Alda is an Icelandic organization promoting reform of Iceland’s government system, including the use of sortition in various ways. A member, Kristinn Már Ársælsson, has an article on the openDemocracy site:

After the crash that destroyed Iceland’s economy, Icelanders started to take an interest in new forms of political and economic governance.


In some respect, Icelanders have made their voices and interests heard in a way people of other countries have not. The protests after the crash got us a new government, the head of the central bank and the financial inspection agency were axed and a process to make a new constitution with the active involvement of the people was initiated.


These are important achievements. Things that other countries could learn from. But frankly, most of these developments were also controversial in Iceland and overall, they could have been executed more efficiently. For example: the idea that the general public should be actively involved in creating a new constitution is indubitably right. But this could have been better carried out. The selection process didn’t have the legitimacy it needed and random selection should have been used as well. The time given to the process was too short. There was not enough debate all over the country and in the media. Of course, in comparison with the constitution being rewritten by a small group of politicians in closed session, as usually happens, the new process was great. But it could have been better.

Alda’s website has a specific proposal for a system that involves randomly selecting part of the legislature as well as allotted citizens’ assemblies:

Election system

In Althingi 42 nationally elected MPs have their seats, in addition to 21 randomly selected MPs from every icelandic citizen of the age 18 to 70.


Alda proposes that Althingi will operate in one house (unicameral). It is also possible to have Althingi to operate in two divisions. In the lower one, only nationally elected MP´s would have their seats, but in the upper only randomly selected MP´s. Research has shown that there is difference in what interests representatives protect, depending on how they were elected. Those chosen randomly generally see themselves representing the general public while those elected from party-lists see themselves as representatives of their political party’s voters. For this reason, it might be a good idea to have MP´s of the upper house elected randomly which will protect everyone’s interests. One should not draw the conclusion here that any one type of representative, political-party, individual or randomly selected, is better than another, but rather that every system has it’s own benefits and drawbacks.



Law and regulation should specify what qualifications one needs to fulfill, in order to be appointed judge. Randomly selected group of supreme court judges decide who of the applicants are qualified, in accordance of those rules. Then a random applicant from those who are deemed qualified is chosen.


Citizens’ assemblies

The legislative power is permitted to hand it’s power partially to citizen’s assemblies, in which MPs do not have any power. Representatives in these assemblies shall be selected in elections or randomly selected. One third of MP´s can demand citizen’s assembly to be held on important issues, or 8 percent of the electorate.


No provisions are in the Icelandic constitution that enable the government to delegate it’s power to democratically chosen citizen’s assemblies. It is imperative that such provisions are part of the constitution, so that citizens can be part of the democratic decision making process (such as randomly selected assemblies). Alda proposes that citizen assemblies can be summoned when controversial and/or important issues need to be decided upon, issues that need more involvement from the public than Althingi is able to generate. Part of the process is meetings and discussion panels around the country, with discussions and information provided by the media. Representatives should be chosen so that they represent different groups of the population and long time is given to finish the process. It is also proposed that assemblies would be summoned in cases involving only local communities.

20 Responses

  1. >Research has shown that there is difference in what interests representatives protect, depending on how they were elected. Those chosen randomly generally see themselves representing the general public while those elected from party-lists see themselves as representatives of their political party’s voters. For this reason, it might be a good idea to have MP´s of the upper house elected randomly which will protect everyone’s interests. One should not draw the conclusion here that any one type of representative, political-party, individual or randomly selected, is better than another, but rather that every system has it’s own benefits and drawbacks.

    It’s hard to understand exactly why we would want to protect the interests of “political party’s voters” as opposed to the interests of the “general public”. Although this might be just a case of pragmatism, another way of making sense of it would be that the lower house (where the majority party/ies dominate) would have the power of initiative (manifesto commitments) and that the upper house would have the power of veto (after deliberation). Note that this is a stronger mandate than that of a revising house (as in Anthony Barnett’s proposal). A division of labour on these lines would respect Madison’s distinction between interests and judgment and ensure that the latter was not corrupted by factional interests. Other than the right of initiative it’s hard to understand the need for a lower house representing factional interests.


  2. Hi,

    Good point about “why to protect interests of political party voters.” Alda thinks that the general idea of political parties, as a field/venue for people to come together as equals and find solutions to social problems, generate ideas how we can make our societies better equipped to increase well-being and happiness, is a good one. And actually we made a proposal on how to organize a political party, with different mechanisms to diffuse power and with a flat structure. We also incorporated sortition in selecting representatives within the party. But the organization is basically a open democratic process (e.g. like participatory budgeting). So Alda thinks that political parties need to be reformed (and other aspects of our democratic systems as well, e.g. interest group influence) so that their interests reflect in a better way the interests of the general public and party members, bottom up.

    But Alda´s main proposal was that representatives were in a single house.


  3. Hi Kristinn,

    It seems to me that a single chamber with both professional politicians and allotted amateurs will likely result in the professionals using their expertise and experience (and superior numbers in your proposal) to manipulate and overpower the amateurs.

    In any system there is an advantage to professionals over amateurs. Institutional arrangements should be set up so as to limit and diffuse this advantage, not to amplify it. If you wish to retain a role for elected officials (at least as an interim arrangement) then having separate chambers, with one of them completely controlled by the the allotted delegates, seems like the better arrangement.

    I am also wondering why the section of your proposal regarding the courts does not advocate doing away with judges and putting power judicial firmly in the hands of randomly selected juries. That seems to me like the natural democratic procedure.


  4. Thanks for your response Yoram.

    I´m not convinced that a parliament composed of three types of representatives, where only 1/3 come from political parties, would overwhelm the randomly selected representatives. That remains to be seen. The dynamics of discussions and decision making would probably be different. Now this might be different in a smaller system/community like Iceland but I would like to see how it pans out (but definitely with some modifications to the system, e.g. transparency, associational democracy etc.)

    The role of political party representatives (which should come from flat structure democratic parties) is in a way different from the randomly selected ones in setting the agenda, promoting ideology etc. Those representatives elected individually are somewhere in between, with expertise (or high level of trust) in certain areas. Perhaps in the system proposed they would align more with the party representatives but then again they might align with the randomly selected ones.

    One new political party in Iceland is based upon our proposals for a democratic political party and they use sortition for their executive board and selection of representatives. They are in some boards in a minority against elected representatives but from the experience so far (about a year) I hear that they hold true to their part as representatives of the members of the political party. No-nonsense and put their foot down when something is not right.

    Now I´m all for more sortition! These are only explanations of how these ideas were formed and the arguments behind them.

    Regarding the judiciary system I would have to think a bit more about whether to dispense completely with judges. That was not an option that came up during our discussions at the time.


  5. Alda’s trinitarian proposal is a marked improvement on the bicameral model adopted by most elective democracies. But if (1) party representatives have a distinctive role in setting the agenda, and (2) directly-elected representatives are more akin to trusted/expert advocates, should there not also be a unique role for (3) randomly-selected members — ie determining the outcome of the deliberative exchange? That would better ensure that the judgment function was not corrupted by ideology or interests. As Madison put it in Federalist 10:

    “No man is allowed to be a judge in his own cause, because his interest would certainly bias his judgment, and, not improbably, corrupt his integrity. With equal, nay with greater reason, a body of men are unfit to be both judges and parties at the same time.”

    Unfortunately Madison failed to put in place the additional separation of powers necessary to distinguish interests and judgment, so Alda would do well to ponder the need to keep the powers of initiation, advocacy and judgment separate, via the erection of “parchment barriers” between the three types of representatives that you (rightly in my view) seek to include in the democratic process. Leaving it all down to sortition would deny the important role of ideology/interests in introducing initiatives and the equally important role of “elect” advocates. But the final judgment should be in the hands of the allotted sample of the demos — who would be truly disinterested (at least at the aggregate level of voting).


  6. Keith: Well, maybe. I would like to create a system were the party and individually elected representatives would not need to be checked by the sortition representatives but rather that the dynamism of the system worked towards the greater good. And there are many different aspects of the system that need to be changed to reach that goal. But perhaps that´s impossible but I would like to try first. I´m not quite ready to give up on the idea that all types of representatives were working together (more or less, constructive criticism and all that) towards a greater good rather than giving up on that an using sortition to be the final judgment, I genuinely think that their input is invaluable at the idea/proposal stage. Cognitive diversity trumps individual ability (Landemore etc.).


  7. Yes I can see why you would want to try to reach consensus, but there would be a considerable asymmetry in the perceived legitimacy and authority of the different categories of representatives. Both party- and individually-elected representatives would be able to claim a direct mandate and might easily intimidate the randomly-selected element, who would naturally defer to the elite members, so you might need to consider granting them some sort of distinctive role. Helene’s argument is a little unclear — in her forthcoming book (I read the m/s for her) she follows Estlund’s and Page’s case for a diversity of cognitive input, but claims (unconvincingly) that this is already the benefit that democracy brings. It strikes me that the evidence for diversity is stronger for the wisdom of crowds (not something she is interested in), but this is an aggregate function that depends on the independent judgment of a large number of people. It’s hard to see how independence of judgment would be maintained in your project without the erection of parchment barriers (consensus being anathema to any Condorcet process). I think you’ve accurately identified the distinctive characteristic of each of the three forms of representation — each has its own unique value — but you do need to think how they will interact as it’s unlikely to be a straightforward summing of three individual virtues, or even an emergent property that is more than the sum of the parts. It’s just as likely that the internal dynamics will lead to the corruption of the distinctive virtue of each element. In the original judicial example we keep the prosecution, defence advocacy and judgment strictly separate, so why not for the political analogy?


  8. Ok. You might be right about the difference in legitimacy but I´m not convinced. Actually I would guess that it would be quite the opposite, at least in Iceland. The political parties (parliament) have a 10% trust rating from the general public. The general sentiment is: away with them all. So I would guess that sortition and individually elected representatives would have more legitimacy than political party representatives in the current state of affairs.

    And, again, I´m also not convinced that sortition representatives would defer to other representatives. That is simply an empirical question and my guess in Iceland today is that they wouldn´t but that could be different elsewhere. But if they defer, that calls, in my view, for adjustments to the deliberation process within parliament. That would be my first option before erecting barriers or removing the sortition representatives from the agenda and proposal setting and to a judgment position.

    I also have Helene´s manuscript, but I haven´t had the time to finish it. Small world :) I think that the theoretical framework is actually quite sound, in that cognitive diversity, even in small numbers is likely to increase the probabilities for “better” solutions. And I also think that the theoretical prediction that sortition representatives will act differently from other types of representatives is worth trying (but on all these points we need further empirical data). AND that those properties will create a different dynamic (which could be greater than the sum of its parts, but not necessarily a proportional sum of it´s parts) within the deliberation process (but there might be mitigating structural effects that might enhance or diminish those effects, and that is something that we need to pay close attention to.) And on those grounds I would like to try to include the sortition representatives in the agenda and proposal making setting (they bring different viewpoints and make checks at that stage, which I believe to be important).

    And Alda argues that open democratic processes (e.g. participatory budgeting, randomly selected citizen assemblies) should play a big part in agenda setting, creating proposals and decision making – effectively tapping into the cognitive diversity and empowering aspects of those processes. And that´s the general idea, to create a robust and resilient system with general public participation and sortition. And I would argue that we need to focus on a great deal of different structural aspects (there might even be differences, at least while changing a system, between countries and cultures) when modeling such a system.

    Alda´s general proposal is sound in my opinion but can and will be improved in the coming months and years.

    I hope your following my drift and that I´m following yours, these comment threads can be a bit constraining (and also I don´t have much time…).


  9. I think we understand each other very well and I agree that these are empirical issues. Helene and I have a long-standing disagreement over whether to focus on the epistemic value of the cognitive diversity of individual allotted members or on the descriptive representativity of a randomly-selected assembly. Ideally one would like to have both, the problem being that it’s hard to see how a statistically-representative mandate can extend to anything other than aggregate functions (the clue being the word “statistical”), thus limiting the epistemic benefit to the wisdom of crowds. It’s interesting that Surowiecki views democratic politics as the one arena where the principle does not apply.

    Disaffection with political parties is widespread (my 2004 book on sortition was called The Party’s Over) and I agree that individuals may initially be elected on the basis of trustee characteristics. Unfortunately this is how elective representation started out (at least in theory), but was corrupted by partisan pressures within a single generation, so it’s realistic to conclude that it’s a built-in consequence of the electoral process. No-one ever intended there to be political parties, indeed their existence was not formally recognised in the British constitution until very recently. That’s why I prefer to focus on the inherent virtue of each selection method and to seek to design structures that preserve each virtue. To my mind anything else is just wishful thinking!


  10. Do you have a paper, book or text somewhere which goes into more detail on the issues you present in the first paragraph (epistemic value vs. descriptive representativity)? Is it in your 2004 book (I just finished ordering from Amazon recently, would have included it)? I would like read more on that…

    About the structures of the political field: I´m not a political scientist, an perhaps this has been tested, whether the electoral process (even though varied) is the sole cause. On face value I would rather doubt it, theoretically. It might be a interaction between different factors, and I would definitely hazard the guess that the structure of our societies (and old social structures, feudalism) and the structures of the economy in modern capitalist societies (at least) play a significant role in the outcome that party-representative democracy went sour. One of the reasons for the increasing importance of the statement “the party´s over” is changes to our societal structure (demographics, transparency, education, interrelations, affluence, organic solidarity etc.). And the call for increased direct public participation (which actually is also inherent in right wing pro-market arguments since the 1980’s) is a result of those changes. We will most likely have to change numerous things within the economic field and in the formal and informal interplay between the economic and political field for different changes made in the political field towards a more participatory democracy take their full effect, BUT I´m not convinced that “anything else is just wishful thinking.” And these questions (both theoretically and empirically) are exactly those that I would, if I ever return to the academic field, like to research.


  11. Manin claims that it’s inherent in the electoral process but, like you, my background is sociology, so I’m very conscious of the other structural factors. Let me ponder on this before replying. PS my 2004 book has been superseded by the 2008 incarnation, A People’s Parliament, in which I do argue for a limited role for the majority political party/ies (the right to introduce manifesto commitments as parliamentary bills for deliberative judgment in an allotted assembly).


  12. Kristinn

    Just uploaded the relevant paper:

    Although Bernard Manin does maintain that the nature of the electoral process inherently led to what he calls the metamorphoses of representative government, I agree with you that it’s a complex mixture of structural factors. If Madison had anticipated the likely corrupting effect of money and communications technology on his “republican” model of government he might well have had second thoughts. Cynics like to argue that the First Amendment was a deliberate ploy to ensure that elections could be bought by the rich and powerful, but this is really just the benefit of hindsight. As for the need to “change numerous things within the economic field” I think this might well be putting the cart before the horse. Let’s get the representative institutions right first and then see if there is call for broader structural change. The attractive thing about your proposal, from a pragmatic perspective, is that it allows due place for our current electoral arrangements and, as such, is less likely to require the turkeys to vote for Christmas. It really could work, but I think you do need to ponder further as to whether each element of your trinity should have uniquely defined powers (that reflect its own virtues).


  13. This is a very good discussion. I did want to mention that there is also a Facebook page run by the same people (informally) responsible for this blog. Our group is called the Kleroterians, and I would encourage anyone on FB to join it.

    I did have one substantive concern myself about the original article. While I do think that participatory democracy and sortition go well together, at least under the right circumstances, I’m not convinced that they represent the same thing. In a large country like the United States–or even a small country like Iceland–most people will never serve in a randomly-selected legislative or constitutional assembly. It might bring in a different group of people (i.e., not just politicians), but that’s rather different from bringing in MORE people, which is what I normally associate with participatory democracy (like the participatory budgeting stuff in Porto Alegre.


  14. Keith: Thank you for the paper. Look forward to finding time to read it :) I agree that it´s important to compartmentalize proposals and ideas so that they can´t be dismissed on objections to “unrelated” issues. But I also think that we need to create/assemble the architecture of institutional design for both the political and economic field – because of the interaction of different structures (and even cultural differences) can and do matter in how these models work in practice. And while we need to use as much we can the empirical evidence available in the end we must make experiments (and remember to analyze and research those to improve further). And Alda is definitely not at the end of the road, but rather at the other end.

    Peter: There are different ways to classify democratic processes/models. Erik Olin Wright has one which I think is simple and revealing. According to that model sortition can be classified either as a deeper version of representative democracy or as part of deeper part of direct democracy (which is basically participatory democracy). But I definitely get your point that there are important differences between sortition and participatory processes. And that it is important to figure out how they work together within a coherent democratic system. Alda was only a few months old when we created the proposals for the Constitutional Council and if we were to do it again, our proposal regarding the participatory aspect would be different. The proposal quoted above is a kind of a mix between random citizen assemblies and participatory budgeting (as a process). Since then Alda has simply differentiated between them and either/or suggested that in e.g. the legal frame work for municipalities there be specific rights granted to the public to hold either a random citizen assembly or a participatory process (in the mold of participatory budgeting). Actually, the interior minister of Iceland is considering adding such rights.


  15. >I also think that we need to create/assemble the architecture of institutional design for both the political and economic field.

    OK, but you run the danger of limiting participation to those who share your views on the economy. Perhaps in a country like Iceland there is broad agreement on this; not so in (say) the US where there appears to be sharp 50/50 split (judging by the popular vote in the presidential election). One of the advantages of sortition is that it is of equal appeal to those for whom social justice is the top priority and those of us who just want to make sure the trains run on time.


  16. As a practical issue concerning the viability of proposals on the political field, yes. Alda is not a political party and not in the “art of what is possible” but in the “art of what is best.” And the political field is in dire need of ideas to change the system (TINA and all that). And therefore I think what “we” (at least part of us that want to change the social system) need to put together a coherent institutional model for the political and the economic field (not least because of the interaction and interdependence of institutions and their functions between the fields).


  17. Where does Erik Olin Wright make that distinction? Sounds useful. Is it in the Deepening Democracy book? Haven’t read that one yet.


  18. Peter: It´s in Envisioning Real Utopias. The best book I have read on “basics” in emancipatory institutional design.


  19. […] Már Ársælsson, of the Icelandic government reform group Alda, advocates sortition in the following […]


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: