Our client wanted to estimate the potential cost of a policy change to free childcare provision in England and Wales.
The project required both a high-level view of the estimated impact and a drill down into which local authorities would be most greatly impacted.
What we did
✔︎ Sourced relevant open government data sources
✔︎ Designed a parameterised model at the authority level, that could be used to test scenarios
✔︎ Estimated parameters for each authority using the data
✔︎ Created an interactive PowerBI dashboard to explore results
✔︎ Produced a whitepaper detailing the approach and findings
Our scenario planner enabled the client to test a number of different scenarios and adjust the parameters of the models to assess the potential cost of the policy changes.
As a result of our work, the client received a bespoke dataset that can be used as a single source of truth for childcare costs across England and Wales, assimilated from a variety of disparate data sources. We took care to explain the joining logic in detail and identify which additional data sources could be used to refine the estimate further if available.
authorities used to estimated the overall impact of the policy change
How we did it
Through collaboration with internal stakeholders, we first sourced the relevant datasets for analysis. The data was disparate, with non-uniform joining fields, so we wrote a reproducible pipeline for cleansing and aggregating the data into a single source of truth.
We then built an algorithmic forecasting model for predicting the potential cost of the new policy, given parameters estimated from the data. This was designed to be flexible to changes in parameters if additional expert knowledge needed to be incorporated.
The analysis was delivered through an interactive dashboard so that the client could observe where the impact would be most greatly felt, and why. Our whitepaper explained the rationale behind our methodology in detail.
Start a conversation
Take the first step by speaking with one of our data experts today.