Part I — Situation overview

This year’s harvest season has closed, and with it the drought has crossed over from the category where it was a forecast into the one where it is measured data. According to Portfolio’s report of 21 August 2026, because of the extraordinary dryness roughly four million tonnes of wheat were produced, a quarter less than usual; the heat and the lack of rainfall forced the plant into premature ripening, so farmers finished the harvest as early as mid-July, two weeks earlier than usual. János Molnár, state secretary for agricultural economy, told RTL Híradó that the crop is amply sufficient for domestic needs for the coming year, and that since there is enough wheat on the world market no further rise in the price of bread and bakery products need be feared. The same report also records the longer-term price path: the price of bakery products has risen one and a half times over the past five years, and a kilogram of half-brown bread recently cost 856 forints on average — twice as fast a rise as the pace at which earnings grew over the same period. Consumer adaptation is measurable too: bread consumption has fallen by 15–20 per cent, while consumers buy roughly the same amount more of pastries instead.

The second thread concerns surface waters. Also on 21 August, Portfolio reported that the water level of Lake Velence has sunk to a historic low: on 16 August the gauge at Agárd measured a mere twenty centimetres, well below the previous lowest value of fifty-three centimetres registered during the 2022 drought. The phenomenon was documented by a Sentinel-2 satellite image from the Copernicus programme, which the European Union shared as its image of the day — the image shows not only the contraction of the lake but also the drying out of the surrounding areas. According to data from the European Environment Agency, 2025 was the fourth driest year since 2000, soil moisture fell across nearly a quarter of the Union’s territory, and within the region Hungary, along with Bulgaria, Cyprus, Greece and Romania, is the most severely affected. The phenomenon is therefore not a Hungarian peculiarity but a central and south-eastern European pattern. In his assessment given to ATV, Zsombor Boromisza, associate professor at the landscape protection department of the Hungarian University of Agriculture and Life Sciences, drew attention to the fact that Lake Velence, with an average depth of barely one and a half metres, is by its nature not a lake of constant water level, and that the present situation therefore signals not only a water shortage but also the problems of the model of use and development built around the lake.

The third thread is intervention. Following the decision of Szabolcs Bóna, minister for agriculture, the hail-suppression system JÉGER has been shut down across most of the country, with four counties excepted. According to Mandiner’s report of 21 August, the decision was preceded by an opinion poll and a farmers’ petition, behind which lies the assumption that intervention with silver iodide reduces the quantity of rainfall; the National Chamber of Agriculture, by contrast, stated that there is no connection between the operation of JÉGER and the drought. Ákos Hadházy, a former member of Parliament, criticised the decision on two grounds: on the one hand he fears that hailstorms will become more frequent, particularly in vineyards, and on the other he objects that the government took a popular decision in response to a scientifically unproven claim. From the nature of the matter, both sides assert something that could be measured — only the measurement is missing.

MIAK’s reading of these three threads runs out into a single proposition: in Hungary drought damage has ceased to be an extraordinary event, yet its handling still works according to the logic of extraordinary events. An extraordinary event receives money after the fact, by an individual decision, under changing rules; a recurring, modellable loss, by contrast, receives a line in the budget in advance and a payment rule fixed in advance. This difference is not a fine point of accounting technique: it determines whether the farmer can plan ahead, whether damage-preventing investment is worth making, and whether the taxpayer’s money is distributed by a predictable rule or by individual discretion. The present state of Lake Velence shows the same thing on the surface-water side: if the water level of a lake sets a record twice in two drought cycles, then the model of development and use around the lake has to be calibrated not to exceptional years but to the new average.

Part II — Foundations in the literature

The conceptual framework of the present situation comes together in three authors. In his work The Economics of Climate Change — The Stern Review, Nicholas Stern (British economist, former chief economist of the World Bank, author of the 2007 government review on the economics of climate change) starts from the premise that emission reduction no longer provides an answer to the effects occurring in the coming decades — adaptation alone does — and that adaptation is stimulated above all by those market and institutional signals which reveal the actual size of the risk; for him risk-proportionate insurance is not a financial product but an information instrument. William D. Nordhaus (American economist, awarded the Nobel Memorial Prize in Economics in 2018 for the macroeconomic modelling of climate change) and Joseph Boyer, in their volume Warming the World, introduce the concept of the damage function: in the model climate damage is not a random event but a continuous series of losses dependent on temperature, which has to be built into the production function — this is the direct theoretical basis of the domestic integrated climate-economy model, programme point K9. And in Governing the Commons, Elinor Ostrom (American political scientist, awarded the Nobel Memorial Prize in Economics in 2009 for her research into the governance of commonly used resources) shows that finite resources used in common — water being a classic case — can be managed sustainably if their use is accompanied by continuous, mutual monitoring and by graduated sanctions; for her, measurement is not an administrative burden but the condition of cooperation. The detailed treatment of the literature — author by author, with quotations — can be found in section 6.4 Literature in detail.

📖 Source: Nicholas Stern: The Economics of Climate Change — The Stern Review; William D. Nordhaus & Joseph Boyer: Warming the World — Economic Models of Global Warming; Elinor Ostrom: Governing the Commons — The Evolution of Institutions for Collective Action

Part III — MIAK’s concrete proposal

MIAK proposes three measurable measures. All three start from the fact that after the close of this year’s season we are working not with estimates but with actual data.

3.1 A named, pre-estimated drought-damage line in the 2027 budget (by submission)

MIAK asks that the 2027 budget bill contain a separate, named appropriation for the handling of drought and dryness damage, and that this amount be not the average of the ex post payments of earlier years but a modelled estimate. The input of the model is the domestic integrated climate-economy framework (K9): crop-loss bands assigned to temperature and rainfall scenarios, and damage values assigned to those. The essence of the proposal is that the budget should recognise the damage in advance, in bands — with amounts corresponding to a lower, a middle and an upper scenario — and that the explanatory memorandum of the bill should show which scenario it reckoned with. This is the joint application of the G1 data-driven budget and the A1 public-money transparency programme points to one concrete expenditure item. The practical yield is twofold. On the one hand the farmer knows, before the sowing decision, what degree of state risk-sharing to expect, and can therefore align their own risk management — irrigation development, choice of variety, insurance — accordingly. On the other hand the annually recurring situation ends in which the amount and the conditions of damage compensation take shape after the damage has occurred, by an individual government decision. Discretionary payment — that is, payment based on free deliberation without criteria fixed in advance — is more expensive and harder to hold to account than a rule fixed in advance.

3.2 A public agricultural damage data platform, broken down by municipality (by the start of the 2027 support year)

The second proposal concerns the data level of damage reporting and damage payment. MIAK asks that the whole process of damage reporting take place through the agricultural data platform under programme point MG2, and that the platform make public an anonymous aggregation broken down by municipality: how many hectares the report concerned, how large the established damage was, and how much was paid out. The protection of personal and business data is not an obstacle here, because an aggregation at municipality level no longer identifies a farmer. Publicity solves three things at once. First: it makes the territorial distribution of the damage visible, which is information in its own right from the point of view of the TE1 micro-regional development index — drought does not strike the country evenly, and the micro-regions with the worst soil conditions bear both the greatest damage and the smallest reserves. Second: it makes the reported damage comparable with the amount paid out, which is the simplest control indicator of how the damage-compensation system works. Third: it supplies data for the model under point 3.1 — without a decade’s worth of municipal damage aggregates the modelled appropriation remains an estimate, whereas with it the model can be calibrated. The same interface can receive the soil-moisture and yield data of the precision farming programme (MG1), so there is no need to build a new system: an existing line of development merely has to be supplemented with a public output.

3.3 An ex post impact assessment protocol for the JÉGER decision — the county boundary as a measurement situation (by the start of the 2027 hail-damage season)

The third proposal starts from the situation which the present decision has unintentionally created. By keeping JÉGER in operation in four counties while stopping it in the rest of the country, a comparison situation has arisen: two adjacent groups of areas with similar climatic conditions, in one of which there is intervention and in the other of which there is not. MIAK asks that, instead of a battle of experts, the government use this situation for measurement, and that before the start of the 2027 hail-damage season it publish the protocol of the ex post impact assessment: which indicators it collects (the number of reported hail damage events, the area affected, the damage value per hectare, broken down by crop), which comparison period it uses, and who carries out the evaluation. The protocol has to be published before the season, because with indicators chosen after the fact any position can be substantiated. The examination follows the methodological logic of the K1 measurable climate targets programme point, and has to answer the question which today both sides assert and neither proves: is there a measurable difference in the extent of hail damage between the areas with and without intervention, and can any difference in rainfall be demonstrated? If the measurement vindicates the shutdown, that strengthens the decision; if it refutes it, then a reversible decision has to be reversed before the damage becomes lasting.

The three proposals are held together by a single principle: recurring damage has to be measured and accounted for, not handled case by case. Ostrom’s proposition on monitoring (see 6.4.3) and Nordhaus’s damage function (see 6.4.2) lead to the same conclusion from two different sides: if the loss is regular, then the institutional response to it also has to be regular, known in advance and verifiable. In the matter of Lake Velence this also means that the model of development and use around the lake has to be adjusted to the natural fluctuation identified by the expert position, not to a water level assumed to be constant — this is the direct application to one concrete lake of the common resource management framework under G28 and MG5.

Part IV — Expected effects and risks

Dimension Expected effect Risk
Budget Drought damage becomes a predictable, pre-planned expenditure item, and the payment rule is not created anew every year A named appropriation creates a level of expectation: if the actual damage is smaller than estimated, pressure may arise to use up the envelope
Agriculture Before the sowing decision the farmer knows the extent of state risk-sharing, and can therefore align their own preventive investment with it If state risk-sharing is too generous and not risk-proportionate, it may crowd out precisely that preventive investment
Data publicity The damage aggregation broken down by municipality makes the working of damage compensation verifiable and calibrates the model Data reporting imposes an administrative burden on smaller farms if it is not built into the existing reporting process
Territorial cohesion It becomes visible that the damage hits hardest the micro-regions with the smallest reserves, which makes more targeted support possible The public damage picture may in the short term worsen the land values and creditworthiness of the micro-regions concerned
Damage prevention The JÉGER decision becomes measurable and, on the basis of the measurement, reversible If the measurement protocol is drawn up after the season, the choice of indicators itself becomes a political question

The main question for deliberation is the degree of risk-sharing. The larger and the more unconditional the ex post state compensation, the smaller the farmer’s incentive to invest in irrigation, soil protection, a drought-tolerant variety or insurance — this is the crowding out of damage prevention. At the same time zero state involvement does not work either: a 25 per cent crop loss takes away the capital of a significant proportion of smaller farms and endangers the following year’s sowing. In MIAK’s view the solution is not to get the degree right but to change the structure: the state contribution should be known in advance, risk-proportionate and tied to preventive investment — that is, whoever has demonstrably acted to reduce the risk should receive compensation at a higher rate. The proposal tips over to the risk side if the named budget line comes into being but the associated data platform and payment rule do not: in that case the system achieves no more than setting aside in advance the payments hitherto made case by case, while the freedom of discretion remains. There is a measure of ideology-free conduct here too: the measurement protocol now proposed makes the present government’s own JÉGER decision verifiable, and that is what gives it its credibility.

Part V — Measurability and summary

5.1 What is worth following? (proposed KPIs)

The performance indicators below (KPIs, Key Performance Indicators) will show in 12 and 24 months whether the damage measurement and damage handling system has genuinely been transformed. These are proposed benchmarks, not government commitments.

  • Named appropriation: whether the 2027 budget bill contains a separate line for drought and dryness damage, and whether the explanatory memorandum names the scenario of the estimate.
  • The divergence of forecast and fact: how large the difference is in 2027 between the damage compensation budgeted for and the amount actually paid out — this is the simplest indicator of the quality of the model.
  • Data publicity: whether an anonymous damage aggregation broken down by municipality is available by mid-2027, and whether it contains both the reported and the paid-out amount.
  • The protocol of the JÉGER measurement: whether the impact assessment protocol appears before the start of the 2027 hail-damage season, and whether it contains a comparison of the counties with and without intervention.
  • Prevention ratio: how, alongside the amount spent on damage compensation, the share of support devoted to prevention — irrigation, soil protection, change of variety — develops; it is worth watching whether the ratio moves towards prevention.

5.2 Summary

MIAK’s request to the decision-maker can be summed up in a single sentence: by the submission of the 2027 budget a named, modelled drought-damage line should be included, by the start of the 2027 support year the public damage picture broken down by municipality should be in place, and before next year’s hail-damage season the measurement protocol of the JÉGER decision should be published. And to the public, one request as to outlook: this year’s wheat crop figure is not disaster news but planning data. Four million tonnes is enough for domestic bread, so the question is not the security of supply but who bears the recurring loss and under what rule.

Two MIAK foundational values are directly in play here. Data-drivenness, because all three proposals answer the same gap: in the dispute around JÉGER, in the distribution of damage compensation and in budgetary planning alike, decisions rest on claims that nobody measures. And accountability, because a rule fixed in advance and a public damage picture together make it verifiable whether the risk-sharing financed from public money reaches the place where the damage arose — in a system built on ex post, individual decisions this question cannot even be put.

Part VI — Justifications and further sources

6.1 The framing of the press, spectrum by spectrum

The economic band presented the topic as a question of quantity and framed it in terms of consumer prices. Two Portfolio pieces concentrated on the figures: one placed at the centre the crop a quarter smaller and the quantity of four million tonnes, with the reassuring framing that domestic supply is not in danger and no further rise in the price of bakery products is expected; the other published the gauge data for Lake Velence and the Copernicus satellite image. The paper also carried a separate piece reporting that even in Europe’s strongest economy it has been admitted that this year’s drought will show in growth. The characteristic of the framing is that it measures the damage as a market consequence — price, quantity, growth — and does not look at it from the side of institutional handling.

The liberal-left and general public-affairs band carried the same story from the side of spectacle and consequence. HVG was present with two pieces: one analysed the agricultural and growth effect of the water shortage, the other carried the satellite image of the drying Lake Velence. ATV broke the Lake Velence matter down into expert explanation and — alone among today’s pieces — made the model of use of the lake the question, rather than the water level itself. 444.hu, by contrast, disputed the tabloid framing: it drew attention to the fact that the most frequently displayed Balaton beach photograph is precisely not typical, because Lake Balaton is in a good position compared with other natural waters. This self-correction within the band is rare and valuable: on the same topic both the mood-based and the fact-based framing appeared.

The conservative band highlighted the political side of the intervention. Mandiner also featured with two pieces: one carried Hadházy’s statement criticising the shutdown of JÉGER, including the charge of populism and the rebuttal by the National Chamber of Agriculture, the other the border-law consequence of the falling level of the Danube. The logic of the framing is clear: what is interesting is not the size of the damage but the quality of the government decision. It is worth recording that among today’s pieces this is the only thread that makes the decision-making procedure the question — the left-wing and the right-wing framing are therefore not arguing about the same thing today, but putting two different questions to the same set of facts.

6.2 Facts and data

Data Value Source
The fall in this year’s wheat crop a quarter less than usual Portfolio, 21 August 2026
The quantity of this year’s wheat crop roughly 4 million tonnes Portfolio, 21 August 2026
The completion of the harvest mid-July, two weeks earlier than usual Portfolio, 21 August 2026
Rise in bakery product prices over five years +150 per cent Portfolio, 21 August 2026
Average price of a kilogram of half-brown bread 856 forints Portfolio, 21 August 2026
Change in bread consumption a fall of 15–20 per cent, a similar rise in pastries Portfolio, 21 August 2026
The water level of Lake Velence at the Agárd gauge 20 centimetres (16 August 2026) Portfolio, 21 August 2026
The previous lowest value 53 centimetres (2022 drought) Portfolio, 21 August 2026
The average depth of Lake Velence barely one and a half metres ATV / Zsombor Boromisza, 22 August 2026
The place of 2025 in the order of dry years the fourth driest since 2000 European Environment Agency (after Portfolio)
EU area with declining soil moisture nearly a quarter of the EU’s territory European Environment Agency (after Portfolio)
The most affected member states Hungary, Bulgaria, Cyprus, Greece, Romania European Environment Agency (after Portfolio)
The operating area of JÉGER after the decision four counties; shut down in the rest of the country Mandiner, 21 August 2026
Southern European water resource and crop yield outlook at 2 °C of warming a fall of some 20 per cent Nicholas Stern: The Economics of Climate Change

The last row of the table yields the non-obvious conclusion of this post. Stern’s southern European crop yield estimate — a fall of some twenty per cent at two degrees Celsius of global warming — is a long-term value relating to a decadal average. This year’s Hungarian wheat crop, by contrast, fell 25 per cent short of the usual in a single season. The two figures are not the same indicator and do not substitute for each other: the present year is an extreme, not a trend. Precisely for this reason the distinction matters — the modelled appropriation should not be set on the basis of a single bad year but in a scenario band, as proposal 3.1 asks.

6.3 Policy dimensions

  • Agriculture (programme points) — placing damage reporting and damage compensation on a data footing, and making use of precision farming data: MG2, MG1, MG5;
  • Environment and climate (programme points) — the modelling of climate damage and the methodology of measurable targets: K9, K1;
  • Economy (programme points) — the framework of the data-driven budget and of common resource management: G1, G28;
  • Territorial inequality and rural policy (programme points) — measuring the territorial distribution of drought damage: TE1;
  • Transparency and anti-corruption policy (programme points) — the public traceability of payments: A1.

6.4 Literature in detail

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

Stern’s review starts from the distinction that emission reduction no longer provides an answer to the effects occurring in the coming decades, because the effect of mitigation appears with a delay in time — until then the only available answer is adaptation. To this the volume adds two findings directly applicable to the present Hungarian situation. One is territorial: southern European water resources and crop yields are estimated to fall by some twenty per cent at two degrees Celsius of global warming, and the regions that are already water-scarce face the greatest and growing costs. The other is institutional, and this is the more important: adaptation is stimulated by those market and regulatory solutions which signal back the actual size of the risk.

“Risk-based insurance schemes, for example, provide strong signals about the size of climate risks and therefore encourage good risk management.”

Stern adds in the same place that a financial safety net may also be needed for the most vulnerable, because they are least able to pay for protection — that is, the risk-proportionate signal and the social safety net are not alternatives to each other but complements. In the case of this year’s Hungarian drought damage this means that damage compensation does not crowd out prevention if the degree of the contribution is aligned with the farmer’s own risk-reducing steps — precisely the risk-proportionality element of proposal 3.1.

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

6.4.2 William D. Nordhaus & Joseph Boyer: Warming the World

The volume by Nordhaus and Boyer describes the modelling solution that treats climate damage not as a separate event but as part of the production function: the damage function assigns to the change in global mean temperature a loss expressed as a proportion of the given region’s output, and it is this loss that is built into the economic model. The authors particularly emphasise that agriculture is one of the two sectors — alongside sea-level rise — where estimation at regional level is methodologically the most advanced, and therefore where modelling is the most reliable. At the same time the volume is self-critical: it records that a significant part of the early estimates was excessively pessimistic, while newer studies that also take adaptation into account paint a less dark picture.

“We derive the global impact from regional impact estimates. These estimates come from an analysis that takes market, non-market and potentially catastrophic impacts equally into account.”

In the case of Hungarian drought damage it follows from this that the domestic model under programme point K9 is not a theoretical luxury: agriculture is precisely the sector in which the connection between temperature and crop loss can best be quantified, and therefore the sector in which a named budget appropriation can be built on a modelled estimate rather than on a mere historical average.

📖 Source: William D. Nordhaus & Joseph Boyer: Warming the World — Economic Models of Global Warming

6.4.3 Elinor Ostrom: Governing the Commons

Ostrom’s work explores the question of how a resource can be managed sustainably when it is used in common by many and when individual property rights cannot be unambiguously separated within it — the cases she examines expressly include water management systems as well. The main claim of the book is that alongside the two usual solutions — privatisation or regulation imposed from outside — a third exists: the self-governing institutions of the users, provided that the system solves the threefold problem of provision, commitment and mutual monitoring. Monitoring for her does not mean external control, but that the users observe the rules visibly to one another as well, and that rule-breaking is followed by graduated sanctions.

“Stable self-governing institutions can be created if certain problems of provision, credibility and monitoring are solved.”

Applied to the Hungarian situation this means two things. On the one hand water resources — whether those of Lake Velence or of irrigation water — are precisely such commonly used, finite resources, so the credibility of the allocation rule stands or falls on public measurement. On the other hand the publicity of the damage reporting system fulfils the same role on the support side: the damage picture broken down by municipality works because farmers see one another’s reports in aggregate as well — this is the practical content of programme points G28 and MG5.

📖 Source: Elinor Ostrom: Governing the Commons — The Evolution of Institutions for Collective Action

6.5 International comparison

Institutionalised solutions for the budgetary handling of recurring drought damage exist in several EU member states. In Spain and in France an agricultural risk management arrangement operates in which the state contribution is linked to the insurance premium rather than to the ex post damage — that is, the state contributes to the sharing of the risk, not to the compensation of the damage. This is precisely the structure described by Stern’s risk-signalling logic: state support then does not crowd out preventive investment, because the degree of support is aligned with the insurance undertaken, not with the damage that has occurred.

In the matter of Lake Velence the international parallel is the management of shallow lakes with high evaporation. In lakes with similar characteristics, management practice treats the fluctuating water level as a starting condition and aligns shoreline use, holiday-zone regulation and beach infrastructure with it, rather than with a constant water level. According to data from the European Environment Agency the present situation is in any case regional: the decline in soil moisture affects nearly a quarter of the Union’s territory, most strongly Hungary, Bulgaria, Cyprus, Greece and Romania — which also means that EU-level funding and comparable methodology are available for adaptation instruments.

Agriculture

  • MG1 — Precision agriculture programme
  • MG2 — Agricultural data platform
  • MG5 — Common resource management framework (commons 2.0)

Environment and climate

  • K1 — Measurable climate targets
  • K5 — Just transition programme
  • K9 — Domestic integrated climate-economy model

Economy

  • G1 — Data-driven budget
  • G28 — Framework for common resource management

Territorial inequality and rural policy

  • TE1 — Micro-regional development index

Transparency and anti-corruption policy

  • A1 — Public money dashboard

Proposed new programme point: Risk-proportionate agricultural damage compensation — for the Agriculture policy area: the degree of the state contribution should be aligned with the farmer’s verified risk-reducing steps (irrigation development, soil protection, drought-tolerant variety, insurance), not with the size of the damage that has occurred.

6.7 List of sources

Press sources (MIAK press monitor, 22 August 2026 — topic 1):

Knowledge base references (literature):

  • 📖 Nicholas Stern: The Economics of Climate Change — The Stern Review
  • 📖 William D. Nordhaus & Joseph Boyer: Warming the World — Economic Models of Global Warming
  • 📖 Elinor Ostrom: Governing the Commons — The Evolution of Institutions for Collective Action

MIAK internal materials:

  • MIAK policy area: Agriculture (programme points; programme point ID: MG2)
  • MIAK policy area: Environment and climate (programme points and background material; programme point ID: K9)
  • MIAK policy area: Economy (programme points; programme point ID: G28)
  • MIAK policy area: Territorial inequality and rural policy (programme points; programme point ID: TE1)
  • MIAK press monitor, 22 August 2026 — topic 1, score: 93/100

Supplementary public data sources:

  • European Environment Agency (EEA) — soil moisture and drought statistics
  • Copernicus programme, Sentinel-2 — satellite monitoring of land cover and water surface
  • KSH (Hungarian Central Statistical Office) — agricultural statistics, cereal crop data

Generation metadata