Tracking Financing to the Sustainable Development Goals
The Sustainable Development Goals (SDGs) are poised to substantially influence the next 15 years of development finance (2015-2030) and, by some estimates, will require the international community to mobilize an additional $1.5 trillion USD per year to meet financing goals. Are development partners living up to their commitments? Where are the greatest shortfalls and surfeits in funding for sustainable development?
Tracking and analyzing funding for the SDGs will be central to measuring progress. However, as aid reporting systems do not currently capture information on the distribution of financing for the SDGs, a coherent methodology is urgently needed. For this reason, AidData is developing a standardized coding schema to systematically track the resource envelope of financing going to each of the sustainable development goals and targets.
AidData’s SDG coding methodology is based on an analysis of the text of development project descriptions. Since 2007, student researchers at AidData have assigned codes to over 800,000 project descriptions through a double-blind coding methodology, providing more granular data on project activities and purposes. This coding schema builds on the Organisation for Economic Co-operation and Development (OECD) Creditor Reporting System (CRS) categories and was designed to improve the quality and usability of our data by adding an additional layer of project-level detail in a standardized way across donors.
Adapting this methodology to measure funding to the SDGs involved three critical steps. The first step wass to map the relationship between existing activity codes and SDG targets. To link AidData activities to targets, a team of student activity coders went through the 544 AidData activity codes and assigned SDG targets to each activity. AidData staff then reviewed the coding and arbitrated cases of disagreement among the coders.
After mapping activity codes to specific SDG targets, we next split aid projects across assigned activities. Using projects that had already been assigned activity codes, we split dollar amounts for a project evenly across all activity codes assigned to it. Although projects will have different distributions of dollar amounts across activities in practice, there is no reliable way to infer this given existing data.
Having split the dollar value of a project across unique activities, the next step was to distribute these activity-dollar amounts across the SDGs, using the mapping developed in step one. If an activity was linked to at least one SDG target, the entire value assigned to that activity was distributed evenly among assigned targets. If an activity was not linked to any targets, then the financing was not counted toward the SDGs. This likely provides a conservative estimate of funding that contributes to the various SDGs.