Overview
EU salaries are not negotiated. They are indexed. Each year, the basic salary tables in Annex XIII to the Staff Regulations are multiplied by a single percentage that comes out of an arithmetic formula. The formula is set in Annex XI to the Staff Regulations and is implemented under Article 65. The Council adopts the result; it does not produce it. The whole point of Annex XI is that the figure is determined by data, not by political bargaining.
This page walks through how that formula works in practice. For the higher-level context — what the adjustment is and why it matters — see the salary adjustments hub. For the current state of the 2026 update, see the 2026 page.
Article 65 of the Staff Regulations
Article 65 sits in Title V, Chapter 1 of the Staff Regulations, which deals with remuneration. Two paragraphs do most of the work.
Paragraph 1 says that the Council shall each year review the level of remuneration of officials and other servants of the Union, and shall update remuneration in light of changes in the cost of living and changes in the purchasing power of national civil servants. Paragraph 2 says that updates take effect from 1 July of the year of effect.
That is all Article 65 itself contains. The mechanics — which national civil servants count, which inflation index to use, how the percentage is computed, and what corrective mechanisms apply — are delegated to Annex XI of the Staff Regulations, which Article 65a treats as the implementing rules of Article 65.
The current Annex XI was adopted as part of the 2014 staff reform under Regulation (EU, Euratom) No 1023/2013. It replaced an earlier version that had been the subject of repeated litigation in the early 2010s. The current version contains a moderation clause, an exception clause and a tighter set of data requirements.
The Formula in Plain Terms
The headline annual update is the product of two factors:
- The specific indicator: the change in the purchasing power of central-government civil servants in the reference Member States, computed by Eurostat.
- The Brussels-Luxembourg inflation factor: the change in the harmonised consumer price index for Brussels (with a Luxembourg correction for staff posted there) over the reference period.
Multiply those two factors and you get the annual update. The two factors are computed over the same twelve-month reference window: from 1 July of the previous year to 30 June of the year of effect.
In equation form, with everything expressed as percentage changes:
Annual update = (1 + specific indicator) × (1 + Brussels-Luxembourg inflation) − 1
In normal years the two factors are small numbers; the product is close to their sum. When the inflation factor is large — as in 2022-2023 — the multiplicative form matters and produces a noticeably higher headline than a simple sum would.
The result is rounded to two decimal places per the rounding rules in Annex XI, and is applied to every line of the salary, allowance and pension tables.
The Reference Member States
The specific indicator is computed from a fixed basket of reference Member States, chosen for their representativeness in EU central-government employment. The basket is set in Annex XI and is reviewed only when Annex XI itself is amended; it does not change year to year. The exact list and the weighting are part of the Annex XI text and apply to every annual update under the current methodology.
For each Member State in the basket, Eurostat collects the gross pay scales of central-government civil servants from the national statistical office. The data series cover all main civil service grades and seniorities, weighted to a common base so that compositional changes do not distort the indicator. The aggregate is then deflated by the corresponding national consumer price index to produce a real-terms figure. The change in that figure is the contribution of that Member State to the specific indicator.
The contributions are combined with the Annex XI weights to produce a single specific indicator for the year. The weights are based on each Member State's share of central-government civil service employment.
The strength of this design is that it pegs EU pay to a representative sample of national civil service pay. The weakness, identified in successive evaluations, is that it relies on national statistical offices having timely and consistent data; in practice, some Member States revise their inputs after the cut-off, which Eurostat then reflects in the next annual update.
The Brussels-Luxembourg Inflation Factor
The second factor in the formula captures the change in the cost of living for staff in Brussels, the reference duty station. Eurostat publishes a harmonised consumer price index for Brussels, computed from data supplied by Statbel. The index is a harmonised HICP, comparable across EU countries. The June 2026 reading versus June 2025 is the relevant change for the 1 July 2026 annual update.
For staff posted in Luxembourg, a separate Luxembourg HICP is used, with the same June-to-June reference period. Where the Luxembourg figure differs from Brussels, Annex XI provides for a Luxembourg correction so that staff in Luxembourg are not under-compensated for cost-of-living changes specific to that duty station.
The Brussels-Luxembourg inflation factor is what made the 1 July 2023 update so large. The June-to-June reading was the highest since the formula was introduced and combined multiplicatively with a positive specific indicator to produce an adjustment of around 8.5%.
The Joint Committee on Remuneration
Annex XI provides for a Joint Committee on Remuneration in which the institutions, the Member States represented by the Council, and the staff representatives examine the Eurostat data ahead of the Commission's formal proposal. The committee does not adopt the figure — that is the Council's role under Article 65 — but it scrutinises the inputs and flags any methodological issue with the specific indicator or with the inflation reading.
In practice, the committee meets in October or November each year, on the basis of the Eurostat interim report. Its discussions are reflected in the explanatory memorandum that accompanies the Commission's proposal. When a Member State has questions about the data series for its national civil service, the committee is the forum where those questions are raised.
The Joint Committee is also a useful early indicator of potential disputes. When the committee's conclusions diverge from the Eurostat numbers, the months that follow tend to involve more intense Council scrutiny of the proposal and, occasionally, recourse to the moderation or exception clauses.
The Moderation Clause
The moderation clause caps the headline annual update when the result of the formula crosses a fixed threshold. Above that threshold, only part of the excess is paid in the current year; the remainder is carried forward to the following annual update. The intent is to smooth large positive shocks across two years rather than concentrate them in one.
The clause works arithmetically: the formula produces a figure, the clause is applied if the figure exceeds the threshold, and the carry-forward amount is added to next year's headline before that year's clause check. The mechanism is symmetric: a carried-forward amount is paid even if the next year's headline is itself low, so the moderation does not strip staff of pay over the cycle, only smooths it.
The political logic of the clause is that very large single-year increases are difficult to integrate into Member State budgets, which fund the EU institutions' staff costs through the annual contribution. By spreading large adjustments across two years, the clause makes the Council's budget conversation easier without breaking the indexation principle.
The Exception Clause
The exception clause allows the Commission to propose a different figure from the one produced by the formula in the case of a sudden and serious deterioration of the EU economic situation. The trigger is defined by reference to specific GDP and unemployment indicators measured at EU level. Both indicators must move past defined thresholds for the clause to be available.
In practice, the exception clause has been the most contested part of Annex XI. The Council invoked an earlier version of the clause in 2011 and 2012 to argue for freezes; the Court of Justice annulled the resulting regulations on the grounds that the trigger conditions were not met. The 2014 staff reform tightened the trigger conditions specifically to avoid that outcome.
Since the 2014 reform, the exception clause has not been used to override a positive headline. The Commission and the Council have, where the underlying conditions appeared close to the trigger, requested additional Eurostat analysis but ultimately adopted the formula's figure unchanged.
Judicial Review and Limits on Council Discretion
The case law on Article 65 is settled in two respects. First, the Council does not have discretion to deviate from the formula on broad policy grounds. The exception clause is the only legal route to a different figure, and its conditions are objective. Second, where the Council adopts a regulation that deviates without proper exception-clause grounds, the Court will annul it and the institutions will pay the missing percentage with interest.
The leading cases are from the early 2010s, in particular Case C-66/12 and the related infringement and annulment proceedings around the 2011 and 2012 updates. They established the principle that Article 65 imposes a procedural obligation on the Council to adopt the result of the formula, save where the exception clause's conditions are met. That principle has shaped every annual update since.
References
- Article 65 and Annex XI of the Staff Regulations — the consolidated version of the legal basis and the implementing methodology.
- Regulation (EU, Euratom) No 1023/2013 — the 2014 staff reform that recast Annex XI.
- Eurostat — the publisher of the specific indicator and the Brussels HICP.
- Court of Justice case law on Article 65 — searchable via the Court of Justice case-law database.