The examples in this article are analytical composites based on recurring patterns in Bulgarian political life. They do not describe or accuse a particular politician, party, or group of voters.

When our values begin following our vote

A voter chooses Candidate A because the candidate appears to defend certain principles: European integration, national sovereignty, lower taxes, social justice, minority rights, judicial reform, traditional values or clean government.

Two years later, Candidate A has changed position. A coalition once described as impossible is suddenly presented as unavoidable. A policy once condemned is now defended. A red line becomes a negotiating position. A promise is postponed, diluted or quietly forgotten.

Yet the voter does not withdraw support.

Instead, the explanation changes:

The situation is complicated.

This is only tactical.

The others are worse.

The leader knows things we do not.

The media are attacking him because he threatens the system.

At some point, the voter is no longer choosing the politician according to the voter’s values. The voter is adjusting those values to protect an earlier political choice.

This is commonly described as commitment bias. But in politics, it is rarely a single bias. It is more accurately understood as a psychological chain involving cognitive dissonance, identity, motivated reasoning and escalation of commitment.

And Bulgaria offers an unusually revealing environment for observing it.

A vote is not always the end of a decision

We normally imagine voting as a simple sequence:

I hold certain views. I compare the candidates. I choose the one closest to my views.

But the process can also run in the opposite direction:

I choose a candidate. I become emotionally invested in that choice. I gradually adjust my views to remain close to the candidate.

Political psychology provides evidence that voting can influence later attitudes rather than merely express attitudes that already existed. One influential study found that the act of being eligible to vote was associated with substantially more polarized later evaluations of presidential candidates, consistent with the idea that people reinterpret subsequent information to make their earlier choice feel correct.

Once a voter has declared support publicly, argued with friends, shared campaign messages, donated money or persuaded relatives, changing position becomes more expensive. The cost is no longer only political. It is psychological and social.

To abandon the candidate may feel like admitting:

I misjudged the person.

I defended something indefensible.

The people I argued with may have been right.

For many people, it is easier to reinterpret the candidate’s behaviour than to reinterpret themselves.

The psychological chain

Commitment bias is the tendency to continue supporting a previous decision even when new evidence suggests that the decision should be reconsidered.

Cognitive dissonance is the discomfort created when two beliefs conflict: “I am a sensible and moral person” and “I may have supported someone who no longer represents what I believe.”

Choice-supportive bias leads people to remember the positive qualities of the option they selected and exaggerate the defects of the alternatives.

Motivated reasoning means that people do not examine every piece of political information with the same purpose. Sometimes they seek accuracy. At other times, often unconsciously, they seek an argument that allows them to preserve an existing belief or identity. Political research distinguishes these “accuracy goals” from “directional goals”, where reasoning is directed toward reaching the preferred conclusion.

Finally, partisan identity can transform a political preference into group membership. Research on affective polarization shows that party identity itself can shape reactions independently of the actual policy involved. The same proposal may be judged differently depending on whose political label is attached to it.

The complete cycle looks like this:

A person votes, publicly defends the vote, encounters contradictory evidence, feels discomfort, rationalizes the contradiction and emerges even more emotionally attached to the original choice.

The more effort invested in defending the candidate, the harder it becomes to leave.

The Bulgarian paradox: volatile politics and immovable cores

At first sight, Bulgaria may appear to be a poor example of political commitment. Parties rise quickly and sometimes decline almost as quickly. Coalitions change. New political projects repeatedly attract voters who have abandoned older ones.

During the intense electoral cycle of 2021–2024, Bulgaria held six parliamentary elections by June 2024. Turnout in the June 2024 parliamentary vote fell to 34.41 percent. International observers described an atmosphere of declining institutional trust, antagonism among major political forces, election fatigue, and campaigns that frequently emphasized damaging information about opponents rather than substantive policy debate.

But high electoral volatility and commitment bias are not opposites.

They can coexist.

The disappointed, weakly attached voter switches parties or stops voting. The strongly attached voter remains. As turnout falls, disciplined political cores constitute a larger share of the votes actually cast.

The result is a paradoxical political system: large parts of society appear detached from politics, while smaller groups become intensely committed to particular leaders, identities and narratives.

Repeated elections may reinforce this division. The exhausted voter withdraws. The loyal voter returns. Politics becomes increasingly dependent on those for whom voting is not merely a choice but an expression of belonging.

The reform project that becomes a personality

Consider a voter who supports a new political project because it promises competence, transparency, and institutional reform.

Initially, the project is a vehicle for principles. Over time, however, its public identity becomes concentrated around one or two personalities. Criticism of the leader begins to feel like criticism of the entire reform cause.

When the project fails to deliver, several explanations are available:

• the institutions resisted;
• coalition partners sabotaged the reforms;
• the media distorted the facts;
• the electorate did not provide a sufficient majority;
• foreign or domestic interests intervened.

Some of these explanations may be true. Governments do face institutional limits, hostile partners and parliamentary arithmetic.

Commitment bias begins not when a voter accepts a reasonable explanation, but when every possible outcome confirms the same conclusion:

• Success proves that the leader is competent.
• Failure proves that the leader is dangerous to the establishment.
• Compromise proves political maturity.
• Refusal to compromise proves moral integrity.
• A policy reversal proves flexibility.
• Maintaining the policy proves consistency.

A belief that explains every result cannot easily be tested by any result.

Research on Bulgarian party development has highlighted the importance of personalized leadership and shown that local branches and territorial organization can stabilize support as a leader’s original popularity begins to decline. Political loyalty, therefore, may become attached simultaneously to a personality, an organization, and a local network.

The coalition that was impossible until it became necessary

Bulgarian voters have repeatedly heard categorical declarations that cooperation with a certain opponent would be impossible.

Then comes an election without a majority.

The impossible coalition is renamed:

• a temporary formula;
• a technical arrangement;
• a constitutional majority;
• a national responsibility;
• the lesser evil;
• the only way to avoid another election.

Coalitions are not inherently betrayals. Parliamentary democracy requires negotiation. A party that never compromises may never govern.

The psychological question is different: Would the voter accept the same explanation from an opposing party?

When a rival enters an unexpected coalition, it is described as corruption, opportunism or betrayal. When one’s own party does it, it becomes statesmanship.

That double standard is one of the clearest indications that political identity has overtaken political principle.

A consistent voter may conclude that both coalitions were justified, that both were unacceptable, or that the circumstances were materially different. A biased voter begins with the party label and constructs the standard afterward.

When the voter changes position after the party does

Imagine that a voter supports a party because it is firmly committed to rapid European integration and adoption of the euro.

Later, the party leadership becomes hesitant, demands a postponement, or begins to emphasize risks it previously dismissed.

There are two possible responses.

The voter may say:

“This was one of the reasons I supported the party. I need a convincing explanation for the change.”

Or the voter may gradually adopt the new position without acknowledging that anything has changed:

“I have always believed we should be cautious.”

The mirror image is equally possible. A voter who originally supported a party because of resistance to deeper European integration may suddenly accept a compromise when that party begins to present it as pragmatic.

Commitment bias has no fixed ideology. It can affect pro-European and Eurosceptic voters, liberals and conservatives, the left and the right.

The revealing question is not whether the voter supports or opposes the euro. It is:

Did the voter’s position change because the evidence changed, or because the party’s position changed?

The party as protector of a community

Another Bulgarian pattern involves parties perceived as protectors of a particular ethnic, religious, regional, professional or social community.

Such loyalty should not automatically be dismissed as irrational. A community with a history of discrimination or political exclusion may reasonably value collective representation. A voter may support an imperfect party because fragmentation appears to weaken the community’s political influence.

But this legitimate concern can also create a powerful psychological shield around party leadership.

Criticism of the organization becomes framed as an attack on the community. Leaving the party is presented not merely as changing one’s vote, but as betraying one’s people. Leadership accountability becomes difficult because voters are asked to choose between internal criticism and collective security.

Commitment bias begins when the claim “this party protects us” becomes impossible to evaluate. If declining local services, questionable alliances, leadership changes and policy reversals can never count as evidence against the claim, the party has ceased to be treated as an instrument. It has become a permanent identity.

The anti-system voter who builds a new system

The same mechanism appears in anti-establishment politics.

A voter supports a movement because it condemns professional politicians, privileges, opaque decisions and backstage bargaining. Once elected, however, the movement must operate inside parliament. Its representatives accept institutional resources, negotiate with opponents, appoint officials and make procedural compromises.

Again, some adaptation is inevitable. Politics cannot function entirely as permanent protest.

But supporters may begin excusing precisely the behavior that originally mobilized them:

• secret negotiations become a necessary strategy;
• centralized leadership becomes discipline;
• attacks on dissenters become defense against infiltration;
• unfulfilled promises become proof that the system is blocking change.

The anti-system identity then becomes self-sealing. Every institutional failure strengthens the belief that the movement is right, while every institutional success strengthens the movement itself.

The municipal network: bias, dependence or rational calculation?

At the local level, repeated political loyalty may have a different explanation.

A family may depend on a municipality, a public enterprise, a local contractor, or a politically connected employer. A voter may believe that changing the governing group could threaten employment, access to services, permits, contracts or informal administrative assistance.

This is not necessarily commitment bias. It may be clientelism, pressure, dependency or calculated self-interest.

Research in Bulgarian and Romanian municipalities has documented how long-term local incumbency can facilitate the politicization of state resources, administrative favors and other non-programmatic forms of mobilization. The authors found such strategies to be more prevalent where political incumbents had remained in power for longer periods.

The distinction matters.

A voter who says, “I know this leadership is poor, but my livelihood depends on it,” is not necessarily deceiving himself. He may be responding rationally to an unhealthy institutional environment.

Commitment bias occurs when dependency is transformed into moral admiration:

“Because this group helped my family, it must also be honest, competent and good for the entire municipality.”

Material dependence explains the vote. Bias rewrites the evaluation.

How political communication strengthens commitment

Political actors do not need to understand academic psychology to benefit from these mechanisms.

The most effective strategies are familiar.

First, politics is personalized. The party becomes the leader, and the leader becomes the cause.

Second, disagreement is moralized. Opponents are not merely mistaken. They are corrupt, foreign-controlled, ignorant, treacherous, or dangerous.

Third, supporters are encouraged to make public commitments. The more publicly someone identifies with a movement, the greater the reputational cost of leaving.

Fourth, failures are attributed to external enemies, while successes are attributed to the leader.

Fifth, standards are made asymmetric. The opponent’s compromise is betrayal; our compromise is responsibility. Their appointment is political capture; ours is necessary reform. Their media supporters are propagandists; ours are independent voices.

This is how a political choice becomes psychologically difficult to reverse.

Loyalty is not automatically a bias

There is nothing irrational about voting repeatedly for the same party.

A party may continue to represent a voter’s interests. A local administration may perform well. A political organization may make reasonable compromises while preserving its central principles. A voter may examine the alternatives and conclude that the familiar option remains preferable.

Research on Bulgaria has also found that voters can evaluate the performance and perceived corruption of their preferred parties. Political commitment is therefore not absolute, and Bulgarian voters should not be portrayed as passive prisoners of identity.

The difference between loyalty and bias is conditionality.

Loyalty says:

“I will continue supporting you while you meet certain standards.”

Commitment bias says:

“Because I support you, whatever you do must somehow meet my standards.”

Healthy loyalty can be withdrawn. Bias cannot imagine the conditions under which withdrawal would be justified.

The mirror test

A voter can test political commitment with a few uncomfortable questions:

1. Would I defend the same action if an opposing party had taken it?
2. Can I state the three principles that originally led me to support this candidate? Are those principles still present?
3. Did my opinion on an issue change before the party changed its position, or afterward?
4. What specific evidence would make me withdraw my support?
5. Am I defending a decision because it is right, or because admitting error would be painful?
6. Would I accept the policy if the party name and leader’s face were removed from it?

The most important question is the fourth. A political belief that cannot be altered by any conceivable evidence is no longer an ordinary political judgment. It has become an identity commitment.

A ballot is a temporary contract

Bulgaria does not suffer from a uniquely Bulgarian form of irrationality. These mechanisms appear across democracies.

What makes them particularly visible in Bulgaria is the combination of personalized political projects, repeated electoral crises, low trust, local dependency, coalition reversals and declining participation. In that environment, politics easily becomes a struggle between identities rather than a comparison of policies.

The antidote is not permanent cynicism. Nor is it changing one’s vote at every election.

It is conditional support.

Before an election, voters can write down the principles and promises that matter to them. After the election, they can compare those commitments with actual decisions. Journalists can revisit previous statements instead of reporting each new position as though it had no history. Parties can publish measurable programs and explicit coalition conditions. Citizens can read credible criticism of their preferred side, not only criticism of its opponents.

Most importantly, changing one’s vote should not be treated as shameful inconsistency.

A ballot is not a marriage vow. It is not a declaration that the chosen candidate is permanently virtuous. It is not a confession of personal identity.

It is a temporary contract.

The mature voter is not the person who never changes sides. It is the person who applies the same standards after voting as before voting.

And perhaps the clearest sign of democratic independence is the ability to say:

“I supported you when you defended these principles. You no longer defend them. Therefore, you no longer have my vote.”

Related Readings:

• Action vs. Inaction: Unraveling the Intriguing World of Action Bias
• Elections and Parliamentary Seats in Bulgaria
• Navigating Feedback Across Cultures: A Guide for Foreign Managers in Bulgaria