Democracy without Elections:
An Argument for Distributing Political Power by Lottery
Name: Rozemarijn Stadens
Student Number: 10763422
Master Thesis Political Science: Political Theory
First Reader: Dr. A. Afsahi
Second Reader: Dr. J.C.A. Olsthoorn
Place and Date: Amsterdam, June 2020
1 Abstract
This paper seeks to further the argument in favour of a sortition democracy. The kind of
sortition democracy envisioned here would distribute the seats in political office through a
nationwide lottery among all citizens. The vast amount of proposals for the use of sortition
has envisioned the drawing of lots, either as an additional democratic tool within electoral
democracy, or as the selecting mechanism for one of two government chambers, existing
alongside an elected chamber. In contrast, this paper argues that elections should be replaced
by sortition entirely. It does so by discussing the various ways in which elections fail to
accomplish democracy as government of, for, and by the people, and furthermore, by
demonstrating that the perceived values of electoral democracy are not being adequately
realized. When looking at the inherent democratic qualities of a sortition democracy, it should
become clear that there is ample motivation to rethink democratic institutions in the absence
2
Contents
1. Introduction……….
2. Electoral Democracy……….
2.1 A Discussion on the history of elections………...
2.1.1 Elections in Athenian democracy………...
2.1.2 Elections as an aristocratic (or oligarchic) method of selection…..
2.1.3 The demise of lottery and triumph of elections.………..
2.2. Inherently undemocratic aspects of elections.……….. 2.2.1 challenges in a situation of choice………..
2.2.2. challenges of competition.……….
2.2.3. challenges of distinction………. 2.3. ‘Indispensable’ values of elections………
2.3.1. The value of consent..………
2.3.2 The value of expertise.………
2.3.3. The value of accountability……… 2.3.4. Getting rid of elections entirely………..
3. Sortition Democracy………
3.1. Political equality in sortition democracy……….
3.2. Deliberation in sortition democracy.………...
3.3. Descriptive representation in sortition democracy.………..
3.4. Basic requirements.………. 4. Conclusion.………... 3 6 7 8 9 10 12 12 13 14 15 16 20 23 26 27 28 29 30 33 35
3
1. Introduction
Free and regular elections occupy a central place in our understanding of democracy. This
paper seeks to shift away from the primacy of elections in democratic theory and instead
propose a new way of selecting political representatives: sortition.
The ways in which elections are failing democracy include, but are certainly not
limited to, decreasing political participation, the rise of anti-democratic parties, the
overrepresentation of a white elite in government and the corrupting influence of money on
both electoral outcomes and policy outcomes (Van Reybrouck 2016: 9-16). Considering the
challenges that elections pose to a healthy democracy I argue to replace them with something
different entirely. I argue that a system in which all citizens participate in a nationwide lottery
where the prize is a seat in government, should replace elections.
The idea of sortition, (or the selection through lottery), has gained attention in
academic literature in the past two decades. It is no longer a fringe subset of democratic
theory. While many theorists have discussed the merits of a sortition system, few have
claimed that elections should be abolished, most proposals steer towards a dual system in
which one chamber of government is selected by random selection, and the other is selected
through elections1. Other then most sortition theorists have argued, I argue that elections
should be completely replaced by random selection.2
Western democracies have become synonymous with representative government.
Whether sortition would still be a form of representative government is not entirely clear, it is
also beyond the scope of this paper. Delannoi and Dowlen have suggested that sortition
democracy is somewhat of a compromise of direct democracy and representative democracy
1 See for example, Dowlen (2017), Gastil & Wright (2018), Malleson (2018), Manin (1997), Mulgan (1984),
Van Reybrouck (2016), Sintomer (2018), Stone (2016).
2 I use the terms, sortition, lot or lottery, and random selection, throughout this paper with more or less the same
connotation. Whereas literature on the modern use of a political lottery often uses the term sortition, its historic use is often designated with the term ‘selection by lot’ or ‘drawing of lots’.
4
(Delannoi and Dowlen 2016: 11). While it requires a more active participation from citizens
than within current representative democracy, it is still based on political delegation and
political deliberation that is cut off from the citizenry at large (ibid). Part of the difficulty of
determining whether sortition is a form of representative democracy lies in the discussion of
the concept of representation.
Arguably the most influential account on representation has been provided by Hanna
F. Pitkin. In the concept of representation she gives a detailed and thorough analyses of
representation. She discusses the different traditional views of representation3 and shows that
none of them are able to grasp the full meaning of representation (Pitkin 1967). Pitkin finds
their biggest flaw in that views of representation focus on how representatives “stand for” the
represented and lack an account of how they “act for” the represented. To account for this, she
puts forward a view of substantive representation (ibid). “representing here means acting in
the interest of the represented, in a manner responsive to them” (idem: 209). For
representative government, this means that government should be in control and promote
public interest while also being responsive to the people. The governed in this context should be capable of “action and judgment” (idem: 233).
Jane Mansbridge has proposed three new views of representation: anticipatory
representation, which comes from the idea of retrospective voting. This means representatives
are more focused on their electorate in the next election rather than the promises they made in
the one that got them elected. The second is gyroscopic representation. This view is based on
the selection of the representative. The representative is simply representing their own beliefs
and principles and they can do so because it was precisely these ideas and principles that got
them elected. Finally, surrogate representation, which is relevant when legislatures have to
3 She discusses the formalistic (or accountability) view, the descriptive view, and the symbolistic view. (Pitkin
5
represent beyond their own constituencies. Mansbridge states that the criteria for judging these views of representation are “almost all deliberative rather than aggregative”
(Mansbridge 2003: 515).
The discussion on what representative government means, and what the criteria for
good representation should be is ongoing which makes it difficult to determine whether
sortition government could be a form of representative government. Moreover, in both Pitkin’s account of representative government, and Mansbridge’s account of representative
government, elections have a central role. This further complicates the question. This might
mean that the language with which representative government has been conceptualized would
have to be reconceptualized in such a way that a sortition government could be included in a
discussion of representative government. It might, on the other hand, mean that sortition
government cannot be seen as a form of representative government, simply because the
concept of representation and representative government have always been tied to elections.
When the latter one is true, the criteria for judging a representative government cannot be
used to judge a system of sortition or to compare its values to that of an electoral system.
Figuring out which of the former two is true, is outside the scope of this paper. Rather, I
provide a critique of elections on the basis of its own promise of realizing democracy, as
government of, for, and by the people, and argue that sortition, regardless of whether it leads
to representative government, responds more adequately to an increasing need for political
equality as well as good political outcomes for the political community as a whole, which are
central values to our current idea of democracy.
In chapter 2, I provide a short genealogical analysis of elections. I demonstrate that for
a long time, elections were viewed as an aristocratic method of selection, rather than a
democratic one. This analysis shows that elections are not inherently democratic and can
6
that challenge democracy. These are challenges of choice, challenges of competition and
challenges of distinction. Because these challenges are inherent to elections, I argue we
should replace elections with a system of sortition. However, there are three democratic
values that are closely associated with elections and make the idea of giving up on elections
entirely, too radical for many theorists. Even proponents of sortition often argue in favour of a
dual system in which one chamber of government is selected through sortition and the other
through elections. These three values of elections are: accountability, consent, and expertise.
In section 2.3 I discuss these values and their importance for democracy. Thereafter, I argue
that while the values of consent and accountability are indeed important for democracy,
elections only provide a weak version of these values. Furthermore, I argue that expertise
should not be seen as a value of democracy in the first place. Therefore, these values do not
constitute a strong argument for maintaining elections, not even in a dual system.
In chapter 3, I discuss the values of a sortition system and how these values could fix
our current problems with elections. The values I discuss here are: political equality, greater
quality of deliberation, and descriptive representation. In section 3.4, I list some basic
requirements that a system of sortition would need. While my aim is not to create an
institutional design for sortition democracy, I do find it important to list several requirements
that this institutional design would need in order to fully realize the values of sortition that I
discussed.
2. Electoral Democracy
In this chapter, I argue that elections are inherently undemocratic and inegalitarian. First, I
discuss the historic roots of elections to demonstrate that elections have not always been tied
7
elections are inherently inegalitarian and undemocratic. I focus on the inequalities produced in
a situation of choice, the competitive nature of elections and the distinction elections create
between representatives and represented. In doing so, I separate elections from representative
democracy. In the final section I discuss three values that are often being put forward in
defence of elections: consent, expertise and accountability. I argue that even these three
values are not being sufficiently realized within elections and therefore not capable of saving
elections from its flaws.
2.1. A discussion on the history of elections.
Today, free elections and democracy are being used almost interchangeably. Van
Reybrouck calls this ‘electoral fundamentalism’ (Van Reybrouck 2016: 42). We no longer
view elections as a method to achieve democracy, it has become a goal in and of itself. Van
Reybrouck demonstrates this with the example of western democracies hoping for
democratization of countries in the ‘global south’. In reality, it is not democracy which is
being advocated. Rather, it is free elections, in the exact same way we have them in the west.
This ignores village councils, traditional conflict mediation, ancient systems of justice etc. Because these forms of organization have nothing to do with elections, they do not ‘count’ as
indicators of democratization (idem: 42). When this system of elections fails, it is never the
system itself that is to blame, it is the country or people that failed to properly execute it
(idem: 43).
It is quite striking that elections are conceived as the only legitimate method of
selecting democratic government, considering the history of elections is relatively young
compared to the long history of democracy. If we look at the origin of the concept of
democracy, ancient Athens, it turns out that election had only a very small role to play in
government. Moreover, the role of elections was not perceived as democratic. I am not
8
democratic government. In Athens there were between 30.000 – 60.000 men who were
eligible to participate in democracy. Men under 30, slaves, and women did not have the
political right to participate. Universal suffrage is needed to give every citizen equal political
rights which is what we now understand democracy to be. Therefore, I would not consider
Athens to be the prototype of a democratic ideal.
Rather than taking Athenian democracy as the ideal democracy, I draw on the history
of ancient Athens, and then the emergence of representative government, to reduce the
primacy that elections have in our democratic imagination. I thus provide a genealogical
critique4 of elections as an inherently democratic method. In doing so, I demonstrate that
elections have not always been interpreted in the way we do now and that in this sense it
cannot be said that elections are essentially democratic.
2.1.1. Elections in Athenian democracy.
Athens’ governance was divided in several political institutions. The principal
assembly is the classic example of direct democracy. Around 6000 civilians would get
together and vote on new laws and vote for magistrates (Van Reybrouck 2016: 65), these men
would apply for it themselves. The selections method here was thus, self-selection. The
second governing institute was the council of 500. It was responsible for making new laws
and policies, preparing the principle assembly, checking the magistrates and engaging in
diplomacy. The council was selected through the drawing of lots. Third, the civic court
dispensed justice and tested the legality of the new laws coming out of the principle assembly.
These 6000 men were selected through the drawing of lots. The fourth institute was a body of
magistrates. The magistrates were responsible for the execution of laws and policies. 600 of
them were selected through the drawing of lots, the remaining 100 were elected by the
9
principal assembly5 (Van Reybrouck 2016: 65).
When looking at these institutes it stands out that most are selected through the
method of sortition rather than elections. The positions of magistrates that were filled by
elections were thought to require a level of expertise that could not be expected from just any
citizen (Mulgan 1984: 546). These were the generals, top military administrators and chief
financial officials (idem: 541). Even though these were influential positions they were
administrators and executors, the magistrates did not have any decisive power, this was in the
hands of the assembly and the courts. That elections were not the most frequent method of
selections in Athenian democracy is not strange considering they viewed it as an oligarchic or
aristocratic method rather than a democratic method. (Manin 1997: 27).
2.1.2. Elections as an aristocratic (or oligarchic) method of selection.
Aristotle discusses this idea of elections as oligarchic in his concept of the mixed
constitution in politics. He believed that combining both democratic as well as oligarchic
elements into a political system led to a better constitution than having only elements of one or
the other. In describing the elements that could be combined, it becomes clear that he sees
elections as an oligarchic method of selection and sortition as a democratic one. For example, he states: “the appointment of magistrates by lot is thought to be democratical and the election
of them oligarchical” (Aristotle 1947: IV,9). Or later, in reference to spartan government: “the
spartan constitution is said to be an oligarchy, because it has many oligarchical elements. That
all offices are filled by election and non by lot, is one of these oligarchical characteristics”
(ibid). Aristotle was not alone in this, influential philosophers of the seventeenth and eighteenth
century had similar thoughts about the use of lot versus the use of elections. Montesquieu and
5 This is a rather simplified account of Athens’ democratic institutions, for a more complex version see Athenian
10
Rousseau also wrote about lot as a democratic method of selection whereas they associated
elections with aristocracy.
Montesquieu stated, “the suffrage by lot is natural to democracy, as that by choice is to aristocracy” (Montesquieu 2004: 55). Rousseau agrees with Montesquieu on the nature of lot
and election (Rousseau 2002: 232). He furthermore states
When choice and lot are combined, the first should be used to fill the posts that require
special talents, such as military appointments: the other is suitable for those in which
good sense, justice, and integrity are sufficient, such as judicial offices, because, in a
well-constituted state, these qualities are common to all citizens. (ibid)
The revolutionaries in England, America and France, all looked at Rousseau and
Montesquieu in guidance of forming new republics, yet nowhere was lot given serious
consideration (Manin 1997: 80-81). The question is, how, when it was clear to theorists at the
time that elections were an aristocratic rather than democratic method of selection, did
elections triumph in the formation of governments as we know them today.
2.1.3 The demise of lottery and triumph of elections.
Manin argues that the demise of the lottery is only in part due to considerations of
feasibility. The most important consideration of feasibility would have been the size of the
new republics which were considerably bigger than the old city republics and Athens where
lot was previously used. However, lot could have been made into a multi-step procedure
where lotteries were first held in districts and a second round of those who had been drawn in
these districts (Manin 1997: 82). It would undoubtfully have been more complicated due to
size than it was in Athens but to say it was only due to considerations of feasibility obscures
that it was to a great extent due to considerations of desirability and a change in the
11
Besides feasibility constraints, Manin argues, it was also due to a shift in the concept
of legitimacy which led to the triumph of elections over lottery. The French, English, and
American revolutions were all accomplished in the name of the principle that “all legitimate
authority stems from the consent of those over whom it is exercised” (idem: 83). The
conception of political citizenship had shifted from the right of a people to rule itself to the
right of a people to choose their rulers. When this principle is held at the heart of legitimacy,
elections appear to be the logical choice. The value of equally distributing political power,
which is accomplished through lot, was set aside for the value of individual consent to
government, which was thought to be accomplished through elections (idem: 91).
Furthermore, there was little objection to a representative government based on elections because “compared to the gap that separated election and heredity, the difference
between the distributive effects of the two non-hereditary procedures (lot and election)
appeared negligible” (ibid). For Yves Sintomer, this account of why there was no objection to
elections is not entirely satisfactory. He states that it fails to explain why radical minorities
who campaigned for descriptive representation did not campaign for lottery (Sintomer 2018:
343). According to Sintomer this can be explained by the lack of statistical knowledge
available at the time. This meant that the connection between descriptive representation and
lottery could not yet be made (ibid).
Another reason lot was not deemed desirable for selecting government, was the belief
that “the representatives be socially superior to those who elect them” (Manin 1997: 94). This
belief can already be seen in the thresholds that existed for those eligible to vote, but even
more so for those who were able to represent. This belief that there should be a distinction
between those who represent and those who are represented shows that elections were never
12
partly chosen not despite, but because of these effects. It had a clear goal to beware of
democracy rather than facilitate it.
2.2. Inherently undemocratic aspects of elections.
The historical account I provided of elections demonstrates that whereas we now see
elections as inherently democratic, they were not always used or perceived as a method of
democracy. This problematizes its central place in democratic theory. To further argue why
elections, even today, are unable to realize an egalitarian democracy, I elaborate on three
kinds of challenges that elections pose to democracy. The first are challenges in a situation of
choice, the second are challenges of competition, the third are challenges of distinction.
2.2. Challenges in a situation of choice.
Within elections voters are put in a situation where they have to make a choice
between different candidates. This situation creates challenges to democracy that arise from
the position of the voter, as well as from the position of the candidate. The first challenge is
the “unequal treatment of candidates by voters” (Manin 1997: 136). This entails that you can
never prevent voters from making arbitrary decisions. The very nature of free elections makes
it so that nothing can prevent voters from discriminating candidates on individual
characteristics (Manin 1997: 136). When a society is racist and sexist there is no way of
stopping people from casting their votes based on sexist and racist beliefs. This means there is
no way of preventing voters from making biased decisions that reproduce structural
inequalities in society. This aspect of elections could attribute to the fact that all over the
world, electoral democracies show very low levels of descriptive representation. “Descriptive
unrepresentativeness appears to be an inherent feature of electoral democracy” (Malleson
2018: 403).
13
can be found on the side of the candidates. Malleson states “those who have the time, money,
connections, and profile required to run successful campaigns are likely to be, on average,
wealthy, educated, and from dominant social positions” (idem: 404). This is the second
challenge that a situation of choice poses to democracy. Because people must choose among
many possible candidates, those who are either already known or those who are best at
disseminating information will have an advantage. The dissemination of information is costly
and therefore, those with money have an advantage in elections (Manin 1997: 144).
Similarly, Malleson argues that electoral reforms can mitigate the effects of money,
but they can never take them away completely. This is because the main condition of success
in getting elected is becoming well known. Communicating your message and mobilizing base will inevitably be easier for those with more money. “Therefore, it seems likely that
electoral democracy, regardless of campaign finance rules, will always be somewhat biased
toward the rich (Malleson 2018: 405). This means that the distribution of political power
within electoral democracy will always be unequal.
2.2.2 Challenges of competition.
Gastil & Wright have argued that there are certain problems within elections that lead
to problematic behaviour within legislation. The biggest problem they point at is that there is
no room for honest deliberation within politics because of electoral competition. Candidates
who have won elections go into office with a set agenda from which it is hard to deviate. “Even when all parties recognize a problem’s optimal solution, one party’s leadership may
decide to block legislation because it cannot afford to grant the other political victory” (Gastil
and Wright 2018: 306). Therefore, policies are not maximizing voters’ interests. Similarly,
Van Reybrouck argues, perhaps a bit more cynical, that electoral politics has become a
popularity contest. Elections have turned into collective hysteria of commercial media, social
14
consequence, long term interests are being sacrificed for short term party interest. (Van
Reybrouck 2016: 52-54).
Another issue with electoral competition is that it might enhance the gender-gap in
politics. Research has shown women to be more hesitant to seek out competitive
environments (Niederle and Vesterlund 2007, 2003). Fox and lawless confirmed this and
stated that “women are significantly more likely than men to view most of the mechanics of
running for office, particularly negative campaigning, as deterrents to a candidacy” (Fox &
Lawless 2011: 66).
One possible way to fix this difference in competitive traits is by socializing girls and
women in the same way we socialize boys and men, thereby making women more prone to
competitive traits. However, While I do believe that a gender-neutral society is desirable, this
particular solution does not seem to be gender-neutral as such. When you socialize girls in a
way that makes them just as prone to seek out competition as boys, you are not striving for a
gender-neutral ideal. This ‘solution’ is just a way of making women more like men, in other
words, men remain the norm. Therefore, I argue that the biggest problem that needs to be
fixed is the system that makes a competition out of the distribution of political power.
2.2.3. Challenges of distinction.
I have argued that elections as a means of distributing political office, came out of the
idea that certain positions needed certain skills and expertise. With the birth of representative
government, it was believed that representatives should be superior to those they represent.
This conception might not be the aim of electoral politics anymore, it is still more or less what
electoral politics creates. Richard Mulgan argues that the prevailing principle of modern
democracy is equality of opportunity rather than strict equality of participation (Mulgan 1984:
556). This still rests upon the belief of people’s unequal abilities and “their relevance to the
15
distinction and distance between representatives and those represented is quite visible. A
study by Peter Kanne in 2011 showed that 87% of Dutch politicians viewed themselves as ‘innovative, progressive and internationally oriented. Whereas 89% of those same politicians
viewed ‘the Dutch people’ as more traditional, conservative, and nationalistic (Van
Reybrouck 2016: 11). This is one example in which politicians clearly believe themselves to
be, if not better, at least different from those they represent. Another example of this
distinction between the representatives and those represented can be found in the fact that
about 90% Dutch parliamentarians have a university degree [WO & HBO] (Volkskrant 2017).
Whereas only 30% of the general population has a university degree (CBS 2018). This
connection between assumed knowledge and experience with being deserving of political
office is problematic in the sense that it is extremely inegalitarian.
2.3. ‘Indispensable’ values of elections.
The most frequently used arguments to keep elections are that without them we would
lose the values of consent, expertise and accountability. These values are deemed to be
essential for a functioning democracy. In order to keep these values, it is often even argued by
sortition theorists themselves that sortition should exist in a system alongside elections. (Van
Reybrouck 2016; Gastil & Wright 2018; Mulgan 1984; Malleson 2018; Sintomer 2018). For
instance, Mulgan states, “no one [..] would dispute that election is the key device of modern
democracies or seriously claim that popular control over governments would be increased by appointing a legislature by lot from among the citizens” (Mulgan 1984: 554). This quote
demonstrates the prevalence of this line of thought. In this section I analyse the importance of
consent, accountability and expertise for representative democracy and how they are
established through elections. I demonstrate that elections are only able to provide a weak
16
‘expertise’ in representative democracy. In doing so, I argue that we should get rid of
elections entirely.
2.3.1. The value of consent.
The first democratic value that is thought to be established through elections is the
value of consent. This entails that democratic government is legitimate because those over
who it holds power, have consented to this power. Consent can be a way to secure our
autonomy while enjoying the benefits of social cooperation (Waldron 1987: 137). This idea
goes back to social contract theorists such as John Locke. Locke has said “men being, as has
been said, by nature all free, equal, and independent, no one can be put out of this estate and subjected to the political power of another without his own consent” (Locke 1967: 240).6 This
Lockean idea that the legitimacy of government is derived from the consent of the governed
has had a strong influence on our current idea of legitimacy. James Gardner has argued that
both the US constitution and the US declaration of independence have been heavily
influenced by a Lockean sense of legitimacy (Gardner 1990: 193). This indeed seems to be
the case in the following passage of the declaration of independence:
We hold these truths to be self-evident, that all men are created equal, that they are
endowed by their Creator with certain unalienable Rights, that among these are Life,
Liberty and the pursuit of Happiness.--That to secure these rights, Governments are
instituted among Men, deriving their just powers from the consent of the governed,
--That whenever any Form of Government becomes destructive of these ends, it is the
Right of the People to alter or to abolish it, and to institute new Government, laying its
6 This seems to indicate that for Locke the only legitimate political power, is that of which men have
consented to. However, Pitkin has argued that contract theorists were not speaking of actual consent but rather hypothetical consent “you are obligated neither by your own consent nor by that of the majority, but by the consent rational men in a hypothetical ‘state of nature’ would have to give” (Pitkin 1965: 999). This would bring the question of legitimate government back to the nature of the government itself and not so much to how it was formed or who actually consented to it (ibid).
17
foundation on such principles and organizing its powers in such form, as to them shall
seem most likely to affect their Safety and Happiness. (US 1776: para 2)
In this passage it becomes clear that the legitimacy of representative government is directly
related to the consent of the governed.
There is a strong assumed link between consent and elections in that “an election is a
specialized sovereign act of choice by the people; it is the means by which the people exercise
their sovereign right to choose the agents who will wield governmental power” (Gardner
1990: 222). This aspect of elections would be lost within a system based on sortition.
However, I argue that elections are still only able to provide a weak version of consent.
Furthermore, consent is not the only value of democracy, and if it can only be provided in a
weak form, it might be preferable to trade it for a strong version of political equality which
can be established through sortition.
The argument that elections legitimize representative government by letting people
consent to it is troubling when we see low voter turn-out percentages all throughout western
democracies. In the United States presidential elections typically have a voter turnout of
around 50%, and mid-term senate and congressional elections have even lower turnouts with
percentages shifting between 30% and 40% (Hill 2006: 207). As Testart has put it, in response to the critique of losing legitimacy of consent in sortition democracy, “this is a laughable
criticism in the context of a system where elected officials are often chosen by as few as 30% of the electorate” (Testart 2011: 121). Europe is doing considerably better regarding voter
turnout, in the first decade of the 21st century, the average voter turnout for national elections
was 77% (Van Reybrouck 2016: 14). Nevertheless, this still comes down to millions of
Europeans not casting their vote. If voting is a form of consent, it can at the very least be
18
turnout shows that elections as they are, do not guarantee legitimacy of consent to
representative government.
It could be said that not-voting is a form of giving consent just as voting is. If people
have the possibility to vote, not-voting could be seen as a way of expressing satisfaction with
government. The assumption is that, because voting can be done at relatively low cost, people
who do not vote would start voting if they wanted things to change. In this view of consent,
the mere right to vote would be sufficient to establish consent (Gewirth 1962: 138).
However, this argument makes voting a mechanism of protest rather than consent because it does not require an active ‘yes’. Elections are not really necessary to maintain this
position. Following this argument, it could be said that within a system of sortition, whenever
people are complying without voicing dissent, it can be taken as consent. Moreover, it is not
at all clear that people who do not vote are motivated by feelings of satisfaction. At times
certain groups do not vote because they have lost faith in a system that has excluded and
oppressed them. (Taft 2006) or because they feel alienated from the political process
(Kimberlee 2002), or simply because none of the available choices appeal to them (Coward &
Mitra 2018). When there are so many different reasons as to why people might not make use
of their right to vote, it cannot convincingly be said that not-voting is a form of consent.
Therefore, low voter-turnouts are a problem for the idea of legitimacy based on consent.
Second, Simmons argues that voting is not a way of giving consent, but rather, of
giving a preference (Simmons 1984: 800). Within elections, voters do not choose between all
possible options, they choose between the options that are offered to them. Simmons states:
if the state gives a group of condemned prisoners the choice of execution by firing
squad or by lethal injection, and all of them vote for the firing squad, we cannot
19
This relates to elections because the candidates that are offered to voters are often not
everyone’s first choice. When people vote, they choose among the options that have been
handed to them. This could mean that they consent to the candidate for whom they vote, but
sometimes it just means that they are preferring a bad candidate over a selection of even
worse candidates. As Schattschneider put it, within representative government the people are “a sovereign whose vocabulary is limited to two words, ‘yes’ and ‘no’.” (Schattschneider
1942: 52). This can hardly count as true and free consent.
Finally, the concept of informed consent further complicates the matter. Informed
consent is most often used as a concept in research concerning research subjects. “an
informed consent occurs if and only if a patient or subject, with substantial understanding, and
in the absence of substantial control by others, intentionally authorizes a professional to do something” (Beauchamp 2010: 57). Autonomy is of vital importance to informed consent.
In a political context this directly relates to the problem of uninformed voters. Ilya
Somin states that an uninformed electorate negatively impacts democracy in two ways. First,
when people do not know what they are voting for, government is unable to reflect the will of the people in any meaningful way, “undercutting the intrinsic case for democracy as a
government that reflects the voluntary decisions of the populace” (Somin 1998: 414). Second,
the ignorance of an electorate opens the door for manipulation by elites of the public (ibid).
Research has shown that the American electorate is far from the ideal of an informed
electorate (Fowler and Margolis 2014: 101). From studies in the 60’s 7 it became clear that
citizens did not know what the parties stood for nor their main points. This did not improve
over the decades, despite an increase in general level of education (Fowler and Margolis
2014: 101). Studies of voter knowledge in other countries have been rare but the results were
20
not much different from the American context (Somin 1998: 417). It has been argued that
voters can use information shortcuts or that uninformed voters cancel each other out in large
electorates, thus minimizing the effects on electoral outcomes. Whether this is the case
remains unsure (Fowler and Margolis 2014: 101). Nevertheless, Whether the outcomes of
elections are significantly different with or without an informed electorate is not entirely
relevant for the concept of consent. When the premise is accepted that you cannot consent to
something when you do not know what it is you are consenting to, and in order to give full
autonomous consent one has to be informed, there can be no consent through elections with
an uninformed electorate.
In sum, elections cannot provide representative government with a strong version of
consent because there are too many people who do not vote for unknown reasons, voting is
sometimes a preference rather than a free choice, and many voters are uninformed. In
showing that consent is currently only provided in a weak form I argue that we should look at
other values of democracy. We need to shift the source of legitimacy and bring it back to
radical political equality. I elaborate on how this is accomplished within sortition democracy
in the third chapter of this paper.
2.3.2. The value of expertise.
The second argument that is used against a system based solely on sortition is that
elections provide government with capable representatives (Grant& Keohane 2005: 32). This
argument seems to be technocratic rather than democratic, but it is frequently used by
sortition theorists themselves who argue for a dual system of sortition and elections (Gastil &
Wright 2018; Mulgan 1984; Malleson 2018). Competence or political expertise is assumed to
be important because policy issues often involve complex trade-offs (Malleson 2018: 410).
Whereas elected politicians have often been involved in politics for a long time, sortition
21
The argument presupposes that there is a certain amount of knowledge and expertise
needed for political office. I argue that this should not be necessary. Mainly because politics
should be about choices, and these choices cannot always be made on a cost-benefit analyses because they often rely on ethical judgments. As Dahl states: “decisions about policies,
whether personal or governmental, almost always require judgments about trade-off’s, a balancing of different ends” (Dahl 2002: 72). Therefore, I want to propose a view of recursive
representation (Mansbridge 2018).
Within recursive representation the representative would serve as an interlocutor.
Meaning, her main job would be to communicate. “the representative as interlocutor links the
representative system together less by making policy herself than by helping those in all the
other parts of the system understand each other” (Mansbridge 2018: 299). This concept would
respond to the perception of many citizens that they’re not being heard and “the reality that in practice representatives very often do not hear those voices” (idem: 305). In Mansbridge’s
view of representative government it is not democratically lacking because representatives do
not stand for the will of the constituents but because they have lost contact with them (idem:
307). Therefore, we should provide representatives with expert staff to make policy and leave
more time for the representative to communicate with each other and their constituents (idem:
310).
I further argue that even the norms for what counts as good political communication
are biased. The rules of reason and deliberation in politics themselves are racist and sexist and
therefore we cannot select people who are considered to be good at playing by these rules to
be representatives. As Iris Marion Young states:
their institutional forms, rules and rhetorical and cultural styles have defined the
meaning of reason itself in the modern world. As ruling institutions, however, they
22
reason and deliberation, both in the institutions and in the rhetorical styles they
represent. Since their enlightenment beginning, they have been male-dominated
institutions and in class and race differentiated societies they have been white- and
upper class- dominated. (Young 1996: 123)
An example of such a norm is the dominant view within political theory that progress
and politics is dependent on reason and separated from emotion (Marcus 2000: 221). This
norm privileges speech that is disembodied and free from emotion (Young 1996: 124). As
Susan Bickford describes it “the most familiar story stresses the antagonism between reason
and emotion and regards emotion as the dark chaotic force that the fully rational self-triumphs over, separates from, or governs” (Bickford 2011: 1026). Furthermore, this dichotomy
between emotion and reason is often gendered and racialised, “this ideology of emotion
interacts with and fortifies ideologies of race and gender, about who counts as the mature self and who is weak, dangerous, and in need of being governed” (ibid).
Exposing the biases in our qualification systems of what counts as good and what
counts as bad is important because it shows that we cannot make an a priori decision of
whether someone will be a good representative. This is not to say that there won’t be differences in people’s performances in office. It is merely to say that we cannot, giving the
inequalities in society at large, predict who will be ‘good’ at politics and who will be ‘bad’.
When these decisions are made within electoral democracies, they inevitably reproduce
structural inequalities in society. Therefore, elections cannot provide us with ‘good’
politicians. The best way of securing good policy outcomes under such circumstances is to
provide the right tools for informed deliberation. In the third chapter of this paper I elaborate
23 2.3.3. The value of accountability.
The last important value of elections is accountability. Politicians have goals and
interests of their own. To avoid politicians from serving their own interests rather than those
of the people they represent it is important to have a system of accountability in place
(Przeworski, Stokes & Manin 1999: 28-29). Warren argues that democracy necessarily needs systems of accountability because there is a “condition of vulnerability to power”, those who
are affected by the exercise of power have to be able to sanction those who wield it (Warren
2014: 3). Within society, social cooperation is necessary but always at risk of exploitation and
oppression. While government is owned by the people, collective decision power is in the
hands of elites (idem: 40). This creates the need for accountability.
One way to achieve accountability is through electoral competition. When
representatives ignore their electorates wishes, the electorate has the power to vote them away
in new elections (Franklin, Soroka and Wlezien 2014: 392-393). Grant and Keohane have
conceptualized two models of accountability, the participation model and the delegation
model (Grant and Keohane 2005: 31). Within the participation model, power is only
legitimate when the outcomes of decisions reflect people’s desires (ibid). it is a form of
legitimacy that rests on full participation and equal treatment of everyone. Direct democracy
would be the ideal form of government in this model and representative government is only
necessary for reasons of scale (ibid). Within this model, power-wielders would be held
accountable through direct participation (idem: 33). The second is the delegation model. This
model grounds the legitimacy of power in the consent of those who delegate it (idem: 32). The advantages of delegating power to those “most able to govern” (ibid) are often noted.
What is less often recognized is the separation between governed and governors, this
24
Elections are a combination of both of these forms of accountability, people participate
in selecting those who should be in government and judge their former delegates (idem: 33).
Franklin, Soroka and Wlezien offer two different models of accountability in elections8. The
first is the responsible party model. This is simply the idea that parties make promises to their
electorate and when they fail to keep these promises the electorate can sanction them by
voting them out of office (Franklin, Soroka and Wlezien 2014: 392 -393). The second
accountability model in elections is the thermostatic model. This works through public
opinion polls. Through these polls, governing parties get an idea of what is considered by the
public as “good policy”. Because re-election is at stake, governing parties have a strong
incentive to pay attention to public opinion. So rather than the actual sanctioning, this model
works through the threat of sanctioning (idem: 394-395).
Whether it is worth sacrificing these forms of accountability that are achieved through
elections, depends on how well elections perform in actually delivering them. I argue that
there are several aspects of elections that complicate its functions of accountability and
therefore elections are only able to provide a weak version of accountability.
The first problem for accountability through elections is that financial elites and
business interests hold too much power over elections and politics in general. It could be
argued that public officials are currently more concerned with their accountability to elites
than with accountability to their electorate. An influential study from Gilens and Page showed
that within American politics, wealthy elites and groups representing business interests have far more influence on American policy than ‘average citizens’ or mass- based interest groups
(Gilens & Page 2014: 564). The fact that public officials are not being held accountable for
8 These models are not in opposition to the participation model and the delegation model from Grant and
Keohane (2005). Whereas the participation and delegation models are based on a distinction between two models of legitimacy and the accountability that flows from it, these are based on the distinct forms of accountability that are specific to elections.
25
this by their constituents could be explained by the argument that it is not in voters’ interest to
do so. Because voters generally do not care, or do not know about policies that have a small
impact on their lives, politicians can sell policies to financial elites. These policies would have
a small cost on each individual voter and concentrate benefits on special interest groups
(Przeworski, Stokes and Manin 1999: 34). This means that it is not in the voter’s interest to
hold public officials accountable for prioritizing special interests over the interests of the
electorate (ibid). This argument mostly affects the responsible party model (Franklin, Soroka
and Wlezien 2014: 392 -393), because it shows that the electorate fails to sanction governing
parties that do not act on their behalf.
To a certain extent, accountability also relies on an involved and informed public. In
general, research of politically informed citizens suggests that levels of informed citizens are
low in most democratic countries (Tyson 2016: 1017). When people are uninformed about policies’ effects, they won’t hold those responsible for them accountable. “Alienating citizens
from the political process (or more specifically, reducing their level of involvement to the
minimum possible) may produce systems where, as Schumpeter points out, vested interests have the sole influence on policy” (Moncrieffe 1998: 397). Przeworski, stokes and Manin
share this concern and argue that citizens never have the same information on policy decisions
that policy makers do. When voters do not have this information there is no way of knowing
whether governments are working in their best interest and they cannot sanction them
appropriately (Przeworski, Stokes and Manin 1999: 40).
Therefore, “elections are not a sufficient mechanism to ensure that government will do
everything they can to maximize citizen’s welfare” (idem: 50). A lack of information is thus a
problem for both the responsible party model and the thermostatic model (Franklin, Soroka
and Wlezien 2014). Parties cannot be appropriately sanctioned without the electorate being
26
Furthermore, Fearon argues that people view elections as a method of selecting ‘good candidates’ much more than as a method of accountability. He sees “the popular dislike of
office seeking, support for term limits, the premium put on politicians being principled and
consistent, and the absence of concern about last period effects” as indicators that voters see
elections primarily as a way of selection (Fearon 1999: 82).
Finally, elections are unable to provide a strong version of accountability because they can
only hold representatives accountable after the effects of unwise or unjust policies are already
being felt. Politicians are not in any way bound to electoral promises. Once they are in office,
they are free to do the exact opposite of what they promised during election campaigns. “voters can sanction deviations from mandates only after their effects have been experienced”
(Manin, Przeworski and Stokes 1999: 39). It could be argued that the thermostatic model
(Franklin, Soroka and Wlezien 2014) would prevent this from happening, that the threat of
sanction is enough, but there is no guarantee that it will be enough as long as there are no
strong institutional mechanisms to hold parties or politicians directly accountable.
All in all, elections are currently not able to provide any strong accountability because
financial elites and business interests have too much influence on electoral politics, the electorate doesn’t have the information needed to judge policies and they do not primarily
view elections as a method of accountability. Thus, while accountability remains an important
value for representative democracy, it does not make a strong case for elections. In the third
chapter of this paper I return to the possibilities for accountability within sortition democracy.
2.3.4. Getting rid of elections entirely.
While even theorists of sortition have been hesitant to say goodbye to elections entirely, I
argue that we should. I have demonstrated that consent cannot be sufficiently realized through
27
namely the value of political equality. Furthermore, selecting ‘skilled politicians’ is not
necessary for representative democracy, moreover, it would enhance inequality. Finally, I
have argued that accountability, whilst remaining an important value for representative
democracy, is also not fully realized in elections. With ‘expertise’ being undesirable and
consent and accountability failing, these values don’t make a strong claim for elections, not
even in combination with sortition.
3. Sortition Democracy
So far, I have argued that elections were historically considered aristocratic. Furthermore, the
very nature of elections poses challenges to democracy. Therefore, I propose replacing
elections with sortition. Selecting representatives through sortition would mean that the seats
in political office which are now being distributed through elections would be distributed
through a lottery among all citizens. While we currently have trouble imagining a government
made up of ordinary people selected by chance, lot has been a widely used method of
selection in the past. It was widely used in Athens but also in the republics of Venice,
Florence, Aragon and to a smaller extent in the roman empire (Manin1997: 44; Dowlen
2008). While its main use in the city republics was to avoid factions, in Athenian democracy
it was also used for the equal distribution of political participation (Dowlen 2008).
In this chapter I argue that a system of sortition should replace elections as the
democratic method of selecting representatives. Within a sortition democracy, governments
get more descriptively representative, the conditions for deliberation improve, and political
28 3.1. Political equality in sortition democracy.
The first argument for a system of sortition is that it creates greater political equality.
Mulgan states that one important reason for using sortition within Athenian democracy was that “it expressed the democratic commitment to political equality” (Mulgan 1984: 545).
Sortition creates the equal right to rule which renders every citizen equally deserving to
govern (Manin 1997: 40-41). This concept relies on the idea that political office can be seen
as a type of good that citizens may have a right or a claim to (Stone 2016: 341). A sortition
government gives everyone the equal right and probability of obtaining this type of good, i.e.
political power 9(Manin 1997: 40). Within electoral democracy people are not seen as equally
deserving of political office (Stone 2016: 341). Whereas elections require and create
superiority, sortition requires and creates equality.
The political equality created by sortition is also a way of realizing democracy as
government by the people. In Athens the use of lot was a way to realize self-government. Due to the scale of today’s democracies this is no longer something that sortition can accomplish
(Sintomer 2018: 338). Chances of being sorted are very small for individual citizens.
However, a modern use of sortition can create government by the people because it selects
average citizens into government rather than career politicians. This removes part of the
distinction within electoral democracy between those who are governed and those who
govern.
An important effect of this equal chance to be sorted is that the manipulation of wealthy
elites on the selection of representatives, that is present in an electoral system, is eliminated
(Dowlen 2008: 224). “choosing political representatives by random selection could
9 This equal probability of obtaining political office is slightly altered when sortition is used as a stratified
sample. While still creating a far greater equality in the chance of being selected as a representative than there exists within elections, it will no longer be perfectly equal.
29
immediately breakup the networks of power, lobbying, and patronage surrounding elections”
(Malleson 2018: 405). Because the sortition mechanism treats everyone as equal, the
advantages that elites hold in elections will disappear and political power will be more evenly
distributed.
3.2. Deliberation in sortition democracy.
There are also good reasons to assume that a sortition chamber would produce a higher
quality of deliberation than we are seeing in elected chambers. I argued that deliberation is
low in elected chambers because politicians are more interested in short term successes than
long term progress. Moreover, they are bound to their parties and their electoral promises,
leaving very little room for genuine deliberation. Sortition can overcome this through
recruiting nonpartisan people who have no interest in advancing their own political career
because there will no longer be such a thing as a political career (Sintomer 2018: 352).
Without being tied to political parties and their constituencies members of a sortition chamber
would be free to enter into open deliberation with one another.
While within an electoral chamber it is nearly impossible for a politician to change
their minds on a subject because they are accountable to their party and their electorate, there
are no constraints for a member of a sortition chamber to change their minds on an issue once they have learned new things or when they are convinced by another member’s argument
(Malleson 2018: 409). Being able to change your mind is essential to good deliberation.
Moreover, a sortition chamber would be far more likely to solve political problems with solutions that can transcend electoral divides because sortition members don’t start their term
with a fixed agenda (Gastil &Wright 2018: 307).
Previous experiences with random sampling have shown that it is indeed true that
average citizens, when handed the right tools, demonstrate high quality deliberation (Sintomer
30
can be found in Fishkin’s deliberative polls. The deliberative poll was a way of combining
deliberation with the equality of random sampling in polls (Fishkin & Luskin 2005: 287).
Within these deliberative polls, randomly selected citizens were asked to discuss policy
proposals as a way to poll public opinion under conditions where citizens were given
sufficient information and a chance at deliberation (idem: 288). The results show that overall,
the average citizen is very capable of high-quality deliberation (idem: 290 -293). Other
examples include the use of mini-publics and citizen juries (Sintomer 2018).
Besides the absence of party discipline and the incentive of re-election, the quality of
deliberation within a sortition chamber is also improved through the diversity of the chamber
(Gastil & Wright 2018: 307). “good deliberation must include diverse points of view, so that the range of arguments considered will be broader and discussion more inclusive” (Sintomer
2018: 353).
3.3. Descriptive representation in sortition democracy.
In section 2.2.1 of this paper, I pointed out that elected chambers are demonstrating
very low levels of descriptive representation. Sortition on the other hand, when used as a
representative sample, will be more descriptively representative of the population at large. In
the introduction I discussed Pitkin’s conceptualization of representation. Pitkin argued that a
descriptive view of representation is inadequate because it merely takes into account in which way a representative “stands for” the represented, whereas a substantive view of
representation takes into account how representatives “act for” the represented (Pitkin 1967).
I agree that descriptive representation in and of itself should not be an end for
representative government. However, descriptive representation can lead to more substantive
representation. I have touched on this quality of sortition while discussing the perceived value
31
politics, good and bad arguments, or good and bad deliberation and that therefore we should
not select our representatives based on a biased judgment of their expertise.
Sortition could make a start at challenging the biased norms through including a more
diverse group of people in political conversation and political power. The bias within the
norms of political communication remain a problem within a system of sortition. However, whereas elections serve to keep people who are not seen as ‘good representatives’ out of
office, sortition lets them in. What sortition can do is create a basis for changing the norms.
Namely, getting people from different social positions in positions of power. Once in office,
there will still need to be attention to the norms of deliberation as not to exclude people whose speech is seen as ‘other’, I do not expect this to happen organically in a system of sortition
either.
I draw on feminist epistemologies for this conception of politics. Donna Haraway
argues that the best way to produce something close to objective knowledge is to include all
situated knowledges and to acknowledge that knowledge is in fact situated and not produced
in a vacuum (Haraway 1998). Knowledge is situated because different social positions produce different experiences. These experiences influence people’s assumptions and
interests (ibid). Young argues that this is relevant for politics because in a situation of “structural relations of privilege and disadvantage” the inclusion of differentiated positions
provides important resources for democratic communication. “it maximizes the social
knowledge available to a democratic public, such that citizens are more likely to make just and wise decisions” (Young 2000: 115).
Anne Phillips makes a similar argument in defence of a politics of presence (1995).
Phillips argues that “representatives only ‘really’ represent their constituents on the issues that were explicitly debated in the course of the election campaign” (idem: 47). On issues that
32
own experiences of, for instance, being a woman, matters (ibid). Moreover, some issues might
never come up without the presence of people that are now excluded (Philips 1995: 51). This
is why descriptive representation can be necessary to accomplish substantive representation.
Everyone brings their own situated knowledge, produced by their own experiences in life, and
all of these experiences are relevant in politics. This politics of presence is what sortition
could provide when used to create a representative sample of society.
The use of sortition as a way of creating a representative sample of society is not
uncontested. Dowlen argues that this is a weak use of lot. Dowlen identifies arationality10 as
the main advantage of lot because it grands impartiality and protects the selection of
representatives from any manipulation (Dowlen 2009: 308). This arationality is compromised
with the rationale of a representative sample (idem: 306). An obvious effect of this is that the
political equality, which I pointed at as one of the advantages of a sortition system, is
compromised. When using sortition as a representative sample, not every individual would
have the same chance of getting sorted. Within a stratified sample the chance of an individual
belonging to a majority group getting sorted is smaller than the chance of an individual from a
minority group getting sorted because a certain percentage of people from this minority group
get sorted into government. This means that even within a sortition government, when using a
stratified sample, the individual chances are not fully equal.
However, in large polities, I argue that the advantages of a descriptively representative
body outweigh this aspect of arationality. In the context of modern nation states, where in
most cases millions of citizens would participate in the lottery, the realistic individual chance
of getting sorted is already very small. Because in large populations the chance of getting
sorted is very small, the difference that a stratified sample makes in individual chances of
10 Dowlen uses arationality rather than irrationality because whereas irrationality has a negative connotation,
33
getting sorted is miniscule. The difference between a government consisting of only white
men and a government that is similar to society at large, on the other hand, could be
enormous. Therefore, I believe the argument of a descriptively representative body is stronger
than that of complete arationality. This does not mean that the value of political equality is no
longer a relevant concept for sortition democracy. Rather, it means that the concept of
political equality is interpreted more broadly. It takes into account, not just the individual
chance of getting sorted into government, but also the relational power between groups in
society. In doing so, it responds more adequately to societal structural inequalities11
3.4. Basic requirements.
The aim of this paper has not been to provide a blueprint of a sortition democracy. Rather,
the aim has been to demonstrate that democracy, as government of the people, by the people,
for the people, is not well achieved through elections and would be better realized through
sortition. Elections inherently create challenges to democratic equality and sortition could
remove these challenges. Moreover, whereas Many advocates of sortition have proposed a
dual system of elections and sortition (Mulgan 1984; Malleson 2018; Gastil and Wright; Van
Reybrouck 2016; Sintomer 2018) I argued elections should be abolished entirely. In the
second chapter of this paper I demonstrated that even those aspects of elections that are
thought to be valuable to democracy, are not fully realized within elections. When combined
with elections, sortition would lose its effect of radical political equality because the
challenges of choice, competition and distinction, as indicated in the second chapter of this
paper, would remain. Thus, to realize its full potential, sortition should function as the only
method of selecting representatives. More effort should be directed at creating an institutional
design based solely on a system of sortition
11 For a more extensive discussion of why we should approach equality this way, see, for example: Anne Phillips
34
Instead of giving a blueprint for sortition government I have provided an argument for
why we might want a blueprint of sortition government. However, throughout the paper I
have hinted at several other requirements a sortition government would have to meet. In this
section I shortly outline some basic requirements that an institutional design of sortition
government would have to take into account.
1. Measures of accountability: In section 2.3.3. of this paper I underlined the importance
of accountability in representative government. I furthermore argued that elections
only provide a weak version of accountability. A system of sortition would not in itself
produce much stronger versions of accountability. While experience with minipublics
has shown that people feel very much responsible to the public at large, arguably more
so than elected politicians (Sintomer 2018: 353), this does not guarantee
accountability. The rotation that takes place in a system of sortition can provide some
sense of accountability. In Athens, rotationascertained that those who governed today,
would be governed tomorrow (Manin 1997: 30). This would ensure that “those in
office had an incentive to take the views of the governed into account: the man giving
the orders one day was discouraged from lording it over his subordinates knowing the next day he would be the subordinate” (idem: 30). While this might indeed be a good
incentive, it is still no guarantee. Further research would have to focus on ways for
citizens to directly hold representatives accountable.
2. Stratified sampling: in order to make sure that a sortition government would be more
descriptively representative, the selecting would have to be slightly more complicated
than a simple lottery. This descriptive representation could be accomplished through
stratified sampling. Sufficient attention must be paid to the process of choosing which
