Abstract
The results of the 2024 Turkish municipal elections surprised many with CHP, the main opposition party, leading the polls for the first time in 35 years and lower turnout rates than one is used to observe in Turkish elections. This paper studies the city council elections, to avoid as much as it can the effects of electoral alliances, in a comparative way, taking the previous municipal elections of 2019 as the base. For each metropolitan city/province, expected votes were calculated based on the expansion of the electoral base and compared with the obtained votes in the 2024 elections.
Keywords: elections, Turkish elections, voter abstention, political science, voter behavior, Turkish politics, municipal elections, local elections, city councils.
1. Background and Questions
The results in the 2024 local elections in Turkey[1] surprised many. Despite the victories of Erdogan and his AK Party since 2001 and the loss of the opposition’s presidential candidate, the main opposition party CHP (Republican People’s Party) lead the election results for the first time in the last 35 years (Sezer & Karakas, 2024). CHP candidates were elected mayors of 36 of the 81 provinces in the surprize win (McLean et al., 2024). The elections also witnessed a low turnout rate of around 76%, compared to last years’ 87% (Fraser & Kiper, 2024). Given the low turnout rate and astounding results, many people have wondered if the elections reflected the opposition’s success, government voters abstaining from voting or a bottom-up alliance resulting from smaller party voters voting for the main parties. This paper tries to compile data in a relevant manner to answer these questions.
In Turkey, geographically and politically, metropolitan cities and provinces are mutually exclusive and exhaustive in the country, whereas districts are mutually exclusive and exhaustive in the metropolitan city or province to which they belong. The mayor of a province is elected by the people in the central district of the province with the same name, whereas the mayor of a metropolitan city is elected by the entire metropolitan population. The status of being a metropolitan city is awarded to a province by law, according to its population.
Elections are carried out on a Sunday and are divided into two broad categories: general and local. During general elections the public votes for the president and the parliamentarians, whereas during local elections it votes for the mayor, the municipal council and other posts regarding the administration of the neighbourhood or the village. The two elections under the general ones and those under the local ones each have a distinct ballot, thus voting for candidates of two different parties for different posts is possible.
2. Data and Methods
Given the situation described in the previous chapter, there was a need to analyse the data to further contemplate on the various arguments. First, the data regarding turnout rates were studied to see if in these elections a smaller fraction of people voted than in previous elections. Then, the fraction of votes and the number of municipalities the parties won were studied to see if these elections were in fact different from the previous ones.
To see if a party got more votes in a locality (metropolitan city, province or district) than before, their current votes in localities were compared with their expected votes. The expected votes were calculated as $E(votes_t)=votes_{t-1}\cdot\frac{\text{voting population}_t}{\text{voting population}_{t-1}}$ for each locality. Therefore, $votes_t-E(votes_t)>0$, or equivalently $\frac{votes_t}{E(votes_t)}>1$ means that the party’s votes outgrew the voting population, whereas $votes_t-E(votes_t)<0$, or equivalently $\frac{votes_t}{E(votes_t)}<1$ means that the voting population outgrew the party’s votes. The higher this fraction, the more a party is assumed to have outperformed itself compared to the previous election. This comparison was done with regards to with regards to 2019 elections (the ones before the most recent in 2024). The city council elections are used for both election periods as for the mayoral elections in 2019 there were unified alliance lists in many cities and in 2024 many people voted for the stronger candidate in the government or the opposition rather than voting their minds.
Political scientists count “making people vote” as one of the responsibilities of political parties (Clarck et al., 2017, p. 759). Relevantly, this index depends not on the fraction but the number of voters who voted for a party. In such a way, this study conducted around this index is believed produce a satisfactory analysis of whether a party outperformed itself compared to the previous election or not. Of course, this analysis cannot bring any dispute to the results obtained in the election. All parties and candidates contested in the same legal and political system, and even if, for example, two parties are both found to have underperformed themselves, one might have underperformed itself more than the other and thus have lost an election.
All the data subject to this study has been obtained directly from the official website of the Open Data Portal of the Supreme Election Council of Turkey, the highest electoral authority in the country (Supreme Electoral Council, 2024).
Only the parties that won at least one metropolitan city or provincial government in the 2024 municipal elections are studied in this paper and they are listed in order of decreasing total obtained votes in the election of the mayors. As an exception, Yeniden Refah (New Welfare Party) was not suitable to the study as it did not contest in the 2019 municipal elections. Additionally, the results of HDP (Peoples’ Democratic Party) in the 2019 elections are used for DEM Party as DEM Party is widely seen and accepted as the political continuation of HDP in Turkey given the frequency of changing parties in the movement to overcome legal obstacles.
3. Comparison on the National Level
Below, one can see the differences between and fractions of the votes a party obtained in the 2024 election and their calculated expected votes ($votes_t-E(votes_t)$ and $\frac{votes_t}{E(votes_t)}$, respectively) for each of the parties subject to the study.
3.1. CHP (Republican People’s Party)
CHP is the only party subject to this study to have obtained a more votes than it was expected to. They obtained 1,501,959 votes more than they were expected to. Fraction wise, it amounts to 10.959% more than their expected votes.
3.2. AK Party (Justice and Development Party)
AK Party is the party that lost the greatest number of votes, but not the greatest fraction, compared to its expected votes. They obtained 5,983,150 votes less than they were expected to. Fraction wise, it amounts to 30.119% less than their expected votes.
3.3. MHP (Nationalist Movement Party)
MHP obtained 969,997 votes less than they were expected to. Fraction wise, it amounts to 27.842% less than their expected votes.
3.4. DEM Party (Peoples’ Equality and Democracy Party)
DEM obtained 200,229 votes less than they were expected to. Fraction wise, it amounts to 7.655% less than their expected votes.
3.5. IYI Party (Good Party)
IYI Party is the party that lost the greatest fraction of votes, but not the greatest number, compared to its expected votes. They obtained 1,442,077 votes less than they were expected to. Fraction wise, it amounts to 42.27% less than their expected votes.
3.6. BBP (Great Unity Party)
BBP obtained 362,684 votes less than they were expected to. Fraction wise, it amounts to 40.8% less than their expected votes.
4. Comparison on the Municipal Level
Below, one can gradient maps showing the differences between and fractions of the votes a party obtained in the 2024 election and their calculated expected votes ($votes_t-E(votes_t)$ and $\frac{votes_t}{E(votes_t)}$, respectively) for each of the parties subject to the study. In both cases, the colour white is used in case of no difference (0 in difference and 1 in fraction). Detailed data are available with the paper on request in CSV format.
4.1. CHP (Republican People’s Party)
4.2. AK Party (Justice and Development Party)
4.3. MHP (Nationalist Movement Party)
4.4. DEM (Peoples’ Equality and Democracy Party)
4.5. IYI Party (Good Party)
4.6. BBP (Great Unity Party)
5. Results and Conclusion
Despite some values not available and some outliers, especially in fractional maps, possibly due to alliances even in city council elections, the data that this paper produced is believed to be meaningful to understand electoral dynamics of Turkey. Especially combined with the differential maps, the paper allows one to observe the rise in the popularity of CHP and the fall in that of the other parties subject to the study. Overall, the data produced by the methods of this paper may be used as a source to many other analyses, most notably provincial ones.
Cited Works
Sezer, C., & Karakas, B. (2024, April 1). Turkey’s resurgent opposition trounces Erdogan in pivotal local elections. Reuters. https://www.reuters.com/world/middle- east/erdogan-battles-key-rival-turkeys-local-elections-2024-03-31/
McLean, S., Tawfeeq, M., & Ebrahim, N. (2024, April 1). Turkey’s Erdogan dealt major election blow as opposition party wins Big Cities. CNN. https://edition.cnn.com/2024/04/01/middleeast/turkey-local-election-blow-erdogan-opposition-mime-intl/index.html
Fraser, S., & Kiper, C. (2024, April 1). In setback to Turkey’s Erdogan, opposition makes huge gains in local election. AP News. https://apnews.com/article/turkey-local-elections-erdogan-popularity-test-8adc4df56cdbce24f78900e1d142ff21
Supreme Electoral Council. (2024). Open Data Portal. https://data.ysk.gov.tr/anasayfa
Clarck, W. R., Golder, M., & Golder, S. N. (2017). Principles of Comparative Politics (3rd ed.). Sage.
[1] In this paper the use of Turkey is preferred over that of Türkiye as the author does not believe in political authorities deciding how something must be called by others.