Democracy divides us

Elections and their discontents

Democracy has achieved some incredible things, making our societies more equal and just. But in recent years rather than being an efficient mechanism for collective decision-making and progress, democracy seems to be fueling discord, division, and distrust of the other side. Do away with elections and democracy itself might be saved, argues Alexander Guerrero.  


Elections are not magical, but they can seem that way.  The last 200 years have seen electoral democracy as the ascendant, most successful kind of political system the world over.  Electoral systems have improved internally as slavery, colonialism, racist and sexist discrimination, and disenfranchisement have gradually if not entirely been challenged and eradicated.  But we are starting to see new trouble.  In many democratic political communities, the leading story is one of division, rage, and irrationality.  It is hard to imagine a path together, forward, in response to the urgent challenges we face.  Elections are not magical.  They don’t automatically make everything better.  They only work well under specific social conditions.  And if those conditions do not obtain, elections can actually make things worse. 

As a teenager attending public schools in the United States, I was taught that our political system was the best and only legitimate form of government.  I was also taught almost nothing else about the system.  There was a Constitution, a Bill of Rights, three branches, some checks and balances (no one knew why the branches would check each other), elections, and freedom.  Our system was considered so much better, that I was taught nothing about other democratic systems.  The options seemed to be: this exact thing we have, some crazy king, or a communist dictator.  My education was not unusual in this regard.  Many of us live in political communities with stunted political imaginations and limited knowledge of how political systems work.  That makes it hard when this apparently magical thing—electoral democracy—is not working well.  It might seem that we don’t have better options.  But we do. 

Elections are not magical. They don’t automatically make everything better. They only work well under specific social conditions.


How elections deepen our divisions

If we have a commitment to work together to solve problems, deep attachments and real affection, a common overarching cause, and basic, unshakable respect for each other, then taking a vote and going with the majority view can be an excellent way of working together, even in the face of disagreement.  But if we do not have those things, elections are dangerously efficient ways of dividing us into groups, intensifying those divisions, and encouraging us to hate those on the other side.

The most important law in many democracies today was not passed by a legislature.  In electoral systems in which one person will be elected to represent a district and the person elected is simply the person with the most votes, there will be two dominant political parties.  This is Duverger’s law, proposed in 1954 by the social scientist Maurice Duverger.  The precise character of the parties might change over time, but there will generally be two stable, dominant political parties. So, elections through systems like those found in the US, UK, Canada, and elsewhere will tend to result in two dominant political parties. 

A significant part of our identity is connected to membership in our group and favorable bias towards it. We create these ingroups and outgroups readily and can do so on the basis of arbitrary and even invented characteristics.

Another important law: human beings categorize everything, including other human beings, in order to understand the world.  In particular, we are drawn to create groups that have an inclusive and exclusive dimension: ingroups and outgroups.  A significant part of our identity is connected to membership in our group and favorable bias towards it.  We create these ingroups and outgroups readily and can do so on the basis of arbitrary and even invented characteristics.  These are central tenets of Henri Tajfel’s influential “social identity theory,” supported and refined by decades of social psychology research.  There is also neuroscientific evidence that how we learn, what we perceive, what emotions we feel, and the intensity of those emotions all are affected by these ingroup/outgroup identities.  These group identities come to structure significant aspects of our life, affecting where we work, how we spend our free time, who we talk to, what media we consume, who we are friends with, who we trust, and so on. 

Taken together, Duverger’s law and social identity theory make evident a significant potential problem with elections.  Our electoral system ensures that there will be two political parties.  If you identify with one of them, as many do, you are part of an ingroup, with a clear outgroup.  You root for your side, take on your group’s views thoroughly and intensely, cheer your group’s victories and their group’s losses, listen to and trust those who echo back your group’s take on political issues, and distrust, discredit, and vilify those who see things differently.  You feel that you are being completely rational in doing all of this, based on the sources and evidence you trust.  Their view of things is bizarre, detached from reality, morally vicious.  It is hard to imagine working together with such people, respecting them even when one disagrees with them, or being fine—merely disappointed, rather than devastated—if they come to power through elections.  But this situation is untenable and unstable.  And it means that our political systems are not helping us address the collective problems we face: poverty and economic hopelessness, unemployment, worsening inequality, climate change, global health crises, and much else.  Liliana Mason, a political scientist who documents these trends in her book Uncivil Agreement, describes “a picture of a nation whose partisan teams are raring to fight, despite an almost total lack of any substantive policy reasons to do so.”[i]

Neither elections nor our human proclivity to ingroup/outgroup sorting are new.  So, what is different now?  There are plenty of potential culprits.  Work and religion no longer provide ‘mixed’ social spaces, as both have become more politically segmented. Our neighborhoods have become more politically homogeneous.  So, we are likely to have very few friends or acquaintances who are members of the outgroup political party.  Increased fragmentation in media markets as a result of cable television news and now social media and algorithmic search have resulted in our living in more robust, durable echo chambers.  Ironically, a long period of relative peace and the absence of even cold war as a direct threat has eliminated one source of felt common cause, leading to more intense internal conflict.  And as the effects of these changes accumulate over time, antipathy toward the outgroup becomes deeper and harder to reconcile with the moral commitment of respect for outgroup members.  We don’t love each other; we don’t even like each other.  We don’t feel part of a shared project.

This division in deeply troubling in its own right, but it also makes it easy for the powerful to distract us by fanning these destructive flames.  Electoral politics becomes a means to divide and distract.  This is not why elections were created, but they can become a very effective means for the elite to retain power and avoid regulation.  All they have to do is throw some red (or blue) meat on the fire.  That is perhaps why, rather than seeing addressing climate change as a common overarching cause shared by all of humanity, it has become highly politicized.  Half of us are screaming about the asteroid hurtling toward Earth; half of us are cheering it on. 

Electoral politics becomes a means to divide and distract. This is not why elections were created, but they can become a very effective means for the elite to retain power and avoid regulation.


Fixing democracy by eliminating elections

Many familiar suggestions for political reform—public financing of campaigns and other strategies for getting money out of politics, reducing lobbyist access, expanding ballot access and voting rights—do nothing to address this deep electoral division.  That’s because elections are at the heart of the problem.

So, if not crazy kings or dictators, what should we do?  One approach would be to focus on rebuilding the background conditions: recreating feelings of respect and connection through creating more informal interaction with outgroup members, inculcating a sense of shared purpose, cultivating tolerance and love for outgroup members, attempting to establish shared sources and mutually trusted testifiers to create some epistemic common ground, and so on.  These are all excellent, vital projects.  But it is very hard to do any of this directly amidst the regular wrenching apart caused by elections.   

A different approach would be to begin considering what democracy might look like if elections were eliminated or reduced in importance.  This might sound paradoxical.  But democracy started with elections at the periphery and lottery selection of political officials at the center.  Aristotle writes in Politics that “the appointment of magistrates by lot is thought to be democratic, and the election of them oligarchic.”  If we associate democracy not with elections but with equality of political power, non-domination, and channels of effective political participation, we can reimagine democratic systems that operate entirely by lot.  That is, using randomly selected political representatives serving on single-issue legislative bodies connected to each other through institutional networks.  I develop the structure around this idea and argue for it at length in articles and a forthcoming book. The basic ideas of lottocracy are simple.  Rather than a single generalist legislature, in a lottocratic system there would be twenty to thirty standing, single-issue legislative bodies, with each legislative institution focusing on one policy area (e.g., agriculture, healthcare, education, energy, etc.).  The members of each single-issue legislature would be chosen by lottery from the relevant political jurisdiction.  And the legislative sessions would be structured so that members of the single-issue legislatures hear from a variety of experts, advocates, and stakeholders on the relevant topic at the beginning of and at various stages throughout each decision-making session, in addition to engaging in in-depth community consultation throughout the process.  At the end of this process, members of the single-issue legislature either have the capacity to directly enact policy or, in some cases, to do so jointly with other single-issue legislatures.[ii]

Without elections, we eliminate the explicit sorting into teams, the regular confrontation of us versus them.

Obviously, there are at least a book’s worth of refinements and potential concerns to discuss. At a more modest and correspondingly more immediately realizable scale, randomly-selected citizens have served on citizens’ assemblies—bodies that focus on particular political issues or political problems.  More than one-hundred Citizens’ Assemblies have been used throughout the world, prominently in Belgium, Canada, France, Germany, Iceland, Ireland, Mongolia, Scotland, and the United Kingdom.  A number of people—including, prominently James Fishkin, John Gastil, Hélène Landemore, and David Van Reybrouck—have been evaluating and reflecting upon these developments and the general use of lottery mechanisms rather than, or as a supplement to, elections.

It has been striking to see, for example, the way in which randomly chosen citizens in Ireland have come together to discuss constitutional reform focused on controversial typically partisan issues such as abortion and same-sex marriage, and to do so thoughtfully, respectfully, and in a way that all sides agree was successful.  It enabled people to better understand the sources of their disagreement, to learn what was most important to each of them, and to hear how their personal experiences affected their views.  As one participant commented:[iii]

“It felt very democratic. All sides were addressed. It very much helped me – not to decide as I already knew how I was voting, but to listen, understand and develop empathy for those who planned to vote the other way. The issue was a very complex and divisive one, and the Citizens’ Assembly helped the issue be seen from all sides.”[i

Lottocratic systems and the use of random selection (sortition) rather than election are attractive for a number of reasons.  They help address concerns about capture and corruption, the role of money in obtaining political power, and the non-representativeness of many elected legislatures in terms of who is in power (often skewing rich, white, and male).  But they are particularly powerful in response to the current division and ingroup/outgroup predicament we face.  Without elections, we eliminate the explicit sorting into teams, the regular confrontation of us versus them.  Moving to single-issue legislative bodies opens up places for us to identify issues on which we agree and where we see common purpose, rather than concentrating political attention on those issues which most deeply divide us.  And if part of our division is a story of manufactured conflict, where the powerful elite impede our working together by amplifying this felt division, then lottery-selection might be a way of taking back real democratic control.  We might become “we the people” once again.   




Liliana Mason, Uncivil Agreement: How Politics Became Our Identity (University of Chicago Press, 2018), p. 88. 

[ii] Alexander Guerrero, “Against Elections: The Lottocratic Alternative,” Philosophy and Public Affairs, Vol. 42, pp. 135-178 (2014); Alexander Guerrero, Lottocracy: A New Kind of Democracy (forthcoming, Oxford University Press).

[iii] Caroline Bannock, “’Transparency and fairness’: Irish readers on why the Citizens’ Assembly worked,” The Guardian (January 22, 2019),; Patrick Chalmers, “How 99 strangers in a Dublin hotel broke Ireland’s abortion deadlock,” The Guardian (March, 8, 2018),

[iv] Reported in Bannock (2019).

Latest Releases
Join the conversation