Sortition, voting, and democratic equality

Abstract

In recent years, democrats both inside and outside the academy have begun to reconsider the merits of the age-old practice of sortition, the random selection of political officials. Despite this fact, however, the comparative assessment of the merits of voting and sortition remains in its infancy. This paper will advance this project by treating the problem of assigning public responsibilities as a problem of allocative justice. To treat the problem in this manner is to treat public office as a type of good to which citizens might have various claims. Random selection is the appropriate method for distributing public office when all citizens have equal claims to that office and there is not enough to go around. Universal distribution is more appropriate when all claimants have equal claims to the office and there is enough to go around (as with universal suffrage, for example). Election (or possibly other procedures, such as appointment) makes sense when citizens do not enjoy equal claims to the office and that office is in scarce supply. This approach captures a crucial component of democratic equality. Different understandings of democratic equality lay behind sortition and election. Each might be appropriate under different circumstances, but both place rights-based constraints on the design of a democratic political system.

Keywords:

• democracy
• election
• equality
• justice
• lottery
• random selection
• sortition
• voting

Acknowledgement

Earlier versions of this paper were presented at the General Conference of the European Consortium for Political Research (ECPR), University of Glasgow, 3–6 September 2014 and the Annual Meeting of the Political Studies Association of Ireland (PSAI), Galway, 17–19 October 2014. I would like to thank participants at these meetings for their helpful comments and suggestions.

Funding

This work was supported by a grant from the Arts and Social Sciences Benefaction Fund at Trinity College Dublin.

Notes

Earlier versions of this paper were presented at the General Conference of the European Consortium for Political Research (ECPR), University of Glasgow, 3–6 September 2014 and the Annual Meeting of the Political Studies Association of Ireland (PSAI), Galway, 17–19 October 2014.

1. See Manin (1997) for a sophisticated account of how democracy became associated, not with sortition, but with voting.

2. Leading books on the theory of sortition include Carson and Martin (1999), Duxbury (1999), Goodwin (2005), Dowlen (2008), and Stone (2011a). Stone (2011b) offers a useful collection of papers on the subject.

3. Political scientists have conducted many experiments to date with randomly selected ‘deliberative opinion polls’ (Fishkin 1991, 1995, 2009), ‘citizen juries’ (Crosby et al. 1986) and ‘planning cells’ (Dienel 1995). Most of these decision-making bodies have held advisory powers only. Some such bodies, however – notably, the British Columbia Citizens’ Assembly on Electoral Reform – have enjoyed a sort of real proposal power (Warren and Pearse 2008).

4. Reform proposals making prominent use of sortition include Leib (2004), Burnheim (2006), Barnett and Carty (2008), Callenbach and Phillips (2008), Sutherland (2008), Buchstein and Hein (2009), Zakaras (2010) and Guerrero (2014).

5. Malkopoulou (2015) is an exception. Malkopoulou, however, attempts a very high-level, comprehensive comparison of the democratic credentials of sortition and voting. My focus here, as will be seen, is on one specific part of the democratic ideal.

6. Many libertarians, notably F.A Hayek, object to the term ‘distributive justice’, because they believe it reflects a prior commitment to the idea that somebody must do the distributing, and decide how things are to be distributed. But liberal egalitarians such as John Rawls have employed the term for some time without in any way committing themselves to this conclusion. For this reason, it makes sense to adopt a different term when explicitly trying to make the commitment that Hayek et al. wish to avoid.

7. Still, it is questionable whether money is the sort of good on which people can have claims that a polity is obliged to satisfy.

8. Obviously, there exist possibilities between these two extremes. A good may be ‘semi-lumpy’, such that it loses some but not all of its value when divided.

9. The example comes from Broome (1990–1991).

10. I assume there are good reasons for not producing more of the good. Perhaps the good is impossible to produce in larger amounts. Perhaps it would be contrary to the public interest to sacrifice too many social resources for the sake of supplying this one good. Or perhaps there exists another good upon which people have claims, and producing more of the first good would impede or limit production of the second good.

11. I refer to this elsewhere as the Just Lottery Rule (Stone 2011a; see also Stone 2009a, b).

12. Broome (1990–1991). Broome complicates his account to deal with some of its more implausible conclusions – e.g. that B might receive the good instead of A even if A’s claim is 100, or 1000, times stronger.

13. Rawls does not employ the term ‘allocative justice’ to refer to these principles – principles relevant when individuals have claims to goods. Interestingly, Rawls does employ the term in A Theory of Justice. There he writes that ‘allocative justice applies when a given collection of goods is to be divided among definite individuals with known desires and needs…Since there are no prior claims on the things to be distributed, it is natural to share them out according to desires and needs, or even to maximise the net balance of satisfaction’ (Rawls 1999, p. 77). Rawls does not explain why such a situation should involve no ‘prior claims’ on the goods in question, or why desire satisfaction should be the natural allocative criterion.

14. For a similar defence of sortition, see Carson and Martin (1999, p. 34).

15. Ancient Athens, on some accounts, had such an understanding of citizen equality (Mulgan 1984). In the next section, I shall discuss this understanding as capturing a prominent understanding of democratic equality.

16. This fact, I believe, helps to account for the way in which the practice of voting has both democratic and aristocratic components, as argued by Manin (1997).

17. The right to vote in a modern democracy is thus awarded using an admission procedure, whereas the right to hold randomly selected office is awarded using a selection procedure. On the distinction between admission and selection, see Hofstee (1990).

18. The argument here does not depend upon the use of ‘winner-take-all’ voting rules, such as plurality rule with single-member districts. All voting rules – including all the various sorts of proportional representation – generate winners and losers. If they did not, there would be no need for elections.

19. The same is true for the most obvious alternative to sortition or voting – appointment, via some sort of non-electoral competitive procedure (like a civil service examination). Space precludes further consideration of this option here.

20. Does attracting votes determine the strength of a candidate’s claim to an office? Or does attracting votes merely indicate that voters believe the candidate has the properties that generate a claim to the office (e.g. intelligence or ability)? The analysis of voting offered here does not depend upon an answer to this question.

21. Note that it is very common for ties in elections to be broken by lottery. For many examples of this practice, see Stone (2011a, pp. 3–5).

22. Lottery voting, otherwise known as the ‘random dictator’ rule, may be viewed as an extreme form of enfranchisement lottery (Gibbard 1977, Amar 1984, Saunders 2010).

23. Indeed, one of the reasons why democracy has struggled to entrench itself in much of the world (e.g. in sub-Saharan Africa) is that political parties all-too-easily become vehicles whereby a dominant ethnic group or tribe uses the state to enrich itself at the expense of everyone else. Stabilising democracy seems to require ‘lowering the stakes’ of politics, such that losing an election does not mean the end of all benefits for a party and its followers. See North, Weingast and Wallis (2012).

24. Of course, if political office were simply a burden, then the allocative justice perspective would also be justified. The absence of a burden can be treated as a benefit or good, the distribution of which will surely involve considerations of justice (cf. Sher 1980). The modern attitude, which treats politics as a profession like any other, is more complicated than this.

25. As Amartya Sen notes, democracy has both an intrinsic and an instrumental value (Sen 1999, ch. 6).

26. Proponents of epistemic democracy locate much of democracy’s appeal in its ability to bring ‘the wisdom of crowds’ to bear upon political problems (e.g. Landemore 2013). But even if epistemic democrats are correct, this does not explain why denial of the right to vote or hold office to a specific individual constitutes a special wrong to that individual, as opposed to the polity as a whole (which thereby loses access to that individual’s contribution to collective wisdom).

27. For a similar effort to parse the various contributions sortition can make to democracy, see Stone (2011a, ch. 6).

28. It should be noted, however, that in ancient Athens the typical jury was comprised of 501 citizens, with even larger juries being employed in cases of great public importance.

29. Should the terms of an office be shortened to the point that every citizen can reasonably expect to take her turn holding it, then random selection has effectively been supplanted by rotation. Rotation and random selection are often conflated, but they work in very different ways, even though they can complement each other under the right circumstances. See Stone (2011a, §5.3) on this point.