Turkey is voting again today after neither Recep Tayyip Erdoğan nor Kemal Kılıçdaroğlu secured an outright majority in the first round of the presidential election two weeks ago. Erdoğan entered the runoff with the advantage after finishing just below the 50% threshold, despite Turkey’s severe inflation, the depreciation of the lira, criticism of his government’s earthquake response and an opposition alliance broader than anything he has faced in years. For many outside observers, this is difficult to reconcile with the assumption that governments presiding over serious economic deterioration should be punished electorally. Turkey is demonstrating why that assumption is too simple.
Economic voting is real. Citizens generally care about inflation, employment, wages and living standards, and governments frequently lose support when those conditions deteriorate. Turkey’s inflation has severely reduced household purchasing power, while Erdoğan’s insistence on low interest rates despite rising prices has placed his economic judgement at the centre of political debate. Yet voters do not enter polling stations with one variable called “the economy” and mechanically reward or punish whoever governs. They compare economic concerns with security, religion, nationalism, party identity, leadership and their judgement of the available alternative.
That last part is crucial. Elections are comparative rather than retrospective. A voter can believe Erdoğan has managed the economy badly without believing Kılıçdaroğlu would govern Turkey better. Dissatisfaction with an incumbent creates an opportunity for opposition; it does not transfer votes automatically. The opposition still has to convince citizens that replacing the government produces an expected improvement large enough to justify the uncertainty of change.
Erdoğan possesses an enormous advantage here because twenty years in national power have made him a known quantity. His opponents can point towards the costs of his economic model, but voters also know roughly what Erdoğan represents on religion, security, infrastructure and foreign policy. Kılıçdaroğlu’s six-party alliance is broader but consequently less coherent. A voter dissatisfied with Erdoğan may still wonder whether parties ranging across secular, nationalist, conservative and liberal traditions can govern effectively once their common opponent is removed.
This is the weakness of a negative coalition. Opposition to Erdoğan can unite parties during an election, but voters are choosing a government rather than merely participating in a referendum on the incumbent. The more ideologically diverse the coalition, the more uncertainty exists about what follows victory. Erdoğan can exploit that uncertainty by presenting himself as stability against an opposition whose principal agreement is that he should leave.
Incumbency also changes how citizens experience economic policy. Governments possess tools unavailable to opposition parties: public-sector wages, pensions, credit policies, reconstruction programmes and infrastructure investment. Such measures can be economically justified while also affecting electoral incentives. The relevant distinction is not whether governments should stop governing before elections, which would be absurd, but whether public resources increasingly become indistinguishable from partisan political resources.
This is one reason long incumbencies can become self-reinforcing even without eliminating elections. A governing movement accumulates organisational networks, relationships with businesses and media, local patronage structures and familiarity among voters. Each additional year in power can make the cost of replacing the incumbent appear larger because more institutions have adapted around the expectation of continuity. Electoral competition remains real, but the challenger is competing not only against another candidate but against an established political ecosystem.
Erdoğan’s appeal also cannot be understood without recognising the identity dimension of economic development. His governments did not simply produce growth during their earlier years. They altered which social groups felt represented by the Turkish state. Conservative and religious citizens who believed the strongly secular Kemalist establishment had historically treated them as culturally inferior gained greater visibility and political power under the AK Party. For those voters, Erdoğan can represent social recognition as well as economic management.
That kind of political loyalty is more durable than support based purely on annual GDP growth. If voters interpret a party as representing who they are, economic deterioration must become severe enough to overcome an identity relationship accumulated over decades. This is why political parties frequently survive periods of poor economic performance when they remain deeply connected to religious, ethnic, regional or cultural identities.
The opposition has increasingly responded by moving towards nationalism itself. The approximately five percent of first-round voters who supported nationalist candidate Sinan Oğan became potentially important in a close runoff. Oğan subsequently endorsed Erdoğan, while Kılıçdaroğlu adopted a considerably harder tone on refugees and promised to send millions of Syrians back to Syria. The campaign between rounds has therefore become more nationalist and security-focused than the first.
This demonstrates how electoral systems shape political positions. In the first round, candidates need to construct their own constituencies. In a runoff, the objective changes: each finalist needs enough of the eliminated candidates’ voters to cross 50%. The preferences of a relatively small bloc can therefore acquire disproportionate influence because those voters become pivotal at the margin.
This resembles coalition bargaining after proportional elections, except the bargaining occurs between candidates and voters rather than parties in parliament. Oğan’s supporters do not need to be the largest political group to matter enormously. They need only to occupy the space between the two finalists’ existing coalitions.
Migration has become particularly important because Turkey hosts millions of Syrian refugees. The country initially accepted large numbers of people fleeing the Syrian civil war, but as economic conditions deteriorated, refugee policy became increasingly politically contentious. Housing costs, employment competition, public services and cultural concerns can all become attached to migration even when the causal relationship between refugees and individual economic problems is complicated.
This creates a recurring political mechanism: economic scarcity increases the salience of distributional identity. When growth is strong, citizens may tolerate competition over resources more easily. When real incomes fall, political questions increasingly become about who receives jobs, housing, welfare and state protection. Migrants become visible participants in those distributional conflicts because they can be represented as people receiving resources without belonging fully to the national political community.
Politicians then face incentives to make migration policy increasingly restrictive even when the underlying economic crisis has much broader causes. Inflation caused partly by monetary policy does not disappear if refugees leave. Housing shortages depend on construction and urban policy as well as migration. Yet migration provides a politically legible explanation for complex problems because it identifies a group rather than an abstract macroeconomic mechanism.
This does not mean every concern about Turkish refugee policy is xenophobic. Hosting millions of refugees imposes genuine fiscal, administrative and social costs, particularly when displacement becomes effectively permanent. Governments have legitimate reasons to discuss how those costs should be distributed and whether eventual return is possible. The danger begins when refugees become a substitute explanation for failures produced primarily by domestic policy.
The runoff therefore illustrates why political narratives can matter as much as economic statistics. Inflation is observable, but voters still need a causal explanation for it. Erdoğan can emphasise global shocks, the pandemic, energy prices and foreign economic pressures. The opposition can emphasise presidential interference with monetary policy and weakened institutions. Both narratives attempt to transform the same price increases into different judgements about responsibility.
Economic voting is therefore mediated by attribution. Citizens punish governments not simply when something bad happens but when they believe government caused it and that an alternative could plausibly have produced a better outcome. That is a much higher informational threshold.
The February earthquake creates the same problem. Nobody can reasonably blame Erdoğan for the earthquake itself. Voters must decide how much responsibility belongs to construction standards, building amnesties, regulatory enforcement and emergency response. Those are complicated counterfactual questions. A government can therefore survive catastrophe politically if enough citizens attribute the underlying shock to forces beyond its control or believe the opposition would not have managed it substantially better.
Erdoğan also benefits from a foreign-policy narrative in which Turkey has become more strategically independent. Ankara remains inside NATO while maintaining relations with Russia, has supplied Ukraine with weapons while refusing to join Western sanctions comprehensively, helped broker the Black Sea grain agreement and has used its position over NATO enlargement as bargaining leverage. Critics may describe this as transactional or unpredictable, but supporters can interpret it as evidence that Turkey no longer simply follows Western preferences.
That matters because foreign policy can become another form of identity politics. A voter may tolerate economic costs if they believe the government has restored national prestige. Governments throughout history have used external status partly to compensate for domestic dissatisfaction. This does not mean citizens irrationally choose nationalism over income; national status can itself be something people value.
The difficulty is that prestige is hard to measure and therefore easy for governments to claim. GDP, inflation and unemployment eventually produce observable numbers. Strategic autonomy can be interpreted differently depending on political preference. One voter sees Turkey skilfully balancing Russia and NATO; another sees diplomatic isolation and unnecessary confrontation. Political narratives therefore have greater room to shape perception where objective measurement is difficult.
Today’s election will ultimately determine more than whether Erdoğan receives another term. Turkey’s presidential system concentrates substantial executive authority in one office, and another Erdoğan victory would extend a political era that began in 2003. His longevity makes the election an institutional test because many younger Turkish voters have never experienced national politics without him at its centre.
Long incumbency changes political expectations. Bureaucracies, businesses, media organisations and foreign governments begin planning around the assumption that one leader will remain. This can create stability, but it can also weaken the distinction between the state and the person governing it. Democratic institutions are strongest when citizens can imagine the country functioning normally under either political coalition.
That is why a close competitive election matters regardless of the winner. Erdoğan does not need to lose for Turkey’s electoral institutions to demonstrate value. He needs to face a genuine possibility of losing, and the opposition needs to retain the ability to organise, campaign, monitor votes and compete again if defeated.
The same standard applies to Kılıçdaroğlu. If he loses, the opposition will need to determine whether the result reflects electoral irregularities significant enough to challenge through lawful procedures or a political strategy that simply failed to persuade enough voters. Those are different conclusions. Democratic learning becomes impossible when every defeat is attributed automatically to unfairness, just as democratic accountability becomes impossible when every criticism of electoral conditions is dismissed because ballots were counted.
Turkey can therefore contain two truths simultaneously: Erdoğan can possess genuine majority support while benefiting from an unequal political environment. Recognising the first prevents analysis from treating millions of Turkish voters as manipulated subjects. Recognising the second prevents election-day competition from becoming the only measure of democratic quality.
The larger lesson concerns economic determinism. Turkey has high inflation, a weakened currency, earthquake reconstruction costs and widespread dissatisfaction. None automatically determines an election because voters do not choose governments according to economic variables alone. They choose between political coalitions representing different combinations of material interests, identities, risks and expectations.
Opposition parties therefore cannot simply wait for governments to fail economically. They have to become credible replacements. An incumbent’s weakness creates political space; only an alternative can occupy it.
That may be the most important lesson from Turkey’s election, whatever tonight’s final result. Erdoğan’s economic vulnerabilities are substantial, but political power is not a spreadsheet in which enough negative indicators automatically produce alternation. Voters also ask who represents them, who can govern, who can protect the country and what uncertainty comes with change.
Economic performance matters enormously.
It just never votes by itself.
