a blog about philosophy in public affairs

Electoral Justice in Pandemic Times

The COVID-19 pandemic has reshaped numerous aspects of our lives during the past few months. Even though it is less immediately felt in our daily routine, one of the most consequential by-products of the pandemic is its impact on the political life of our communities. Elections, in particular, have been heavily affected throughout the world, with the International Foundation for Electoral Systems reporting postponments of elections in 51 countries, as of the 27th of April 2020. In other cases, elections have been held in the traditional fashion, sometimes under heavy criticism and resulting in a severely depressed turnout. All of this prompts the question of what is the appropriate governmental response in respect to holding elections during pandemic times? Should we continue to go about our electoral business as usual? Or should we postpone elections until the outbreak is over? Or should we still hold elections, but under alternative mechanisms, such as postal voting or e-voting?

Milwaukee (Wisconsin) resident in line at a polling station, 7 April 2020. Photo by Patricia McKnight / Milwaukee Journal Sentinel

An answer to these questions is not easy to provide. In effect, the circumstances of the pandemic outbreak give rise to an ethical tension, with each response having some important detrimental effects on core electoral values. In what is probably the most influential book on the topic, Dennis Thompson has outlined three principles of electoral justice: equality, freedom, and popular sovereignty. Since these principles are not particularly controversial, we may deploy them in order to illustrate the problem, which may be called the Pandemic Electoral Trilemma:

P1: Reasonably just elections should adequately satisfy the principles of equality, freedom, and popular sovereignty.

P2: Traditional voting during a pandemic fails to adequately satisfy the principle of equality.

P3: Postponing elections during a pandemic fails to adequately satisfy the principle of popular sovereignty.

P4: Distance voting during a pandemic fails to adequately satisfy the principle of freedom.


C: No electoral policy during pandemic outbreaks can ensure that elections are reasonably just.

Let’s look at P2, P3, and P4, in turn. Consider, first, the option of holding elections in a traditional fashion, with in-person attendance at the polling station required in order to cast a vote. This option may immediately strike us as the least palatable, because of its expected effects on public health. With weeks or even months of restrictive measures taken in order to combat the outbreak, the social interaction required for holding elections would threaten to undo all progress obtained. But putting the public health issue aside, holding traditional elections would undermine the equality principle of electoral justice. Since COVID-19 disproportionately affects senior citizens and people who already have certain pre-existing conditions, the expected costs of attending the polls (risking to catch the virus in the process) would be disproportionately high for them, in comparison to other groups. This substantive inequality in the opportunity to cast a vote is a serious concern, particularly as the groups most likely to abstain are, by and large, more vulnerable to begin with and widespread abstention at the polls could lead to a further compounding of their disadvantage.

Consider, then, the option of postponing elections, which appears to be most widely deployed, at least for the time being. While public health concerns would suggest this option as the one which is least bad, political concerns regarding the postponement of elections abound, since the rules of the democratic game stipulate that positions of power are to be occupied by elected officials on a clearly time-limited basis. The principle of popular sovereignty is undermined by postponing elections, since without them the voice of the demos cannot be heard at the appropriate time. Without this voice, holding power becomes democratically illegitimate, a problem which is particularly worrisome if no time frame is to be given for the length of the postponement.

Perhaps, then, we should endorse a third alternative, which is to hold elections, but under distance voting mechanisms such as voting by mail or e-voting, rather than by in-person casting of ballots at the polling station. This is the option favoured, for instance, by Lee Drutman in the recent conversation had with Andrei Poama on the subject*. And it is an option which is not novel, since some elections already allow or are fully held through such mechanisms. But distance voting comes with its own drawbacks. The most obvious one is that of electoral fraud, which might not necessarily be widespread in current elections that use distance voting, but might be seriously worrisome if they are introduced for the first time in the publicly opaque circumstances of a pandemic, which do not allow for proper scrutiny or piloting. But putting this aside, distance voting also fails to protect one of the principles of electoral justice outlined by Thompson, namely that of freedom. Since ballot secrecy cannot be thoroughly ensured through e-voting or postal voting, the freedom to exercise one’s vote according to one’s own conscience is curtailed in clientelistic communities or patriarchal families, as power relations would be able to dictate how more vulnerable agents will vote, instead of their own reasoned preferences.

The Pandemic Electoral Trilemma suggests a grim conclusion. Of course, some might object to the existence of the trilemma to begin with. Perhaps P1 does not adequately capture what just elections are; or perhaps one of the subsequent claims are false; or perhaps there is a fourth category of policy options not yet explored. But what if we think that the trilemma does indeed exist, in the way outlined above?

My contention is that the right response would be to choose the electoral policy which can be supplemented with other institutional actions, that are able to go sufficiently far in alleviating the concerns raised by that policy option for electoral justice. Because the problems in P2 and P4 do not appear to be readily solvable through institutional measures, I suggest that the most plausible path would be to target the problem of popular sovereignty, caused by postponed elections. Without being able to develop it in much greater detail here, my tentative proposal is that the postponement of elections should be followed by a norm of forming national unity governments, which would include a coalition of political parties with parliamentary representation which is as broad as possible. Without fully extinguishing the problem of popular sovereignty, since the demos has not spoken in favour of this political configuration, it could be temporarily sufficient, as power would be shared in such a way so as to represent a wide range of interests within the demos, rather than the much narrower and possibly sub-majoritarian representation of interests under a one-party or even an approximately minimal winning coalition. Furthermore, in times of political uncertainty, a national unity government could also (1) be in the interest of both governmental and opposition forces, who would share the political costs and benefits both of holding power now and of electoral scrutiny for handling the crisis down the road, (2) be useful in staving off some of the more populist political measures, which could lead to worse outcomes, and (3) receive widespread support from citizens, as a recent YouGov poll suggests.

A few caveats are worth mentioning. First, it is possible that under a closer examination the trilemma would be softened either by the conclusion that the satisfaction of all principles is not equally important, or that the three principles are not equally affected by the various policy proposals. If this is the case, we may have reasons to favor a different policy proposal. Second, I do not mean to suggest that the proposal outlined above is necessarily optimal in all cases. Political culture and traditions as well as regime type are important features, since they may affect the degree to which problems generated by unveiling the vote will arise. Third, the proposed solution can only be pitched at some levels of the political system. Direct elections for executive offices (e.g. mayors) cannot technically be addressed in this fashion.


* The conversation is part of a series of Insights developed as part of the “Reconstructing Democracy in Times of Crisis: A Voter-Centred Perspective (REDEM)” project (http://www.redem-h2020.eu/).

Alexandru Volacu

Alexandru Volacu is an Associate Professor at the University of Bucharest and Director of the Bucharest Center for Political Theory. His research interests mainly revolve around several topics: the ethics of voting, theories of justice, and the ideal/non-ideal theory debate.



A new virus, an old problem: Why lessons from poverty and gender inequality in Brazil matter


Should Economics be Pluralist?


  1. This was really interesting! What about holding traditional elections but giving people an opportunity to distance vote? This seems to mitigate the equality loss in traditional elections. And, assuming that those who are most vulnerable to the virus are not the same group as those who are most susceptible to have their freedom curtailed, it seems to promise a near-realisation of all the three desiderata.

    • Alexandru Volacu

      Hi Anca,

      Thanks for the comment. This is an interesting proposal, and might ultimately be one that minimizes the most important problems. But right now I’m not entirely convinced it will. As Andrei privately commented on my own proposal, the devil might be in the details (which I think is right for all of these proposals). If we’re talking about e-voting, I doubt it would be able to solve the problem since the plausible assumption is that it would be difficult to have widespread usage of this mechanism for senior citizens right now (which would only increase the inequality). If we’re talking about postal voting, I’m not sure how that can be done in order to simultaneously: (1) avoid social interaction and (2) ensure some modicum of secrecy. I think that ultimately the choice boils down to (a) holding elections under inadequate conditions, which might be biased in a lot of different ways and may not adequately reflect the preferences of the electorate overall and (b) extending current mandates for a reasonable time frame (6-12 months) until elections can be held under adequate conditions. Is (a) or (b) better from a democratic perspective? I tend to think (b) is generally better, but also that we should look at it on a case-by-case basis. The problem is that the states where more serious violations of equality and freedom would expectedly occur if they’re doing (a) are the same ones where extending current mandates are expectedly going to be used for power grabs and attacks on democracy. So for countries like Belgium or Germany, for instance, I don’t think the trilemma is that pressing to begin with. But for countries like Hungary or Turkey it is.

  2. Not sure about the principle of national elections. The Conservatives in the UK at one point are busy trying to get Labour involved in a government of national unity. But there are good reasons not to go there, from the Labour perspective, so this ‘solution’ to the trilemma seems to ignore important dimensions of political conflict and disagreement. Governments of national unity have to be voluntary. My hunch is that the best option is to postpone long enough to organise them safely. If we can organise trips to the super market, this must be in the power of governments to manage; and perhaps it should be possible to create mobile secret ballots, rather like little ‘sandwich shops/burger shops on wheels’. They arrrange to be on a street/neighbourhood at a particular hour/set of hours; old people and the disabled are given priority slots (as happens at many supermarkets), and then the rest of people can go to normal electoral sites, arranged to space people out. Again, if you have to spread the process over a couple of days in order to make distancing possible, so be it. But unless I am really thick, generally safe elections shouldn’t be beyond the bounds of human ingenuity to arrange.

  3. Alexandru Volacu

    Thanks Annabelle. I agree with your points. I would definitely not suggest any form of compulsion in regard to the formation of governments, so it would be strictly voluntary. Rather, the thought is only that the voluntary formation of national unity governments would diminish some of our more serious concerns regarding the postponement of elections. Particularly if organizing elections safely would not be possible within some reasonable time frame (a year maybe?) from their initial schedules. I take your point about political conflict and disagreement – perhaps one of the underlying assumptions of such an argument is that there is sufficient overlap between the parties on the specific matter of handling the crisis so that their conventional political disagreements can temporarily be set aside. But this assumption might be problematic (and surely is in many cases).

  4. Dear Alexu,
    I doubt that national unity governments are a substitute for elections, because by implication NONE of the people in office is particular legitimate by that point, so uniting on that basis doesn’t seem likely to do much for legitimacy. The case for national unity governments, I would have thought, is based on some national crises so severe that you could expect people to prefer unity to party politics, in the circumstances, even if elections were possible. Perhaps I’m wrong about this, but the case for National Unity Governments seems independent of whether or not you can hold an election.
    The question then is, what happens when you can’t hold an election? And I suppose what I would like more information on the circumstances you have in mind. If we think of current circumstances for most of Europe and America, say, I’m not really sure why elections couldn’t be held once you are easing constraints on lock-down. It seems more an issue of organisation and willingness to let an election go on for several days or a week etc. But in principle that does not seem objectionable. When you think about the Herculean organisation required for Indian elections, normally, I don’t really see that present challenges are overwhelming.
    If there is a spike and we all have to go back to strict quarantine, that is another matter. But then you obviously have no choice but to delay, since you can’t really campaign properly in those circumstances. So the trilemma seems a bit moot, as the postal ballot issue can’t really arise – in fact, you just lack most of the conditions for an election. Nor would you want people in power to be distracted….so it seems, the burden is on securing as much transparency and accountability in government as possible until elections can be called, whoever is in power/whether a party or some national coalition. The parallel would be to an earthquake or other catastrophe at election time, no? So there would be nothing special about it being a pandemic and quite possibly no trilemma either.

    The difficult cases, I imagine, are for places like India, where even ordinary elections are incredibly complex, difficult and often dangerous operations. But probably we in Europe and North America might have something to learn from the techniques used in such places to make elections possible over extended distances and time-spans. And I suspect experiences of emergency rule, military rule etc will frame the compromises that are possible safely, and without sacrificing legitimacy, faced with the specific challenges of a pandemic, as opposed to other types of emergency.

    I do, though, have a question about how serious a threat we should treat postal voting. I’ve just been reading the Elster collection on secrecy and elections, and it seems that until recently Germany treated postal votes as justified only in exceptional circumstances and, so it seems, students in Oxford had to go to London (twice) to vote in the French 2 step Presidential elections.

    I can see that some families will simply not be conducive to secret voting, if voting happens at home. And clearly, we should neither minimise the extent of domestic violence nor the possibility that people who are not actually violent may nonetheless intimidate their spouses and children to vote the way that they desire. But I wonder whether it is not a mistake to draw such a sharp line between postal voting, assumed to be presumptively inconsistent with secrecy, and a secret ballot in a public place, for eg. a school. After all, people who are willing to intimidate their families into voting the way they wish, if voting occurs at home, are likely if they care that much about politics, to strictly control their family’s access to political information. In which case, their problem for fair elections (and, more fundamentally, I would have thought, for the lives and well being of those family members) exists wherever voting occurs and however secret the moment of voting itself. So perhaps your trilemma depends on exaggerating the dangers of postal voting for secrecy? The specific dangers of internet voting I leave to the more tech-savvy to discuss.

Leave a Reply

Your email address will not be published. Required fields are marked *

Powered by WordPress & Theme by Anders Norén