The Magenta Book evaluation guidance update 2026 has arrived with serious advice on public money and a notebook demanding a control group for the biscuit tin. The refreshed guidance covers value for money evaluation, test and learn approaches, and the alignment of benefits management with monitoring and evaluation.
Magenta Book evaluation guidance update 2026 meets the notebook
The Magenta Book is central government guidance on evaluation. It helps policy, delivery, finance and analytical teams ask whether an intervention worked, why it worked, who benefited and whether the result justified the cost.
That matters because public money remains serious, even after someone places it inside a business case and teaches it to sound confident. The guidance encourages teams to plan evaluation early. It also supports testing, monitoring and adjustment before a policy grows too large to question politely.
The system is being encouraged to learn before it repeats the same mistake nationwide and commissions a celebratory report about the journey. This is a sensible ambition. It is also a dangerous one for any spreadsheet that has spent years presenting confidence as evidence.
Three questions and one suspicious teabag
The guidance distinguishes process, impact and value for money evaluation. Process evaluation asks how something was delivered. Impact evaluation asks what difference it made. Value for money evaluation asks whether the result justified the resources.
The notebook has added tea break evaluation. It asks whether anyone explained the form clearly enough for a human being to complete it without consulting a second notebook. The answer is currently trapped in an annex.
The Magenta Book evaluation guidance update 2026 also stresses transparency and the difference between correlation and causation. This is awkward news for the successful looking graph. It has spent years entering official statements through the side door while the underlying question waits outside with no pass.
A chart can show movement. It cannot prove that a policy caused the movement by itself, unless the chart has recently been appointed permanent secretary.
The evidence must survive the paperwork
Evaluation can expose unintended consequences, improve decisions and make spending more accountable. It cannot guarantee that every policy works. It can make it harder to rename uncertainty as a communications challenge and send the resulting fog to a design agency.
The notebook has completed its own assessment. The teabag delivers strong colour, weak structure and no measurable improvement after five minutes. The final recommendation is that the tea be escalated to a cross department working group, where it can spend three years proving that boiling water had an impact.
Source: GOV.UK.
