Negotiations on the EU’s Multi-annual Financial Framework (MFF) take place within the General Affairs Council of the European Union. Ultimately, the final decisions are taken by the European Council of Heads of State and Government and must also receive the consent of the European Parliament. There are many points of friction, including the overall size of the MFF, its allocation across headings, and the question of own resources and how to finance the budget. Traditionally, there is a specific focus on the pre-allocated amounts as these are a significant determinant of the net balance between receipts and contributions for individual Member States – often referred to as the focus on juste retour in the MFF negotiations.
In the Commission’s proposal for the next MFF, the pre-allocated funds have all been merged into a single European Fund for economic, social and territorial cohesion, agriculture and rural, fisheries and maritime, prosperity and security. This allocates funding for the National and Regional Partnership (NRP) Plans, the EU Facility and the Interreg Plan (Article 10, Budget). The NRP Plan allocation, in turn, is made up of the General Allocation, the Home Affairs allocation covering migration, security and home affairs, and the Social Climate Fund allocation. The General Allocation covers all the priorities previously covered by cohesion, agriculture, fisheries and just transition funds, but it now also includes support to security and defence capabilities.
The Commission has published a Fact Sheet setting out the Member State allocations under the three headings. The formulas it has used to allocate the General Allocation and the Home Affairs allocation are set out in Annex 1 to the European Fund Regulation (although not explicitly confirmed in the Fact Sheet, I assume that the amounts allocated to each Member State under the Social Climate Fund are based on the formula included in Annex 1 to the Social Climate Fund Regulation (EU) 2023/955 even though that formula applies to the period 2026-2032 and the NRP Fund applies to the period 2028-2034).
In this post, I comment on the allocation formula for the General Allocation share of the NRP Plans which replaces the cohesion, agriculture, fisheries and just transition funds. To date, there has been limited interrogation of this formula despite its significance for individual Member States (one exception concentrating on its implications for cohesion policy is this policy brief published by the Conference of Peripheral Maritime Regions). There are several interesting features including some problematic ones that are worth commenting on.
Roles of the Council and Parliament
The new single fund model appears to give the Parliament at least in theory a greater say in how the MFF resources will be allocated. In previous MFFs with a larger number of MFF headings, the Council (and ultimately the European Council) decided not only on the overall size of the MFF but also on its allocation between headings. These included specifying the amounts to be available for cohesion policy, for agricultural policy and for other headings. Now in the Commission’s MFF proposal and specifically the Annex to that proposal, there are only four headings for the Council to decide, with a single heading covering the European Fund. The Council will almost certainly make changes to the Commission proposal, and these will most likely be in a downward direction.
However, how this Fund will be allocated is set out in the European Fund Regulation, not the MFF Regulation. The European Fund Regulation will be decided by co-decision and thus any budgetary allocations, or allocation formulas, in this Regulation must be agreed by both the Council and Parliament. Thus, for example, either the Council or Parliament could seek to increase the minimum ring-fenced amount for CAP income support and fisheries. They can also seek to amend the allocation formula in Annex 1 which distributes the NRP Plan funding to Member States.
For the Parliament, this power may be more theoretical than real. This is because the European Council in its MFF conclusions (which must be reached unanimously) could well decide to insert specific numbers for sub-headings and text which should actually be determined by co-decision. This was done most extensively in the 2013 European Council MFF conclusions affecting the 2014-2020 CAP reform (see Matthews, 2014 for full details) but certain elements dealt with in the CAP regulation were also carried forward into the 2020 European Council conclusions for the 2021-2027 MFF (see Martinos, Matthews et al., 2022 pp. 26-27). In the past, when CAP legislation was being negotiated in trilogues, the Council representatives tended to argue that, where numbers or text had been agreed in the European Council conclusions, these numbers and text were non-negotiable. We will have to wait and see how these issues are resolved in the coming negotiations.
The allocation formula for the General Allocation amount in the NRP Plan Fund
Here I quote the exact text from Annex 1 of the draft European Fund Regulation. It is a complicated beast, and I break it down into its component parts in subsequent sections. In this post, we are only interested in the amounts distributed according to the allocation key Ai in the following formula.
The financial contribution available for each Member State (FCi) is the consolidated amount for the implementation of the plan established as follows:
FCi =
Ai × amount available for Member States NRPs excluding amounts set out in Article 4 of Regulations [Migration], Article 4 of Regulation [Borders], Article 4 of Regulation [Internal Security] and Regulation (EU) 2023/955+
Bi × amounts set out in Article 4 of Regulations [Migration], Article 4 of Regulation [Borders], Article 4 of Regulation [Internal Security]+
Ci × amount available for the Social Climate Fund in accordance with Article 10(3) of this Regulation.

where for each Member State i and NUTS level 3 region r
– Pop is the population at 1 January 2024 (Eurostat online database code: demo_gind, tps00001);
– AROPE ra is the population at risk of poverty or social exclusion in rural areas in 2024 (Eurostat online database code: https://ec.europa.eu/eurostat/databrowser/view/ilc_peps13n/default/table?lang=enilc_peps13n, 2024);
– GNI pc PPS is gross national income (GNI) per capita measured in purchasing power standards (Eurostat online database code: nama_10_pp, 2023);
– GDP pc PPS r is the regional gross domestic product (GDP) per capita measured in purchasing power standards (Eurostat online database code: nama_10r_3gdp, average 2021-23);
– DPi is the amount of direct payments estimated in budget year 2027 (excluding POSEI/SAI);
– ha is hectares declared as eligible for support under the (“potential eligible area”; claim year 2022).
The ai of all Member States is normalised to ensure that the sum of all ai equals 100%.
To avoid an excessive concentration of resources, a cap and safety net apply for the general allocation key Ai:
– For all Member States, the allocation share ai cannot be lower than 80% and not be higher than 105% of its allocation share in the 2021-2027 total of all relevant pre-allocated funds under shared management, as calculated by the Commission based on the initial 2020 allocation of pre-allocated funds before transfers. The ai of all Member States is proportionally adjusted to ensure that the sum of all ai equals 100%.
(Note: the ai terms in the last three paragraphs should be written with the Greek letter alpha rather than ‘a’ but I currently have a problem in rendering the Greek letter).
General structure of the allocation formula
There are six elements to the general allocation formula, as follows:
- A population factor. Both the share of a country’s population in the total EU population and the share of its population in rural areas at risk of poverty or social exclusion are weighted equally and the average of these shares is the first element in the formula.
- Adjustment factor squared. This population share variable is then multiplied by an adjustment factor in square brackets which is squared (raised to the power of two). This squaring gives greater weight to the extremes for the variables included in the square brackets.
- Relative prosperity. This is the first variable in the adjustment factor, defined as the ratio of average EU GDP per head in purchasing power standard terms to the average GDP per head in the specific country. For example, if the country has a GDP per capita in purchasing power standard that is 66% of the EU average, this variable would take the value of 1.5 and contributes positively to the adjustment factor.
- Regional prosperity gap. This variable gives greater weight to a country that has a larger share of its population in NUTS3 regions with a GDP per capita in purchasing power standard below 75% of the EU average. Suppose we have a country with two NUTS3 regions with equal population, one of which has a GDP per capita in purchasing power parity equal to 90% of the EU average and the other with a GDP per capita equal to 50% of the EU average. The regional prosperity gap formula resolves to 0 for the first region (as 75% – 90% is a negative value) and to 25% for the second region. Multiplying by the shares in the total population gives a regional prosperity gap for this country of (25% x 50%) or 12.5% (0.125) and contributes positively to the adjustment factor by this amount.
- Agricultural prosperity gap. This variable gives weight to the demand for external convergence in the allocation of CAP direct payments. Those countries whose average value of direct payments per hectare in 2027, the last year of the current CAP, are below 90% of the EU average will get an upward boost in their allocation share equivalent to the percentage difference between these two values. For example, the direct payments in the Baltic countries in 2027 will be about 15% below 90% of the EU average. Therefore, their adjustment factor will be increased by 0.15 in calculating their allocation share.
- Safety net, capping and normalisation. To avoid any excessive swing in the amounts received by a Member State in the next MFF compared to the current one, a cap and safety net apply to the allocation key Ai, in that it cannot be lower than 80% nor higher than 105% of its allocation share in the 2021-2027 total of all relevant pre-allocated funds under shared management. The text specifically notes that the comparison is made with the 2021-2027 total “based on the initial 2020 allocation of pre-allocated funds before transfers”. I take this to mean top-ups to these funds from the European Recovery Instrument and Next Generation EU are not included in this comparison.
The fact that these percentages are not symmetrical around 100% reflects the fact that the overall total allocated to the NRP Plans is less than what was allocated to the relevant pre-allocated funds in the current MFF (for the figures, see my previous post on this topic). I find it confusing that the text refers to both the allocation keys and allocation shares but what is meant by allocation share is never defined. Also, the text uses different symbols to refer to these shares (Ai to refer to the allocation key, and ai to refer to the allocation share). In my interpretation, I have assumed that both shares refer to the same overall budget share, or allocation key, which determines the money a Member State gets from the General Allocation part of the Fund, but I stand to be corrected on that.
Finally, normalisation of the allocation key shares is undertaken twice. Normalisation is necessary because, when the individual Ai’s for Member States are calculated according to the formula for the General Key, they will add up to more than 100%. So the first normalisation is a proportional adjustment of all the individual shares to ensure they add up to 100%.
Following this normalisation, the safety net and capping limits kick in and it is possible the allocation key shares for some Member States are adjusted as a result. This means that, once again, the sum of the shares no longer sums to 100%, so a second normalisation is undertaken to make sure this is the case.
The General Key formula inherits several characteristics from the formulas used to distribute resources under the cohesion funds as set out in Annex XXVI to the Common Provisions Regulation (EU) 2021/1060. But it is considerably simplified with fewer variables and these play a different role in the formula. For a thorough analysis of the CPR formula, see the European Court of Auditors rapid case review of the Commission’s initial proposal for the allocation of cohesion policy funding to Member States for 2021-2027. It distinguished between main factors, adjusting coefficients, and additional premiums (variables that gave extra brownie points) in the overall formula. We identify a similar categorisation in our evaluation of the formula in the next section.
Some comments on the allocation formula
Relative prosperity is the key variable differentiating Member State shares. If we start by assuming that the fund could be distributed on the basis of equal per capita amounts, then differentiation is achieved through three strands (i) modulating the population shares in the main factor by averaging these with the share of people in rural areas at risk of poverty or risk of social exclusion; (ii) modulating the population shares by the relative prosperity gap by multiplying by the square of the percentage gap between an country’s GDI per capita and the EU average (reducing the share of wealthier countries and increasing the share of poorer countries); (iii) modulating the relative prosperity gap in turn by additional premia for a regional prosperity gap and an agricultural prosperity gap. In practice, the additional premia are small relative to the contribution made by the relative prosperity gap.
The AROPE variable only takes account of rural areas. The AROPE variable, measuring the share of the population at risk of poverty or social exclusion, has previously been used in the allocation formula for the European Social Fund+. However, in the General Key formula it is confined to the AROPE share in rural areas. The CMPR paper notes that this appears to neglect poverty and social exclusion risks in urban and intermediate areas. Once the money is allocated, of course, Member States are free to prioritise interventions in these areas if they wish. But one can understand that Member States that lose out because of the focus on AROPE in rural areas only may well press for using the more comprehensive variable.
The agricultural prosperity variable does not make sense. The two premia variables, the regional and agricultural prosperity gaps, are given equal weight, reflecting the roughly equal shares of cohesion policy and agricultural policy in the shared management funds. The regional prosperity gap gives bonus points to Member States with significant shares of their population in lagging regions (defined as regions below 75% of the average EU GDP per head). This at least can be taken as an objective indicator of need, even if the precise measurement formula can be debated. The same cannot be said of the agricultural prosperity gap. It is based on the external convergence formula for direct payments in the current CAP. But any inequity in the distribution of direct payments in the current CAP has no bearing when constructing a completely new allocation key for the next MFF, unless the Commission has the intention that the new allocation key should also atone for sins of the past! The formula used for the agricultural prosperity gap is also not an indicator of need. It reflects the circumstances in which certain Member States acceded to the Union in the past. While I do not underestimate the difficulty of defining an indicator to measure the need for agricultural support across countries, this particular definition of the agricultural prosperity gap is not a measure of need.
The absence of non-economic variables in the formula. With separate allocation keys for individual funds as under the current MFF, it is possible to play with the allocation formulas to reflect the specificities in the objectives of the different funds. For example, criteria for the allocation of European Social Fund+ funds can take account of the number of unemployed, youth unemployment rates, rates of early school leaving, share of persons with tertiary eduction, and so on. For a purely agricultural fund, one could envisage allocation criteria such as the share of land under nature protection, the share of land under organic farming, or other specific criteria. For the just transition fund, one could allocate funds based on the risk of losing jobs as a result of decarbonisation. When all these funds are put together, it is no longer possible to take these specificities into account unless we want to end up with a totally incomprehensible formula.
We should be clear that the variables that enter the allocation formula do not in any way constrain how Member States can use the funds they are allocated. These uses are established in the specific priorities laid down in the European Fund Regulation. The criteria used in the allocation key do not determine the uses to which the funds can be put.
The allocation key does not reward performance. In the run-up to the publication of its MFF proposal, the Commission talked a lot about the need to make the budget more performance-oriented (Commission, 2025). However, it tends to use performance in the sense of aligning EU policy objectives and EU spending. In this context, I use performance to mean rewarding results (as in payment by results agri-environment schemes). If anything, the allocation formula generates perverse incentives. For example, if a Member State with lagging regions uses its funding to invest heavily in those regions, with the successful result that their average GDP per capita is lifted above the 75% threshold mark, this Member State would lose money in the next MFF. This is the opposite of a results-based performance-based budget.
Here there is scope to think outside the box. Why not design an allocation formula that rewards those Member States that make the greatest strides towards achieving the EU priorities? Why not give more money to those countries that succeed in reducing GHG emissions the fastest, that succeed the most in reversing the decline in biodiversity, that achieve the fastest degree of regional convergence? The budget would then become a real incentive mechanism, and not just an instrument of juste retour.
Conclusions
In this post we have explored the formula used by the Commission to allocate resources to Member States from the General Allocation heading for the fund for the National and Regional Partnership Plans (thus leaving to one side the formula for the Home Affairs heading and the pre-existing formula for the Social Climate Fund which are the other headings in the Plans). One thing to make clear is that the formula should not be taken too literally. As a CPMR paper describing the ‘Berlin formula’ used heretofore to distribute cohesion funding states, “the allocation methodology’s main purpose is to justify politically acceptable outcomes in terms of the distribution of funds”. It is often the start of the bargaining process around money rather than the end. For example, the budget for cohesion funding has usually been ‘topped up’ by special gifts to individual countries outside of the general methodology. In the 2014-2020 cohesion budget, these side payments totalled €9.9 billion. Thus, whatever the shape of the final formula used, we should also expect to see the continuation of these side payments in order to reach a unanimous agreement around the Council.
What I have not done in this post is to simulate the formula with the data sources provided to see if I arrive at the same outcome as the Commission. What would interest me is how big is the role played by the capping and safety net limits in constraining the allocation shares for individual Member States. Would the shares be very different without these limits? Or does the formula already closely replicate the Member State shares in the pre-allocated amounts in the current MFF?
Instead, with this post I have tried to understand the logic of the allocation formula. Apart from population size, relative prosperity is the variable with the greatest impact on the formula. Additional premia can be earned based on the regional and agricultural prosperity gaps, but the definition of the agricultural prosperity gap is not a specific indicator of need and hardly makes economic sense. There is no role for non-economic variables, and there is no performance result-based incentive built into the formula. The formula, along with other monetary amounts in the European Fund Regulation, must be decided by co-decision thus giving both the Council and the Parliament a voice in the final outcome. It will be interesting to see if the European Council, which usually has the biggest say on financial matters, will attempt to lay down markers in its MFF conclusions which might constrain the Parliament’s role in the trilogue negotiations.
This post was written by Alan Matthews.
O artigo foi publicado originalmente em CAP Reform.