Skip to main content

Grant-application essentials — child poverty

For Lisa, tilskuddsansvarlig

Grant-application essentials — child poverty

The task
Writing a child-poverty grant application that names a kommune, backs the claim with real numbers, and survives a fact-check by a directorate's reviewer.
A focused starter set for grant applications about child poverty at kommune resolution. Ordered from headline → ground truth → temporal depth → lived-experience proxy → Atlas synthesis. Together these five datasets answer the four questions every directorate reviewer asks: how many, how persistent, how that compares to the neighbours, and what the policy gap looks like. Every dataset here is at kommune resolution. They share a join key — `kommune_nr`, the 4-digit zero-padded kommune number — so a single query stitches the whole evidence base together.

Datasets in this collection​

  1. 1
    Barne-, ungdoms- og familiedirektoratet logo
    Barnefattigdom kommunemonitor
    Barne-, ungdoms- og familiedirektoratet · 💰 Income & poverty
    Stable
    The headline indicator most directorates already cite. Bufdir's Barnefattigdom kommunemonitor is the standard reference for "how many children live in low-income households in this kommune".
  2. 2
    Statistisk sentralbyrå logo
    08764: Personer under 18 år i husholdninger med lavinntekt (EU- og OECD-skala). Antall og prosent (K) (B) 2005-2024
    Statistisk sentralbyrå · 💰 Income & poverty
    Stable
    The SSB ground truth. Bufdir's monitor is derived from this table. Cite SSB when a reviewer wants the canonical source; cite Bufdir when you want the editorial framing.
  3. 3
    Statistisk sentralbyrå logo
    12944: Personer i husholdninger med vedvarende lavinntekt (EU-skala 60 prosent), etter alder. Treårsperiode (K) (B) (2011-2013)-(2022-2024)
    Statistisk sentralbyrå · 💰 Income & poverty
    Stable
    Persistent low-income (3-year average, EU-60 scale). The difference between *transient* and *persistent* poverty is the difference between a household between jobs and a household stuck. Reviewers read persistence as a stronger evidence base.
  4. 4
    Folkehelseinstituttet logo
    Trangbodd_UTDANN
    Folkehelseinstituttet · 🎓 Education & youth
    Stable
    A lived-experience proxy: children in overcrowded housing. Closely correlates with low income but adds the "what does poverty feel like" dimension that grant narratives need. Pairs well with the income tables above.
  5. 5
    Child-poverty coverage gap (kommune)
    Atlas view (api_v1.*)
    The Atlas synthesis view. Combines Bufdir's monitor with SSB's coverage to surface the kommuner where Bufdir didn't publish and where, if your application is about underserved data, you have a story to tell.

How to use these together​

All five datasets join on kommune_nr — the 4-digit, zero-padded kommune number. A minimal query that pulls one kommune's evidence base (PostgREST):

curl -sS 'https://api-atlas.urbalurba.com/indicator_latest_values?kommune_nr=eq.0301&select=kommune_name,source_id,contents_code,value,year'

Per-kommune values live in indicator_latest_values. indicator_summary is one row per (source, measure) across the whole country — coverage and value range, no kommune column — so it answers "what does this indicator look like nationally", not "what is the number for my kommune".

To compare kommuner within a fylke, use the two-digit fylke prefix: kommune_nr=like.11* is Rogaland's 23 kommuner.

Oslo is the exception, and it is the example directly above: fylke 03 contains only kommune 0301, so like.03* returns Oslo by itself. If you try the pattern on Oslo first, one kommune is the correct answer, not a broken filter.

Watch out: the number is zero-padded, so Oslo is 0301, not 301. And kommune numbers are reissued by reform: Bærum was 3024 under Viken (2020-2023) and is 3201 today. A retired number returns 200 with an empty array rather than an error — dim_kommune carries both, distinguished by is_active, so filter is_active=eq.true when you join against it. Concretely: dim_kommune?kommune_nr=like.11* returns 57 rows for Rogaland — 23 active kommuner and 34 retired numbers. Without the filter you would count Rogaland's kommuner as 57.

Only the curated api_v1 relations are public. The upstream source tables are not exposed through this API; use meta_sources to see which relations publish a given source.