Every Day an Election Day

II.A Epistemic Manipulation of Voters

Epistemic manipulation of voters occurs when information (or misinformation) is strategically withheld or deployed during an electoral period in order to influence the outcome. It is not just politicians but media organizations, advocacy groups, and other political agents that are incentivized to manipulate the timing and salience of public information.¹⁰ Of course, electoral campaigning involves a host of strategies to persuade voters ahead of an election. These include strategically timing announcements, debates, advertising, and other media and public relations strategies. Many of these acts are innocuous and necessary for informing voters as they make up their minds about which candidates or policies to support. However, some elements of political campaigning are objectionable from the perspective of a voter who wants to rely on accurate, relevant, and timely information before voting. Specifically, the temporality of elections creates problems because both the salience and the timing of public information is manipulable by political actors.

Salience is a property of information that makes it striking or perceptible to voters.¹¹ Voters not only have policy preferences but also policy priorities. Voters have limited cognitive resources they must deploy strategically in order to make their choice. Focusing on striking information that aligns with their issue priorities while ignoring less noticeable information makes the electoral process cognitively manageable for voters. Problems arise, however, because the salience of an issue can be manipulated.¹² Changes to salience can be driven by factors that are exogenous or endogenous to the agency of political actors. Exogenous sources of salience include financial crises, security threats such as terror attacks, and long-term structural changes beyond party-political control.¹³ Endogenous factors, such as significant political decisions taken by a government, can also affect salience. A notable example is a declaration of war. However, more prosaic information can become salient ahead of an election by the policy response taken by elected officials.

Take, for instance, the response of the Trump Administration to a large “migrant caravan” that formed in Central America in the summer of 2018. Migrants typically travel together in caravans as mutual protection against criminals. The migrant caravan formed in Honduras in 2018 had from one to as many as four thousand individuals, making it the largest such caravan to be assembled.¹⁴ However, this must be put in the context of the nearly 400 thousand migrants apprehended by US border agents in 2018.¹⁵ In the unlikely event that the caravan made it to the US border intact, it would represent less than 1 percent of those who were detained, not counting those who successfully cross. What made this particular migration story notable is that President Trump raised its salience. “Migrant caravan” entered the public discussion like never before amid news reports about its size and progress, rampant speculation about its potential composition, and Trump’s association of its existence with weak border laws he claimed were supported by his Democratic opponents. The saga culminated with the decision of the Department of Defense, under White House directive, to deploy 5,200 active duty troops to the southern border on October 29, days before the midterm congressional elections.¹⁶ Trump’s unprecedented decision to deploy active duty troops to the southern border was widely criticized by his political opponents as an attempt to mobilize support for his party ahead of the 2018 midterm congressional elections. Whether his intention was electoral manipulation or a sincere effort to discharge his duty, the episode illustrates incumbent governments’ ability to use its authority to raise the salience of a policy issue.¹⁷

The susceptibility of voters to salience manipulation is heightened by the narrowness of the electoral period. The ability of misinformation to be dispelled, better arguments to win the day, and more thoughtful outcomes to emerge are hampered by the short window in which elections occur. Voters also find themselves inundated with a large volume of information over a short period of time ahead of elections. As a result, “voters may find it difficult to process all relevant information and instead rely on cognitive shortcuts when faced with difficult problems.”¹⁸ Using short-cuts and heuristics is often an efficient way to make optimal decisions. However, as I show in the next section, the narrowness of the election period diminishes the effectiveness of these cognitive strategies.

II.B Retrospective Voting and the Temporality Problem

In a representative democracy, citizens exercise their political authority through their elected representatives. There are various competing accounts of the nature and role of representation in democracy, and how elections realize the ideals of representation.¹⁹ Theories variously identify representatives as trustees,²⁰ advocates,²¹ mediators,²² among others. Furthermore, the fact that representatives are distinct from and, in some sense and by some standard, have to act on behalf of citizens opens up robust debate on the institutional design of democracies. Nonetheless, a core aspect of any ideal of democratic representation is that representatives are selected and authorized by voters and accountable to them. Elections play a better or worse role to the extent they empower voters to realize ideals of fair representation.²³ Periodic voting undermines the ability of voters to exercise these functions. While I cannot survey how temporality interacts with, and undermines, the various ideals of representation, I use the retrospective theory of voting to illustrate the temporality problem. The retrospective theory of voting is ideal for exploring the temporality problem because: (i) it captures a core aspect of the popular understanding of democracy wherein voters “hire” and “fire” their representatives based on their performance and (ii) retrospective voting assumes that voters make their decisions without the benefit of deep policy expertise. As such, the fact that the temporality of periodic voting undermines the effectiveness of elections (in realizing democratic aims) when voting retrospectively suggests temporality imposes even more constraints on accounts where voters are expected to have higher levels of sophistication and knowledge.

According to retrospective theory, voters cast a ballot on the basis of the past performance of the government and their representatives.²⁴ That is, voters “exert substantial control over their leaders, despite knowing little about the details of public policy, simply by assessing the performance of incumbent officials, rewarding success and punishing failure.”²⁵ This makes retrospective voting a core mechanism for realizing democratic accountability of political institutions. The retrospective theory of political accountability offers a less epistemically demanding account of how citizens ultimately control the democratic political system.²⁶ When voting retrospectively, voters need not assess the technical details of government policy. Rather, voters respond to policy outcomes—such as economic indicators including GDP growth, income growth, and the unemployment rate—which are a far more transparent signal for the average citizen. Retrospective theory therefore preserves the normatively attractive division of labor between representatives and citizens. On the retrospective account, representatives pursue policies that advance the interests of citizens. Voters respond to the positive effects of government policy (or lack thereof) in making their electoral choices.

However, the interaction of voter psychology and the temporal properties of election time disable the effectiveness of this mechanism. As has been well-documented in the literature on voter behavior, incumbents may find themselves benefiting or being penalized by voters for actions over which they have little or no control simply because they occur in proximity to an election. These include natural disasters, price increases due to exogenous events, and even putatively irrelevant occurrences such as the recent performance of a local beloved sports team.²⁷

For our purposes, the main challenge the temporality of elections poses to effective retrospective voting is that voters overwhelmingly rely on public information available close to an election to determine how to vote. For instance, multiple studies on US politics have established that the election-year economy influences presidential election results more than cumulative economic growth over the full presidential term. Healy and Lenz find that voters intend to judge presidential performance by using cumulative growth. However, that information is not easily accessible. As a result, voters use the election-year economy as a proxy for cumulative performance.²⁸ The unintended myopia of voters is such that “objective changes in economic well-being seem to matter significantly only if they occur in close proximity to Election Day.”²⁹ Consequently, myopic retrospection severely hinders voters’ ability to select for or sanction economic performance by incumbents. Given the electoral value of election year economic performance, myopic retrospection incentivizes incumbents to pursue policies that boost economic activity in the months leading up to an election potentially to the detriment of long-term economic performance.

William Nordhaus’ influential work on the “political business cycle” first drew significant scholarly attention to political manipulation of economic policy for electoral gain.³⁰ Nordhaus posited that the constraints on policymakers in a democratic political system would result in a business cycle, i.e., cycles of economic growth and contraction. Policymakers operate within a macro-economic framework where there is a trade-off between inflation and unemployment and voters have preferences about both (although they are ignorant of the tradeoff). Nordhaus’ model has the result that “democratic systems will choose a policy on the long-run trade-off that has lower unemployment and higher inflation than is optimal.”³¹ This is due to the myopia of voters incentivizing expansionary policy ahead of elections to boost employment followed by contractionary policy after elections to tame inflation.

Political business cycles subsequently became one of the most widely studied phenomena in economics and political science. However, subsequent empirical work has revealed much less empirical evidence for political business cycles in mature democracies than the theoretical models suggest.³² Achen and Bartels suggest the limited empirical evidence for political business cycles is due to the difficulty of detecting pre-election economic fluctuations with short time series and few elections, among other challenges posed by variations in democratic electoral systems.³³ Regardless of whether political business cycles are well-founded, there is considerable evidence that governments target income growth in the form of tax cuts, subsidies, and other policies ahead of elections. For instance, in the United States, there is a significant increase in income growth in presidential election years.³⁴

In sum, epistemic manipulation of voters and myopic retrospection undermine or disable crucial democratic values such as responsiveness and the ability of citizens to hold representatives to account. It empowers incumbents to shape the narrow electoral time period to their electoral benefit or at least incentivizes them to do so. Providing voters with more information so as to enable them to avoid these manipulative strategies will merely exacerbate the cognitive task voters are asked to perform over a short period of time. Below, I outline registral voting and show how it blocks the mechanisms that endanger responsiveness and accountability of representatives.

IV.A Registral Voting and the Value of Democracy

With an outline of registral voting in place, we can see how it disables or minimizes the effects of epistemic manipulation and myopic retrospection. Under periodic voting, cognitive limitations combined with the range of manipulable signals available to political actors (especially incumbents) put voters in an unenviable epistemic position ahead of an election. Voters want to make informed evaluations of political performance. However, as discussed in section II, they are ill-served in their capacity to do. The incentive of political actors to manipulate the timing and salience of political information and the myopic retrospection of voters are consequences of the temporality of periodic elections. Registral voting equalizes the importance of every day in determining the outcome of an election. As a result, information being made available to the public at one time rather than another has dramatically reduced election-specific importance. Registral voting therefore minimizes the ability of political actors to manipulate the salience of information or events for electoral ends.

Myopic retrospection is eliminated by registral voting. Voters are able to assess the performance of their representatives in “real time” and vote accordingly. The cognitive load of the information they must consider when voting is significantly reduced. Voters need not remember all the information over the term of office they view as being important to making voting decisions, they can vote as and when salient information becomes available. There is no period close to Election Day during which events have overwhelming importance.

Of course, it is inevitable, and not objectionable, for politicians to attempt to capture voters’ attention. Indeed, registral voting will tend to enhance political competition by making every day electorally valuable. What registral voting dampens is the ability to translate attentional capture into electoral benefit on the cheap. For instance, tax cuts, short-term subsidies, reduced petrol prices at state owned pumps,⁴⁶ among other policies, can be enacted for relatively low fiscal cost since their effect need only be felt for a short period ahead of an election. The electoral benefit of pump priming hinges on the limited, short-term boost to voter consumption (and therefore voter’s perception of government performance) ahead of an election. Such expansionary pre-election budgets are frequently followed by post-election contractionary measures to stabilize public finances.⁴⁷ The elimination of non-electoral periods by registral voting means that incumbents pursuing such policies over long periods would be forced to commit much larger fiscal resources to rapidly diminish electoral returns.⁴⁸ The reputational cost is therefore much higher. Incumbents would open themselves up to more effective charges of fiscal irresponsibility.

The structure of election time inevitably results in a particular distribution of political power. As Elizabeth Cohen notes, “in the same way that decision making is considered democratized when it is dispersed over a larger rather than smaller portion of the population, so too might we think of boundaries and rights as more democratically structured when they are built using multiple points in time rather than single dates.”⁴⁹ Registral voting realizes a new form of election time. By enfranchising voters across the whole term rather than a single date, registral voting enhances the ability of voters to resist the ability of governments and elites to wield durational time for their own ends. This, among other things, serves as a democratic counterweight to a form of temporally distributed binding constraint on governments, namely, markets. Capital markets provide instantaneous feedback to government policy. But binding feedback from voters potentially takes years. Policy therefore will tend to be more receptive to capital than to ordinary voters. Registral voting rectifies this imbalance by providing a daily democratic counterweight to markets.

IV.B Saving Democracy from Elections

Prominent defenses of democracy highlight its capacity to secure values such as political equality, deliberation, autonomy, among others. On egalitarian accounts, democracy is justified by the fact that it treats everyone as equals. In a democracy, everyone has an equal ability to influence the political process and shape public policy.⁵⁰ Democracy prevents the subordination of some citizens to others. It is not a system where some rule while others are ruled but rather one in which citizens relate to one another as equals.⁵¹ A (democratic) political system is justified insofar as it realizes these egalitarian aims. Another influential account of democracy centers the value of deliberation to public justification.⁵² Democracy enables the public use of reason by which the laws and institutions that govern citizens conceived as free and equal are justified.⁵³

The promise of registral voting lies in its capacity to better realize these core democratic values. Electoral democracy’s putative failure to do so has recently become a flashpoint for proposals to restore democracy by doing away with elections. Alexander Guerrero and Hélène Landemore advance prominent lottocractic alternatives to elections, where representatives are randomly selected from among the citizenry by a lottery system, to fully replace or complement electoral representation.⁵⁴ For Guerrero the “primary purpose of elections is to help measure the extent to which candidates are actually supported and authorized to govern […].”⁵⁵ Different electoral systems can do better or worse at realizing this aim. However, Guerrero argues elections are inherently limited in their ability to deliver a responsive and good government. Elections fail to secure “meaningful accountability” of representatives because of the widely established ignorance of voters on many aspects of government policy as well as the capacity of powerful interests to capture the electoral process. For instance, financial constraints and the corporate control of mainstream media impose limits on the range of candidates and policies that are viable.

Landemore argues that elections are ill-suited to realizing the aims of democracy, including ideals of deliberation, political equality, and popular rule. For Landemore, representative democracy engenders “the separation of a ruling elite of elected officials, appointed courts, and administrative bodies on the one hand and the mass of ordinary citizens on the other.”⁵⁶ According to the most prominent accounts of democracy, ordinary citizens control the ruling elite through periodic free and fair elections. However, Landemore argues that elections fail to empower all citizens equally. For Landemore:

Elections introduce systematic discriminatory effects in terms of who has access to power, specifically agenda-setting power. By so doing, elections skew the type of perspectives and input that shape law-making, likely resulting in suboptimal results. Second, elections entail a type of party politics that is itself not all that conducive to deliberation or its prerequisite virtues, such as open-mindedness, rather than partisanship.⁵⁷

The immense influence elites have over crucial institutions in an electoral system, from political parties to news organizations, render elections an ambiguously democratic form of popular rule. The reliance on political parties to mediate the relationship between voters and their elected representatives engenders partisanship, which is inimical to the open-minded, undominated, and free dialogue that is central to deliberative accounts of democracy. As such, Landemore argues, elections prevent rather than facilitate popular rule.⁵⁸

Guerrero and Landemore argue that existing proposals to “save democracy” (e.g., campaign finance reform) are ill-matched to the scale of problems posed by modern electoral democracies. They propose moving away from elections to randomly selected representative assemblies to secure the core values of democracy. While analysis of their specific proposals is beyond the scope of this paper, there are two reasons to resist the move to lottocracy made possible by registral voting. First, Guerrero and Landemore partly draw on the empirical evidence on the ignorance or myopia of voters (discussed in section 2) to highlight the far-reaching and objectionable divergence between democratic theory and practice. However, as I have argued, these facts are exacerbated by the temporality of elections. Insofar as registral voting can disable many of these objectionable features of election time, we have reason to try this transformative reform before taking the more radical step of abolishing elections altogether.

Second, as Kevin J. Elliot argues, lottocratic proposals come at the cost of making the political process much more complex than in modern electoral democracies.⁵⁹ The participation costs of randomly selected citizens assemblies are far greater than those exacted by elections. As Elliot notes, the potent value of elections rests in the fact that “elections […] are a powerful instance of the principle of concentrating power where participation costs are low.”⁶⁰ If there is a way to preserve elections while disabling the anomalies of election time, such a proposal would combine the relatively low participation cost of elections with a more responsive political system. I argue registral voting does just that. Registral voting does increase the participation costs of elections by increasing their frequency, at least for persuadable voters. However, this is balanced by the reduction in cognitive and epistemic costs associated with periodic elections. Registral voters do not have to consider an immense amount of cumulative information (representing years of governance) as well as the highly dense near-election media environment. Registral voting eases this cognitive burden by temporally decompressing the electoral period.

Elliot also identifies election time as the source of many of the political anomalies that plague democracies. His proposed solution, mandatory annual elections, does better than other proposals in tackling election time. But, as I argue below, it does so at the expense of the effective ability of elected representatives to direct the state.

IV.C Why Not Annual Elections?

Elliot argues that the “unequal busyness” of democratic citizens poses a serious and underexplored challenge to democratic equality. Elliot contends that recent lottocratic proposals for replacing elected legislatures increase the complexity of the political system and thereby render political participation even more demanding for the average citizen. Elliot defends elections on grounds that they are uniquely well-suited to serve as participatory institutions for busy people. For Elliot, “the brute fact of voting being inexpensive and yet also decisively powerful makes democracies that center elections much more likely than conceivable alternatives to empower busy people and avoid the pathologies of inequality generated by more demanding modes of participation.”⁶¹ That is, elections fulfill the design principle that realize democratic equality among citizens who are unequal in the time and attention they are able to give to politics. Among these design principles is that political participation be “simple, easy, and undemanding to leave citizens with as much time under their control as possible.”⁶²

Elliot’s main proposals, mandatory voting and annual elections, aim to empower ordinary, busy citizens by creating an institutional and cultural environment of low-cost, participatory elections. Elliot’s proposal for annual elections directly addresses the politics of election time. Elliot argues that “longer durations between elections mean fewer opportunities to change political direction. This concentrates more power in the hands of representatives because they get to hold power for a longer period of time. It also reduces their accountability […]”⁶³ Elliot interprets the literature on retrospective voting as having the upshot “that we can divide the term of office between an electorally accountable period crowded temporally close to the next election and the rest of it as a period of unaccountable power.”⁶⁴ Representatives are untethered from voters in the non-electoral period over which voters are blinded by myopic retrospection.

Annual elections circumvent the anomalies of election time generated by myopic retrospection. If elections are held every year, there is no need for voters to rely on a proxy to gauge the performance of the government. The amount of information they need to consider or recall is significantly reduced. Consequently, the ability of voters to hold representatives accountable is magnified. However, annual elections come with a cost. I appeal not to the cost of frequent elections, but to the cost of frequent changes of government or representatives. Longer terms of office may insulate representatives from the influence of their voters. But they also engender the development of deep political expertise in politicians. Additionally, regardless of whether incumbents are re-elected, it gives them the opportunity to enact policies whose benefits have a longer time horizon without worrying that they will not receive the credit. Annual elections incentivize politicians to campaign on and enact policies with immediate or short-term benefits potentially to the detriment of longer-term gains.

Elliot cites annual elections for state legislatures in the early American republic as a real-world case of how annual elections contribute to responsive and accountable government.⁶⁵ However, the lessons to be drawn from this case are tempered by the evolution of government over the intervening centuries. The activities of government and its social, economic, and legal role have grown immensely complex in the past century. Among member states of the Organization for Economic Cooperation and Development, the average government spent 40.8 percent of GDP in 2019.⁶⁶ In advanced modern economies, sectors such as education, health and social care, energy, and banking, among others, have extensive state involvement. In this context, having annual terms of office might perversely limit the responsiveness of government not by insulating representatives from voters but by shrinking the role of elected representatives in the state. It would potentially empower the permanent civil service vis-à-vis elected officials.

Fortunately, registral voting is able to side-step these limitations of annual elections without sacrificing responsiveness and accountability. Under registral voting, while votes are recorded daily, representatives serve over longer terms of office (perhaps four or five years as is typical in most democracies). Consequently, even a newly formed government will have a significant period of time to become adept at directing the operations of the state. However, the longer term of office does not shield incumbents from popular control. The fate of the government after its term of office expires is being written every day.

V.A Hybrid models

Registral voting can be paired with periodic elections to combine elements of both models. For instance, in ranked-choice registral voting, registral voting can be used to determine the top two candidates (through ranked-choice registral voting). These two candidates will then proceed to a run-off, periodic election to determine the winner. In ranking-based registral voting (e.g., Borda Count or Hare Rule), voters do not select a single candidate or party in their registral ballots but rather rank the parties or candidates in order of preference.⁶⁹ Among other advantages, ranking-based voting methods prevent wastage of votes. Even if electoral candidates withdraw or are removed well into the electoral term (and voting period), the voting rules will reallocate those votes to candidates that are still active.

Registral voting is also an ideal system for implementing provisions for recall elections in jurisdictions that have them.⁷⁰ In the United States, thirty-nine states have provisions for recalling elected officials at some administrative level. Twenty US States make provisions for the governor to face a recall election. Typically, recall elections are triggered by gathering valid signatures on a recall petition.⁷¹ On the registral recall model, at the beginning of the term after every periodic election for executive office, voters register a vote of confidence on the eBoard. If the confidence vote share of the incumbent falls below a threshold over a specified period of time, a recall election will be held according to provisions in the electoral laws.

V.B Local Experimentation

In federal or devolved political systems, sub-national units can serve as genuine “laboratories of democracy” through implementation of registral voting. I propose that municipalities are the ideal political unit for initially implementing registral voting, for instance through registral ranked-choice voting for the election of city councilors. The size and scale of municipalities, along with their more limited political role compared to national political institutions, makes them a favorable testing ground for emergent political dynamics. Lower-level political units also tend to suffer from lower voter engagement and turnout when political competition is limited.⁷² Since municipalities are providers of crucial services such as education, public safety, water and sanitation, among others, the enhanced responsiveness secured by the registral model is a welcome empowerment of residents.

V.C Non-State Registral Voting

This paper has almost exclusively discussed elections as mechanisms for the democratic governance of states. However, non-state institutions—such as labor unions, universities, corporations, co-operatives, among others—conduct elections. These elections can be conducted through registral voting. In particular, the governance of public corporations is well-served by registral voting. Under US laws, shareholders have governance rights over corporations which they exercise through electing members of the board of directors and advancing and voting on proposals at annual meetings.⁷³ As they have a direct financial interest, shareholders are incentivized to be well-informed about the firms in which they invest. Investors pay large sums for high-quality information from financial newspapers, proxy advisors, trading platforms, among others. Public markets provide daily information aggregating market sentiment about firm performance. These are ideal epistemic conditions for governance. Registral voting enhances shareholder control over corporations, a “fundamental element of corporate ownership.”⁷⁴

Even with cautious implementation, there are still challenges to the registral model that need to be addressed. Below, I discuss objections to registral voting and respond.

VI.A The Benefits of Temporality

A challenge to registral voting argues that while it disables problems caused by election time, it also undermines positive values secured by the temporality of the election cycle. I consider two objections to registral voting drawing on the literature on the temporality of election cycles: the responsibility objection and the electoral fatigue objection.

First, the responsibility objection holds that while register voting enhances the responsiveness of representatives, it for the same reason reduces their responsibility. Recently, political scientists studying the relationship between party and government (particularly in Europe) have highlighted what they call the responsiveness-responsibility dilemma.⁷⁵ There is a trade-off between the extent to which governing parties adhere to the preferences of their voters and the need for governments to pursue policies that transcend election cycles. On this framework, responsiveness is the tendency of political parties and leaders to “sympathetically respond to the short-term demands of voters, public opinion, interest groups, and the media” while responsibility concerns “(a) the long-term needs of their people and countries” and “(b) the claims of audiences other than the national electoral audience.”⁷⁶ The relationship between responsiveness and responsibility is highly complex.⁷⁷ Yet it is plausible that on many occasions the constraints imposed by voters (responsiveness) will go against the constraints imposed on governments by international treaties, financial markets, and long-term planning requirements (responsibility). In such cases, registral voting can be said to have a responsiveness bias by making voter preferences more binding.

The temporality of periodic elections has features that facilitate responsibility by shielding governing parties from electoral penalties of acting ‘responsibly’. The period immediately after an election is typically a time of intense government activity. Newly formed governments enjoy a “honeymoon period” of popularity. This early period serves as cover for governments as they push through policies that may otherwise be unpopular but that they reasonably judge to be in the long-term interests of society. On the periodic model, incumbents hope that some of the fruits of this early activity will be apparent before the next election or be forgotten by the time elections are held. Without this “cooling-off period,” governments might become reluctant to enact bold but “responsible” policies for fear of a backlash that would, under registral voting, immediately register an electoral penalty.

Second, the electoral fatigue objection highlights the benefits of the election cycle in providing political warming-up and cooling-off periods. The interval between elections offers rest for voters. On the other hand, an election campaign is an opportunity for voters to re-engage with the political process more fully. Campaign rallies, debates, political canvassing, and other forms of electioneering bring politics to the forefront of citizen’s minds for a relatively contained period ahead of an election. Whereas in the period immediately after an election, voters can partially disengage and reduce partisan political activity. A citizenry that is constantly electorally engaged, with no warming-up or cooling-off period, might become exhausted by the political process. Furthermore, election cycles influence economic and social behavior. For instance, strong partisanship affects voters’ emotions and market pricing activity.⁷⁸ If voters continuously see one another as either co-partisans or political opponents, this has the potential to distort other social and economic relations between fellow citizens.

The responsibility and electoral fatigue objections raise reasonable worries. The temporality problem must be weighed against the benefits of electoral cycles. However, I argue registral voting can realize both responsibility and civic rest. First, in so far as there is a trade-off between responsiveness and responsibility, registral voting shifts the balance toward responsiveness. Yet these concepts need not be characterized as purely antagonistic.⁷⁹ Responsibility is itself a species of responsiveness—to non-citizens and foreign states, to the long-term interests of citizens, to future generations, and to outcomes that outlive a single parliamentary term. The practical tension, therefore, is not between being responsive or not, but between being responsive to short-run preferences and being responsive to long-term (or non-electoral) interests. Registral voting can lessen rather than exacerbate that tension. By widening the electorate’s sense of ownership over governmental direction, the reform supplies mainstream parties with strong incentives to articulate and defend credible long-horizon policymaking. When citizens no longer feel disempowered, the appetite for “populist” entrepreneurs—whose platforms typically jettison fiscal, environmental, or constitutional restraints—declines.⁸⁰ In this way, greater day-to-day responsiveness can indirectly fortify long-term responsibility.

Peter Mair’s influential analysis of European party government underscores this dynamic: decades of dwindling responsiveness have eroded the capacity of governing elites in Europe to pursue responsible policies because they face constant insurgencies that deny the very legitimacy of responsibility-constraints.⁸¹ Registral voting counteracts that erosion by restoring the sense that voters can meaningfully steer the ship of state, thereby insulating leaders who adopt forward-looking policies against the charge of technocratic detachment.

Robert Dahl famously defined democracy as “the continuing responsiveness of the government to the preferences of its citizens, considered as political equals.”⁸² Responsiveness is a constitutive requirement of democratic legitimacy, whereas responsibility is a prudential virtue that depends on continual legitimation. However, characterized in this way, there occasionally remains a trade-off between the short-term preferences of voters and their long-term interests. Where the two conflict, registral voting will tend to incentivize politicians to pursue the shorter-term. While I cannot fully defend that bias here, it is worth noting that it realizes the Dahlian conception of democracy even if it might trade that for some level of policy effectiveness. For democratic theorists worried about a contemporary “democracy deficit” or the “democratic malaise” fueling extremist movements, the responsiveness bias of registral voting is a welcome feature.

Nonetheless, if responsibility and civic rest require cool-off and warm-up periods provided by electoral cycles, these can be accommodated by registral voting. A registral model could incorporate a formula that puts less weight on votes registered on, say, the first three months of a term and puts higher weight on the votes of the last three months. The exact figures don’t matter for the purpose of the argument, but a weighted formula could instantiate cooling-off and warming-up periods. I am not claiming that such a formula is necessary. Rather, weighted registral voting is an alternative for those who argue electoral cycles are indispensable for responsibility and civic rest. Weighted registral voting allows for cooling-off and warming-up without many of the drawbacks of periodic elections highlighted in this paper. It could minimize the ability of incumbents to bank insurmountable support in their honeymoon period, provide an electorally less costly time to enact bold policies, and allow voters to reduce engagement early in the term and increase it towards the end of the term without fear of losing out on influence.

VI.B Voter Regret

I have argued that registral voting eliminates myopic retrospection. Nonetheless, one may object that voters who change their votes may come to regret their registered votes. Take the case of Joe, who registers a vote for candidate A in 1,300 out of 1,460 days of a term of office. On day 1,301 a news story reveals that Joe’s favored candidate is corrupt. Joe is mortified and changes his vote to candidate B. However, he sees that 89% of his vote has already gone to candidate A, with only 11% remaining to be allocated to another candidate. Knowing what he knows now, Joe would not have voted the way he did. Joe feels that his vote has been wasted. Joe may also regret his lack of foresight. For instance, suppose Joe switched his vote from candidate A to candidate B in response to his anger at candidate A, and later switches back his support to candidate A because it turns out his anger was unwarranted. He may regret the many votes he has deprived his candidate (or given to a candidate he doesn’t support).⁸³

Registral voting does not eliminate voter regret. However, this is not a feature that is unique to registral voting. In fact, registral voting may end up doing better than periodic systems on that front. After all, a voter like Joe may come to learn something about a candidate after election day such that, had he had that information before election day, his vote would have been different. Voters wishing to change their vote in periodic voting systems will at best get the chance during the next election many months or years later. Under registral voting, potentially only fractions of one’s vote are regretted. In periodic electoral systems, necessarily the whole of one’s vote is regretted if one finds oneself in Joe’s situation. Similarly, voters on election day may vote against an incumbent to signal their displeasure and later come to regret their lack of foresight if the opposition party or candidate they do not support overall is victorious.

However, there is an asymmetry between voter regret under periodic and registral voting. Voter regret in periodic elections concerns an outcome that is in the past. In the registral model, the failure of the incoming government to reflect the current preference of voters is something that’s yet to be realized. Voters might be frustrated in cases where the late-winning candidate seems ideal for the political moment. The weighted formula discussed in section VI.A goes some way to strengthening late-term votes but does not guarantee the late-term vote share leader will win.

This objection is then a question of whether the preferences of voters proximate to the end of term ought to prevail in all cases. In this paper I have argued that there are reasons to hold it should not. I grant that there are circumstances where the ideal candidate is one with wind in its sails late in the term. But this is sometimes a serendipitous outcome. The limitations of the temporality of elections on the other hand are well-established and persistent across elections (e.g., myopic retrospection). Given these limitations, I argue it is better to distribute the franchise throughout the whole term. The loss in the occasional serendipitous electoral outcome is made up for by the persistent enhancement of the responsiveness of the political system. A core feature of registral voting is that it eliminates the significance of any one day. Consequently, voters have greater flexibility in assigning their votes between candidates. While eliminating voter regret is not possible, registral voting provides more opportunities for voters to change their vote and still influence the electoral outcome without having to wait until the next election.