Members' expenditure · April 2025 to March 2026
View 01
A party’s total spend mostly just restates how many seats it holds, so this view divides by seats instead. On that basis the parties turn out to be remarkably close together, and the interesting variation sits between individuals rather than between parties.
Drill down
Click any bar above to see that party's averages, range and full member list.
View 02
Each blob is one member who served the full year. The useful question is not who tops a league table, it is whether your MLA sits anywhere unusual. Most of them don't.
Hover for a name, click for the full breakdown. The vertical rule is the median. Colour marks nothing except what you have selected, on purpose.
View 03
Staff cost is 82% of the whole thing. Once you see that, most arguments about MLA expenses turn out to be arguments about how many people they employ.
View 04
Five members each, everywhere. Pick a constituency and compare the people who represent the same streets as each other.
View 05
Ministers cost slightly less than backbenchers, not more. And you can identify every single office holder from the allowances column alone, because a government car means they stop claiming travel.
Each office holder's allowances against the average claimed by the other members of their own constituency.
| Member | Role | Claimed | Neighbours | Gap |
|---|
Constituency neighbours means the other full-year members returned for the same constituency.
View 06
Members make two separate spending decisions. Constituency office spend is rent, rates and equipment: the shopfront. Staff cost is how many people they employ. You might expect these to rise together, on the theory that some members are simply bigger spenders than others. They don’t. A member with an expensive office is no more likely to have a large team than a member with a cheap one, which means there is no such thing as a high-spending MLA, only MLAs who spend on different things.
Each dot is one full-year member. Click one for the full breakdown. The dashed lines split the chamber at the halfway point on each measure, so the corner someone sits in tells you which of the two they lean towards. For anyone who wants the statistic: the relationship between the two measures scores 0.11, where 0 would mean no connection at all and 1 would mean they rise and fall together.
View 07
Worth showing precisely because there is no story here. Split the chamber by designation and the cost per seat lands within about 1.5% across all three. Anyone arriving at this data hoping to find that one side costs more will leave disappointed.
Designation follows each party's own declared position. Independent and People Before Profit are grouped with Other, which is where they sit in the chamber's own arithmetic.
View 08
Six of the 94 records are not a normal year in post. Read them as frugality or extravagance and you will be wrong about all six. This view exists so nobody has to.
View 09
That depends entirely on what you divide it by, which is the whole problem with big numbers in headlines. Here are the divisions that seem fair.
Every figure comes from the Assembly’s published Members’ Expenditure return for April 2025 to March 2026. All 94 records and all five columns reconcile exactly to the published totals, so nothing here has been estimated.
Party affiliation is as at 31 March 2026, the last day of the period. Doug Beattie is recorded as UUP because he sat as a UUP member throughout this year; he left the party on 31 May 2026, after the period closed. Gary Middleton is recorded as a full year for the same reason, having resigned in April 2026.
Per-seat figures divide by the party’s current seat count, not by the number of records, since three seats changed hands mid-year and would otherwise be counted twice.
Colour here marks selection and category, never party. Assigning parties their usual colours would have made a chart about payroll look like a chart about something else.