There are not many institutional reforms that you can implement with results visible in a short period of time. Electoral system reform is one such reform because the overall legitimacy of the system, the degree of political inclusion of different social groups, as well as the method of conflict resolution in a society depend on the design of electoral rules. The electoral architecture will influence which politicians will represent us, which parties will be represented in parliament and whether it will be representative, who will be able to form what kind of government, what political parties will look like, and finally, what public policies our representatives will adopt. And our electoral building is built on shaky foundations.
A proportional electoral system, with a single electoral unit, closed blocked candidate lists, a 3% electoral threshold applied both to individual parties and to broad coalition lists, and affirmative action measures for parties representing the interests of national minorities and members of the underrepresented gender, produces a few good and a larger number of disastrous consequences for the functioning of our political institutions and the political system as a whole.
Is there anything good?
Serbia’s electoral system is representative, at least in terms of the share of votes and the share of mandates won by parties. This is especially true after the lowering of the electoral threshold to 3% immediately before the 2020 elections. What does that actually mean? SNS won slightly over 44% of the votes and 48% of the mandates in the April elections, the United Opposition just above 14% of the votes and 15.2% of the mandates, Moramo 4.84% of the votes and 5.2% of the mandates. To some people, proportionality (fair distribution) is not important, but there are more who consider it a key advantage of proportional electoral systems.
At the same time, the abolition of the electoral threshold for national minorities and the increase of their coefficients in the allocation of mandates by 35% (more simply, in practice, the number of votes for minority lists is artificially increased by 35%) ensures that every minority list that wins 0.3% of the votes is certain to receive at least one mandate (the share can be even lower than 0.3% if there are “wasted” votes).
Finally, the rules according to which there must be at least two representatives of the underrepresented gender for every five spots on an electoral list (currently in our country this always means women) ensure at least descriptive representation of women in national and local parliaments. And that is where the list of major advantages of the electoral system ends.
And what is wrong, and why?
The two electoral mechanisms that in conjunction produce the greatest number of defects in the functioning of political institutions in Serbia are the single nationwide electoral unit and the closed electoral list. What do we actually do at polling stations? Do we choose who will represent us, or do we merely vote for those who have been preselected by others? Whomever you voted for in April, I am almost certain you cannot name more than ten names from the electoral list you circled. Few of us even looked at the names on the electoral list. Since Serbia is constituted as a single electoral unit in which all 250 MPs are elected, and we exclusively vote for the electoral list as a whole, we end up in a situation where with a single stroke of the pen we vote for all 250 candidates on the list. Later, those 250 MPs in parliament theoretically represent all citizens of the Republic of Serbia, whether they are from Subotica, Novi Pazar, Crna Trava, Trstenik, or Vračar. Theoretically they represent everyone, but in practice they most often represent none of us. An electoral model that does not allow us to choose our representatives leads to poor communication between citizens and MPs and the complete depersonalization of those to whom we have transferred sovereignty. A single electoral unit also leads to poor territorial/geographic representativeness of the parliament and the overrepresentation of Belgrade in parliamentary seats.
And perhaps all of this would not be so problematic if the electoral system did not affect other political institutions. If it does not matter what citizens think about the (in)action of MPs, but only what political party leaders think, you will see the strengthening of undemocratic tendencies in parties, and all intra-party disagreements will end in two scenarios: an individual leaving politics or joining a new/different party. “Elected” MPs will thus sit in a parliament that serves strictly as a rubber stamp for laws proposed by the Government, and the subordinate role of parliament in the political system along with undemocratically structured parties will ultimately lead to citizen passivization and flight from politics. The defects multiply further because the same electoral model is used in parliamentary, provincial, and all local elections in Serbia.
What could be the solution?
As always, there are multiple solutions, and which one is best depends on where you want to go and what the goal of electoral reform is. You can read more about the advantages and disadvantages of different variations of the mixed electoral system, what we would gain and lose by increasing the number of electoral units, and what opening electoral lists to voters would look like at https://biramkogabiram.rs/istrazivanja-i-analize/sest-alternativnih-izbornih-modela/
.
One model, however, frequently stands out in the proposals of expert and academic circles, as well as occasionally and timidly in the reform proposals of certain politicians and political parties. The model for this system is Slovenia’s electoral system, and it would be most accurately described as a proportional electoral system with direct voting for candidates in single-member electoral districts (single-member because voters in them vote exclusively for one candidate).
How all of this looks in a simplified form in Slovenia: The lower and only important chamber of the Slovenian parliament has 90 MPs. One representative each represents the Hungarian and Italian ethnic communities, while the remaining 88 are elected across eight electoral units (11 in each). What sets the Slovenian electoral system apart is that each of these electoral units is divided into 11 small single-member electoral districts. For example, Electoral Unit 2 Postojna is divided into 11 electoral districts (Tolmin, Nova Gorica 1, Nova Gorica 2, Ajdovščina, Postojna, Ilirska Bistrica, Sežana, Koper 1, Koper 2, Izola, and Piran).
If you live in Piran, you will receive a ballot with the names of candidates in your specific electoral district and vote for one of them. Once the elections are over, the votes cast for candidates of a party across all 11 electoral districts are added together at the level of Electoral Unit Postojna. And so on for every party. The mandates, 11 of them, are distributed proportionally at the level of Electoral Unit Postojna. And so on for all eight electoral units. The electoral model, therefore, attempts to reconcile the advantages of majoritarian and proportional electoral systems: voting for candidates in single-member districts and proportional mandate distribution in electoral units.
How, however, is it determined which party candidate wins a mandate? If Party A in Electoral Unit Postojna wins, for example, four mandates, it will not be able to assign them to just anyone, but rather to those candidates of its own who achieved the best result in the electoral districts. In order for your candidate from Piran to win a mandate, they will have to defeat their party colleagues in other electoral districts, and how many of them they will need to outperform to enter parliament will depend on the number of mandates their party won in the electoral unit. Citizen support thus becomes important not only in inter-party distribution of mandates, but also in intra-party distribution. Citizens choose both the parties that will represent them in parliament and the candidates who will sit in parliamentary seats.
You can see the map of electoral units and electoral districts in Slovenia at the following link: https://volitve.dvk-rs.si/#/rezultati
.
How could all of this look in Serbia?
The consequences of an electoral system depend on the way different electoral mechanisms are combined. Even if there were agreement on a desirable electoral model, within it, dozens of major and minor adjustments can be made depending on what we want to achieve through reform. In normal democratic states, these adjustments are carried out through a broad public debate involving all interested parties, all with the goal of obtaining the best possible solution.
By way of illustration, Serbia could be divided into seven electoral units – Belgrade, Vojvodina, Šumadija, Western Serbia, Southern Serbia, Eastern Serbia, and Kosovo and Metohija (the number of electoral units could be greater or smaller than seven). In these units, an equal number of mandates would not be assigned as in the eight Slovenian ones; rather, it would be proportional to the number of voters/residents living in them. In Vojvodina, 25.6% of the total electorate is registered on voter lists, so that electoral unit would elect 64 MPs to the National Assembly, Belgrade would elect 61 MPs, etc. Each of the seven electoral units would be divided into single-member electoral districts (fears of so-called gerrymandering are not justified in this situation because mandates are still distributed at the level of electoral units, so parties cannot profit from redrawing district boundaries). Thus, in Vojvodina there would be 64 electoral districts. In the territory of Novi Sad there would be 12, Subotica five, Zrenjanin and Pančevo four each, and Bačka Topola one.
If you live in Bačka Topola, you would receive a ballot with the names and surnames of your fellow citizens and circle one of them. Voters in Bačka Palanka will make their decisions guided by different motives. Some will vote specifically because of the candidate and their personal traits. Some might not be thrilled with the candidate, but will circle them anyway because of the platform of the party that nominated them. A third group won’t be thrilled with the candidate, will know nothing about the party platform, but will make their decision out of love for the party leader. However this trilemma is resolved, parties will be motivated to field better candidates because a portion of the votes they win will depend on whom they have offered us.
The number of mandates won by each electoral list would be allocated as before, using the D’Hondt method with a natural threshold and a 35% quotient increase for minority parties (though other mandate allocation methods could also be used). Let us assume that the number of votes won by each list in elections under the new model in Vojvodina was the same as in the April elections. The list centered around SNS would have 33 mandates, United Opposition nine, SPS six, SVM five, Nada three, Dveri, Zavetnici, Moramo, and the minority list Together for Vojvodina two each. Who do those mandates belong to? Unlike in the April elections, under the new model, the Together for Vojvodina list would not be able to assign its two mandates to Aleksandar Olenik and Tomislav Žigmanov unless the two of them were precisely the ZZV candidates with the highest share of votes in the electoral districts in which they ran. Since ZZV achieved its best results in April in the municipalities of Bač, Bački Petrovac, Kovačica, Subotica, and Bela Crkva, the two ZZV candidates with the best results would likely come from those municipalities. Unless ZZV nominated someone in another municipality who inspired trust in a larger number of fellow citizens. The Dveri-POKS list achieved an above-average result in Sombor, so one elected MP would likely be Žika Gojković. Unless Dveri made an effort to find better candidates in other single-member districts who would outperform Gojković’s result. The electoral model “rewards” the party’s most successful candidates, which motivates parties to seek the best possible individuals.
What happens to the advantages of the current electoral model after the reform?
As you have noticed, the new electoral system remains representative. Parties that cross the 3% threshold will win mandates in all electoral units, except Kosovo and Metohija. The electoral model would refresh party dynamics at the regional level, decentralize parties, and also make it easier for political parties without national coverage to enter parliament. For example, the Do Not Let Belgrade Drown movement would not have to worry about a national threshold and subsequent entry into parliament, because a result of around ten percent in the Belgrade electoral unit would be sufficient to win 7-8 mandates.
Representation of national minorities would also not be compromised. Although the natural electoral threshold is higher when a smaller number of mandates is assigned (in the Vojvodina example above it is 1.34%), Hungarian, Bosniak, Albanian, and Slovak parties would cross the natural threshold in their respective electoral units (moreover, the D’Hondt formula and the 35% quotient increase could be replaced by the Sainte-Laguë formula, which would further lower the natural threshold and favor smaller parties).
The achieved level of parliamentary representation of women could be compromised because male candidates could potentially achieve better results. This shortcoming, however, is easily solvable and is already used in comparative practice with the implementation of open electoral lists. If it were to happen that of the five mandates assigned to SVM in Vojvodina all five, based on votes, belonged to male candidates, the same rule currently applied during the nomination process could be applied to intra-party mandate distribution. The first three mandates would be won by the men with the highest number of votes, and the fourth and fifth would go to the top-ranked women. The need for such a solution would almost certainly be temporary, as female candidates are already in no way lagging behind their male colleagues. The reform, on the other hand, would strengthen the position of those elected, so descriptive representation of the underrepresented gender would be supplemented by substantive representation. Instead of 40% women in parliament, we would get 40% women in parliament elected directly by voters’ votes. The same, of course, applies to men.
What happens to the shortcomings of the current model after the reform?
The previous “dynamics” in political parties and political institutions would be radically altered. The institutional incentives that the new model sends to all electoral actors (voters, candidates, elected MPs, political parties, party leaders) are completely different. Although some voters will likely continue to vote after the reform because of Vučić, Đilas, Dačić, Jeremić, Obradović, Jovanović… some of them will not. If you want to achieve the best possible result in an electoral unit and win the highest number of mandates, you will be motivated to seek out the best candidates for single-member electoral districts.
Citizens will finally be able to choose who represents them in parliament, MPs will be empowered by voters’ votes and will be connected to the narrower territorial unit from which they were elected. Intra-party competition for mandates will further refresh the almost completely stagnant party scene in Serbia. Questions of party leadership will partly be resolved in parliamentary elections as well. If, as a currently irreplaceable party leader, you achieve a disastrous result in your single-member electoral district, it will be clear that it is time to yield the position to someone else.
The new model would also correct the previous distortion of territorial representation. Not only would Belgrade have 61 MPs, but economically underdeveloped local communities within other electoral units would finally get the opportunity to have representatives. The new electoral rules would strengthen parliament relative to the Government and the President of the Republic, and the opportunity to compete in elections in one’s local communities would attract new young faces into politics. Such people would become desirable not only to voters, but also to rational individuals within political parties, because every good candidate who is ready and capable of organizing a good local campaign would increase the number of votes and mandates that would accrue to the political party.
Some reform risks
There is no perfect electoral model. If one existed, there would likely be no debates around the world about which electoral system is best. A proportional electoral system with direct candidate voting in single-member electoral districts eliminates the key shortcomings of the current electoral model without compromising its advantages. However, it potentially introduces several new problems, some of which are characteristic of majoritarian systems, while others are its specific characteristics:
Voting for individuals in relatively small electoral districts would inevitably raise the question of strengthening clientelistic practices, but this time at the local level. If winning a parliamentary mandate depends on a few hundred or even dozens of votes, it is to be expected that certain candidates will try to exploit this. However, it is not as if the current electoral model prevents vote-buying practices, so that the new one would cause a disaster in that regard.
Certain political parties would face challenges in the process of candidate selection and nomination across a large number of electoral districts. To achieve the best possible electoral result, it is beneficial to have candidates in all single-member electoral districts at the level of the electoral unit. Some political parties are likely unable to do this. However, that should not primarily be a problem for citizens to worry about, but rather party leaders who will be incentivized to strengthen their party organizations. And if they are incapable of doing so, we will have an organic consolidation of the party scene. If you don’t have 250 quality candidates for 250 single-member districts, you will have to unite. Or continue as before and suffer electoral defeats.
Certain single-member districts will not be represented by candidates who won the highest number of votes, perhaps not even the runner-up, which is expected because otherwise it would be a majoritarian electoral system. Also, a smaller number of single-member electoral districts would be left without a representative, while some would have two. However, if it were stipulated that the ranking on the post-election intra-party list is determined based on the candidate’s share of votes out of the total number of registered voters on the voter list in each single-member district—rather than those who voted—local communities with higher voter turnout would be rewarded with a larger number of MPs. Turn out to vote in as large numbers as possible, and your municipality will be represented in parliament by two or perhaps even three MPs. Stay home on election day, and there will be no one to represent your local community in parliament.
Finally, the key risk relates to the context in which we find ourselves. There is no electoral system that will function well in an undemocratic political environment. Even if we imagine that the key political actors—ruling coalition and opposition—reach agreement on the desired direction of changes, questions remain as to how and who will conduct the elections, who will count votes at polling stations, and complete the official minutes of polling boards. For this reason, changing the electoral system would have to be accompanied by the professionalization of the electoral administration, as well as changes to a set of laws directly or indirectly related to elections.