Part I — Situation overview

On 7 September the average price of diesel at Hungarian filling stations rose to 697 forints a litre — 104 forints higher than a year earlier — and that of petrol to 617 forints. Diesel was last this expensive in December 2022. The cause of the rise is multi-layered: the Brent quotation has climbed above 97 dollars a barrel because of the escalating conflict between the United States and Iran, Russian diesel exports are restricted, output at the Russian refineries is faltering, and Mol’s refinery at Százhalombatta has been operating at fifty per cent capacity since the accident last October — full restoration is expected by the end of September. In an interview given to ATV the energy expert Attila Holoda stressed that the problem is not a shortage of crude oil but the tightness of refining capacity in producing diesel, and that this phenomenon is not specifically Hungarian, nor even European.

On the same day two competing policy answers were put on the table. At the sitting of Parliament Prime Minister Péter Magyar announced that the government will decide on Wednesday about targeted support: not at the pumps but through another channel, and not for everyone — according to the announcement, for the owners of lower-powered vehicles used for commuting and for taking children to school, and for the agricultural sector. He rejected restoring the price cap on the ground that within a week diesel would run out at Hungarian filling stations, because once the strategic reserve was emptied nobody would import; and on the cost of the protected price he said that with a difference of a hundred forints a litre it would mean at least 50 billion forints of public money. István Kapitány, the minister for the economy and energy, confirmed this: the compensation will not be implemented at the petrol stations. In parallel the parliamentary groups of Fidesz and the KDNP tabled the bill on restoring the regulated fuel price: with a regulated price of 595 forints a litre for 95-octane petrol and 615 for diesel, with the stipulation that — subject to specified exceptions — only fuel put directly into the tank of a vehicle with Hungarian registration plates would qualify as being at the regulated price. To compensate the losses of small filling stations the proposal would prescribe support of 20 forints a litre on 80 per cent of the quantity sold, and its explanatory memorandum also refers to the increase in agricultural diesel use caused by the drought.

MIAK’s reading: this is the cleanest policy debate of the week, because two instruments differing in structure have been put on the table for the same problem — price setting and a targeted income transfer — and the difference between them is not a matter of world view but measurable. The question is which instrument involves what budgetary outlay, what efficiency loss and what distributional effect. The Hungarian experience of the price cap in 2022–2023 — supply disruptions, retailers withdrawing — is not an opinion but an empirical fact, and the prime minister’s argument was built on it too. At the same time MIAK applies the same yardstick to the other side: targeted support is in principle the better instrument, but this superiority in principle is not in itself a result. A targeted transfer can also be badly targeted, burdened with administrative obstacles, without an expiry date and impossible to measure afterwards — and then it commits exactly the same error in another form. Telex’s data analysis also records a connection that is seldom mentioned: according to the August statistics of the Central Statistical Office (KSH) Hungarian fuel was on average 29–35 forints a litre cheaper than in the neighbouring countries, which over the longer term is in itself a supply risk.

Part II — Foundations in the literature

Behind the choice of instrument stand three classic propositions. The OECD’s Economic Outlook is the most direct: in its analysis of governmental responses to the surge in energy costs the organisation names three conditions — the support should be well targeted, it should preserve the incentive to reduce energy consumption, and it should contain a clear expiry mechanism. It expressly contrasts these with broad-based subsidies, tax cuts and price caps, which can be introduced quickly but are more expensive and weaken the incentive to economise. Nicholas Stern, the British economist who led the British government report on the economics of climate change, shows in the Stern Review the order of magnitude of the distortion: price support for fossil energy was globally an item of some 250 billion dollars a year. This sum is roughly equal to what a two-decade climate investment programme would cost — the support is therefore not only expensive, it also draws the resources away from structural transformation. William Nordhaus, the American economist who founded the cost-benefit modelling of climate policy, formalises the same contradiction in his work Warming the World: the carbon price and fossil price support give price signals in opposite directions, that is, the support is not a neutral intervention but modifies the long-run transition path as well. The detailed treatment of the literature — author by author, with quotations — can be found in section 6.4 Literature in detail.

Part III — MIAK’s concrete proposal

MIAK proposes three measurable measures.

3.1 A public eligibility rule and automatic take-up (simultaneously with the promulgation of the government decision)

The weak point of targeted support is always the same: those who need it most often do not ask for it. The reason for this is not a lack of information but the administrative burden — programme point G26, built on the research of the World Bank, records precisely that material scarcity imposes a cognitive burden, that is, it worsens the decisions taken in complicated administrative situations. MIAK therefore proposes that the public eligibility rule be published simultaneously with the promulgation of the government decision, with three elements: (1) who is eligible — defined exactly by the power of the vehicle, the age of the vehicle, the income position of the owner or the status of agricultural producer, and not on an equitable basis; (2) what the upper limit of the support per eligible person is, expressed monthly or in litres; (3) how the eligible person obtains it. On this last point MIAK’s proposal is concrete: the support should as far as possible start automatically, from the linking of the already existing state registers — the vehicle register, tax authority data, the agricultural support register — without a separate application. Where an application is unavoidable, there should be a single electronic form with pre-filled data. The targeted support logic of programme point SZ1 and the behavioural impact assessment requirement of G26 together provide the frame; the first condition of the OECD recommendation (see 6.4.1) — good targeting — is measured not by the narrowness of the circle of eligibility but by the take-up rate.

3.2 An expiry date and a phase-out path fixed in advance (in the legislation on the support)

The international experience of energy price support is unambiguous: introduction is politically easy, phasing out is politically hard — the expiry must therefore be fixed at the time of introduction, when nobody is yet eligible. MIAK proposes that the legislation on the support contain a specific calendar expiry date, and that it attach to this two automatic review points: one on the 30th day following the return of the Százhalombatta refinery to full capacity, the other at the close of the winter heating season. The outcome of the review may be of three kinds — the support ceases, continues unchanged, or continues at a reduced amount — but in all three cases with public reasoning. The energy price shock preparedness plan of programme point G25 describes the same logic at system level: automatic compensation is tied to a threshold value, and when the level falls below the threshold it switches off by itself. And programme point K7 (energy market shock resilience) warns that the support must not preserve diesel dependence: if the phase-out is not planned in, the support turns into structural expenditure, and it will draw away the resources of the just transition programme in K5. Stern’s order of magnitude (see 6.4.2) is precisely about this relationship of substitution.

3.3 A compulsory subsequent impact assessment broken down by distribution (within 12 months of introduction)

MIAK’s third proposal is the one that makes today’s debate decidable afterwards. Within twelve months of the introduction of the support a public impact assessment should be prepared communicating four data points: (1) how many were eligible and how many actually took it up — that is, what the take-up rate is; (2) what the total budgetary outlay was; (3) how the support was distributed among the income deciles of households; (4) whether an effect on the volume of diesel consumption can be shown. The subsequent impact assessment system of programme point G20 — fixing the expected result in advance, then comparing it with the actual result after 12–18 months — provides a ready methodology for this, and would in any case make it compulsory for measures above 50 billion forints. The value of the assessment is not the rating of the given measure but the grounding of the next decision: once there is comparable data on the distributional effect of the targeted transfer and — from the 2022–2023 period — of the price cap, then at the time of the next energy price shock the debate will not start from zero.

The three proposals are bound together by the same principle: the choice of instrument cannot be decided by theoretical superiority, only by measurement. The eligibility rule makes visible for whom we intend the money; the expiry date guarantees that the crisis measure does not become permanent expenditure; and the impact assessment tells us whether the money reached where it was intended. MIAK’s position is therefore not that targeted support is better than the price cap — but that with these three conditions it is better, and without them the same structural error is repeated in different packaging.

Part IV — Expected effects and risks

Dimension Expected effect Risk
Economy The budgetary exposure of the targeted transfer is orders of magnitude smaller than that of a general price cap, which on the prime minister’s estimate would be at least 50 billion forints at a difference of a hundred forints a litre; the price signal survives, so the incentive to economise does not disappear The administrative burden may reduce the take-up rate; distorting adaptation may develop around the eligibility thresholds (for example the re-registration of vehicles), which worsens the targeting of the support
Security of supply The support is not built into the pump price, so it does not generate a disincentive to import, and it does not attract cross-border refuelling demand The domestic price remains below the regional average, which if it persists may in itself cause supply tension, irrespective of the form of the support
Society and environment The targeted form makes it possible for the support to reach lower-income households dependent on a car, rather than the largest consumers Without an expiry date the support preserves diesel dependence, and draws resources away from energy efficiency and transport transition programmes

The main tension of the package runs between speed and accuracy. Targeted support is good if it is well targeted, but good targeting requires data, a rule and time — while the price pressure is now, not in three months. This is a real dilemma, and it cannot be resolved by denying one of the considerations. MIAK’s proposal is therefore that rapid introduction and an accurate rule should not be alternatives: the support should start as soon as possible, but the eligibility rule and the expiry date should be public simultaneously with promulgation, and the fine-tuning should take place after the 12-month impact assessment. The second question for consideration is that of the small filling stations. According to Attila Holoda’s warning, alongside the present net retail margin of 10–16 forints 600–700 smaller filling stations could find themselves in difficulty. This consideration speaks for the price cap proposal in so far as that at least names the compensation — whereas in the targeted support model the margin of the stations remains untouched, which is a more favourable starting point for the smaller players. The third is the relationship of the competing proposals: the regulated price and the targeted transfer cannot be combined, because one intervenes in the price and the other in income, and applied together the support would pay twice for the same thing.

Part V — Measurability and summary

5.1 What is worth following? (suggested KPIs)

MIAK proposes four performance indicators (KPIs, in full: Key Performance Indicators) from which it will be visible in 6 and 12 months whether the decision has brought a policy result:

  • The take-up rate: what percentage of those eligible actually obtained the support. The suggested target is a rate above 80 per cent; below this the targeting of the support fails not on the construction in principle but on the obstacles of administration.
  • The distribution of the support by income deciles: the suggested target is that at least half the sum should reach households in the bottom five income deciles. This is the indicator that actually distinguishes the targeted transfer from the price cap — with a price cap this ratio is typically reversed, because of the distribution of consumption.
  • The difference between the domestic and the regional fuel price: on the basis of the KSH’s monthly statistics. It is worth following, because a domestic price persistently below that of the neighbouring countries — in August 29 forints a litre for diesel and 35 for petrol — signals a supply risk.
  • The actual expiry of the support: whether the termination date fixed in legislation has been met, or whether it was extended. This is the simplest and the most telling indicator: the international life course of energy price support is typically characterised not by its introduction but by its phasing out.

5.2 Summary

MIAK’s request in a single sentence: Wednesday’s government decision should promulgate not only the fact of the support but also the eligibility rule and the expiry date, and from the day of introduction it should be fixed what has to be measured in 12 months’ time. Concretely: a public eligibility rule with automatic take-up simultaneously with promulgation, a calendar expiry date with two review points, and a compulsory subsequent impact assessment broken down by income deciles.

Two MIAK foundational values are in play here. One is being ideology-free: the choice between the price cap and the targeted transfer has long appeared in Hungary as a question of political identity, although the difference between them is entirely measurable — budgetary outlay, distributional effect, supply consequence. MIAK therefore judges neither proposal by who tabled it: the argument for the regulated price — that it takes effect immediately and for everyone, without administration — is a real advantage, only dearly paid for. The other is data-drivenness: even today there is no public, income-broken-down distributional analysis of the experience of the 2022–2023 price cap, so the present debate is largely conducted from memory. If compulsory measuring back is attached to the support now starting, then at the time of the next energy price shock there will for the first time be Hungarian data to which both sides can refer.

Part VI — Reasoning and further sources

6.1 The framing of the press by spectrum

The economic band carried the policy content, and its framing was that of the technical implementation of the measure. Portfolio followed the announcement in two separate articles, and in both the emphasis was that the compensation will not be implemented at the pumps — that is, the paper highlighted the structural difference that separates the two proposals. The same paper also published the world market background: the rise in the Brent quotation, the restriction of Russian diesel exports and the role of refinery outages. This framing gives the debate the context that is typically missing from the other two bands: that the overwhelming part of the price rise is not the consequence of a Hungarian policy decision.

In the liberal-left band the emphasis fell on setting the two proposals against each other. 444.hu’s report followed the dynamics of the parliamentary exchange and highlighted two elements of the prime minister’s argument: the risk of supply shortage and the 50 billion cost estimate — the framing here was that of the political dispute. HVG, by contrast, put the content of the measure first, and carried the broader energy policy question further in a separate podcast: the expert argument about the long-run harm of freezing household energy prices, which affects a far wider circle than the present debate. Telex ran not on the announcement but on the data: from the KSH’s August statistics it highlighted the connection — one that fits awkwardly into both political narratives — that Hungarian fuel has for months been cheaper than the regional average. This article is the most valuable contribution to the debate, because it brings in a fact on which neither side’s argument was built.

The conservative band carried its own bill on this day, primarily by reporting the fact of its tabling; 24.hu — from the public affairs band — gave the most detailed account of the substantive elements of the proposal, including the compensation of 20 forints a litre for small filling stations and the restriction to Hungarian registration plates. ATV’s interview with Attila Holoda represented the expert band, and brought in two considerations that appeared in neither political framing: refining capacity as the real bottleneck, and the livelihood risk of small filling stations. By MIAK’s yardstick of being ideology-free the argument for the regulated price must also be stated: it takes effect immediately, requires no administration, and leaves out nobody who would be eligible — this is a real advantage, whose price is the supply risk and the higher budgetary outlay.

6.2 Facts and data

Data Value Source
The national average price of diesel 697 HUF/litre (7 September 2026) — 104 HUF higher than a year ago Portfolio, 7 September 2026
The national average price of 95-octane petrol 617 HUF/litre Portfolio, 7 September 2026
Daily price rise on 7 September petrol +8 HUF, diesel +9 HUF compared with Friday’s price 24.hu, HVG, 7 September 2026
The last comparable diesel price level December 2022 24.hu, 7 September 2026
Brent quotation 97.48 USD/barrel (morning of 7 September 2026), +7.8% in the previous week HVG, 7 September 2026
The proposed regulated price 95-octane petrol 595 HUF/l, diesel 615 HUF/l 24.hu, 7 September 2026
The proposed compensation for small filling stations 20 HUF/litre on 80% of the quantity sold 24.hu, 7 September 2026
The estimated cost of the protected price at least 50 billion HUF at a difference of 100 HUF/litre (the prime minister’s statement) 444.hu, 7 September 2026
The capacity of the Százhalombatta refinery 50% since the accident of October 2025; full capacity expected from the end of September HVG, 444.hu, 7 September 2026
August domestic average prices (KSH) diesel 672 HUF, petrol 589 HUF Telex, 7 September 2026
Difference from the average of the nine countries examined diesel −13 HUF, petrol −23 HUF Telex, 7 September 2026
Difference from the neighbouring countries diesel −29 HUF, petrol −35 HUF Telex, 7 September 2026
Retail margin net 10–16 HUF/litre; 600–700 smaller filling stations could find themselves in difficulty ATV, 8 September 2026
The lag of the inflationary pass-through typically a month and a half to two months ATV, 8 September 2026

One item in the table requires a separate note: the cost estimate of 50 billion forints is a figure stated in the prime minister’s parliamentary speech, which assumes a price difference of a hundred forints a litre — it is not an independent analyst’s calculation, and it does not include the cost of compensating the small filling stations.

6.3 Policy dimensions

  • Economy (programme points) — the frame of instrument choice and of measuring back: G25 (energy price shock preparedness plan) provides the model of threshold-based compensation, G20 (subsequent impact assessment) the obligation of measurement at 12 months, and G26 (behavioural public policy design) the methodology of reducing the burden of take-up;
  • Social policy (programme points) — the targeting of the transfer: the provisions of programme point SZ1 on eligibility thresholds and administrative burden are directly applicable to the rule now to be shaped;
  • Environment and climate (programme points) — the relationship of the price signal and the transition: K7 (energy market shock resilience) and K5 (just transition) provide the frame in which the support does not preserve but bridges diesel dependence;
  • Transport and infrastructure (background material) — from the point of view of the electric mobility plan of KO3, the duration and the phase-out path of the support determine how far it distorts household decisions on the vehicle fleet.

6.4 Literature in detail

6.4.1 OECD: Economic Outlook

The OECD evaluates governmental responses to the surge in energy costs from the point of view of budgetary sustainability and of incentives together, and formulates a recommendation that fits today’s Hungarian debate almost item by item:

“Any new discretionary measures should be well targeted on the most needy households and viable businesses, preserve incentives to reduce energy consumption, and contain a clear expiry mechanism. […] Broad-based subsidies and transfers, tax cuts and price caps can be introduced quickly, but entail a higher budgetary cost and weaken the incentives to reduce energy use.”

The report adds that where fiscal room for manoeuvre is narrow, such broad-based measures are more likely to force expenditure cuts in other areas.

For the Hungarian decision this gives three clear yardsticks. The targeted support now announced structurally meets the first condition — targeting — and the second as well — the preservation of the incentive — since the pump price does not change, so economising continues to pay. The third condition, the expiry mechanism, is the one about which the announcement has so far said nothing. The second element of MIAK’s package of proposals asks precisely for this missing third condition — not because of any special suspicion about the Hungarian measure, but because on the OECD’s recommendation these three conditions work only together.

📖 Source: OECD: Economic Outlook

6.4.2 Nicholas Stern: The Economics of Climate Change — The Stern Review

Stern treats price support for fossil energy not primarily as a climate question but as a question of economic efficiency. The report records that many countries have long subsidised particular energy carriers — coal, oil, nuclear energy, rural electricity supply — and that, with the exception of research and development support, these are sources of economic distortion and loss, and that historically the distortion was strongly in favour of the more polluting fuels. On the order of magnitude the report writes:

“The liberalisation of energy markets […] was seen as a means of reducing these subsidies, which in some cases had reached extraordinary levels. By 1998 they had fallen worldwide, but still amounted to almost 250 billion dollars a year. […] These transfers are of roughly the same order of magnitude as the average additional cost of the investment programme required for the world to move onto a meaningful climate policy path over the next twenty years.”

Stern adds: one of the indirect benefits of climate policy is precisely that it provides an occasion to eliminate the existing distortions of the energy market.

For the Hungarian situation this parallel yields the most important lesson. Of the two instruments now in dispute it is the price cap that falls into the category Stern describes: general fossil price support built into the price, which gives the most support to the largest consumer. The targeted income transfer is structurally different — it does not distort the price but supplements income — and therefore does not fall under Stern’s criticism, provided that it does not become permanent. If, however, it survives without expiry, the difference is blurred: a lasting transfer draws resources away from transition investments in just the way Stern describes for price support.

📖 Source: Nicholas Stern: The Economics of Climate Change — The Stern Review

6.4.3 William Nordhaus: Warming the World

Nordhaus is the founder of the cost-benefit modelling of climate policy; Warming the World presents the results of the DICE and RICE family of models, in which alternative policy scenarios are compared uniformly through the price burden placed on carbon dioxide emissions — in the language of the model, the carbon tax. One of the central results of the model family is that the efficiency of the various climate policy paths depends essentially on what price signal the user of fossil energy receives: the same emission reduction target can be achieved at economic costs differing by orders of magnitude, depending on whether the price signal is uniform and predictable, or differs by sector and by period.

From this follows the consideration that is not voiced in today’s Hungarian debate. Price support for fossil energy and the carbon price act in the same decision situation in opposite directions: one makes the same litre of diesel cheaper, the other more expensive. Whoever applies both — and Hungary, as a member state of the Union, is a participant in the European emissions trading system — is working against their own policy instruments, and the net price signal becomes unpredictable. In Nordhaus’s frame this is not a moral but an efficiency question: an investment decision taken alongside an uncertain price signal is worse than one taken alongside either a high or a low, but predictable, price.

The practical conclusion for MIAK’s proposal: the duration and the phase-out path of the support are not an administrative detail. A bridging transfer announced in advance for six months does not influence a household’s decision on changing its vehicle, because it does not alter expectations about long-run running costs. Support without an expiry, by contrast, does — and it steers precisely the lower-income households dependent on diesel towards a vehicle fleet whose running costs jump suddenly when the support ceases.

📖 Source: William Nordhaus: Warming the World

6.5 International comparison

The European responses to the energy price shocks between 2021 and 2023 tried both instruments, and the lessons are relatively uniform. The general interventions built into the price — the German fuel tax cut, the French pump discount, the Spanish support per litre — took effect quickly, but had three common features: the budgetary outlay far exceeded the prior estimate, a significant part of the effect appeared in the retail margin rather than in the consumer price, and phasing out proved harder in every case than planned. The targeted solutions — the income-linked French fuel voucher, or the Italian household energy bonus of similar logic — were cheaper, but consistently produced a low take-up rate where they required a separate application, and a high rate where they ran automatically from the existing tax or social registers.

This last difference is the most important Hungarian lesson. The take-up rate depends not on the strictness of the eligibility rule but on the form of the administration — and support that half of those in need do not take up is not targeted, merely narrow. The first element of MIAK’s proposal, automatic take-up built on the existing registers, is therefore not a matter of convenience: this is the decision that determines whether the superiority in principle of the targeted construction also appears in practice.

On the side of the price cap the Hungarian experience of 2022–2023 was extreme even in international comparison: the supply disruptions that developed alongside the regulated price and the withdrawal of retailers were sharper than elsewhere because the measure did not affect the wholesale price, so the retail stage bore the whole loss. The proposal now tabled processes this experience in so far as it builds in compensation per litre for small filling stations — this is a substantive difference compared with the 2022 construction, and in the debate on the proposal it deserves an objective evaluation.

Economy

  • G20 — Economic policy impact assessment system
  • G25 — Energy price shock preparedness plan
  • G26 — Behavioural economics in public policy design

Social policy

  • SZ1 — Targeted support

Environment and climate

  • K5 — Just transition programme
  • K7 — Energy market shock resilience

Transport and infrastructure

  • KO3 — Electric mobility plan

Suggested new programme point: A phase-out clause for energy price support — for the Economy area: the legislation on every support measure targeting a price level or an energy cost must contain a compulsory calendar expiry date and at least one automatic review point.

6.7 List of sources

Press sources (MIAK press monitor, 8 September 2026 — topic 3):

Knowledge base references (specialist books):

  • 📖 OECD: Economic Outlook
  • 📖 Nicholas Stern: The Economics of Climate Change — The Stern Review
  • 📖 William Nordhaus: Warming the World

MIAK internal materials:

  • MIAK policy area: Economy (programme points; programme point ID: G20, G25, G26)
  • MIAK policy area: Social policy (programme points; programme point ID: SZ1)
  • MIAK policy area: Environment and climate (programme points; programme point ID: K5, K7)
  • MIAK policy area: Transport and infrastructure (background material; programme point ID: KO3)
  • MIAK press monitor, 8 September 2026 — topic 3, score: 92/100

Supplementary public data sources:

  • KSH fuel price statistics (monthly, with regional comparison)
  • European Commission Weekly Oil Bulletin
  • Data of the Hungarian Energy and Public Utility Regulatory Authority
  • International Energy Agency (IEA) Oil Market Report

Generation metadata