Data · · 4 min · Dr. Krishna Hari Pushkar
More than a quarter of Catalonia’s residents were born outside Spain. The figure is accurate, widely quoted, and almost useless for planning a service, because the thing it averages away is the thing you need.
25.1% of Catalonia's residents were born outside Spain: 2.04 million people, on the 2025 Idescat series. It is a real number from a good source and you will see it in every deck about Spanish demographics, including ours.
It is also the wrong resolution for almost every decision a municipality actually makes.
A regional average of 25.1% is compatible with a very large number of underlying distributions. It is compatible with every neighbourhood in Catalonia sitting at 25%. It is also compatible with half of them at 8% and the other half at 42%. The reality is closer to the second, and the operational consequences of the two scenarios have nothing in common.
Thirty-six Catalan municipalities already exceed a 25% foreign-born share (Idescat, 2025). Within several of those municipalities, individual barris are well above and well below the municipal figure. A school catchment, a primary care centre, and a social services district all operate at that resolution. None of them operate at the regional one.
Three reasons, in increasing order of how much they cost.
Services are sited, not averaged. A municipality does not deliver integration support to a regional average. It delivers it at a specific address, with a specific catchment, and a specific staffing level. A figure that cannot be resolved to that address cannot inform the decision.
Averages hide the direction of travel. A stable 25% and a 25% that was 18% three years ago are the same number and completely different problems. Stock tells you where you are. Flow tells you what is about to happen. Most published series are stock.
The tails are where the work is. The neighbourhoods that need attention are, by definition, not at the average. A metric that describes the centre of a distribution is silent about exactly the places a director is accountable for.
This is not an argument that the regional number is worthless. It is good for three things and we use it for all three.
What it cannot do is tell a social services chief in a particular municipality what changed in their district last month. Nothing published currently does that, which is the actual problem.
| Question | Resolution needed | Resolution published |
|---|---|---|
| Is this a region-scale phenomenon? | Regional | Regional, annual |
| Which municipalities are under most pressure? | Municipal | Municipal, annual |
| Which barris in my municipality moved this quarter? | Sub-municipal | Not routinely published |
| Which service is about to be over capacity? | Service catchment | Not published at all |
The bottom two rows are the gap. They are not published because sub-municipal aggregation at short intervals runs into two genuine problems: cell sizes get small enough to raise real disclosure risk, and boundary definitions are unstable enough that a series can move because a line moved rather than because anything happened.
Both problems are solvable and neither is solved by wishing. Small cell sizes require a minimum-cell-size floor applied at query time across every filter combination, including exports, which is an architectural decision rather than a policy one. Boundary instability requires pinning and versioning the boundary definition and recomputing history when it changes, rather than silently redefining a series.
We quote 25.1% where it belongs, which is in the paragraph establishing that the phenomenon is real and at scale, with the source and the year attached. We do not quote it as evidence about any particular neighbourhood, because it is not evidence about any particular neighbourhood.
And when someone shows you a slide where a regional average is used to justify a local intervention, the useful question is: what is the dispersion, and what is the lag?
Every figure on this page describes populations, not individuals, and describes a system, not a group of people. A neighbourhood where the foreign-born share rose is not a neighbourhood with a problem. It is a neighbourhood where the instrumentation should be able to tell a director whether the local school, the local health centre, and the local employment office are sized for who lives there now. That is a question about public administration, and it is the only question this data is entitled to answer.
Follow the work
We publish when there is something to publish, which is roughly monthly. If you would like the next piece, write to hello@humaniwork.com with the word Signals. No list, no automation, no sequence. One of us adds you and one of us sends it.
Field brief
15 JUL 2026 · 4 min
The interval between a neighbourhood signal appearing in administrative data and the point where a response still changes the outcome is about six weeks. Most municipalities are working eighteen months behind it.
Dr. Krishna Hari Pushkar
Government
12 JUN 2026 · 4 min
Six questions decide whether a municipal AI pilot happens. They are asked in the same order every time, and a vendor who cannot answer the first two rarely gets to the third.
Desh Deepak
Case study
22 APR 2026 · 4 min
We took the diagnostic requirements of a live Catalan funding instrument and checked each one against what a municipality can actually produce from published sources today. Roughly half are answerable. The unanswerable half is consistently the operational half.
Dr. Krishna Hari Pushkar