Employment & labour market

In-depth analysis

FO1 — Employment data platform

• Mechanism: The platform integrates three data streams in real time: (1) employment filings from NAV — the Hungarian National Tax and Customs Administration (anonymised, district-level aggregate); (2) jobseeker registrations and vacancies from NFSZ — the National Employment Service; (3) demand data scraped from online job portals (Profession.hu, LinkedIn, Indeed) via NLP analysis. Using a machine-learning model, the platform produces 6-12 month employment forecasts at district level, by skill category. The forecast feeds the design of retraining programmes (FO3).
• Quantified target: By 2028, a monthly-updated labour-market dashboard for each of the 174 districts; forecast accuracy >75% (on a 6-month horizon); a 30% reduction in the mismatch between training supply and labour-market demand.
• International precedent: Singapore — the SkillsFuture system links labour-market demand data directly to training supply and gives workers individual “skills credits” that can only be redeemed for training in in-demand fields. The result: training participation doubled in 5 years and relevance improved.
• Trade-off / risk: Online job-advertisement data are biased: highly qualified, formal-sector roles are over-represented, while rural, informal and manual jobs are under-represented. As a result the model produces less accurate forecasts for disadvantaged districts — compensating data sources (manual data collection by NFSZ local offices) are required.

FO2 — Data-driven overhaul of the public-works scheme

• Mechanism: A gradual transformation of the public-works scheme (~80-100,000 people) in three phases: (1) Impact assessment: measuring at district level what percentage of public-works participants moved on to the primary labour market within 1 and 3 years (current estimate: <10%). (2) Raising the share of “productive public works”: converting public works into genuine work performed for municipal services and local cooperatives (e.g. renovating municipal buildings, forestry, social services). (3) Exit incentives: a 6-month “transitional wage supplement” (50% of the gap between the public-works wage and the market minimum wage) for those moving into the private sector.
• Quantified target: The share of public-works participants moving on to the primary labour market rises from <10% to 25% by 2030; the number of public-works participants falls from 80,000 to 40,000, with the remaining programme focused on productive activities.
• International precedent: India — MGNREGA (Mahatma Gandhi National Rural Employment Guarantee Act) is the world’s largest public-works programme (~50 million households). Evaluations are mixed: the poverty-reducing effect is measurable, but the labour-market integration effect is weak. The lesson: public works can serve as a social safety net but are not effective as an employment-policy instrument — the two functions must be explicitly separated.
• Trade-off / risk: Dismantling the public-works scheme means short-term income losses for the municipalities concerned, especially where the local government is the largest “employer”. Rapid downsizing may cause a social crisis — gradualism and the simultaneous build-up of a market alternative are critical. Politically, public works are the power base of local mayors, so transformation meets resistance.

FO3 — Targeted retraining programme

• Mechanism: A regionally differentiated retraining programme whose content is determined by the demand forecast of the employment data platform (FO1). Training is modular: 3-6 month intensive vocational modules + 1-2 months of workplace practice. During training, participants receive a living allowance (80% of the minimum wage). Training providers are incentivised by outcome-based financing: 50% of the fee is paid at training completion and 50% six months after training if the participant is employed. Priority fields: healthcare (nurses, assistants), IT (basic programming, systems administration), construction (energy-efficiency upgrades), social care.
• Quantified target: 20,000 retrained workers per year; 60% of those completing training are employed within 6 months; a 10% annual reduction in the mismatch between training content and local demand.
• International precedent: Denmark — the essence of the “flexicurity” model is that easy dismissal is combined with high unemployment benefits and intensive retraining. Active labour market policy (ALMP) spending is ~2% of GDP (EU average: ~0.5%). Result: one of the lowest long-term unemployment rates in the EU, but the system is extremely expensive.
• Trade-off / risk: Outcome-based financing encourages “creaming”: training providers select the most easily placed participants to improve their metrics and filter out the most disadvantaged. The solution: differentiated fees, with a higher payment for placing a disadvantaged participant.

FO4 — Measuring labour-market discrimination

• Mechanism: Annual paired audit studies (correspondence testing) in three areas: (1) job applications (submitting fictitious CVs with identical qualifications but Roma/non-Roma names, older/younger, male/female); (2) housing rentals (similar method); (3) credit applications (mystery shopping at banks). The research is carried out by an independent academic institute (MTA TK / CEU) and the results are published in an annual “Discrimination Report”. The report does not impose sanctions but creates a measurable baseline.
• Quantified target: By 2028, the annual discrimination audit covers the 10 largest cities; the “callback gap” for Roma-origin applicants (the difference in call-back rates) is measurable and public; a 5-year reduction target for the gap: −20%.
• International precedent: Sweden — since 2006, regular correspondence-testing studies have shown that applicants with Arab names receive ~50% lower call-back rates. Publication of the results triggered public debate and led to the pilot introduction of “anonymous applications” — with mixed results.
• Trade-off / risk: Publication of discrimination audits is politically sensitive, especially on the ethnic dimension. “Anonymous application” is an intuitive remedy, but the Swedish experience shows that discrimination reappears in later phases of the selection process (interview) — the structural problem cannot be solved by a single technical tool.

FO5 — Social economy and cooperatives

• Mechanism: Three pillars: (1) Legal framework: statutory regulation of the legal status of social enterprises (currently there is no dedicated legal form in Hungary) — defined social purpose, restricted profit distribution, democratic governance. (2) Financial support: a dedicated social-enterprise fund (5 Bn HUF per year) providing subsidised loans and guarantee-fee support. (3) Market creation: amendment of the Public Procurement Act — reserved public procurement procedures for social enterprises (the EU Public Procurement Directive allows this).
• Quantified target: By 2032, 500 registered social enterprises (current: ~50-100 informally); 5,000 people employed in the social economy; the share of reserved public procurement reaches 3%.
• International precedent: Italy — Law 381 of 1991 (cooperativa sociale) created two types of social cooperatives: type A (provision of social services) and type B (employing disadvantaged people). Italy today has ~15,000 social cooperatives with ~400,000 employees. Key to the success: the social cooperative received an independent, clearly defined legal status rather than being subsumed under traditional cooperative law.
• Trade-off / risk: Social enterprises are prone to dependency on state support: if market revenue is insufficient, the enterprise remains permanently subsidy-dependent. In the Italian model, around 30% of type B cooperatives struggle with sustainability. The solution: tapered support (subsidy decreasing year by year) and a minimum market-revenue ratio (≥50% of revenue from market sources from year 3 onwards).

FO6 — Demographic response strategy

• Mechanism: The EU’s working-age population (20-64) is falling from its peak (272M, 2009) to 258M by 2030 — Hungary shows a similar trend, and further increases in the employment rate are limited. Four pillars: (1) Managed labour migration: a list of shortage occupations (IT, healthcare, construction) and a targeted visa programme for workers — attracting 20,000+ workers per year. (2) Raising female labour-market participation: 30% expansion of nursery places, incentivising flexible working hours, supporting return after parental leave (6-month “reintegration mentoring”). (3) Active ageing: increasing healthy life years, making work more flexible at age 60+ (part-time work, mentor roles), flexibilising the retirement age (not raising it but incentivising). (4) Nursery and kindergarten capacity expansion: the institutional care rate for the 0-3 age group rises from 17% to 33% (EU Barcelona target).
• Quantified target: The female employment rate (ages 25-54) rises from 78% to 85% by 2032; managed labour migration brings in 20,000+ shortage-occupation workers per year; the employment rate of the 60-64 age group rises from 40% to 55%; the number of nursery places rises from 50,000 to 70,000.
• International precedent: Canada’s Express Entry system: a points-based labour-migration programme aligning immigration with economic needs — 300,000+ economic immigrants per year, 90% of whom are employed within 6 months. Sweden: the childcare system (förskola) is universal, the female employment rate is 80%+ — the link between nursery capacity and labour-market participation is empirically established. Japan: the third pillar of “Abenomics” was “womenomics” — the female employment rate rose from 60% to 72% between 2013 and 2023.
• Trade-off / risk: Labour migration is politically sensitive, especially in Hungary. The data-driven approach helps: not an ideological debate, but based on labour-market data — we attract labour to where there are shortages. Raising female labour-market participation is only successful if household burdens (childcare, elderly care) are also redistributed — this also requires cultural change, which policy can only encourage, not force. 📖 Source: EC: 2025 Euro Area Report (IP304)

FO7 — Countercyclical employment fund

• Mechanism: Keynes: unemployment is caused by a lack of effective demand. Countercyclical employment fund: GDP-linked employment spending rises automatically in recession (green investments, infrastructure, community development) and falls in upswings. Maximising the multiplier effect: spending should reach the strata with the highest propensity to consume.
• Quantified target: The fund’s size is 0.5-1% of GDP; it activates within 6 months of a recession; measurement and public reporting of the multiplier effect; 50,000+ transitional jobs during recessionary periods. 📖 Source: Keynes: The General Theory (chapters 10, 24)

FO8 — Human capital investment framework (vocational training guarantee)

• Mechanism: Smith: “the acquired and useful abilities of the workman are capital, just as a machine is” — spending on training people is investment. Lee Kuan Yew: the key to Singapore’s success was targeted workforce training. Every Hungarian citizen aged 18-55 is guaranteed one retraining opportunity per year (40-120 hour modules). Outcome-based financing: the training provider receives full payment only if the participant enters employment within 6 months.
• Quantified target: 100,000+ retraining participations per year; a 6-month placement rate of 70%+; the ROI of training expenditure is publicly measured. 📖 Source: Smith: The Wealth of Nations (II/1); Lee Kuan Yew: From Third World to First

FO9 — Productivity movement — workplace innovation incentives

• Mechanism: Lee Kuan Yew: the Japanese-style productivity movement and quality control circles (QCC) raised productivity in Singapore. National Productivity Programme: a workplace innovation-proposal system for SMEs, with a state award and a tax allowance for implemented improvements. Onboarding 5,000 enterprises per year.
• Quantified target: 5,000 enrolled enterprises per year; 10,000+ implemented improvement proposals per year; participating SMEs’ productivity rises measurably by 5%+. 📖 Source: Lee Kuan Yew: From Third World to First; Smith: The Wealth of Nations (I/1)

FO10 — Wage-bargaining modernisation — tripartite wage forum

• Mechanism: Lee Kuan Yew: Singapore created the National Wages Council in 1972 — it issues annual wage guidelines tied to productivity data. Smith: “the wages of labour rise with the wealth of the nation.” National Wage Review Forum: annual wage guidelines based on productivity and cost-of-living data, with sector-differentiated recommendations.
• Quantified target: Publication of annual wage guidelines; the real-wage/productivity gap narrows to below 2 percentage points; the number of labour disputes falls by 15%. 📖 Source: Lee Kuan Yew: From Third World to First; Smith: The Wealth of Nations (I/8)

FO11 — Self-sufficiency incentives — preventing welfare dependency

• Mechanism: Lee Kuan Yew: “when the government took over the basic responsibilities of the head of the family, people’s motivation weakened.” “Exit incentives” built into the social-benefits system: tapered support (gradual reduction as income rises, rather than abrupt loss), a 12-month work-entry bonus, and individual savings incentives.
• Quantified target: The share of those moving from social benefits into work rises by 20%; tapered support fully replaces abrupt loss (100%); the work-entry bonus reaches 20,000+ people per year. 📖 Source: Lee Kuan Yew: From Third World to First; Keynes: The General Theory (chapter 24)

FO12 — AI labour-market impact analysis and preparedness programme

• Mechanism: The OECD’s 2026 Economic Outlook showed that AI investments are already generating measurable productivity effects, but this also means jobs are being transformed. Programme: (1) A biannual “AI Labour-Market Impact Report”: sectoral and occupational analysis — which jobs can be automated, which are transformed, which are created? (2) Expanding the FO1 (Employment data platform) with AI-impact indicators. (3) A targeted “AI retraining programme”: in the most affected jobs (data entry, bookkeeping, customer service, manufacturing inspection), proactive retraining for AI-augmented work. (4) An AI skills-development loan for employers who provide AI training to their staff.
• Quantified target: A biannual AI Impact Report from 2027; 10,000+ AI retrainings per year; 50% of workers in the roles most threatened by AI are retrained within 3 years; 1,000+ enterprises per year take up the AI skills-development loan.
• Trade-off / risk: Forecasting AI’s labour-market impact is uncertain — projections often over- or under-estimate. Flexibility is key: the programme is adaptive, not deterministic. The “AI is taking your job” narrative can cause unnecessary panic — communication should focus on transformation, not disappearance. 📖 Source: OECD: Economic Outlook, Interim Report — Testing Resilience (March 2026)

FO13 — Labour-market programme for the trade realignment

• Mechanism: The IMF’s WEO 2025 documented that tariff wars and the trade realignment have pushed tariffs “to levels not seen in a century” — the impact is uneven, with some sectors and regions disproportionately affected. The return of industrial policy brings new employment patterns: building local supply chains creates jobs, but workers in previously export-oriented sectors may be pushed out. Hungarian programme: (1) A “Trade Realignment Impact Report” — an annual analysis of which Hungarian sectors are most affected by the global trade realignment (automotive, electronics, agriculture). (2) Targeted retraining: an FO3-style programme for workers in affected sectors, tailored to the specific skill needs arising from the trade realignment. (3) Support for regional mobility: relocation grants, remote-work incentives, so that workers do not remain stuck in sectorally declining regions.
• Quantified target: An annual Trade Realignment Impact Report from 2027; 40% of workers in affected sectors are retrained within 3 years; 5,000+ workers per year take up regional mobility support.
• Trade-off / risk: The direction of the trade realignment is uncertain — geopolitics is shifting rapidly. The programme must not turn into protectionist industrial policy; data-driven targeting is key. 📖 Source: IMF: World Economic Outlook 2025 — Global Economy in Flux, Prospects Remain Dim (chapter 3, Box 1.1)