156 billion forints for a programme where 3 in 4 never find a real job.

Rationale

This is the largest active-spending line in the chapter and the one where the gap between the seen and the unseen is widest. The seen is direct: roughly 67,000 people were in public employment in early 2026[^1], paid a public-works wage, doing locally-organised work in Hungary's poorest settlements. The unseen is the worker whose payroll levy funds the line, and — more sharply — the public-works participant whose years inside the programme make open-market employment *less* likely, not more. The mechanism is now well documented. Public works is a state-run parallel labour market. It does not respond to a price signal for labour; a settlement's mayor and the county government office decide how many places to open and what work they do. Without a wage set by what an employer would actually pay for the output, there is no signal that the activity creates value at its cost — and the programme's own evaluations show it frequently does not. Analysis published in the Hungarian labour-market literature found that long, high-hour public-works participation measurably *reduces* the probability of moving to the open labour market and *raises* the probability of cycling back into public works[^2]. The transition rate is the central number: of those in public works in 2012, only about 15% were working outside the programme a year later; by 2018, in a far stronger labour market, this had risen to roughly 25%[^3]. A programme that returns three in four participants to dependency, or to itself, is not a bridge into employment. It is a holding pattern that the levy-payer funds. Hungary's spending on this is an international outlier. Around 2016 Hungary directed roughly 0.7% of GDP to public works; the OECD average for comparable active-labour-market public-employment spending was about 0.05% — and among the very few countries at a comparable level, those that spend this share do so in far poorer economies on rural agricultural programmes rather than urban-adjacent public-works[^3]. The question this raises is not whether Hungary's poorest settlements need a labour-market policy. It is whether 156,000.0 millió Ft of payroll levy, administratively allocated, with a documented one-in-four exit rate, is that policy. An honest reform does not strand the people currently in the programme. The protected party is real: roughly 67,000 participants, many in settlements where the open labour market is thin and the public-works place is the household's only formal income. Abrupt removal would push them onto passive benefits or into the informal economy with no transition. The phase-out therefore runs five years, with the envelope redirected as it declines into two channels that the evidence supports better than parallel-market employment: direct wage subsidies tied to placement in actual private-sector jobs (the "Közfoglalkoztatásból a versenyszférába" placement benefit, currently 45,600 Ft monthly, is the existing instrument and the design template[^4]), and a needs-based income transfer for those in settlements where no open labour market realistically exists within the horizon. The first channel pays for results an employer has validated with a real job offer; the second is honest about the fact that it is income support, not employment, rather than disguising a transfer as work. Neither requires the state to run a parallel labour market it cannot price.