Custodians, not managers: who runs Romania’s state-owned companies

Originally published in Romanian in Economedia on 3 November 2026.

This autumn, the Bolojan government launched, at the initiative of Deputy Prime Minister Oana Gheorghiu and her office, the first digital platform listing the leadership of state-owned companies: positions, pay, political affiliation and CVs. I analysed the professional backgrounds of the 557 directors and board members of 121 central-government state-owned companies who have a public CV, and set them against the governance standard Romania has committed to as part of its OECD accession. The conclusion is uncomfortable: the state does not select managers, it selects custodians — people who look after a company, not leaders who take it forward. Not because the rules are missing: on paper, our law is in good shape. But because, somewhere between the law and the appointment, we lose exactly what the standard demands: proven competence and distance from politics.

Where the stakes really are

Around 90% of turnover and almost 90% of net profit of the companies analysed come from monopolies, unique infrastructure and companies with a dominant position in liberalised markets. At the other end, state-owned companies that actually compete in an open market account for just over 3% of turnover and, taken together (including TAROM, the national airline), post a net loss. The OECD’s 2023 review of Romania reached the same conclusion on different data: the return of the state portfolio comes almost entirely from the five most profitable companies.

When results are secured by tariffs, market position and assets, management seems less relevant; the data show that the profitability of monopolies is not influenced by the profile of those who run them. And what is not tested through market activity is not required in the selection of executives either.

Who is in charge

• Missing sector expertise. 51% of those analysed have never worked in the sector of the company they run or oversee. On boards, the share rises to 61% of seats.
• Limited scale. At most 22% have ever led an organisation of more than 500 people. Only 16 people have led more than 5,000.
• Duties, not results. 79% of CVs list duties only. Just 7% contain a result with figures: growth, a turnaround, an investment carried through.
• Interim appointments are down, but not gone. 16% of mandates are provisional or interim. The progress is real — in 2021, according to the OECD, 72% of board seats in the 50 largest companies were provisional — but the 12-month legal limit is still being breached: AMEPIP, the agency that monitors state-owned enterprises, went as far as fining a minister (in Romanian) over extended provisional mandates at a company run on an interim basis since 2016.

Politics: what the guvernanta.gov.ro platform says and what the CV shows

The platform records the political affiliation reported by the ministries. According to it, 103 people (18%) have a party affiliation: 44 PSD (Social Democrats), 41 PNL (National Liberals), 17 USR (Save Romania Union) and one UDMR (the Hungarian minority party). Another 355 are listed as having “no political affiliation”, and for 99 people the field is blank.

The CVs, however, say more. A political or politico-administrative track record — from state secretary or ministerial adviser to elected local official or party positions — appears in 34% of those analysed, almost double the recorded party affiliation. And among those listed with “no political affiliation”, more than 90 people, over one in four, show such a track record in their own CV. The platform field shows a label; the CV shows the network through which the appointment comes.

The highest share of political history is found on the boards of monopolies: 47% of board seats. Precisely where performance can be neither proven nor challenged.

Seven career patterns

I did not start from predefined categories: I grouped people statistically by the similarity of their career paths, and seven profiles emerged from the data.

Only two profiles resemble what the private sector would call a company manager: the home-grown insider and the corporate manager. Together, they account for less than 20%. The others are not professionally weak people; they are competent in something else: law, audit, administration, their own small business or political negotiation. The problem is not incompetence, but a mismatch with the role.

The system rewards conformity, not leadership competence. From the perspective of organisational psychology and of the 25 years I have spent assessing leaders, the best and most readily available predictor of future performance is behaviour demonstrated in comparable situations. A CV that lists duties does not say what the person did, only what the job description said. When 79% of CVs look like this, the selection committee does not judge the candidate’s current and potential performance, but conformity with a form.

This is why the monopoly and the custodian fit together so well. The selection process does not change, errors are not visible, so we learn nothing, and the pattern reproduces itself. Such a company resembles an old, well-kept house: it holds its value through the land and the walls, but no one asks any more whether it should be renovated or extended. The custodian preserves what was handed over; the manager builds what comes next.

Portrait of the decision-maker: the CEO and the board

The typical CEO is an engineer or an economist with a master’s degree. Almost four in ten grew up in the same company or group, and 60% know the sector. Three in four have led something before, most often a department or a branch of 50–500 people, usually also in the state sector. One in three has a political track record in their CV. Not an amateur, but a home-grown custodian: the role they actually play is to keep the company running and hand it on, when the time comes or when asked.

The typical board member is an economist or a lawyer, coming from a small firm, their own practice or the public administration. Six in ten have never worked in the sector of the company they oversee, and almost four in ten have a political history. This is the weak link of the system: the board — although its essential role is to set objectives, assess the CEO and remove them if necessary — is, paradoxically, the body with the least technical competence in the company’s specific field of activity.

The pay paradox: big money where management matters least

According to the platform, a CEO earns a median of about 32,000 lei gross per month (one euro is roughly five lei), more than three times the average gross wage (in Romanian) in Romania, of about 9,500 lei. A board seat brings about 13,700 lei for a mandate that requires periodic attendance at meetings — more than the salary of a full-time employee. One in six people holds two or three mandates at once; with three such positions, the median monthly income reaches about 59,000 lei gross. In total, the remuneration listed on the platform adds up to more than 130 million lei a year.

Normally, market logic dictates that the highest pay should go to the manager who manages to attract customers and win market share in a competitive environment. In state-owned companies, however, this principle works exactly the other way round.

At the median, a board seat at a monopoly is worth almost four times as much as one at a competitive company. Part of the difference is down to size, since the competitive companies also include small firms and mines being closed down. The conclusion stands: where the state most needs a manager, it offers the least money and attracts the weakest profiles.

Within boards, too, the money is not distributed at random. Political history rises with pay: it is present among 31% of members in the lowest-paid quartile and 47% of those in the highest-paid quartile. For CEOs, the trend is reversed: the best paid more often have sector expertise. The board remains a safe zone: a place where an appointment can be both well paid and shielded from performance pressure.

The mandate as currency

A board seat works like an annuity: it generates substantial income for part-time work, carries limited accountability and, when provisional, can be revoked at any moment. It can reward proven loyalty or secure future loyalty. The OECD notes that interim board members are replaced in line with political cycles, and agency theory describes the mechanism: the appointee serves the interests of whoever has the power to remove them. Although the data cannot prove the intent behind each individual appointment, they clearly reflect the system of interests and rewards that governs these decisions.

The comparison with other countries is telling. In Sweden, the state ownership policy provides that government officials who sit on the boards of state-owned companies are, as a rule, not paid a fee. In Germany, a federal civil servant sent to a supervisory board may keep at most between €3,700 and €6,100 a year (in German), depending on pay grade, with the rest paid back to the state. There, the board seat is an official duty. Here, it can become a second salary.

The standard exists. Enforcement does not

The OECD Guidelines on Corporate Governance of State-Owned Enterprises, revised in 2024 and officially translated into Romanian, require board members, including any civil servants, to be appointed on the basis of qualifications relevant to the company’s field of activity and commercial profile. Politicians who can significantly influence the conditions in which the company operates should not sit on its board, and former politicians should be subject to a cooling-off period set in advance.

Romania has obtained 24 of the 25 formal opinions of the OECD committees; “technically speaking, we are ready” (in Romanian), said Luca Niculescu, the accession coordinator who has since become prime minister, in early October 2026. On paper, our law is sound: seven years of experience, an independent majority on boards, a competency matrix, interim mandates capped at 12 months and a ban on politicians currently in office. Romania does not have a standards problem; it has an enforcement problem. The law bans a current political office, not the political track record behind it. The competency matrix exists, but risks remaining a bureaucratic formality. And without evidence of performance, even the best procedure will not succeed in selecting the manager a state-owned company needs.

What the same job looks like where the rules are applied

• What the board needs to know: In Romania, 6 in 10 board members have never worked in the company’s sector. In Sweden, each member is required to bring industry knowledge or directly relevant expertise. In Germany, the supervisory board as a whole must cover all the technical and operational knowledge required.
• Politics. In Romania, 34% of those analysed have a political history in their CV, almost double the share officially declared on the platform. Sweden does not ban active politicians by law, but in practice it does not nominate them. Germany includes members of parliament and ministry representatives on boards, but their appointment is openly declared, made in the name of and to represent the state.
• Who gets to the top. At Vattenfall (Sweden), the CEO appointed in 2020 was promoted directly from the position of the company’s chief financial officer. At Deutsche Bahn (Germany), the CEO appointed in 2025 has a career of almost 30 years in industry, energy and rail transport.

Let us not idealise: these countries have their setbacks too. The board of the Swedish mining company LKAB, although chaired by a former finance minister, was publicly criticised for its lack of mining expertise (in Swedish). At Deutsche Bahn, the supervisory board is chaired by a former state secretary and includes members of parliament from the governing coalition; even so, the government appointed three managers from private business (in German), ensuring that half of the shareholder representatives come from the leadership of large companies.

The fundamental difference lies in transparency and in the system’s capacity to correct itself. There, political involvement is openly declared, a lack of competence quickly becomes a matter of public debate and is corrected, and companies are run by professionals with long experience in their field. In Romania, the order is reversed: political affiliation is visible only in the CVs, not on the public platform, and sector competence is missing precisely from the body that should demand it rigorously: the board.

Solutions: closing the gap between law and practice

What is needed is not a reform from scratch, but the rigorous application and calibration of the rules that already exist.

1. The profile follows the company’s role. A monopoly needs an excellent custodian, able to hand over the infrastructure in better condition than it was received, with performance assessed on investments delivered, efficiency and quality of service, not just on a profit guaranteed by tariffs. A company in a competitive market, by contrast, needs a manager who has already been directly accountable for an organisation’s results in an open market.
2. Pay and its variable component follow the company’s time horizon: In the competitive sector, the state must look for the best profiles and pay market rates; otherwise, it will not attract competence. Executives in all companies should have a real variable component, capped and public — the law has provided for this since 2011 (in Romanian). At an infrastructure monopoly (e.g. the railways), indicators are measured over 3–5 years (investment, safety, punctuality). At a market-facing company (e.g. TAROM), indicators are annual (operating result, load factor, cost per seat-kilometre), but part of the bonus is deferred and paid in instalments, so that short-term profit does not sacrifice the company’s future. Bonuses paid on results later proven false are clawed back. Board members receive only a fixed fee, with no bonus, as the OECD recommends, so that they remain independent of the executives they oversee.

Diverging international practice: International models differ, but both work because of their rigour. In Finland, at the electricity transmission system operator, the CEO’s variable pay is capped at 40% of fixed salary (80% in exceptional cases) and is partly calculated over three years. Sweden, by contrast, rules out variable pay entirely for state-company executives and compensates with extremely strict selection. The failure of our model lies in a toxic combination: high pay where the tariff secures the result, low pay where the manager should be building it, bonuses invisible to the public and a recruitment process emptied of its selective value.

3. A scale threshold. A simple condition that is easy to verify: anyone applying to lead a large company must already have successfully led an organisation of similar complexity.
4. Evidence, not duties. Replacing a checklist of past responsibilities with a standard application file built on verifiable results: what the company looked like when the candidate took over, how they left it, and what the concrete figures of their mandate were.
5. A political cooling-off period based on the actual track record. This is the only major requirement of the OECD Guidelines that is entirely absent from Romanian law. The restriction must cover the real political track record (former state secretary, ministerial adviser, elected local official) for a fixed number of years set in advance. The public transparency platform should display this complete history, not just the label of the current party.
6. A public competency matrix and a real board evaluation. Every vacant board seat should be filled strictly to cover the gaps in the board’s competency matrix. The annual board evaluation, carried out with the involvement of external experts, should directly determine whether mandates are renewed. According to the OECD’s 2026 report, 71% of the countries analysed already use evaluation results for new nominations. Exceeding the legal limit on interim mandates must also be firmly sanctioned.
7. Professional assessment of the manager. Selection cannot stop at reviewing a CV and a formal interview. Judgement in ambiguous situations, integrity, decision-making under pressure and the ability to build teams must be measured with scientifically validated instruments, by independent assessors. Their role is not to replace the decision of the state as shareholder, but to base it on evidence.

The test of the next appointment

For the decision-maker who signs the next appointment, three questions are enough to bring clarity to the process:

1. Does the company run a monopoly or compete in a market?
2. Has the candidate already led a structure of this scale, with proven results?
3. Who assessed the candidate, and with what instruments, before they were proposed?

The performance of any organisation is fundamentally determined by the quality of its human capital, and it is the state that sets its own standard through its appointments to public enterprises. That is where the governance chain breaks today, and only there can it be rebuilt. A manager-custodian will, at best, hand over a company at the same level — but entirely unprepared for tomorrow’s market.

With OECD accession on the home straight and with the transparency brought by the new public platform, we have, for the first time, both the context and the data needed. It remains to be seen whether the Romanian state will choose to use them for the next appointment.

Claudia Indreica, PhD is an organisational psychologist and the founder of Psihoselect. For more than 25 years, she has supported Romanian and international companies operating in Romania in their toughest decisions about people: psychological profiling, executive search and organisational diagnosis & intervention. She holds a PhD in organisational psychology from Babeș-Bolyai University.

She leads the Labour Market Task Force of Romanian Business Leaders, is a board member of DWNT and coordinates the debates of the Râșnov Film and History Festival. She writes about the labour market and leadership in Economedia, on Substack and on LinkedIn.

Methodological note and limitations

Data. The guvernanta.gov.ro platform, updated on 24–25 September 2026: 121 central-government companies, 585 people, 697 mandates. Turnover and net result for 2025 come from the ranking published on the same site (114 companies). For TAROM, which is missing from the ranking, I used the turnover from the 2024 financial statements filed with the Ministry of Finance (1.23 billion lei) and the estimated 2025 loss of 186 million lei (in Romanian) made public by the government. Pay figures are medians of the gross monthly amounts shown on the platform, calculated across 570 mandates with a positive amount; the platform does not specify whether they include the variable component.

Method. I coded the CVs of the 557 people against 12 criteria, with the support of AI-assisted analysis tools, using a grid defined by the author and calibrated on a pilot of 20 manually coded CVs. A blind manual recoding of 50 randomly selected CVs produced 90% agreement (between 82% and 98% across criteria). The seven patterns come from hierarchical cluster analysis (Ward’s method). The classification of companies by market type — monopoly, dominant position, competitive market, captive market — is the author’s own.

Limitations. CVs are self-declared and non-standardised; 28 people have no public CV. Political track records are read from CVs, so they are a minimum estimate. The size of organisations previously led was coded generously, so 22% is an upper bound. The financial data cover two years and mandates change often, so we are talking about associations, not causation. This is a system-level analysis: it does not judge any individual person or company.