![Sample our Politics & International Relations journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5947/TOC-Subject-Sample-2021-Politics-International-Relations.jpg"})
ABSTRACT
In response to widespread voter ignorance, Jason Brennan argues for a voting ethics that can be summarized as one negative duty: do not vote badly. The implication that abstaining is always permissible entails no incentive for citizens to become competent voters or to vote once competent. Following the Condorcet Jury Theorem, this can lead to suboptimal outcomes, suggesting that voter turnout should concern instrumentalist epistemic accounts of democratic legitimacy. This could be addressed by adding two positive voting duties: to make an effort to become a competent voter; and, once competent, to vote.
Notes
1 It bears clarifying that by narrowing the scope of this paper, I do not intend to identify myself as a pure instrumentalist.
2 For recent overviews of empirical research documenting voter ignorance, see Somin Citation2013, Brennan Citation2016, and Friedman Citation2019, ch. 6.
3 See Friedman Citation2019, 282-83.
4 If voters cannot adequately assess their own competence, both Brennan’s voting ethics and the positive voting duties I argue for might be seen as self-undermining. I will say more about this in the last section of the paper, where I reflect on the realism of these theories.
5 For example, deliberative democrats argue that there should be more facilitation of public debate, while Kristoffer Ahlstrom-Vij (Citation2012), for one, advocates information markets as a means of combating public ignorance.
6 See Blais and Achen Citation2019 for an overview of how the concept of civic duty appears in the literature on voting behavior, and for empirical results supporting the notion of civic duty promoting voter turnout.
7 This excludes the aforementioned exception of a minimum voting age.
8 Later, in Against Democracy (2016), Brennan argues against rule by the people and in favor of epistocracy, or rule by the knowledgeable. However, in The Ethics of Voting (Citation2011), he still accepted the democratic framework.
9 The criterion for justified beliefs shouldn’t be conflated with the claim that any justified belief about political candidates makes one politically competent. One must have an understanding of how the belief relates to the candidate being more likely to benefit the common good.
10 Paul Gunn (Citation2019) rightly criticizes Brennan for often portraying certain views as obviously right when in fact social scientists are in disagreement about them, the protectionist versus free trade debate being one of these. I agree with Gunn regarding Brennan’s negligence in not addressing such disagreements among experts, and that this undercuts his argument for epistocracy in Against Democracy. However, I don’t think this omission affects the core of Brennan’s claim in The Ethics of Voting, where Jane is considered a “bad voter” not because she is wrong, but because her beliefs are unjustified. In the example cited, Jane is consistently confronted with counterevidence against the claim that protectionist economies are effective in promoting the common good and encounters no evidence in favor of protectionist policies. This makes it epistemically vicious for her to hold onto her conviction that protectionist policies are effective. This leaves open the possibility that Jane might actually be right or that there is also good evidence in favor of protectionist policies available, even if Jane hasn’t encountered it. I agree with Gunn that this question is more nuanced than Brennan acknowledges. Still, the fact remains that Jane has no justification for her convictions at this moment, and thus should not vote in favor of the protectionist candidate. In short, I agree with Brennan that voters should have epistemically justified reasons for favoring one candidate over the other, but this is not the same as claiming that one should possess all relevant information or be right, as demonstrated by the examples of Elly and Tommy. I also agree with Gunn that justified political beliefs needn’t necessarily point in the same direction, nor that social scientists are immune to cognitive errors. Accepting Brennan’s negative voting duties need not commit me to his later argument in favor of epistocracy, which I do not endorse.
11 This is not to say that all proceduralists are concerned with participation rates, but to point out that those who are concerned with participation rates are typically proceduralists rather than instrumentalists.
12 For example, consider the principle of fairness: why value fair decision making if it does not lead to a manifestation of fairness in society because, for instance, a majority of voters are not concerned with distributive justice? Under Brennan’s scheme, this discrepancy would not appear, since an account of distributive justice must be incorporated into a justified interpretation of the common good.
13 Note that by explaining voter ignorance through a lack of incentives, Downs (like Downsian scholars after him) implicitly treats ignorance as something that is permanent. Brennan does seem to accept the theory of rational ignorance (Brennan Citation2011, 33). More specifically, citizens are not punished for voting based on irrational beliefs (Brennan Citation2016, 23). However, the fact that he provides suggestions on how to counteract biases and thereby remedy some ignorance suggests that Brennan does not take ignorance to be permanent, in the sense of incurable.
14 For a helpful and more elaborate explanation of the theorem, see Goodin and Spiekerman Citation2018, 17-22.
15 Following the CJT, this would imply that an increase of N approaches certainty that the majority chooses the lesser option.
16 E.g., according to Fabienne Peter (Citation2016), democratic decision making seems appropriate exactly when there is no third-personal authority who can identify the correct decision. I will say more about the absence of political “truth” about outcomes in the final section of this paper.
17 This might seem an arbitrary cutoff, as it remains a vague notion at this point what it means to be competent, i.e., what type of level of political knowledge is required of citizens. One might take issue with the assumption that this competence is attainable at all (see concluding remarks for more on this). I do not mean to go into this question in much detail here. In this section, I merely aim to show that whatever competence entails, if we assume this competence is needed for voting well, there is a case to be made for the epistemic value of having more people reaching this competence and for exercising this competence by voting.
18 Note that Brennan’s notion of excused harmful voting implies that epistemically justified votes (made by competent voters) can still be mistaken (Brennan Citation2011, 69-70). For this reason, we should be concerned with voter turnout in relation to maximizing the possibility of the right option being selected.
19 As noted above, for a majority decision-making process to lead to the right decision, it is sufficient that p > (1/k) and not necessarily that p > 0.5. The level of competence required to reach epistemically best outcomes might be lower depending on the number of voting options available. This seems to imply that under different conditions, different people have an obligation to vote (e.g., if there are three options, everyone with p > 0.34 will contribute to the likelihood of the best option being chosen). However, since this paper also complies with Brennan’s idea that voting well requires making epistemically justified decisions, an independent value of p > 0.5 seems a more appropriate threshold than a dependent value of p > (1/k). Additionally, the threshold of p > 0.5 provides more incentive to those citizens who live in democracies where there are many options available (with no two-party system) to improve the epistemic quality of their political beliefs.
20 This move echoes Julia Maskivker’s argument for a duty to vote well as a Samaritan duty of aid (Maskivker Citation2018).
21 I will say more about this in the concluding remarks.
22 To be more precise: Brennan (Citation2009, 545) seems to acknowledge this when he states that by abstaining from voting we don’t give up our political power: we are allowing those who know better to make the decision. However, while the impact of one bad vote is morally significant enough to call for a negative voting duty, apparently the impact of one good vote is not enough for Brennan to call for a positive voting duty, even though the impact of one vote among the competent voters (a minority of citizens, according to Brennan) is higher relative to the impact of one bad vote among all citizens.
23 I must stress that the latter duty refers to actual votes and not blank votes, which do not improve the epistemic quality of the outcome. When there is no “good” option available, the lesser of two evils is still the optimal option available. I should also mention that, as pointed out by Valeria Ottonelli in response to an earlier version of this paper, one could wonder whether the positive duty to vote when competent could be conditional on the competence of other voters. Consider a quorum with an average of p = 0.8. In such a case, the addition of a less competent vote (e.g., with p = 0.6) may lower the group’s probability of achieving the right outcome by majority voting. Should this voter therefore refrain from voting? I am inclined to say no. I suspect that the range in which this adverse effect will occur would be very narrow. Accepting, for the sake of argument, Brennan’s pessimism regarding average citizen competence, it is more likely for those who are competent to be on the lower side of p = 0.5–1. Furthermore, under non-ideal circumstances (whereby not everyone complies with the Clean Hands Principle), chances are even greater that someone with p > 0.5 will most likely contribute positively to the average p. Still, since one is only responsible for own’s own role in the collective action of democratic decision making, we should design our positive voting duties based on what would be desirable behavior from everyone. After all, concerns about an adverse effect under non-ideal circumstances would apply to Brennan’s Clean Hands Principle as well, which holds that one should abstain from voting if one has p < 0.5. For example, someone with p = 0.4 should not vote. Yet if a substantial number of people with p < 0.3 vote, the addition of a voter with p = 0.4 would actually increase the average p and positively contribute to the likelihood of the epistemically best option being chosen. Still, one should act as we want everyone to act (both the person with p = 0.4 and other people with p < 0.3 ought to abstain). If the person with p = 0.4 abstains from voting, he/she has complied with voting ethics. Those who vote despite their low competence have not and are therefore blameworthy.
24 In epistemologist terms: reductionist and direct view accounts of testimony both deem testimony as a valid source of knowledge.
25 Again, this is not to claim that experts have some special epistemic ability to know the right outcome of political decision making (this would be a reason to endorse epistocracy, which I do not). However, it should be clear that when we are inquiring about, for example, empirical data or the understanding/interpretation of that data, certain information sources are more credible than others. Notwithstanding that it can be difficult at times to determine the trustworthiness of sources, it seems to me that there are certain reliable objective norms one can appeal to in deciding who (not) to trust; recognizing whether someone has expertise is one helpful means for determining the trustworthiness of a source.
26 E.g. see Goldman Citation2001 and Citation2018, Scholz Citation2009, and Croce Citation2019 for discussions of how to recognize experts. See Goldman Citation2002 and Coady Citation2006 and Citation2012 for discussions of what to do when experts disagree.
27 Furthermore, let me also emphasize that wrongness and blame are not always perfectly correlated. We can deem uninformed voting as bad moral practice even if barely anyone is ever blameworthy for it. E.g., most people think there is something intrinsically wrong with doing harm. At the same time, we might not blame someone for doing harm to an attacker when it is in self defense and is the only available option.