Skip to content
Jan 22 / Barry Clemson

An Idealized Design for the Legislative Branch of Government

Terry Bouricius and David Schecter. Volume 2. Issue 1. January 22, 2013

Abstract. This paper presents an idealized design for a legislative system. The concept of idealized design is explained. The paper critiques two critical (and often taken for granted) features of the legislative branches of most contemporary democratic governments: legislators are chosen by election, and the same bodies perform all legislative and meta-legislative functions, for all laws. Seven problems with these two features are described. A new model of lawmaking is proposed, based on three concepts from ancient Athenian democracy — random selection, dividing legislative functions among multiple bodies, and the use of temporary bodies (like contemporary juries) for final decision making. The benefits of the model are laid out, and likely objections are addressed.



Russell Ackoff developed a profoundly useful concept for systems thinking – Idealized Design (Ackoff, 1974, Ackoff, Magdison & Addison, 2006). An Idealized Design is developed based on what the designers really want, rather than working incrementally from current reality. There are two constraints imposed on idealized designs,and one important requirement. First, the design must be technologically feasible—no science fiction. This constraint does not preclude innovation, but it does restrict innovations to what we currently know we can develop. An idealized design might not be implementable for economic, social, or political reasons, but it must be technologically feasible to operate if it were implemented. The second constraint is that the design, if implemented, must be capable of surviving in the current environment. This does not mean that the design must be capable of being implemented now. Finally, there is the important requirement that the process must be capable of being improved over time. It should be ready, willing, and able to change itself or be changed (Ackoff, 1974).

In this paper, we propose an idealized design for the legislative function of government. The design could theoretically be adapted for use in any country with a democratic form of government, and could be implemented in ways that range in scope from a small incremental change to a fundamental reform, at any level of government.

The design is based on three concepts adapted from ancient Athenian democracy, as described in Mogens Herman Hansen’s authoritative history, The Athenian Democracy in the Age of Demosthenes (Hansen, 1999),and Paul Woodruff’s First Democracy: the Challenge of an Ancient Idea (Woodruff, 2005).  First, instead of concentrating lawmaking responsibilities in one or two all-purpose legislative houses, these responsibilities are divided among six types of more specialized bodies, with limited powers. Second, the members of these bodies aren’t elected – they are selected by lot. Third, the bodies that make the final decisions are temporary, like juries.

We are not alone in this project. In the last few decades, there has been a revival of interest worldwide in the practices of ancient Athenian democracy – especially, selecting decision makers by lot (also known as sortition). There have been many successful experiments with randomly selected groups deliberating about public policy (Crosby, Kelly, and Schaeffer, 1986, Fishkin, 2009). The Citizens’ Assembly in British Columbia, a Budget Jury established by a suburb of Sydney, Australia, and the Citizens’ Initiative Review Panel in Oregon are just a few examples of randomly selected representative bodies being incorporated into the legislative process.

There have also been a number of proposals for incorporating randomly selected citizens into the legislative process, including the “Citizen Legislature” (Callenbach and Phillips, 1985), “Popular Branch” (Leib, 2004) and “People’s House” (O’Leary, 2006). Our ideas have been greatly influenced by these authors. However, their proposals each include only one randomly selected body, whereas we believe that a multi-body design is much better able to reduce opportunities for corruption, provide checks and balances, and increase opportunities for meaningful participation (see “Three Strategies for a Better Model” below).

In this paper, we describe seven problems with elected, all-purpose legislatures (by “all-purpose” we mean that the same one or two legislative chambers are responsible for virtually all parts of the lawmaking process). One of the authors (Terry Bouricius) developed the new model that is presented. Over a 30-year career as an elected legislator, Bouricius closely observed the problems that will be described. His years of experience further validates the new model that is presented; one which we believe could solve these problems. This is followed by a discussion of how the model could be implemented, as well as responses to possible objections.

Seven problems with elected, all-purpose legislatures

1.     Descriptive representation – Elected legislatures usually include a combination of perspectives that is very different from the perspectives of the people they are supposed to represent. They are wealthier, better educated, with fewer women, fewer working class people, fewer people of color, and so on. They are not “representative” of the people who elected them in terms of their demographics and points of view – what the political scientist Hannah Pitkin calls “descriptive representation” (Pitkin, 1972).

2.     Diversity – Elected legislatures are not only different in their perspectives compared to the people they represent – they are also less diverse. There is an overriding homogeneity of perspectives, even among elected legislators from opposite ends of the political spectrum (Carnes, 2012).

3.     Opportunity to participate — With elected, all-purpose legislatures, there are few opportunities for participation in lawmaking, and most citizens (including people who might make outstanding citizen-lawmakers) have virtually no chance of becoming legislators (Crenson & Ginsberg, 2002).

4.     Not enough time and attention to do their jobs well – Elected legislators typically spend a large portion of their time on fundraising, campaigning, constituent relations, and cultivating the media. Former U.S. Senator Ernest Hollings (2006) estimated that nearly one third of a Senator’s time is spent fund-raising. In addition, representatives in all-purpose legislatures have too many tasks to do, about too many bills, to pay significant attention to more than a small portion of them. That’s one reason why legislators often vote on bills they haven’t even read.

5.     Accountability – Holding elected legislators accountable is crucial because elected legislators generally have different interests than the people they represent. Having one opportunity every 2-6 years to “throw the bums out,” when legislators may vote on hundreds of bills each year, is an accountability mechanism that cannot possibly work.

6.     Corruption – Elected legislators face heavy pressure to vote in ways that will satisfy campaign donors, political party leaders, interest groups, the media, and small groups of swing voters. In addition, concentrating all the legislative power in just one or two bodies is an invitation to corruption and abuse of power.

7.     Deliberation – Elected legislators face heavy pressure to speak in ways that “score points” for themselves and their allies, and that make their opponents look bad. When legislators speak to score points rather than to solve problems, it can become a serious obstacle to meaningful deliberation.

The first five problems can be traced to a more fundamental problem, one that is familiar to readers of the pioneering cyberneticians, Ross Ashby and Stafford Beer (Ashby, 1956; Beer, 1979). The legislatures that we have today do not have “requisite variety” to do their jobs effectively.

In cybernetic terms, the effectiveness of any regulatory system depends on the extent to which it has sufficient complexity (“requisite variety” in cybernetic terms) in comparison with the complexity of whatever it is supposed to regulate. In our current legislative systems, a very small number of people (the legislators), who tend to be a very homogenous group compared to the people as a whole, have responsibility for nearly alllegislative tasks, for allissues. Even with the help of staff members (and let us not forget, lobbyists), it would generally be impossible for legislators to pay adequate attention to all their tasks about all the issues – even if they didn’t have to spend a large portion of their time on fundraising, campaigning, and media relations. In cybernetic terms, the legislature does not have “requisite variety” in comparison to its issues and tasks.

There is also a second “dimension” in which our legislatures don’t have requisite variety – in relation to the perspectives of the people. In a democracy, a legislature is responsible for representing the will of the people. The elected legislatures we have today – even in countries with proportional representation – are far less diverse in their range of perspectives than the people they are supposed to represent.

Three strategies for a better model

Terry Bouricius has developed a model (described below) that we believe can solve these problems. The model is based on three strategies adapted from the first well-documented democracy – ancient Athens (Ober, 2008; Woodruff, 2005).

First, instead of one or two all-purpose legislative houses, the functions and powers of lawmaking are divided among multiple bodies. This makes it much more difficult for wealthy and powerful factions to corrupt the whole legislative process — even if they manage to gain control over one or two of the bodies, there are others than can provide checks and balances. Having multiple bodies also greatly increases opportunities for citizen participation in lawmaking, and makes it easier for each body to do its job well.

Second, the members of these bodies aren’t elected. Instead, they are selected by lot, in order to be more representative of the people, less beholden to donors and other powerful players in the political game, and more diverse.

Third, while members of the bodies that prepare and review legislation serve multi-year terms, the bodies that make the final decisions about laws are temporary, like our juries. Each one meets for a very short time about only one bill, hears pro and con presentations, votes, and then disbands. This makes it much easier for ordinary people to participate.

Taken together, these three strategies give the legislature (actually, a larger set of bodies that carry out the legislative process) much more complexity (“variety”) to match the complexity of issues and the complexity of perspectives among the people, while reducing the complexity of the tasks required of each body and each individual.

Actors and roles: six types of randomly selected bodies, and support staff

Agenda Council: The Agenda Council creates and updates a list of issue areas (environment, health, public safety, etc.).It decides which topics in each area need new laws written, or which existing laws need changing – but doesn’t draft bills or vote on them. The Agenda Council proposes objectives and criteria for bills, but Review Panels can change these. This body also provides overall legislative coordination. The Agenda Council’s members are selected by lot among interested people. They serve 3-year terms (with one third of members replaced each year), but they are not allowed to serve consecutive terms. Agenda Council members are well compensated.

Interest Panels: Each Interest Panel produces a draft bill about one issue. Their draft bills are reviewed by Review Panels, and voted on by Policy Juries. They can also propose changes to objectives and criteria, but Review Panels make the final decisions about these. Interest Panels are composed of volunteers, 12 to 20 members each, formed into panels by lot or self organized, serving long enough to draft a bill. Unlike members of the other bodies, Interest Panel members are not compensated, because there will probably be large numbers of members who may serve for long terms, and the cost would be prohibitive.

Review Panels: Each Review Panel is responsible for only one issue area, so that panel members can concentrate on one subject and become knowledgeable about it. Within this area, they review draft bills from Interest Panels, amend and combine bills, and produce final proposed bills that will go to a vote – but they don’t initiate bills or cast the final votes. Review Panel members are selected by lot from a pool of volunteers. They don’t choose what issue area they will be assigned to, in order to avoid possible domination of the panels by special interests. Review Panel members serve rotating terms of a year or more. They are well compensated.

Policy Juries: Each Policy Jury hears pro and con presentations about one bill, and makes the final decision, in a week or less – but they don’t set agendas or write bills. Their members are selected by lot to be as statistically representative of the whole people as possible (the short term of service makes this much easier). Policy jury members are paid a small stipend.

Rules Council: The Rules Council handles work that is not part of making laws, but that is about the lawmaking process – for example, deciding the procedural and ethics rules to be used by the other bodies. Its members are selected by lot from a pool of volunteers, and they serve 3-year terms (with one third of members replaced each year), but they are not allowed to serve consecutive terms. They are well compensated.

Oversight Council: This body handles oversight of the lawmaking process, including the work of support staff. It also evaluates implementation of laws by the Executive Branch. The Council’s members are selected by lot from a pool of volunteers, and they serve 3-year terms (with one third of members replaced each year), but they are not allowed to serve consecutive terms. Oversight Council members are well compensated.

Support staff: Like today’s legislatures, these new legislative bodies would require the assistance of a permanent support staff for tasks such as researching issues, setting up testimony from experts and members of affected groups, documentation of deliberations and actions, and providing technical support for electronic communications. Unlike today’s legislative staffs, these staff members would be “shared” among entire bodies instead of being dedicated to individual legislators or political party caucuses, and they would not work on political campaigns, fundraising, or media relations.

The legislative process

Choosing issues:The Agenda Council gets ideas for priority issues from multiple sources (staff, Review Panels, Policy Juries, citizens, interest groups), and then decides which issues need laws written or changed. If citizens are unhappy with the Agenda Council’s decisions, they can initiate legislation themselves by petition.

Developing bills: Once a topic is selected, there is a call for volunteers for Interest Panels. Each Interest Panel attempts to produce one draft bill.If there are many Interest Panels working on the same topic, there could be a super-majority requirement to advance a bill to the reviewing stage. Since the members know they must draft a bill that can pass a Review Panel and a Policy Jury, they have an incentive to draft bills with the potential for broad appeal. This process might also incorporate Internet-based “crowd-sourcing” methods, which allow large number of people to collaborate on proposals without having to meet face-to-face, instead of relying entirely on small groups who have the discretionary time to meet together.

Reviewing bills: For each issue area, a Review Panel reviews draft bills from Interest Panels, takes expert testimony, holds hearings, amends and combines bills, and decides by majority vote on final proposed bills that will go to the voting stage.

Voting: Each bill is voted on by a one-time Policy Jury. Policy Jury members meet for up to a week, listen to pro and con arguments, and vote by secret ballot, without internal debate, in order to avoid “group-think” or domination by those with high social status. The secret ballot minimizes the social pressure of voting, and the risk of vote buying.

The whole process would look like this:

leg process

Figure 1. Lawmaking process. This figure visually depicts the idealized design discussed in this article.

How the model solves the problems

1.     Descriptive representation – elected legislatures usually include a combination of perspectives that’s very different from the perspectives of the people they represent. In contrast, a multi-body legislature selected by lot would be, in the words of John Adams, a “portrait of the people in miniature.” It would have a mix of perspectives that is much less “elite,” and much closer to the perspectives of the people. Policy Juries, with their very short terms of service for busy people, will be even more descriptively representative.

2.     Diversity – elected legislatures are usually much less diverse in their perspectives than the people they represent. The proposed mix of bodies selected by lot would include much more diversity of perspectives than an elected legislature. Recent research indicates that diversity of perspectives and “cognitive style” is a critical ingredient in good decision-making – in fact, for non-technical decisions, diverse groups of ordinary people generally outperform non-diverse groups of experts (Surowiecki, 2004).

3.     Opportunity to participate – With elected, all-purpose legislatures, there are relatively few opportunities available for participation in lawmaking, and most citizens (including people who might make outstanding citizen lawmakers) have virtually no chance of becoming legislators. In this model, all citizens have a more or less equal opportunity to serve. Because there would be more bodies with many more members, there would be many more opportunities for citizens to participate than there are with all-purpose legislatures. Citizens who have “no time for politics” could still participate in a policy jury for a few days, and play a critical role in deciding whether a bill becomes law.

4.     More time and attention to spend doing their jobs – elected legislators typically spend a large portion of their time on fundraising, campaigning, constituent relations, and cultivating the media. Citizen-legislators chosen by lot wouldn’t have to do these things – instead, they could focus their time and attention on the work of making laws. In addition, each member of any of the bodies in this model would have a much smaller area of responsibility than a member of an all-purpose legislature, so it would be much easier for them to do their job well. Each Interest Panel could focus on one bill, each Review Panel could focus on the review of bills for one issue, and each Policy Jury could focus on the final vote about one bill.

5.     Accountability – holding elected legislators accountable is crucial because elected legislators generally have different interests than the people they represent. A statistically representative sample of the population would inevitably share the interests of the population from which they were drawn. Each member of a body selected by lot would simply be accountable to his or her own conscience, as if the entire population were directly making the same decision.

6.     Corruption – elected legislators face heavy pressure to vote in ways that will please their donors, political party leaders, interest groups, the media, and small groups of swing voters. Randomly selected citizen-legislators would face none of these pressures. Concentrating all the legislative power in just one or two bodies is an invitation to corruption and abuse of power. Dividing powers into multiple bodies allows checks and balances that make corruption more difficult. Having each Policy Jury vote on a different bill eliminates the problem of “vote trading,” and having them vote by secret ballot reduces the possibility of special interests trying to buy their votes.

7.     Deliberation – elected legislators face heavy pressure to speak in ways that “score points” for themselves and their allies, and that make their opponents look bad. Citizen lawmakers chosen by lot would be more able to speak and vote based on what they really think about the issues. They would also be freer to change their minds based on new information, unlike politicians elected with pre-committed policy positions. There would still be lots of conflict, and occasional egotism and petty arguments, but much less of the “point scoring game” that characterizes much of what passes for public deliberation in our legislatures today.

How the model could adapt to change

How could this model be capable of rapidly adapting to changing conditions, and to the changing values of future generations of citizens? This is a critically important question, and we can’t fully answer it yet. However, there are two features of the model that are designed to make it more adaptable to change than most current legislative systems are. First, the Agenda Council is responsible for reviewing and updating the framework of issue areas for potential legislation. As new issues emerge, the Agenda Council has the authority and responsibility to insert them into the framework, which in turn will allow for Interest Panels to propose laws within these new areas. Second, the Rules Council is responsible for reviewing the overall process and changing it as needed. Note that the Rules Council is randomly selected, serves limited terms, and does not itself propose legislation, all of which helps it to focus on adaptation of the process while avoiding conflicts of interest.

How the model could be implemented

This model could be implemented in a variety of ways, ranging from small incremental changes to fundamental basic reforms. For example, different variations of the model could be used to:

1. Make one law (like the British Columbia Citizens’ Assembly)

2. Make all laws within one issue area (for example, an area that is so controversial that elected officials would actually prefer to delegate it to citizens, or where legislators have a conflict of interest, as in term limits, legislative salaries, or election law)

3. Enhance the deliberative quality of an initiative and referendum process

4. Replace one elected house of a bicameral legislature

5. Carry out the entire legislative process in place of an elected legislature

Possible objections, and our answers

“Good governance requires special ability, expertise, and experience. Ordinary people selected by lot couldn’t possibly be competent to make laws.”

The job of any citizen-lawmaker in this design would be much simpler than the job of an elected legislator today. For true subject matter expertise, they would rely on staff and consultations with experts, just as elected legislators do. Randomly selected law-makers would not need to devote vast amounts of time to mastering public relations, fund-raising, and parliamentary maneuvering skills. And a diverse group of citizens working together will likely have a good set of strengths that complement each other. As in most groups, for the group to be competent as a whole, it is not necessary for every member of the group to be competent individually. As a final note, if our legislators are that much more competent than the rest of us, why do they consistently get such low approval ratings from the public?

“These randomly selected people won’t be able to stay in their legislative jobs long enough to develop expertise.”

Each body has a much narrower range of topics to handle than an all-purpose legislature would. The terms for each body are designed to allow members to become competent at the tasks at hand, without wasting time on things like fund-raising. If they want to stay involved in politics over the long term, they can become government staff, serve repeatedly on Interest Panels, or work for a civil society organization or the media. The benefits from having politicians in office for multiple terms are less than the benefits from giving many more people a term in office, along with reducing the risk of corruption.

“Most people won’t have time for this kind of commitment.”

This is true. That’s why the design includes different terms for different kinds of bodies. Many people could spare a few days to a week to serve on a policy jury, just as they do with trial juries today. Others could make a part-time commitment to serve on an Interest Panel drafting a bill, just until the job is finished. Most people have enough time at some point over their lifetime to get involved if they wish. Fewer people would be willing to serve on the bodies with longer terms, but there should be enough of them.

“We need to make government simpler! Creating all these different bodies makes things much too complicated.”

The people whose jobs really are too complicated are today’s elected legislators. In contrast, a member of any of the proposed bodies would only deal with a fraction of the responsibilities of an elected legislator – and besides, they wouldn’t have to spend endless hours campaigning and fundraising.

“This would cost a lot of money – and government is too expensive already!”

Elections are very expensive, both in terms of public money and private money, not to mention the costs of legislation designed to cater to the interests of major contributors. Elected all-purpose legislatures, with their salaries and perks and staffs, are also very expensive. This model proposes many more “legislators” than the current model, and if it was added to an existing legislature, it would indeed cost more. However, the total cost could be less than the cost of current systems if elected legislatures were replaced by the model we have described.


We have criticized two common, important, and frequently taken for granted features of contemporary legislatures — that legislators are chosen by election, and that the same legislative bodies perform all tasks for all legislation. We identified seven problems caused by these features, and then proposed a new model with the potential to solve these problems, based on three strategies adapted from ancient Athenian democracy: random selection of legislators, dividing power among multiple bodies, and the use of temporary bodies for final decision making. We described how the model could work, how it could adapt to change, different ways in which it could be implemented, and how it would solve the problems we identified with elected, all-purpose legislatures.

We believe that our proposed model of lawmaking would be superior to the current models. However, our intention goes beyond merely advocating for our model. More importantly, we hope to open up new possibilities for conversation among the community of systems thinkers about idealized design of the legislative function of government — and government in general.

 Author Bios

Terry Bouricius received a bachelor of arts in political science from Middlebury College in 1976. In 1981, he was elected to the Burlington (VT) City council, becoming the first Citizens Party candidate elected in the United States. After a decade on the City council, he was elected, as a Progressive, to the Vermont House of Representatives. After stepping down in 2001, he worked with FairVote, the Center for Voting and Democracy, as an election reform policy analyst, focusing on ranked choice voting. He is currently an election administrator for a non-profit organization, and is working on a book about sortition.

David Schecter is a San Francisco-based consultant to nonprofit organizations and government agencies, specializing in conceptual mapping and strategic planning. He has two over-arching passions. One is developing, implementing, and popularizing social innovations, especially in democratic governance and cooperative economics. The other is developing and using tools and methods that help people think more effectively.

We would welcome comments on this article. Readers can email Terry Bouricius at and David Schecter at

Staff helping with this article: Editor: Dr. Steven Schneider.  Reviewers: Gene Bellinger, Barry Clemson, Richard Wright.


Ackoff, R. (1974). Redesigning the future: A systems approach to societal problems.

New York: John Wiley

Ackoff, R., Magidson, J., & Addison, H (2006). Idealized design: How to dissolve tomorrow’s

crisis . . . today. Philadelphia: Wharton School Publishing.

Ashby, R. (1956): An introduction to cybernetics. London: Chapman & Hall.

Beer, S. (1979). The heart of enterprise. London: John Wiley.

Callenbach, R., & Phillips, M. 1985. A citizen legislature. Berkeley, California: Banyan Tree


Carnes, N. (2012). Does the numerical underrepresentation of the working class in Congress

matter? Legislative Studies Quarterly. 37(1), 5-34.

Crenson, M. & Ginsberg, B. (2002). Downsizing democracy: How America sidelined its

citizens and privatized its public. Baltimore: Johns Hopkins University Press.

Crosby, N., Kelly, J., & Schaefer, P. (1986). Citizens panels: A new approach to citizen

participation. Public Administration Review. 46 (2).

Fishkin, J. (2009). When the people speak: Deliberative democracy and public consultation.

Oxford: Oxford University Press.

Hansen, M. H. (1999). Athenian democracy in the age of Demosthenes. Norman, Oklahoma:

University of Oklahoma Press.

Hollings, E. (2006, Feb. 19). Stop the Money Chase. Washington Post, B07.

Leib, E. J. (2004). Deliberative democracy in America: A proposal for a popular branch of

government. University Park, Pennsylvania: Penn State Press.

Ober, J. (2008). Democracy and knowledge: Innovation and learning in classical Athens.

Princeton, New Jersey: Princeton University Press.

O’Leary, K. (2006). Saving democracy: A plan for real representation in America. Stanford,

California: Stanford University Press.

Pitkin, H. (1972). The concept of representation. Berkeley, California: University of California


Surowiecki, J. (2004). The wisdom of crowds: Why the many are smarter than the few and

how collective wisdom shapes business, economies, societies, and nations. New York,

London, Toronto, Sidney, Auckland: Doubleday.

Woodruff, P.  (2005). First democracy: The challenge of an ancient idea. Oxford, New York.

Oxford University Press.

Reference details for this article

Bouricius, Terry and Schecter, David. (2013). An Idealized Design for the Legislative Branch of government. Systems Thinking World Journal: Reflection in Action. [Online Journal]. 2(1). [Referred 2013-01-22]. Available: . ISSN-L 2242-8577  ISSN 2242-8577






Print Friendly
Leave a Comment