🔵 Forecasting

Probabilistic forecasting, model evaluation, and where forecasting stops being causal inference.

Forecasting answers a different question than causal inference — not what would have happened otherwise, but what is likely to happen next. Both matter. Neither replaces the other. The posts on this page walk through the methods I use and, more importantly, the ones I choose not to use.

If you’re coming from the causal-inference arc, start with Forecasting Is Not Causal Inference — it’s the bridge between the two.

Posts in this series

Forecasting is not causal inference: a practitioner’s bridge

A practitioner’s guide to a distinction that quietly wrecks business decisions: predicting what will happen next is not the same task as predicting what would happen if you changed something. The bridge between this blog’s causal inference series and its forecasting series.

Aug 3, 2026

51 min

Projecting teacher requirements in mainland Portugal to 2040

A sixteen-year projection disaggregated by NUTS III region and schooling cycle. Cumulative recruitment reaches 90,575 teachers, an average of 5,661 a year, against a historical mean of 4,537. Between 58 and 64 per cent of the 2024 workforce retires within the horizon, and the system recruits fewer teachers than it loses because the number of posts is falling.

Sep 21, 2026

12 min