The tool has so far been implemented in two Latin American countries through the United Nations Development Program (UNDP), with ambitions to use it even more in the region and bring it to other regions.
To help policy makers prioritize their public spending on the road to achieving the Sustainable Development Goals (SDGs) - especially as new challenges approach - an artificial intelligence approach called Policy Priority Inference (PPI) has been developed as a collaboration between researchers in the UK and Mexico, supported by the United Nations Development Program and the Alan Turing Institute.
The PPI approach is a unique combination of economic theory, behavioral economics, network science and agent-based modeling, which aims to support governments in determining budgetary priorities to achieve national development goals. How long will it take to reach development goals, given a current trajectory? How viable are the objectives that have been set? Which policy areas can accelerate progress towards the SDGs? These are just some of the questions for which the PPI is able to provide evidence-based answers.
“Recently, governments around the world have had to devote substantial resources to combat the COVID-19 pandemic, preventing them achieving their original goals. In this context, PPI can be used to keep governments on track, despite the setbacks of the virus, ”explains Omar Guerrero, ESRC / Turing member at University College London.
The tool has so far been implemented in two Latin American countries through the United Nations Development Program (UNDP), with ambitions to use it even more in the region and bring it to other regions. Turing has just published an impact story detailing this new instrument, which has the potential to overwhelm the effectiveness of sustainable development supported by the government for the benefit of billions of people - and the planet itself. The PPI is based on a computational behavioral model, taking into account the learning process of civil servants, coordination problems, incomplete information and imperfect government monitoring mechanisms.
“The tool has the potential to provide governments with concrete information on how to increase the effectiveness of public spending and accelerate the achievement of development goals,” says Annabelle Sulmont, Public Policy Projects Coordinator at the UNDP Mexico office. "The model also provides a common language that allows its implementation in other parts of the world and facilitates the sharing and comparison of results between regions and countries".
Prioritizing issues for maximum impact is a huge challenge for governments. The range of development policy options is uncountable, often with unforeseen inefficiencies that waste resources. And, crucially, there are complex interdependencies between policies that must be taken into account (for example, investing in industrialization also tends to produce negative results for the environment, while investing in public transport can also increase education outcomes, because more children become able to access the school.) Modeling these complex scenarios is exactly the type of `` perverse '' problem in which UNDP and the Turing Public Policy program are committed to working with policy makers around the world and that cutting-edge data science and AI technology can make a huge impact.
“Government spending data will take this technology to a whole new level and PPI is not just about government, but also about accountability. We also want to take these tools to NGOs, because it is useful for assessing government actions. NGOs can verify that governments are prioritizing the right policies. ”, Added Guerrero.