Part I — Situation overview

On Tuesday, 29 September 2026, the government published on the Kormánydöntések 2010–2026 (Government decisions 2010–2026) page of kormany.hu the list of the so-called 2000-series government decisions adopted between 2010 and 2026, in yearly PDF files. According to the government’s announcement, 3,598 decisions were adopted in this period that were not published in the Hungarian Gazette (Magyar Közlöny). They were sent directly to those concerned. They typically assign tasks to government, law enforcement or other state leaders, and they are not the same as government decisions containing classified information. The publication covers the serial numbers and titles of the decisions. Their text — which by HVG’s estimate may run to as many as ten thousand pages — still cannot be accessed. According to HVG’s summary, 2000-series decisions are as a rule non-public for twenty years (previously it was ten years), because they qualify as material supporting decision-making. And until the end of 2008 they were accessible to anyone in the official journal Határozatok Tára (Register of Decisions) — so secrecy is not an old tradition, but a practice dating from 2009.

28 days ago — in connection with a package released under a freedom of information request, containing financial decisions worth more than 500 billion forints — MIAK asked for three things: an itemised register of every non-public decision, a compulsory expiry period and audit access. The current list of titles partly fulfils the first of these, retroactively: the existence and subject of the decisions are now public. It does not, however, fulfil what turns a register into a rule. It contains no date, amount or circle of beneficiaries, no legal basis for the confidentiality, and no obligation for the future. And according to HVG, 2000-series decisions were adopted even after the change of government on 9 May 2026.

MIAK’s reading: the greatest value of the list is not the revealing power of individual titles, but the fact that for the first time it is possible to measure what this form of decision was used for. And the measurement brings two previously unseen facts to the surface. Non-public governance became widespread after 2020, and was increasingly used for distributing money. In addition, there are nearly five hundred gaps in the list’s serial numbers, for which the publication offers no explanation. The problem is therefore no longer that the number of secret decisions is unknown — but that there is no rule that would say which of them should not have been secret.

Part II — Foundations in the literature

Three sources provide the framework for interpreting the topic. The first is the constitutional text itself: under Article 39 of the Fundamental Law of Hungary, every organisation managing public funds is obliged to account publicly for them, and data on public funds and national assets are data of public interest. State aid awarded to a company or foundation therefore belongs, in the constitutional sense, to the public sphere from the outset. The second is the volume The Authority of Law by Joseph Raz (Israeli-born British legal philosopher, a defining figure of the Oxford school of legal theory). Raz sets out as one of the basic principles of the rule of law that the making of particular laws should be guided by open, stable, clear and general rules. The problem is not the individual decision itself, but the absence of a public, previously known framework behind it. The third is Corruption and Government by Susan Rose-Ackerman (American lawyer-economist, professor at Yale University, one of the founders of the institutional economics of corruption). On subsidy programmes and individual favours granted to firms, it argues that structures allowing unlimited discretion are the most likely sources of abuse. Ultimately, it is the disclosure obligation that draws the line between legal and illegitimate advantage. The detailed discussion of the literature — 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, which build on the package of proposals of 2 September 2026 but take individual points of it further. There, the stake was a register at the level of a government decree and the expiry period. Now that the retroactive list is available, the question is whether publicity becomes a lasting rule, and whether the content of the most important past decisions can also be known.

3.1 Exhaustive statutory grounds and a ban on secrecy for individual subsidy decisions (bill by the end of 2026)

The conditions for a non-public government decision should be laid down not in the Government’s own rules but in an act, so that a later government cannot amend them by a unilateral decision. Under the proposal, the Government submits a bill, which the National Assembly adopts. It must list exhaustively the grounds on which a government decision may remain non-public: national security and defence interests, ongoing international negotiations, ongoing law enforcement proceedings, and strictly time-bound financial operations capable of influencing the market. The same act states the prohibition too. No non-public government decision may be adopted on an individual subsidy, capital injection, asset transfer, loan, bond subscription or guarantee for a business company, foundation, association, church legal person or other organisation. This is not a new principle but the consistent application of Article 39 of the Fundamental Law (see 6.4.1): if data on public funds are of public interest, then a decision on awarding public funds cannot be hidden either. In Raz’s framework (see 6.4.2) this means that the individual decision remains in the Government’s toolkit, but within limits bound by a public general rule. The draft act is also worth submitting to public, structured consultation under MIAK’s I9 programme point.

3.2 Prioritised publication of the full texts (first round within 90 days, the entire stock within 12 months)

The full texts of the 3,598 decisions need not be published all at once, but publication must be carried out on a schedule and with deadlines. In the first round, within 90 days, the decisions on financial matters should be published in full: subsidies, investment promotion, capital increases, property purchases, loans, bond subscriptions, asset transfers. These can be filtered out of the list with good approximation by automated search on their titles. In the second round, within 12 months, the entire stock should follow. If a decision remains closed after that, the confidentiality must be justified individually by reference to one of the grounds under 3.1, and the confidentiality must be given an expiry date. The release is the task of the Prime Minister’s Office. The personal data of those concerned — for example the names of private individuals appearing in a decision — must be handled in accordance with data protection rules at the time of publication, but this cannot justify withholding the financial content. This order builds on Rose-Ackerman’s argument (see 6.4.3): where an individual advantage arises, disclosure is the only boundary between the acceptable and the unacceptable — which is why that is where we must start.

3.3 A decision-by-decision, machine-readable register — for the current government too (with quarterly updates, from 1 January 2027)

The current PDF list is an enumeration of titles by year; a working register needs more than that. MIAK proposes that every 2000-series serial number issued — past and future alike — should have a machine-readable (CSV or JSON) record. It should contain the serial number, the date of adoption, the title, the responsible ministry, the legal basis for confidentiality, the order of magnitude of the financial impact and the type of beneficiary. It should also contain the status: non-public, published since, withdrawn, or the serial number was never issued. The latter is needed because in the published list — by MIAK’s own count — 491 serial numbers are missing from the continuous numbering (in detail in point 6.2). These gaps may have arisen for many reasons, even entirely harmless ones, but the list does not reveal which. The register must also be updated quarterly for the current government’s own non-public decisions adopted since May 2026. This is the direct application of MIAK’s A1 public money dashboard programme point and the D2 open data programme point.

The three measures are linked by a single principle: secrecy should be an exception that has a legal basis, a duration and a trace — and the awarding of public money should not be able to be such an exception. The proposal of 2 September established that secrecy may apply to the content, not to the existence of the decision. The current list has fulfilled this for the past. The next step is for the rule to be binding for the future as well, and to remove decisions on financial matters from the scope of what can be classified.

Part IV — Expected effects and risks

Dimension Expected effect Risk
Public finances The full range of individual subsidy decisions becomes public; beneficiaries and amounts can be checked retrospectively Because of the public form, some investment promotion negotiations may become more difficult if the partner requests confidentiality
Public administration Indicating the legal basis for confidentiality disciplines practice; the decision-by-decision register eliminates the obscurity in the numbering Reviewing the 3,598 decisions is considerable administrative work; with poor scheduling even the first round may slip
Legal system Because it is at statutory level, a later government cannot override the rule by its own decision By broad interpretation of the exhaustive grounds (e.g. everything being “national security”), the prohibition can be circumvented
Political competition The rule also applies to the current government, so it is not a tool against the predecessor If only past decisions are opened up, the step is spent as a one-sided political gesture

The most important question to weigh is investment promotion. It may be argued that during negotiations on a large corporate investment, the publicity of the subsidy amount plays into the hands of competitors or other countries. This may be true for the duration of the negotiations, but no longer once the decision has been made: an awarded subsidy is not a business secret but a use of public money. That is why MIAK allows the strictly time-bound exception under 3.1 for the negotiation phase, but not for the subsidy decision. The proposal becomes risky if, after the act is adopted, financial decisions migrate from the 2000-series numbering into other non-public forms — for example individual decisions at ministry level or internal resolutions of state-owned companies. This can only be prevented if the prohibition is tied not to the form of the decision but to its content: every individual subsidy decision by any state body, in whatever form it is taken, should be public.

Part V — Measurability and summary

5.1 What is worth following? (suggested KPIs)

MIAK proposes the following performance indicators (KPIs — Key Performance Indicators). These are suggested indicators, not government commitments.

  • The proportion of 2000-series decisions on financial matters whose full text is published within 90 days. It is worth taking 100 per cent as the target; anything within this round that remains closed must receive an individual, public justification.
  • The number of non-public decisions adopted by the current government each quarter, and within that the proportion marked with a legal basis. The indicator is worth something only if the number itself is public — this is the open question today.
  • The number of unexplained gaps in the serial numbers. It is worth reducing this from today’s 491 to zero within one year: for every serial number it should be known what happened to it.
  • The number of non-public government decisions on individual subsidies adopted after the act enters into force. The target is zero; any other value signals circumvention of the prohibition.

5.2 Summary

MIAK’s request is addressed to the current government. The publication of the list of titles is a substantive step forward, because it has made accessible data that could previously be obtained only piece by piece, through freedom of information requests and lawsuits. The list, however, describes only the past, and only at the level of titles. The next step is a bill stating that no secret government decision may be taken on individual subsidies, and a timetable under which the full text of decisions on financial matters appears within 90 days. And MIAK asks the public not to stop at the individual juicy titles: the most important message of the list is the time series, not any single item.

Of MIAK’s foundational values, transparency and data-drivenness move together here. Transparency means that the existence and content of decisions can be known. Data-drivenness means that the knowable data are available in an analysable form. The current publication has taken a big step on the first, but still a small one on the second: a yearly PDF list is readable, but hard to analyse. The 491 unexplained serial numbers became visible only because someone counted through the list. In a well-built register this question would not even arise.

Part VI — Reasoning and further sources

6.1 Press framing by spectrum

The left-liberal and public affairs band carried the topic, but with two different emphases. Telex and 24.hu largely followed the structure of the government announcement: they highlighted the total of 3,598, the 2021 peak and the rise in 2014. Telex also added its own selection. It pointed out that from 2020 onwards individual subsidies for private companies and foundations were also decided in this form, and cited as an example the subsidy granted to the Samsung factory in Göd. It also touched on the decisions related to the military engagement in Chad. HVG’s framing is more critical and more historical. According to the paper, the list of titles is “not worth much on its own”, because the editorial office had previously requested the annual numbers and published them regularly. The article presents the four types of decision and the discontinuation of the Határozatok Tára in 2008, as well as the cases whose 2000-series decisions HVG had earlier obtained through litigation. 444.hu worked with a subjective selection by year, and focused on decisions concerning businesses close to the government, sports organisations and foundations. It is also there that one can read that non-public decisions concerning large companies appeared as early as 2013.

The conservative band did not raise the topic among the highlighted topics in MIAK’s press monitor on this day. From MIAK’s point of view this is interesting because the proposed rules — especially the quarterly register under 3.3, which also applies to the current government — address precisely the objection that the publication only reflects on the previous government. Another Telex piece is indirectly linked to the topic: minister Viktória Lőrincz filed a criminal complaint in the case of a 1.8 billion forint subsidy granted to a municipally owned company. Former minister Tibor Navracsics responded that the subsidy had been decided by public government decision 1380/2025, published in the Hungarian Gazette. The case illustrates the value of publicity well: with a public decision, the debate can be about the content of the decision, not about whether it existed at all.

6.2 Facts and data

The table below is based on MIAK’s count of the yearly lists published on kormany.hu. The sum of the “decisions in the list” column is exactly 3,598, matching the government’s announcement. The “highest serial number” is the largest 2000-series serial number in the list; the “missing serial numbers” is the number of numbers from the continuous numbering that do not appear in the list.

Year Decisions in the list Highest serial number Missing serial numbers
2010 21 2021 0
2011 19 2019 0
2012 10 2010 0
2013 21 2021 0
2014 33 2036 3
2015 43 2047 4
2016 74 2081 7
2017 117 2122 5
2018 142 2155 13
2019 187 2215 28
2020 385 2445 60
2021 679 2882 203
2022 470 2503 33
2023 635 2678 43
2024 273 2319 46
2025 419 2452 33
2026 70 2083 13
Total 3598 — 491

Three observations follow from the data:

  • The real break is 2020, not 2014. The rise in 2014 (from 21 to 33) was small. The year 2020, however, doubled the number in a single year (from 187 to 385), and in 2021 it rose to 679. The seven years between 2020 and 2026 produced 81 per cent of the 3,598 decisions (2,931 of them).
  • The content has also shifted towards financial decisions. The Hungarian word stem “támogat” (subsidy/support) occurs twice in the titles of the 2014 list, 125 times in 2021, 246 times in 2023 and 191 times in 2025 (MIAK word count on the lists of titles; a decision’s title may contain several occurrences, so this is an indication of direction, not a count of decisions).
  • The gaps appear from 2014 and peak in 2021. Between 2010 and 2013 the numbering is continuous; in 2021, 203 serial numbers are missing from the list. For 2015, HVG wrote of 47 non-public decisions, while the list contains 43 — the difference is exactly the missing serial numbers. The published list does not explain the reason for the gaps.

The announcement’s wording of “10–20 a year” is a rounding for the years 2010–2013: the list contains 21 items in 2010 and 2013, 19 in 2011 and 10 in 2012. The 2010 list also includes a few decisions from the beginning of the year, adopted before the change of government (for example a decision related to a NATO crisis management exercise in March 2010). HVG’s earlier series, based on freedom of information requests, showed 272 decisions for 2024, 419 for 2025 and 70 for 2026 up to the change of government, which matches the list with a difference of one item.

6.3 Policy dimensions

  • Transparency and anti-corruption policy (programme points) — machine-readable, public disclosure of the use of public money (programme point ID: A1) and strengthening the control function of constitutional institutions (programme point ID: A6);
  • Public administration and e-government (background material) — the order of government decision-making, the types and registration of government decisions; the decision-by-decision register is an implementation task of the Prime Minister’s Office;
  • Justice (programme points) — preparing the statutory rule through public consultation (programme point ID: I9);
  • Economy (programme points) — public justification and ex-post evaluation of economic decisions (programme point ID: G19);
  • Digitalisation and AI regulation (programme points) — machine-readable publication of public interest state data (programme point ID: D2).

6.4 Literature in detail

6.4.1 National Assembly: Fundamental Law of Hungary

The Fundamental Law directly concerns the question of non-public government decisions at two points. Article VI establishes as a fundamental right the right to access and disseminate data of public interest, and entrusts the supervision of its enforcement to an independent authority. Article 39 is more concrete. Under paragraph (1), support from the central budget may be granted only to an organisation whose ownership structure, organisation and activities aimed at using the support are transparent. Paragraph (2) reads as follows:

“Every organisation managing public funds shall be obliged to publicly account for its management of public funds. […] Data relating to public funds and to national assets shall be data of public interest.”

The reasoning of the Fundamental Law adds that transparency enables verifiability by the state bodies charged with this task and by the public. In the case of 2000-series decisions this creates a tension within the legal system: at the constitutional level, data on the awarding of public money are of public interest, whereas under government practice the decision itself can be locked away for twenty years. Access to data of public interest may be restricted by law, so this is not necessarily unconstitutional — but it shows that the exception has become the rule. The statutory prohibition under 3.1 resolves this tension in favour of the constitutional text.

📖 Source: National Assembly: Fundamental Law of Hungary

6.4.2 Joseph Raz: The Authority of Law

Raz describes the rule of law not as a moral ideal but as a formal virtue of law: law must be capable of guiding those to whom it applies. From this he derives a series of principles, two of which apply directly to the present topic. The first is publicity: law must be open and adequately publicised, because if it is to guide people, they must be able to know it. The second concerns particular acts:

“The making of particular laws (particular legal orders) should be guided by open, stable, clear, and general rules.”

Raz stresses that it is impossible to govern without particular decisions, and that this is not even desirable. The rule of law does not prohibit the individual decision, but requires that it be bounded by the framework of more durable, general, public rules. In the Hungarian practice of 2000-series decisions this framework is missing. The grounds for confidentiality are not set out publicly and exhaustively, so it cannot be verified from outside why a given decision is not public. The exhaustive statutory grounds under 3.1 would create exactly this missing general rule: the individual decision may remain closed, but only on the basis of a rule known to everyone.

📖 Source: Joseph Raz: The Authority of Law

6.4.3 Susan Rose-Ackerman: Corruption and Government

Rose-Ackerman’s book touches in two places on the question that emerges from the time series of the current list: what happens when closed, individual decisions are made on state subsidies. In the part on anti-corruption reform strategies she writes:

“Structures that allow officials unfettered discretion are particularly likely sources of payoffs, especially if citizens and firms have no recourse.”

And in the chapter on political finance she points out that the assessment of the relationship between firms and politics depends on how far politicians are able to grant individual favours to particular companies. If such favours are not prohibited or otherwise controlled, the line between legal and illegitimate advantage is drawn primarily by reporting obligations and the reaction of voters. The time series of the Hungarian list is a textbook example of this situation. After 2020 the non-public form increasingly served individual financial decisions, and these were attached neither to tender criteria nor to disclosure obligations. According to Rose-Ackerman’s logic, this is why the publication of the full texts (3.2) must begin with the decisions on subsidies: that is where the room for discretion is greatest, and where a counterweight is most lacking.

📖 Source: Susan Rose-Ackerman: Corruption and Government — Causes, Consequences, and Reform

6.5 International comparison

In European practice, the publicity of government documents is typically the rule, and confidentiality an exception listed in law. In Sweden the principle of public access to official documents has had constitutional rank since the 18th century, and classification is possible only on grounds specifically listed in the secrecy act. In Finland the act on the openness of government activities likewise makes publicity the default state, and sets statutory time limits for secret documents. Regulation (EC) No 1049/2001, which governs access to the documents of the institutions of the European Union, lists the grounds for refusal in a closed list and limits the application of the exceptions in time. The common element is that the ground for confidentiality depends not on the discretion of the decision-maker but on a public statutory list. This is the structure that MIAK’s point 3.1 would apply to Hungarian government decisions. On the side of government subsidies, EU state aid rules in any case require the data of individual aid above certain thresholds to be entered in a public EU register. A non-public government decision is therefore hard to reconcile with the principle of publicity that EU competition law expects of the member states.

Transparency and anti-corruption policy

  • A1 — Public money dashboard
  • A6 — Strengthening checks and balances

Justice

  • I9 — Popular-sovereignty audit — legislation with citizen feedback

Economy

  • G19 — Radical transparency in economic decision-making

Digitalisation and AI regulation

  • D2 — Open data programme

Proposed new programme point: Publicity obligation for individual subsidy decisions and exhaustive statutory grounds for non-public government decisions — for the Transparency and anti-corruption policy area (as a supplement to the programme point on an itemised register and expiry period proposed in the analysis of 2 September 2026).

6.7 List of sources

Press sources (MIAK press monitor, 30 September 2026 — topic 2):

Knowledge base references (books):

  • 📖 National Assembly: Fundamental Law of Hungary
  • 📖 Joseph Raz: The Authority of Law
  • 📖 Susan Rose-Ackerman: Corruption and Government — Causes, Consequences, and Reform

MIAK internal materials:

Supplementary public data sources:

  • Government of Hungary: Kormánydöntések 2010–2026 (Government decisions 2010–2026) — yearly lists of non-public government decisions — https://kormany.hu/kormanydontesek-20102026
  • Hungarian Gazette (Magyar Közlöny) archive — serial numbers of public government decisions
  • Regulation (EC) No 1049/2001 of the European Parliament and of the Council regarding public access to institutional documents

Generation metadata