Active employment policy, or passive policy of dependence
The Employment Strategy foresees 25,740 people in 2026 covered by financial measures. The question is not how many, but what those numbers produce in the structure of work.
The son of a friend of mine from Belgrade, who graduated in computer science in 2024, waited nine months for a call from the local branch of the National Employment Service. He was eventually informed that he had been put forward for a “self-employment subsidy”, a programme by which the state helps unemployed persons to open a firm. The subsidy ranges between 220,000 and 380,000 dinars, depending on profile and activity. The son accepted the programme, opened a web application development agency, received the funds, and three months later closed the agency, because he had no clients. He did not have to return the money. The statistics registered him as a “successfully employed person”.
This case is not isolated, it is instructive. The Employment Strategy of the Republic of Serbia for the period 2021-2026, like the Action Plan for the period 2024-2026, defines the coverage of active employment policy financial measures: 25,280 persons in 2024, 25,210 in 2025, and 25,740 in 2026. These include self-employment subsidies, public works, professional training, programmes for hard-to-employ persons. The figures are clear, the measures are clear, the execution is reported. The question sociology must ask, and economics need not, is what those programmes produce in the long-term structure of work, not merely in the table of this year’s report.
The difference between a measure and a structure
In the book “The Post-Socialist Transition in Serbia: State, Classes, Social Structure” (Slobodan Cvejic, Plato 2016), the author shows that since 2008 the Serbian model of active employment policy has developed in a direction that could sociologically be described as “compensatory clientelism”. The term does not belong to Cvejic, but it describes the phenomenon he has mapped well: the state, unable to guarantee stable growth of formal employment in the private sector, uses public resources to compensate for the structural gap through programmes that statistically produce “the employed”, and that structurally deepen the dependence of individuals on state channels.
My friend’s son is an example of the first kind. Statistically, in 2024 he entered the category of the employed. Structurally, in 2026 he is an unemployed programmer who tried to work privately, saw that the market was closed (because large integrators do business with the state through tenders that are structurally inaccessible to independent agencies without political connections), and gave up. The subsidy was not a mistake, but it was not a measure that solved the problem either. It solved what the statistics consider a problem (the figure), but did not touch what sociology considers a problem (the structure).
The second type of example is even more complicated. The “public works” programme, used above all in smaller communities, engages unemployed persons in communal, park, cultural or social tasks, with allowances slightly above the minimum wage, for a limited period. In the structure of places like Aleksinac, Pirot, Knjazevac, that programme produces the following sociological consequence: in every round of public works, local committees of the ruling party have an informal influence on who will be put forward. This need not be a formal criterion (it is not, officially), but in practice it is, because the National Employment Service in smaller communities shares a building, a neighbourhood and an institutional culture with the local political machines. A person who enters the programme knows to whom they must be loyal. A person who does not enter knows it too.
Now I must bring in a counter-argument, which I must offer honestly. There is a reading according to which any active employment policy is better than a passive one. A self-employment subsidy, even when it fails, gives an individual a chance that otherwise would not exist. A public works job, even with a clientelistic filter, secures real income for real families. By that logic, criticism of “compensatory clientelism” is an academic luxury. Partly true. But the truth of that criticism does not mean the programme is good, only that it is less bad than the alternative called “nothing”. The difference between less bad and good is what politics must, in the long run, move.
In media sociology, as Pierre Bourdieu would put it in “Distinction” (1979), a clientelistic structure reproduces itself not only through material exchange, but through symbolic exchange, through speaking. A clientelistic system requires the recipient, after receiving resources, to “understand” the exchange, to verbally confirm it, to translate it into gratitude in everyday conversation. My friend’s son did not enter that system, because after closing the agency he went abroad (he is now in Berlin). The recipient of public works in Knjazevac does, because his next round depends on the same filter. The difference between the two is not economic, it is political. The labour market in Serbia in 2026 operates as a combination of two markets, one in which economic selection is at work (who is profitable), and another in which political confirmation is at work (who is loyal).
None of this need be answered with cynicism. The Employment Strategy for the period after 2026, already being prepared, has the opportunity to set a different framework. Three concrete measures could shift that structure, if there were the political will: first, handing management of active measures to an independent agency without ministerial control over personnel decisions. Second, public publication, at the municipal level, of who received which measure and with what outcome, with precise data on economic results six months, one year and two years later. Third, excluding subsidised self-employment programmes that do not come with at least a 12-month mentoring framework and access to the domestic market.
I return to my friend’s son. He is in Berlin, working for a fintech company as a senior developer, earning around 5,500 euros net per month. The 280,000-dinar subsidy he received from the Republic of Serbia has remained in his memory as a curiosity, not as an investment that changed his trajectory. Sociologically, that subsidy is, like his departure, two forms of failure of the same state strategy. Statistically, the one and the other are recorded in different tables. Only when we begin to read the tables together, not separately, will we begin to understand where Serbian labour policy really stands.
