Part I — Situation overview

On 12 September 2026 a 220-page report on the state of the child protection system appeared on the government’s website. The document was presented by Vilmos Kátai-Németh, Minister for Social and Family Affairs, Szilvia Gyurkó, State Secretary for Children’s Rights and Child Protection, and Prime Minister Péter Magyar at a press conference held at the Kossuth Lajos Children’s Home in Budapest. The diagnosis, prepared between May and August 2026, is based on on-site visits to 19 institutions or professional units in 12 counties and on interviews with some 500 professionals. Its main figures: nearly 8 per cent of children living in Hungary, almost 130,000 children, are registered as at risk; including aftercare, some 23,000 children and young people live in state care; roughly 2,700 children live in an unsuitable placement or are waiting in an endangering family; on average 17,000 runaway incidents occur every year, and some 800 children are on the run long-term. On behalf of the government the prime minister apologised and announced an immediate crisis programme of 9.5 billion forints, which creates a total of 184 new or repurposed places. He also said that the government would submit a child protection legislative package to Parliament the following week. The law is adopted by Parliament; in this form the announcement is an intention to submit, not a change in the law in force.

The report has a history. MIAK’s analysis of 19 August 2026 dealt with the reports of the children’s home inspections ordered in 2024, which were released only after two court judgments, and at that point the ministry promised the first part of the present diagnosis for 31 August. The present document is thus a substantive step towards data publicity: not a forced but a voluntary publication, and system-level rather than institution-level. At the same press conference two matters unrelated to the policy content were also raised. The prime minister announced that the government intends to amend the rules restricting adoption by same-sex couples; this is a question at the level of an act, so Parliament will also decide on that change. The 2016 state decoration of the director of the Bicske children’s home, who was later convicted with final effect, also came up; according to the documents made public, the nomination chain ran from the then head of the Directorate-General for Social Affairs and Child Protection through the Minister of Human Capacities to the prime minister. State decorations are conferred by the President of the Republic, on the government’s proposal.

MIAK’s reading: the crisis programme is a capacity response, whereas the report describes a system failure. The diagnosis itself states that the biggest problem is not the shortage of places but the shortage of professionals: in institutions maintained by the Directorate-General for Social Affairs and Child Protection the vacancy rate was 17.2 per cent at the end of 2025, and professional turnover exceeded 24 per cent. If 184 new places open but the staff shortage does not decrease, the new places will operate with the same shortage. The question is therefore not whether 9.5 billion is enough, but whether the legislative package to be submitted fixes a yardstick from which it will be clear in a year’s time whether the system has changed — or merely grown.

Part II — Foundations in the literature

Three sources provide the frame in which the relationship between the report and the crisis programme can be interpreted. The 2025 policy brief on vulnerable children by the European Observatory on Health Systems and Policies, which works alongside the WHO Regional Office for Europe, examines the practice of five European countries. Its conclusion is that data collected on vulnerability does not in itself bring change: it becomes action when it is regularly linked to the evaluation of programmes. The World Bank’s World Development Report 2015, the founding document of the behavioural turn in development policy, offers two theses: prolonged stress in early childhood causes lasting damage to the developing brain, and good policy is not a one-off design exercise but a continuous learning process built on feedback. The European Pillar of Social Rights, proclaimed by the European Parliament, the Council and the Commission in Gothenburg in 2017, establishes in its principle 11 that children have the right to protection from poverty, and children from disadvantaged backgrounds have the right to specific measures to enhance equal opportunities. This is the yardstick to which the domestic indicator system can be tied. The detailed treatment of the literature — by author, with quotations — is in section 6.4 Literature in detail.

Part III — MIAK’s concrete proposal

MIAK proposes three measurable measures for the legislative package to be submitted. None of them slows down the crisis programme; all three ensure that the programme’s results can be judged afterwards.

3.1 A professional staffing indicator broken down by territory (fixed in the legislative package, first publication by 31 March 2027)

MIAK proposes that the legislative package require quarterly publication of child protection staffing, broken down by county and by maintaining body, with at least four data points: the number of children cared for per professional by form of care, the vacancy rate, professional turnover, and the share of staff receiving supervision. The breakdown should also cover church and foundation maintainers: according to the report, part of the foster-parent networks was transferred to non-state maintainers without the state requiring professional transparency, and today the ministry is “blind” over a significant part of this area. The rationale follows from the structure of the crisis programme itself: the impact of the 1.5 billion forints earmarked for supporting professionals and of the announced two-stage pay settlement can be measured if the baseline values of turnover and shortage are public. The European Observatory’s brief (see 6.4.1) names exactly the error this proposal eliminates: a great deal of data is produced, but without an action plan. The proposal applies the efficiency-measurement principle of KI8 to child protection, and makes the targeting of the KI6 public-service pay system verifiable.

3.2 Tracking young people who have left care between the ages of 18 and 25 (built by the end of 2027)

One of the report’s gravest findings is that the system reproduces itself: often the children and grandchildren of young people who grew up in child protection come back into it. Citing a Budapest report, Mayor Gergely Karácsony stated that there is an area where 90 per cent of street drug users came out of child protection institutions. Nobody can currently confirm or refute this figure nationally, because the system does not track what happens to young people who have left care. MIAK proposes an annual statement, in aggregated form unsuitable for personal identification, on young people who have left care after the age of 18, in five dimensions: housing, employment or education, entry into healthcare (including addiction treatment), involvement in crime, and child protection involvement of their own children. The data can be produced by linking existing administrative registers, with a data protection impact assessment. This extends the logic of the KB3 prevention-versus-punishment impact assessment and of the DM2 family-support impact assessment to the group for which the cost of failed prevention is highest. The World Bank’s thesis on prolonged early-childhood stress (see 6.4.2) explains why this life course is no accident.

3.3 A child protection diagnosis repeated every year with the same methodology, including the processing time for babies stuck in hospital (first repetition September 2027)

The present report becomes an institutional turning point if it does not remain a one-off. MIAK proposes that the legislative package require the diagnosis to be repeated and published annually with the same methodology and with core indicators fixed in advance. These should include: the number of children at risk and of children in state care, the number living in unsuitable placements, the number on the run long-term, the share returned to their birth families, the average waiting time for child psychiatric care, and the number of babies stuck in hospital together with the time elapsed until their placement. For the latter the government has already given a yardstick: according to the announcement, the requirement for recognising the situation will be cut to eight days. The report’s methodology (on-site visits, professional interviews, national data) can be repeated the following year, and thus becomes comparable. The proposal applies the annual, indicator-based cycle of the SZ12 national action plan to child protection, tied to Gothenburg principle 11 (see 6.4.3).

The three proposals are bound together by one principle: the place is the system’s input; the child’s fate is its output. The crisis programme expands the input, and this is needed. But the report shows that for years the system knew about its problems and did not change; what makes the difference now is whether the output becomes measurable and public. According to the World Bank’s thesis on adaptive design, the real failure is the one from which no learning comes — and that requires data comparable from year to year.

Part IV — Expected effects and risks

Dimension Expected effect Risk
Situation of children Annual measurement of runaways, family reunification and time spent in hospital shows where the child’s path gets stuck; tracking young people who have left care provides a basis for targeting aftercare Formal fulfilment of indicators: long-term running away may “decrease” through statistical reclassification without real improvement
Social care system The staffing indicator steers the pay settlement and the supervision programme to where turnover is highest Public territorial data may in the short term worsen the recruitment prospects of the weakest counties
Budget With an annual cost of care of around 8 million forints per child, the return on prevention and family reunification becomes quantifiable Measurement itself ties up funds and professionals in a system where both are lacking
Public security and healthcare Tracking young people who have left care makes the risk of drug use and crime manageable from the prevention side Data protection risk and stigmatisation if individuals can be identified from the linked data

The main trade-off lies between speed and measurability. The government justified the crisis programme by saying there is no time to wait for next year’s budget — which is justified from the perspective of children in residential care. The risk arises if rapid capacity expansion crowds out building the measurement, and delivering the 184 places becomes the only proof of success. According to the report, the number of children removed from their families rose by 17 per cent between 2010 and 2025, and the recent stalling of admissions signals not improvement but the system becoming clogged: in this situation every new place fills up quickly. The proposal tips to the risk side if the indicator system is used to evaluate and sanction maintainers; the purpose of the indicators is to target resources, not to punish the weakest institutions.

Part V — Measurability and summary

5.1 What is worth following? (suggested KPIs)

MIAK proposes the following performance indicators (KPIs, in English: Key Performance Indicator) for monitoring. These are proposals, not government commitments:

  • Professional turnover and vacancies: according to the report the baseline is turnover above 24 per cent and a vacancy rate of 17.2 per cent (end of 2025, state maintenance). Suggested target: a decrease in both indicators 12 months after the first stage of the pay settlement, also in the county breakdown.
  • Number of children on the run long-term: baseline around 800. It is worth following whether it falls after the opening of the 80 new crisis-care places, and after how many runaway incidents repeat running away does not occur.
  • Number of babies stuck in hospital and their placement time: the figure given at the press conference was 358. Suggested yardstick: the compliance rate with the announced eight-day recognition deadline and the average time until placement.
  • Situation of 18–25-year-olds who have left care: there is no baseline data today. Suggested target: by the end of 2027 the first national aggregated statement on housing, employment, and health and criminal involvement should be completed.
  • Repetition of the diagnosis: the 2027 report should appear with the same methodology and the same core indicators, and include a comparison with the 2026 values.

5.2 Summary

MIAK’s request for the legislative package to be submitted, in a single sentence: alongside the 9.5 billion forints, the law should also contain how it will be known in a year’s time whether it worked. Concretely: a quarterly staffing indicator broken down by territory, aggregated tracking of young people who have left care up to the age of 25, and a diagnosis repeated annually with the same methodology, including the processing time for babies stuck in hospital. MIAK asks the Ministry of Social and Family Affairs to include these indicators with baseline values in the concept promised for the end of the year; and asks Parliament not to leave the measurement obligation to an implementing decree in the debate on the bill.

Two MIAK foundational values are at stake. Transparency, because publishing the report is a real turning point after the inspection reports kept secret for two years — but transparency becomes institutional when it depends not on the decision of the government of the day but is required by law year after year. And data-drivenness, because the diagnosis itself shows that the system did not deteriorate for lack of data: the problems were known, they simply were not followed by measurable targets and mandatory feedback. The present step will be more than a well-meaning rescue package if the child’s fate becomes the yardstick, not the number of places.

Part VI — Reasoning and further sources

6.1 Press framing by spectrum

The left-liberal band chose two different emphases. Alongside the announcement and the interview, Telex foregrounded the internal professional perspective: from Szilvia Gyurkó’s studio interview it headlined that child protection is a “traumatised system”, and separately highlighted that according to the report only one in four or five children returns to their family — that is, the framing questioned the system’s output. HVG processed the details of the report item by item (shortage of places, placements 500 kilometres away, the different admission obligations of non-state maintainers), and in two accompanying articles pursued the threads of responsibility: Gergely Karácsony’s drug-related connection and the justification for the decoration of the Bicske director. 444.hu published the itemised content of the crisis programme, pursuing the political thread rather through a data request for the decoration document.

In the mainstream public-affairs band, 24.hu made the nomination chain for the decoration the backbone of its article, with the policy content pushed into the background. ATV, by contrast, carried the numerical findings of the diagnosis most fully — it was the only source read that reported the shrinking of the foster-parent network by 2,000 places, the rise in the number of children in children’s homes and the turnover figure. In the economic band Portfolio wrote twice: about the publication of the report and about the funding breakdown of the crisis programme, essentially in a budgetary frame.

The conservative band, through two Magyar Nemzet articles, put the government’s credibility and the adoption announcement at the centre. According to one article, instead of the immediate 25 per cent pay rise promised in the campaign, after four months staff received a report and another promise, and the paper also reported the statement of the Fidesz parliamentary group on the circumstances of an earlier interview with a little boy living in a child protection institution; the other article disputed research on adoption by same-sex couples. MIAK regards the gap between the pay-rise promise and the actual pay decision as a substantive, verifiable criticism: this is precisely why it proposes that the impact of the pay settlement be measurable through the turnover indicator. Népszava headlined the text of the decoration justification. Mandiner did not serve as a fresh source that day.

6.2 Facts and data

Datum Value Source
Publication of the report 12 September 2026, on the government’s website Telex, Portfolio, ATV, 12 September 2026
Length of the report 220 pages Portfolio, 444.hu, ATV, 12 September 2026
Research method May–August 2026; 12 counties, 19 institutions or professional units, some 500 professionals HVG, ATV, 12 September 2026
Children registered as at risk almost 130,000 (nearly 8 per cent of children) HVG, Telex, 12 September 2026
Living in state care (including aftercare) around 23,000 HVG, Portfolio, 12 September 2026
Children in unsuitable placements or waiting around 2,700 HVG, ATV, 12 September 2026
Runaway incidents per year / on the run long-term 17,000 on average / around 800 HVG, Telex, 12 September 2026
Children removed from their families per 10,000 minors 99 (2010) → 127 (2024) HVG, 12 September 2026
Share of children in state care Hungary 1.4%; EU average approx. 1%; Austria 0.8%; Sweden 0.66% HVG, 12 September 2026
Number of foster parents since 2023 400 fewer (loss of around 2,000 places) ATV, Telex, 12 September 2026
Children cared for in children’s homes rose from 5,518 to 6,197 ATV, 12 September 2026
Vacancies / professional turnover (state maintenance) 17.2% / above 24% (end of 2025) ATV, 12 September 2026
Babies stuck in hospital 358 Telex, 12 September 2026
Police complaints for child abuse 2,249 (2024–2025) Portfolio, 12 September 2026
Annual cost of state care for one child around 8 million forints (8–12 million according to the state secretary) HVG, Telex, 12 September 2026
Crisis programme HUF 9.5 billion: around 6 billion for professional programmes, 3.4–3.5 billion for development; 184 new or repurposed places; crisis care for 80 children; strengthening 15 special children’s homes for 120 children; HUF 440 million to support at-risk adolescent girls and young mothers; HUF 1.5 billion for professional and psychological support of professionals Portfolio, 444.hu, 12 September 2026
Announced legislative package submission to Parliament the following week; an 8-day recognition deadline in the cases of children stuck in hospital Telex, Portfolio, 12 September 2026

Two data points require a separate note. Portfolio’s detailed article reported the development part of the crisis programme as 3.5 billion forints, while 444.hu and Portfolio’s press-conference summary gave 3.4 billion; Telex wrote of a ten-billion programme, and in an interview the state secretary mentioned a further one billion earmarked for the situation of babies stuck in hospital. Several figures also appear for the rate of staff shortage: HVG gave 16 per cent for children’s homes, the press conference an average shortage of around 20 per cent, in places 60–65 per cent, and ATV a 17.2 per cent vacancy rate for state maintenance. The differences probably stem from different base populations — the exact values should be checked against the original text of the report, and this is precisely the rationale for proposal 3.1.

6.3 Policy dimensions

  • Social policy (programme points) — specialised child protection care as a field for implementing principle 11 of social rights; the annually repeated, indicator-based diagnosis fits into the action-plan cycle of SZ12, the minimum level of care belongs to SZ7, and the poverty-related causes of family risk to the frame of SZ10;
  • Demography (programme points) — according to the report, family-policy instruments do not reach the most vulnerable children; the DM2 impact assessment must measure this group separately;
  • Public security and law enforcement (programme points) — the drug-use and crime risk of young people who have left care, and the new procedure for abuse cases, connect to the prevention impact assessment of KB3 and the victim-protection minimum of KB8;
  • Justice (programme points) — the enforcement of rights of abused children, which the analysis of 19 August 2026 discussed in detail on the basis of I6;
  • Healthcare (programme points) — waiting times for child psychiatry and the prevention of early-childhood harm can be measured as part of the E4 prevention data programme;
  • Public administration and e-government (programme points) — measuring staffing and targeting the pay settlement within KI8 and KI6.

A clarification of competences: the report was prepared by the State Secretariat for Children’s Rights and Child Protection of the Ministry of Social and Family Affairs, the government decided on the crisis programme, and the government submits the legislative package, but only Parliament can adopt it. In abuse cases, investigation is the competence of the police, indictment of the prosecution service and adjudication of the courts — the announced “new procedure” can therefore relate to the reporting and cooperation protocol of child protection professionals, not to the rules of criminal procedure, which are likewise set at the level of an act.

6.4 Literature in detail

6.4.1 European Observatory: Improving reach and access to health promotion and preventive services for vulnerable children and adolescents

The 2025 policy brief of the European Observatory on Health Systems and Policies (authors: Eriksen, De Bock, Alayli and colleagues) examines the practice of Finland, Germany, the Netherlands, Portugal and England from the angle of how vulnerable children can be reached. It names three routes: strategic cooperation between the health, social and education sectors, reliable data collection and data sharing, and targeted, personalised outreach. For today’s topic the second is decisive. One of the document’s sharpest findings:

“Often, extensive data are collected to highlight health inequities, but without a clear and systematic plan for action. This means that the data risk falling short of driving meaningful change.”

The brief also warns that there is a significant lack of disaggregated data on vulnerable children, so their needs are difficult to estimate. Within this frame the Hungarian report is at once progress and a warning. Progress, because the 220-page diagnosis is the first system-level, public data survey. A warning, because the report itself records that the 2024 inspection reports also gave an accurate picture — and nothing followed from them. Proposals 3.1 and 3.3 therefore do not ask for further data collection, but for the data to be linked to a mandatory, recurring decision cycle.

📖 Source: European Observatory on Health Systems and Policies (Eriksen et al.): Improving reach and access to health promotion and preventive services for vulnerable children and adolescents (Policy Brief 67, 2025)

6.4.2 World Bank: World Development Report 2015

The World Bank’s 2015 development report contributes two independent theses to the topic. The first concerns early childhood development: according to the report, prolonged, unbuffered stress — when the body’s stress-response system is strongly and continuously activated — results in fewer neural connections in the areas of the developing brain responsible for learning and behaviour, and this later shows up in school performance and social relationships. This provides the professional rationale for proposal 3.2: the difficulties of young people who have left care are not a series of individual failures but a predictable consequence of early harm, which is why they must be tracked.

The second thesis concerns policy design. According to the report, the initial diagnosis of programmes is often incorrect or only partially correct, and this becomes clear only during implementation; this requires an institutional mechanism that provides effective feedback. In the report’s words:

“However, instead of penalizing failure or burying findings of failure, organizations need to recognize that the real failures are policy interventions in which learning from experience does not happen.”

This fits the Hungarian child protection system to the letter: according to the report, flagging problems brought field workers not support but reproach, and the 2024 findings never reached the point of decision. The annual diagnosis with the same methodology under proposal 3.3 is the institutional form of this learning cycle.

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

6.4.3 European Parliament, Council and Commission: European Pillar of Social Rights

The document jointly proclaimed in Gothenburg on 17 November 2017 sets out 20 principles of social rights. Principle 11 (childcare and support to children) states two rights: children have the right to affordable early childhood education and care of good quality, and to protection from poverty. The continuation of the second point is what matters for today’s topic:

“Children from disadvantaged backgrounds have the right to specific measures to enhance equal opportunities.”

The principle does not prescribe a specific institutional solution — it respects member-state competence — but it sets a yardstick: the effect of the specific measure owed to a disadvantaged child must be measurable, otherwise it cannot be determined whether the right is being realised. In the Hungarian situation this means that for the 130,000 children at risk and the 23,000 children in state care, the “specific measure” is not the amount of the crisis programme but whether the children’s prospects — family reunification, progress at school, the transition to adult life — actually converge with those of other children. The SZ12 programme point turns this principle into an indicator-based action plan updated every year; proposal 3.3 is its child protection application.

📖 Source: European Parliament, Council and Commission: European Pillar of Social Rights (Social Summit Booklet, Gothenburg, 2017)

6.5 International comparison

The most direct model is the Finnish practice described in the European Observatory’s brief. In Finland the national institute for health and welfare surveys some 260,000 children and adolescents every two years on well-being, health and learning, and the data are publicly available broken down by region, year and sex for municipal and regional planning. The Finnish government’s 2022 national action plan linked to the European Child Guarantee set the goal of extending the survey to children in vulnerable situations, including those living with foster parents. This is exactly the step missing from the Hungarian system today: a child in care is not a separate statistical island but appears in the same measurement as other children, so the gap in opportunities can be shown.

In England the public health data platform publishes a child and maternal health profile for every local authority, which, compared with national and regional averages, signals where intervention is needed. The Hungarian equivalent could be the county-level staffing indicator of proposal 3.1. At the same time, the EU comparison quoted by HVG from the report — 1.4 per cent of children in Hungary live in state care, against an EU average of roughly 1 per cent, 0.8 per cent in Austria and 0.66 per cent in Sweden — is a reminder that international comparison does not in itself pass judgement: a lower rate may come from stronger prevention, but also from insufficient detection. That is why, alongside the admission rate, an output indicator is needed — family reunification, running away and the fate of young people who have left care.

Social policy

  • SZ7 — “Dignity-based” social policy
  • SZ10 — The cognitive burden of poverty — behavioural-science social policy
  • SZ12 — National action plan of the European Pillar of Social Rights

Demography

  • DM2 — Impact assessment of family support

Public security and law enforcement

  • KB3 — Crime prevention vs. punishment impact assessment
  • KB8 — Victim-protection service minimum

Justice

  • I6 — Lowering the threshold of citizens’ enforcement of rights

Healthcare

  • E4 — Prevention data programme

Public administration and e-government

  • KI6 — Competitive public-service pay system
  • KI8 — Drucker-style efficiency measurement in public administration

Proposed new programme point: Child protection outcome indicator system and annually repeated diagnosis — for the Social policy area: core indicators fixed in law and published every year with the same methodology (staffing and turnover, long-term running away, family reunification, placement time of babies stuck in hospital), together with aggregated tracking of young people who have left care up to the age of 25.

6.7 List of sources

Press sources (MIAK press monitor, 13 September 2026 — topic 1):

Knowledge-base references (books and official documents):

  • 📖 European Observatory on Health Systems and Policies (Eriksen et al.): Improving reach and access to health promotion and preventive services for vulnerable children and adolescents (Policy Brief 67, 2025)
  • 📖 World Bank: World Development Report 2015 — Mind, Society, and Behavior
  • 📖 European Parliament, Council and Commission: European Pillar of Social Rights (Social Summit Booklet, Gothenburg, 2017)

MIAK internal materials:

  • MIAK policy area: Social policy (programme points; programme point ID: SZ7, SZ10, SZ12)
  • MIAK policy area: Demography (programme points; programme point ID: DM2)
  • MIAK policy area: Public security and law enforcement (programme points; programme point ID: KB3, KB8)
  • MIAK policy area: Justice (programme points; programme point ID: I6)
  • MIAK policy area: Healthcare (programme points; programme point ID: E4)
  • MIAK policy area: Public administration and e-government (programme points; programme point ID: KI6, KI8)
  • MIAK press monitor, 13 September 2026 — topic 1, score: 97/100

Supplementary public data sources:

  • The government’s child protection report (12 September 2026) — full text published on the government’s website
  • Hungarian Central Statistical Office (KSH) — child protection statistics
  • Children’s rights reports of the Commissioner for Fundamental Rights
  • Eurostat — share of children at risk of poverty or social exclusion

Generation metadata