notably, the UK should use its vast data resources from monitoring and evaluation of past aid-funded projects to facilitate this targeting.

Recommendations:

• The UK Government should analyse monitoring and evaluation data from previous UK ODA projects to identify success and failure conditions for different interventions.
• The UK Government should apply geospatial modelling and predictive analysis to rapidly identify locations where enabling conditions coincide, providing an efficient and evidence-based means of intervention targeting.
• The UK Government should target future ODA-funded interventions to areas where alignment with local needs, resource availability, and socio-economic realities maximises likelihood of uptake and development impact.

How should UK ODA best be deployed to empower local delivery of aid and decision making around development?

Predictive targeting should never overrule the lived knowledge of local institutions and people. While it can increase spending efficiency by identifying areas where likely high-impact interventions align with data-driven representations of local context, community engagement on-the-ground is still necessary.

Although predictive models can enable faster, cheaper, and potentially more accurate identification of high-impact beneficiaries than traditional approaches, their limitations must be acknowledged. Widespread real-world use of predictive algorithms has raised concerns regarding algorithmic ethics, fairness, treatment of individual rights, and overall legitimacy[3]. Consequently, predictive targeting should be understood as a tool to support decision-making processes rather than a substitute for local institutions or the contextual knowledge held by affected communities[4].

UK ODA should therefore empower local delivery and decision-making by working with and through existing decentralised decision-making structures in target areas. By doing so, costs to the UK taxpayer can be decreased (i.e., as fewer human resources are needed) while local buy-in existing local decision-making capacity are increased. An added benefit is the reduction of interview fatigue among communities, particularly in locations and sectors where there is already a high presence of international actors repeatedly conducting surveys and consultations with minimal visible impact[5]. Importantly, the efficiency of local delivery mechanisms should be evaluated prior to engagement.

Recommendations:

• The UK Government should never overrule local decision-making using data-driven ODA targeting.
• The UK Government should conduct community engagement in identified priority areas for ODA interventions.
• The UK Government should support existing decentralised development committees in-country via ODA-funded engagement to further increase efficiency and buy-in.

How can equity and accountability be better embedded into the way the UK Government delivers its aid budget?

The impact of ODA projects should be monitored with a quantitative intersectional approach to ensure equitable impact. Not everyone benefits equally from the same “amount” of support: for instance, a widowed unemployed mother will benefit differently from a $50 cash transfer than a married childless entrepreneur. As such, a purely utilitarian view of the impact of UK ODA is inadequate during monitoring. Budget reductions may encourage spending on low-hanging fruit over those who are most vulnerable; impact must therefore be monitored with an intersectional lens.

While this may sound onerous, it need not be. Standard demographic survey data can be analysed to uncover differences in impact across vulnerable intersectional groups[6]. Development planning models can be configured to take demographically disaggregated inputs or to run equity-focused scenarios (e.g., in energy and transport planning[7],[8]). If quantifying impact in monetary terms is important, standard methodologies like social return on investment[9] can be applied.

Combining demographic and geolocation data can furthermore enable assessment of whether aid is effectively directed toward the most vulnerable areas using relatively straightforward GIS analysis. Notably, this is not always the case at present: taking World Bank–financed projects as an example, in certain Sub-Saharan African countries, there is a negative correlation between the geographical distribution of projects and the share of the population belonging to the poorest 40%[10].

Recommendations:

• The UK Government should collect demographically disaggregated data during ODA-funded project monitoring to enable the calculation of intersectional impact metrics.
• The UK Government should account for relative disparities in initial living conditions when quantifying impact potential of ODA-funded interventions.