Lottery and Legislative Powers: A Reply to Yoram Gat

In his recent blog post, “The Elected Legislator’s Burden,” Yoram Gat challenges one of the arguments of my essay, “Lot and Democratic Representation.”  In that essay, I argue that the U.S. Senate (along with state Senates) should be abolished and replaced with a citizens’ chamber, with its members chosen by lottery. In short, I propose that we preserve bicameral legislatures, but with one chamber filled through election and the other by lot. I argue, however, that the citizens’ chamber should have fewer powers and responsibilities than the elective chamber. It should have the power to veto any legislation ratified by the elective chamber; it should also have the power to draw district boundaries for the elective chamber and to compel a floor vote in that chamber on any legislation introduced there.

            Gat challenges my reluctance to grant the citizens’ chamber “full parliamentary powers – to set its own agenda, initiate legislation and draft its own legislative proposals.”  He suggests that citizens chosen by lottery are capable of wielding these powers responsibly—or, at least, that there is every reason to expect that they will do so as responsibly as elected legislators. He lays out several arguments in support of this claim, and I will consider each in turn.

 (1) His first argument, if I understand it correctly, is that the skills needed to wield a responsible veto are precisely the skills needed to draft and revise legislation. In the paper, I argued that citizens chosen by lot for one-year terms would generally lack the policy expertise of elected officials, and “would have virtually no experience assessing the likely consequences of different policy alternatives.” Gat argues exercising a responsible veto requires, precisely, that citizens assess the consequences of policy alternatives. 

            Here I think Gat is mistaken. In order to draft and revise legislation skillfully, a legislator must be able to choose, out of all possible policy alternatives, one that is well-suited (if not best-suited) to accomplish the relevant goal. She must therefore have an exhaustive understanding of the possible alternatives. She must also understand the “craft” of policy design—how to use incentives effectively, how to avoid exposing the law to obvious legal challenges, etc. And she must know how to ensure that the new law does not conflict with or frustrate existing law, etc. She must therefore know a great deal about the existing shape of the law.

            But veto power can be responsibly exercised without all of this expertise. Imagine a group of citizens who get together to debate healthcare reform, and who decide after considerable reading and discussion that they will be satisfied only with a reform proposal that meets the following two objectives: (1) it gives all Americans access to affordable healthcare within a reasonable time frame, and (2) it controls costs by changing incentives for doctors, so that they are not rewarded simply for prescribing more medical treatments and procedures. Though these citizens lack the policy expertise they would need to draft a competent law, they are—in my view—more than capable of exercising a principled veto. And were they empowered to do so, their veto would help define the objectives that elected legislators would have to meet in framing a reform proposal.

            Citizens who confined their deliberations strictly to the proper aims of healthcare reform would not, of course, be in good position to discriminate between the many possible policies that could be designed to achieve those aims. They would not be able to determine whether any particular proposal was in fact the best available. They would, however, even without a great deal of policy expertise, be able to recognize those proposals that are not in fact structured around the aims they endorse. They could therefore rule certain proposals out on principled grounds—and this would, in itself, be an important exercise of popular power.

In some cases, of course, citizens might articulate aims that are simply not feasible, or that would cost too much to be acceptable to them—in such cases, they would likely find reason, over time, to amend their list of aims. And of course, the more citizens knew about healthcare policy, the more discriminating their veto power would be, and the more control they could responsibly exercise over the shape of the law. My main point here is that it is possible for a citizen to lack the expertise required to write a good law, but to possess knowledge enough to exercise a responsible veto.

 (2) Gat argues, secondly, that “there is no reason to assume that a few years’ experience as an elected legislator would be necessary and sufficient training for writing legislation.”  This is certainly true. Elected officials often lack the expertise they would need to draft good laws (or even to preside intelligently over policy hearings, as late-night CSPAN viewers often discover to their chagrin). But there is good reason, I think, to expect that they will accumulate more expertise, on average, than citizens chosen by lot for relatively brief terms.

            First, elected officials serve longer terms in office, and since incumbents enjoy substantial advantages at the polls, they often stay in office for a long time. Second and relatedly, elected legislators often retain committee appointments for many years, allowing them to develop expertise in specific policy areas. Third, elected legislators must at least be competent enough to run—or at least preside over—a successful campaign, which is no small or simple feat.

It could be argued, of course, that elected legislators in fact delegate policy-writing to skilled aides, and that citizens chosen by lot would do the same. Still, the task of choosing the right aides and supervising and directing them effectively requires, in itself, no small measure of accumulated experience. Citizens recently chosen for office, without any political background, might be easy marks for career policy specialists with their own agendas, and would not be around long enough to learn the ropes and correct their mistakes.

 (3) Why, then, shouldn’t we extend citizens’ terms in office, or extend their training periods, to allow them to acquire enough expertise to handle “full parliamentary powers”? I am less certain of my conclusions here, but let me at least explain my inclination to resist this suggestion.

            First, to ask citizens to leave their lives for a certain stretch of time, move to a new place (which may be far from friends, family, children’s schools, etc.), commit themselves to public service, and put their professional lives on hold, is to ask a great deal. I can imagine asking citizens to do this for a single legislative session. I can also imagine asking employers to grant them leave for one session without forfeiting their jobs or seniority. It is harder for me to imagine asking this—either of citizens or their employers—for much longer.  

            The more demanding it is to serve in a citizens’ chamber, the less likely citizens will be to accept the office. And this would, in my view, begin to undermine lottery’s democratic credentials. As I explain in the paper, it is important that allotted chambers be “descriptively” representative of the general population. Members of the citizens’ chamber should reflect the general demographic and economic characteristics of the population at large. If they do, they can claim to represent their constituents in a meaningful sense (the sense that John Adams invoked when he wrote that legislatures “should be an exact portrait, in miniature, of the people at large”). Part of the problem with elections is that those elected tend to be elites (rich, white, male). If citizens’ chambers demanded many years of citizens’ lives, then I worry that they would attract only a minority of citizens. If anything, then, I would be more inclined to shorten, rather than lengthen, citizens’ terms in office, to ensure that service in a citizens’ chamber was a duty that all citizens, from all backgrounds, could reasonably be asked to fulfill.

(4) Gat suggests, finally, that even if members of citizens’ chambers were not themselves competent enough to draft legislation, they could assign this responsibility to panels of experts who they selected for the task. They could thus avoid relying on elected legislators to write and revise laws.

            I am not sure exactly how to understand Gat’s suggestion here. He might be suggesting that elected legislators be replaced altogether with panels of policy experts chosen by allotted chambers. If so, I think there are good democratic reasons to reject his proposal. As I argue in the paper, elections do serve an important democratic function—a function that lotteries cannot serve. Elections allow constituents to choose public officials who are formally accountable to them, in the sense of having to answer to them during election season. In this sense, elections give rise to what Philip Pettit has called “responsive representation,” which is an important element of democratic accountability. Legislators chosen by lottery for single terms are simply not accountable to constituents in the same way.  This is why I have proposed that we combine elections and lotteries rather than simply replacing elections with lotteries.

            Gat might, however, be suggesting that we preserve the bicameral structure of our legislatures but grant the allotted chamber full legislative powers, provided that they delegate the drafting and revision of laws to panels of experts. This is an interesting idea.  It would not greatly expand the powers available to members of citizens’ chambers under my proposal. It would also further complicate the legislative process. But I would certainly be interested in seeing such a proposal developed and defended in more detail.

17 Responses

  1. One concern expressed is the lack of expert knowledge inherent in a one-year-term sortition legislature. I am not going to give a full elaboration here, but it seems to me that we should not assume that we would want a one-for-one replacement of legislative chambers (replacing an elected chamber with a sortitioned chamber). I think the model of the British Columbia Citizens Assembly is instructive. Each sortitioned body could focus on one defined area, and thus become expert in that area, rather than tackle all sorts of legislation. Thus we might replace elected legislators serving two or six year terms with scores of short-term or part-time sortition bodies… One on health care, one on immigration, one on energy, etc. I would also imagine a coordinating sortition body akin to the Athenian Council of 500, that made no legislative decisions, but instead decided what issues warranted calling a sorition body into existence to work on. Such a “meta-legislature” could set the procedural rules, oversee the selection of staff, and balanced expert panels for the policy bodies. My notion is that sortitioned bodies at local and state levels would be brief, and common, with most citizens serving on many sorition bodies for a few months at a time throughout their lifetime. The pool for drawing memebers of national sortition bodies would be made up of those who had served on a state or local body previously.


  2. Unlike Yoram, I haven’t read your original paper, but from the description in this post it sounds like a development of the bicameral constitution in Harrington’s Commonwealth of Oceana (although the terminology has been inverted). In Harrington’s view the role of the Senate was to propose legislation, which then could be vetoed by the lower house (the Prerogative Tribe). The Senate was composed of the “natural aristocracy” that resulted from an elective process (in Manin’s sense), whereas the lower “popular” house was chosen by a combination of sortition and election. Although the only power of the lower house was one of veto, the Senate was constrained to only propose such measures that were not likely to be vetoed.

    I have a lot of sympathy with this proposal although I would prefer a unicameral solution with the elective element debating the proposals in front of the “Tribe” so that they could benefit from hearing the arguments. I think Fishkin’s work has shown that this model is feasible.

    I would agree with you that Yoram’s proposal to grant all powers to the sortitioned chamber is profoundly undemocratic.

    Keith Sutherland


  3. > I would agree with you that Yoram’s proposal to grant all powers to the sortitioned chamber is profoundly undemocratic.

    I see no evidence that Alex shares your conspiratorial suspicions that an allotted chamber would be somehow controlled by a shadowy cabal made of a handful of activist members. His reservations are based on less fanciful concerns regarding competence. (I’ll continue the discussion of those matters soon.)


  4. What if we imagine a system exactly like Alex Zakaras, except for one thing: The elected chamber was not elected by the general population, but only that representative sample of it which sits in the allotted chamber (I’ll call it the s-chamber for short).

    In this system, the elected representatives would perform a role similar to Yoram Gat’s s-chamber-appointed experts – they might have slightly more power, since the s-chamber could presumably throw them out if they refused to cooperate/do as they’re told under Yoram Gat’s system.

    The question to Alex Zakaras would be this: Is such a system, where the elected chamber is elected by the s-chamber, bad? If so, how does it become better by the voting public doing the election?


  5. To followup myself here, I think Alex Zakara’s system would lead to the following situation:

    The lots are drawn, and A and B didn’t come to sit in the S-chamber. So now it’s time for them to elect representatives to the other chamber.

    A thinks as follows: My interests are fairly represented and adequately defended by the s-chamber, so for the other I will choose a representative solely based on his competence, intelligence, ability to get things done and work constructively.

    B thinks as follows: My interests are fairly represented in the s-chamber – but to hell with fair! I’m right, and I know it. I will choose a representative who’s willing to push hard, engage in brinkmanship on the s-chamber veto, use any procedural trick or loophole to get an advantage, and in general be a badass in promoting my interests at all costs.

    Now, B de facto have more political power than A, don’t you agree? And isn’t this a problem?

    I think it is. It also seems to me to be one all electoral democracies struggles with, when voters are supposed to choose according to both competence and interest. You can sacrifice good government in order to win advantage for yourself – especially in bicameral systems with a veto, such as the US. There, the rewards for brinkmanship and obstructionism are especially high. B’s ideal candidate sounds a lot like a lawyer, and a lot like many US congressmen and senators (who tend to be lawyers), and that’s no coincidence.


  6. Harald –

    I do not think that it is really a matter of good-citizen-vs.-selfish-interest conflict. To me, the very idea that selecting experts is best done as some kind of a mass activity is just absurd: a small group would be able to, and would be motivated to, interview more candidates, and evaluate them much more thoroughly than the entire population could and would (or does). It would also be able to monitor the experts more closely and respond more promptly to corruption.

    Since this is the case, advocates of elections have to resort to formalisms such as “consent” and “electoral accountability”.

    (More in this vein in an upcoming response post.)


  7. Well, I hope I was clear enough: I agree that there’s little or no advantage in using the voting population to decide who the experts/traditional chamber representatives would be.

    I wouldn’t exactly call it a good-citizen-vs.-selfish-interest conflict. It’s just that all conventional elections (even our rather benign PR elections) reward brinkmanship to some degree. I seem to recall you touched upon the issue in your old post about elections as ultimatum games.

    Multiple layers of veto increases the rewards for brinkmanship, so we can expect adding a second chamber – even one with sortition – will have undesirable consequences for the the first chamber. To the degree that they’re elected in order to represent interests rather than for competence (and we can’t count on them to not be), they will just undermine the sortitioned chamber.


  8. My view is that conflicts of interest are legitimately an important factor in the selection of experts. What is illegitimate is the experts using their privileged position to promote their own interests. This was also the nature of the brinkmanship that I was discussing in the post you refer to: elections are a game of brinkmanship between experts (i.e., politicians) on one side and voters on the other side. That game does not reflect conflict within the group of voters.


  9. Thanks to all the participants here for this interesting discussion. I want to briefly consider the suggestion that Harald made, which is that legislators be selected indirectly, by a group of citizens chosen by lottery, rather than by general election.

    I would not support this kind of change because (as Yoram anticipates) it would undermine the special sort of accountability that elections create. As a citizen, I have a vote in the upcoming November elections. Because of this fact, candidates for office in my state (local, state, and federal) have a concrete incentive to persuade me to vote for them, and to do this by speaking to my needs and concerns. Moreover, November’s winners, because they will be seeking re-election within a few years, will have a concrete incentive to continue paying attention to what I and other constituents say and need. In my view, this basic form of electoral accountability is very important: it forces elected officials to care about how ordinary people in their district are doing. It also gives ordinary citizens a reason to believe that their voices, their participation, will be heard–and so it gives citizens an incentive to participate.

    In my view, then, the main function of elections is to create and sustain accountability, not to choose experts (though, as I’ve written above, there’s some reason to hope that elected officials will be competent and develop policy expertise over time). We all agree that elections have their flaws–especially in a country in which they’re saturated with private money. My paper argues that lotteries can be used to correct or offset some of the chronic problems with elections–mainly, the problem that elected officials in virtually all constituencies tend to be richer and more privileged than their constituents. But I don’t think we should replace general elections altogether. Elections are an indispensable democratic tool.


  10. Alex Zakaras,

    The people are in fact represented through the alloted assembly. They will in practice care for your needs and concerns, not out of a desire to serve you or please you, but because they are you, in a statistical sense. You and I know that. But you expect that regular people won’t understand it, and will refuse to accept a government they didn’t vote for?

    About participation, I believe – and maybe I disagree with Yoram Gat here? – that parties will continue to exist and be very influental if sortition is implemented. It’s just that their efforts would be directed at the people directly, and broadly, in the hope that someone who has heard your political sermons is drawn. People will probably feel that their opinions matter, when parties are courting them constantly for their minds instead of just their votes.

    I’ll let the brinkmanship arguments lie for now; since there’s already so much of it in the US’ bicameral system – I still think it has the potential to undermine the positive effect of an alloted assembly, but it wouldn’t be anything new.


  11. > Moreover, November’s winners, because they will be seeking re-election within a few years, will have a concrete incentive to continue paying attention to what I and other constituents say and need. In my view, this basic form of electoral accountability is very important: it forces elected officials to care about how ordinary people in their district are doing.

    This standard “rewards based” justification of elections does not stand scrutiny.

    The flaw most commonly pointed out, I think, is that the voters really know very little about how well the elected officials represent their interests. Obtaining an informed opinion about such a matter is quite difficult and the effort is hardly justified by the diffuse reward of being able to express that informed opinion as one vote among a myriad votes once every few years.

    Fatal as this flaw is, however, there is an even more fundamental flaw in the rewards-based theory of elections – a self-contradiction in the conception itself: the conception assumes that politicians seek re-election and are only willing to serve the public because of their belief that it would increase their chances of being re-elected. But the ambition for being re-elected must then be due to some personal benefit that would be derived from success in this attempt. That potential, uncertain benefit in an upcoming term must be high enough to outweigh the benefits the can surely be derived from corruption in the current term.

    But if the benefits of service are high enough so they are comparable to the benefits of corruption, then the cost to public of providing those benefits is not much lower than the cost of having corrupt politicians. This, of course, means that the rewards-based conception of elections not only fails to provide a promise of high-quality government, it in fact provides an expectation of poor government – government which is either corrupt or imposes costs that are very similar to those of corruption.


  12. > About participation, I believe – and maybe I disagree with Yoram Gat here? – that parties will continue to exist and be very influental if sortition is implemented. It’s just that their efforts would be directed at the people directly, and broadly, in the hope that someone who has heard your political sermons is drawn. People will probably feel that their opinions matter, when parties are courting them constantly for their minds instead of just their votes.

    This is certainly a possibility, but I would not take it for granted. The desirability of this development would depend on how democratic those parties are internally. As I already indicated I think that the establishment of institutionalized democratic media would be a useful, substantial (as opposed to formalistic) form of mass participation.


  13. I think the following points reply to the objections raised in this thread.

    1.) A Citizen Legislature would be, finally, unassailably a *legitimate* body … fulfilling the phrase ‘by the people’.
    “Government” was never intended to be ‘them’. ‘Interests’ were supposed to have been ‘ours’.

    2.) A recognition of service must needs be restored. Accommodations can be arranged for longer terms (for return to jobs, for example).
    The spread of leadership experience throughout the citizenry would, over time, strengthen all of society and lead to less need for governmental supports.

    3.) Three year terms, staggered entries, for sortitioned citizen legislators. At least three months of pre-service training.

    4.) Requirements to be placed in the lottery pool: a.) registering (service is voluntary); b.) passing a basic civics test (no more difficult than getting a driver’s license in the USA).

    5.) Recompense at 150% of the national median.

    6.) Additional aspect of point #1: More fundamentally important than concerns about competence, the weight of money or the influence of media is the fact that electioneering produces a skewed psychological profile of legislators. Elections select *only* for those personality types willing to enter the fray. The loss of the quietly thoughtful — as well as, yes, the occasional guy gone bonkers — means a greatly diminished conceptual breadth and, I would argue, depth. Not to mention the automatic selection for competitive aggressiveness through elections.


  14. David –

    I generally agree with your points, with the exception of the matter of qualification. I think that if the qualification barrier is set so that most people would have no problem passing it then it is unnecessary, and if it is set so that it could actually exclude many people then it is dangerous. In any case, the act of exclusion, and even the potential for exclusion, is counter-democratic.

    Regarding the other points: I think the exact parameters – length of service, length of training, level and type of rewards – are best left for future determination through theorizing and experimentation when circumstances allow for better understanding, but general considerations show that:

    a. A period of service and training measured in years is necessary for achieving effective government

    b. The rewards for service need to be high enough to indicate the importance society places on the function, but not so high as to turn the group of delegates into an elite circle.

    c. Putting the allotted body at a power disadvantage compared to an elitist body, such as an elected chamber, would have the potential of nullifying any democratizing effects of the use of sortition.


  15. […] our recent exchange (1, 2), Alex Zakaras and I debated whether an allotted chamber should be given the full legislative […]


  16. […] take on this proposal and exchange with Zakaras are here: The elected legislator’s burden, Lottery and Legislative Powers: A Reply to Yoram Gat, and Limiting the allotted chamber’s powers – a foundational […]


  17. […] Alex. 2010b. Lottery and Legislative Powers: A Reply to Yoram Gat. En Equality by Lot: […]


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: