I would like to thank Fabio Wolkenstein for his kind response to my comments on his paper “Epistemic Barriers to Rational Voting: The Case of European Parliament Elections” (2020). Below, I will make several clarifications and expansions for my previous comments in response to Professor Wolkenstein’s response… [please read below the rest of the article].
Aksoy, Faruk. 2020. “Further Remarks on ‘Epistemic Barriers to Rational Voting: The Case of European Parliament Elections’.” Social Epistemology Review and Reply Collective 9 (9): 40-45. https://wp.me/p1Bfg0-5mm.
🔹 The PDF of the article gives specific page numbers.
❧ Wolkenstein, Fabio. 2020b. “Further Thoughts on ‘Epistemic Barriers to Rational Voting:’ A Response to Faruk Aksoy.” Social Epistemology Review and Reply Collective 9 (7): 1-5.
❦ Aksoy, Faruk. 2020. “Rational Voting in European Parliament Elections: A Critical Reply to Wolkenstein.” Social Epistemology Review and Reply Collective 9 (4): 10-15.
❦ Wolkenstein, Fabio. 2020a. “Epistemic Barriers to Rational Voting: The Case of European Parliament Elections.” Social Epistemology 34 (3): 294-308.
As was indicated by Professor Wolkenstein, I raised three issues in my earlier reply. It seems that we agree on the conception of rationality, and I agree with him that the issues that I raised on the definition of rational voting pose no problem for the discussion on the scenarios and the remaining analyses. Therefore, I will elaborate on the other two issues; namely, the relationship between strategic voting and epistemic barriers, and the transnationalisation of European Parliament elections.
Strategic Voting and Epistemic Barriers
Regarding my argument on the strategic voting, Professor Wolkenstein indicated that “… I agree that I should have spent more time discussing such cases, it seems to me that I could equally aim at very specific results in normal voting, where I vote for my preferred party. Whenever that is the case, then there is no relevant difference between strategic voting and normal voting as far as voters’ epistemic burdens are concerned.” (Wolkenstein 2020b, 4). Also, he said that “ One argument for why strategic voting is more epistemically demanding than normal voting could be that, when I vote for a non-preferred party in order for that party to perform a particular function that I deem important, I need to make some judgments about the likely behavior of an agent that I may not know well” (Wolkenstein 2020b, 4).
I agree with his argument about gathering information about non-”most” preferred parties is epistemically demanding in case of voting strategically. Nonetheless, as he also pointed out, my argument on the severity of the epistemic barriers for strategic voting refers to the calculation and inclusion of the expected vote share of parties to the voting decision, which corresponds to the first quote in the previous paragraph. I have realized that my argumentation was not explicit enough. Let me unpack my thoughts on the relationship between epistemic barriers and strategic voting in a more structured manner.
To analyze the strategic voting in any context, two overarching questions might be asked with respect to voters’ expectations for the vote share of parties. Note that the literature offers varieties of tactics for strategic voting, which assign different motivations to voters. Some examples might be to protest or send a message to the most preferred party (Meirowitz and Tucker 2007) or to help another party to exceed an electoral threshold (Cox 1997; Freden 2014). Nonetheless, for the parsimony, I will focus on not wasting the vote as the voters’ motivation, following the classical definition of strategic voting (Cox 1997; Lago 2012). Additionally, I will restrict the correlates of voting behavior as voters’ preference ranking for parties and their expectations for the parties’ vote shares.
Let me start with the first question: Whether voters expect that their most preferred party has a realistic chance to win the election or not ? If the voters expect that their most preferred has a realistic chance to win the elections, then they do not have an incentive to vote for any other party if we constrain the idea of strategic voting in the classical sense. So, the necessary condition for strategic voting is that voters assume no realistic chance to their most preferred party for winning the election. This necessary condition might sound deterministic, and I admit that it is. Using an expected utility model in which voter’s preference for the party and voters’ expectations for winning interact for each party might be mushrooming; yet, my aim here is, firstly, to underline the unique status of the most preferred party while presenting my argument more clearly without overextension.
The second overarching question is whether voters include their expectations for parties’ vote shares in the calculus of their voting decision or not. In this respect, a group consists of voters who vote for their most preferred party without considering any condition other than their party affiliation. These are partisan and “habitual” voters who vote for the same party in each election, and the primary correlate of their voting behavior is their party affiliation, so their preference ranking for parties. Meanwhile, another group consists of voters who include calculations regarding their expectations for the vote shares of parties to decide which party to vote for. In other words, their voting calculus depends both on their preference ranking and expected vote shares for parties.
This brief classification produces four types of voters.
❧ The first type includes voters who do not use their expectations for the vote shares of parties and assumes that their most preferred party has a realistic chance to win the elections. Under these conditions, the viable option for those voters is a sincere vote, so voting for their most preferred party.
❧ The second type consists of voters who do not use their expectations for the vote shares of parties and assumes that their most preferred party does not have a realistic chance to win the elections. Under these circumstances, since voters do not include the expected vote shares into their decision-making calculus; in fact, they do not prefer to make these calculations, the viable option for those voters is again a sincere vote.
❧ The third type represents voters who use expected vote shares of parties to decide whom to vote for and predicts that their most preferred party has a realistic chance to win the elections. For those voters, the options of sincere and strategic vote correspond to each other. In the literature, it is defined as a straightforward vote (Farquaharson, 1969; Kriesi, 1998 ).
❧ The fourth type includes voters who use their expectations for the vote shares of parties and anticipate that their most preferred party does not have a realistic chance to win the elections. In fact, these are potential strategic voters.
My point in the previous reply anchored on the third and the fourth types of voters. Arguably, the third and the fourth types of voters are those who face epistemic barriers on rational voting. Nevertheless, epistemic barries are more challenging for the fourth group of voters (potential strategic voters) in comparison to the third group of voters (whose voting behavior can be defined as “straightforward”). To better explain my point, It is required to clarify the conception of “winning.” Therefore, I will continue my argumentation with two alternative conceptions of “winning” with respect to different electoral rules.
The first electoral rule in which I will discuss the conception of winning is Single Member District (SMD), which refers to a system in which magnitude is 1 for all districts. Simply, in SMD elections, the “winning” means gaining the only seat of the district. In this regard, voters might apply their expected vote share for parties in their voting decision in two steps. Initially, they need to determine whether their most preferred party has a realistic chance to win the seat or not. If their answer is yes, then they belong to the third group and, as “straightforward” voters, they do not need to make additional calculations. In other words, they only need to be sure that their most preferred party will be at least the runner up in the elections. However, if their answer is no, then they are required to make further calculations to decide which party they might vote for. They need first to estimate the probability of being in an efficacious set if they will vote strategically. Concomitantly, they need to estimate the chance of winning of other parties with respect to their preference ranking. For instance, they may require to compare the chance of their second most preferred party and their third-most preferred party to win the elections before deciding to vote. Therefore, the fourth group of voters is more likely to need additional calculations, which makes them fragile against the epistemic barriers more.
The calculus for strategic voting turns out to be more complicated if the elections are held under a proportional representation rule (PR). In the electoral systems in which the seats are allotted with respect to the vote shares proportionally, as like European Parliament elections, the definition of the “winning” turns out to be less straightforward.
Under these circumstances, at first glance, epistemic barriers seem equally severe for the third and fourth groups of voters. Since multiple seats (instead of one) will be allotted to the parties (instead of the party that first past the post) with respect to the volume of their vote share, voters should have information on the rules of aggregation and allotment, parties’ vote shares, the volume of vote share differences between parties. Therefore, one may argue that the third groups of voters have to consider all options simultaneously to calculate the allotment of multiple seats; hence, epistemic barriers are equally severe for them. Nevertheless, especially for partisan small party voters, I think epistemic barriers affect them more if they are the fourth type in PR systems, similar to the SMD elections. Let me illustrate my point with an example.
Suppose a highly affiliated voter to a party that has the potential to gain only one seat in a district in which district magnitude is 7. Probably, “winning” for this voter means that a seat gain by her most preferred party. Since the number of potential seats which can be gained by her most preferred party is 1, her calculus for voting is similar to an SMD election. So, if this voter expects that her most preferred party might gain this potential seat in the elections, then she is the third type of voter. On the other hand, if she thinks that her most preferred party cannot gain a seat, then she is the fourth type of voter. Under these circumstances, the only calculation that she needs to make is whether her most preferred party will gain a seat or not to be a third type of voter. However, if she is a fourth type of voter, which means that she, firstly, anticipates her most preferred party cannot gain a seat, she needs to make additional calculations with respect to her preference ranking and expectation for vote shares, according to which, 7 seats will be allotted.
As I have mentioned earlier, this classification is deterministic; therefore, one may argue that a probabilistic model, like expected utility, may result that the third and the fourth groups of voters suffer from the epistemic barriers at a similar level, especially under PR rule. My response to this is the following: I think deciding to cast a strategic vote is a two-level calculation process; instead of calculating all voting options simultaneously. It means that voters are reflexively inclined to vote for their most preferred party, especially if their party affiliation is sufficiently high. If they realize that their most preferred party cannot win the elections, then they consider voting for other parties with respect to their preference and expected votes shares.
Transnationalising European Parliament Elections
Professor Wolkenstien’s question that “who is this ‘we’ “ is a clear indication of the lacuna in my question, which was “ In what ways can we encourage the political actor (mainly individual politicians and political parties) to cooperate more in transnational party groups” (Aksoy 2020, 14), and I think, his answer to “who is this ‘we’ “ perfectly depicts the problems to encourage the transnational party groups through institutional innovations alongside hardship of engaging the same task through a bottom-up process.
My intention to come up with this question was theoretical. In other words, I implied a theoretical discussion on the effect of political institutions on alleviating epistemic barriers by incentivizing cooperation in transnational party groups. Yet, in his reply, Professor Wolkenstein rightfully underlines another essential aspect, which is the impediments on institutional innovations and the bottom-up process for encouraging transnational party groups. I agree with him in terms of obstacles and problems that he suggested. Nevertheless, let me explain why I find a theoretical discussion on institutional innovation productive, particularly under the structure of the European Union; and let me say a couple of words about its practical implications.
I partly agree with Professor Wolkenstein in terms of the top-down transnationalisation of European elections, in the sense that, for instance, unilateral actions of the European Commission unsound normatively and practically. In fact, I think an exclusive feature of the European Commission (EC), in comparison to the executive branch of a state, opens an avenue for institutional innovations that diminish the epistemic barriers without consulting unilateral action. This is; independence of the formation of the European Commission from the EP elections. This feature of EC, and its’ substantial power on the legislative initiative, alter a fundamental dynamic of the process of institutional change.
Arguably, institutional innovations, or in a more general sense setting new institutions, occurs in an asymmetric structure. Most of the time, the rules of the game (institutions) are set or changed by those who have played (and will play) the game. In this regard, any institutional change reflects the balance of power among the actors who are entitled to take part in the decision-making process of institutional change. Of course, it does not necessarily mean that the powerful actors will benefit more from new institutions due to several reasons such as unintended consequences of institutions and probable altruistic or principled behavior of actors. Nevertheless, the literature on institutional change suggests that the new institutions are likely to represent the interest of the actors with mirroring their relative power (Hall and Taylor 1996; Negretto 2012). Furthermore, of course, the probability of institutional change increases when the interest of the powerful actors cannot be pursed by current institutional settings.
In this regard, EC has the power to shape a game, the EP elections, in which it is not a direct player. The agenda-setting power of EC, which entitles it to propose legislation and its partial independence, pave the way for preparing legislation proposal with a lesser impact of member states and political parties- on the other hand- whose support is necessary to enact the proposals. Note that it does not mean that EC does not have an interest or not an actor in E.U politics; contrary, theoretically, EC has the interest to expand the impact and relevance of the E.U by decreasing the epistemic barriers in EP elections, as Professor Wolkenstein’s example on EC’s proposal to introduce transnational party list shadow out. Obviously, the rejection of the proposal by the European People’s Party shows that even though the E.C has partial independence from member states and parties, any proposal should take account of their interests. Yet, it also shows that E.C has the potential and interest to make institutional innovations to decrease the epistemic barriers. Proposals’ probability of turning legislation might depend on the timing of introducing proposals and proper calculations and considerations for other actors’ interests that enable E.C to form a minimum winning coalition within the E.U to enact the proposal.
I thank the SERRC for providing this opportunity of deliberation and Fabio Wolkenstein for engaging in this discussion with me.
Faruk Aksoy, Sabancı University, firstname.lastname@example.org
Aksoy, Faruk. 2020. “Rational Voting in European Parliament Elections: A Critical Reply to Wolkenstein.” Social Epistemology Review and Reply Collective 9 (4): 10-15.
Cox, G.W. 1997. Making Votes Count: Strategic Coordination in the World’s Electoral Systems: Cambridge University Press.
Farquharson, Robin. 1969. Theory of Voting. Blackwell.
Fredén, Annika. 2014. “Threshold Insurance Voting in PR Systems: A Study of Voters’ Strategic Behavior in the 2010 Swedish General Election.” Journal of Elections, Public Opinion and Parties 24 (4): 473-492.
Hall, Peter A., and Rosemary C. R. Taylor. 1996. “Political Science and the Three New Institutionalisms.” Political Studies 44 (5): 936-957.
Kriesi, Hanspeter. 1998. “Straightforward and Strategic Voting in the Elections for the Swiss Council of States in 1995.” Electoral Studies 17 (1):45-59.
Lago, Ignacio. 2012. “Strategic Voting in Proportional Representation Systems:Evidence from a Natural Experiment.” Party Politics 18 (5):653-665.
Meirowitz, Adam, and Joshua A. Tucker. 2007. “Run Boris Run: Strategic Voting in Sequential Elections.” Journal of Politics 69 (1): 88-99.
Negretto, Gabriel L. 2012. “Replacing and Amending Constitutions: The Logic of Constitutional Change in Latin America.” Law & Society Review 46 (4): 749-779.
Wolkenstein, Fabio. 2020a. “Epistemic Barriers to Rational Voting: The Case of European Parliament Elections.” Social Epistemology 34 (3): 294-308.
Wolkenstein, Fabio. 2020b. “Further Thoughts on ‘Epistemic Barriers to Rational Voting:’ A Response to Faruk Aksoy.” Social Epistemology Review and Reply Collective 9 (7): 1-5.
Categories: Critical Replies