How Socyti works

Ranking methodology

Socyti compares Sweden's 290 municipalities across 15 life domains. For each domain you will see three things. They are related, but they do not mean the same thing:

Rank (e.g. "Rank 34 of 290") – the municipality's position when all 290 municipalities are sorted by their score in that domain. Rank 1 is the strongest relative position, rank 290 the weakest. A rank is always relative: it shows where a municipality stands compared with the others, not how "good" something is in absolute terms. A municipality at rank 145 of 290 is in the middle, not halfway to a goal.

Band (e.g. "Top 25% nationally") – the broader national grouping the rank falls into. A band is a descriptive grouping, nothing more: it exists because a single rank is easily over-read. The difference between rank 140 and rank 155 is often small.

Score (0–100) – the combined value the rank is based on. Two municipalities can be close in score yet several ranks apart.

How the score is calculated. Scores and ranks are calculated deterministically according to a fixed method, without AI and without manual adjustments of individual municipalities. Within each domain, the key figures are normalised so they can be compared on the same scale, and each key figure is treated as favourable or unfavourable for the domain based on what it measures, never as a blanket rule for a whole domain. The key figures are then weighted equally within the domain: the domain score is the mean of the normalised key figures. The overall position follows the same principle: the mean of all 15 domain positions, each domain weighing equally.

We have no thumb on the scale. No factor has secretly been given more importance than any other. If you want to weight by what matters most to you, you do that yourself in the Flyttprofil (section 08).

A worked example. Take Lerum in the Income domain. The domain is built from four key figures from Kolada, including mean and median earned income for ages 20–64 and income distribution. Each key figure is normalised, and the domain score is their mean: 75.3. When all 290 municipalities are sorted by their income score, Lerum lands at rank 34, corresponding to the band Top 25% nationally. The example illustrates the chain step by step: public key figures → normalisation → equal weighting → score → rank → band.

What a rank means and what it does not

A rank describes measured conditions, not future outcomes.

If a municipality is ranked 34 of 290 in Education, its measured indicators (such as qualification rates, results, and resources) place it there compared with the other municipalities, using the latest data in Socyti's current data version.

It does not mean your child will do better in school there, or that moving there will change your family's outcomes. The statistics describe the municipality as a whole, not individual schools, workplaces, or households. Nor do they describe cause and effect. A high rank in education does not necessarily mean better teachers. And a rank is never a judgment of a municipality's effort, leadership, or residents. Measured differences often reflect structural conditions, in both directions.

Use a rank for what it is designed for: a starting point for comparison and your own questions, not a prediction.

Data sources

All data comes from public Swedish authorities and registers. We use no private data sources.

• Kolada (the council for the promotion of municipal analysis, RKA) – primary source, municipal key figures across all 15 domains
• SCB (Statistics Sweden) – population, income, tax, labour market, demographics
• Skolverket (the National Agency for Education) – compulsory and upper secondary school results, teacher density

Socyti maps 230 key figures in total, of which 163 are score-forming. These build the scores and ranks. The remaining 67 are contextual key figures shown as information but never affecting any rank. Data is updated annually as the authorities publish new figures, usually in the first half of the year.

The source data is public: anyone can retrieve the same figures directly from the authorities' own APIs.

15 life domains

Municipalities are compared across 15 thematic domains covering the most important aspects of life in a municipality, from childcare and education to transport, safety, and sustainability. Each domain card on the municipality pages shows which key figures are included, how many are score-forming, and how current the data is.

AI-generated texts

The descriptive texts on municipality pages are AI-generated, under strict conditions. A pipeline of specialised steps interprets the request, retrieves the municipality's key figures and ranks from the database, generates the text from that data alone, and automatically checks the finished text against the source records. The AI model has no access to the internet or external information. It works exclusively with the traceable source records our pipeline provides.

The check is designed so that every factual claim can be traced to a source record. Texts where the check detects claims that cannot be traced to the source are not published. The check reduces the risk of incorrect or untraceable claims, but it does not replace your own judgment or the option to request a correction (section 10).

Most importantly: the texts never influence scores or ranks. The calculation is deterministic and entirely without AI (section 01); the texts exist to make the numbers readable. Every AI-generated text on the municipality pages is also labelled as such where it appears.

Data quality and limitations

No data source is perfect, and no method is free of choices. These are the most important limitations to know about:

Time lag. Data is published with a 6–18 month delay. Figures labelled "2024" usually refer to the operational year 2023. Each domain card shows the newest and oldest data year.

Missing values. Not all municipalities report all key figures. Where a key figure is missing, the national median is used for that specific measure, rather than treating the value as zero or excluding the municipality entirely. This reduces the risk that missing reporting in itself produces extreme swings, but it also means the rank then rests on fewer observed values. Each domain card openly shows how many key figures the domain is built on and how current the data is.

Contextual differences. Small municipalities and large cities operate under fundamentally different conditions. The ranking compares measured key figures the same way for all 290 municipalities. It is descriptive, not a model that adjusts for structural starting points. This cuts both ways: a smaller municipality ranking above a large city does not make the city worse, and a municipality with tough structural conditions ranking low is not a judgment of the municipality or its residents. Comparing municipalities with similar conditions is often most informative.

Indicators, not causes. Key figures show results and conditions, not cause and effect.

Method dependence. A rank depends on which key figures are included and how they are made comparable. Our selection is based on availability and relevance, not completeness, and all combined comparisons are sensitive to such method choices. That is why the method is fixed: it changes only in versioned updates that are openly documented.

We work continuously to improve data quality. If you find an error, you can request a review (section 10).

Our philosophy, and what neutrality means here

Socyti presents data as a mirror, not a judge. We do not create winners and losers among Sweden's municipalities, never use letter grades or words like "best" and "worst", and show every municipality with its strengths first, without hiding weaker positions. The colours represent relative position, not traffic lights; trend arrows show direction without judgment.

Neutrality does not mean the method is free of choices. Which key figures are included and how they are made comparable are choices we have made and disclose. What neutrality means at Socyti is verifiable:

• The same calculation for all 290 municipalities, no exceptions.
• Equal weighting: no factor has been given hidden extra weight.
• No manual adjustments of individual municipalities' results.
• No one can pay for a rank, whether municipalities, companies, or anyone else.
• Traceable sources: every figure traces back to a public source.
• A versioned method: changes to the method are documented openly, never silently.

The Flyttprofil: your priorities, the same data

The Flyttprofil is Socyti's personal matching tool. It uses exactly the same data and the same method described above, with one difference: you set the weights. You choose the life domains that matter most to you and how much they should count, and the Flyttprofil calculates a personal comparison of all 290 municipalities based on your choices. It is designed especially for anyone comparing municipalities ahead of a move, a housing choice, or other family decisions.

This is also the answer to a fair objection to equal weighting: not everyone prioritises equally. That is precisely why Socyti adds no judgment of its own on top of the data in the base ranking, and gives you the tool to add yours.

A tool for comparison

Socyti is a basis for comparison, not a complete basis for a decision. A moving decision weighs many things no statistic captures: people, distance, work, feeling. Use Socyti to ask better questions and compare on equal terms. Visit the municipalities. The decisions are yours.

Corrections and review

If you believe a figure, description, or interpretation is incorrect, you can request a review via kontakt@socyti.se. We then review the source value, the calculation step, and the published text. If an error is confirmed, it is corrected and the change is documented in the relevant data version or method note.