How to use Nomis to find local jobs, pay and benefits data
Nomis is the ONS service for local labour market statistics. The quickest route is a Local Authority Profile, which puts population, employment, unemployment, earnings, benefit claimants and jobs for one council area on a single page beside the region and Great Britain. The main skill is telling apart measures that look alike: survey unemployment, model-based unemployment and the claimant count are three different numbers, and pay can be measured where people live or where they work.
What Nomis is and when to use it
Nomis describes itself as "a service provided by Office for National Statistics (ONS)" that publishes statistics on population, society and the labour market at national, regional and local levels. Use it when a claim is about a particular town, council or constituency: "unemployment here is double the national rate", "wages in our city are below average", "more people are claiming benefits". For UK-wide headline rates, the monthly ONS labour market bulletin is the better starting point; see our guide to labour market statistics .
Source: Nomis
Step 1: open the right ready-made report
From the Nomis menu, choose Reports . The area reports most people need are:
Local Authority Profile for a district, unitary or county council area;
2024 Parliamentary Constituencies Profile for a constituency on its current boundaries;
Local Area Report for smaller areas such as wards, mostly built from census data.
Pick the area from the list. Each profile is laid out in the same order: resident population, labour supply (employment, unemployment, inactivity), qualifications, earnings, out-of-work benefits, jobs and businesses. Every table shows the local figure next to the region and Great Britain, and has links to view the time series, compare other areas or download the data.
Sources: Local Authority Profile, 2024 Parliamentary Constituencies Profile, Local Area Report
Step 2: read the source line under every table
Each table in a profile comes from a different source, with its own period. Note both before quoting a figure:
Employment and unemployment come from the Annual Population Survey, a sample survey covering a rolling 12 months.
Model-based unemployment combines that survey with benefit data to steady small-area estimates.
Claimant count is a monthly administrative count of people claiming unemployment-related benefits; see claimant count .
Earnings come from the Annual Survey of Hours and Earnings (ASHE), which the profile describes as "based on a 1 per cent sample of employees" and which does not cover the self-employed.
So one page can mix a single month, a 12-month survey period and a calendar year. A claim that sets one against another should say so.
Worked example: three unemployment numbers for Manchester
On 13 October 2026 we opened the Nomis labour market profile for Manchester . It showed a resident population of 588,256 in 2025. It also showed three figures that could each be described as "unemployment in Manchester":
Survey unemployment (April 2025 to March 2026): 12,400 people, 4.0% of the economically active.
Model-based unemployment (same period): 15,800 people, 5.0%.
Claimant count (August 2026, not seasonally adjusted): 25,185 people, 6.0%, against 4.2% for the North West and 3.9% for Great Britain.
These are not three estimates of the same thing. The profile itself says that because of small sample sizes, "ONS recommends people use the model-based version, which makes use of both survey and administrative data." The claimant count is a different measure with a different denominator: its note says the percentage is "the number of claimants as a proportion of resident population of area aged 16-64", not of the economically active. It also warns that under Universal Credit "a broader span of claimants are required to look for work", so the count can include people in work on low earnings.
So if someone says "6% of Manchester is unemployed", the source is probably the claimant count, and the accurate description is "6.0% of residents aged 16 to 64 were claiming unemployment-related benefits in August 2026". For the ILO unemployment rate the profile points to 5.0% (model-based, April 2025 to March 2026). See ILO unemployment and how UK unemployment is calculated .
Source: Nomis labour market profile for Manchester
Worked example: pay where people live or where they work
The same Manchester profile has two earnings tables for 2025, both median gross weekly pay for full-time workers:
By place of residence (people who live in Manchester): £715.60 a week, against £766.60 for Great Britain.
By place of work (people whose job is in Manchester): £794.80 a week, against £766.60 for Great Britain.
One table puts the city below the national median, the other above it. Both are accurate. Many higher-paid jobs in a city centre are filled by people who commute in from elsewhere. A claim that "Manchester pay is below average" and one that "Manchester pay is above average" can both cite this page, so check which table each one used. Both are medians; see median .
Step 3: build your own table with the query tool
For anything a profile does not show, such as a longer time series or a comparison of several councils, use Query data . The steps are always the same:
Choose a dataset by theme, for example Claimant Count or Annual Survey of Hours and Earnings.
Choose the geography type and tick the areas you want.
Choose the dates. For survey data, pick periods that do not overlap if you want independent comparisons.
Choose the variables (for example sex, age, or median rather than mean pay).
Choose the format and download.
Each data source also has a page explaining what it measures, such as the claimant count , Annual Population Survey and ASHE pages. Census tables from 2021 back to 1921 are listed on the Nomis census pages ; see also where to find census data .
Sources: Query data, claimant count, Annual Population Survey, ASHE, Nomis census pages
Common traps
Great Britain, not UK. Profile comparison columns are for Great Britain, so Northern Ireland is not included. See Great Britain and the UK .
Small samples. Some survey cells are marked as too small for a reliable estimate. Survey figures for one council can move a lot from one period to the next without any real change; see confidence intervals .
Not seasonally adjusted. Local claimant counts are usually unadjusted, so compare the same month in different years rather than consecutive months. See seasonally adjusted .
Counts against rates. A big city will always have more claimants than a small town. Compare rates, and check that both use the same denominator.
Checklist
Right report chosen (council, constituency or small area)
Source and period noted for each figure
Survey, model-based and claimant measures kept separate
Denominator checked (economically active or population aged 16 to 64)
Residence or workplace earnings identified
Comparison area noted (region, Great Britain)
Same months compared for unadjusted data
Official sources used in this guide
Sources: Nomis, Nomis: local authority profiles, Nomis: labour market profile for Manchester, Nomis: 2024 parliamentary constituency profiles, Nomis: query data by theme, Nomis: claimant count source information, ONS: Labour market overview, UK
Questions
Is Nomis official?
Yes. It is provided by the ONS, and the profiles show the original source and period for each table.
Source: Nomis
Which unemployment figure should I quote for my council?
For unemployment on the ILO definition, the ONS recommends the model-based estimate for local authorities. Quote the claimant count only as a count of benefit claimants, with its month.
Source: Nomis
Can I get data for my constituency?
Yes. Use the 2024 Parliamentary Constituencies Profile, which uses the boundaries in force since the 2024 general election.
Source: Nomis
Does Nomis cover Northern Ireland?
Some datasets do, but local authority profiles compare with Great Britain. For Northern Ireland, see our NISRA guide .
Source: Nomis