Part I — Situation overview

Romania began on 7 August, and continued on the 8th, the operation in which it sinks barges loaded with stones into the bed of the Danube. The aim is to build an artificial underwater sill in the bed of the Bala branch, some 60 kilometres from Cernavodă, so that more water is diverted into the Old Danube leading towards the power plant. By dawn on Saturday the controlled flooding of two barges had been completed; two more were planned for that afternoon. The stake is concrete: at Cernavodă a water level of minus 222 centimetres was measured on Saturday, and if the level reaches minus 233 centimetres, the still-operating, 700-megawatt unit two of the Cernavodă nuclear power plant will also have to be shut down. From the intervention the authorities hope for a respite of three to four days. The flow of the river at the entry point of the Romanian reach, at Baziaș, sank in the middle of the week to the historic low recorded four decades ago, to 1,400 cubic metres per second; the water authorities expect no substantive change before 14 August. The Romanian government declared a state of readiness and authorised the operator of the electricity grid to disconnect large industrial consumers as necessary. On the Hungarian reach, meanwhile — over recent days — a slight rise has begun: at Paks the water level on Saturday was minus 127 centimetres, and the temperature of the river 28.3 degrees Celsius, which makes continuous monitoring of the effect of power plant cooling on wildlife necessary.

On the same weekend one of the most precisely documented cases of domestic drought damage became known. In the fish ponds of Rétimajor and Rétszilas some 150 tonnes of fish died; according to the operator’s statement the value of the direct stock loss approaches 165 million forints. The ponds concerned had been left without substantive water replenishment for six to eight weeks, while on the upper reach of the same water system irrigated crop production continued — with reduced volumes, but continued. The Rétszilas ponds are a nature conservation area of national importance, a Natura 2000 site and, under the Ramsar Convention, a wetland of international importance; the presence of some 220 bird species has been documented in the region. Nationally there are about 27,000 hectares of operating fish pond area, typically in shallow systems exposed to evaporation and supplied through canals.

The third thread is the least spectacular and the longest-term. According to a statement of the Hungarian Academy of Sciences prepared with the involvement of three experts, the record low water level is traceable back not to one but to two mutually reinforcing processes. One is indeed climate change, which is transforming the spatial and temporal distribution of precipitation. The other is riverbed subsidence: on the Hungarian reach of the Danube the bed has subsided by an average of some two metres over the past 60 years, while on the Tisza and the Drava levels have fallen by an average of some one and a half metres over the past century. The cause of this is the river regulation of the previous century for shipping development and flood protection purposes: the main beds were stabilised and narrowed, side branches were closed off, high-water beds were confined between embankments — so the quantity of sediment travelling in the rivers fell, while the erosive capacity of the bed grew. A deeper bed contributes to a lower water level just as much as the decrease in the quantity of water.

Domestic political debate, by contrast, looks backwards: onto the agenda came the minutes of the committee hearing last December of former investment minister János Lázár, in which he spoke of a reservoir stock unsuited to water retention and of the government having made a mistake when it did not build large reservoirs ten years earlier. MIAK’s reading: the retrospective allocation of responsibility is misleading because the present damage does not follow from a single missed investment. Three mutually independent system failures run together: there is no public order of allocation for the case where there is less water than demand; there is no consultation mechanism for unilateral riverbed interventions carried out on a shared river; and there is no independent, climate-independent handling of bed subsidence. All three failures can be remedied, and none of them depends on the weather.

Part II — Literature foundation

The situation raises two clearly distinguishable economic questions. The first is that of timing: when is it worth intervening if the damage arises later and uncertainly? Nicholas Stern (British economist, former chief economist of the World Bank, author of the large-scale government report on the economics of climate change) derives in the work known as The Stern Review that adaptation has limits: it is not capable of reducing the net costs of climate change to zero, residual damage always remains, and adaptation itself carries a cost. Moreover, spontaneous, individual adaptation regularly achieves less than would be necessary, so policy instruments are needed to dismantle the obstacles. This is precisely the structure of the Rétimajor case: the individual actors each acted rationally, yet at the level of the system damage arose. The second question is that of cross-sectoral feedbacks: how can it be measured what loss a water abstraction causes elsewhere? The work Warming the World by William Nordhaus (American economist, awarded the Nobel Memorial Prize in Economics in 2018 for the economic modelling of climate change) describes the construction of integrated assessment models: these models link end to end economic activity, the physical environmental effect and the feedback to human systems. Such a model would give what is missing today: by how much the abstraction of irrigation water reduces the value of the fish ponds and habitats lying further downstream. The detailed treatment of the literature — author by author, with quotations — can be found in the 6.4 Literature in detail section.

📖 Source: Nicholas Stern: The Economics of Climate Change — The Stern Review; William Nordhaus: Warming the World

Part III — MIAK’s concrete proposal

MIAK proposes three measurable measures. All three can be started even in the present water-scarce period, and none of them requires a new large investment.

3.1 A public order of water allocation and thresholds (by 30 November 2026)

The most urgent shortage is not of water but of rules. There is today no publicly published, pre-fixed order setting out which group of users has priority in a water-scarce period, and at what measured water level or flow threshold each degree of restriction comes into force. The proposal: the ministry responsible for river basin management and the water directorates should publish the degrees before the start of the next irrigation season — similarly to the way the consumption reduction degrees worked during the energy crisis. The order cannot be “agriculture” or “industry” in general; the Rétimajor case shows precisely that the real conflict runs within one and the same sector, between the upper and lower reaches of the same water system. The degrees therefore have to be at the level of the water system, and they also have to contain, as an independent item, the minimum water demand of fish farming needed to keep the stock alive. The measure builds on the data infrastructure of K3 pollution monitoring, and in its decision logic it is the application of the G28 community resource management framework to a single resource. Stern’s argument (see 6.4.1) is directly applicable here: the obstacles to individual adaptation have to be removed with a policy instrument, because they do not resolve themselves.

3.2 A protocol of prior consultation on Danube riverbed interventions (initiative in autumn 2026)

The Romanian sinking of barges is not necessarily unlawful, but it appeared to the public as a unilateral intervention in the bed of a shared river with international legal status. MIAK proposes that Hungary initiate the adoption, among the Danube riparian states, of a simple, mandatory procedure of prior information and consultation: before every riverbed intervention which modifies water sharing durably or to a significant extent, the state concerned informs the downstream and upstream parties a specified time in advance, publishes the extent of the intervention and its expected effect, and receives observations. The proposal is deliberately modest: it asks not for a veto right but for information and time. Accordingly its diplomatic cost is also low, while it would already work in the next drought year. The natural forum of the initiative is the International Commission for the Protection of the Danube River; on the Hungarian side a joint submission by the foreign affairs and water management ministries is needed. This is the cross-border extension of the G28 programme point, and on the measurement side it follows the logic of K1 measurable climate targets.

3.3 Launching a riverbed subsidence monitoring and slowing programme (first report by mid-2027)

Riverbed deepening is the only component of the present situation which is not a function of the weather but a consequence of earlier engineering decisions — and therefore directly influenceable. MIAK’s proposal: the water service should begin annual measurement of the bed subsidence of the main river reaches with a uniform methodology, and publish it cross-section by cross-section; then, on the basis of the first report, an intervention plan should be prepared on the possibilities of sediment replenishment, the reopening of side branches and the widening of the high-water bed, with cost and impact estimates by reach. The aim is not to reverse river regulation but the measurable slowing of the pace of subsidence. The output of the programme can be fed directly into the K9 domestic integrated climate-economic model (see 6.4.2), and it grounds the application of the G20 impact assessment system to every future water management investment: when a reservoir or a shipping development is assessed, the riverbed effect has to become a quantified item.

The three proposals are bound together by a single principle: drought is not an event but a state, and states have to be handled not with crisis management but with pre-fixed decision rules. The order of allocation says what happens if there is little water; the consultation protocol says what happens if the neighbour intervenes; and the riverbed monitoring says by how much there will be less water even if the same amount of rain falls. All three documents do the same thing: they bring the weighing forward to before the crisis.

Part IV — Expected effects and risks

Dimension Expected effect Risk
Economy Predictable degrees of restriction reduce uncertainty for producers; irrigation and fish farming investment decisions become plannable A public order inevitably also names a loser; if the degrees turn out to be the results of political bargains, the credibility of the system is lost immediately
Society The priority of household drinking water supply becomes explicit, which also simplifies crisis communication Presenting the demand of fish ponds and wetlands as an independent item gives occasion for misreadings of the “fish matter more than people” type
Public administration Publication of riverbed subsidence data makes river regulation decisions accountable The data also qualify earlier investments retrospectively, which may provoke political resistance to publication
External relations As initiator of the prior consultation protocol Hungary’s position strengthens on the regional water management agenda If the neighbouring states read the proposal as preparation for Hungarian river barrage ambitions, the credibility of the initiative suffers

The most sensitive point of judgement is the relation between the order of allocation and local room for manoeuvre. A centrally fixed, rigid system of degrees may disregard local conditions well known to the specialists of the water directorate — in which case the rule forces a worse decision than judgement would. The proposal therefore works well if the degrees fix the thresholds and the minimums to be protected, while in the intermediate range room for judgement remains, the exercise of which has to be documented and justified afterwards. The second sensitive point is timing: the order of allocation will be credible if it is born not in the middle of a drought, under the pressure of an individual conflict, but before the season, in the cold. In the present situation this means that the work has to be done in the autumn, when the pressure eases — this is the hardest part, because by then public attention will have turned elsewhere.

Part V — Measurability and summary

5.1 What is worth tracking? (proposed KPIs)

Four proposed performance indicators (KPIs, Key Performance Indicators) from which, in the next drought period, it will be visible whether preparedness has improved:

  • Whether the public order of water allocation appeared by 30 November 2026, by water system, with quantified thresholds. A binary, document-based indicator.
  • The size in hectares of fish pond area left without water replenishment, compared with the figure for the 2026 season. This is the only indicator which directly measures whether the situation of the lower reaches has improved.
  • The share of cross-border riverbed interventions accompanied by prior consultation relative to all such interventions. It starts from zero, so any movement is measurable.
  • The annual pace of bed subsidence of the main river reaches in centimetres, with a uniform methodology, published cross-section by cross-section. This is the slowest-moving indicator, and therefore the most suitable for judging the long-term direction.

5.2 Summary

MIAK’s message: in the present water crisis the most expensive shortage is not of precipitation but of a decision rule. Concretely, it asks the government to publish, before the next irrigation season, the order of allocation by water system and the restriction thresholds; to initiate with the Danube riparian states prior consultation on riverbed interventions; and to start the uniform measurement of bed subsidence. And of the public it asks that instead of the retrospective allocation of responsibility it demand these three documents — their existence is more simply verifiable than any political claim about who failed to do what ten years ago.

Two MIAK foundational values move in this matter. Data-drivenness, because the case of riverbed deepening shows precisely that an existing, measurable cause can remain off the agenda for decades simply because there is no regular, public time series about it — while climate change has become a frame that explains everything, and has thereby covered over the part that can be intervened in. And universal representation, because in the absence of an order of allocation water goes where access is stronger: by the structure of the Rétimajor case the upper reach was irrigated and the lower reach dried out, and in this there was no decision, only the absence of an order. A public rule does precisely this: it makes the demand of the user in the weaker position a named item as well.

Part VI — Justifications and further sources

6.1 The press framing by spectrum

The liberal-left band concentrated on the intervention next door and on the consequences for households. Telex published a factual, figure-rich report on the Romanian sinking of barges — water level, threshold, flow, the estimated length of the respite — without political assessment; the same paper carried in a separate article the disaster management service’s statement on the fire situation, which presented the other consequence of the heat. HVG used two differing registers: one article asked service providers and specialists about the time horizon of drinking water supply, the other, a report from the Great Plain, carried the perspective of farmers — the latter being the only material in the day’s selection which presented the producers’ decision situation from the inside. In a third, summarising piece the same paper noted ironically that Hungary had become at once the country of a million water management experts — the framing placed at the centre the sudden expertification of public discourse.

The public-affairs band gave the most new factual material. 24.hu published the detailed data of the Rétimajor fish kill on the basis of the operator’s statement, including the coincidence of upper-reach irrigation and lower-reach water shortage, together with the sector’s proposals; the same paper presented the academy statement on riverbed deepening, and this material is the most important professional yield of the whole day’s selection. In a third article the paper worked up the political debate, also approaching the previous government’s state secretary responsible for water management, who admitted that more could have been done but invoked the lack of funds. ATV made news of water running to waste beside an industrial investment in Debrecen, which raises the question of proportionality at a time of household restrictions.

In the economic band Portfolio covered the widest spectrum: it published material on the security of the capital’s water supply, on the water balance of the soil, on the long-term prospects of the Danube and on the data of Lake Balaton as well — the framing was consistently built around the question of “what is the new normal”, not around responsibility. The conservative band on this day narrowed the topic to meteorological data and the daily movement of the water level: Magyar Nemzet published a piece on the severity of the July data and on the fall of the water level at Paks, while Mandiner carried the preparation of summer festivals in relation to energy and water. The most striking difference between the bands is therefore not the assessment of the facts but the time horizon of the perspective: the public-affairs and economic bands sought structural causes, the conservative band stayed with the current measured data, and the liberal-left band highlighted the question of political responsibility.

6.2 Facts and data

Item Value Source
Water level at Cernavodă −222 cm (8 August 2026, 4 cm lower than the previous day) Telex / MTI, 8 August 2026
Shutdown threshold of Cernavodă unit two −233 cm Telex / MTI, 8 August 2026
Capacity of Cernavodă unit two 700 MW Telex / MTI, 8 August 2026
Flow of the Danube at Baziaș 1,400 m³/s — the historic low recorded four decades ago Telex / MTI, 8 August 2026
Respite expected from the sinking of barges 3–4 days Telex / MTI, 8 August 2026
Water level at Paks −127 cm (8 August 2026), expected return to close to −97 cm Telex / MTI, 8 August 2026
Water temperature of the Danube at Paks 28.3 °C Telex / MTI, 8 August 2026
Fish kill at Rétimajor and Rétszilas approx. 150 tonnes Statement of Aranyponty Halászati Zrt., 24.hu, 8 August 2026
Value of the direct stock loss approx. 165 million forints Statement of Aranyponty Halászati Zrt., 24.hu, 8 August 2026
Period without water replenishment at the Rétimajor ponds 6–8 weeks Statement of Aranyponty Halászati Zrt., 24.hu, 8 August 2026
Hungary’s operating fish pond area approx. 27,000 hectares 24.hu, 8 August 2026
Documented bird species at Rétimajor approx. 220 species 24.hu, 8 August 2026
Bed subsidence on the Hungarian reach of the Danube on average approx. 2 metres over the past 60 years Statement of the Hungarian Academy of Sciences, 24.hu, 8 August 2026
Bed subsidence on the Tisza and the Drava on average approx. 1.5 metres over the past century Statement of the Hungarian Academy of Sciences, 24.hu, 8 August 2026

Two pairs of figures deserve separate attention. The first is the relation of the 150 tonnes of fish and the 6–8 weeks without water replenishment: the damage is not the result of a sudden event but of a weeks-long, foreseeable process — that is, there was time to intervene, there was simply no rule about whose job it was. The second is the relation of the two-metre bed subsidence and the daily water level data: alongside the news flow about daily centimetres it goes almost unnoticed that the reference base itself has shifted. If the bed has subsided by an average of two metres over 60 years, then the same flow gives a lower water level today than before — the records are therefore produced in part by a slow process of engineering origin, not only by the drought.

6.3 Policy dimensions

  • Environment and climate (programme points) — the data basis of the allocation thresholds, the monitoring of bed subsidence and the inputs of the domestic climate-economic model (programme point ID: K1, K3, K9);
  • Agriculture (programme points) — measurement of irrigation water abstraction and the naming of the water demand of fish farming (programme point ID: MG1, MG2, MG5);
  • Transport and infrastructure (background material) — the connection between navigability and riverbed deepening, and the ex post riverbed effect of shipping development interventions;
  • Territorial inequality and rural policy (programme points) — the public service minimum of the Great Plain regions most affected by water shortage (programme point ID: TE4);
  • Economy (programme points) — the frame of common resource management and the impact assessment obligation for investments (programme point ID: G28, G20).

6.4 Literature in detail

6.4.1 Nicholas Stern: The Stern Review

The chapter of Stern’s report which discusses the economic frame of adaptation makes two findings that fit the present situation directly. The first is that adaptation is not a panacea: the report emphasises that “adaptation is unlikely to reduce the net costs of climate change to zero — that is, it has limits”, and that residual damage often remains, while adaptation itself also carries a cost. The second is that adaptation does not get going by itself to the necessary extent. Analysing modelling practice, the report records that local-level cost-benefit studies generally show a net benefit, so some adaptation will certainly take place — “although policy measures are often needed to overcome the barriers”. These two statements together explain the Rétimajor case. The individual producers each adapted: on the upper reach they irrigated with reduced volumes, which from their own point of view was rational. For the fish ponds on the lower reach, however, this decision meant water shortage, and their “adaptation” was a non-existent option, because the water physically did not arrive. At the level of the system, therefore, spontaneous adaptation did not bring a good result — precisely the case for which Stern recommends policy intervention. The order of allocation proposed by MIAK in point 3.1 removes this obstacle: it does not create more water but decides who may abstract how much, before the damage occurs.

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

6.4.2 William Nordhaus: Warming the World

Nordhaus’s work is about how the economic and physical connections of climate change can be handled in a single, computable model. According to the book’s description, integrated assessment models — among them the widely used DICE and RICE models he developed — link end to end the various sides of global warming: economic activity, the carbon cycle, the radiative balance of the atmosphere, and then the feedback to human societies and natural ecosystems. Nordhaus at the same time frankly indicates the difficulty as well: one of the hardest tasks of model building is finding parsimonious, manageable relationships between economic activity and climatic change. Translated to the Hungarian situation this means that the value of the model lies not in predicting the future but in making the trade-offs between sectors quantifiable. Today, when an irrigation water abstraction permit is issued, there is no figure to say by how much this reduces the value of the fish ponds and wetlands lying downstream and — in the longer term — of the drinking water base. The programme point on a domestic integrated climate-economic model targets precisely this gap, and the bed subsidence time series proposed in point 3.3 is one of the inputs without which the Hungarian model could not reproduce the real development of the water level.

📖 Source: William Nordhaus: Warming the World

6.5 International comparison

A pre-fixed order of allocation is not a Hungarian innovation: in regions regularly suffering severe droughts it is established practice to tie the degrees of water shortage to measured thresholds, and to assign publicly to each degree what extent of restriction each group of users can expect. The common feature is that household drinking water supply and use linked to human health enjoy priority everywhere, while the contested question is the ranking of agricultural, industrial and ecological use relative to one another. Experience shows that the system which works well is the one which treats the ecological minimum — the quantity of water needed for the survival of the river or the lake — not as a residual but as an independent, protected item, because otherwise the restriction always reaches the user with the weakest capacity to assert interests first.

On the question of cross-border riverbed interventions the general principle of international water law is the prohibition of significant harm and the obligation of prior information, but practice is strongly uneven: it works where there is a joint, continuously sitting technical body which also shares the data. The Danube basin is in a relatively good position in this respect, because a joint protection organisation and joint data collection exist — yet the present Romanian operation took place without the Hungarian public knowing of it in advance. This indicates that the gap is not on the side of the forum but on the side of the procedure; this is why MIAK proposes a protocol, not a new institution.

Environment and climate

  • K1 — Measurable climate targets
  • K3 — Pollution monitoring
  • K9 — Domestic integrated climate-economic model (DICE-HUNGARY)

Agriculture

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

Economy

  • G20 — Economic policy impact assessment system (Drucker audit)
  • G28 — Framework of community resource management (Ostrom-CPR)

Territorial inequality and rural policy

  • TE4 — Rural public service minimum

Proposed new programme point: Riverbed subsidence monitoring and slowing programme — for the Environment and climate area, with an annual, cross-section-by-cross-section public time series and an intervention plan by reach.

6.7 List of sources

Press sources (MIAK press monitor, 9 August 2026 — topic 2):

Knowledge-base references (books):

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

Note: the local file path of the books does not appear in the visible text of the blog — only the author and the title. The file path is an internal matter of the generation process, not the reader’s.

MIAK internal materials:

  • MIAK policy area: Environment and climate (programme points; programme point ID: K9)
  • MIAK policy area: Agriculture (programme points; programme point ID: MG2)
  • MIAK policy area: Economy (programme points; programme point ID: G28)
  • MIAK policy area: Transport and infrastructure (background material)
  • MIAK press monitor, 9 August 2026 — topic 2, score: 93/100

Additional public data sources (where used):

  • General Directorate of Water Management water level and flow data; International Commission for the Protection of the Danube River; Copernicus European Drought Observatory; KSH agricultural statistics

Generation metadata