The national Social Support Act prediction modelPublished on: Author: Team Communication Category: News
In 2019, the municipality of The Hague started an innovative project: developing a prediction model for future Social Support Act (Wet maatschappelijke ondersteuning) applications. Policy officers immediately reacted positively. The model is of great value for every Dutch municipality, hence why its development will be in cooperation with the Association of Netherlands Municipalities (VNG). Rotterdam joins the project early 2020 and appoints data scientist Amor Frans. Amor has been seconded by Qualogy at the municipality of Rotterdam since 2018.
Aside from The Hague and Rotterdam, municipalities Velsen, Schiedam, Zoetermeer, Amersfoort, and Roosendaal also joined the project. They have all appointed a data scientist. A project leader will be in charge. Amor: “It was amazing to be able to work on a project with great social value. I enjoyed the challenge of working together with multiple municipalities and data scientists. During the project, my responsibility was the prognosis. To accomplish this, I deployed machine learning. With multilevel modeling in programming language R in particular.”
In short: The Social Support Act
The Social Support Act (Wet maatschappelijke ondersteuning) has become effective as of 2015. It enables municipalities to help people live at home as long as possible so they can actively take part in society. Examples of the municipalities’ tasks are installing a shower seat or arranging transport to a club or association. Municipalities pay (part of) the costs. Because of i.e. an aging society, costs have been rising for years. That is why municipalities want to gain insight into the number of prospective clients.
The national Social Support Act prediction model
For this purpose, the national Social Support Act prediction model has been developed. The goal is to give insight into the progression of the number of Social Support Act clients per neighborhood. This is essential for the policy makers. If they know how the numbers develop, they can compose a good policy.
The prediction model uses different data streams, such as the demographics of various neighborhoods. But less obvious data is also used to make the best possible prediction.
Project transferred to the VNG
In June, the project was transferred to the Association of Netherlands Municipalities (VNG) by alderman Kavita Parbhudayal of the municipality of The Hague. Multiple interested municipalities attended the presentation. Their responses to the prediction model were enthusiastic. VNG aspires to have every municipality use the model.
The Association of Netherlands Municipalities (VNG) is responsible for the development and the national implementation of the Social Support Act prediction model. The third development stage will soon start. The aim is to have finished the development phase by the end of 2020, and to start implementing the prediction model in multiple municipalities.
Proud of the project
Amor is proud of his contribution to the project. He says: “I’m very happy with the result. It always remains to be seen how a project will be carried out with so many different stakeholders and team members. The project leader did a great job in ensuring the cooperation went smoothly. An added challenge was working from home. We worked in Microsoft Teams during the entire project. This went perfectly. The result: a great model with a high social value.”
Amor is seconded to the municipality of Rotterdam until the end of 2020 at least. Aside from the national Social Support Act prediction model, he worked on different, interesting data science projects, like improving the accessibility of the municipality with Textmining, predicting the costs of youth care, and making transparent which factors are important for the development of the youth of Rotterdam.
Curious to know what Qualogy’s data science team could mean for your organization? Contact us via email@example.com or 070 319 5000.