European Sovereignty Monitor

Across the four areas

The European Sovereignty Index

One number per area and one for Europe, saying which way European technology sovereignty moved, built from every graded news item this publication has published. The key between the four areas is the editor's, and you can move it.

The index is drawn from the publication's own file, index.json, and needs JavaScript to draw here. The file itself is open.

How to read it

Every published news item on this site carries a reading, shown on the item under "How this item counts": the direction it pushes Europe's dependence, which of the four questions it changes (who owns, who operates, whose law governs, who made the parts), how binding it is, how big, and how sure we are. The weight is the product of those, at most 36. The bars above are the sum of the month's signed weights per area, divided by 36, so one maximal item equals 1.0. The line is the running total from the start date. The pipeline proposes each reading and the editor approves it with the grade; the method page has the full rubric.

What the numbers mean

There are two scales on this page, and they answer different questions.

The level, 0 to 100, answers "where does Europe stand?" Each pillar's level is the plain average of five indicators, and each indicator is scored the same way: distance to the EU's own stated target where one exists, or the European share where only a control reading exists. So the scale reads:

  • 100 means the EU meets its own target in full, or the control test finds full European control. It does not mean sovereign: the ceiling of this scale is the ambition the EU set for itself (20 percent of chips by 2030, 50 percent of defence procurement at home, every directive transposed), and a row can sit at 100 while real exposure remains; the "does not test" line on each row says where.
  • 75 means three quarters of the way there: the capability or rule exists and covers most of what the EU says is needed, with a named remainder (six of the eight launches, twenty of the twenty-seven transpositions).
  • 50 means halfway: real capability, real rules, covering about half of the stated need. A pillar at 50 is not failing; it is exposed, and the exposure is specific and listed.
  • 25 means the ambition is mostly on paper: the target exists, a first slice is real, and the bulk has not happened.
  • 0 means nothing of the target is met, or the control test finds nothing European. The zero on this page is real: none of the binding instruments on the regulation timeline tests who made the parts.

The flow, the signed numbers on the charts, answers "which way did it move?" One maximal development (a Verified, in-force, large development that changes all four sovereignty questions) counts 1.0 in its month; ordinary items count 0.05 to 0.35. A month at zero means nothing graded moved, whatever the level; a negative month means the graded developments deepened dependence. The running total accumulates from the start date: +1.0 reads as one maximal development's worth of net movement towards autonomy, however many ordinary items produced it, and it moves with every published item.

The two scales are deliberately separate: news can flow towards autonomy for months while the level barely moves, because levels measure stock against targets and the flow measures graded events. The level is reassessed quarterly with every source named and dated; the flow updates with every published item.

The key

Nobody agrees how much law weighs against compute, or defence against industry. The editor's default is compute 30, policy 25, defence 25, industry 20, for the reasons printed under each slider. It is a decision of this publication, not a finding, which is why the sliders are there: set your own and the combined number redraws, and the address carries your key so you can send it to someone who disagrees.

What it does not test

The flow says how much graded movement there was, not how much sovereignty Europe has. A quiet month scores 0 whatever the state. Two items of equal weight in opposite directions cancel, which is the honest reading of a mixed month, not a claim that nothing happened. The items are the ones this publication picked, and the picks lean towards European moves, so the flow leans positive. The corrective is the level layer, twenty indicators scored against the EU's own targets, which follows. The start date is 1 January 2026 for now and moves back as older items enter the archive.

The data

The numbers come from index.json, rebuilt with every publication, and the readings from news.json. Both are open, CC BY 4.0, and the connector and API serve them to your own tools.