Part I — Situation overview

At its meeting of 3 September 2026 the government decided on the continuation of the Mohács Danube bridge investment, and the government decision recording it appeared late in the evening of 4 September. The content runs in two directions. The bridge itself will be built: according to the statement of the minister for transport and investment the review established that the technical content of the structure can no longer be substantively reshaped — it can neither be made suitable for rail traffic nor narrowed to one lane in each direction — and reversing it would not be reasonable. On the section between the M6 motorway and the Danube bridge the original two-times-two-lane technical content also remains. What is dropped is the four-laning of the roughly 20-kilometre road section planned on the Great Plain side: this was ordered by the previous government in March 2026, raising the cost of the investment by more than 76 billion forints. According to the minister, on the basis of the traffic forecasts — taking the Serbian developments into account as well — a four-lane layout on this section would not be justified even up to 2050, so the road will be built as one lane in each direction, as in the original plans. The government decision prescribes the contract amendment with a deadline at the end of October, whereas agreement on the original technical content of the Danube bridge, the connection of Homorúd and a roundabout junction has to be reached only by 30 November 2027.

The decision ties together two earlier strands. One is the history of the Mohács bridge: in recent years the investment grew step by step, while the rail connection between Pécs and Szeged, which according to the minister could have fitted into the project, was not realised. The other is the wider government review programme: the same week saw the waste management system, the casino and tobacco concessions, and — according to Saturday’s announcement — the 35-year motorway concession also come onto the agenda. The Mohács decision is exceptional in this series because it is not about cancellation but about resizing, and because the government reported the saving in figures. On the authorisation of the government decision the minister also committed to making a proposal on the use of the funds saved: the aim is the renovation of neglected sections of the state main and minor road network, road safety and traffic developments. On the same day the ministry also announced the preparation of a tram-train network in Borsod-Abaúj-Zemplén county, which would build on the existing rail infrastructure — that is, a lower-cost solution that makes better use of an existing asset.

MIAK’s reading: this is the first major investment decision that is not about the political fate of the project but about the relationship between technical content and expected traffic — and precisely for that reason it matters that the calculation behind it should be public. With an inherited megaproject the easiest route is always either to finish it unchanged (because “the money is already in it”) or to cancel it as a political gesture. Resizing is the third, professionally most demanding route: it requires somebody to calculate where the boundary of genuinely justified content lies within the investment. MIAK supports this direction. The question is not whether the decision is right — on the arguments available it is defensible — but whether it remains a one-off decision or becomes a rule.

Part II — Foundations in the literature

Two sources give the interpretive frame. The World Bank’s report World Development Report 2015 — Mind, Society, and Behavior discusses in a separate chapter the bias that leads decision makers to continue projects already begun even when the arguments are against them. This is sunk cost bias: the tendency to persist with a decision because of the money already poured in, even though past expenditure ought not to influence a future decision. In its own experiment the report showed that the effect operates among development professionals as well, and that the greater the share already spent, the greater the willingness to continue. The International Monetary Fund’s (IMF) Fiscal Monitor 2025 approaches the same problem from the institutional side: on international data it identifies the most frequent weak points of public investment management as project appraisal and selection, inadequate maintenance funding, and the monitoring of existing public assets. Together the two sources give the yardstick for the present decision: the Mohács decision is a step out of a cognitive trap, but it will bring lasting improvement only if it becomes an institutional procedure. The detailed treatment of the literature — source by source, with quotations — can be found in section 6.4 Literature in detail.

Part III — MIAK’s concrete proposal

MIAK proposes three measurable steps. The starting point is that the present decision is defensible in substance, and that the line of reasoning it follows — on what volume of traffic the investment is built, and which part is actually justified — is exactly what MIAK expects of investment policy. The proposals therefore do not dispute the decision, they make it verifiable and repeatable.

3.1 Publication of the traffic calculation behind the decision and of the derivation of the saving (within 30 days)

The ministry should publish the traffic forecast on which it based the dropping of the four-lane layout, and an itemised derivation of the 76 billion saving. The traffic material should show the forecast of the average daily vehicle count on the section concerned, the path up to 2050, the assumption relating to the Serbian developments, and at what traffic level widening would become justified — this last is the most important figure, because it is what later makes it verifiable whether the decision estimated demand well. The derivation of the saving should also include any costs arising from the contract amendment, so that the gross and net figures do not become conflated. Publication is not a formality: the credibility of the figure depends on the derivation, and in transport policy credibility is the most important and hardest to restore capital. The proposal is a direct application of MIAK’s programme point G19 (Radical transparency in economic decision-making).

3.2 A mandatory project checklist for every state investment above 20 billion forints (from the 2027 budget cycle)

MIAK proposes that for every state investment above 20 billion forints a checklist fixed in advance and public should be mandatory, consisting of four elements: (1) an ex-ante cost-benefit analysis under a uniform methodology; (2) a comparison of at least three variants, among them a variant with lower technical content and a “do not build” base case; (3) a 30-year estimate of maintenance and operating costs, because the real price of a state investment is not the construction sum; (4) reasons for the decision if the variant chosen is not the one with the best cost-benefit ratio — under the “the ranking may be departed from, but it has to be justified in writing” principle of MIAK’s programme point KO4. The IMF analysis (see 6.4.2) identifies exactly this absence as the most frequent weak point in public investment management, and shows with Togolese and British examples that methodological standardisation brings a measurable improvement in efficiency. The proposal builds on programme points G20 (Economic policy impact assessment system) and EP4 (Construction industry transparency).

3.3 A three-year outcome review after handover, and traceable reuse of the saving

For every large investment a public outcome review should be prepared in the third year after handover: how much the actual traffic is compared with the forecast, how much the construction cost compared with the plan, and how much the actual maintenance cost is. This is the Drucker audit of MIAK’s programme point G20 — that is, an ex-post impact assessment comparing expected and actual results — applied at project level. In the case of the Mohács decision the traffic on the planned single-lane-per-direction section therefore also has to be measured: if the estimate was wrong, that has to be visible, and widening should be reopened on the basis of the measurement, not under political pressure. In addition, the use of the 76 billion forints saved should follow the aim set down in the government decision, and should be traceable item by item — with a project list, sums and named road sections. If a year later the saving has not turned from a large announced sum into identifiable concrete renovations, then the step remains a communications result, not a policy one. The proposal connects to programme points A1 (Public money dashboard) and TE4 (Rural public service minimum).

The three proposals are bound together by a single principle: a good investment decision is worth something if it is repeatable — and for that what is needed is not resolve but methodology. Publishing the calculation, a uniform checklist and a three-year outcome review together make it possible that at the next megaproject one does not have to rely again on the personal professional judgement of the decision maker.

Part IV — Expected effects and risks

Dimension Expected effect Risk
Budget 76 billion forints is freed up; the order of magnitude of the sum is comparable with the annual maintenance envelope of the national road network The contract amendment may involve costs or dispute; the net saving may be lower than the announced gross sum
Territorial accessibility The bridge is completed, the connection between Southern Transdanubia and the Southern Great Plain improves; the funds freed up can be directed to neglected minor roads If the traffic estimate underestimated demand, the two-lane section could later become a bottleneck, and subsequent widening is more expensive
Investment culture A precedent that an inherited project can be resized, not merely continued or cancelled Without a derivation the decision looks like a political gesture, and the next government could reach into a running investment in the same way, with the same lack of support
Construction industry Lower technical content means more predictable, lower-risk implementation On the contractor side a shortfall arises on the 20-kilometre section already ordered; the contract amendment could trigger a damages claim

The main question of judgement is the relationship between sizing and future expandability. An under-sized road can later be widened at a greater cost than the initial saving was worth — which is why present traffic is not enough: the path up to 2050 and the reserve for widening (alignment, land safeguarding, preparation of structures) also matter. The public traffic material is precisely for this reason not a luxury: without it there is no deciding whether the decision was thrifty or merely tight-fisted. The second point of judgement is the timing asymmetry in the government decision: the contract amendment on the narrowing has to be carried out by the end of October, whereas agreement on the original technical content of the bridge and the associated junctions is required only by November 2027. This is understandable — the saving is urgent, the details are not — but it also carries a risk: the questions left open are under political attention at the time the decision is taken, but a year later they no longer are.

Part V — Measurability and summary

5.1 What is worth following? (suggested KPIs)

The performance indicators below (KPIs, Key Performance Indicators) are suggestions, not government decisions — MIAK considers them suitable for judging in 12 and 36 months’ time what the decision has delivered:

  • The traceability of the use of the saving: how much of the 76 billion forints has appeared in itemised, identifiable road renovation projects, with named road sections. Suggested target: 100 per cent in the first two years.
  • The divergence of forecast and actual traffic: the average daily traffic measured in the third year after handover compared with the forecast. This tells whether the sizing decision was good.
  • The share of large investments starting with a public cost-benefit analysis: for what percentage of new state investments above 20 billion forints an ex-ante, published analysis was prepared. Suggested target: 100 per cent by 2028.
  • The length of state minor roads renovated annually: this is the indicator that connects the saving with the end result that the road user actually perceives.

5.2 Summary

MIAK’s request can be summed up in a single sentence: the ministry should publish the traffic calculation behind the Mohács decision and the derivation of the saving, and should raise to a rule what it did rightly in this case. The decision is in itself a good direction: it was taken not on the political meaning of the investment but on the relationship between expected traffic and technical content. That is rare in Hungarian investment policy, and precisely for that reason it would be a pity to leave it as a one-off event.

Two MIAK foundational values are at play here. Data-drivenness, because the substantive content of the decision is a traffic forecast — that is, exactly the kind of argument that MIAK asks for in place of prestige or symbolic arguments; yet data-drivenness is a yardstick only if the data itself is visible, not merely the conclusion that invokes it. And accountability, because without a three-year outcome review it will never emerge whether the saving was a genuine improvement in efficiency or a quiet lowering of the service level — and that difference can only be established afterwards by measurement.

Part VI — Reasoning and further sources

6.1 The framing of the press by spectrum

The public affairs and liberal-left band processed the decision factually, concentrating on the details. HVG gave the most precise institutional framing: it took as its basis not the ministerial announcement but the text of the government decision, and from it highlighted what other papers did not carry — the end-of-October deadline for the contract amendment, the November 2027 deadline for the further technical questions, and the authorisation under which the minister has to make a proposal on the reuse of the funds saved. This framing matters because the fate of the saving is the element of the news most often lost. 24.hu carried the related rail development announcement as a separate item, and recalled that the region had had a previously closed railway line restored the same week.

The economic band focused on the figures: alongside the Mohács decision Portfolio also covered the M6 operating saving and the Borsod-Abaúj-Zemplén tram-train preparation, that is, it treated the three as elements of a turn in transport policy. The framing of the conservative band diverged most: Mandiner described the news as a “pruning”, and put the emphasis on a significant part of the investment being dropped — that is, it chose the loss reading instead of the saving reading. This is a legitimate framing, and it also raises a real question that MIAK likewise considers important: lower technical content can become a bottleneck in the long run. The political form of the debate was documented by ATV, which carried the public exchange between the minister and a government-independent analyst on the size of the saving — but the debate was about the credibility of the figure, not about the methodology behind it. It is precisely this difference that justifies MIAK’s proposal 3.1: as long as the derivation is not public, the debate has no means of resolution, only volume.

6.2 Facts and data

Data Value Level of evidence
The day of the government decision 3 September 2026 (cabinet meeting) ministerial statement
The publication of the government decision 4 September 2026, late evening official gazette (Magyar Közlöny)
The section taken out of the project some 20 km, two-times-two-lane layout on the Great Plain side government decision, ministerial statement
The date of ordering the section taken out March 2026 (previous government) press report, consistent across several papers
The saving 74–76 billion forints ministerial statement; slight divergence between papers
The deadline for the contract amendment end of October 2026 government decision
The deadline for agreement on the further technical questions 30 November 2027 government decision
The time horizon of the traffic justification the four-lane layout is not justified even up to 2050 ministerial statement, the underlying calculation is not public

One remark on the data: the papers do not report an entirely identical figure for the amount of the saving — some accounts mention 74, others 76 billion forints, and give the original value of the order as “more than 76 billion forints”. The blog therefore uses the announced order of magnitude. The exact value will be establishable after the closing of the contract amendment, and this is itself an argument for publishing the derivation.

6.3 Policy dimensions

  • Transport and infrastructure (programme points) — data-based project ranking and carrying over the “comply or explain” principle to road investments (programme point ID: KO4);
  • Construction (programme points) — real-time monitoring of investment cost and deadline, transparency of the subcontractor chain (programme point ID: EP4);
  • Economy (programme points) — mandatory ex-post impact assessment and publicity for the reasoning behind decisions (programme point ID: G19, G20);
  • Territorial inequality and rural policy (programme points) — the condition of the minor road network as a basic precondition of rural accessibility (programme point ID: TE4);
  • Transparency and anti-corruption policy (programme points) — itemised traceability of the use of the saving (programme point ID: A1).

A clarification of competences: amending the technical content of the investment is a matter of agreement between the contracting parties, which the government decision merely prescribes as a task of the minister — the decision itself does not amend the contract. The legal closing of the amendment and the settlement of any contractor claims therefore take place within the contractual procedure, and in case of dispute this too is decided by a court.

6.4 Literature in detail

6.4.1 World Bank: World Development Report 2015 — Mind, Society, and Behavior

The tenth chapter of the report devotes a separate analysis to the biases that affect development and policy decision makers. The most important for the present case is sunk cost bias, which the report defines as follows:

“Policy makers can also be influenced by the sunk cost bias, which is the tendency of individuals to continue a project once an initial investment of resources has been made. To stop a project is to acknowledge that past efforts and resources have been wasted; thus the bias may arise from the cultural admonition not to appear wasteful.”

The report also publishes its own experimental evidence on this: the authors asked World Bank staff about continuing a five-year programme that had run into a dead end, with varying shares of funds already spent. According to the result, the greater the share already spent, the greater the willingness to continue — that is, the bias affects professionally trained decision makers as well. Read in this frame the Mohács decision is noteworthy: the government made exactly the distinction that the logic of the report requires. For the structure already built, reversal really would be unreasonable — here continuation is not bias but correct judgement. For the section not yet started, however, there is no sunk cost, so the decision has to be taken purely on the basis of expected traffic. Yet according to the report, defence against the bias does not depend on individual vigilance but on procedure: this is why MIAK proposes institutionalising an ex-ante checklist.

📖 Source: World Bank: World Development Report 2015 — Mind, Society, and Behavior

6.4.2 IMF: Fiscal Monitor 2025 — Public Spending Efficiency

The 2025 edition of the International Monetary Fund (IMF) examines how much more could be achieved from the same amount of state spending. According to the analysis the efficiency gap in public spending — the difference between the actual result and the best result achievable from the same funds — is some 31 per cent even in advanced economies. On the investment side the report names the most frequent weak points concretely:

“Common weaknesses in the management of public investment include those involving project appraisal and selection, adequacy of maintenance funding, and monitoring of public assets.”

All three of these elements are demonstrable in the case of the Hungarian road network: project appraisal is missing or not public, maintenance funding is disproportionately low compared with the sum spent on the expressway network, and there is no public, aggregated report on the condition of existing public assets. The report also brings two reform case studies: in Togo the introduction of a multi-year public investment programme, the standardisation of project appraisal methodology and mandatory cost-benefit analysis for every project reduced the investment efficiency gap by five percentage points in eight years; in the United Kingdom the infrastructure supervisory office set up in 2016 improved project selection and cost control. Both examples show that the source of improvement is not extra money but the decision procedure — this is the direct basis of MIAK’s proposal 3.2.

📖 Source: IMF: Fiscal Monitor 2025 — Public Spending Efficiency

6.5 International comparison (where relevant)

Disputes over the sizing of megaprojects recur in every European country, and good practice can be narrowed to two elements. One is a uniform, mandatory project appraisal methodology: many countries issue central guidance on how the costs and benefits of a state investment are to be accounted for, and departures from it have to be justified — this is what makes the projects of different ministries comparable. The other is independent project supervision with public status reporting: in the United Kingdom the status of every large state investment is available with a red-amber-green classification, and in Chile the contracts and amendments of state construction investments are public in real time — MIAK’s programme point EP4 builds on these patterns. On the rail side the Swiss “Bahn 2000” programme serves as the most instructive example: the country optimised not for maximum speed but for a clock-face timetable and interchanges at nodes, and thus doubled rail passenger traffic in twenty years at a fraction of the cost of a high-speed network. The logic of the present Hungarian decision — making better use of existing infrastructure instead of greater technical content, exemplified by the announced preparation of the tram-train network as well — points in the same direction. What is missing from it is the other half of Swiss and British practice: the public, comparable calculation.

Transport and infrastructure

  • KO4 — Rail development with data-based priorities

Construction

  • EP1 — Digital construction permitting
  • EP4 — Construction industry transparency

Economy

  • G19 — Radical transparency in economic decision-making
  • G20 — Economic policy impact assessment system
  • G21 — Systematic review of state expenditure

Territorial inequality and rural policy

  • TE4 — Rural public service minimum

Transparency and anti-corruption policy

  • A1 — Public money dashboard

Suggested new programme point: A mandatory project checklist for state investments above 20 billion forints — ex-ante cost-benefit analysis, three-variant comparison and a three-year outcome review — for the Transport and infrastructure area.

6.7 List of sources

Press sources (MIAK press monitor, 6 September 2026 — topic 2):

Knowledge base references (specialist books):

  • 📖 World Bank: World Development Report 2015 — Mind, Society, and Behavior
  • 📖 IMF: Fiscal Monitor 2025 — Public Spending Efficiency

MIAK internal materials:

  • MIAK policy area: Transport and infrastructure (programme points; programme point ID: KO4)
  • MIAK policy area: Construction (programme points; programme point ID: EP1, EP4)
  • MIAK policy area: Economy (programme points; programme point ID: G19, G20, G21)
  • MIAK policy area: Territorial inequality and rural policy (programme points; programme point ID: TE4)
  • MIAK press monitor, 6 September 2026 — topic 2, score: 90/100

Supplementary public data sources (where used):

  • Magyar Közlöny — the government decision on the Mohács Danube bridge
  • Eurostat — regional accessibility indicators

Generation metadata