Clientelism

The Albanian protests are widely representative, a broad cross-section of society has been coming out to protest, and one of the main messages has been: “A pox on both their houses”, meaning both Berisha (the leader of the largest opposition party) and Rama, the current prime minister. One of the main chants is: “Rama n’burg Berisha n’burg” (Rama and Berisha in jail). Albanians are certainly tired of the two party system, and they’ve been tired of it for quite a while. With every corruption case, and there have been many, everyone expected some form of popular eruption. But each case came and went and nothing really happened. The talk in Tirana cafes was mostly of exasperation. Then the elections of 2025 came and went, the Socialist Party won 83 of 140 seats, while the Democratic Party-led alliance won 50. People were incredulous, many complained of the system, but still the system prevailed. Most Albanians blamed the fact that the opposition is weak and uninspiring, led by an octogenarian who has been in politics since the early nineties. Personal note: I left Albania in 1995 and Berisha was president at the time. When I came back, nearly thirty years later, he’s still the leader of his party and vying to be the country’s next prime minister. While it is true that the opposition is weak and Berisha a much disliked figure, there is something systemically broken in Albanian politics that goes beyond the personalities and the political platforms of the two major parties. The system itself is built around clientelism, and it is surprisingly robust to outside challenges for a variety of reasons, most having to do with the lack of non-clientelist alternatives, voter cynicism and a well-oiled clientelist machine.

That Albania is a clientelist state is not much in doubt to anyone who has lived here for longer than a year, or anyone living abroad who has paid a modicum of attention. The word patronazhist — meaning a politically connected broker who mediates between voters, parties, and the state — is often heard at the protests and has become a term of contempt. But what is clientelism exactly? It’s a state where access to public goods becomes politically mediated, citizens do not encounter the state primarily as a provider of rights but as a gatekeeper. To get a job, a permit, a public contract or help with an administrative problem, you often need access to someone who has access. In this way political loyalty has become entangled with the ordinary navigation of the state. In a clientelist state, the citizen does not ask what the law says; he asks who can make the law work for him. Rights become favours and public jobs become political rewards. This creates an environment where the relationship of the individual with the state becomes purely transactional and mediated through political parties rather than voting. Want a permit? Better call your local political broker, he’ll get it done for you. Good luck trying through the normal channels. Naturally you need to grease the wheels of the machine, permits ain’t cheap.

But how is the system so resilient and what does it take to change it? It is frustrating to get stuck in a bad equilibrium that everyone realizes is destructive, and no one can get out of. Albanian politics is spinning its wheels at the moment, and the protests are an indication that almost everyone, even those who voted for the system, have had it. Let’s dive into how the Albanian variety of clientelism works, try to understand the disease before we can offer any cures. Side note: I am going to keep using the word patronazhist liberally without translation. It’s such a great appropriation of a foreign term into Albanian, it drips with such contempt that it begs to be used.

As always, for those with little time to spare, I’ll give you a quick synopsis of how Albanian clientelism works. First, the state makes things opaque, vague. For a clientelist state to work, the rules of the game must be difficult to navigate so that patronazhists (for the Albos reading, I know, I laughed too) can be gainfully employed. If the state had clear rules, there would be no need for them. So the necessary condition is that direct access to the state for the provision of public goods must be difficult, either through a series of hurdles or through vagueness. This outsources the provision of public good to politically connected people. Ok, but how exactly do they make sure you vote for their party? They do it by mapping out exactly who is dependent on the state. Then they create a relationship with the voter, making the voter reliant on them. Lastly, they make the following argument to the voter: “If you vote for the other guy, who knows who you’re going to get? I am a tried and tested person, you know me, I’ve solved many problems for you”. Suppose the voter hates the system and wants to change parties. He can, but the other big party does exactly the same thing, so competing for votes mostly comes down to how good your local patronazhist is. You vote for him or her, not out of some ideological conviction. There are smaller parties you can vote for, but electoral law puts heavy barriers on them, so you think that your vote would be wasted. Plus, you’ve never really seen any party do things differently and you’ve become cynical. So in the end you go with your local patronazhist, even though you know this is a crappy system. The whole thing is oiled by corruption and appointments in administrative positions, so your local patronazhist has enough to keep himself fat and happy. That’s all in a nutshell. In what follows I will give you a longer and more empirically based argument, so if you have the time, do read on.

But how do we even know that patronazhists exist? Those who live in Albania know because, well because they live in Albania. But in April 2021, just before the parliamentary elections, the media outlet Lapsi.al[1] exposed a database containing personal information on around 910,000 voters in the Tirana region. BIRN reported that the database purportedly belonged to the Socialist Party and included names, addresses, dates of birth, personal ID numbers, employment data, family background records, past voting behaviour, likely current political preference, and comments written by party officials. Each voter was assigned to a patronazhist. BIRN’s follow-up investigation found 910,061 lines in the database and identified 9,027 patronazhists in the Tirana dataset, roughly one for every hundred voters. It also found that many of them appeared to work in public administration, municipalities, and other public bodies.

The database did at least five things. First, it personalised every voter. Each voter was attached to a workplace, household, voting centre, political preference, and assigned party monitor. That made the voter legible to the party. Second, it assigned responsibility to patronazhists. In effect, the party could ask: who is responsible for this household, this building, this street? Third, it recorded needs and vulnerabilities. BIRN reported that the comments column included notes such as a voter requesting employment for his wife, another note complaining that a voter had not voted and had not even thanked them after receiving the house title, a business owner who should be contacted through his employees, and a voter whose mother was employed in the municipality. Fourth, it monitored social and political behaviour outside the ballot box. BIRN reported notes referring to Facebook profiles, previous party preferences, and people who should be “kept under monitoring.” This is exactly the mechanism: the machine cannot see the ballot directly, so it reads the social traces around the ballot. Fifth, it blurred state and party boundaries. The controversy was especially serious because many people believed that some data may have come from state sources. Transparency International urged Albanian authorities to determine whether the ruling party had obtained voters’ data from government registries.

Let’s unpack each mechanism individually.

1. Opacity

Suppose you notice that your fence has deteriorated. You and your neighbour decide to rebuild it.[2] You start the process but the municipal inspectors come by and tell you that you need a permit to build the fence. You ask around and find out that the permit needs a lot of paperwork: you need to hire a civil engineer to design the wall and map out the borders exactly, then he needs to do a budget so the municipality knows how much to tax you, then he needs to put in the paperwork with the municipality after notarizing it, then put in the same paperwork electronically with the central system, then the municipality has to approve it and then you can start work. You start with the first step and you realize that even if you do all that, once it reaches the municipality, even after you’ve loaded your paperwork electronically, the file will sit there for ages, and you won’t get a response. That’s when your local patronazhist becomes valuable. He knows people at the municipality, you just have to give him the file, with a “small” honorarium for his troubles and those of the municipal workers that will have to “expedite” your permit[3], and voila, within a couple of days you’re ready to begin building. In the old days you could just go directly and bribe the municipal worker, but that system was too decentralized for the main parties, they needed to grab control.

This whole system relies on the procedure being as opaque and as difficult as possible. As a last resort, the government can just tell its workers to approve projects at a snail’s pace. Either way, the system is built around the patronazhist. No one in that chain has any incentive to change the system. The government doesn’t, because it’s how it wins elections, the patronazhist lives off it, and the municipal worker gets a taste of the action and has no incentive to work diligently. In fact, every link in this chain has an incentive to make the system as difficult to navigate as possible and the voter as dependent on it. The only person with any real incentive to change the system is the voter, but he has quite a hill to climb.

2. The other party is the same

Who am I going to vote for is the tiresome but reasonable complaint of the average Albanian. It is reasonable because the model is the same for both parties, and political competition is not over who is better at governing but who is better at clientelism. When both major parties are clientelist, elections do not offer voters a clean choice between patronage and public goods, but between rival machines of access. The Socialist Party and the Democratic Party are not identical, and incumbency matters because the party in power has greater control over state resources, but both have long operated through patronage, brokers, personal leadership, and selective distribution. This changes what political “competence” means. In a normal system, voters might ask which party has the better plan for taxation, education, health. In a clientelist system, many are pushed to ask a narrower and more immediate question: which party can get my relative a job, protect my business, move my paperwork, shield me from punishment, or deliver something tangible to my family? The result is that parties are judged less by their capacity to govern than by their capacity to mediate access effectively. A party can appear electorally competent not because it produces better public goods, but because it has deeper broker networks, stronger control over public employment, closer ties to business and media, and a more credible ability to reward loyalty or punish defection. The voter who chooses such a party is not necessarily naïve; he may understand perfectly well that the system is bad. But inside a clientelist order, voting often becomes a survival calculation rather than a programmatic choice.

Does it work? Do voters actually get the benefit of voting this way? A recent study on Albanian elections found that:

supporting the ruling party significantly increases individuals’ employment and earnings. This labor market premium is particularly large among individuals with low costs of campaign participation, while patronage jobs are concentrated in lower-tier public sector positions. Administrative data further reveal that the allocation of jobs to party supporters is strongly associated with a higher vote share for the incumbent.

3. High barriers to entry for smaller parties

The question is then: Why don’t Albanians vote for other, smaller parties? Here again, the country’s two main parties came together in an extremely rare moment of unison and changed electoral law together. This was striking because if you know anything about Albanian politics you’ll know that these two parties will argue about whether the earth is flat or not, their conflict is that deep. But they agreed on this.

Albania’s electoral system does not make small parties impossible, but it does make alternatives appear risky and often futile. Because MPs are elected through regional proportional contests rather than a single national pool, small parties must concentrate enough support in specific districts to win seats; in smaller constituencies, the effective threshold can be much higher than the formal one. This creates powerful strategic-voting pressure. A voter may sympathise with a new anti-clientelist party but still conclude that supporting it would waste their vote, especially if the party’s backing is spread too thinly to secure representation. The 2025 result illustrates this dynamic: the Socialist Party and the Democratic Party-led alliance won almost all seats, while all other parties together secured only a small remainder. The deeper problem is behavioural as much as institutional. Electoral rules create coordination problems, pushing voters toward contenders seen as viable. In Albania, that logic reinforces the clientelist duopoly: small parties look weak, so voters avoid them; because voters avoid them, they remain weak. Clientelism makes alternatives materially risky, while the electoral system makes them look electorally wasted.

4. Voter cynicism

This is where Albania’s recent history comes into play. The average Albanian has never experienced a state that provides public goods in an efficient way, free of political meddling. No one can point to some golden period where the state worked for everyone, where your political leanings or party membership had nothing to do with whether you received benefits from the state or whether you could apply for a job. It’s hard to convince someone that has never experienced this, that another world is possible. That’s because in this environment cynicism is rational, even though everyone knows the outcome is bad for all. Voting for smaller parties is a risk. How do you know that they won’t turn around and do exactly the same thing? Voters in this case are asked to take a risk and change a system that they have known forever. If they have enough to lose, that risk becomes prohibitive.

Clientelism survives not only because parties impose it from above, but because many voters have adapted to it from below. When citizens have little experience of a state that works through neutral, impersonal rules, politics begins to look like a permanent world of connections, favours, informal payments, party lists, and personal intervention. Voters may dislike the system, but still believe that every party will steal, that anti-corruption promises are mostly theatre, and that clean alternatives will either lose or become corrupt once in power. In that environment, a scandal simply confirms what voters already assume, that everyone is corrupt. The electoral question then shifts from “who is clean?” or “who will govern well?” to “who can still help, protect, deliver, or share?” Reform parties need voters to believe that rules can replace favours, but clientelist parties only need voters to believe that favours are the only reality. This is why cynicism stabilizes the system: it lowers expectations, normalizes dependency, and makes clientelism appear unavoidable.

5. Corruption oils the whole machine

In terms of corruption scandals, Albania has had some pretty egregious cases. We have an ex deputy Prime Minister in exile (he escaped) accused of embezzling millions of euros from public projects. Another former deputy Prime Minister that has been accused of violating public tender procedures, whose immunity the parliament refuses to revoke. Tirana mayor Erion Veliaj was arrested and later sent to trial on corruption, money-laundering, asset-declaration and abuse-of-office charges. A large number of government ministers, MPs and other public officials have either been accused, are standing trial or are actually getting out of jail since sentence ranges are determined by parliament and surprisingly, they’re fairly light. This is not a coincidence in a clientelist system.

Corruption is what keeps the machine moving. Clientelism requires resources to distribute: jobs, contracts, cash, permits, protection, media influence, and favours. Corruption generates those resources by turning public office into a source of rents, which can then be recycled into political loyalty. Inflated contracts, procurement abuse, regulatory favours, informal payments, create money for campaigns, rewards for brokers and dependence among business elites. A clean bureaucracy would make clientelism harder because rules would be visible, automatic, and enforceable. A corrupt bureaucracy makes clientelism easier because rules become flexible for allies and rigid for opponents. In this sense, corruption is not merely a moral failure or an occasional scandal; it is the lubricant of the clientelist state, converting public resources into private obligations and political loyalty.

There you have it. This is how Albanian clientelism works. People have known this for a while, but the fact that they’re all up in arms now, years after all the corruption cases were revealed and the patronazhists data dump was published, shows that the economic pain from this system has reached deep. This is a protest heavily supported by a young generation that does not carry the cynicism of their parents and are explicitly rejecting it. It has now attracted all generations in the hope that another Albania is possible, and they’re intent on making it happen.

(c) 2026 Elton Dusha. All rights belong to the author. Cover image generated with AI.

---

[1] The original link to the article has disappeared. I used the internet archive to recover a snapshot.

[2] This is not exactly a hypothetical. It’s happened to someone I know.

[3] Albanians call this “having a coffee”